Release 1.4.1 (#6874)

Fix GPU AUC.

.github/FUNDING.yml

@@ -1 +1,2 @@
 open_collective: xgboost
+custom: https://xgboost.ai/sponsors

.github/workflows/main.yml

@@ -0,0 +1,313 @@
+# This is a basic workflow to help you get started with Actions
+
+name: XGBoost-CI
+
+# Controls when the action will run. Triggers the workflow on push or pull request
+# events but only for the master branch
+on: [push, pull_request]
+
+# A workflow run is made up of one or more jobs that can run sequentially or in parallel
+jobs:
+  gtest-cpu:
+    name: Test Google C++ test (CPU)
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [macos-10.15]
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+    - name: Install system packages
+      run: |
+        brew install lz4 ninja libomp
+    - name: Build gtest binary
+      run: |
+        mkdir build
+        cd build
+        cmake .. -DGOOGLE_TEST=ON -DUSE_OPENMP=ON -DUSE_DMLC_GTEST=ON -DPLUGIN_LZ4=ON -DPLUGIN_DENSE_PARSER=ON -GNinja
+        ninja -v
+    - name: Run gtest binary
+      run: |
+        cd build
+        ctest --exclude-regex AllTestsInDMLCUnitTests --extra-verbose
+
+  gtest-cpu-nonomp:
+    name: Test Google C++ unittest (CPU Non-OMP)
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest]
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+    - name: Install system packages
+      run: |
+        sudo apt-get install -y --no-install-recommends ninja-build
+    - name: Build and install XGBoost
+      shell: bash -l {0}
+      run: |
+        mkdir build
+        cd build
+        cmake .. -GNinja -DGOOGLE_TEST=ON  -DUSE_DMLC_GTEST=ON -DUSE_OPENMP=OFF
+        ninja -v
+    - name: Run gtest binary
+      run: |
+        cd build
+        ctest --extra-verbose
+
+  python-sdist-test:
+    name: Test installing XGBoost Python source package
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ubuntu-latest, macos-10.15, windows-latest]
+        python-version: ["3.8"]
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+    - name: Install osx system dependencies
+      if: matrix.os == 'macos-10.15'
+      run: |
+        brew install ninja libomp
+    - name: Install Ubuntu system dependencies
+      if: matrix.os == 'ubuntu-latest'
+      run: |
+        sudo apt-get install -y --no-install-recommends ninja-build
+    - uses: conda-incubator/setup-miniconda@v2
+      with:
+        auto-update-conda: true
+        python-version: ${{ matrix.python-version }}
+        activate-environment: test
+    - name: Display Conda env
+      shell: bash -l {0}
+      run: |
+        conda info
+        conda list
+    - name: Build and install XGBoost
+      shell: bash -l {0}
+      run: |
+        cd python-package
+        python --version
+        python setup.py sdist
+        pip install -v ./dist/xgboost-*.tar.gz
+        cd ..
+        python -c 'import xgboost'
+
+  c-api-demo:
+    name: Test installing XGBoost lib + building the C API demo
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: ["ubuntu-latest"]
+        python-version: ["3.8"]
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+    - name: Install system packages
+      run: |
+        sudo apt-get install -y --no-install-recommends ninja-build
+    - uses: conda-incubator/setup-miniconda@v2
+      with:
+        auto-update-conda: true
+        python-version: ${{ matrix.python-version }}
+        activate-environment: test
+    - name: Display Conda env
+      shell: bash -l {0}
+      run: |
+        conda info
+        conda list
+    - name: Build and install XGBoost
+      shell: bash -l {0}
+      run: |
+        mkdir build
+        cd build
+        cmake .. -DBUILD_STATIC_LIB=ON -DCMAKE_INSTALL_PREFIX=$CONDA_PREFIX -GNinja
+        ninja -v install
+    - name: Build and run C API demo
+      shell: bash -l {0}
+      run: |
+        cd demo/c-api/
+        mkdir build
+        cd build
+        cmake .. -GNinja -DCMAKE_PREFIX_PATH=$CONDA_PREFIX
+        ninja -v
+        cd ..
+        ./build/api-demo
+
+  test-with-jvm:
+    name: Test JVM on OS ${{ matrix.os }}
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [windows-latest, ubuntu-latest]
+
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+
+    - uses: actions/setup-python@v2
+      with:
+        python-version: '3.8'
+        architecture: 'x64'
+
+    - uses: actions/setup-java@v1
+      with:
+        java-version: 1.8
+
+    - name: Install Python packages
+      run: |
+        python -m pip install wheel setuptools
+        python -m pip install awscli
+
+    - name: Cache Maven packages
+      uses: actions/cache@v2
+      with:
+        path: ~/.m2
+        key: ${{ runner.os }}-m2-${{ hashFiles('./jvm-packages/pom.xml') }}
+        restore-keys: ${{ runner.os }}-m2
+
+    - name: Test XGBoost4J
+      run: |
+        cd jvm-packages
+        mvn test -B -pl :xgboost4j_2.12
+
+    - name: Extract branch name
+      shell: bash
+      run: echo "##[set-output name=branch;]$(echo ${GITHUB_REF#refs/heads/})"
+      id: extract_branch
+      if: |
+        (github.ref == 'refs/heads/master' || contains(github.ref, 'refs/heads/release_')) &&
+        matrix.os == 'windows-latest'
+
+    - name: Publish artifact xgboost4j.dll to S3
+      run: |
+        cd lib/
+        Rename-Item -Path xgboost4j.dll -NewName xgboost4j_${{ github.sha }}.dll
+        dir
+        python -m awscli s3 cp xgboost4j_${{ github.sha }}.dll s3://xgboost-nightly-builds/${{ steps.extract_branch.outputs.branch }}/ --acl public-read
+      if: |
+        (github.ref == 'refs/heads/master' || contains(github.ref, 'refs/heads/release_')) &&
+        matrix.os == 'windows-latest'
+      env:
+        AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID_IAM_S3_UPLOADER }}
+        AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY_IAM_S3_UPLOADER }}
+
+
+    - name: Test XGBoost4J-Spark
+      run: |
+        rm -rfv build/
+        cd jvm-packages
+        mvn -B test
+      if: matrix.os == 'ubuntu-latest'  # Distributed training doesn't work on Windows
+      env:
+        RABIT_MOCK: ON
+
+  lint:
+    runs-on: ubuntu-latest
+    name: Code linting for Python and C++
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+    - uses: actions/setup-python@v2
+      with:
+        python-version: '3.7'
+        architecture: 'x64'
+    - name: Install Python packages
+      run: |
+        python -m pip install wheel setuptools
+        python -m pip install pylint cpplint numpy scipy scikit-learn
+    - name: Run lint
+      run: |
+        make lint
+
+  mypy:
+    runs-on: ubuntu-latest
+    name: Type checking for Python
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+    - uses: actions/setup-python@v2
+      with:
+        python-version: '3.7'
+        architecture: 'x64'
+    - name: Install Python packages
+      run: |
+        python -m pip install wheel setuptools mypy dask[complete] distributed
+    - name: Run mypy
+      run: |
+        make mypy
+
+  doxygen:
+    runs-on: ubuntu-latest
+    name: Generate C/C++ API doc using Doxygen
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+    - uses: actions/setup-python@v2
+      with:
+        python-version: '3.7'
+        architecture: 'x64'
+    - name: Install system packages
+      run: |
+        sudo apt-get install -y --no-install-recommends doxygen graphviz ninja-build
+        python -m pip install wheel setuptools
+        python -m pip install awscli
+    - name: Run Doxygen
+      run: |
+        mkdir build
+        cd build
+        cmake .. -DBUILD_C_DOC=ON -GNinja
+        ninja -v doc_doxygen
+    - name: Extract branch name
+      shell: bash
+      run: echo "##[set-output name=branch;]$(echo ${GITHUB_REF#refs/heads/})"
+      id: extract_branch
+      if: github.ref == 'refs/heads/master' || contains(github.ref, 'refs/heads/release_')
+    - name: Publish
+      run: |
+        cd build/
+        tar cvjf ${{ steps.extract_branch.outputs.branch }}.tar.bz2 doc_doxygen/
+        python -m awscli s3 cp ./${{ steps.extract_branch.outputs.branch }}.tar.bz2 s3://xgboost-docs/doxygen/ --acl public-read
+      if: github.ref == 'refs/heads/master' || contains(github.ref, 'refs/heads/release_')
+      env:
+        AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID_IAM_S3_UPLOADER }}
+        AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY_IAM_S3_UPLOADER }}
+
+  sphinx:
+    runs-on: ubuntu-latest
+    name: Build docs using Sphinx
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+    - uses: actions/setup-python@v2
+      with:
+        python-version: '3.8'
+        architecture: 'x64'
+    - name: Install system packages
+      run: |
+        sudo apt-get install -y --no-install-recommends graphviz
+        python -m pip install wheel setuptools
+        python -m pip install -r doc/requirements.txt
+    - name: Extract branch name
+      shell: bash
+      run: echo "##[set-output name=branch;]$(echo ${GITHUB_REF#refs/heads/})"
+      id: extract_branch
+      if: github.ref == 'refs/heads/master' || contains(github.ref, 'refs/heads/release_')
+    - name: Run Sphinx
+      run: |
+        make -C doc html
+      env:
+        SPHINX_GIT_BRANCH: ${{ steps.extract_branch.outputs.branch }}

.github/workflows/r_nold.yml

@@ -0,0 +1,44 @@
+# Run R tests with noLD R. Only triggered by a pull request review
+# See discussion at https://github.com/dmlc/xgboost/pull/6378
+
+name: XGBoost-R-noLD
+
+on:
+  pull_request_review_comment:
+    types: [created]
+
+env:
+  R_PACKAGES: c('XML', 'igraph', 'data.table', 'magrittr', 'ggplot2', 'DiagrammeR', 'Ckmeans.1d.dp', 'vcd', 'testthat', 'lintr', 'knitr', 'rmarkdown', 'e1071', 'cplm', 'devtools', 'float', 'titanic')
+
+jobs:
+  test-R-noLD:
+    if: github.event.comment.body == '/gha run r-nold-test' && contains('OWNER,MEMBER,COLLABORATOR', github.event.comment.author_association)
+    timeout-minutes: 120
+    runs-on: ubuntu-latest
+    container: rhub/debian-gcc-devel-nold
+    steps:
+    - name: Install git and system packages
+      shell: bash
+      run: |
+        apt-get update && apt-get install -y git libcurl4-openssl-dev libssl-dev libssh2-1-dev libgit2-dev libxml2-dev
+
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+
+    - name: Install dependencies
+      shell: bash
+      run: |
+        cat > install_libs.R <<EOT
+          install.packages(${{ env.R_PACKAGES }},
+                           repos = 'http://cloud.r-project.org',
+                           dependencies = c('Depends', 'Imports', 'LinkingTo'))
+        EOT
+        /tmp/R-devel/bin/Rscript install_libs.R
+
+    - name: Run R tests
+      shell: bash
+      run: |
+        cd R-package && \
+        /tmp/R-devel/bin/R CMD INSTALL . && \
+        /tmp/R-devel/bin/R -q -e "library(testthat); setwd('tests'); source('testthat.R')"

.github/workflows/r_tests.yml

@@ -0,0 +1,116 @@
+name: XGBoost-R-Tests
+
+on: [push, pull_request]
+
+env:
+  R_PACKAGES: c('XML', 'igraph', 'data.table', 'magrittr', 'ggplot2', 'DiagrammeR', 'Ckmeans.1d.dp', 'vcd', 'testthat', 'lintr', 'knitr', 'rmarkdown', 'e1071', 'cplm', 'devtools', 'float', 'titanic')
+
+jobs:
+  lintr:
+    runs-on: ${{ matrix.config.os }}
+    name: Run R linters on OS ${{ matrix.config.os }}, R ${{ matrix.config.r }}, Compiler ${{ matrix.config.compiler }}, Build ${{ matrix.config.build }}
+    strategy:
+      matrix:
+        config:
+          - {os: windows-latest, r: 'release', compiler: 'mingw', build: 'autotools'}
+    env:
+      R_REMOTES_NO_ERRORS_FROM_WARNINGS: true
+      RSPM: ${{ matrix.config.rspm }}
+
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+
+    - uses: r-lib/actions/setup-r@master
+      with:
+        r-version: ${{ matrix.config.r }}
+
+    - name: Install dependencies
+      shell: Rscript {0}
+      run: |
+        install.packages(${{ env.R_PACKAGES }},
+                         repos = 'http://cloud.r-project.org',
+                         dependencies = c('Depends', 'Imports', 'LinkingTo'))
+
+    - name: Run lintr
+      run: |
+        cd R-package
+        R.exe CMD INSTALL .
+        Rscript.exe tests/helper_scripts/run_lint.R
+
+  test-with-R:
+    runs-on: ${{ matrix.config.os }}
+    name: Test R on OS ${{ matrix.config.os }}, R ${{ matrix.config.r }}, Compiler ${{ matrix.config.compiler }}, Build ${{ matrix.config.build }}
+    strategy:
+      fail-fast: false
+      matrix:
+        config:
+          - {os: windows-2016, r: 'release', compiler: 'mingw', build: 'autotools'}
+          - {os: windows-2016, r: 'release', compiler: 'msvc', build: 'cmake'}
+          - {os: windows-2016, r: 'release', compiler: 'mingw', build: 'cmake'}
+    env:
+      R_REMOTES_NO_ERRORS_FROM_WARNINGS: true
+      RSPM: ${{ matrix.config.rspm }}
+
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+
+    - uses: r-lib/actions/setup-r@master
+      with:
+        r-version: ${{ matrix.config.r }}
+
+    - name: Install dependencies
+      shell: Rscript {0}
+      run: |
+        install.packages(${{ env.R_PACKAGES }},
+                         repos = 'http://cloud.r-project.org',
+                         dependencies = c('Depends', 'Imports', 'LinkingTo'))
+
+    - uses: actions/setup-python@v2
+      with:
+        python-version: '3.7'
+        architecture: 'x64'
+
+    - name: Test R
+      run: |
+        python tests/ci_build/test_r_package.py --compiler='${{ matrix.config.compiler }}' --build-tool='${{ matrix.config.build }}'
+
+  test-R-CRAN:
+    runs-on: ubuntu-latest
+
+    strategy:
+      fail-fast: false
+      matrix:
+        config:
+          - {r: 'release'}
+
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: 'true'
+
+    - uses: r-lib/actions/setup-r@master
+      with:
+        r-version: ${{ matrix.config.r }}
+
+    - uses: r-lib/actions/setup-tinytex@master
+
+    - name: Install system packages
+      run: |
+        sudo apt-get update && sudo apt-get install libcurl4-openssl-dev libssl-dev libssh2-1-dev libgit2-dev pandoc pandoc-citeproc
+
+    - name: Install dependencies
+      shell: Rscript {0}
+      run: |
+        install.packages(${{ env.R_PACKAGES }},
+                         repos = 'http://cloud.r-project.org',
+                         dependencies = c('Depends', 'Imports', 'LinkingTo'))
+
+    - name: Check R Package
+      run: |
+        # Print stacktrace upon success of failure
+        make Rcheck || tests/ci_build/print_r_stacktrace.sh fail
+        tests/ci_build/print_r_stacktrace.sh success

.gitignore

@@ -51,6 +51,7 @@ Debug
 #.Rbuildignore
 R-package.Rproj
 *.cache*
+.mypy_cache/
 # java
 java/xgboost4j/target
 java/xgboost4j/tmp
@@ -70,6 +71,7 @@ build
 build_plugin
 recommonmark/
 tags
+TAGS
 *.class
 target
 *.swp
@@ -92,6 +94,7 @@ metastore_db
 # files from R-package source install
 **/config.status
 R-package/src/Makevars
+*.lib
 
 # Visual Studio Code
 /.vscode/
@@ -102,4 +105,23 @@ R-package/src/Makevars
 /cmake-build-debug/
 
 # GDB
-.gdb_history
+.gdb_history
+
+# Python joblib.Memory used in pytest.
+cachedir/
+
+# Files from local Dask work
+dask-worker-space/
+
+# Jupyter notebook checkpoints
+.ipynb_checkpoints/
+
+# credentials and key material
+config
+credentials
+credentials.csv
+*.env
+*.pem
+*.pub
+*.rdp
+*_rsa

.gitmodules

@@ -1,9 +1,10 @@
 [submodule "dmlc-core"]
 	path = dmlc-core
 	url = https://github.com/dmlc/dmlc-core
-[submodule "rabit"]
-	path = rabit
-	url = https://github.com/dmlc/rabit
+	branch = main
 [submodule "cub"]
 	path = cub
 	url = https://github.com/NVlabs/cub
+[submodule "gputreeshap"]
+	path = gputreeshap
+	url = https://github.com/rapidsai/gputreeshap.git

.travis.yml

@@ -1,34 +1,22 @@
-# disable sudo for container build.
 sudo: required
 
-# Enabling test OS X
-os:
-  - linux
-  - osx
-
-osx_image: xcode10.3
 dist: bionic
 
-# Use Build Matrix to do lint and build seperately
 env:
-  matrix:
-    # python package test
-    - TASK=python_test
-    # test installation of Python source distribution
-    - TASK=python_sdist_test
-    # java package test
-    - TASK=java_test
-    # cmake test
-    - TASK=cmake_test
+  global:
+    - secure: "PR16i9F8QtNwn99C5NDp8nptAS+97xwDtXEJJfEiEVhxPaaRkOp0MPWhogCaK0Eclxk1TqkgWbdXFknwGycX620AzZWa/A1K3gAs+GrpzqhnPMuoBJ0Z9qxXTbSJvCyvMbYwVrjaxc/zWqdMU8waWz8A7iqKGKs/SqbQ3rO6v7c="
+    - secure: "dAGAjBokqm/0nVoLMofQni/fWIBcYSmdq4XvCBX1ZAMDsWnuOfz/4XCY6h2lEI1rVHZQ+UdZkc9PioOHGPZh5BnvE49/xVVWr9c4/61lrDOlkD01ZjSAeoV0fAZq+93V/wPl4QV+MM+Sem9hNNzFSbN5VsQLAiWCSapWsLdKzqA="
 
-matrix:
-  exclude:
-    - os: linux
+jobs:
+  include:
+    - os: osx
+      arch: amd64
+      osx_image: xcode10.2
       env: TASK=python_test
-    - os: linux
+    - os: osx
+      arch: amd64
+      osx_image: xcode10.2
       env: TASK=java_test
-    - os: linux
-      env: TASK=cmake_test
 
 # dependent brew packages
 addons:
@@ -39,9 +27,14 @@ addons:
       - graphviz
       - openssl
       - libgit2
+      - lz4
       - wget
       - r
     update: true
+  apt:
+    packages:
+      - snapd
+      - unzip
 
 before_install:
   - source tests/travis/travis_setup_env.sh

CMakeLists.txt

@@ -1,9 +1,10 @@
 cmake_minimum_required(VERSION 3.13)
-project(xgboost LANGUAGES CXX C VERSION 1.1.0)
+project(xgboost LANGUAGES CXX C VERSION 1.4.1)
 include(cmake/Utils.cmake)
 list(APPEND CMAKE_MODULE_PATH "${xgboost_SOURCE_DIR}/cmake/modules")
 cmake_policy(SET CMP0022 NEW)
 cmake_policy(SET CMP0079 NEW)
+set(CMAKE_POLICY_DEFAULT_CMP0063 NEW)
 cmake_policy(SET CMP0063 NEW)
 
 if ((${CMAKE_VERSION} VERSION_GREATER 3.13) OR (${CMAKE_VERSION} VERSION_EQUAL 3.13))
@@ -23,17 +24,23 @@ write_version()
 set_default_configuration_release()
 
 #-- Options
+## User options
 option(BUILD_C_DOC "Build documentation for C APIs using Doxygen." OFF)
 option(USE_OPENMP "Build with OpenMP support." ON)
 option(BUILD_STATIC_LIB "Build static library" OFF)
+option(RABIT_BUILD_MPI "Build MPI" OFF)
 ## Bindings
 option(JVM_BINDINGS "Build JVM bindings" OFF)
 option(R_LIB "Build shared library for R package" OFF)
 ## Dev
 option(USE_DEBUG_OUTPUT "Dump internal training results like gradients and predictions to stdout.
 Should only be used for debugging." OFF)
+option(FORCE_COLORED_OUTPUT "Force colored output from compilers, useful when ninja is used instead of make." OFF)
+option(ENABLE_ALL_WARNINGS "Enable all compiler warnings. Only effective for GCC/Clang" OFF)
+option(LOG_CAPI_INVOCATION "Log all C API invocations for debugging" OFF)
 option(GOOGLE_TEST "Build google tests" OFF)
 option(USE_DMLC_GTEST "Use google tests bundled with dmlc-core submodule" OFF)
+option(USE_DEVICE_DEBUG "Generate CUDA device debug info." OFF)
 option(USE_NVTX "Build with cuda profiling annotations. Developers only." OFF)
 set(NVTX_HEADER_DIR "" CACHE PATH "Path to the stand-alone nvtx header")
 option(RABIT_MOCK "Build rabit with mock" OFF)
@@ -57,6 +64,10 @@ address, leak, undefined and thread.")
 ## Plugins
 option(PLUGIN_LZ4 "Build lz4 plugin" OFF)
 option(PLUGIN_DENSE_PARSER "Build dense parser plugin" OFF)
+option(PLUGIN_RMM "Build with RAPIDS Memory Manager (RMM)" OFF)
+## TODO: 1. Add check if DPC++ compiler is used for building
+option(PLUGIN_UPDATER_ONEAPI "DPC++ updater" OFF)
+option(ADD_PKGCONFIG "Add xgboost.pc into system." ON)
 
 #-- Checks for building XGBoost
 if (USE_DEBUG_OUTPUT AND (NOT (CMAKE_BUILD_TYPE MATCHES Debug)))
@@ -65,6 +76,9 @@ endif (USE_DEBUG_OUTPUT AND (NOT (CMAKE_BUILD_TYPE MATCHES Debug)))
 if (USE_NCCL AND NOT (USE_CUDA))
   message(SEND_ERROR "`USE_NCCL` must be enabled with `USE_CUDA` flag.")
 endif (USE_NCCL AND NOT (USE_CUDA))
+if (USE_DEVICE_DEBUG AND NOT (USE_CUDA))
+  message(SEND_ERROR "`USE_DEVICE_DEBUG` must be enabled with `USE_CUDA` flag.")
+endif (USE_DEVICE_DEBUG AND NOT (USE_CUDA))
 if (BUILD_WITH_SHARED_NCCL AND (NOT USE_NCCL))
   message(SEND_ERROR "Build XGBoost with -DUSE_NCCL=ON to enable BUILD_WITH_SHARED_NCCL.")
 endif (BUILD_WITH_SHARED_NCCL AND (NOT USE_NCCL))
@@ -78,6 +92,23 @@ endif (R_LIB AND GOOGLE_TEST)
 if (USE_AVX)
   message(SEND_ERROR  "The option 'USE_AVX' is deprecated as experimental AVX features have been removed from XGBoost.")
 endif (USE_AVX)
+if (PLUGIN_RMM AND NOT (USE_CUDA))
+  message(SEND_ERROR "`PLUGIN_RMM` must be enabled with `USE_CUDA` flag.")
+endif (PLUGIN_RMM AND NOT (USE_CUDA))
+if (PLUGIN_RMM AND NOT ((CMAKE_CXX_COMPILER_ID STREQUAL "Clang") OR (CMAKE_CXX_COMPILER_ID STREQUAL "GNU")))
+  message(SEND_ERROR "`PLUGIN_RMM` must be used with GCC or Clang compiler.")
+endif (PLUGIN_RMM AND NOT ((CMAKE_CXX_COMPILER_ID STREQUAL "Clang") OR (CMAKE_CXX_COMPILER_ID STREQUAL "GNU")))
+if (PLUGIN_RMM AND NOT (CMAKE_SYSTEM_NAME STREQUAL "Linux"))
+  message(SEND_ERROR "`PLUGIN_RMM` must be used with Linux.")
+endif (PLUGIN_RMM AND NOT (CMAKE_SYSTEM_NAME STREQUAL "Linux"))
+if (ENABLE_ALL_WARNINGS)
+  if ((NOT CMAKE_CXX_COMPILER_ID MATCHES "Clang") AND (NOT CMAKE_CXX_COMPILER_ID STREQUAL "GNU"))
+    message(SEND_ERROR "ENABLE_ALL_WARNINGS is only available for Clang and GCC.")
+  endif ((NOT CMAKE_CXX_COMPILER_ID MATCHES "Clang") AND (NOT CMAKE_CXX_COMPILER_ID STREQUAL "GNU"))
+endif (ENABLE_ALL_WARNINGS)
+if (BUILD_STATIC_LIB AND (R_LIB OR JVM_BINDINGS))
+  message(SEND_ERROR "Cannot build a static library libxgboost.a when R or JVM packages are enabled.")
+endif (BUILD_STATIC_LIB AND (R_LIB OR JVM_BINDINGS))
 
 #-- Sanitizer
 if (USE_SANITIZER)
@@ -92,11 +123,20 @@ if (USE_CUDA)
   message(STATUS "Configured CUDA host compiler: ${CMAKE_CUDA_HOST_COMPILER}")
 
   enable_language(CUDA)
+  if (${CMAKE_CUDA_COMPILER_VERSION} VERSION_LESS 10.0)
+    message(FATAL_ERROR "CUDA version must be at least 10.0!")
+  endif()
   set(GEN_CODE "")
   format_gencode_flags("${GPU_COMPUTE_VER}" GEN_CODE)
-  message(STATUS "CUDA GEN_CODE: ${GEN_CODE}")
+  add_subdirectory(${PROJECT_SOURCE_DIR}/gputreeshap)
 endif (USE_CUDA)
 
+if (FORCE_COLORED_OUTPUT AND (CMAKE_GENERATOR STREQUAL "Ninja") AND
+    ((CMAKE_CXX_COMPILER_ID STREQUAL "GNU") OR
+      (CMAKE_CXX_COMPILER_ID STREQUAL "Clang")))
+  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fdiagnostics-color=always")
+endif()
+
 find_package(Threads REQUIRED)
 
 if (USE_OPENMP)
@@ -112,63 +152,59 @@ endif (USE_OPENMP)
 msvc_use_static_runtime()
 add_subdirectory(${xgboost_SOURCE_DIR}/dmlc-core)
 set_target_properties(dmlc PROPERTIES
-  CXX_STANDARD 11
+  CXX_STANDARD 14
   CXX_STANDARD_REQUIRED ON
   POSITION_INDEPENDENT_CODE ON)
-list(APPEND LINKED_LIBRARIES_PRIVATE dmlc)
+if (MSVC)
+  target_compile_options(dmlc PRIVATE
+                         -D_CRT_SECURE_NO_WARNINGS -D_CRT_SECURE_NO_DEPRECATE)
+  if (TARGET dmlc_unit_tests)
+    target_compile_options(dmlc_unit_tests PRIVATE
+                           -D_CRT_SECURE_NO_WARNINGS -D_CRT_SECURE_NO_DEPRECATE)
+  endif (TARGET dmlc_unit_tests)
+endif (MSVC)
+if (ENABLE_ALL_WARNINGS)
+  target_compile_options(dmlc PRIVATE -Wall -Wextra)
+endif (ENABLE_ALL_WARNINGS)
 
 # rabit
-set(RABIT_BUILD_DMLC OFF)
-set(DMLC_ROOT ${xgboost_SOURCE_DIR}/dmlc-core)
-set(RABIT_WITH_R_LIB ${R_LIB})
 add_subdirectory(rabit)
 
-if (RABIT_MOCK)
-  list(APPEND LINKED_LIBRARIES_PRIVATE rabit_mock_static)
-else()
-  list(APPEND LINKED_LIBRARIES_PRIVATE rabit)
-endif(RABIT_MOCK)
-foreach(lib rabit rabit_base rabit_empty rabit_mock rabit_mock_static)
-  # Explicitly link dmlc to rabit, so that configured header (build_config.h)
-  # from dmlc is correctly applied to rabit.
-  if (TARGET ${lib})
-    target_link_libraries(${lib} dmlc ${CMAKE_THREAD_LIBS_INIT})
-    if (HIDE_CXX_SYMBOLS)  # Hide all C++ symbols from Rabit
-      set_target_properties(${lib} PROPERTIES CXX_VISIBILITY_PRESET hidden)
-    endif (HIDE_CXX_SYMBOLS)
-  endif (TARGET ${lib})
-endforeach()
+# core xgboost
+add_subdirectory(${xgboost_SOURCE_DIR}/src)
+target_link_libraries(objxgboost PUBLIC dmlc)
 
 # Exports some R specific definitions and objects
 if (R_LIB)
   add_subdirectory(${xgboost_SOURCE_DIR}/R-package)
 endif (R_LIB)
 
-# core xgboost
-list(APPEND LINKED_LIBRARIES_PRIVATE Threads::Threads ${CMAKE_THREAD_LIBS_INIT})
+# Plugin
 add_subdirectory(${xgboost_SOURCE_DIR}/plugin)
-add_subdirectory(${xgboost_SOURCE_DIR}/src)
-target_link_libraries(objxgboost PUBLIC dmlc)
-set(XGBOOST_OBJ_SOURCES "${XGBOOST_OBJ_SOURCES};$<TARGET_OBJECTS:objxgboost>")
 
 #-- library
 if (BUILD_STATIC_LIB)
-  add_library(xgboost STATIC ${XGBOOST_OBJ_SOURCES})
+  add_library(xgboost STATIC)
 else (BUILD_STATIC_LIB)
-  add_library(xgboost SHARED ${XGBOOST_OBJ_SOURCES})
+  add_library(xgboost SHARED)
 endif (BUILD_STATIC_LIB)
+target_link_libraries(xgboost PRIVATE objxgboost)
+
+if (USE_CUDA)
+  xgboost_set_cuda_flags(xgboost)
+endif (USE_CUDA)
 
 #-- Hide all C++ symbols
 if (HIDE_CXX_SYMBOLS)
-  set_target_properties(objxgboost PROPERTIES CXX_VISIBILITY_PRESET hidden)
-  set_target_properties(xgboost PROPERTIES CXX_VISIBILITY_PRESET hidden)
+  foreach(target objxgboost xgboost dmlc)
+    set_target_properties(${target} PROPERTIES CXX_VISIBILITY_PRESET hidden)
+  endforeach()
 endif (HIDE_CXX_SYMBOLS)
 
 target_include_directories(xgboost
   INTERFACE
-  $<INSTALL_INTERFACE:${CMAKE_INSTALL_PREFIX}/include>
+  $<INSTALL_INTERFACE:$<INSTALL_PREFIX>/include>
   $<BUILD_INTERFACE:${CMAKE_CURRENT_LIST_DIR}/include>)
-target_link_libraries(xgboost PRIVATE ${LINKED_LIBRARIES_PRIVATE})
 
 # This creates its own shared library `xgboost4j'.
 if (JVM_BINDINGS)
@@ -177,18 +213,21 @@ endif (JVM_BINDINGS)
 #-- End shared library
 
 #-- CLI for xgboost
-add_executable(runxgboost ${xgboost_SOURCE_DIR}/src/cli_main.cc ${XGBOOST_OBJ_SOURCES})
+add_executable(runxgboost ${xgboost_SOURCE_DIR}/src/cli_main.cc)
+target_link_libraries(runxgboost PRIVATE objxgboost)
+if (USE_NVTX)
+  enable_nvtx(runxgboost)
+endif (USE_NVTX)
 
 target_include_directories(runxgboost
   PRIVATE
   ${xgboost_SOURCE_DIR}/include
   ${xgboost_SOURCE_DIR}/dmlc-core/include
   ${xgboost_SOURCE_DIR}/rabit/include)
-target_link_libraries(runxgboost PRIVATE ${LINKED_LIBRARIES_PRIVATE})
 set_target_properties(
   runxgboost PROPERTIES
   OUTPUT_NAME xgboost
-  CXX_STANDARD 11
+  CXX_STANDARD 14
   CXX_STANDARD_REQUIRED ON)
 #-- End CLI for xgboost
 
@@ -199,11 +238,12 @@ add_dependencies(xgboost runxgboost)
 
 #-- Installing XGBoost
 if (R_LIB)
+  include(cmake/RPackageInstallTargetSetup.cmake)
   set_target_properties(xgboost PROPERTIES PREFIX "")
   if (APPLE)
     set_target_properties(xgboost PROPERTIES SUFFIX ".so")
   endif (APPLE)
-  setup_rpackage_install_target(xgboost ${CMAKE_CURRENT_BINARY_DIR})
+  setup_rpackage_install_target(xgboost "${CMAKE_CURRENT_BINARY_DIR}/R-package-install")
   set(CMAKE_INSTALL_PREFIX "${CMAKE_CURRENT_BINARY_DIR}/dummy_inst")
 endif (R_LIB)
 if (MINGW)
@@ -220,7 +260,20 @@ include(GNUInstallDirs)
 install(DIRECTORY ${xgboost_SOURCE_DIR}/include/xgboost
   DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
 
-install(TARGETS xgboost runxgboost
+# Install libraries. If `xgboost` is a static lib, specify `objxgboost` also, to avoid the
+# following error:
+#
+#  > install(EXPORT ...) includes target "xgboost" which requires target "objxgboost" that is not
+#  > in any export set.
+#
+# https://github.com/dmlc/xgboost/issues/6085
+if (BUILD_STATIC_LIB)
+  set(INSTALL_TARGETS xgboost runxgboost objxgboost dmlc)
+else (BUILD_STATIC_LIB)
+  set(INSTALL_TARGETS xgboost runxgboost)
+endif (BUILD_STATIC_LIB)
+
+install(TARGETS ${INSTALL_TARGETS}
   EXPORT XGBoostTargets
   ARCHIVE DESTINATION ${CMAKE_INSTALL_LIBDIR}
   LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
@@ -274,3 +327,12 @@ endif (GOOGLE_TEST)
 # replace /MD with /MT. See https://github.com/dmlc/xgboost/issues/4462
 # for issues caused by mixing of /MD and /MT flags
 msvc_use_static_runtime()
+
+# Add xgboost.pc
+if (ADD_PKGCONFIG)
+  configure_file(${xgboost_SOURCE_DIR}/cmake/xgboost.pc.in ${xgboost_BINARY_DIR}/xgboost.pc @ONLY)
+
+  install(
+    FILES ${xgboost_BINARY_DIR}/xgboost.pc
+    DESTINATION ${CMAKE_INSTALL_LIBDIR}/pkgconfig)
+endif (ADD_PKGCONFIG)

CONTRIBUTORS.md

@@ -10,8 +10,8 @@ The Project Management Committee(PMC) consists group of active committers that m
   - Tianqi is a Ph.D. student working on large-scale machine learning. He is the creator of the project.
 * [Michael Benesty](https://github.com/pommedeterresautee)
   - Michael is a lawyer and data scientist in France. He is the creator of XGBoost interactive analysis module in R.
-* [Yuan Tang](https://github.com/terrytangyuan), Ant Financial
-  - Yuan is a software engineer in Ant Financial. He contributed mostly in R and Python packages.
+* [Yuan Tang](https://github.com/terrytangyuan), Ant Group
+  - Yuan is a software engineer in Ant Group. He contributed mostly in R and Python packages.
 * [Nan Zhu](https://github.com/CodingCat), Uber
   - Nan is a software engineer in Uber. He contributed mostly in JVM packages.
 * [Jiaming Yuan](https://github.com/trivialfis)
@@ -37,6 +37,8 @@ Committers are people who have made substantial contribution to the project and
   - Sergei is a software engineer in Criteo. He contributed mostly in JVM packages.
 * [Scott Lundberg](http://scottlundberg.com/), University of Washington
   - Scott is a Ph.D. student at University of Washington. He is the creator of SHAP, a unified approach to explain the output of machine learning models such as decision tree ensembles. He also helps maintain the XGBoost Julia package.
+* [Egor Smirnov](https://github.com/SmirnovEgorRu), Intel
+  - Egor has led a major effort to improve the performance of XGBoost on multi-core CPUs.
 
 
 Become a Committer

Jenkinsfile

@@ -6,6 +6,9 @@
 // Command to run command inside a docker container
 dockerRun = 'tests/ci_build/ci_build.sh'
 
+// Which CUDA version to use when building reference distribution wheel
+ref_cuda_ver = '10.0'
+
 import groovy.transform.Field
 
 @Field
@@ -31,29 +34,19 @@ pipeline {
 
   // Build stages
   stages {
-    stage('Jenkins Linux: Get sources') {
-      agent { label 'linux && cpu' }
+    stage('Jenkins Linux: Initialize') {
+      agent { label 'job_initializer' }
       steps {
         script {
+          def buildNumber = env.BUILD_NUMBER as int
+          if (buildNumber > 1) milestone(buildNumber - 1)
+          milestone(buildNumber)
+
           checkoutSrcs()
           commit_id = "${GIT_COMMIT}"
         }
+        sh 'python3 tests/jenkins_get_approval.py'
         stash name: 'srcs'
-        milestone ordinal: 1
-      }
-    }
-    stage('Jenkins Linux: Formatting Check') {
-      agent none
-      steps {
-        script {
-          parallel ([
-            'clang-tidy': { ClangTidy() },
-            'lint': { Lint() },
-            'sphinx-doc': { SphinxDoc() },
-            'doxygen': { Doxygen() }
-          ])
-        }
-        milestone ordinal: 2
       }
     }
     stage('Jenkins Linux: Build') {
@@ -61,17 +54,23 @@ pipeline {
       steps {
         script {
           parallel ([
+            'clang-tidy': { ClangTidy() },
             'build-cpu': { BuildCPU() },
+            'build-cpu-arm64': { BuildCPUARM64() },
             'build-cpu-rabit-mock': { BuildCPUMock() },
-            'build-cpu-non-omp': { BuildCPUNonOmp() },
-            'build-gpu-cuda9.0': { BuildCUDA(cuda_version: '9.0') },
+            // Build reference, distribution-ready Python wheel with CUDA 10.0
+            // using CentOS 6 image
             'build-gpu-cuda10.0': { BuildCUDA(cuda_version: '10.0') },
+            // The build-gpu-* builds below use Ubuntu image
             'build-gpu-cuda10.1': { BuildCUDA(cuda_version: '10.1') },
-            'build-jvm-packages': { BuildJVMPackages(spark_version: '2.4.3') },
+            'build-gpu-cuda10.2': { BuildCUDA(cuda_version: '10.2', build_rmm: true) },
+            'build-gpu-cuda11.0': { BuildCUDA(cuda_version: '11.0') },
+            'build-gpu-rpkg': { BuildRPackageWithCUDA(cuda_version: '10.0') },
+            'build-jvm-packages-gpu-cuda10.0': { BuildJVMPackagesWithCUDA(spark_version: '3.0.0', cuda_version: '10.0') },
+            'build-jvm-packages': { BuildJVMPackages(spark_version: '3.0.0') },
             'build-jvm-doc': { BuildJVMDoc() }
           ])
         }
-        milestone ordinal: 3
       }
     }
     stage('Jenkins Linux: Test') {
@@ -80,19 +79,19 @@ pipeline {
         script {
           parallel ([
             'test-python-cpu': { TestPythonCPU() },
-            'test-python-gpu-cuda9.0': { TestPythonGPU(cuda_version: '9.0') },
-            'test-python-gpu-cuda10.0': { TestPythonGPU(cuda_version: '10.0') },
-            'test-python-gpu-cuda10.1': { TestPythonGPU(cuda_version: '10.1') },
-            'test-python-mgpu-cuda10.1': { TestPythonGPU(cuda_version: '10.1', multi_gpu: true) },
-            'test-cpp-gpu': { TestCppGPU(cuda_version: '10.1') },
-            'test-cpp-mgpu': { TestCppGPU(cuda_version: '10.1', multi_gpu: true) },
-            'test-jvm-jdk8': { CrossTestJVMwithJDK(jdk_version: '8', spark_version: '2.4.3') },
+            'test-python-cpu-arm64': { TestPythonCPUARM64() },
+            // artifact_cuda_version doesn't apply to RMM tests; RMM tests will always match CUDA version between artifact and host env
+            'test-python-gpu-cuda10.2': { TestPythonGPU(artifact_cuda_version: '10.0', host_cuda_version: '10.2', test_rmm: true) },
+            'test-python-gpu-cuda11.0-cross': { TestPythonGPU(artifact_cuda_version: '10.0', host_cuda_version: '11.0') },
+            'test-python-gpu-cuda11.0': { TestPythonGPU(artifact_cuda_version: '11.0', host_cuda_version: '11.0') },
+            'test-python-mgpu-cuda10.2': { TestPythonGPU(artifact_cuda_version: '10.0', host_cuda_version: '10.2', multi_gpu: true, test_rmm: true) },
+            'test-cpp-gpu-cuda10.2': { TestCppGPU(artifact_cuda_version: '10.2', host_cuda_version: '10.2', test_rmm: true) },
+            'test-cpp-gpu-cuda11.0': { TestCppGPU(artifact_cuda_version: '11.0', host_cuda_version: '11.0') },
+            'test-jvm-jdk8': { CrossTestJVMwithJDK(jdk_version: '8', spark_version: '3.0.0') },
             'test-jvm-jdk11': { CrossTestJVMwithJDK(jdk_version: '11') },
-            'test-jvm-jdk12': { CrossTestJVMwithJDK(jdk_version: '12') },
-            'test-r-3.5.3': { TestR(use_r35: true) }
+            'test-jvm-jdk12': { CrossTestJVMwithJDK(jdk_version: '12') }
           ])
         }
-        milestone ordinal: 4
       }
     }
     stage('Jenkins Linux: Deploy') {
@@ -100,10 +99,9 @@ pipeline {
       steps {
         script {
           parallel ([
-            'deploy-jvm-packages': { DeployJVMPackages(spark_version: '2.4.3') }
+            'deploy-jvm-packages': { DeployJVMPackages(spark_version: '3.0.0') }
           ])
         }
-        milestone ordinal: 5
       }
     }
   }
@@ -124,13 +122,17 @@ def checkoutSrcs() {
   }
 }
 
+def GetCUDABuildContainerType(cuda_version) {
+  return (cuda_version == ref_cuda_ver) ? 'gpu_build_centos6' : 'gpu_build'
+}
+
 def ClangTidy() {
-  node('linux && cpu') {
+  node('linux && cpu_build') {
     unstash name: 'srcs'
     echo "Running clang-tidy job..."
     def container_type = "clang_tidy"
     def docker_binary = "docker"
-    def dockerArgs = "--build-arg CUDA_VERSION=10.1"
+    def dockerArgs = "--build-arg CUDA_VERSION_ARG=10.1"
     sh """
     ${dockerRun} ${container_type} ${docker_binary} ${dockerArgs} python3 tests/ci_build/tidy.py
     """
@@ -138,48 +140,6 @@ def ClangTidy() {
   }
 }
 
-def Lint() {
-  node('linux && cpu') {
-    unstash name: 'srcs'
-    echo "Running lint..."
-    def container_type = "cpu"
-    def docker_binary = "docker"
-    sh """
-    ${dockerRun} ${container_type} ${docker_binary} make lint
-    """
-    deleteDir()
-  }
-}
-
-def SphinxDoc() {
-  node('linux && cpu') {
-    unstash name: 'srcs'
-    echo "Running sphinx-doc..."
-    def container_type = "cpu"
-    def docker_binary = "docker"
-    def docker_extra_params = "CI_DOCKER_EXTRA_PARAMS_INIT='-e SPHINX_GIT_BRANCH=${BRANCH_NAME}'"
-    sh """#!/bin/bash
-    ${docker_extra_params} ${dockerRun} ${container_type} ${docker_binary} make -C doc html
-    """
-    deleteDir()
-  }
-}
-
-def Doxygen() {
-  node('linux && cpu') {
-    unstash name: 'srcs'
-    echo "Running doxygen..."
-    def container_type = "cpu"
-    def docker_binary = "docker"
-    sh """
-    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/doxygen.sh ${BRANCH_NAME}
-    """
-    echo 'Uploading doc...'
-    s3Upload file: "build/${BRANCH_NAME}.tar.bz2", bucket: 'xgboost-docs', acl: 'PublicRead', path: "doxygen/${BRANCH_NAME}.tar.bz2"
-    deleteDir()
-  }
-}
-
 def BuildCPU() {
   node('linux && cpu') {
     unstash name: 'srcs'
@@ -191,15 +151,15 @@ def BuildCPU() {
       # This step is not necessary, but here we include it, to ensure that DMLC_CORE_USE_CMAKE flag is correctly propagated
       # We want to make sure that we use the configured header build/dmlc/build_config.h instead of include/dmlc/build_config_default.h.
       # See discussion at https://github.com/dmlc/xgboost/issues/5510
-    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/build_via_cmake.sh
-    ${dockerRun} ${container_type} ${docker_binary} build/testxgboost
+    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/build_via_cmake.sh -DPLUGIN_LZ4=ON -DPLUGIN_DENSE_PARSER=ON
+    ${dockerRun} ${container_type} ${docker_binary} bash -c "cd build && ctest --extra-verbose"
     """
     // Sanitizer test
     def docker_extra_params = "CI_DOCKER_EXTRA_PARAMS_INIT='-e ASAN_SYMBOLIZER_PATH=/usr/bin/llvm-symbolizer -e ASAN_OPTIONS=symbolize=1 -e UBSAN_OPTIONS=print_stacktrace=1:log_path=ubsan_error.log --cap-add SYS_PTRACE'"
     sh """
     ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/build_via_cmake.sh -DUSE_SANITIZER=ON -DENABLED_SANITIZERS="address;leak;undefined" \
       -DCMAKE_BUILD_TYPE=Debug -DSANITIZER_PATH=/usr/lib/x86_64-linux-gnu/
-    ${docker_extra_params} ${dockerRun} ${container_type} ${docker_binary} build/testxgboost
+    ${docker_extra_params} ${dockerRun} ${container_type} ${docker_binary} bash -c "cd build && ctest --exclude-regex AllTestsInDMLCUnitTests --extra-verbose"
     """
 
     stash name: 'xgboost_cli', includes: 'xgboost'
@@ -207,6 +167,35 @@ def BuildCPU() {
   }
 }
 
+def BuildCPUARM64() {
+  node('linux && arm64') {
+    unstash name: 'srcs'
+    echo "Build CPU ARM64"
+    def container_type = "aarch64"
+    def docker_binary = "docker"
+    def wheel_tag = "manylinux2014_aarch64"
+    sh """
+    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/build_via_cmake.sh --conda-env=aarch64_test -DOPEN_MP:BOOL=ON -DHIDE_CXX_SYMBOL=ON
+    ${dockerRun} ${container_type} ${docker_binary} bash -c "cd build && ctest --extra-verbose"
+    ${dockerRun} ${container_type} ${docker_binary} bash -c "cd python-package && rm -rf dist/* && python setup.py bdist_wheel --universal"
+    ${dockerRun} ${container_type} ${docker_binary} python tests/ci_build/rename_whl.py python-package/dist/*.whl ${commit_id} ${wheel_tag}
+    ${dockerRun} ${container_type} ${docker_binary} auditwheel repair --plat ${wheel_tag} python-package/dist/*.whl
+    mv -v wheelhouse/*.whl python-package/dist/
+    # Make sure that libgomp.so is vendored in the wheel
+    ${dockerRun} ${container_type} ${docker_binary} bash -c "unzip -l python-package/dist/*.whl | grep libgomp  || exit -1"
+    """
+    echo 'Stashing Python wheel...'
+    stash name: "xgboost_whl_arm64_cpu", includes: 'python-package/dist/*.whl'
+    if (env.BRANCH_NAME == 'master' || env.BRANCH_NAME.startsWith('release')) {
+      echo 'Uploading Python wheel...'
+      path = ("${BRANCH_NAME}" == 'master') ? '' : "${BRANCH_NAME}/"
+      s3Upload bucket: 'xgboost-nightly-builds', path: path, acl: 'PublicRead', workingDir: 'python-package/dist', includePathPattern:'**/*.whl'
+    }
+    stash name: 'xgboost_cli_arm64', includes: 'xgboost'
+    deleteDir()
+  }
+}
+
 def BuildCPUMock() {
   node('linux && cpu') {
     unstash name: 'srcs'
@@ -222,48 +211,100 @@ def BuildCPUMock() {
   }
 }
 
-def BuildCPUNonOmp() {
-  node('linux && cpu') {
+def BuildCUDA(args) {
+  node('linux && cpu_build') {
     unstash name: 'srcs'
-    echo "Build CPU without OpenMP"
-    def container_type = "cpu"
+    echo "Build with CUDA ${args.cuda_version}"
+    def container_type = GetCUDABuildContainerType(args.cuda_version)
     def docker_binary = "docker"
+    def docker_args = "--build-arg CUDA_VERSION_ARG=${args.cuda_version}"
+    def arch_flag = ""
+    if (env.BRANCH_NAME != 'master' && !(env.BRANCH_NAME.startsWith('release'))) {
+      arch_flag = "-DGPU_COMPUTE_VER=75"
+    }
+    def wheel_tag = "manylinux2010_x86_64"
     sh """
-    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/build_via_cmake.sh -DUSE_OPENMP=OFF
-    """
-    echo "Running Non-OpenMP C++ test..."
-    sh """
-    ${dockerRun} ${container_type} ${docker_binary} build/testxgboost
+    ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/build_via_cmake.sh -DUSE_CUDA=ON -DUSE_NCCL=ON -DOPEN_MP:BOOL=ON -DHIDE_CXX_SYMBOLS=ON ${arch_flag}
+    ${dockerRun} ${container_type} ${docker_binary} ${docker_args} bash -c "cd python-package && rm -rf dist/* && python setup.py bdist_wheel --universal"
+    ${dockerRun} ${container_type} ${docker_binary} ${docker_args} python tests/ci_build/rename_whl.py python-package/dist/*.whl ${commit_id} ${wheel_tag}
     """
+    if (args.cuda_version == ref_cuda_ver) {
+      sh """
+      ${dockerRun} auditwheel_x86_64 ${docker_binary} auditwheel repair --plat ${wheel_tag} python-package/dist/*.whl
+      mv -v wheelhouse/*.whl python-package/dist/
+      # Make sure that libgomp.so is vendored in the wheel
+      ${dockerRun} auditwheel_x86_64 ${docker_binary} bash -c "unzip -l python-package/dist/*.whl | grep libgomp  || exit -1"
+      """
+    }
+    echo 'Stashing Python wheel...'
+    stash name: "xgboost_whl_cuda${args.cuda_version}", includes: 'python-package/dist/*.whl'
+    if (args.cuda_version == ref_cuda_ver && (env.BRANCH_NAME == 'master' || env.BRANCH_NAME.startsWith('release'))) {
+      echo 'Uploading Python wheel...'
+      path = ("${BRANCH_NAME}" == 'master') ? '' : "${BRANCH_NAME}/"
+      s3Upload bucket: 'xgboost-nightly-builds', path: path, acl: 'PublicRead', workingDir: 'python-package/dist', includePathPattern:'**/*.whl'
+    }
+    echo 'Stashing C++ test executable (testxgboost)...'
+    stash name: "xgboost_cpp_tests_cuda${args.cuda_version}", includes: 'build/testxgboost'
+    if (args.build_rmm) {
+      echo "Build with CUDA ${args.cuda_version} and RMM"
+      container_type = "rmm"
+      docker_binary = "docker"
+      docker_args = "--build-arg CUDA_VERSION_ARG=${args.cuda_version}"
+      sh """
+      rm -rf build/
+      ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/build_via_cmake.sh --conda-env=gpu_test -DUSE_CUDA=ON -DUSE_NCCL=ON -DPLUGIN_RMM=ON ${arch_flag}
+      ${dockerRun} ${container_type} ${docker_binary} ${docker_args} bash -c "cd python-package && rm -rf dist/* && python setup.py bdist_wheel --universal"
+      ${dockerRun} ${container_type} ${docker_binary} ${docker_args} python tests/ci_build/rename_whl.py python-package/dist/*.whl ${commit_id} manylinux2010_x86_64
+      """
+      echo 'Stashing Python wheel...'
+      stash name: "xgboost_whl_rmm_cuda${args.cuda_version}", includes: 'python-package/dist/*.whl'
+      echo 'Stashing C++ test executable (testxgboost)...'
+      stash name: "xgboost_cpp_tests_rmm_cuda${args.cuda_version}", includes: 'build/testxgboost'
+    }
     deleteDir()
   }
 }
 
-def BuildCUDA(args) {
-  node('linux && cpu') {
+def BuildRPackageWithCUDA(args) {
+  node('linux && cpu_build') {
     unstash name: 'srcs'
-    echo "Build with CUDA ${args.cuda_version}"
-    def container_type = "gpu_build"
+    def container_type = 'gpu_build_r_centos6'
     def docker_binary = "docker"
-    def docker_args = "--build-arg CUDA_VERSION=${args.cuda_version}"
-    sh """
-    ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/build_via_cmake.sh -DUSE_CUDA=ON -DUSE_NCCL=ON -DOPEN_MP:BOOL=ON -DHIDE_CXX_SYMBOLS=ON
-    ${dockerRun} ${container_type} ${docker_binary} ${docker_args} bash -c "cd python-package && rm -rf dist/* && python setup.py bdist_wheel --universal"
-    ${dockerRun} ${container_type} ${docker_binary} ${docker_args} python3 tests/ci_build/rename_whl.py python-package/dist/*.whl ${commit_id} manylinux2010_x86_64
-    """
-    // Stash wheel for CUDA 9.0 target
-    if (args.cuda_version == '9.0') {
-      echo 'Stashing Python wheel...'
-      stash name: 'xgboost_whl_cuda9', includes: 'python-package/dist/*.whl'
+    def docker_args = "--build-arg CUDA_VERSION_ARG=10.0"
+    if (env.BRANCH_NAME == 'master' || env.BRANCH_NAME.startsWith('release')) {
+      sh """
+      ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/build_r_pkg_with_cuda.sh ${commit_id}
+      """
+      echo 'Uploading R tarball...'
       path = ("${BRANCH_NAME}" == 'master') ? '' : "${BRANCH_NAME}/"
-      s3Upload bucket: 'xgboost-nightly-builds', path: path, acl: 'PublicRead', workingDir: 'python-package/dist', includePathPattern:'**/*.whl'
-      echo 'Stashing C++ test executable (testxgboost)...'
-      stash name: 'xgboost_cpp_tests', includes: 'build/testxgboost'
+      s3Upload bucket: 'xgboost-nightly-builds', path: path, acl: 'PublicRead', includePathPattern:'xgboost_r_gpu_linux_*.tar.gz'
     }
     deleteDir()
   }
 }
 
+def BuildJVMPackagesWithCUDA(args) {
+  node('linux && mgpu') {
+    unstash name: 'srcs'
+    echo "Build XGBoost4J-Spark with Spark ${args.spark_version}, CUDA ${args.cuda_version}"
+    def container_type = "jvm_gpu_build"
+    def docker_binary = "nvidia-docker"
+    def docker_args = "--build-arg CUDA_VERSION_ARG=${args.cuda_version}"
+    def arch_flag = ""
+    if (env.BRANCH_NAME != 'master' && !(env.BRANCH_NAME.startsWith('release'))) {
+      arch_flag = "-DGPU_COMPUTE_VER=75"
+    }
+    // Use only 4 CPU cores
+    def docker_extra_params = "CI_DOCKER_EXTRA_PARAMS_INIT='--cpuset-cpus 0-3'"
+    sh """
+    ${docker_extra_params} ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/build_jvm_packages.sh ${args.spark_version} -Duse.cuda=ON $arch_flag
+    """
+    echo "Stashing XGBoost4J JAR with CUDA ${args.cuda_version} ..."
+    stash name: 'xgboost4j_jar_gpu', includes: "jvm-packages/xgboost4j-gpu/target/*.jar,jvm-packages/xgboost4j-spark-gpu/target/*.jar"
+    deleteDir()
+  }
+}
+
 def BuildJVMPackages(args) {
   node('linux && cpu') {
     unstash name: 'srcs'
@@ -290,15 +331,17 @@ def BuildJVMDoc() {
     sh """
     ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/build_jvm_doc.sh ${BRANCH_NAME}
     """
-    echo 'Uploading doc...'
-    s3Upload file: "jvm-packages/${BRANCH_NAME}.tar.bz2", bucket: 'xgboost-docs', acl: 'PublicRead', path: "${BRANCH_NAME}.tar.bz2"
+    if (env.BRANCH_NAME == 'master' || env.BRANCH_NAME.startsWith('release')) {
+      echo 'Uploading doc...'
+      s3Upload file: "jvm-packages/${BRANCH_NAME}.tar.bz2", bucket: 'xgboost-docs', acl: 'PublicRead', path: "${BRANCH_NAME}.tar.bz2"
+    }
     deleteDir()
   }
 }
 
 def TestPythonCPU() {
   node('linux && cpu') {
-    unstash name: 'xgboost_whl_cuda9'
+    unstash name: "xgboost_whl_cuda${ref_cuda_ver}"
     unstash name: 'srcs'
     unstash name: 'xgboost_cli'
     echo "Test Python CPU"
@@ -306,78 +349,72 @@ def TestPythonCPU() {
     def docker_binary = "docker"
     sh """
     ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/test_python.sh cpu
-    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/test_python.sh cpu-py35
+    """
+    deleteDir()
+  }
+}
+
+def TestPythonCPUARM64() {
+  node('linux && arm64') {
+    unstash name: "xgboost_whl_arm64_cpu"
+    unstash name: 'srcs'
+    unstash name: 'xgboost_cli_arm64'
+    echo "Test Python CPU ARM64"
+    def container_type = "aarch64"
+    def docker_binary = "docker"
+    sh """
+    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/test_python.sh cpu-arm64
     """
     deleteDir()
   }
 }
 
 def TestPythonGPU(args) {
-  nodeReq = (args.multi_gpu) ? 'linux && mgpu' : 'linux && gpu'
+  def nodeReq = (args.multi_gpu) ? 'linux && mgpu' : 'linux && gpu'
+  def artifact_cuda_version = (args.artifact_cuda_version) ?: ref_cuda_ver
   node(nodeReq) {
-    unstash name: 'xgboost_whl_cuda9'
+    unstash name: "xgboost_whl_cuda${artifact_cuda_version}"
+    unstash name: "xgboost_cpp_tests_cuda${artifact_cuda_version}"
     unstash name: 'srcs'
-    echo "Test Python GPU: CUDA ${args.cuda_version}"
+    echo "Test Python GPU: CUDA ${args.host_cuda_version}"
     def container_type = "gpu"
     def docker_binary = "nvidia-docker"
-    def docker_args = "--build-arg CUDA_VERSION=${args.cuda_version}"
-    if (args.multi_gpu) {
-      echo "Using multiple GPUs"
-      sh """
-      ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/test_python.sh mgpu
-      """
-      if (args.cuda_version != '9.0') {
-        echo "Running tests with cuDF..."
-        sh """
-        ${dockerRun} cudf ${docker_binary} ${docker_args} tests/ci_build/test_python.sh mgpu-cudf
-        """
-      }
-    } else {
-      echo "Using a single GPU"
-      sh """
-      ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/test_python.sh gpu
-      """
-      if (args.cuda_version != '9.0') {
-        echo "Running tests with cuDF..."
-        sh """
-        ${dockerRun} cudf ${docker_binary} ${docker_args} tests/ci_build/test_python.sh cudf
-        """
-      }
+    def docker_args = "--build-arg CUDA_VERSION_ARG=${args.host_cuda_version}"
+    def mgpu_indicator = (args.multi_gpu) ? 'mgpu' : 'gpu'
+    // Allocate extra space in /dev/shm to enable NCCL
+    def docker_extra_params = (args.multi_gpu) ? "CI_DOCKER_EXTRA_PARAMS_INIT='--shm-size=4g'" : ''
+    sh "${docker_extra_params} ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/test_python.sh ${mgpu_indicator}"
+    if (args.test_rmm) {
+      sh "rm -rfv build/ python-package/dist/"
+      unstash name: "xgboost_whl_rmm_cuda${args.host_cuda_version}"
+      unstash name: "xgboost_cpp_tests_rmm_cuda${args.host_cuda_version}"
+      sh "${docker_extra_params} ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/test_python.sh ${mgpu_indicator} --use-rmm-pool"
     }
-    // For CUDA 10.0 target, run cuDF tests too
-    deleteDir()
-  }
-}
-
-def TestCppRabit() {
-  node(nodeReq) {
-    unstash name: 'xgboost_rabit_tests'
-    unstash name: 'srcs'
-    echo "Test C++, rabit mock on"
-    def container_type = "cpu"
-    def docker_binary = "docker"
-    sh """
-    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/runxgb.sh xgboost tests/ci_build/approx.conf.in
-    """
     deleteDir()
   }
 }
 
 def TestCppGPU(args) {
-  nodeReq = (args.multi_gpu) ? 'linux && mgpu' : 'linux && gpu'
+  def nodeReq = 'linux && mgpu'
+  def artifact_cuda_version = (args.artifact_cuda_version) ?: ref_cuda_ver
   node(nodeReq) {
-    unstash name: 'xgboost_cpp_tests'
+    unstash name: "xgboost_cpp_tests_cuda${artifact_cuda_version}"
     unstash name: 'srcs'
-    echo "Test C++, CUDA ${args.cuda_version}"
+    echo "Test C++, CUDA ${args.host_cuda_version}"
     def container_type = "gpu"
     def docker_binary = "nvidia-docker"
-    def docker_args = "--build-arg CUDA_VERSION=${args.cuda_version}"
-    if (args.multi_gpu) {
-      echo "Using multiple GPUs"
-      sh "${dockerRun} ${container_type} ${docker_binary} ${docker_args} build/testxgboost --gtest_filter=*.MGPU_*"
-    } else {
-      echo "Using a single GPU"
-      sh "${dockerRun} ${container_type} ${docker_binary} ${docker_args} build/testxgboost --gtest_filter=-*.MGPU_*"
+    def docker_args = "--build-arg CUDA_VERSION_ARG=${args.host_cuda_version}"
+    sh "${dockerRun} ${container_type} ${docker_binary} ${docker_args} build/testxgboost"
+    if (args.test_rmm) {
+      sh "rm -rfv build/"
+      unstash name: "xgboost_cpp_tests_rmm_cuda${args.host_cuda_version}"
+      echo "Test C++, CUDA ${args.host_cuda_version} with RMM"
+      container_type = "rmm"
+      docker_binary = "nvidia-docker"
+      docker_args = "--build-arg CUDA_VERSION_ARG=${args.host_cuda_version}"
+      sh """
+      ${dockerRun} ${container_type} ${docker_binary} ${docker_args} bash -c "source activate gpu_test && build/testxgboost --use-rmm-pool --gtest_filter=-*DeathTest.*"
+      """
     }
     deleteDir()
   }
@@ -405,30 +442,13 @@ def CrossTestJVMwithJDK(args) {
   }
 }
 
-def TestR(args) {
-  node('linux && cpu') {
-    unstash name: 'srcs'
-    echo "Test R package"
-    def container_type = "rproject"
-    def docker_binary = "docker"
-    def use_r35_flag = (args.use_r35) ? "1" : "0"
-    def docker_args = "--build-arg USE_R35=${use_r35_flag}"
-    sh """
-    ${dockerRun} ${container_type} ${docker_binary} ${docker_args} tests/ci_build/build_test_rpkg.sh || tests/ci_build/print_r_stacktrace.sh
-    """
-    deleteDir()
-  }
-}
-
 def DeployJVMPackages(args) {
   node('linux && cpu') {
     unstash name: 'srcs'
     if (env.BRANCH_NAME == 'master' || env.BRANCH_NAME.startsWith('release')) {
       echo 'Deploying to xgboost-maven-repo S3 repo...'
-      def container_type = "jvm"
-      def docker_binary = "docker"
       sh """
-      ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/deploy_jvm_packages.sh ${args.spark_version}
+      ${dockerRun} jvm_gpu_build docker --build-arg CUDA_VERSION_ARG=10.0 tests/ci_build/deploy_jvm_packages.sh ${args.spark_version}
       """
     }
     deleteDir()

Jenkinsfile-win64

@@ -10,17 +10,29 @@ def commit_id   // necessary to pass a variable from one stage to another
 
 pipeline {
   agent none
+
+  // Setup common job properties
+  options {
+    timestamps()
+    timeout(time: 240, unit: 'MINUTES')
+    buildDiscarder(logRotator(numToKeepStr: '10'))
+    preserveStashes()
+  }
+
   // Build stages
   stages {
-    stage('Jenkins Win64: Get sources') {
-      agent { label 'win64 && build' }
+    stage('Jenkins Win64: Initialize') {
+      agent { label 'job_initializer' }
       steps {
         script {
+          def buildNumber = env.BUILD_NUMBER as int
+          if (buildNumber > 1) milestone(buildNumber - 1)
+          milestone(buildNumber)
           checkoutSrcs()
           commit_id = "${GIT_COMMIT}"
         }
+        sh 'python3 tests/jenkins_get_approval.py'
         stash name: 'srcs'
-        milestone ordinal: 1
       }
     }
     stage('Jenkins Win64: Build') {
@@ -28,10 +40,9 @@ pipeline {
       steps {
         script {
           parallel ([
-            'build-win64-cuda9.0': { BuildWin64() }
+            'build-win64-cuda10.1': { BuildWin64() }
           ])
         }
-        milestone ordinal: 2
       }
     }
     stage('Jenkins Win64: Test') {
@@ -39,13 +50,9 @@ pipeline {
       steps {
         script {
           parallel ([
-            'test-win64-cpu': { TestWin64CPU() },
-            'test-win64-gpu-cuda9.0': { TestWin64GPU(cuda_target: 'cuda9') },
-            'test-win64-gpu-cuda10.0': { TestWin64GPU(cuda_target: 'cuda10_0') },
-            'test-win64-gpu-cuda10.1': { TestWin64GPU(cuda_target: 'cuda10_1') }
+            'test-win64-cuda10.1': { TestWin64() },
           ])
         }
-        milestone ordinal: 3
       }
     }
   }
@@ -67,14 +74,18 @@ def checkoutSrcs() {
 }
 
 def BuildWin64() {
-  node('win64 && build') {
+  node('win64 && cuda10_unified') {
     unstash name: 'srcs'
     echo "Building XGBoost for Windows AMD64 target..."
     bat "nvcc --version"
+    def arch_flag = ""
+    if (env.BRANCH_NAME != 'master' && !(env.BRANCH_NAME.startsWith('release'))) {
+      arch_flag = "-DGPU_COMPUTE_VER=75"
+    }
     bat """
     mkdir build
     cd build
-    cmake .. -G"Visual Studio 15 2017 Win64" -DUSE_CUDA=ON -DCMAKE_VERBOSE_MAKEFILE=ON -DGOOGLE_TEST=ON -DUSE_DMLC_GTEST=ON
+    cmake .. -G"Visual Studio 15 2017 Win64" -DUSE_CUDA=ON -DCMAKE_VERBOSE_MAKEFILE=ON -DGOOGLE_TEST=ON -DUSE_DMLC_GTEST=ON ${arch_flag} -DCMAKE_UNITY_BUILD=ON
     """
     bat """
     cd build
@@ -92,8 +103,11 @@ def BuildWin64() {
     """
     echo 'Stashing Python wheel...'
     stash name: 'xgboost_whl', includes: 'python-package/dist/*.whl'
-    path = ("${BRANCH_NAME}" == 'master') ? '' : "${BRANCH_NAME}/"
-    s3Upload bucket: 'xgboost-nightly-builds', path: path, acl: 'PublicRead', workingDir: 'python-package/dist', includePathPattern:'**/*.whl'
+    if (env.BRANCH_NAME == 'master' || env.BRANCH_NAME.startsWith('release')) {
+      echo 'Uploading Python wheel...'
+      path = ("${BRANCH_NAME}" == 'master') ? '' : "${BRANCH_NAME}/"
+      s3Upload bucket: 'xgboost-nightly-builds', path: path, acl: 'PublicRead', workingDir: 'python-package/dist', includePathPattern:'**/*.whl'
+    }
     echo 'Stashing C++ test executable (testxgboost)...'
     stash name: 'xgboost_cpp_tests', includes: 'build/testxgboost.exe'
     stash name: 'xgboost_cli', includes: 'xgboost.exe'
@@ -101,51 +115,29 @@ def BuildWin64() {
   }
 }
 
-def TestWin64CPU() {
-  node('win64 && cpu') {
+def TestWin64() {
+  node('win64 && cuda10_unified') {
     unstash name: 'srcs'
     unstash name: 'xgboost_whl'
     unstash name: 'xgboost_cli'
-    echo "Test Win64 CPU"
-    echo "Installing Python wheel..."
-    bat "conda activate && (python -m pip uninstall -y xgboost || cd .)"
-    bat """
-    conda activate && for /R %%i in (python-package\\dist\\*.whl) DO python -m pip install "%%i"
-    """
-    echo "Installing Python dependencies..."
-    bat """
-     conda activate && conda upgrade scikit-learn pandas numpy
-    """
-    echo "Running Python tests..."
-    bat "conda activate && python -m pytest -v -s --fulltrace tests\\python"
-    bat "conda activate && python -m pip uninstall -y xgboost"
-    deleteDir()
-  }
-}
-
-def TestWin64GPU(args) {
-  node("win64 && gpu && ${args.cuda_target}") {
-    unstash name: 'srcs'
-    unstash name: 'xgboost_whl'
     unstash name: 'xgboost_cpp_tests'
-    echo "Test Win64 GPU (${args.cuda_target})"
+    echo "Test Win64"
     bat "nvcc --version"
     echo "Running C++ tests..."
     bat "build\\testxgboost.exe"
-    echo "Installing Python wheel..."
-    bat "conda activate && (python -m pip uninstall -y xgboost || cd .)"
-    bat """
-    conda activate && for /R %%i in (python-package\\dist\\*.whl) DO python -m pip install "%%i"
-    """
     echo "Installing Python dependencies..."
+    def env_name = 'win64_' + UUID.randomUUID().toString().replaceAll('-', '')
+    bat "conda env create -n ${env_name} --file=tests/ci_build/conda_env/win64_test.yml"
+    echo "Installing Python wheel..."
     bat """
-     conda activate && conda upgrade scikit-learn pandas numpy
+    conda activate ${env_name} && for /R %%i in (python-package\\dist\\*.whl) DO python -m pip install "%%i"
     """
     echo "Running Python tests..."
+    bat "conda activate ${env_name} && python -m pytest -v -s -rxXs --fulltrace tests\\python"
     bat """
-    conda activate && python -m pytest -v -s --fulltrace -m "(not slow) and (not mgpu)" tests\\python-gpu
+    conda activate ${env_name} && python -m pytest -v -s -rxXs --fulltrace -m "(not slow) and (not mgpu)" tests\\python-gpu
     """
-    bat "conda activate && python -m pip uninstall -y xgboost"
+    bat "conda env remove --name ${env_name}"
     deleteDir()
   }
 }

Makefile

@@ -44,7 +44,7 @@ export CXX = g++
 endif
 endif
 
-export CFLAGS= -DDMLC_LOG_CUSTOMIZE=1 -std=c++11 -Wall -Wno-unknown-pragmas -Iinclude $(ADD_CFLAGS)
+export CFLAGS= -DDMLC_LOG_CUSTOMIZE=1 -std=c++14 -Wall -Wno-unknown-pragmas -Iinclude $(ADD_CFLAGS)
 CFLAGS += -I$(DMLC_CORE)/include -I$(RABIT)/include -I$(GTEST_PATH)/include
 
 ifeq ($(TEST_COVER), 1)
@@ -86,6 +86,15 @@ cover: check
 	)
 endif
 
+
+# dask is required to pass, others are not
+# If any of the dask tests failed, contributor won't see the other error.
+mypy:
+	cd python-package; \
+	mypy ./xgboost/dask.py ../tests/python/test_with_dask.py --follow-imports=silent; \
+	mypy ../tests/python-gpu/test_gpu_with_dask.py --follow-imports=silent; \
+	mypy . || true ;
+
 clean:
 	$(RM) -rf build lib bin *~ */*~ */*/*~ */*/*/*~ */*.o */*/*.o */*/*/*.o #xgboost
 	$(RM) -rf build_tests *.gcov tests/cpp/xgboost_test
@@ -134,14 +143,18 @@ Rpack: clean_all
 	sed -i -e 's/@OPENMP_LIB@//g' xgboost/src/Makevars.win
 	rm -f xgboost/src/Makevars.win-e   # OSX sed create this extra file; remove it
 	bash R-package/remove_warning_suppression_pragma.sh
+	bash xgboost/remove_warning_suppression_pragma.sh
 	rm xgboost/remove_warning_suppression_pragma.sh
+	rm -rfv xgboost/tests/helper_scripts/
+
+R ?= R
 
 Rbuild: Rpack
-	R CMD build --no-build-vignettes xgboost
+	$(R) CMD build xgboost
 	rm -rf xgboost
 
 Rcheck: Rbuild
-	R CMD check xgboost*.tar.gz
+	$(R) CMD check --as-cran xgboost*.tar.gz
 
 -include build/*.d
 -include build/*/*.d

NEWS.md

@@ -3,6 +3,639 @@ XGBoost Change Log
 
 This file records the changes in xgboost library in reverse chronological order.
 
+## v1.3.0 (2020.12.08)
+
+### XGBoost4J-Spark: Exceptions should cancel jobs gracefully instead of killing SparkContext (#6019).
+* By default, exceptions in XGBoost4J-Spark causes the whole SparkContext to shut down, necessitating the restart of the Spark cluster. This behavior is often a major inconvenience.
+* Starting from 1.3.0 release, XGBoost adds a new parameter `killSparkContextOnWorkerFailure` to optionally prevent killing SparkContext. If this parameter is set, exceptions will gracefully cancel training jobs instead of killing SparkContext.
+
+### GPUTreeSHAP: GPU acceleration of the TreeSHAP algorithm (#6038, #6064, #6087, #6099, #6163, #6281, #6332)
+* [SHAP (SHapley Additive exPlanations)](https://github.com/slundberg/shap) is a game theoretic approach to explain predictions of machine learning models. It computes feature importance scores for individual examples, establishing how each feature influences a particular prediction. TreeSHAP is an optimized SHAP algorithm specifically designed for decision tree ensembles.
+* Starting with 1.3.0 release, it is now possible to leverage CUDA-capable GPUs to accelerate the TreeSHAP algorithm. Check out [the demo notebook](https://github.com/dmlc/xgboost/blob/master/demo/gpu_acceleration/shap.ipynb).
+* The CUDA implementation of the TreeSHAP algorithm is hosted at [rapidsai/GPUTreeSHAP](https://github.com/rapidsai/gputreeshap). XGBoost imports it as a Git submodule.
+
+### New style Python callback API (#6199, #6270, #6320, #6348, #6376, #6399, #6441)
+* The XGBoost Python package now offers a re-designed callback API. The new callback API lets you design various extensions of training in idomatic Python. In addition, the new callback API allows you to use early stopping with the native Dask API (`xgboost.dask`). Check out [the tutorial](https://xgboost.readthedocs.io/en/release_1.3.0/python/callbacks.html) and [the demo](https://github.com/dmlc/xgboost/blob/master/demo/guide-python/callbacks.py).
+
+### Enable the use of `DeviceQuantileDMatrix` / `DaskDeviceQuantileDMatrix` with large data (#6201, #6229, #6234).
+* `DeviceQuantileDMatrix` can achieve memory saving by avoiding extra copies of the training data, and the saving is bigger for large data. Unfortunately, large data with more than 2^31 elements was triggering integer overflow bugs in CUB and Thrust. Tracking issue: #6228.
+* This release contains a series of work-arounds to allow the use of `DeviceQuantileDMatrix` with large data:
+  - Loop over `copy_if` (#6201)
+  - Loop over `thrust::reduce` (#6229)
+  - Implement the inclusive scan algorithm in-house, to handle large offsets (#6234)
+
+### Support slicing of tree models (#6302)
+* Accessing the best iteration of a model after the application of early stopping used to be error-prone, need to manually pass the `ntree_limit` argument to the `predict()` function.
+* Now we provide a simple interface to slice tree models by specifying a range of boosting rounds. The tree ensemble can be split into multiple sub-ensembles via the slicing interface. Check out [an example](https://xgboost.readthedocs.io/en/release_1.3.0/python/model.html).
+* In addition, the early stopping callback now supports `save_best` option. When enabled, XGBoost will save (persist) the model at the best boosting round and discard the trees that were fit subsequent to the best round.
+
+### Weighted subsampling of features (columns) (#5962)
+* It is now possible to sample features (columns) via weighted subsampling, in which features with higher weights are more likely to be selected in the sample. Weighted subsampling allows you to encode domain knowledge by emphasizing a particular set of features in the choice of tree splits. In addition, you can prevent particular features from being used in any splits, by assigning them zero weights.
+* Check out [the demo](https://github.com/dmlc/xgboost/blob/master/demo/guide-python/feature_weights.py).
+
+### Improved integration with Dask
+* Support reverse-proxy environment such as Google Kubernetes Engine (#6343, #6475)
+* An XGBoost training job will no longer use all available workers. Instead, it will only use the workers that contain input data (#6343).
+* The new callback API works well with the Dask training API.
+* The `predict()` and `fit()` function of `DaskXGBClassifier` and `DaskXGBRegressor` now accept a base margin (#6155).
+* Support more meta data in the Dask API (#6130, #6132, #6333).
+* Allow passing extra keyword arguments as `kwargs` in `predict()` (#6117)
+* Fix typo in dask interface: `sample_weights` -> `sample_weight` (#6240)
+* Allow empty data matrix in AFT survival, as Dask may produce empty partitions (#6379)
+* Speed up prediction by overlapping prediction jobs in all workers (#6412)
+
+### Experimental support for direct splits with categorical features (#6028, #6128, #6137, #6140, #6164, #6165, #6166, #6179, #6194, #6219)
+* Currently, XGBoost requires users to one-hot-encode categorical variables. This has adverse performance implications, as the creation of many dummy variables results into higher memory consumption and may require fitting deeper trees to achieve equivalent model accuracy.
+* The 1.3.0 release of XGBoost contains an experimental support for direct handling of categorical variables in test nodes. Each test node will have the condition of form `feature_value \in match_set`, where the `match_set` on the right hand side contains one or more matching categories. The matching categories in `match_set` represent the condition for traversing to the right child node. Currently, XGBoost will only generate categorical splits with only a single matching category ("one-vs-rest split"). In a future release, we plan to remove this restriction and produce splits with multiple matching categories in `match_set`.
+* The categorical split requires the use of JSON model serialization. The legacy binary serialization method cannot be used to save (persist) models with categorical splits.
+* Note. This feature is currently highly experimental. Use it at your own risk. See the detailed list of limitations at [#5949](https://github.com/dmlc/xgboost/pull/5949).
+
+### Experimental plugin for RAPIDS Memory Manager (#5873, #6131, #6146, #6150, #6182)
+* RAPIDS Memory Manager library ([rapidsai/rmm](https://github.com/rapidsai/rmm)) provides a collection of efficient memory allocators for NVIDIA GPUs. It is now possible to use XGBoost with memory allocators provided by RMM, by enabling the RMM integration plugin. With this plugin, XGBoost is now able to share a common GPU memory pool with other applications using RMM, such as the RAPIDS data science packages.
+* See [the demo](https://github.com/dmlc/xgboost/blob/master/demo/rmm_plugin/README.md) for a working example, as well as directions for building XGBoost with the RMM plugin.
+* The plugin will be soon considered non-experimental, once #6297 is resolved.
+
+### Experimental plugin for oneAPI programming model (#5825)
+* oneAPI is a programming interface developed by Intel aimed at providing one programming model for many types of hardware such as CPU, GPU, FGPA and other hardware accelerators.
+* XGBoost now includes an experimental plugin for using oneAPI for the predictor and objective functions. The plugin is hosted in the directory `plugin/updater_oneapi`.
+* Roadmap: #5442
+
+### Pickling the XGBoost model will now trigger JSON serialization (#6027)
+* The pickle will now contain the JSON string representation of the XGBoost model, as well as related configuration.
+
+### Performance improvements
+* Various performance improvement on multi-core CPUs
+  - Optimize DMatrix build time by up to 3.7x. (#5877)
+  - CPU predict performance improvement, by up to 3.6x. (#6127)
+  - Optimize CPU sketch allreduce for sparse data (#6009)
+  - Thread local memory allocation for BuildHist, leading to speedup up to 1.7x. (#6358)
+  - Disable hyperthreading for DMatrix creation (#6386). This speeds up DMatrix creation by up to 2x.
+  - Simple fix for static shedule in predict (#6357)
+* Unify thread configuration, to make it easy to utilize all CPU cores (#6186)
+* [jvm-packages] Clean the way deterministic paritioning is computed (#6033)
+* Speed up JSON serialization by implementing an intrusive pointer class (#6129). It leads to 1.5x-2x performance boost.
+
+### API additions
+* [R] Add SHAP summary plot using ggplot2 (#5882)
+* Modin DataFrame can now be used as input (#6055)
+* [jvm-packages] Add `getNumFeature` method (#6075)
+* Add MAPE metric (#6119)
+* Implement GPU predict leaf. (#6187)
+* Enable cuDF/cuPy inputs in `XGBClassifier` (#6269)
+* Document tree method for feature weights. (#6312)
+* Add `fail_on_invalid_gpu_id` parameter, which will cause XGBoost to terminate upon seeing an invalid value of `gpu_id` (#6342)
+
+### Breaking: the default evaluation metric for classification is changed to `logloss` / `mlogloss` (#6183)
+* The default metric used to be accuracy, and it is not statistically consistent to perform early stopping with the accuracy metric when we are really optimizing the log loss for the `binary:logistic` objective.
+* For statistical consistency, the default metric for classification has been changed to `logloss`. Users may choose to preserve the old behavior by explicitly specifying `eval_metric`.
+
+### Breaking: `skmaker` is now removed (#5971)
+* The `skmaker` updater has not been documented nor tested.
+
+### Breaking: the JSON model format no longer stores the leaf child count (#6094).
+* The leaf child count field has been deprecated and is not used anywhere in the XGBoost codebase.
+
+### Breaking: XGBoost now requires MacOS 10.14 (Mojave) and later.
+* Homebrew has dropped support for MacOS 10.13 (High Sierra), so we are not able to install the OpenMP runtime (`libomp`) from Homebrew on MacOS 10.13. Please use MacOS 10.14 (Mojave) or later.
+
+### Deprecation notices
+* The use of `LabelEncoder` in `XGBClassifier` is now deprecated and will be removed in the next minor release (#6269). The deprecation is necessary to support multiple types of inputs, such as cuDF data frames or cuPy arrays.
+* The use of certain positional arguments in the Python interface is deprecated (#6365). Users will use deprecation warnings for the use of position arguments for certain function parameters. New code should use keyword arguments as much as possible. We have not yet decided when we will fully require the use of keyword arguments.
+
+### Bug-fixes
+* On big-endian arch, swap the byte order in the binary serializer to enable loading models that were produced by a little-endian machine (#5813).
+* [jvm-packages] Fix deterministic partitioning with dataset containing Double.NaN (#5996)
+* Limit tree depth for GPU hist to 31 to prevent integer overflow (#6045)
+* [jvm-packages] Set `maxBins` to 256 to align with the default value in the C++ code (#6066)
+* [R] Fix CRAN check (#6077)
+* Add back support for `scipy.sparse.coo_matrix` (#6162)
+* Handle duplicated values in sketching. (#6178)
+* Catch all standard exceptions in C API. (#6220)
+* Fix linear GPU input (#6255)
+* Fix inplace prediction interval. (#6259)
+* [R] allow `xgb.plot.importance()` calls to fill a grid (#6294)
+* Lazy import dask libraries. (#6309)
+* Deterministic data partitioning for external memory (#6317)
+* Avoid resetting seed for every configuration. (#6349)
+* Fix label errors in graph visualization (#6369)
+* [jvm-packages] fix potential unit test suites aborted issue due to race condition (#6373)
+* [R] Fix warnings from `R check --as-cran` (#6374)
+* [R] Fix a crash that occurs with noLD R (#6378)
+* [R] Do not convert continuous labels to factors (#6380)
+* [R] remove uses of `exists()` (#6387)
+* Propagate parameters to the underlying `Booster` handle from `XGBClassifier.set_param` / `XGBRegressor.set_param`. (#6416)
+* [R] Fix R package installation via CMake (#6423)
+* Enforce row-major order in cuPy array (#6459)
+* Fix filtering callable objects in the parameters passed to the scikit-learn API. (#6466)
+
+### Maintenance: Testing, continuous integration, build system
+* [CI] Improve JVM test in GitHub Actions (#5930)
+* Refactor plotting test so that it can run independently (#6040)
+* [CI] Cancel builds on subsequent pushes (#6011)
+* Fix Dask Pytest fixture (#6024)
+* [CI] Migrate linters to GitHub Actions (#6035)
+* [CI] Remove win2016 JVM test from GitHub Actions (#6042)
+* Fix CMake build with `BUILD_STATIC_LIB` option (#6090)
+* Don't link imported target in CMake (#6093)
+* Work around a compiler bug in MacOS AppleClang 11 (#6103)
+* [CI] Fix CTest by running it in a correct directory (#6104)
+* [R] Check warnings explicitly for model compatibility tests (#6114)
+* [jvm-packages] add xgboost4j-gpu/xgboost4j-spark-gpu module to facilitate release (#6136)
+* [CI] Time GPU tests. (#6141)
+* [R] remove warning in configure.ac (#6152)
+* [CI] Upgrade cuDF and RMM to 0.16 nightlies; upgrade to Ubuntu 18.04 (#6157)
+* [CI] Test C API demo (#6159)
+* Option for generating device debug info. (#6168)
+* Update `.gitignore` (#6175, #6193, #6346)
+* Hide C++ symbols from dmlc-core (#6188)
+* [CI] Added arm64 job in Travis-CI (#6200)
+* [CI] Fix Docker build for CUDA 11 (#6202)
+* [CI] Move non-OpenMP gtest to GitHub Actions (#6210)
+* [jvm-packages] Fix up build for xgboost4j-gpu, xgboost4j-spark-gpu (#6216)
+* Add more tests for categorical data support (#6219)
+* [dask] Test for data initializaton. (#6226)
+* Bump junit from 4.11 to 4.13.1 in /jvm-packages/xgboost4j (#6230)
+* Bump junit from 4.11 to 4.13.1 in /jvm-packages/xgboost4j-gpu (#6233)
+* [CI] Reduce testing load with RMM (#6249)
+* [CI] Build a Python wheel for aarch64 platform (#6253)
+* [CI] Time the CPU tests on Jenkins. (#6257)
+* [CI] Skip Dask tests on ARM. (#6267)
+* Fix a typo in `is_arm()` in testing.py (#6271)
+* [CI] replace `egrep` with `grep -E` (#6287)
+* Support unity build. (#6295)
+* [CI] Mark flaky tests as XFAIL (#6299)
+* [CI] Use separate Docker cache for each CUDA version (#6305)
+* Added `USE_NCCL_LIB_PATH` option to enable user to set `NCCL_LIBRARY` during build  (#6310)
+* Fix flaky data initialization test. (#6318)
+* Add a badge for GitHub Actions (#6321)
+* Optional `find_package` for sanitizers. (#6329)
+* Use pytest conventions consistently in Python tests (#6337)
+* Fix missing space in warning message (#6340)
+* Update `custom_metric_obj.rst` (#6367)
+* [CI] Run R check with `--as-cran` flag on GitHub Actions (#6371)
+* [CI] Remove R check from Jenkins (#6372)
+* Mark GPU external memory test as XFAIL. (#6381)
+* [CI] Add noLD R test (#6382)
+* Fix MPI build. (#6403)
+* [CI] Upgrade to MacOS Mojave image (#6406)
+* Fix flaky sparse page dmatrix test. (#6417)
+* [CI] Upgrade cuDF and RMM to 0.17 nightlies (#6434)
+* [CI] Fix CentOS 6 Docker images (#6467)
+* [CI] Vendor libgomp in the manylinux Python wheel (#6461)
+* [CI] Hot fix for libgomp vendoring (#6482)
+
+### Maintenance: Clean up and merge the Rabit submodule (#6023, #6095, #6096, #6105, #6110, #6262, #6275, #6290)
+* The Rabit submodule is now maintained as part of the XGBoost codebase.
+* Tests for Rabit are now part of the test suites of XGBoost.
+* Rabit can now be built on the Windows platform.
+* We made various code re-formatting for the C++ code with clang-tidy.
+* Public headers of XGBoost no longer depend on Rabit headers.
+* Unused CMake targets for Rabit were removed.
+* Single-point model recovery has been dropped and removed from Rabit, simplifying the Rabit code greatly. The single-point model recovery feature has not been adequately maintained over the years.
+* We removed the parts of Rabit that were not useful for XGBoost.
+
+### Maintenance: Refactor code for legibility and maintainability
+* Unify CPU hist sketching (#5880)
+* [R] fix uses of 1:length(x) and other small things (#5992)
+* Unify evaluation functions. (#6037)
+* Make binary bin search reusable. (#6058)
+* Unify set index data. (#6062)
+* [R] Remove `stringi` dependency (#6109)
+* Merge extract cuts into QuantileContainer. (#6125)
+* Reduce C++ compiler warnings (#6197, #6198, #6213, #6286, #6325)
+* Cleanup Python code. (#6223)
+* Small cleanup to evaluator. (#6400)
+
+### Usability Improvements, Documentation
+* [jvm-packages] add example to handle missing value other than 0 (#5677)
+* Add DMatrix usage examples to the C API demo (#5854)
+* List `DaskDeviceQuantileDMatrix` in the doc. (#5975)
+* Update Python custom objective demo. (#5981)
+* Update the JSON model schema to document more objective functions. (#5982)
+* [Python] Fix warning when `missing` field is not used. (#5969)
+* Fix typo in tracker logging (#5994)
+* Move a warning about empty dataset, so that it's shown for all objectives and metrics (#5998)
+* Fix the instructions for installing the nightly build. (#6004)
+* [Doc] Add dtreeviz as a showcase example of integration with 3rd-party software (#6013)
+* [jvm-packages] [doc] Update install doc for JVM packages (#6051)
+* Fix typo in `xgboost.callback.early_stop` docstring (#6071)
+* Add cache suffix to the files used in the external memory demo. (#6088)
+* [Doc] Document the parameter `kill_spark_context_on_worker_failure` (#6097)
+* Fix link to the demo for custom objectives (#6100)
+* Update Dask doc. (#6108)
+* Validate weights are positive values. (#6115)
+* Document the updated CMake version requirement. (#6123)
+* Add demo for `DaskDeviceQuantileDMatrix`. (#6156)
+* Cosmetic fixes in `faq.rst` (#6161)
+* Fix error message. (#6176)
+* [Doc] Add list of winning solutions in data science competitions using XGBoost (#6177)
+* Fix a comment in demo to use correct reference (#6190)
+* Update the list of winning solutions using XGBoost (#6192)
+* Consistent style for build status badge (#6203)
+* [Doc] Add info on GPU compiler (#6204)
+* Update the list of winning solutions (#6222, #6254)
+* Add link to XGBoost's Twitter handle (#6244)
+* Fix minor typos in XGBClassifier methods' docstrings (#6247)
+* Add sponsors link to FUNDING.yml (#6252)
+* Group CLI demo into subdirectory. (#6258)
+* Reduce warning messages from `gbtree`. (#6273)
+* Create a tutorial for using the C API in a C/C++ application (#6285)
+* Update plugin instructions for CMake build (#6289)
+* [doc] make Dask distributed example copy-pastable (#6345)
+* [Python] Add option to use `libxgboost.so` from the system path (#6362)
+* Fixed few grammatical mistakes in doc (#6393)
+* Fix broken link in CLI doc (#6396)
+* Improve documentation for the Dask API (#6413)
+* Revise misleading exception information: no such param of `allow_non_zero_missing` (#6418)
+* Fix CLI ranking demo. (#6439)
+* Fix broken links. (#6455)
+
+### Acknowledgement
+**Contributors**: Nan Zhu (@CodingCat), @FelixYBW, Jack Dunn (@JackDunnNZ), Jean Lescut-Muller (@JeanLescut),  Boris Feld (@Lothiraldan), Nikhil Choudhary (@Nikhil1O1), Rory Mitchell (@RAMitchell), @ShvetsKS, Anthony D'Amato (@Totoketchup), @Wittty-Panda, neko (@akiyamaneko), Alexander Gugel (@alexanderGugel), @dependabot[bot], DIVYA CHAUHAN (@divya661), Daniel Steinberg (@dstein64), Akira Funahashi (@funasoul), Philip Hyunsu Cho (@hcho3), Tong He (@hetong007), Hristo Iliev (@hiliev), Honza Sterba (@honzasterba), @hzy001, Igor Moura (@igormp), @jameskrach, James Lamb (@jameslamb), Naveed Ahmed Saleem Janvekar (@janvekarnaveed), Kyle Nicholson (@kylejn27), lacrosse91 (@lacrosse91), Christian Lorentzen (@lorentzenchr), Manikya Bardhan (@manikyabard), @nabokovas, John Quitto-Graham (@nvidia-johnq), @odidev, Qi Zhang (@qzhang90), Sergio Gavilán (@sgavil), Tanuja Kirthi Doddapaneni (@tanuja3), Cuong Duong (@tcuongd), Yuan Tang (@terrytangyuan), Jiaming Yuan (@trivialfis), vcarpani (@vcarpani), Vladislav Epifanov (@vepifanov), Vitalie Spinu (@vspinu), Bobby Wang (@wbo4958), Zeno Gantner (@zenogantner), zhang_jf (@zuston)
+
+**Reviewers**: Nan Zhu (@CodingCat), John Zedlewski (@JohnZed), Rory Mitchell (@RAMitchell), @ShvetsKS, Egor Smirnov (@SmirnovEgorRu), Anthony D'Amato (@Totoketchup), @Wittty-Panda, Alexander Gugel (@alexanderGugel), Codecov Comments Bot (@codecov-commenter), Codecov (@codecov-io), DIVYA CHAUHAN (@divya661), Devin Robison (@drobison00), Geoffrey Blake (@geoffreyblake), Mark Harris (@harrism), Philip Hyunsu Cho (@hcho3), Honza Sterba (@honzasterba), Igor Moura (@igormp), @jakirkham, @jameskrach, James Lamb (@jameslamb), Janakarajan Natarajan (@janaknat), Jake Hemstad (@jrhemstad), Keith Kraus (@kkraus14), Kyle Nicholson (@kylejn27), Christian Lorentzen (@lorentzenchr), Michael Mayer (@mayer79), Nikolay Petrov (@napetrov), @odidev, PSEUDOTENSOR / Jonathan McKinney (@pseudotensor), Qi Zhang (@qzhang90), Sergio Gavilán (@sgavil), Scott Lundberg (@slundberg), Cuong Duong (@tcuongd), Yuan Tang (@terrytangyuan), Jiaming Yuan (@trivialfis), vcarpani (@vcarpani), Vladislav Epifanov (@vepifanov), Vincent Nijs (@vnijs), Vitalie Spinu (@vspinu), Bobby Wang (@wbo4958), William Hicks (@wphicks)
+
+## v1.2.0 (2020.08.22)
+
+### XGBoost4J-Spark now supports the GPU algorithm (#5171)
+* Now XGBoost4J-Spark is able to leverage NVIDIA GPU hardware to speed up training.
+* There is on-going work for accelerating the rest of the data pipeline with NVIDIA GPUs (#5950, #5972).
+
+### XGBoost now supports CUDA 11 (#5808)
+* It is now possible to build XGBoost with CUDA 11. Note that we do not yet distribute pre-built binaries built with CUDA 11; all current distributions use CUDA 10.0.
+
+### Better guidance for persisting XGBoost models in an R environment (#5940, #5964)
+* Users are strongly encouraged to use `xgb.save()` and `xgb.save.raw()` instead of `saveRDS()`. This is so that the persisted models can be accessed with future releases of XGBoost.
+* The previous release (1.1.0) had problems loading models that were saved with `saveRDS()`. This release adds a compatibility layer to restore access to the old RDS files. Note that this is meant to be a temporary measure; users are advised to stop using `saveRDS()` and migrate to `xgb.save()` and `xgb.save.raw()`.
+
+### New objectives and metrics
+* The pseudo-Huber loss `reg:pseudohubererror` is added (#5647). The corresponding metric is `mphe`. Right now, the slope is hard-coded to 1.
+* The Accelerated Failure Time objective for survival analysis (`survival:aft`) is now accelerated on GPUs (#5714, #5716). The survival metrics `aft-nloglik` and `interval-regression-accuracy` are also accelerated on GPUs.
+
+### Improved integration with scikit-learn
+* Added `n_features_in_` attribute to the scikit-learn interface to store the number of features used (#5780). This is useful for integrating with some scikit-learn features such as `StackingClassifier`.  See [this link](https://scikit-learn-enhancement-proposals.readthedocs.io/en/latest/slep010/proposal.html) for more details.
+* `XGBoostError` now inherits `ValueError`, which conforms scikit-learn's exception requirement (#5696).
+
+### Improved integration with Dask
+* The XGBoost Dask API now exposes an asynchronous interface (#5862). See [the document](https://xgboost.readthedocs.io/en/latest/tutorials/dask.html#working-with-asyncio) for details.
+* Zero-copy ingestion of GPU arrays via `DaskDeviceQuantileDMatrix` (#5623, #5799, #5800, #5803, #5837, #5874, #5901): Previously, the Dask interface had to make 2 data copies: one for concatenating the Dask partition/block into a single block and another for internal representation. To save memory, we introduce `DaskDeviceQuantileDMatrix`. As long as Dask partitions are resident in the GPU memory, `DaskDeviceQuantileDMatrix` is able to ingest them directly without making copies. This matrix type wraps `DeviceQuantileDMatrix`.
+* The prediction function now returns GPU Series type if the input is from Dask-cuDF (#5710). This is to preserve the input data type.
+
+### Robust handling of external data types (#5689, #5893)
+- As we support more and more external data types, the handling logic has proliferated all over the code base and became hard to keep track. It also became unclear how missing values and threads are handled. We refactored the Python package code to collect all data handling logic to a central location, and now we have an explicit list of of all supported data types.
+
+### Improvements in GPU-side data matrix (`DeviceQuantileDMatrix`)
+* The GPU-side data matrix now implements its own quantile sketching logic, so that data don't have to be transported back to the main memory (#5700, #5747, #5760, #5846, #5870, #5898). The GK sketching algorithm is also now better documented.
+  - Now we can load extremely sparse dataset like URL, although performance is still sub-optimal.
+* The GPU-side data matrix now exposes an iterative interface (#5783), so that users are able to construct a matrix from a data iterator. See the [Python demo](https://github.com/dmlc/xgboost/blob/release_1.2.0/demo/guide-python/data_iterator.py).
+
+### New language binding: Swift (#5728)
+* Visit https://github.com/kongzii/SwiftXGBoost for more details.
+
+### Robust model serialization with JSON (#5772, #5804, #5831, #5857, #5934)
+* We continue efforts from the 1.0.0 release to adopt JSON as the format to save and load models robustly.
+* JSON model IO is significantly faster and produces smaller model files.
+* Round-trip reproducibility is guaranteed, via the introduction of an efficient float-to-string conversion algorithm known as [the Ryū algorithm](https://dl.acm.org/doi/10.1145/3192366.3192369). The conversion is locale-independent, producing consistent numeric representation regardless of the locale setting of the user's machine.
+* We fixed an issue in loading large JSON files to memory.
+* It is now possible to load a JSON file from a remote source such as S3.
+
+### Performance improvements
+* CPU hist tree method optimization
+  - Skip missing lookup in hist row partitioning if data is dense. (#5644)
+  - Specialize training procedures for CPU hist tree method on distributed environment. (#5557)
+  - Add single point histogram for CPU hist.  Previously gradient histogram for CPU hist is hard coded to be 64 bit, now users can specify the parameter `single_precision_histogram` to use 32 bit histogram instead for faster training performance. (#5624, #5811)
+* GPU hist tree method optimization
+  - Removed some unnecessary synchronizations and better memory allocation pattern. (#5707)
+  - Optimize GPU Hist for wide dataset.  Previously for wide dataset the atomic operation is performed on global memory, now it can run on shared memory for faster histogram building. But there's a known small regression on GeForce cards with dense data. (#5795, #5926, #5948, #5631)
+
+### API additions
+* Support passing fmap to importance plot (#5719). Now importance plot can show actual names of features instead of default ones.
+* Support 64bit seed. (#5643)
+* A new C API `XGBoosterGetNumFeature` is added for getting number of features in booster (#5856).
+* Feature names and feature types are now stored in C++ core and saved in binary DMatrix (#5858).
+
+### Breaking: The `predict()` method of `DaskXGBClassifier` now produces class predictions (#5986). Use `predict_proba()` to obtain probability predictions.
+* Previously, `DaskXGBClassifier.predict()` produced probability predictions. This is inconsistent with the behavior of other scikit-learn classifiers, where `predict()` returns class predictions. We make a breaking change in 1.2.0 release so that `DaskXGBClassifier.predict()` now correctly produces class predictions and thus behave like other scikit-learn classifiers. Furthermore, we introduce the `predict_proba()` method for obtaining probability predictions, again to be in line with other scikit-learn classifiers.
+
+### Breaking: Custom evaluation metric now receives raw prediction (#5954)
+* Previously, the custom evaluation metric received a transformed prediction result when used with a classifier. Now the custom metric will receive a raw (untransformed) prediction and will need to transform the prediction itself.  See [demo/guide-python/custom\_softmax.py](https://github.com/dmlc/xgboost/blob/release_1.2.0/demo/guide-python/custom_softmax.py) for an example.
+* This change is to make the custom metric behave consistently with the custom objective, which already receives raw prediction (#5564).
+
+### Breaking: XGBoost4J-Spark now requires Spark 3.0 and Scala 2.12 (#5836, #5890)
+* Starting with version 3.0, Spark can manage GPU resources and allocate them among executors.
+* Spark 3.0 dropped support for Scala 2.11 and now only supports Scala 2.12. Thus, XGBoost4J-Spark also only supports Scala 2.12.
+
+### Breaking: XGBoost Python package now requires Python 3.6 and later (#5715)
+* Python 3.6 has many useful features such as f-strings.
+
+### Breaking: XGBoost now adopts the C++14 standard (#5664)
+* Make sure to use a sufficiently modern C++ compiler that supports C++14, such as Visual Studio 2017, GCC 5.0+, and Clang 3.4+.
+
+### Bug-fixes
+* Fix a data race in the prediction function (#5853). As a byproduct, the prediction function now uses a thread-local data store and became thread-safe.
+* Restore capability to run prediction when the test input has fewer features than the training data (#5955). This capability is necessary to support predicting with LIBSVM inputs. The previous release (1.1) had broken this capability, so we restore it in this version with better tests.
+* Fix OpenMP build with CMake for R package, to support CMake 3.13 (#5895).
+* Fix Windows 2016 build (#5902, #5918).
+* Fix edge cases in scikit-learn interface with Pandas input by disabling feature validation. (#5953)
+* [R] Enable weighted learning to rank (#5945)
+* [R] Fix early stopping with custom objective (#5923)
+* Fix NDK Build (#5886)
+* Add missing explicit template specializations for greater portability (#5921)
+* Handle empty rows in data iterators correctly (#5929). This bug affects file loader and JVM data frames.
+* Fix `IsDense` (#5702)
+* [jvm-packages] Fix wrong method name `setAllowZeroForMissingValue` (#5740)
+* Fix shape inference for Dask predict (#5989)
+
+### Usability Improvements, Documentation
+* [Doc] Document that CUDA 10.0 is required (#5872)
+* Refactored command line interface (CLI). Now CLI is able to handle user errors and output basic document. (#5574)
+* Better error handling in Python: use `raise from` syntax to preserve full stacktrace (#5787).
+* The JSON model dump now has a formal schema (#5660, #5818). The benefit is to prevent `dump_model()` function from breaking. See [this document](https://xgboost.readthedocs.io/en/latest/tutorials/saving_model.html#difference-between-saving-model-and-dumping-model) to understand the difference between saving and dumping models.
+* Add a reference to the GPU external memory paper (#5684)
+* Document more objective parameters in the R package (#5682)
+* Document the existence of pre-built binary wheels for MacOS (#5711)
+* Remove `max.depth` in the R gblinear example. (#5753)
+* Added conda environment file for building docs (#5773)
+* Mention dask blog post in the doc, which introduces using Dask with GPU and some internal workings. (#5789)
+* Fix rendering of Markdown docs (#5821)
+* Document new objectives and metrics available on GPUs (#5909)
+* Better message when no GPU is found. (#5594)
+* Remove the use of `silent` parameter from R demos. (#5675)
+* Don't use masked array in array interface. (#5730)
+* Update affiliation of @terrytangyuan: Ant Financial -> Ant Group (#5827)
+* Move dask tutorial closer other distributed tutorials (#5613)
+* Update XGBoost + Dask overview documentation (#5961)
+* Show `n_estimators` in the docstring of the scikit-learn interface (#6041)
+* Fix a type in a doctring of the scikit-learn interface (#5980)
+
+### Maintenance: testing, continuous integration, build system
+* [CI] Remove CUDA 9.0 from CI (#5674, #5745)
+* Require CUDA 10.0+ in CMake build (#5718)
+* [R] Remove dependency on gendef for Visual Studio builds (fixes #5608) (#5764). This enables building XGBoost with GPU support with R 4.x.
+* [R-package] Reduce duplication in configure.ac (#5693)
+* Bump com.esotericsoftware to 4.0.2 (#5690)
+* Migrate some tests from AppVeyor to GitHub Actions to speed up the tests. (#5911, #5917, #5919, #5922, #5928)
+* Reduce cost of the Jenkins CI server (#5884, #5904, #5892). We now enforce a daily budget via an automated monitor. We also dramatically reduced the workload for the Windows platform, since the cloud VM cost is vastly greater for Windows.
+* [R] Set up automated R linter (#5944)
+* [R] replace uses of T and F with TRUE and FALSE (#5778)
+* Update Docker container 'CPU' (#5956)
+* Simplify CMake build with modern CMake techniques (#5871)
+* Use `hypothesis` package for testing (#5759, #5835, #5849).
+* Define `_CRT_SECURE_NO_WARNINGS` to remove unneeded warnings in MSVC (#5434)
+* Run all Python demos in CI, to ensure that they don't break (#5651)
+* Enhance nvtx support (#5636). Now we can use unified timer between CPU and GPU. Also CMake is able to find nvtx automatically.
+* Speed up python test. (#5752)
+* Add helper for generating batches of data. (#5756)
+* Add c-api-demo to .gitignore (#5855)
+* Add option to enable all compiler warnings in GCC/Clang (#5897)
+* Make Python model compatibility test runnable locally (#5941)
+* Add cupy to Windows CI (#5797)
+* [CI] Fix cuDF install; merge 'gpu' and 'cudf' test suite (#5814)
+* Update rabit submodule (#5680, #5876)
+* Force colored output for Ninja build. (#5959)
+* [CI] Assign larger /dev/shm to NCCL (#5966)
+* Add missing Pytest marks to AsyncIO unit test (#5968)
+* [CI] Use latest cuDF and dask-cudf (#6048)
+* Add CMake flag to log C API invocations, to aid debugging (#5925)
+* Fix a unit test on CLI, to handle RC versions (#6050)
+* [CI] Use mgpu machine to run gpu hist unit tests (#6050)
+* [CI] Build GPU-enabled JAR artifact and deploy to xgboost-maven-repo (#6050)
+
+### Maintenance: Refactor code for legibility and maintainability
+* Remove dead code in DMatrix initialization. (#5635)
+* Catch dmlc error by ref. (#5678)
+* Refactor the `gpu_hist` split evaluation in preparation for batched nodes enumeration. (#5610)
+* Remove column major specialization. (#5755)
+* Remove unused imports in Python (#5776)
+* Avoid including `c_api.h` in header files. (#5782)
+* Remove unweighted GK quantile, which is unused. (#5816)
+* Add Python binding for rabit ops. (#5743)
+* Implement `Empty` method for host device vector. (#5781)
+* Remove print (#5867)
+* Enforce tree order in JSON (#5974)
+
+### Acknowledgement
+**Contributors**: Nan Zhu (@CodingCat), @LionOrCatThatIsTheQuestion, Dmitry Mottl (@Mottl), Rory Mitchell (@RAMitchell), @ShvetsKS, Alex Wozniakowski (@a-wozniakowski), Alexander Gugel (@alexanderGugel), @anttisaukko, @boxdot, Andy Adinets (@canonizer), Ram Rachum (@cool-RR), Elliot Hershberg (@elliothershberg), Jason E. Aten, Ph.D. (@glycerine), Philip Hyunsu Cho (@hcho3), @jameskrach, James Lamb (@jameslamb), James Bourbeau (@jrbourbeau), Peter Jung (@kongzii), Lorenz Walthert (@lorenzwalthert), Oleksandr Kuvshynov (@okuvshynov), Rong Ou (@rongou), Shaochen Shi (@shishaochen), Yuan Tang (@terrytangyuan), Jiaming Yuan (@trivialfis), Bobby Wang (@wbo4958), Zhang Zhang (@zhangzhang10)
+
+**Reviewers**: Nan Zhu (@CodingCat), @LionOrCatThatIsTheQuestion, Hao Yang (@QuantHao), Rory Mitchell (@RAMitchell), @ShvetsKS, Egor Smirnov (@SmirnovEgorRu), Alex Wozniakowski (@a-wozniakowski), Amit Kumar (@aktech), Avinash Barnwal (@avinashbarnwal), @boxdot, Andy Adinets (@canonizer), Chandra Shekhar Reddy (@chandrureddy), Ram Rachum (@cool-RR), Cristiano Goncalves (@cristianogoncalves), Elliot Hershberg (@elliothershberg), Jason E. Aten, Ph.D. (@glycerine), Philip Hyunsu Cho (@hcho3), Tong He (@hetong007), James Lamb (@jameslamb), James Bourbeau (@jrbourbeau), Lee Drake (@leedrake5), DougM (@mengdong), Oleksandr Kuvshynov (@okuvshynov), RongOu (@rongou), Shaochen Shi (@shishaochen), Xu Xiao (@sperlingxx), Yuan Tang (@terrytangyuan), Theodore Vasiloudis (@thvasilo), Jiaming Yuan (@trivialfis), Bobby Wang (@wbo4958), Zhang Zhang (@zhangzhang10)
+
+## v1.1.1 (2020.06.06)
+This patch release applies the following patches to 1.1.0 release:
+
+* CPU performance improvement in the PyPI wheels (#5720)
+* Fix loading old model (#5724)
+* Install pkg-config file (#5744)
+
+## v1.1.0 (2020.05.17)
+
+### Better performance on multi-core CPUs (#5244, #5334, #5522)
+* Poor performance scaling of the `hist` algorithm for multi-core CPUs has been under investigation (#3810). #5244 concludes the ongoing effort to improve performance scaling on multi-CPUs, in particular Intel CPUs. Roadmap: #5104
+* #5334 makes steps toward reducing memory consumption for the `hist` tree method on CPU.
+* #5522 optimizes random number generation for data sampling.
+
+### Deterministic GPU algorithm for regression and classification (#5361)
+* GPU algorithm for regression and classification tasks is now deterministic.
+* Roadmap: #5023. Currently only single-GPU training is deterministic. Distributed training with multiple GPUs is not yet deterministic.
+
+### Improve external memory support on GPUs (#5093, #5365)
+* Starting from 1.0.0 release, we added support for external memory on GPUs to enable training with larger datasets. Gradient-based sampling (#5093) speeds up the external memory algorithm by intelligently sampling a subset of the training data to copy into the GPU memory. [Learn more about out-of-core GPU gradient boosting.](https://arxiv.org/abs/2005.09148)
+* GPU-side data sketching now works with data from external memory (#5365).
+
+### Parameter validation: detection of unused or incorrect parameters (#5477, #5569, #5508)
+* Mis-spelled training parameter is a common user mistake. In previous versions of XGBoost, mis-spelled parameters were silently ignored. Starting with 1.0.0 release, XGBoost will produce a warning message if there is any unused training parameters. The 1.1.0 release makes parameter validation available to the scikit-learn interface (#5477) and the R binding (#5569).
+
+### Thread-safe, in-place prediction method (#5389, #5512)
+* Previously, the prediction method was not thread-safe (#5339). This release adds a new API function `inplace_predict()` that is thread-safe. It is now possible to serve concurrent requests for prediction using a shared model object.
+* It is now possible to compute prediction in-place for selected data formats (`numpy.ndarray` / `scipy.sparse.csr_matrix` / `cupy.ndarray` / `cudf.DataFrame` / `pd.DataFrame`) without creating a `DMatrix` object.
+
+### Addition of Accelerated Failure Time objective for survival analysis (#4763, #5473, #5486, #5552, #5553)
+* Survival analysis (regression) models the time it takes for an event of interest to occur. The target label is potentially censored, i.e. the label is a range rather than a single number. We added a new objective `survival:aft` to support survival analysis. Also added is the new API to specify the ranged labels. Check out [the tutorial](https://xgboost.readthedocs.io/en/release_1.1.0/tutorials/aft_survival_analysis.html) and the [demos](https://github.com/dmlc/xgboost/tree/release_1.1.0/demo/aft_survival).
+* GPU support is work in progress (#5714).
+
+### Improved installation experience on Mac OSX (#5597, #5602, #5606, #5701)
+* It only takes two commands to install the XGBoost Python package: `brew install libomp` followed by `pip install xgboost`. The installed XGBoost will use all CPU cores. Even better, starting with this release, we distribute pre-compiled binary wheels targeting Mac OSX. Now the install command `pip install xgboost` finishes instantly, as it no longer compiles the C++ source of XGBoost. The last three Mac versions (High Sierra, Mojave, Catalina) are supported.
+* R package: the 1.1.0 release fixes the error `Initializing libomp.dylib, but found libomp.dylib already initialized` (#5701)
+
+### Ranking metrics are now accelerated on GPUs (#5380, #5387, #5398)
+
+### GPU-side data matrix to ingest data directly from other GPU libraries (#5420, #5465)
+* Previously, data on GPU memory had to be copied back to the main memory before it could be used by XGBoost. Starting with 1.1.0 release, XGBoost provides a dedicated interface (`DeviceQuantileDMatrix`) so that it can ingest data from GPU memory directly. The result is that XGBoost interoperates better with GPU-accelerated data science libraries, such as cuDF, cuPy, and PyTorch.
+* Set device in device dmatrix. (#5596)
+
+### Robust model serialization with JSON (#5123, #5217)
+* We continue efforts from the 1.0.0 release to adopt JSON as the format to save and load models robustly. Refer to the release note for 1.0.0 to learn more.
+* It is now possible to store internal configuration of the trained model (`Booster`) object in R as a JSON string (#5123, #5217).
+
+### Improved integration with Dask
+* Pass through `verbose` parameter for dask fit (#5413)
+* Use `DMLC_TASK_ID`. (#5415)
+* Order the prediction result. (#5416)
+* Honor `nthreads` from dask worker. (#5414)
+* Enable grid searching with scikit-learn. (#5417)
+* Check non-equal when setting threads. (#5421)
+* Accept other inputs for prediction. (#5428)
+* Fix missing value for scikit-learn interface. (#5435)
+
+### XGBoost4J-Spark: Check number of columns in the data iterator (#5202, #5303)
+* Before, the native layer in XGBoost did not know the number of columns (features) ahead of time and had to guess the number of columns by counting the feature index when ingesting data. This method has a failure more in distributed setting: if the training data is highly sparse, some features may be completely missing in one or more worker partitions. Thus, one or more workers may deduce an incorrect data shape, leading to crashes or silently wrong models.
+* Enforce correct data shape by passing the number of columns explicitly from the JVM layer into the native layer.
+
+### Major refactoring of the `DMatrix` class
+* Continued from 1.0.0 release.
+* Remove update prediction cache from predictors. (#5312)
+* Predict on Ellpack. (#5327)
+* Partial rewrite EllpackPage (#5352)
+* Use ellpack for prediction only when sparsepage doesn't exist. (#5504)
+* RFC: #4354, Roadmap: #5143
+
+### Breaking: XGBoost Python package now requires Pip 19.0 and higher (#5589)
+* Your Linux machine may have an old version of Pip and may attempt to install a source package, leading to long installation time. This is because we are now using `manylinux2010` tag in the binary wheel release. Ensure you have Pip 19.0 or newer by running `python3 -m pip -V` to check the version. Upgrade Pip with command
+```
+python3 -m pip install --upgrade pip
+```
+Upgrading to latest pip allows us to depend on newer versions of system libraries. [TensorFlow](https://www.tensorflow.org/install/pip) also requires Pip 19.0+.
+
+### Breaking: GPU algorithm now requires CUDA 10.0 and higher (#5649)
+* CUDA 10.0 is necessary to make the GPU algorithm deterministic (#5361).
+
+### Breaking: `silent` parameter is now removed (#5476)
+* Please use `verbosity` instead.
+
+### Breaking: Set `output_margin` to True for custom objectives (#5564)
+* Now both R and Python interface custom objectives get un-transformed (raw) prediction outputs.
+
+### Breaking: `Makefile` is now removed. We use CMake exclusively to build XGBoost (#5513)
+* Exception: the R package uses Autotools, as the CRAN ecosystem did not yet adopt CMake widely.
+
+### Breaking: `distcol` updater is now removed (#5507)
+* The `distcol` updater has been long broken, and currently we lack resources to implement a working implementation from scratch.
+
+### Deprecation notices
+* **Python 3.5**. This release is the last release to support Python 3.5. The following release (1.2.0) will require Python 3.6.
+* **Scala 2.11**. Currently XGBoost4J supports Scala 2.11. However, if a future release of XGBoost adopts Spark 3, it will not support Scala 2.11, as Spark 3 requires Scala 2.12+. We do not yet know which XGBoost release will adopt Spark 3.
+
+### Known limitations
+* (Python package) When early stopping is activated with `early_stopping_rounds` at training time, the prediction method (`xgb.predict()`) behaves in a surprising way. If XGBoost runs for M rounds and chooses iteration N (N < M) as the best iteration, then the prediction method will use M trees by default. To use the best iteration (N trees), users will need to manually take the best iteration field `bst.best_iteration` and pass it as the `ntree_limit` argument to `xgb.predict()`. See #5209 and #4052 for additional context.
+* GPU ranking objective is currently not deterministic (#5561).
+* When training parameter `reg_lambda` is set to zero, some leaf nodes may be assigned a NaN value. (See [discussion](https://discuss.xgboost.ai/t/still-getting-unexplained-nans-new-replication-code/1383/9).) For now, please set `reg_lambda` to a nonzero value.
+
+### Community and Governance
+* The XGBoost Project Management Committee (PMC) is pleased to announce a new committer: Egor Smirnov (@SmirnovEgorRu). He has led a major initiative to improve the performance of XGBoost on multi-core CPUs.
+
+### Bug-fixes
+* Improved compatibility with scikit-learn (#5255, #5505, #5538)
+* Remove f-string, since it's not supported by Python 3.5 (#5330). Note that Python 3.5 support is deprecated and schedule to be dropped in the upcoming release (1.2.0).
+* Fix the pruner so that it doesn't prune the same branch twice (#5335)
+* Enforce only major version in JSON model schema (#5336). Any major revision of the model schema would bump up the major version.
+* Fix a small typo in sklearn.py that broke multiple eval metrics (#5341)
+* Restore loading model from a memory buffer (#5360)
+* Define lazy isinstance for Python compat (#5364)
+* [R] fixed uses of `class()` (#5426)
+* Force compressed buffer to be 4 bytes aligned, to keep cuda-memcheck happy (#5441)
+* Remove warning for calling host function (`std::max`) on a GPU device (#5453)
+* Fix uninitialized value bug in xgboost callback (#5463)
+* Fix model dump in CLI (#5485)
+* Fix out-of-bound array access in `WQSummary::SetPrune()` (#5493)
+* Ensure that configured `dmlc/build_config.h` is picked up by Rabit and XGBoost, to fix build on Alpine (#5514)
+* Fix a misspelled method, made in a git merge (#5509)
+* Fix a bug in binary model serialization (#5532)
+* Fix CLI model IO (#5535)
+* Don't use `uint` for threads (#5542)
+* Fix R interaction constraints to handle more than 100000 features (#5543)
+* [jvm-packages] XGBoost Spark should deal with NaN when parsing evaluation output (#5546)
+* GPU-side data sketching is now aware of query groups in learning-to-rank data (#5551)
+* Fix DMatrix slicing for newly added fields (#5552)
+* Fix configuration status with loading binary model (#5562)
+* Fix build when OpenMP is disabled (#5566)
+* R compatibility patches (#5577, #5600)
+* gpu\_hist performance fixes (#5558)
+* Don't set seed on CLI interface (#5563)
+* [R] When serializing model, preserve model attributes related to early stopping (#5573)
+* Avoid rabit calls in learner configuration (#5581)
+* Hide C++ symbols in libxgboost.so when building Python wheel (#5590). This fixes apache/incubator-tvm#4953.
+* Fix compilation on Mac OSX High Sierra (10.13) (#5597)
+* Fix build on big endian CPUs (#5617)
+* Resolve crash due to use of `vector<bool>::iterator` (#5642)
+* Validation JSON model dump using JSON schema (#5660)
+
+### Performance improvements
+* Wide dataset quantile performance improvement (#5306)
+* Reduce memory usage of GPU-side data sketching (#5407)
+* Reduce span check overhead (#5464)
+* Serialise booster after training to free up GPU memory (#5484)
+* Use the maximum amount of GPU shared memory available to speed up the histogram kernel (#5491)
+* Use non-synchronising scan in Thrust (#5560)
+* Use `cudaDeviceGetAttribute()` instead of `cudaGetDeviceProperties()` for speed (#5570)
+
+### API changes
+* Support importing data from a Pandas SparseArray (#5431)
+* `HostDeviceVector` (vector shared between CPU and GPU memory) now exposes `HostSpan` interface, to enable access on the CPU side with bound check (#5459)
+* Accept other gradient types for `SplitEntry` (#5467)
+
+### Usability Improvements, Documentation
+* Add `JVM_CHECK_CALL` to prevent C++ exceptions from leaking into the JVM layer (#5199)
+* Updated Windows build docs (#5283)
+* Update affiliation of @hcho3 (#5292)
+* Display Sponsor button, link to OpenCollective (#5325)
+* Update docs for GPU external memory (#5332)
+* Add link to GPU documentation (#5437)
+* Small updates to GPU documentation (#5483)
+* Edits on tutorial for XGBoost job on Kubernetes (#5487)
+* Add reference to GPU external memory (#5490)
+* Fix typos (#5346, #5371, #5384, #5399, #5482, #5515)
+* Update Python doc (#5517)
+* Add Neptune and Optuna to list of examples (#5528)
+* Raise error if the number of data weights doesn't match the number of data sets (#5540)
+* Add a note about GPU ranking (#5572)
+* Clarify meaning of `training` parameter in the C API function `XGBoosterPredict()` (#5604)
+* Better error handling for situations where existing trees cannot be modified (#5406, #5418). This feature is enabled when `process_type` is set to `update`.
+
+### Maintenance: testing, continuous integration, build system
+* Add C++ test coverage for data sketching (#5251)
+* Ignore gdb\_history (#5257)
+* Rewrite setup.py. (#5271, #5280)
+* Use `scikit-learn` in extra dependencies (#5310)
+* Add CMake option to build static library (#5397)
+* [R] changed FindLibR to take advantage of CMake cache (#5427)
+* [R] fixed inconsistency in R -e calls in FindLibR.cmake (#5438)
+* Refactor tests with data generator (#5439)
+* Resolve failing Travis CI (#5445)
+* Update dmlc-core. (#5466)
+* [CI] Use clang-tidy 10 (#5469)
+* De-duplicate code for checking maximum number of nodes (#5497)
+* [CI] Use Ubuntu 18.04 LTS in JVM CI, because 19.04 is EOL (#5537)
+* [jvm-packages] [CI] Create a Maven repository to host SNAPSHOT JARs (#5533)
+* [jvm-packages] [CI] Publish XGBoost4J JARs with Scala 2.11 and 2.12 (#5539)
+* [CI] Use Vault repository to re-gain access to devtoolset-4 (#5589)
+
+### Maintenance: Refactor code for legibility and maintainability
+* Move prediction cache to Learner (#5220, #5302)
+* Remove SimpleCSRSource (#5315)
+* Refactor SparsePageSource, delete cache files after use (#5321)
+* Remove unnecessary DMatrix methods (#5324)
+* Split up `LearnerImpl` (#5350)
+* Move segment sorter to common (#5378)
+* Move thread local entry into Learner (#5396)
+* Split up test helpers header (#5455)
+* Requires setting leaf stat when expanding tree (#5501)
+* Purge device\_helpers.cuh (#5534)
+* Use thrust functions instead of custom functions (#5544)
+
+### Acknowledgement
+**Contributors**: Nan Zhu (@CodingCat), Rory Mitchell (@RAMitchell), @ShvetsKS, Egor Smirnov (@SmirnovEgorRu), Andrew Kane (@ankane), Avinash Barnwal (@avinashbarnwal), Bart Broere (@bartbroere), Andy Adinets (@canonizer), Chen Qin (@chenqin), Daiki Katsuragawa (@daikikatsuragawa), David Díaz Vico (@daviddiazvico), Darius Kharazi (@dkharazi), Darby Payne (@dpayne), Jason E. Aten, Ph.D. (@glycerine), Philip Hyunsu Cho (@hcho3), James Lamb (@jameslamb), Jan Borchmann (@jborchma), Kamil A. Kaczmarek (@kamil-kaczmarek), Melissa Kohl (@mjkohl32), Nicolas Scozzaro (@nscozzaro), Paul Kaefer (@paulkaefer), Rong Ou (@rongou), Samrat Pandiri (@samratp), Sriram Chandramouli (@sriramch), Yuan Tang (@terrytangyuan), Jiaming Yuan (@trivialfis), Liang-Chi Hsieh (@viirya), Bobby Wang (@wbo4958), Zhang Zhang (@zhangzhang10),
+
+**Reviewers**: Nan Zhu (@CodingCat), @LeZhengThu, Rory Mitchell (@RAMitchell), @ShvetsKS, Egor Smirnov (@SmirnovEgorRu), Steve Bronder (@SteveBronder), Nikita Titov (@StrikerRUS), Andrew Kane (@ankane), Avinash Barnwal (@avinashbarnwal), @brydag, Andy Adinets (@canonizer), Chandra Shekhar Reddy (@chandrureddy), Chen Qin (@chenqin), Codecov (@codecov-io), David Díaz Vico (@daviddiazvico), Darby Payne (@dpayne), Jason E. Aten, Ph.D. (@glycerine), Philip Hyunsu Cho (@hcho3), James Lamb (@jameslamb), @johnny-cat, Mu Li (@mli), Mate Soos (@msoos), @rnyak, Rong Ou (@rongou), Sriram Chandramouli (@sriramch), Toby Dylan Hocking (@tdhock), Yuan Tang (@terrytangyuan), Oleksandr Pryimak (@trams), Jiaming Yuan (@trivialfis), Liang-Chi Hsieh (@viirya), Bobby Wang (@wbo4958),
+
+## v1.0.2 (2020.03.03)
+This patch release applies the following patches to 1.0.0 release:
+
+* Fix a small typo in sklearn.py that broke multiple eval metrics (#5341)
+* Restore loading model from buffer (#5360)
+* Use type name for data type check (#5364)
+
+## v1.0.1 (2020.02.21)
+This release is identical to the 1.0.0 release, except that it fixes a small bug that rendered 1.0.0 incompatible with Python 3.5. See #5328.
+
 ## v1.0.0 (2020.02.19)
 This release marks a major milestone for the XGBoost project.
 

R-package/CMakeLists.txt

@@ -6,8 +6,11 @@ file(GLOB_RECURSE R_SOURCES
   ${CMAKE_CURRENT_LIST_DIR}/src/*.c)
 # Use object library to expose symbols
 add_library(xgboost-r OBJECT ${R_SOURCES})
-
-set(R_DEFINITIONS
+if (ENABLE_ALL_WARNINGS)
+  target_compile_options(xgboost-r PRIVATE -Wall -Wextra)
+endif (ENABLE_ALL_WARNINGS)
+target_compile_definitions(xgboost-r
+  PUBLIC
   -DXGBOOST_STRICT_R_MODE=1
   -DXGBOOST_CUSTOMIZE_GLOBAL_PRNG=1
   -DDMLC_LOG_BEFORE_THROW=0
@@ -15,24 +18,27 @@ set(R_DEFINITIONS
   -DDMLC_LOG_CUSTOMIZE=1
   -DRABIT_CUSTOMIZE_MSG_
   -DRABIT_STRICT_CXX98_)
-target_compile_definitions(xgboost-r
-  PRIVATE ${R_DEFINITIONS})
 target_include_directories(xgboost-r
   PRIVATE
   ${LIBR_INCLUDE_DIRS}
   ${PROJECT_SOURCE_DIR}/include
   ${PROJECT_SOURCE_DIR}/dmlc-core/include
   ${PROJECT_SOURCE_DIR}/rabit/include)
+target_link_libraries(xgboost-r PUBLIC ${LIBR_CORE_LIBRARY})
+if (USE_OPENMP)
+  find_package(OpenMP REQUIRED)
+  target_link_libraries(xgboost-r PUBLIC OpenMP::OpenMP_CXX OpenMP::OpenMP_C)
+endif (USE_OPENMP)
 set_target_properties(
   xgboost-r PROPERTIES
-  CXX_STANDARD 11
+  CXX_STANDARD 14
   CXX_STANDARD_REQUIRED ON
   POSITION_INDEPENDENT_CODE ON)
 
-set(XGBOOST_DEFINITIONS "${XGBOOST_DEFINITIONS};${R_DEFINITIONS}" PARENT_SCOPE)
-set(XGBOOST_OBJ_SOURCES $<TARGET_OBJECTS:xgboost-r> PARENT_SCOPE)
-set(LINKED_LIBRARIES_PRIVATE ${LINKED_LIBRARIES_PRIVATE} ${LIBR_CORE_LIBRARY} PARENT_SCOPE)
+# Get compilation and link flags of xgboost-r and propagate to objxgboost
+target_link_libraries(objxgboost PUBLIC xgboost-r)
+# Add all objects of xgboost-r to objxgboost
+target_sources(objxgboost INTERFACE $<TARGET_OBJECTS:xgboost-r>)
 
-if (USE_OPENMP)
-  target_link_libraries(xgboost-r PRIVATE OpenMP::OpenMP_CXX)
-endif ()
+set(LIBR_HOME "${LIBR_HOME}" PARENT_SCOPE)
+set(LIBR_EXECUTABLE "${LIBR_EXECUTABLE}" PARENT_SCOPE)

R-package/DESCRIPTION

@@ -1,8 +1,8 @@
 Package: xgboost
 Type: Package
 Title: Extreme Gradient Boosting
-Version: 1.1.0.1
-Date: 2020-02-21
+Version: 1.4.1.1
+Date: 2020-08-28
 Authors@R: c(
   person("Tianqi", "Chen", role = c("aut"),
          email = "tianqi.tchen@gmail.com"),
@@ -31,9 +31,9 @@ Authors@R: c(
   )
 Description: Extreme Gradient Boosting, which is an efficient implementation
     of the gradient boosting framework from Chen & Guestrin (2016) <doi:10.1145/2939672.2939785>.
-    This package is its R interface. The package includes efficient linear 
-    model solver and tree learning algorithms. The package can automatically 
-    do parallel computation on a single machine which could be more than 10 
+    This package is its R interface. The package includes efficient linear
+    model solver and tree learning algorithms. The package can automatically
+    do parallel computation on a single machine which could be more than 10
     times faster than existing gradient boosting packages. It supports
     various objective functions, including regression, classification and ranking.
     The package is made to be extensible, so that users are also allowed to define
@@ -53,8 +53,9 @@ Suggests:
     testthat,
     lintr,
     igraph (>= 1.0.1),
-    jsonlite,
-    float
+    float,
+    crayon,
+    titanic
 Depends:
     R (>= 3.3.0)
 Imports:
@@ -62,6 +63,6 @@ Imports:
     methods,
     data.table (>= 1.9.6),
     magrittr (>= 1.5),
-    stringi (>= 0.5.2)
-RoxygenNote: 7.1.0
-SystemRequirements: GNU make, C++11
+    jsonlite (>= 1.0),
+RoxygenNote: 7.1.1
+SystemRequirements: GNU make, C++14

R-package/NAMESPACE

@@ -36,8 +36,10 @@ export(xgb.create.features)
 export(xgb.cv)
 export(xgb.dump)
 export(xgb.gblinear.history)
+export(xgb.get.config)
 export(xgb.ggplot.deepness)
 export(xgb.ggplot.importance)
+export(xgb.ggplot.shap.summary)
 export(xgb.importance)
 export(xgb.load)
 export(xgb.load.raw)
@@ -46,10 +48,12 @@ export(xgb.plot.deepness)
 export(xgb.plot.importance)
 export(xgb.plot.multi.trees)
 export(xgb.plot.shap)
+export(xgb.plot.shap.summary)
 export(xgb.plot.tree)
 export(xgb.save)
 export(xgb.save.raw)
 export(xgb.serialize)
+export(xgb.set.config)
 export(xgb.train)
 export(xgb.unserialize)
 export(xgboost)
@@ -76,14 +80,11 @@ importFrom(graphics,lines)
 importFrom(graphics,par)
 importFrom(graphics,points)
 importFrom(graphics,title)
+importFrom(jsonlite,fromJSON)
+importFrom(jsonlite,toJSON)
 importFrom(magrittr,"%>%")
 importFrom(stats,median)
 importFrom(stats,predict)
-importFrom(stringi,stri_detect_regex)
-importFrom(stringi,stri_match_first_regex)
-importFrom(stringi,stri_replace_all_regex)
-importFrom(stringi,stri_replace_first_regex)
-importFrom(stringi,stri_split_regex)
 importFrom(utils,head)
 importFrom(utils,object.size)
 importFrom(utils,str)

R-package/R/callbacks.R

@@ -62,11 +62,11 @@ cb.print.evaluation <- function(period = 1, showsd = TRUE) {
   callback <- function(env = parent.frame()) {
     if (length(env$bst_evaluation) == 0 ||
         period == 0 ||
-        NVL(env$rank, 0) != 0 )
+        NVL(env$rank, 0) != 0)
       return()
 
     i <- env$iteration
-    if ((i-1) %% period == 0 ||
+    if ((i - 1) %% period == 0 ||
         i == env$begin_iteration ||
         i == env$end_iteration) {
       stdev <- if (showsd) env$bst_evaluation_err else NULL
@@ -115,7 +115,7 @@ cb.evaluation.log <- function() {
       stop("bst_evaluation must have non-empty names")
 
     mnames <<- gsub('-', '_', names(env$bst_evaluation))
-    if(!is.null(env$bst_evaluation_err))
+    if (!is.null(env$bst_evaluation_err))
       mnames <<- c(paste0(mnames, '_mean'), paste0(mnames, '_std'))
   }
 
@@ -123,12 +123,12 @@ cb.evaluation.log <- function() {
     env$evaluation_log <- as.data.table(t(simplify2array(env$evaluation_log)))
     setnames(env$evaluation_log, c('iter', mnames))
 
-    if(!is.null(env$bst_evaluation_err)) {
+    if (!is.null(env$bst_evaluation_err)) {
       # rearrange col order from _mean,_mean,...,_std,_std,...
       # to be _mean,_std,_mean,_std,...
       len <- length(mnames)
-      means <- mnames[seq_len(len/2)]
-      stds <- mnames[(len/2 + 1):len]
+      means <- mnames[seq_len(len / 2)]
+      stds <- mnames[(len / 2 + 1):len]
       cnames <- numeric(len)
       cnames[c(TRUE, FALSE)] <- means
       cnames[c(FALSE, TRUE)] <- stds
@@ -144,7 +144,7 @@ cb.evaluation.log <- function() {
       return(finalizer(env))
 
     ev <- env$bst_evaluation
-    if(!is.null(env$bst_evaluation_err))
+    if (!is.null(env$bst_evaluation_err))
       ev <- c(ev, env$bst_evaluation_err)
     env$evaluation_log <- c(env$evaluation_log,
                             list(c(iter = env$iteration, ev)))
@@ -351,13 +351,19 @@ cb.early.stop <- function(stopping_rounds, maximize = FALSE,
 
   finalizer <- function(env) {
     if (!is.null(env$bst)) {
-      attr_best_score = as.numeric(xgb.attr(env$bst$handle, 'best_score'))
-      if (best_score != attr_best_score)
-        stop("Inconsistent 'best_score' values between the closure state: ", best_score,
-             " and the xgb.attr: ", attr_best_score)
-      env$bst$best_iteration = best_iteration
-      env$bst$best_ntreelimit = best_ntreelimit
-      env$bst$best_score = best_score
+      attr_best_score <- as.numeric(xgb.attr(env$bst$handle, 'best_score'))
+      if (best_score != attr_best_score) {
+        # If the difference is too big, throw an error
+        if (abs(best_score - attr_best_score) >= 1e-14) {
+          stop("Inconsistent 'best_score' values between the closure state: ", best_score,
+               " and the xgb.attr: ", attr_best_score)
+        }
+        # If the difference is due to floating-point truncation, update best_score
+        best_score <- attr_best_score
+      }
+      env$bst$best_iteration <- best_iteration
+      env$bst$best_ntreelimit <- best_ntreelimit
+      env$bst$best_score <- best_score
     } else {
       env$basket$best_iteration <- best_iteration
       env$basket$best_ntreelimit <- best_ntreelimit
@@ -372,9 +378,9 @@ cb.early.stop <- function(stopping_rounds, maximize = FALSE,
       return(finalizer(env))
 
     i <- env$iteration
-    score = env$bst_evaluation[metric_idx]
+    score <- env$bst_evaluation[metric_idx]
 
-    if (( maximize && score > best_score) ||
+    if ((maximize && score > best_score) ||
         (!maximize && score < best_score)) {
 
       best_msg <<- format.eval.string(i, env$bst_evaluation, env$bst_evaluation_err)
@@ -500,7 +506,7 @@ cb.cv.predict <- function(save_models = FALSE) {
     for (fd in env$bst_folds) {
       pr <- predict(fd$bst, fd$watchlist[[2]], ntreelimit = ntreelimit, reshape = TRUE)
       if (is.matrix(pred)) {
-        pred[fd$index,] <- pr
+        pred[fd$index, ] <- pr
       } else {
         pred[fd$index] <- pr
       }
@@ -613,9 +619,7 @@ cb.gblinear.history <- function(sparse=FALSE) {
 
   init <- function(env) {
     if (!is.null(env$bst)) { # xgb.train:
-      coef_path <- list()
     } else if (!is.null(env$bst_folds)) { # xgb.cv:
-      coef_path <- rep(list(), length(env$bst_folds))
     } else stop("Parent frame has neither 'bst' nor 'bst_folds'")
   }
 
@@ -705,11 +709,11 @@ xgb.gblinear.history <- function(model, class_index = NULL) {
   if (!is_cv) {
     # extract num_class & num_feat from the internal model
     dmp <- xgb.dump(model)
-    if(length(dmp) < 2 || dmp[2] != "bias:")
+    if (length(dmp) < 2 || dmp[2] != "bias:")
       stop("It does not appear to be a gblinear model")
-    dmp <- dmp[-c(1,2)]
+    dmp <- dmp[-c(1, 2)]
     n <- which(dmp == 'weight:')
-    if(length(n) != 1)
+    if (length(n) != 1)
       stop("It does not appear to be a gblinear model")
     num_class <- n - 1
     num_feat <- (length(dmp) - 4) / num_class
@@ -732,9 +736,9 @@ xgb.gblinear.history <- function(model, class_index = NULL) {
   if (!is.null(class_index) && num_class > 1) {
     coef_path <- if (is.list(coef_path)) {
       lapply(coef_path,
-             function(x) x[, seq(1 + class_index, by=num_class, length.out=num_feat)])
+             function(x) x[, seq(1 + class_index, by = num_class, length.out = num_feat)])
     } else {
-      coef_path <- coef_path[, seq(1 + class_index, by=num_class, length.out=num_feat)]
+      coef_path <- coef_path[, seq(1 + class_index, by = num_class, length.out = num_feat)]
     }
   }
   coef_path

R-package/R/utils.R

@@ -20,6 +20,12 @@ NVL <- function(x, val) {
   stop("typeof(x) == ", typeof(x), " is not supported by NVL")
 }
 
+# List of classification and ranking objectives
+.CLASSIFICATION_OBJECTIVES <- function() {
+  return(c('binary:logistic', 'binary:logitraw', 'binary:hinge', 'multi:softmax',
+           'multi:softprob', 'rank:pairwise', 'rank:ndcg', 'rank:map'))
+}
+
 
 #
 # Low-level functions for boosting --------------------------------------------
@@ -69,23 +75,23 @@ check.booster.params <- function(params, ...) {
 
   if (!is.null(params[['monotone_constraints']]) &&
       typeof(params[['monotone_constraints']]) != "character") {
-        vec2str = paste(params[['monotone_constraints']], collapse = ',')
-        vec2str = paste0('(', vec2str, ')')
-        params[['monotone_constraints']] = vec2str
+        vec2str <- paste(params[['monotone_constraints']], collapse = ',')
+        vec2str <- paste0('(', vec2str, ')')
+        params[['monotone_constraints']] <- vec2str
   }
 
   # interaction constraints parser (convert from list of column indices to string)
   if (!is.null(params[['interaction_constraints']]) &&
       typeof(params[['interaction_constraints']]) != "character"){
     # check input class
-    if (!identical(class(params[['interaction_constraints']]),'list')) stop('interaction_constraints should be class list')
-    if (!all(unique(sapply(params[['interaction_constraints']], class)) %in% c('numeric','integer'))) {
+    if (!identical(class(params[['interaction_constraints']]), 'list')) stop('interaction_constraints should be class list')
+    if (!all(unique(sapply(params[['interaction_constraints']], class)) %in% c('numeric', 'integer'))) {
       stop('interaction_constraints should be a list of numeric/integer vectors')
     }
 
     # recast parameter as string
-    interaction_constraints <- sapply(params[['interaction_constraints']], function(x) paste0('[', paste(x, collapse=','), ']'))
-    params[['interaction_constraints']] <- paste0('[', paste(interaction_constraints, collapse=','), ']')
+    interaction_constraints <- sapply(params[['interaction_constraints']], function(x) paste0('[', paste(x, collapse = ','), ']'))
+    params[['interaction_constraints']] <- paste0('[', paste(interaction_constraints, collapse = ','), ']')
   }
   return(params)
 }
@@ -145,7 +151,8 @@ xgb.iter.update <- function(booster_handle, dtrain, iter, obj = NULL) {
   if (is.null(obj)) {
     .Call(XGBoosterUpdateOneIter_R, booster_handle, as.integer(iter), dtrain)
   } else {
-    pred <- predict(booster_handle, dtrain, outputmargin = TRUE, training = TRUE)
+    pred <- predict(booster_handle, dtrain, outputmargin = TRUE, training = TRUE,
+                    ntreelimit = 0)
     gpair <- obj(pred, dtrain)
     .Call(XGBoosterBoostOneIter_R, booster_handle, dtrain, gpair$grad, gpair$hess)
   }
@@ -166,13 +173,12 @@ xgb.iter.eval <- function(booster_handle, watchlist, iter, feval = NULL) {
   evnames <- names(watchlist)
   if (is.null(feval)) {
     msg <- .Call(XGBoosterEvalOneIter_R, booster_handle, as.integer(iter), watchlist, as.list(evnames))
-    msg <- stri_split_regex(msg, '(\\s+|:|\\s+)')[[1]][-1]
-    res <- as.numeric(msg[c(FALSE,TRUE)]) # even indices are the values
-    names(res) <- msg[c(TRUE,FALSE)]      # odds are the names
+    mat <- matrix(strsplit(msg, '\\s+|:')[[1]][-1], nrow = 2)
+    res <- structure(as.numeric(mat[2, ]), names = mat[1, ])
   } else {
     res <- sapply(seq_along(watchlist), function(j) {
       w <- watchlist[[j]]
-      preds <- predict(booster_handle, w) # predict using all trees
+      preds <- predict(booster_handle, w, outputmargin = TRUE, ntreelimit = 0) # predict using all trees
       eval_res <- feval(preds, w)
       out <- eval_res$value
       names(out) <- paste0(evnames[j], "-", eval_res$metric)
@@ -187,13 +193,23 @@ xgb.iter.eval <- function(booster_handle, watchlist, iter, feval = NULL) {
 # Helper functions for cross validation ---------------------------------------
 #
 
+# Possibly convert the labels into factors, depending on the objective.
+# The labels are converted into factors only when the given objective refers to the classification
+# or ranking tasks.
+convert.labels <- function(labels, objective_name) {
+  if (objective_name %in% .CLASSIFICATION_OBJECTIVES()) {
+    return(as.factor(labels))
+  } else {
+    return(labels)
+  }
+}
+
 # Generates random (stratified if needed) CV folds
 generate.cv.folds <- function(nfold, nrows, stratified, label, params) {
 
   # cannot do it for rank
-  if (exists('objective', where = params) &&
-      is.character(params$objective) &&
-      strtrim(params$objective, 5) == 'rank:') {
+  objective <- params$objective
+  if (is.character(objective) && strtrim(objective, 5) == 'rank:') {
     stop("\n\tAutomatic generation of CV-folds is not implemented for ranking!\n",
          "\tConsider providing pre-computed CV-folds through the 'folds=' parameter.\n")
   }
@@ -206,19 +222,16 @@ generate.cv.folds <- function(nfold, nrows, stratified, label, params) {
     #  - For classification, need to convert y labels to factor before making the folds,
     #    and then do stratification by factor levels.
     #  - For regression, leave y numeric and do stratification by quantiles.
-    if (exists('objective', where = params) &&
-        is.character(params$objective)) {
-      # If 'objective' provided in params, assume that y is a classification label
-      # unless objective is reg:squarederror
-      if (params$objective != 'reg:squarederror')
-        y <- factor(y)
+    if (is.character(objective)) {
+      y <- convert.labels(y, params$objective)
     } else {
       # If no 'objective' given in params, it means that user either wants to
       # use the default 'reg:squarederror' objective or has provided a custom
       # obj function.  Here, assume classification setting when y has 5 or less
       # unique values:
-      if (length(unique(y)) <= 5)
+      if (length(unique(y)) <= 5) {
         y <- factor(y)
+      }
     }
     folds <- xgb.createFolds(y, nfold)
   } else {
@@ -307,6 +320,68 @@ xgb.createFolds <- function(y, k = 10)
 #' @name xgboost-deprecated
 NULL
 
+#' Do not use \code{\link[base]{saveRDS}} or \code{\link[base]{save}} for long-term archival of
+#' models. Instead, use \code{\link{xgb.save}} or \code{\link{xgb.save.raw}}.
+#'
+#' It is a common practice to use the built-in \code{\link[base]{saveRDS}} function (or
+#' \code{\link[base]{save}}) to persist R objects to the disk. While it is possible to persist
+#' \code{xgb.Booster} objects using \code{\link[base]{saveRDS}}, it is not advisable to do so if
+#' the model is to be accessed in the future. If you train a model with the current version of
+#' XGBoost and persist it with \code{\link[base]{saveRDS}}, the model is not guaranteed to be
+#' accessible in later releases of XGBoost. To ensure that your model can be accessed in future
+#' releases of XGBoost, use \code{\link{xgb.save}} or \code{\link{xgb.save.raw}} instead.
+#'
+#' @details
+#' Use \code{\link{xgb.save}} to save the XGBoost model as a stand-alone file. You may opt into
+#' the JSON format by specifying the JSON extension. To read the model back, use
+#' \code{\link{xgb.load}}.
+#'
+#' Use \code{\link{xgb.save.raw}} to save the XGBoost model as a sequence (vector) of raw bytes
+#' in a future-proof manner. Future releases of XGBoost will be able to read the raw bytes and
+#' re-construct the corresponding model. To read the model back, use \code{\link{xgb.load.raw}}.
+#' The \code{\link{xgb.save.raw}} function is useful if you'd like to persist the XGBoost model
+#' as part of another R object.
+#'
+#' Note: Do not use \code{\link{xgb.serialize}} to store models long-term. It persists not only the
+#' model but also internal configurations and parameters, and its format is not stable across
+#' multiple XGBoost versions. Use \code{\link{xgb.serialize}} only for checkpointing.
+#'
+#' For more details and explanation about model persistence and archival, consult the page
+#' \url{https://xgboost.readthedocs.io/en/latest/tutorials/saving_model.html}.
+#'
+#' @examples
+#' data(agaricus.train, package='xgboost')
+#' bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2,
+#'                eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
+#'
+#' # Save as a stand-alone file; load it with xgb.load()
+#' xgb.save(bst, 'xgb.model')
+#' bst2 <- xgb.load('xgb.model')
+#'
+#' # Save as a stand-alone file (JSON); load it with xgb.load()
+#' xgb.save(bst, 'xgb.model.json')
+#' bst2 <- xgb.load('xgb.model.json')
+#' if (file.exists('xgb.model.json')) file.remove('xgb.model.json')
+#'
+#' # Save as a raw byte vector; load it with xgb.load.raw()
+#' xgb_bytes <- xgb.save.raw(bst)
+#' bst2 <- xgb.load.raw(xgb_bytes)
+#'
+#' # Persist XGBoost model as part of another R object
+#' obj <- list(xgb_model_bytes = xgb.save.raw(bst), description = "My first XGBoost model")
+#' # Persist the R object. Here, saveRDS() is okay, since it doesn't persist
+#' # xgb.Booster directly. What's being persisted is the future-proof byte representation
+#' # as given by xgb.save.raw().
+#' saveRDS(obj, 'my_object.rds')
+#' # Read back the R object
+#' obj2 <- readRDS('my_object.rds')
+#' # Re-construct xgb.Booster object from the bytes
+#' bst2 <- xgb.load.raw(obj2$xgb_model_bytes)
+#' if (file.exists('my_object.rds')) file.remove('my_object.rds')
+#'
+#' @name a-compatibility-note-for-saveRDS-save
+NULL
+
 # Lookup table for the deprecated parameters bookkeeping
 depr_par_lut <- matrix(c(
   'print.every.n', 'print_every_n',
@@ -315,8 +390,8 @@ depr_par_lut <- matrix(c(
   'with.stats', 'with_stats',
   'numberOfClusters', 'n_clusters',
   'features.keep', 'features_keep',
-  'plot.height','plot_height',
-  'plot.width','plot_width',
+  'plot.height', 'plot_height',
+  'plot.width', 'plot_width',
   'n_first_tree', 'trees',
   'dummy', 'DUMMY'
 ), ncol = 2, byrow = TRUE)
@@ -329,20 +404,20 @@ colnames(depr_par_lut) <- c('old', 'new')
 check.deprecation <- function(..., env = parent.frame()) {
   pars <- list(...)
   # exact and partial matches
-  all_match <- pmatch(names(pars), depr_par_lut[,1])
+  all_match <- pmatch(names(pars), depr_par_lut[, 1])
   # indices of matched pars' names
   idx_pars <- which(!is.na(all_match))
   if (length(idx_pars) == 0) return()
   # indices of matched LUT rows
   idx_lut <- all_match[idx_pars]
   # which of idx_lut were the exact matches?
-  ex_match <- depr_par_lut[idx_lut,1] %in% names(pars)
+  ex_match <- depr_par_lut[idx_lut, 1] %in% names(pars)
   for (i in seq_along(idx_pars)) {
     pars_par <- names(pars)[idx_pars[i]]
     old_par <- depr_par_lut[idx_lut[i], 1]
     new_par <- depr_par_lut[idx_lut[i], 2]
     if (!ex_match[i]) {
-      warning("'", pars_par, "' was partially matched to '", old_par,"'")
+      warning("'", pars_par, "' was partially matched to '", old_par, "'")
     }
     .Deprecated(new_par, old = old_par, package = 'xgboost')
     if (new_par != 'NULL') {

R-package/R/xgb.Booster.R

@@ -1,6 +1,7 @@
 # Construct an internal xgboost Booster and return a handle to it.
 # internal utility function
-xgb.Booster.handle <- function(params = list(), cachelist = list(), modelfile = NULL) {
+xgb.Booster.handle <- function(params = list(), cachelist = list(),
+                               modelfile = NULL) {
   if (typeof(cachelist) != "list" ||
       !all(vapply(cachelist, inherits, logical(1), what = 'xgb.DMatrix'))) {
     stop("cachelist must be a list of xgb.DMatrix objects")
@@ -10,6 +11,7 @@ xgb.Booster.handle <- function(params = list(), cachelist = list(), modelfile =
     if (typeof(modelfile) == "character") {
       ## A filename
       handle <- .Call(XGBoosterCreate_R, cachelist)
+      modelfile <- path.expand(modelfile)
       .Call(XGBoosterLoadModel_R, handle, modelfile[1])
       class(handle) <- "xgb.Booster.handle"
       if (length(params) > 0) {
@@ -62,8 +64,8 @@ is.null.handle <- function(handle) {
   return(FALSE)
 }
 
-# Return a verified to be valid handle out of either xgb.Booster.handle or xgb.Booster
-# internal utility function
+# Return a verified to be valid handle out of either xgb.Booster.handle or
+# xgb.Booster internal utility function
 xgb.get.handle <- function(object) {
   if (inherits(object, "xgb.Booster")) {
     handle <- object$handle
@@ -110,6 +112,8 @@ xgb.get.handle <- function(object) {
 #'                eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
 #' saveRDS(bst, "xgb.model.rds")
 #'
+#' # Warning: The resulting RDS file is only compatible with the current XGBoost version.
+#' # Refer to the section titled "a-compatibility-note-for-saveRDS-save".
 #' bst1 <- readRDS("xgb.model.rds")
 #' if (file.exists("xgb.model.rds")) file.remove("xgb.model.rds")
 #' # the handle is invalid:
@@ -369,8 +373,8 @@ predict.xgb.Booster <- function(object, newdata, missing = NA, outputmargin = FA
       matrix(ret, nrow = n_row, byrow = TRUE, dimnames = list(NULL, cnames))
     } else {
       arr <- array(ret, c(n_col1, n_group, n_row),
-                   dimnames = list(cnames, NULL, NULL)) %>% aperm(c(2,3,1)) # [group, row, col]
-      lapply(seq_len(n_group), function(g) arr[g,,])
+                   dimnames = list(cnames, NULL, NULL)) %>% aperm(c(2, 3, 1)) # [group, row, col]
+      lapply(seq_len(n_group), function(g) arr[g, , ])
     }
   } else if (predinteraction) {
     n_col1 <- ncol(newdata) + 1
@@ -379,11 +383,11 @@ predict.xgb.Booster <- function(object, newdata, missing = NA, outputmargin = FA
     ret <- if (n_ret == n_row) {
       matrix(ret, ncol = 1, dimnames = list(NULL, cnames))
     } else if (n_group == 1) {
-      array(ret, c(n_col1, n_col1, n_row), dimnames = list(cnames, cnames, NULL)) %>% aperm(c(3,1,2))
+      array(ret, c(n_col1, n_col1, n_row), dimnames = list(cnames, cnames, NULL)) %>% aperm(c(3, 1, 2))
     } else {
       arr <- array(ret, c(n_col1, n_col1, n_group, n_row),
-                   dimnames = list(cnames, cnames, NULL, NULL)) %>% aperm(c(3,4,1,2)) # [group, row, col1, col2]
-      lapply(seq_len(n_group), function(g) arr[g,,,])
+                   dimnames = list(cnames, cnames, NULL, NULL)) %>% aperm(c(3, 4, 1, 2)) # [group, row, col1, col2]
+      lapply(seq_len(n_group), function(g) arr[g, , , ])
     }
   } else if (reshape && npred_per_case > 1) {
     ret <- matrix(ret, nrow = n_row, byrow = TRUE)
@@ -656,7 +660,7 @@ print.xgb.Booster <- function(x, verbose = FALSE, ...) {
 
   if (!is.null(x$params)) {
     cat('params (as set within xgb.train):\n')
-    cat( '  ',
+    cat('  ',
          paste(names(x$params),
                paste0('"', unlist(x$params), '"'),
                sep = ' = ', collapse = ', '), '\n', sep = '')
@@ -669,9 +673,9 @@ print.xgb.Booster <- function(x, verbose = FALSE, ...) {
   if (length(attrs) > 0) {
     cat('xgb.attributes:\n')
     if (verbose) {
-      cat( paste(paste0('  ',names(attrs)),
-                 paste0('"', unlist(attrs), '"'),
-                 sep = ' = ', collapse = '\n'), '\n', sep = '')
+        cat(paste(paste0('  ', names(attrs)),
+                  paste0('"', unlist(attrs), '"'),
+                  sep = ' = ', collapse = '\n'), '\n', sep = '')
     } else {
       cat('  ', paste(names(attrs), collapse = ', '), '\n', sep = '')
     }
@@ -693,7 +697,7 @@ print.xgb.Booster <- function(x, verbose = FALSE, ...) {
   #cat('ntree: ', xgb.ntree(x), '\n', sep='')
 
   for (n in setdiff(names(x), c('handle', 'raw', 'call', 'params', 'callbacks',
-                                'evaluation_log','niter','feature_names'))) {
+                                'evaluation_log', 'niter', 'feature_names'))) {
     if (is.atomic(x[[n]])) {
       cat(n, ':', x[[n]], '\n', sep = ' ')
     } else {

R-package/R/xgb.DMatrix.R

@@ -15,8 +15,7 @@
 #'
 #' @examples
 #' data(agaricus.train, package='xgboost')
-#' train <- agaricus.train
-#' dtrain <- xgb.DMatrix(train$data, label=train$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
 #' xgb.DMatrix.save(dtrain, 'xgb.DMatrix.data')
 #' dtrain <- xgb.DMatrix('xgb.DMatrix.data')
 #' if (file.exists('xgb.DMatrix.data')) file.remove('xgb.DMatrix.data')
@@ -27,6 +26,7 @@ xgb.DMatrix <- function(data, info = list(), missing = NA, silent = FALSE, ...)
     if (length(data) > 1)
       stop("'data' has class 'character' and length ", length(data),
            ".\n  'data' accepts either a numeric matrix or a single filename.")
+    data <- path.expand(data)
     handle <- .Call(XGDMatrixCreateFromFile_R, data, as.integer(silent))
   } else if (is.matrix(data)) {
     handle <- .Call(XGDMatrixCreateFromMat_R, data, missing)
@@ -65,6 +65,7 @@ xgb.get.DMatrix <- function(data, label = NULL, missing = NA, weight = NULL) {
       warning("xgboost: label will be ignored.")
     }
     if (is.character(data)) {
+      data <- path.expand(data)
       dtrain <- xgb.DMatrix(data[1])
     } else if (inherits(data, "xgb.DMatrix")) {
       dtrain <- data
@@ -171,8 +172,7 @@ dimnames.xgb.DMatrix <- function(x) {
 #'
 #' @examples
 #' data(agaricus.train, package='xgboost')
-#' train <- agaricus.train
-#' dtrain <- xgb.DMatrix(train$data, label=train$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
 #'
 #' labels <- getinfo(dtrain, 'label')
 #' setinfo(dtrain, 'label', 1-labels)
@@ -224,8 +224,7 @@ getinfo.xgb.DMatrix <- function(object, name, ...) {
 #'
 #' @examples
 #' data(agaricus.train, package='xgboost')
-#' train <- agaricus.train
-#' dtrain <- xgb.DMatrix(train$data, label=train$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
 #'
 #' labels <- getinfo(dtrain, 'label')
 #' setinfo(dtrain, 'label', 1-labels)
@@ -257,8 +256,6 @@ setinfo.xgb.DMatrix <- function(object, name, info, ...) {
     return(TRUE)
   }
   if (name == "weight") {
-    if (length(info) != nrow(object))
-      stop("The length of weights must equal to the number of rows in the input data")
     .Call(XGDMatrixSetInfo_R, object, name, as.numeric(info))
     return(TRUE)
   }
@@ -292,8 +289,7 @@ setinfo.xgb.DMatrix <- function(object, name, info, ...) {
 #'
 #' @examples
 #' data(agaricus.train, package='xgboost')
-#' train <- agaricus.train
-#' dtrain <- xgb.DMatrix(train$data, label=train$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
 #'
 #' dsub <- slice(dtrain, 1:42)
 #' labels1 <- getinfo(dsub, 'label')
@@ -322,7 +318,7 @@ slice.xgb.DMatrix <- function(object, idxset, ...) {
     for (i in seq_along(ind)) {
       obj_attr <- attr(object, nms[i])
       if (NCOL(obj_attr) > 1) {
-        attr(ret, nms[i]) <- obj_attr[idxset,]
+        attr(ret, nms[i]) <- obj_attr[idxset, ]
       } else {
         attr(ret, nms[i]) <- obj_attr[idxset]
       }
@@ -349,8 +345,7 @@ slice.xgb.DMatrix <- function(object, idxset, ...) {
 #'
 #' @examples
 #' data(agaricus.train, package='xgboost')
-#' train <- agaricus.train
-#' dtrain <- xgb.DMatrix(train$data, label=train$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
 #'
 #' dtrain
 #' print(dtrain, verbose=TRUE)
@@ -359,10 +354,10 @@ slice.xgb.DMatrix <- function(object, idxset, ...) {
 #' @export
 print.xgb.DMatrix <- function(x, verbose = FALSE, ...) {
   cat('xgb.DMatrix  dim:', nrow(x), 'x', ncol(x), ' info: ')
-  infos <- c()
-  if(length(getinfo(x, 'label')) > 0) infos <- 'label'
-  if(length(getinfo(x, 'weight')) > 0) infos <- c(infos, 'weight')
-  if(length(getinfo(x, 'base_margin')) > 0) infos <- c(infos, 'base_margin')
+  infos <- character(0)
+  if (length(getinfo(x, 'label')) > 0) infos <- 'label'
+  if (length(getinfo(x, 'weight')) > 0) infos <- c(infos, 'weight')
+  if (length(getinfo(x, 'base_margin')) > 0) infos <- c(infos, 'base_margin')
   if (length(infos) == 0) infos <- 'NA'
   cat(infos)
   cnames <- colnames(x)

R-package/R/xgb.DMatrix.save.R

@@ -1,14 +1,13 @@
 #' Save xgb.DMatrix object to binary file
-#' 
+#'
 #' Save xgb.DMatrix object to binary file
-#' 
+#'
 #' @param dmatrix the \code{xgb.DMatrix} object
 #' @param fname the name of the file to write.
-#' 
+#'
 #' @examples
 #' data(agaricus.train, package='xgboost')
-#' train <- agaricus.train
-#' dtrain <- xgb.DMatrix(train$data, label=train$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
 #' xgb.DMatrix.save(dtrain, 'xgb.DMatrix.data')
 #' dtrain <- xgb.DMatrix('xgb.DMatrix.data')
 #' if (file.exists('xgb.DMatrix.data')) file.remove('xgb.DMatrix.data')
@@ -18,7 +17,8 @@ xgb.DMatrix.save <- function(dmatrix, fname) {
     stop("fname must be character")
   if (!inherits(dmatrix, "xgb.DMatrix"))
     stop("dmatrix must be xgb.DMatrix")
-  
+
+  fname <- path.expand(fname)
   .Call(XGDMatrixSaveBinary_R, dmatrix, fname[1], 0L)
   return(TRUE)
 }

R-package/R/xgb.config.R

@@ -0,0 +1,38 @@
+#' Global configuration consists of a collection of parameters that can be applied in the global
+#' scope. See \url{https://xgboost.readthedocs.io/en/stable/parameter.html} for the full list of
+#' parameters supported in the global configuration. Use \code{xgb.set.config} to update the
+#' values of one or more global-scope parameters. Use \code{xgb.get.config} to fetch the current
+#' values of all global-scope parameters (listed in
+#' \url{https://xgboost.readthedocs.io/en/stable/parameter.html}).
+#'
+#' @rdname xgbConfig
+#' @title Set and get global configuration
+#' @name xgb.set.config, xgb.get.config
+#' @export xgb.set.config xgb.get.config
+#' @param ... List of parameters to be set, as keyword arguments
+#' @return
+#' \code{xgb.set.config} returns \code{TRUE} to signal success. \code{xgb.get.config} returns
+#' a list containing all global-scope parameters and their values.
+#'
+#' @examples
+#' # Set verbosity level to silent (0)
+#' xgb.set.config(verbosity = 0)
+#' # Now global verbosity level is 0
+#' config <- xgb.get.config()
+#' print(config$verbosity)
+#' # Set verbosity level to warning (1)
+#' xgb.set.config(verbosity = 1)
+#' # Now global verbosity level is 1
+#' config <- xgb.get.config()
+#' print(config$verbosity)
+xgb.set.config <- function(...) {
+  new_config <- list(...)
+  .Call(XGBSetGlobalConfig_R, jsonlite::toJSON(new_config, auto_unbox = TRUE))
+  return(TRUE)
+}
+
+#' @rdname xgbConfig
+xgb.get.config <- function() {
+  config <- .Call(XGBGetGlobalConfig_R)
+  return(jsonlite::fromJSON(config))
+}

R-package/R/xgb.create.features.R

@@ -1,87 +1,87 @@
 #' Create new features from a previously learned model
-#' 
+#'
 #' May improve the learning by adding new features to the training data based on the decision trees from a previously learned model.
-#' 
+#'
 #' @param model decision tree boosting model learned on the original data
 #' @param data original data (usually provided as a \code{dgCMatrix} matrix)
 #' @param ... currently not used
-#' 
+#'
 #' @return \code{dgCMatrix} matrix including both the original data and the new features.
 #'
-#' @details 
+#' @details
 #' This is the function inspired from the paragraph 3.1 of the paper:
-#' 
+#'
 #' \strong{Practical Lessons from Predicting Clicks on Ads at Facebook}
-#' 
-#' \emph{(Xinran He, Junfeng Pan, Ou Jin, Tianbing Xu, Bo Liu, Tao Xu, Yan, xin Shi, Antoine Atallah, Ralf Herbrich, Stuart Bowers, 
+#'
+#' \emph{(Xinran He, Junfeng Pan, Ou Jin, Tianbing Xu, Bo Liu, Tao Xu, Yan, xin Shi, Antoine Atallah, Ralf Herbrich, Stuart Bowers,
 #' Joaquin Quinonero Candela)}
-#'  
+#'
 #' International Workshop on Data Mining for Online Advertising (ADKDD) - August 24, 2014
-#' 
+#'
 #' \url{https://research.fb.com/publications/practical-lessons-from-predicting-clicks-on-ads-at-facebook/}.
-#' 
+#'
 #' Extract explaining the method:
-#' 
+#'
 #' "We found that boosted decision trees are a powerful and very
 #' convenient way to implement non-linear and tuple transformations
 #' of the kind we just described. We treat each individual
 #' tree as a categorical feature that takes as value the
-#' index of the leaf an instance ends up falling in. We use 
-#' 1-of-K coding of this type of features. 
-#' 
-#' For example, consider the boosted tree model in Figure 1 with 2 subtrees, 
+#' index of the leaf an instance ends up falling in. We use
+#' 1-of-K coding of this type of features.
+#'
+#' For example, consider the boosted tree model in Figure 1 with 2 subtrees,
 #' where the first subtree has 3 leafs and the second 2 leafs. If an
 #' instance ends up in leaf 2 in the first subtree and leaf 1 in
 #' second subtree, the overall input to the linear classifier will
 #' be the binary vector \code{[0, 1, 0, 1, 0]}, where the first 3 entries
 #' correspond to the leaves of the first subtree and last 2 to
 #' those of the second subtree.
-#' 
+#'
 #' [...]
-#' 
+#'
 #' We can understand boosted decision tree
 #' based transformation as a supervised feature encoding that
 #' converts a real-valued vector into a compact binary-valued
 #' vector. A traversal from root node to a leaf node represents
 #' a rule on certain features."
-#' 
+#'
 #' @examples
 #' data(agaricus.train, package='xgboost')
 #' data(agaricus.test, package='xgboost')
-#' dtrain <- xgb.DMatrix(data = agaricus.train$data, label = agaricus.train$label)
-#' dtest <- xgb.DMatrix(data = agaricus.test$data, label = agaricus.test$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
+#' dtest <- with(agaricus.test, xgb.DMatrix(data, label = label))
 #'
 #' param <- list(max_depth=2, eta=1, silent=1, objective='binary:logistic')
 #' nrounds = 4
 #'
 #' bst = xgb.train(params = param, data = dtrain, nrounds = nrounds, nthread = 2)
-#' 
+#'
 #' # Model accuracy without new features
 #' accuracy.before <- sum((predict(bst, agaricus.test$data) >= 0.5) == agaricus.test$label) /
 #'                    length(agaricus.test$label)
-#' 
+#'
 #' # Convert previous features to one hot encoding
 #' new.features.train <- xgb.create.features(model = bst, agaricus.train$data)
 #' new.features.test <- xgb.create.features(model = bst, agaricus.test$data)
-#' 
+#'
 #' # learning with new features
 #' new.dtrain <- xgb.DMatrix(data = new.features.train, label = agaricus.train$label)
 #' new.dtest <- xgb.DMatrix(data = new.features.test, label = agaricus.test$label)
 #' watchlist <- list(train = new.dtrain)
 #' bst <- xgb.train(params = param, data = new.dtrain, nrounds = nrounds, nthread = 2)
-#' 
+#'
 #' # Model accuracy with new features
 #' accuracy.after <- sum((predict(bst, new.dtest) >= 0.5) == agaricus.test$label) /
 #'                   length(agaricus.test$label)
-#' 
+#'
 #' # Here the accuracy was already good and is now perfect.
 #' cat(paste("The accuracy was", accuracy.before, "before adding leaf features and it is now",
 #'           accuracy.after, "!\n"))
-#' 
+#'
 #' @export
 xgb.create.features <- function(model, data, ...){
   check.deprecation(...)
   pred_with_leaf <- predict(model, data, predleaf = TRUE)
   cols <- lapply(as.data.frame(pred_with_leaf), factor)
-  cbind(data, sparse.model.matrix( ~ . -1, cols))
+  cbind(data, sparse.model.matrix(~ . -1, cols)) # nolint
 }

R-package/R/xgb.cv.R

@@ -2,12 +2,15 @@
 #'
 #' The cross validation function of xgboost
 #'
-#' @param params the list of parameters. Commonly used ones are:
+#' @param params the list of parameters. The complete list of parameters is
+#'   available in the \href{http://xgboost.readthedocs.io/en/latest/parameter.html}{online documentation}. Below
+#'   is a shorter summary:
 #' \itemize{
 #'   \item \code{objective} objective function, common ones are
 #'   \itemize{
-#'     \item \code{reg:squarederror} Regression with squared loss
-#'     \item \code{binary:logistic} logistic regression for classification
+#'     \item \code{reg:squarederror} Regression with squared loss.
+#'     \item \code{binary:logistic} logistic regression for classification.
+#'     \item See \code{\link[=xgb.train]{xgb.train}()} for complete list of objectives.
 #'   }
 #'   \item \code{eta} step size of each boosting step
 #'   \item \code{max_depth} maximum depth of the tree
@@ -33,6 +36,8 @@
 #'   \item \code{error} binary classification error rate
 #'   \item \code{rmse} Rooted mean square error
 #'   \item \code{logloss} negative log-likelihood function
+#'   \item \code{mae} Mean absolute error
+#'   \item \code{mape} Mean absolute percentage error
 #'   \item \code{auc} Area under curve
 #'   \item \code{aucpr} Area under PR curve
 #'   \item \code{merror} Exact matching error, used to evaluate multi-class classification
@@ -76,7 +81,7 @@
 #'
 #' All observations are used for both training and validation.
 #'
-#' Adapted from \url{http://en.wikipedia.org/wiki/Cross-validation_\%28statistics\%29#k-fold_cross-validation}
+#' Adapted from \url{https://en.wikipedia.org/wiki/Cross-validation_\%28statistics\%29}
 #'
 #' @return
 #' An object of class \code{xgb.cv.synchronous} with the following elements:
@@ -107,7 +112,7 @@
 #'
 #' @examples
 #' data(agaricus.train, package='xgboost')
-#' dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
 #' cv <- xgb.cv(data = dtrain, nrounds = 3, nthread = 2, nfold = 5, metrics = list("rmse","auc"),
 #'                   max_depth = 3, eta = 1, objective = "binary:logistic")
 #' print(cv)
@@ -134,20 +139,20 @@ xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing =
   #  stop("Either 'eval_metric' or 'feval' must be provided for CV")
 
   # Check the labels
-  if ( (inherits(data, 'xgb.DMatrix') && is.null(getinfo(data, 'label'))) ||
-       (!inherits(data, 'xgb.DMatrix') && is.null(label))) {
+  if ((inherits(data, 'xgb.DMatrix') && is.null(getinfo(data, 'label'))) ||
+      (!inherits(data, 'xgb.DMatrix') && is.null(label))) {
     stop("Labels must be provided for CV either through xgb.DMatrix, or through 'label=' when 'data' is matrix")
   } else if (inherits(data, 'xgb.DMatrix')) {
     if (!is.null(label))
       warning("xgb.cv: label will be ignored, since data is of type xgb.DMatrix")
-    cv_label = getinfo(data, 'label')
+    cv_label <- getinfo(data, 'label')
   } else {
-    cv_label = label
+    cv_label <- label
   }
 
   # CV folds
-  if(!is.null(folds)) {
-    if(!is.list(folds) || length(folds) < 2)
+  if (!is.null(folds)) {
+    if (!is.list(folds) || length(folds) < 2)
       stop("'folds' must be a list with 2 or more elements that are vectors of indices for each CV-fold")
     nfold <- length(folds)
   } else {
@@ -162,7 +167,7 @@ xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing =
 
   # verbosity & evaluation printing callback:
   params <- c(params, list(silent = 1))
-  print_every_n <- max( as.integer(print_every_n), 1L)
+  print_every_n <- max(as.integer(print_every_n), 1L)
   if (!has.callbacks(callbacks, 'cb.print.evaluation') && verbose) {
     callbacks <- add.cb(callbacks, cb.print.evaluation(print_every_n, showsd = showsd))
   }
@@ -193,20 +198,20 @@ xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing =
   bst_folds <- lapply(seq_along(folds), function(k) {
     dtest  <- slice(dall, folds[[k]])
     # code originally contributed by @RolandASc on stackoverflow
-    if(is.null(train_folds))
+    if (is.null(train_folds))
        dtrain <- slice(dall, unlist(folds[-k]))
     else
        dtrain <- slice(dall, train_folds[[k]])
     handle <- xgb.Booster.handle(params, list(dtrain, dtest))
-    list(dtrain = dtrain, bst = handle, watchlist = list(train = dtrain, test=dtest), index = folds[[k]])
+    list(dtrain = dtrain, bst = handle, watchlist = list(train = dtrain, test = dtest), index = folds[[k]])
   })
   rm(dall)
   # a "basket" to collect some results from callbacks
   basket <- list()
 
   # extract parameters that can affect the relationship b/w #trees and #iterations
-  num_class <- max(as.numeric(NVL(params[['num_class']], 1)), 1)
-  num_parallel_tree <- max(as.numeric(NVL(params[['num_parallel_tree']], 1)), 1)
+  num_class <- max(as.numeric(NVL(params[['num_class']], 1)), 1) # nolint
+  num_parallel_tree <- max(as.numeric(NVL(params[['num_parallel_tree']], 1)), 1) # nolint
 
   # those are fixed for CV (no training continuation)
   begin_iteration <- 1
@@ -223,7 +228,7 @@ xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing =
     })
     msg <- simplify2array(msg)
     bst_evaluation <- rowMeans(msg)
-    bst_evaluation_err <- sqrt(rowMeans(msg^2) - bst_evaluation^2)
+    bst_evaluation_err <- sqrt(rowMeans(msg^2) - bst_evaluation^2) # nolint
 
     for (f in cb$post_iter) f()
 
@@ -282,10 +287,10 @@ print.xgb.cv.synchronous <- function(x, verbose = FALSE, ...) {
     }
     if (!is.null(x$params)) {
       cat('params (as set within xgb.cv):\n')
-      cat( '  ',
-           paste(names(x$params),
-                 paste0('"', unlist(x$params), '"'),
-                 sep = ' = ', collapse = ', '), '\n', sep = '')
+      cat('  ',
+          paste(names(x$params),
+                paste0('"', unlist(x$params), '"'),
+                sep = ' = ', collapse = ', '), '\n', sep = '')
     }
     if (!is.null(x$callbacks) && length(x$callbacks) > 0) {
       cat('callbacks:\n')

R-package/R/xgb.dump.R

@@ -1,15 +1,15 @@
 #' Dump an xgboost model in text format.
-#' 
+#'
 #' Dump an xgboost model in text format.
-#' 
+#'
 #' @param model the model object.
-#' @param fname the name of the text file where to save the model text dump. 
+#' @param fname the name of the text file where to save the model text dump.
 #'        If not provided or set to \code{NULL}, the model is returned as a \code{character} vector.
 #' @param fmap feature map file representing feature types.
-#'        Detailed description could be found at 
+#'        Detailed description could be found at
 #'        \url{https://github.com/dmlc/xgboost/wiki/Binary-Classification#dump-model}.
 #'        See demo/ for walkthrough example in R, and
-#'        \url{https://github.com/dmlc/xgboost/blob/master/demo/data/featmap.txt} 
+#'        \url{https://github.com/dmlc/xgboost/blob/master/demo/data/featmap.txt}
 #'        for example Format.
 #' @param with_stats whether to dump some additional statistics about the splits.
 #'        When this option is on, the model dump contains two additional values:
@@ -27,18 +27,18 @@
 #' data(agaricus.test, package='xgboost')
 #' train <- agaricus.train
 #' test <- agaricus.test
-#' bst <- xgboost(data = train$data, label = train$label, max_depth = 2, 
+#' bst <- xgboost(data = train$data, label = train$label, max_depth = 2,
 #'                eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
 #' # save the model in file 'xgb.model.dump'
 #' dump_path = file.path(tempdir(), 'model.dump')
 #' xgb.dump(bst, dump_path, with_stats = TRUE)
-#' 
+#'
 #' # print the model without saving it to a file
 #' print(xgb.dump(bst, with_stats = TRUE))
-#' 
+#'
 #' # print in JSON format:
 #' cat(xgb.dump(bst, with_stats = TRUE, dump_format='json'))
-#' 
+#'
 #' @export
 xgb.dump <- function(model, fname = NULL, fmap = "", with_stats=FALSE,
                      dump_format = c("text", "json"), ...) {
@@ -50,22 +50,23 @@ xgb.dump <- function(model, fname = NULL, fmap = "", with_stats=FALSE,
     stop("fname: argument must be a character string (when provided)")
   if (!(is.null(fmap) || is.character(fmap)))
     stop("fmap: argument must be a character string (when provided)")
-  
+
   model <- xgb.Booster.complete(model)
   model_dump <- .Call(XGBoosterDumpModel_R, model$handle, NVL(fmap, "")[1], as.integer(with_stats),
                       as.character(dump_format))
 
-  if (is.null(fname)) 
-    model_dump <- stri_replace_all_regex(model_dump, '\t', '')
-  
+  if (is.null(fname))
+    model_dump <- gsub('\t', '', model_dump, fixed = TRUE)
+
   if (dump_format == "text")
-    model_dump <- unlist(stri_split_regex(model_dump, '\n'))
-  
+    model_dump <- unlist(strsplit(model_dump, '\n', fixed = TRUE))
+
   model_dump <- grep('^\\s*$', model_dump, invert = TRUE, value = TRUE)
-  
+
   if (is.null(fname)) {
     return(model_dump)
   } else {
+    fname <- path.expand(fname)
     writeLines(model_dump, fname[1])
     return(TRUE)
   }

R-package/R/xgb.ggplot.R

@@ -3,9 +3,9 @@
 
 #' @rdname xgb.plot.importance
 #' @export
-xgb.ggplot.importance <- function(importance_matrix = NULL, top_n = NULL, measure = NULL, 
+xgb.ggplot.importance <- function(importance_matrix = NULL, top_n = NULL, measure = NULL,
                                   rel_to_first = FALSE, n_clusters = c(1:10), ...) {
-  
+
   importance_matrix <- xgb.plot.importance(importance_matrix, top_n = top_n, measure = measure,
                                            rel_to_first = rel_to_first, plot = FALSE, ...)
   if (!requireNamespace("ggplot2", quietly = TRUE)) {
@@ -14,21 +14,21 @@ xgb.ggplot.importance <- function(importance_matrix = NULL, top_n = NULL, measur
   if (!requireNamespace("Ckmeans.1d.dp", quietly = TRUE)) {
     stop("Ckmeans.1d.dp package is required", call. = FALSE)
   }
-  
+
   clusters <- suppressWarnings(
     Ckmeans.1d.dp::Ckmeans.1d.dp(importance_matrix$Importance, n_clusters)
   )
   importance_matrix[, Cluster := as.character(clusters$cluster)]
 
   plot <-
-    ggplot2::ggplot(importance_matrix, 
+    ggplot2::ggplot(importance_matrix,
                     ggplot2::aes(x = factor(Feature, levels = rev(Feature)), y = Importance, width = 0.5),
-                    environment = environment()) + 
-    ggplot2::geom_bar(ggplot2::aes(fill = Cluster), stat = "identity", position = "identity") + 
-    ggplot2::coord_flip() + 
-    ggplot2::xlab("Features") + 
-    ggplot2::ggtitle("Feature importance") + 
-    ggplot2::theme(plot.title = ggplot2::element_text(lineheight = .9, face = "bold"), 
+                    environment = environment()) +
+    ggplot2::geom_bar(ggplot2::aes(fill = Cluster), stat = "identity", position = "identity") +
+    ggplot2::coord_flip() +
+    ggplot2::xlab("Features") +
+    ggplot2::ggtitle("Feature importance") +
+    ggplot2::theme(plot.title = ggplot2::element_text(lineheight = .9, face = "bold"),
                    panel.grid.major.y = ggplot2::element_blank())
   return(plot)
 }
@@ -42,7 +42,7 @@ xgb.ggplot.deepness <- function(model = NULL, which = c("2x1", "max.depth", "med
     stop("ggplot2 package is required for plotting the graph deepness.", call. = FALSE)
 
   which <- match.arg(which)
-  
+
   dt_depths <- xgb.plot.deepness(model = model, plot = FALSE)
   dt_summaries <- dt_depths[, .(.N, Cover = mean(Cover)), Depth]
   setkey(dt_summaries, 'Depth')
@@ -60,30 +60,30 @@ xgb.ggplot.deepness <- function(model = NULL, which = c("2x1", "max.depth", "med
         axis.ticks = ggplot2::element_blank(),
         axis.text.x = ggplot2::element_blank()
       )
-  
-    p2 <- 
+
+    p2 <-
       ggplot2::ggplot(dt_summaries) +
-      ggplot2::geom_bar(ggplot2::aes(x = Depth, y = Cover), stat = "Identity") + 
+      ggplot2::geom_bar(ggplot2::aes(x = Depth, y = Cover), stat = "Identity") +
       ggplot2::xlab("Leaf depth") +
       ggplot2::ylab("Weighted cover")
-  
+
     multiplot(p1, p2, cols = 1)
     return(invisible(list(p1, p2)))
-    
+
   } else if (which == "max.depth") {
     p <-
       ggplot2::ggplot(dt_depths[, max(Depth), Tree]) +
       ggplot2::geom_jitter(ggplot2::aes(x = Tree, y = V1),
-                           height = 0.15, alpha=0.4, size=3, stroke=0) + 
+                           height = 0.15, alpha = 0.4, size = 3, stroke = 0) +
       ggplot2::xlab("tree #") +
       ggplot2::ylab("Max tree leaf depth")
     return(p)
-    
+
   } else if (which == "med.depth") {
     p <-
       ggplot2::ggplot(dt_depths[, median(as.numeric(Depth)), Tree]) +
       ggplot2::geom_jitter(ggplot2::aes(x = Tree, y = V1),
-                           height = 0.15, alpha=0.4, size=3, stroke=0) + 
+                           height = 0.15, alpha = 0.4, size = 3, stroke = 0) +
       ggplot2::xlab("tree #") +
       ggplot2::ylab("Median tree leaf depth")
     return(p)
@@ -92,24 +92,102 @@ xgb.ggplot.deepness <- function(model = NULL, which = c("2x1", "max.depth", "med
     p <-
       ggplot2::ggplot(dt_depths[, median(abs(Weight)), Tree]) +
       ggplot2::geom_point(ggplot2::aes(x = Tree, y = V1),
-                          alpha=0.4, size=3, stroke=0) + 
+                          alpha = 0.4, size = 3, stroke = 0) +
       ggplot2::xlab("tree #") +
       ggplot2::ylab("Median absolute leaf weight")
     return(p)
   }
 }
 
+#' @rdname xgb.plot.shap.summary
+#' @export
+xgb.ggplot.shap.summary <- function(data, shap_contrib = NULL, features = NULL, top_n = 10, model = NULL,
+                                    trees = NULL, target_class = NULL, approxcontrib = FALSE, subsample = NULL) {
+  data_list <- xgb.shap.data(
+    data = data,
+    shap_contrib = shap_contrib,
+    features = features,
+    top_n = top_n,
+    model = model,
+    trees = trees,
+    target_class = target_class,
+    approxcontrib = approxcontrib,
+    subsample = subsample,
+    max_observations = 10000  # 10,000 samples per feature.
+  )
+  p_data <- prepare.ggplot.shap.data(data_list, normalize = TRUE)
+  # Reverse factor levels so that the first level is at the top of the plot
+  p_data[, "feature" := factor(feature, rev(levels(feature)))]
+  p <- ggplot2::ggplot(p_data, ggplot2::aes(x = feature, y = p_data$shap_value, colour = p_data$feature_value)) +
+    ggplot2::geom_jitter(alpha = 0.5, width = 0.1) +
+    ggplot2::scale_colour_viridis_c(limits = c(-3, 3), option = "plasma", direction = -1) +
+    ggplot2::geom_abline(slope = 0, intercept = 0, colour = "darkgrey") +
+    ggplot2::coord_flip()
+
+  p
+}
+
+#' Combine and melt feature values and SHAP contributions for sample
+#' observations.
+#'
+#' Conforms to data format required for ggplot functions.
+#'
+#' Internal utility function.
+#'
+#' @param data_list List containing 'data' and 'shap_contrib' returned by
+#'   \code{xgb.shap.data()}.
+#' @param normalize Whether to standardize feature values to have mean 0 and
+#'   standard deviation 1 (useful for comparing multiple features on the same
+#'   plot). Default \code{FALSE}.
+#'
+#' @return A data.table containing the observation ID, the feature name, the
+#'   feature value (normalized if specified), and the SHAP contribution value.
+prepare.ggplot.shap.data <- function(data_list, normalize = FALSE) {
+  data <- data_list[["data"]]
+  shap_contrib <- data_list[["shap_contrib"]]
+
+  data <- data.table::as.data.table(as.matrix(data))
+  if (normalize) {
+    data[, (names(data)) := lapply(.SD, normalize)]
+  }
+  data[, "id" := seq_len(nrow(data))]
+  data_m <- data.table::melt.data.table(data, id.vars = "id", variable.name = "feature", value.name = "feature_value")
+
+  shap_contrib <- data.table::as.data.table(as.matrix(shap_contrib))
+  shap_contrib[, "id" := seq_len(nrow(shap_contrib))]
+  shap_contrib_m <- data.table::melt.data.table(shap_contrib, id.vars = "id", variable.name = "feature", value.name = "shap_value")
+
+  p_data <- data.table::merge.data.table(data_m, shap_contrib_m, by = c("id", "feature"))
+
+  p_data
+}
+
+#' Scale feature value to have mean 0, standard deviation 1
+#'
+#' This is used to compare multiple features on the same plot.
+#' Internal utility function
+#'
+#' @param x Numeric vector
+#'
+#' @return Numeric vector with mean 0 and sd 1.
+normalize <- function(x) {
+  loc <- mean(x, na.rm = TRUE)
+  scale <- stats::sd(x, na.rm = TRUE)
+
+  (x - loc) / scale
+}
+
 # Plot multiple ggplot graph aligned by rows and columns.
 # ... the plots
 # cols number of columns
 # internal utility function
 multiplot <- function(..., cols = 1) {
   plots <- list(...)
-  num_plots = length(plots)
-  
+  num_plots <- length(plots)
+
   layout <- matrix(seq(1, cols * ceiling(num_plots / cols)),
                    ncol = cols, nrow = ceiling(num_plots / cols))
-  
+
   if (num_plots == 1) {
     print(plots[[1]])
   } else {
@@ -118,7 +196,7 @@ multiplot <- function(..., cols = 1) {
     for (i in 1:num_plots) {
       # Get the i,j matrix positions of the regions that contain this subplot
       matchidx <- as.data.table(which(layout == i, arr.ind = TRUE))
-      
+
       print(
         plots[[i]], vp = grid::viewport(
           layout.pos.row = matchidx$row,
@@ -131,5 +209,5 @@ multiplot <- function(..., cols = 1) {
 
 globalVariables(c(
   "Cluster", "ggplot", "aes", "geom_bar", "coord_flip", "xlab", "ylab", "ggtitle", "theme",
-  "element_blank", "element_text", "V1", "Weight"
+  "element_blank", "element_text", "V1", "Weight", "feature"
 ))

R-package/R/xgb.importance.R

@@ -1,66 +1,66 @@
 #' Importance of features in a model.
-#' 
+#'
 #' Creates a \code{data.table} of feature importances in a model.
-#' 
+#'
 #' @param feature_names character vector of feature names. If the model already
 #'       contains feature names, those would be used when \code{feature_names=NULL} (default value).
 #'       Non-null \code{feature_names} could be provided to override those in the model.
 #' @param model object of class \code{xgb.Booster}.
 #' @param trees (only for the gbtree booster) an integer vector of tree indices that should be included
 #'          into the importance calculation. If set to \code{NULL}, all trees of the model are parsed.
-#'          It could be useful, e.g., in multiclass classification to get feature importances 
+#'          It could be useful, e.g., in multiclass classification to get feature importances
 #'          for each class separately. IMPORTANT: the tree index in xgboost models
 #'          is zero-based (e.g., use \code{trees = 0:4} for first 5 trees).
 #' @param data deprecated.
 #' @param label deprecated.
 #' @param target deprecated.
 #'
-#' @details 
-#' 
+#' @details
+#'
 #' This function works for both linear and tree models.
-#' 
-#' For linear models, the importance is the absolute magnitude of linear coefficients. 
-#' For that reason, in order to obtain a meaningful ranking by importance for a linear model, 
-#' the features need to be on the same scale (which you also would want to do when using either 
+#'
+#' For linear models, the importance is the absolute magnitude of linear coefficients.
+#' For that reason, in order to obtain a meaningful ranking by importance for a linear model,
+#' the features need to be on the same scale (which you also would want to do when using either
 #' L1 or L2 regularization).
-#' 
+#'
 #' @return
-#' 
+#'
 #' For a tree model, a \code{data.table} with the following columns:
 #' \itemize{
 #'   \item \code{Features} names of the features used in the model;
 #'   \item \code{Gain} represents fractional contribution of each feature to the model based on
-#'        the total gain of this feature's splits. Higher percentage means a more important 
+#'        the total gain of this feature's splits. Higher percentage means a more important
 #'        predictive feature.
 #'   \item \code{Cover} metric of the number of observation related to this feature;
 #'   \item \code{Frequency} percentage representing the relative number of times
 #'        a feature have been used in trees.
 #' }
-#' 
+#'
 #' A linear model's importance \code{data.table} has the following columns:
 #' \itemize{
 #'   \item \code{Features} names of the features used in the model;
 #'   \item \code{Weight} the linear coefficient of this feature;
 #'   \item \code{Class} (only for multiclass models) class label.
 #' }
-#' 
-#' If \code{feature_names} is not provided and \code{model} doesn't have \code{feature_names}, 
+#'
+#' If \code{feature_names} is not provided and \code{model} doesn't have \code{feature_names},
 #' index of the features will be used instead. Because the index is extracted from the model dump
 #' (based on C++ code), it starts at 0 (as in C/C++ or Python) instead of 1 (usual in R).
-#' 
+#'
 #' @examples
-#' 
+#'
 #' # binomial classification using gbtree:
 #' data(agaricus.train, package='xgboost')
-#' bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2, 
+#' bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2,
 #'                eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
 #' xgb.importance(model = bst)
-#' 
+#'
 #' # binomial classification using gblinear:
-#' bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, booster = "gblinear", 
+#' bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, booster = "gblinear",
 #'                eta = 0.3, nthread = 1, nrounds = 20, objective = "binary:logistic")
 #' xgb.importance(model = bst)
-#' 
+#'
 #' # multiclass classification using gbtree:
 #' nclass <- 3
 #' nrounds <- 10
@@ -73,7 +73,7 @@
 #' xgb.importance(model = mbst, trees = seq(from=0, by=nclass, length.out=nrounds))
 #' xgb.importance(model = mbst, trees = seq(from=1, by=nclass, length.out=nrounds))
 #' xgb.importance(model = mbst, trees = seq(from=2, by=nclass, length.out=nrounds))
-#' 
+#'
 #' # multiclass classification using gblinear:
 #' mbst <- xgboost(data = scale(as.matrix(iris[, -5])), label = as.numeric(iris$Species) - 1,
 #'                booster = "gblinear", eta = 0.2, nthread = 1, nrounds = 15,
@@ -83,33 +83,33 @@
 #' @export
 xgb.importance <- function(feature_names = NULL, model = NULL, trees = NULL,
                            data = NULL, label = NULL, target = NULL){
-  
+
   if (!(is.null(data) && is.null(label) && is.null(target)))
     warning("xgb.importance: parameters 'data', 'label' and 'target' are deprecated")
-  
+
   if (!inherits(model, "xgb.Booster"))
     stop("model: must be an object of class xgb.Booster")
-  
+
   if (is.null(feature_names) && !is.null(model$feature_names))
     feature_names <- model$feature_names
-  
+
   if (!(is.null(feature_names) || is.character(feature_names)))
     stop("feature_names: Has to be a character vector")
 
   model_text_dump <- xgb.dump(model = model, with_stats = TRUE)
-  
+
   # linear model
-  if(model_text_dump[2] == "bias:"){
+  if (model_text_dump[2] == "bias:"){
     weights <- which(model_text_dump == "weight:") %>%
                {model_text_dump[(. + 1):length(model_text_dump)]} %>%
                as.numeric
-    
+
     num_class <- NVL(model$params$num_class, 1)
-    if(is.null(feature_names)) 
+    if (is.null(feature_names))
       feature_names <- seq(to = length(weights) / num_class) - 1
     if (length(feature_names) * num_class != length(weights))
       stop("feature_names length does not match the number of features used in the model")
-    
+
     result <- if (num_class == 1) {
       data.table(Feature = feature_names, Weight = weights)[order(-abs(Weight))]
     } else {
@@ -117,18 +117,17 @@ xgb.importance <- function(feature_names = NULL, model = NULL, trees = NULL,
                  Weight = weights,
                  Class = seq_len(num_class) - 1)[order(Class, -abs(Weight))]
     }
-  } else { 
-  # tree model
-    result <- xgb.model.dt.tree(feature_names = feature_names,
-                                text = model_text_dump,
-                                trees = trees)[
-      Feature != "Leaf", .(Gain = sum(Quality), 
-                           Cover = sum(Cover), 
-                           Frequency = .N), by = Feature][
-      ,`:=`(Gain = Gain / sum(Gain), 
-            Cover = Cover / sum(Cover),
-            Frequency = Frequency / sum(Frequency))][
-      order(Gain, decreasing = TRUE)]
+  } else { # tree model
+      result <- xgb.model.dt.tree(feature_names = feature_names,
+                                  text = model_text_dump,
+                                  trees = trees)[
+          Feature != "Leaf", .(Gain = sum(Quality),
+                               Cover = sum(Cover),
+                               Frequency = .N), by = Feature][
+        , `:=`(Gain = Gain / sum(Gain),
+               Cover = Cover / sum(Cover),
+               Frequency = Frequency / sum(Frequency))][
+          order(Gain, decreasing = TRUE)]
   }
   result
 }

R-package/R/xgb.model.dt.tree.R

@@ -1,12 +1,12 @@
 #' Parse a boosted tree model text dump
-#' 
+#'
 #' Parse a boosted tree model text dump into a \code{data.table} structure.
-#' 
+#'
 #' @param feature_names character vector of feature names. If the model already
 #'          contains feature names, those would be used when \code{feature_names=NULL} (default value).
 #'          Non-null \code{feature_names} could be provided to override those in the model.
 #' @param model object of class \code{xgb.Booster}
-#' @param text \code{character} vector previously generated by the \code{xgb.dump} 
+#' @param text \code{character} vector previously generated by the \code{xgb.dump}
 #'          function  (where parameter \code{with_stats = TRUE} should have been set).
 #'          \code{text} takes precedence over \code{model}.
 #' @param trees an integer vector of tree indices that should be parsed.
@@ -18,11 +18,11 @@
 #'          represented as integers (when FALSE) or as "Tree-Node" character strings (when FALSE).
 #' @param ... currently not used.
 #'
-#' @return 
+#' @return
 #' A \code{data.table} with detailed information about model trees' nodes.
 #'
 #' The columns of the \code{data.table} are:
-#' 
+#'
 #' \itemize{
 #'  \item \code{Tree}: integer ID of a tree in a model (zero-based index)
 #'  \item \code{Node}: integer ID of a node in a tree (zero-based index)
@@ -36,109 +36,111 @@
 #'  \item \code{Quality}: either the split gain (change in loss) or the leaf value
 #'  \item \code{Cover}: metric related to the number of observation either seen by a split
 #'                      or collected by a leaf during training.
-#' } 
-#' 
+#' }
+#'
 #' When \code{use_int_id=FALSE}, columns "Yes", "No", and "Missing" point to model-wide node identifiers
-#' in the "ID" column. When \code{use_int_id=TRUE}, those columns point to node identifiers from 
+#' in the "ID" column. When \code{use_int_id=TRUE}, those columns point to node identifiers from
 #' the corresponding trees in the "Node" column.
-#' 
+#'
 #' @examples
 #' # Basic use:
-#' 
+#'
 #' data(agaricus.train, package='xgboost')
-#' 
-#' bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2, 
+#'
+#' bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2,
 #'                eta = 1, nthread = 2, nrounds = 2,objective = "binary:logistic")
-#' 
+#'
 #' (dt <- xgb.model.dt.tree(colnames(agaricus.train$data), bst))
-#' 
-#' # This bst model already has feature_names stored with it, so those would be used when 
+#'
+#' # This bst model already has feature_names stored with it, so those would be used when
 #' # feature_names is not set:
 #' (dt <- xgb.model.dt.tree(model = bst))
-#' 
+#'
 #' # How to match feature names of splits that are following a current 'Yes' branch:
-#' 
+#'
 #' merge(dt, dt[, .(ID, Y.Feature=Feature)], by.x='Yes', by.y='ID', all.x=TRUE)[order(Tree,Node)]
-#'  
+#'
 #' @export
 xgb.model.dt.tree <- function(feature_names = NULL, model = NULL, text = NULL,
                               trees = NULL, use_int_id = FALSE, ...){
   check.deprecation(...)
-  
+
   if (!inherits(model, "xgb.Booster") && !is.character(text)) {
     stop("Either 'model' must be an object of class xgb.Booster\n",
          "  or 'text' must be a character vector with the result of xgb.dump\n",
          "  (or NULL if 'model' was provided).")
   }
-  
+
   if (is.null(feature_names) && !is.null(model) && !is.null(model$feature_names))
     feature_names <- model$feature_names
-  
+
   if (!(is.null(feature_names) || is.character(feature_names))) {
     stop("feature_names: must be a character vector")
   }
-  
+
   if (!(is.null(trees) || is.numeric(trees))) {
     stop("trees: must be a vector of integers.")
   }
-  
+
   if (is.null(text)){
     text <- xgb.dump(model = model, with_stats = TRUE)
   }
-  
+
   if (length(text) < 2 ||
-      sum(stri_detect_regex(text, 'yes=(\\d+),no=(\\d+)')) < 1) {
+      sum(grepl('yes=(\\d+),no=(\\d+)', text)) < 1) {
     stop("Non-tree model detected! This function can only be used with tree models.")
   }
-  
-  position <- which(!is.na(stri_match_first_regex(text, "booster")))
-  
+
+  position <- which(grepl("booster", text, fixed = TRUE))
+
   add.tree.id <- function(node, tree) if (use_int_id) node else paste(tree, node, sep = "-")
-  
+
   anynumber_regex <- "[-+]?[0-9]*\\.?[0-9]+([eE][-+]?[0-9]+)?"
-  
+
   td <- data.table(t = text)
   td[position, Tree := 1L]
   td[, Tree := cumsum(ifelse(is.na(Tree), 0L, Tree)) - 1L]
-  
+
   if (is.null(trees)) {
     trees <- 0:max(td$Tree)
   } else {
     trees <- trees[trees >= 0 & trees <= max(td$Tree)]
   }
   td <- td[Tree %in% trees & !grepl('^booster', t)]
-  
-  td[, Node := stri_match_first_regex(t, "(\\d+):")[,2] %>% as.integer ]
+
+  td[, Node := as.integer(sub("^([0-9]+):.*", "\\1", t))]
   if (!use_int_id) td[, ID := add.tree.id(Node, Tree)]
-  td[, isLeaf := !is.na(stri_match_first_regex(t, "leaf"))]
+  td[, isLeaf := grepl("leaf", t, fixed = TRUE)]
 
   # parse branch lines
   branch_rx <- paste0("f(\\d+)<(", anynumber_regex, ")\\] yes=(\\d+),no=(\\d+),missing=(\\d+),",
                       "gain=(", anynumber_regex, "),cover=(", anynumber_regex, ")")
   branch_cols <- c("Feature", "Split", "Yes", "No", "Missing", "Quality", "Cover")
-  td[isLeaf == FALSE, 
+  td[isLeaf == FALSE,
      (branch_cols) := {
-      # skip some indices with spurious capture groups from anynumber_regex
-      xtr <- stri_match_first_regex(t, branch_rx)[, c(2,3,5,6,7,8,10), drop = FALSE]
-      xtr[, 3:5] <- add.tree.id(xtr[, 3:5], Tree)
-      lapply(seq_len(ncol(xtr)), function(i) xtr[,i])
+       matches <- regmatches(t, regexec(branch_rx, t))
+       # skip some indices with spurious capture groups from anynumber_regex
+       xtr <- do.call(rbind, matches)[, c(2, 3, 5, 6, 7, 8, 10), drop = FALSE]
+       xtr[, 3:5] <- add.tree.id(xtr[, 3:5], Tree)
+       as.data.table(xtr)
     }]
   # assign feature_names when available
   if (!is.null(feature_names)) {
     if (length(feature_names) <= max(as.numeric(td$Feature), na.rm = TRUE))
       stop("feature_names has less elements than there are features used in the model")
-    td[isLeaf == FALSE, Feature := feature_names[as.numeric(Feature) + 1] ]
+    td[isLeaf == FALSE, Feature := feature_names[as.numeric(Feature) + 1]]
   }
-  
+
   # parse leaf lines
   leaf_rx <- paste0("leaf=(", anynumber_regex, "),cover=(", anynumber_regex, ")")
   leaf_cols <- c("Feature", "Quality", "Cover")
   td[isLeaf == TRUE,
      (leaf_cols) := {
-      xtr <- stri_match_first_regex(t, leaf_rx)[, c(2,4)]
-      c("Leaf", lapply(seq_len(ncol(xtr)), function(i) xtr[,i]))
+       matches <- regmatches(t, regexec(leaf_rx, t))
+       xtr <- do.call(rbind, matches)[, c(2, 4)]
+       c("Leaf", as.data.table(xtr))
     }]
-  
+
   # convert some columns to numeric
   numeric_cols <- c("Split", "Quality", "Cover")
   td[, (numeric_cols) := lapply(.SD, as.numeric), .SDcols = numeric_cols]
@@ -146,14 +148,14 @@ xgb.model.dt.tree <- function(feature_names = NULL, model = NULL, text = NULL,
     int_cols <- c("Yes", "No", "Missing")
     td[, (int_cols) := lapply(.SD, as.integer), .SDcols = int_cols]
   }
-  
+
   td[, t := NULL]
   td[, isLeaf := NULL]
-  
+
   td[order(Tree, Node)]
 }
 
 # Avoid error messages during CRAN check.
 # The reason is that these variables are never declared
 # They are mainly column names inferred by Data.table...
-globalVariables(c("Tree", "Node", "ID", "Feature", "t", "isLeaf",".SD", ".SDcols"))
+globalVariables(c("Tree", "Node", "ID", "Feature", "t", "isLeaf", ".SD", ".SDcols"))

R-package/R/xgb.plot.deepness.R

@@ -2,48 +2,48 @@
 #'
 #' Visualizes distributions related to depth of tree leafs.
 #' \code{xgb.plot.deepness} uses base R graphics, while \code{xgb.ggplot.deepness} uses the ggplot backend.
-#' 
+#'
 #' @param model either an \code{xgb.Booster} model generated by the \code{xgb.train} function
 #'        or a data.table result of the \code{xgb.model.dt.tree} function.
-#' @param plot (base R barplot) whether a barplot should be produced. 
+#' @param plot (base R barplot) whether a barplot should be produced.
 #'        If FALSE, only a data.table is returned.
 #' @param which which distribution to plot (see details).
 #' @param ... other parameters passed to \code{barplot} or \code{plot}.
-#' 
+#'
 #' @details
-#' 
+#'
 #' When \code{which="2x1"}, two distributions with respect to the leaf depth
 #' are plotted on top of each other:
 #' \itemize{
 #'  \item the distribution of the number of leafs in a tree model at a certain depth;
-#'  \item the distribution of average weighted number of observations ("cover") 
+#'  \item the distribution of average weighted number of observations ("cover")
 #'        ending up in leafs at certain depth.
 #' }
-#' Those could be helpful in determining sensible ranges of the \code{max_depth} 
+#' Those could be helpful in determining sensible ranges of the \code{max_depth}
 #' and \code{min_child_weight} parameters.
-#' 
+#'
 #' When \code{which="max.depth"} or \code{which="med.depth"}, plots of either maximum or median depth
 #' per tree with respect to tree number are created. And \code{which="med.weight"} allows to see how
 #' a tree's median absolute leaf weight changes through the iterations.
 #'
 #' This function was inspired by the blog post
 #' \url{https://github.com/aysent/random-forest-leaf-visualization}.
-#' 
+#'
 #' @return
-#' 
+#'
 #' Other than producing plots (when \code{plot=TRUE}), the \code{xgb.plot.deepness} function
 #' silently returns a processed data.table where each row corresponds to a terminal leaf in a tree model,
 #' and contains information about leaf's depth, cover, and weight (which is used in calculating predictions).
-#' 
+#'
 #' The \code{xgb.ggplot.deepness} silently returns either a list of two ggplot graphs when \code{which="2x1"}
 #' or a single ggplot graph for the other \code{which} options.
 #'
-#' @seealso 
-#' 
+#' @seealso
+#'
 #' \code{\link{xgb.train}}, \code{\link{xgb.model.dt.tree}}.
-#' 
+#'
 #' @examples
-#' 
+#'
 #' data(agaricus.train, package='xgboost')
 #'
 #' # Change max_depth to a higher number to get a more significant result
@@ -53,16 +53,16 @@
 #'
 #' xgb.plot.deepness(bst)
 #' xgb.ggplot.deepness(bst)
-#' 
+#'
 #' xgb.plot.deepness(bst, which='max.depth', pch=16, col=rgb(0,0,1,0.3), cex=2)
-#' 
+#'
 #' xgb.plot.deepness(bst, which='med.weight', pch=16, col=rgb(0,0,1,0.3), cex=2)
 #'
 #' @rdname xgb.plot.deepness
 #' @export
 xgb.plot.deepness <- function(model = NULL, which = c("2x1", "max.depth", "med.depth", "med.weight"),
                               plot = TRUE, ...) {
-  
+
   if (!(inherits(model, "xgb.Booster") || is.data.table(model)))
     stop("model: Has to be either an xgb.Booster model generaged by the xgb.train function\n",
          "or a data.table result of the xgb.importance function")
@@ -71,32 +71,32 @@ xgb.plot.deepness <- function(model = NULL, which = c("2x1", "max.depth", "med.d
     stop("igraph package is required for plotting the graph deepness.", call. = FALSE)
 
   which <- match.arg(which)
-  
+
   dt_tree <- model
   if (inherits(model, "xgb.Booster"))
     dt_tree <- xgb.model.dt.tree(model = model)
-  
+
   if (!all(c("Feature", "Tree", "ID", "Yes", "No", "Cover") %in% colnames(dt_tree)))
     stop("Model tree columns are not as expected!\n",
          "  Note that this function works only for tree models.")
-  
+
   dt_depths <- merge(get.leaf.depth(dt_tree), dt_tree[, .(ID, Cover, Weight = Quality)], by = "ID")
   setkeyv(dt_depths, c("Tree", "ID"))
   # count by depth levels, and also calculate average cover at a depth
   dt_summaries <- dt_depths[, .(.N, Cover = mean(Cover)), Depth]
   setkey(dt_summaries, "Depth")
-  
+
   if (plot) {
     if (which == "2x1") {
       op <- par(no.readonly = TRUE)
-      par(mfrow = c(2,1),
-          oma = c(3,1,3,1) + 0.1,
-          mar = c(1,4,1,0) + 0.1)
+      par(mfrow = c(2, 1),
+          oma = c(3, 1, 3, 1) + 0.1,
+          mar = c(1, 4, 1, 0) + 0.1)
 
       dt_summaries[, barplot(N, border = NA, ylab = 'Number of leafs', ...)]
 
       dt_summaries[, barplot(Cover, border = NA, ylab = "Weighted cover", names.arg = Depth, ...)]
-    
+
       title("Model complexity", xlab = "Leaf depth", outer = TRUE, line = 1)
       par(op)
     } else if (which == "max.depth") {
@@ -123,14 +123,14 @@ get.leaf.depth <- function(dt_tree) {
       dt_tree[Feature != "Leaf", .(ID, To = No, Tree)]
     ))
   # whether "To" is a leaf:
-  dt_edges <- 
+  dt_edges <-
     merge(dt_edges,
           dt_tree[Feature == "Leaf", .(ID, Leaf = TRUE)],
           all.x = TRUE, by.x = "To", by.y = "ID")
   dt_edges[is.na(Leaf), Leaf := FALSE]
 
   dt_edges[, {
-    graph <- igraph::graph_from_data_frame(.SD[,.(ID, To)])
+    graph <- igraph::graph_from_data_frame(.SD[, .(ID, To)])
     # min(ID) in a tree is a root node
     paths_tmp <- igraph::shortest_paths(graph, from = min(ID), to = To[Leaf == TRUE])
     # list of paths to each leaf in a tree

R-package/R/xgb.plot.importance.R

@@ -92,28 +92,27 @@ xgb.plot.importance <- function(importance_matrix = NULL, top_n = NULL, measure
     importance_matrix <- head(importance_matrix, top_n)
   }
   if (rel_to_first) {
-    importance_matrix[, Importance := Importance/max(abs(Importance))]
+    importance_matrix[, Importance := Importance / max(abs(Importance))]
   }
   if (is.null(cex)) {
-    cex <- 2.5/log2(1 + nrow(importance_matrix))
+    cex <- 2.5 / log2(1 + nrow(importance_matrix))
   }
 
   if (plot) {
-    op <- par(no.readonly = TRUE)
-    mar <- op$mar
+    original_mar <- par()$mar
+
+    # reset margins so this function doesn't have side effects
+    on.exit({par(mar = original_mar)})
+
+    mar <- original_mar
     if (!is.null(left_margin))
       mar[2] <- left_margin
     par(mar = mar)
 
     # reverse the order of rows to have the highest ranked at the top
-    importance_matrix[nrow(importance_matrix):1,
+    importance_matrix[rev(seq_len(nrow(importance_matrix))),
                       barplot(Importance, horiz = TRUE, border = NA, cex.names = cex,
                               names.arg = Feature, las = 1, ...)]
-    grid(NULL, NA)
-    # redraw over the grid
-    importance_matrix[nrow(importance_matrix):1,
-                      barplot(Importance, horiz = TRUE, border = NA, add = TRUE)]
-    par(op)
   }
 
   invisible(importance_matrix)

R-package/R/xgb.plot.multi.trees.R

@@ -9,7 +9,7 @@
 #' @param plot_height height in pixels of the graph to produce
 #' @param render a logical flag for whether the graph should be rendered (see Value).
 #' @param ... currently not used
-#' 
+#'
 #' @details
 #'
 #' This function tries to capture the complexity of a gradient boosted tree model
@@ -67,58 +67,55 @@ xgb.plot.multi.trees <- function(model, feature_names = NULL, features_keep = 5,
 
   # first number of the path represents the tree, then the following numbers are related to the path to follow
   # root init
-  root.nodes <- tree.matrix[stri_detect_regex(ID, "\\d+-0"), ID]
+  root.nodes <- tree.matrix[Node == 0, ID]
   tree.matrix[ID %in% root.nodes, abs.node.position := root.nodes]
 
   precedent.nodes <- root.nodes
 
-  while(tree.matrix[,sum(is.na(abs.node.position))] > 0) {
+  while (tree.matrix[, sum(is.na(abs.node.position))] > 0) {
     yes.row.nodes <- tree.matrix[abs.node.position %in% precedent.nodes & !is.na(Yes)]
     no.row.nodes <- tree.matrix[abs.node.position %in% precedent.nodes & !is.na(No)]
     yes.nodes.abs.pos <- yes.row.nodes[, abs.node.position] %>% paste0("_0")
     no.nodes.abs.pos <- no.row.nodes[, abs.node.position] %>% paste0("_1")
-    
+
     tree.matrix[ID %in% yes.row.nodes[, Yes], abs.node.position := yes.nodes.abs.pos]
     tree.matrix[ID %in% no.row.nodes[, No], abs.node.position := no.nodes.abs.pos]
     precedent.nodes <- c(yes.nodes.abs.pos, no.nodes.abs.pos)
   }
-  
+
   tree.matrix[!is.na(Yes), Yes := paste0(abs.node.position, "_0")]
   tree.matrix[!is.na(No), No := paste0(abs.node.position, "_1")]
-  
-  remove.tree <- . %>% stri_replace_first_regex(pattern = "^\\d+-", replacement = "")
-  
-  tree.matrix[,`:=`(abs.node.position = remove.tree(abs.node.position),
-                    Yes = remove.tree(Yes),
-                    No = remove.tree(No))]
-  
+
+  for (nm in c("abs.node.position", "Yes", "No"))
+    data.table::set(tree.matrix, j = nm, value = sub("^\\d+-", "", tree.matrix[[nm]]))
+
   nodes.dt <- tree.matrix[
         , .(Quality = sum(Quality))
         , by = .(abs.node.position, Feature)
       ][, .(Text = paste0(Feature[1:min(length(Feature), features_keep)],
                           " (",
-                          format(Quality[1:min(length(Quality), features_keep)], digits=5),
+                          format(Quality[1:min(length(Quality), features_keep)], digits = 5),
                           ")") %>%
                    paste0(collapse = "\n"))
         , by = abs.node.position]
-  
+
   edges.dt <- tree.matrix[Feature != "Leaf", .(abs.node.position, Yes)] %>%
-    list(tree.matrix[Feature != "Leaf",.(abs.node.position, No)]) %>%
+    list(tree.matrix[Feature != "Leaf", .(abs.node.position, No)]) %>%
     rbindlist() %>%
     setnames(c("From", "To")) %>%
     .[, .N, .(From, To)] %>%
-    .[, N:=NULL]
-  
+    .[, N := NULL]
+
   nodes <- DiagrammeR::create_node_df(
     n = nrow(nodes.dt),
-    label = nodes.dt[,Text]
+    label = nodes.dt[, Text]
   )
-  
+
   edges <- DiagrammeR::create_edge_df(
-    from = match(edges.dt[,From], nodes.dt[,abs.node.position]),
-    to = match(edges.dt[,To], nodes.dt[,abs.node.position]),
+    from = match(edges.dt[, From], nodes.dt[, abs.node.position]),
+    to = match(edges.dt[, To], nodes.dt[, abs.node.position]),
     rel = "leading_to")
-  
+
   graph <- DiagrammeR::create_graph(
       nodes_df = nodes,
       edges_df = edges,

R-package/R/xgb.plot.shap.R

@@ -81,6 +81,7 @@
 #' xgb.plot.shap(agaricus.test$data, model = bst, features = "odor=none")
 #' contr <- predict(bst, agaricus.test$data, predcontrib = TRUE)
 #' xgb.plot.shap(agaricus.test$data, contr, model = bst, top_n = 12, n_col = 3)
+#' xgb.ggplot.shap.summary(agaricus.test$data, contr, model = bst, top_n = 12)  # Summary plot
 #'
 #' # multiclass example - plots for each class separately:
 #' nclass <- 3
@@ -99,6 +100,7 @@
 #'               n_col = 2, col = col, pch = 16, pch_NA = 17)
 #' xgb.plot.shap(x, model = mbst, trees = trees0 + 2, target_class = 2, top_n = 4,
 #'               n_col = 2, col = col, pch = 16, pch_NA = 17)
+#' xgb.ggplot.shap.summary(x, model = mbst, target_class = 0, top_n = 4)  # Summary plot
 #'
 #' @rdname xgb.plot.shap
 #' @export
@@ -109,69 +111,33 @@ xgb.plot.shap <- function(data, shap_contrib = NULL, features = NULL, top_n = 1,
                           plot_NA = TRUE, col_NA = rgb(0.7, 0, 1, 0.6), pch_NA = '.', pos_NA = 1.07,
                           plot_loess = TRUE, col_loess = 2, span_loess = 0.5,
                           which = c("1d", "2d"), plot = TRUE, ...) {
-
-  if (!is.matrix(data) && !inherits(data, "dgCMatrix"))
-    stop("data: must be either matrix or dgCMatrix")
-
-  if (is.null(shap_contrib) && (is.null(model) || !inherits(model, "xgb.Booster")))
-    stop("when shap_contrib is not provided, one must provide an xgb.Booster model")
-
-  if (is.null(features) && (is.null(model) || !inherits(model, "xgb.Booster")))
-    stop("when features are not provided, one must provide an xgb.Booster model to rank the features")
-
-  if (!is.null(shap_contrib) &&
-      (!is.matrix(shap_contrib) || nrow(shap_contrib) != nrow(data) || ncol(shap_contrib) != ncol(data) + 1))
-    stop("shap_contrib is not compatible with the provided data")
-
-  nsample <- if (is.null(subsample)) min(100000, nrow(data)) else as.integer(subsample * nrow(data))
-  idx <- sample(1:nrow(data), nsample)
-  data <- data[idx,]
-
-  if (is.null(shap_contrib)) {
-    shap_contrib <- predict(model, data, predcontrib = TRUE, approxcontrib = approxcontrib)
-  } else {
-    shap_contrib <- shap_contrib[idx,]
-  }
+  data_list <- xgb.shap.data(
+    data = data,
+    shap_contrib = shap_contrib,
+    features = features,
+    top_n = top_n,
+    model = model,
+    trees = trees,
+    target_class = target_class,
+    approxcontrib = approxcontrib,
+    subsample = subsample,
+    max_observations = 100000
+  )
+  data <- data_list[["data"]]
+  shap_contrib <- data_list[["shap_contrib"]]
+  features <- colnames(data)
 
   which <- match.arg(which)
   if (which == "2d")
     stop("2D plots are not implemented yet")
 
-  if (is.null(features)) {
-    imp <- xgb.importance(model = model, trees = trees)
-    top_n <- as.integer(top_n[1])
-    if (top_n < 1 && top_n > 100)
-      stop("top_n: must be an integer within [1, 100]")
-    features <- imp$Feature[1:min(top_n, NROW(imp))]
-  }
-
-  if (is.character(features)) {
-    if (is.null(colnames(data)))
-      stop("Either provide `data` with column names or provide `features` as column indices")
-    features <- match(features, colnames(data))
-  }
-
   if (n_col > length(features)) n_col <- length(features)
-
-  if (is.list(shap_contrib)) { # multiclass: either choose a class or merge
-    shap_contrib <- if (!is.null(target_class)) shap_contrib[[target_class + 1]]
-                    else Reduce("+", lapply(shap_contrib, abs))
-  }
-
-  shap_contrib <- shap_contrib[, features, drop = FALSE]
-  data <- data[, features, drop = FALSE]
-  cols <- colnames(data)
-  if (is.null(cols)) cols <- colnames(shap_contrib)
-  if (is.null(cols)) cols <- paste0('X', 1:ncol(data))
-  colnames(data) <- cols
-  colnames(shap_contrib) <- cols
-
   if (plot && which == "1d") {
     op <- par(mfrow = c(ceiling(length(features) / n_col), n_col),
-              oma = c(0,0,0,0) + 0.2,
-              mar = c(3.5,3.5,0,0) + 0.1,
+              oma = c(0, 0, 0, 0) + 0.2,
+              mar = c(3.5, 3.5, 0, 0) + 0.1,
               mgp = c(1.7, 0.6, 0))
-    for (f in cols) {
+    for (f in features) {
       ord <- order(data[, f])
       x <- data[, f][ord]
       y <- shap_contrib[, f][ord]
@@ -192,7 +158,7 @@ xgb.plot.shap <- function(data, shap_contrib = NULL, features = NULL, top_n = 1,
       grid()
       if (plot_loess) {
         # compress x to 3 digits, and mean-aggredate y
-        zz <- data.table(x = signif(x, 3), y)[, .(.N, y=mean(y)), x]
+        zz <- data.table(x = signif(x, 3), y)[, .(.N, y = mean(y)), x]
         if (nrow(zz) <= 5) {
           lines(zz$x, zz$y, col = col_loess)
         } else {
@@ -216,3 +182,108 @@ xgb.plot.shap <- function(data, shap_contrib = NULL, features = NULL, top_n = 1,
   }
   invisible(list(data = data, shap_contrib = shap_contrib))
 }
+
+#' SHAP contribution dependency summary plot
+#'
+#' Compare SHAP contributions of different features.
+#'
+#' A point plot (each point representing one sample from \code{data}) is
+#' produced for each feature, with the points plotted on the SHAP value axis.
+#' Each point (observation) is coloured based on its feature value. The plot
+#' hence allows us to see which features have a negative / positive contribution
+#' on the model prediction, and whether the contribution is different for larger
+#' or smaller values of the feature. We effectively try to replicate the
+#' \code{summary_plot} function from https://github.com/slundberg/shap.
+#'
+#' @inheritParams xgb.plot.shap
+#'
+#' @return A \code{ggplot2} object.
+#' @export
+#'
+#' @examples # See \code{\link{xgb.plot.shap}}.
+#' @seealso \code{\link{xgb.plot.shap}}, \code{\link{xgb.ggplot.shap.summary}},
+#'   \url{https://github.com/slundberg/shap}
+xgb.plot.shap.summary <- function(data, shap_contrib = NULL, features = NULL, top_n = 10, model = NULL,
+                                  trees = NULL, target_class = NULL, approxcontrib = FALSE, subsample = NULL) {
+  # Only ggplot implementation is available.
+  xgb.ggplot.shap.summary(data, shap_contrib, features, top_n, model, trees, target_class, approxcontrib, subsample)
+}
+
+#' Prepare data for SHAP plots. To be used in xgb.plot.shap, xgb.plot.shap.summary, etc.
+#' Internal utility function.
+#'
+#' @inheritParams xgb.plot.shap
+#' @keywords internal
+#'
+#' @return A list containing: 'data', a matrix containing sample observations
+#'   and their feature values; 'shap_contrib', a matrix containing the SHAP contribution
+#'   values for these observations.
+xgb.shap.data <- function(data, shap_contrib = NULL, features = NULL, top_n = 1, model = NULL,
+                          trees = NULL, target_class = NULL, approxcontrib = FALSE,
+                          subsample = NULL, max_observations = 100000) {
+  if (!is.matrix(data) && !inherits(data, "dgCMatrix"))
+    stop("data: must be either matrix or dgCMatrix")
+
+  if (is.null(shap_contrib) && (is.null(model) || !inherits(model, "xgb.Booster")))
+    stop("when shap_contrib is not provided, one must provide an xgb.Booster model")
+
+  if (is.null(features) && (is.null(model) || !inherits(model, "xgb.Booster")))
+    stop("when features are not provided, one must provide an xgb.Booster model to rank the features")
+
+  if (!is.null(shap_contrib) &&
+      (!is.matrix(shap_contrib) || nrow(shap_contrib) != nrow(data) || ncol(shap_contrib) != ncol(data) + 1))
+    stop("shap_contrib is not compatible with the provided data")
+
+  if (is.character(features) && is.null(colnames(data)))
+    stop("either provide `data` with column names or provide `features` as column indices")
+
+  if (is.null(model$feature_names) && model$nfeatures != ncol(data))
+    stop("if model has no feature_names, columns in `data` must match features in model")
+
+  if (!is.null(subsample)) {
+    idx <- sample(x = seq_len(nrow(data)), size = as.integer(subsample * nrow(data)), replace = FALSE)
+  } else {
+    idx <- seq_len(min(nrow(data), max_observations))
+  }
+  data <- data[idx, ]
+  if (is.null(colnames(data))) {
+    colnames(data) <- paste0("X", seq_len(ncol(data)))
+  }
+
+  if (!is.null(shap_contrib)) {
+    if (is.list(shap_contrib)) { # multiclass: either choose a class or merge
+      shap_contrib <- if (!is.null(target_class)) shap_contrib[[target_class + 1]] else Reduce("+", lapply(shap_contrib, abs))
+    }
+    shap_contrib <- shap_contrib[idx, ]
+    if (is.null(colnames(shap_contrib))) {
+      colnames(shap_contrib) <- paste0("X", seq_len(ncol(data)))
+    }
+  } else {
+    shap_contrib <- predict(model, newdata = data, predcontrib = TRUE, approxcontrib = approxcontrib)
+    if (is.list(shap_contrib)) { # multiclass: either choose a class or merge
+      shap_contrib <- if (!is.null(target_class)) shap_contrib[[target_class + 1]] else Reduce("+", lapply(shap_contrib, abs))
+    }
+  }
+
+  if (is.null(features)) {
+    if (!is.null(model$feature_names)) {
+      imp <- xgb.importance(model = model, trees = trees)
+    } else {
+      imp <- xgb.importance(model = model, trees = trees, feature_names = colnames(data))
+    }
+    top_n <- top_n[1]
+    if (top_n < 1 | top_n > 100) stop("top_n: must be an integer within [1, 100]")
+    features <- imp$Feature[1:min(top_n, NROW(imp))]
+  }
+  if (is.character(features)) {
+    features <- match(features, colnames(data))
+  }
+
+  shap_contrib <- shap_contrib[, features, drop = FALSE]
+  data <- data[, features, drop = FALSE]
+
+  list(
+    data = data,
+    shap_contrib = shap_contrib
+  )
+}

R-package/R/xgb.plot.tree.R

@@ -1,7 +1,7 @@
 #' Plot a boosted tree model
-#' 
-#' Read a tree model text dump and plot the model. 
-#' 
+#'
+#' Read a tree model text dump and plot the model.
+#'
 #' @param feature_names names of each feature as a \code{character} vector.
 #' @param model produced by the \code{xgb.train} function.
 #' @param trees an integer vector of tree indices that should be visualized.
@@ -14,10 +14,10 @@
 #' @param show_node_id a logical flag for whether to show node id's in the graph.
 #' @param ... currently not used.
 #'
-#' @details 
-#' 
+#' @details
+#'
 #' The content of each node is organised that way:
-#' 
+#'
 #' \itemize{
 #'  \item Feature name.
 #'  \item \code{Cover}: The sum of second order gradient of training data classified to the leaf.
@@ -27,21 +27,21 @@
 #'  \item \code{Gain} (for split nodes): the information gain metric of a split
 #'        (corresponds to the importance of the node in the model).
 #'  \item \code{Value} (for leafs): the margin value that the leaf may contribute to prediction.
-#' } 
+#' }
 #' The tree root nodes also indicate the Tree index (0-based).
-#' 
+#'
 #' The "Yes" branches are marked by the "< split_value" label.
 #' The branches that also used for missing values are marked as bold
 #' (as in "carrying extra capacity").
-#' 
+#'
 #' This function uses \href{http://www.graphviz.org/}{GraphViz} as a backend of DiagrammeR.
-#'  
+#'
 #' @return
-#' 
+#'
 #' When \code{render = TRUE}:
 #' returns a rendered graph object which is an \code{htmlwidget} of class \code{grViz}.
 #' Similar to ggplot objects, it needs to be printed to see it when not running from command line.
-#' 
+#'
 #' When \code{render = FALSE}:
 #' silently returns a graph object which is of DiagrammeR's class \code{dgr_graph}.
 #' This could be useful if one wants to modify some of the graph attributes
@@ -49,23 +49,23 @@
 #'
 #' @examples
 #' data(agaricus.train, package='xgboost')
-#' 
+#'
 #' bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 3,
 #'                eta = 1, nthread = 2, nrounds = 2,objective = "binary:logistic")
 #' # plot all the trees
 #' xgb.plot.tree(model = bst)
 #' # plot only the first tree and display the node ID:
 #' xgb.plot.tree(model = bst, trees = 0, show_node_id = TRUE)
-#' 
+#'
 #' \dontrun{
-#' # Below is an example of how to save this plot to a file. 
+#' # Below is an example of how to save this plot to a file.
 #' # Note that for `export_graph` to work, the DiagrammeRsvg and rsvg packages must also be installed.
 #' library(DiagrammeR)
 #' gr <- xgb.plot.tree(model=bst, trees=0:1, render=FALSE)
 #' export_graph(gr, 'tree.pdf', width=1500, height=1900)
 #' export_graph(gr, 'tree.png', width=1500, height=1900)
 #' }
-#' 
+#'
 #' @export
 xgb.plot.tree <- function(feature_names = NULL, model = NULL, trees = NULL, plot_width = NULL, plot_height = NULL,
                           render = TRUE, show_node_id = FALSE, ...){
@@ -77,18 +77,18 @@ xgb.plot.tree <- function(feature_names = NULL, model = NULL, trees = NULL, plot
   if (!requireNamespace("DiagrammeR", quietly = TRUE)) {
     stop("DiagrammeR package is required for xgb.plot.tree", call. = FALSE)
   }
-  
+
   dt <- xgb.model.dt.tree(feature_names = feature_names, model = model, trees = trees)
 
-  dt[, label:= paste0(Feature, "\nCover: ", Cover, ifelse(Feature == "Leaf", "\nValue: ", "\nGain: "), Quality)]
+  dt[, label := paste0(Feature, "\nCover: ", Cover, ifelse(Feature == "Leaf", "\nValue: ", "\nGain: "), Quality)]
   if (show_node_id)
     dt[, label := paste0(ID, ": ", label)]
   dt[Node == 0, label := paste0("Tree ", Tree, "\n", label)]
-  dt[, shape:= "rectangle"][Feature == "Leaf", shape:= "oval"]
-  dt[, filledcolor:= "Beige"][Feature == "Leaf", filledcolor:= "Khaki"]
+  dt[, shape := "rectangle"][Feature == "Leaf", shape := "oval"]
+  dt[, filledcolor := "Beige"][Feature == "Leaf", filledcolor := "Khaki"]
   # in order to draw the first tree on top:
   dt <- dt[order(-Tree)]
-  
+
   nodes <- DiagrammeR::create_node_df(
     n         = nrow(dt),
     ID        = dt$ID,
@@ -97,7 +97,7 @@ xgb.plot.tree <- function(feature_names = NULL, model = NULL, trees = NULL, plot
     shape     = dt$shape,
     data      = dt$Feature,
     fontcolor = "black")
-  
+
   edges <- DiagrammeR::create_edge_df(
     from  = match(dt[Feature != "Leaf", c(ID)] %>% rep(2), dt$ID),
     to    = match(dt[Feature != "Leaf", c(Yes, No)], dt$ID),
@@ -126,9 +126,9 @@ xgb.plot.tree <- function(feature_names = NULL, model = NULL, trees = NULL, plot
       attr_type = "edge",
       attr  = c("color", "arrowsize", "arrowhead", "fontname"),
       value = c("DimGray", "1.5", "vee", "Helvetica"))
-  
+
   if (!render) return(invisible(graph))
-  
+
   DiagrammeR::render_graph(graph, width = plot_width, height = plot_height)
 }
 

R-package/R/xgb.save.R

@@ -1,29 +1,33 @@
 #' Save xgboost model to binary file
-#' 
+#'
 #' Save xgboost model to a file in binary format.
-#' 
+#'
 #' @param model model object of \code{xgb.Booster} class.
 #' @param fname name of the file to write.
-#' 
-#' @details 
-#' This methods allows to save a model in an xgboost-internal binary format which is universal 
+#'
+#' @details
+#' This methods allows to save a model in an xgboost-internal binary format which is universal
 #' among the various xgboost interfaces. In R, the saved model file could be read-in later
-#' using either the \code{\link{xgb.load}} function or the \code{xgb_model} parameter 
+#' using either the \code{\link{xgb.load}} function or the \code{xgb_model} parameter
 #' of \code{\link{xgb.train}}.
-#' 
-#' Note: a model can also be saved as an R-object (e.g., by using \code{\link[base]{readRDS}} 
-#' or \code{\link[base]{save}}). However, it would then only be compatible with R, and 
-#' corresponding R-methods would need to be used to load it.
-#' 
-#' @seealso 
-#' \code{\link{xgb.load}}, \code{\link{xgb.Booster.complete}}. 
-#' 
+#'
+#' Note: a model can also be saved as an R-object (e.g., by using \code{\link[base]{readRDS}}
+#' or \code{\link[base]{save}}). However, it would then only be compatible with R, and
+#' corresponding R-methods would need to be used to load it. Moreover, persisting the model with
+#' \code{\link[base]{readRDS}} or \code{\link[base]{save}}) will cause compatibility problems in
+#' future versions of XGBoost. Consult \code{\link{a-compatibility-note-for-saveRDS-save}} to learn
+#' how to persist models in a future-proof way, i.e. to make the model accessible in future
+#' releases of XGBoost.
+#'
+#' @seealso
+#' \code{\link{xgb.load}}, \code{\link{xgb.Booster.complete}}.
+#'
 #' @examples
 #' data(agaricus.train, package='xgboost')
 #' data(agaricus.test, package='xgboost')
 #' train <- agaricus.train
 #' test <- agaricus.test
-#' bst <- xgboost(data = train$data, label = train$label, max_depth = 2, 
+#' bst <- xgboost(data = train$data, label = train$label, max_depth = 2,
 #'                eta = 1, nthread = 2, nrounds = 2,objective = "binary:logistic")
 #' xgb.save(bst, 'xgb.model')
 #' bst <- xgb.load('xgb.model')
@@ -38,6 +42,7 @@ xgb.save <- function(model, fname) {
          if (inherits(model, "xgb.DMatrix")) " Use xgb.DMatrix.save to save an xgb.DMatrix object." else "")
   }
   model <- xgb.Booster.complete(model, saveraw = FALSE)
+  fname <- path.expand(fname)
   .Call(XGBoosterSaveModel_R, model$handle, fname[1])
   return(TRUE)
 }

R-package/R/xgb.train.R

@@ -3,9 +3,9 @@
 #' \code{xgb.train} is an advanced interface for training an xgboost model.
 #' The \code{xgboost} function is a simpler wrapper for \code{xgb.train}.
 #'
-#' @param params the list of parameters.
-#'        The complete list of parameters is available at \url{http://xgboost.readthedocs.io/en/latest/parameter.html}.
-#'        Below is a shorter summary:
+#' @param params the list of parameters. The complete list of parameters is
+#'   available in the \href{http://xgboost.readthedocs.io/en/latest/parameter.html}{online documentation}. Below
+#'   is a shorter summary:
 #'
 #' 1. General Parameters
 #'
@@ -15,7 +15,7 @@
 #'
 #' 2. Booster Parameters
 #'
-#' 2.1. Parameter for Tree Booster
+#' 2.1. Parameters for Tree Booster
 #'
 #' \itemize{
 #'   \item \code{eta} control the learning rate: scale the contribution of each tree by a factor of \code{0 < eta < 1} when it is added to the current approximation. Used to prevent overfitting by making the boosting process more conservative. Lower value for \code{eta} implies larger value for \code{nrounds}: low \code{eta} value means model more robust to overfitting but slower to compute. Default: 0.3
@@ -24,12 +24,14 @@
 #'   \item \code{min_child_weight} minimum sum of instance weight (hessian) needed in a child. If the tree partition step results in a leaf node with the sum of instance weight less than min_child_weight, then the building process will give up further partitioning. In linear regression mode, this simply corresponds to minimum number of instances needed to be in each node. The larger, the more conservative the algorithm will be. Default: 1
 #'   \item \code{subsample} subsample ratio of the training instance. Setting it to 0.5 means that xgboost randomly collected half of the data instances to grow trees and this will prevent overfitting. It makes computation shorter (because less data to analyse). It is advised to use this parameter with \code{eta} and increase \code{nrounds}. Default: 1
 #'   \item \code{colsample_bytree} subsample ratio of columns when constructing each tree. Default: 1
+#'   \item \code{lambda} L2 regularization term on weights. Default: 1
+#'   \item \code{alpha} L1 regularization term on weights. (there is no L1 reg on bias because it is not important). Default: 0
 #'   \item \code{num_parallel_tree} Experimental parameter. number of trees to grow per round. Useful to test Random Forest through Xgboost (set \code{colsample_bytree < 1}, \code{subsample  < 1}  and \code{round = 1}) accordingly. Default: 1
 #'   \item \code{monotone_constraints} A numerical vector consists of \code{1}, \code{0} and \code{-1} with its length equals to the number of features in the training data. \code{1} is increasing, \code{-1} is decreasing and \code{0} is no constraint.
 #'   \item \code{interaction_constraints} A list of vectors specifying feature indices of permitted interactions. Each item of the list represents one permitted interaction where specified features are allowed to interact with each other. Feature index values should start from \code{0} (\code{0} references the first column).  Leave argument unspecified for no interaction constraints.
 #' }
 #'
-#' 2.2. Parameter for Linear Booster
+#' 2.2. Parameters for Linear Booster
 #'
 #' \itemize{
 #'   \item \code{lambda} L2 regularization term on weights. Default: 0
@@ -43,13 +45,23 @@
 #' \item \code{objective} specify the learning task and the corresponding learning objective, users can pass a self-defined function to it. The default objective options are below:
 #'   \itemize{
 #'     \item \code{reg:squarederror} Regression with squared loss (Default).
+#'     \item \code{reg:squaredlogerror}: regression with squared log loss \eqn{1/2 * (log(pred + 1) - log(label + 1))^2}. All inputs are required to be greater than -1. Also, see metric rmsle for possible issue with this objective.
 #'     \item \code{reg:logistic} logistic regression.
+#'     \item \code{reg:pseudohubererror}: regression with Pseudo Huber loss, a twice differentiable alternative to absolute loss.
 #'     \item \code{binary:logistic} logistic regression for binary classification. Output probability.
 #'     \item \code{binary:logitraw} logistic regression for binary classification, output score before logistic transformation.
-#'     \item \code{num_class} set the number of classes. To use only with multiclass objectives.
+#'     \item \code{binary:hinge}: hinge loss for binary classification. This makes predictions of 0 or 1, rather than producing probabilities.
+#'     \item \code{count:poisson}: poisson regression for count data, output mean of poisson distribution. \code{max_delta_step} is set to 0.7 by default in poisson regression (used to safeguard optimization).
+#'     \item \code{survival:cox}: Cox regression for right censored survival time data (negative values are considered right censored). Note that predictions are returned on the hazard ratio scale (i.e., as HR = exp(marginal_prediction) in the proportional hazard function \code{h(t) = h0(t) * HR)}.
+#'     \item \code{survival:aft}: Accelerated failure time model for censored survival time data. See \href{https://xgboost.readthedocs.io/en/latest/tutorials/aft_survival_analysis.html}{Survival Analysis with Accelerated Failure Time} for details.
+#'     \item \code{aft_loss_distribution}: Probabilty Density Function used by \code{survival:aft} and \code{aft-nloglik} metric.
 #'     \item \code{multi:softmax} set xgboost to do multiclass classification using the softmax objective. Class is represented by a number and should be from 0 to \code{num_class - 1}.
 #'     \item \code{multi:softprob} same as softmax, but prediction outputs a vector of ndata * nclass elements, which can be further reshaped to ndata, nclass matrix. The result contains predicted probabilities of each data point belonging to each class.
 #'     \item \code{rank:pairwise} set xgboost to do ranking task by minimizing the pairwise loss.
+#'     \item \code{rank:ndcg}: Use LambdaMART to perform list-wise ranking where \href{https://en.wikipedia.org/wiki/Discounted_cumulative_gain}{Normalized Discounted Cumulative Gain (NDCG)} is maximized.
+#'     \item \code{rank:map}: Use LambdaMART to perform list-wise ranking where \href{https://en.wikipedia.org/wiki/Evaluation_measures_(information_retrieval)#Mean_average_precision}{Mean Average Precision (MAP)} is maximized.
+#'     \item \code{reg:gamma}: gamma regression with log-link. Output is a mean of gamma distribution. It might be useful, e.g., for modeling insurance claims severity, or for any outcome that might be \href{https://en.wikipedia.org/wiki/Gamma_distribution#Applications}{gamma-distributed}.
+#'     \item \code{reg:tweedie}: Tweedie regression with log-link. It might be useful, e.g., for modeling total loss in insurance, or for any outcome that might be \href{https://en.wikipedia.org/wiki/Tweedie_distribution#Applications}{Tweedie-distributed}.
 #'   }
 #'   \item \code{base_score} the initial prediction score of all instances, global bias. Default: 0.5
 #'   \item \code{eval_metric} evaluation metrics for validation data. Users can pass a self-defined function to it. Default: metric will be assigned according to objective(rmse for regression, and error for classification, mean average precision for ranking). List is provided in detail section.
@@ -120,16 +132,18 @@
 #' Note that when using a customized metric, only this single metric can be used.
 #' The following is the list of built-in metrics for which Xgboost provides optimized implementation:
 #'   \itemize{
-#'      \item \code{rmse} root mean square error. \url{http://en.wikipedia.org/wiki/Root_mean_square_error}
-#'      \item \code{logloss} negative log-likelihood. \url{http://en.wikipedia.org/wiki/Log-likelihood}
-#'      \item \code{mlogloss} multiclass logloss. \url{http://wiki.fast.ai/index.php/Log_Loss}
+#'      \item \code{rmse} root mean square error. \url{https://en.wikipedia.org/wiki/Root_mean_square_error}
+#'      \item \code{logloss} negative log-likelihood. \url{https://en.wikipedia.org/wiki/Log-likelihood}
+#'      \item \code{mlogloss} multiclass logloss. \url{https://scikit-learn.org/stable/modules/generated/sklearn.metrics.log_loss.html}
 #'      \item \code{error} Binary classification error rate. It is calculated as \code{(# wrong cases) / (# all cases)}.
 #'            By default, it uses the 0.5 threshold for predicted values to define negative and positive instances.
 #'            Different threshold (e.g., 0.) could be specified as "error@0."
 #'      \item \code{merror} Multiclass classification error rate. It is calculated as \code{(# wrong cases) / (# all cases)}.
-#'      \item \code{auc} Area under the curve. \url{http://en.wikipedia.org/wiki/Receiver_operating_characteristic#'Area_under_curve} for ranking evaluation.
+#'      \item \code{mae} Mean absolute error
+#'      \item \code{mape} Mean absolute percentage error
+#'      \item \code{auc} Area under the curve. \url{https://en.wikipedia.org/wiki/Receiver_operating_characteristic#'Area_under_curve} for ranking evaluation.
 #'      \item \code{aucpr} Area under the PR curve. \url{https://en.wikipedia.org/wiki/Precision_and_recall} for ranking evaluation.
-#'      \item \code{ndcg} Normalized Discounted Cumulative Gain (for ranking task). \url{http://en.wikipedia.org/wiki/NDCG}
+#'      \item \code{ndcg} Normalized Discounted Cumulative Gain (for ranking task). \url{https://en.wikipedia.org/wiki/NDCG}
 #'   }
 #'
 #' The following callbacks are automatically created when certain parameters are set:
@@ -181,8 +195,8 @@
 #' data(agaricus.train, package='xgboost')
 #' data(agaricus.test, package='xgboost')
 #'
-#' dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
-#' dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
+#' dtrain <- with(agaricus.train, xgb.DMatrix(data, label = label))
+#' dtest <- with(agaricus.test, xgb.DMatrix(data, label = label))
 #' watchlist <- list(train = dtrain, eval = dtest)
 #'
 #' ## A simple xgb.train example:
@@ -268,7 +282,7 @@ xgb.train <- function(params = list(), data, nrounds, watchlist = list(),
 
   # evaluation printing callback
   params <- c(params)
-  print_every_n <- max( as.integer(print_every_n), 1L)
+  print_every_n <- max(as.integer(print_every_n), 1L)
   if (!has.callbacks(callbacks, 'cb.print.evaluation') &&
       verbose) {
     callbacks <- add.cb(callbacks, cb.print.evaluation(print_every_n))
@@ -318,12 +332,9 @@ xgb.train <- function(params = list(), data, nrounds, watchlist = list(),
       niter_init <- xgb.ntree(bst) %/% (num_parallel_tree * num_class)
     }
   }
-  if(is_update && nrounds > niter_init)
+  if (is_update && nrounds > niter_init)
     stop("nrounds cannot be larger than ", niter_init, " (nrounds of xgb_model)")
 
-  # TODO: distributed code
-  rank <- 0
-
   niter_skip <- ifelse(is_update, 0, niter_init)
   begin_iteration <- niter_skip + 1
   end_iteration <- niter_skip + nrounds
@@ -335,7 +346,6 @@ xgb.train <- function(params = list(), data, nrounds, watchlist = list(),
 
     xgb.iter.update(bst$handle, dtrain, iteration - 1, obj)
 
-    bst_evaluation <- numeric(0)
     if (length(watchlist) > 0)
       bst_evaluation <- xgb.iter.eval(bst$handle, watchlist, iteration - 1, feval)
 
@@ -350,7 +360,7 @@ xgb.train <- function(params = list(), data, nrounds, watchlist = list(),
   bst <- xgb.Booster.complete(bst, saveraw = TRUE)
 
   # store the total number of boosting iterations
-  bst$niter = end_iteration
+  bst$niter <- end_iteration
 
   # store the evaluation results
   if (length(evaluation_log) > 0 &&

R-package/R/xgb.unserialize.R

@@ -6,7 +6,26 @@
 xgb.unserialize <- function(buffer) {
   cachelist <- list()
   handle <- .Call(XGBoosterCreate_R, cachelist)
-  .Call(XGBoosterUnserializeFromBuffer_R, handle, buffer)
+  tryCatch(
+    .Call(XGBoosterUnserializeFromBuffer_R, handle, buffer),
+    error = function(e) {
+      error_msg <- conditionMessage(e)
+      m <- regexec("(src[\\\\/]learner.cc:[0-9]+): Check failed: (header == serialisation_header_)",
+                   error_msg, perl = TRUE)
+      groups <- regmatches(error_msg, m)[[1]]
+      if (length(groups) == 3) {
+        warning(paste("The model had been generated by XGBoost version 1.0.0 or earlier and was ",
+                      "loaded from a RDS file. We strongly ADVISE AGAINST using saveRDS() ",
+                      "function, to ensure that your model can be read in current and upcoming ",
+                      "XGBoost releases. Please use xgb.save() instead to preserve models for the ",
+                      "long term. For more details and explanation, see ",
+                      "https://xgboost.readthedocs.io/en/latest/tutorials/saving_model.html",
+                      sep = ""))
+        .Call(XGBoosterLoadModelFromRaw_R, handle, buffer)
+      } else {
+        stop(e)
+      }
+    })
   class(handle) <- "xgb.Booster.handle"
   return (handle)
 }

R-package/R/xgboost.R

@@ -91,11 +91,8 @@ NULL
 #' @importFrom data.table setkeyv
 #' @importFrom data.table setnames
 #' @importFrom magrittr %>%
-#' @importFrom stringi stri_detect_regex
-#' @importFrom stringi stri_match_first_regex
-#' @importFrom stringi stri_replace_first_regex
-#' @importFrom stringi stri_replace_all_regex
-#' @importFrom stringi stri_split_regex
+#' @importFrom jsonlite fromJSON
+#' @importFrom jsonlite toJSON
 #' @importFrom utils object.size str tail
 #' @importFrom stats predict
 #' @importFrom stats median

R-package/configure

@@ -613,6 +613,7 @@ infodir
 docdir
 oldincludedir
 includedir
+runstatedir
 localstatedir
 sharedstatedir
 sysconfdir
@@ -682,6 +683,7 @@ datadir='${datarootdir}'
 sysconfdir='${prefix}/etc'
 sharedstatedir='${prefix}/com'
 localstatedir='${prefix}/var'
+runstatedir='${localstatedir}/run'
 includedir='${prefix}/include'
 oldincludedir='/usr/include'
 docdir='${datarootdir}/doc/${PACKAGE_TARNAME}'
@@ -934,6 +936,15 @@ do
   | -silent | --silent | --silen | --sile | --sil)
     silent=yes ;;
 
+  -runstatedir | --runstatedir | --runstatedi | --runstated \
+  | --runstate | --runstat | --runsta | --runst | --runs \
+  | --run | --ru | --r)
+    ac_prev=runstatedir ;;
+  -runstatedir=* | --runstatedir=* | --runstatedi=* | --runstated=* \
+  | --runstate=* | --runstat=* | --runsta=* | --runst=* | --runs=* \
+  | --run=* | --ru=* | --r=*)
+    runstatedir=$ac_optarg ;;
+
   -sbindir | --sbindir | --sbindi | --sbind | --sbin | --sbi | --sb)
     ac_prev=sbindir ;;
   -sbindir=* | --sbindir=* | --sbindi=* | --sbind=* | --sbin=* \
@@ -1071,7 +1082,7 @@ fi
 for ac_var in	exec_prefix prefix bindir sbindir libexecdir datarootdir \
 		datadir sysconfdir sharedstatedir localstatedir includedir \
 		oldincludedir docdir infodir htmldir dvidir pdfdir psdir \
-		libdir localedir mandir
+		libdir localedir mandir runstatedir
 do
   eval ac_val=\$$ac_var
   # Remove trailing slashes.
@@ -1224,6 +1235,7 @@ Fine tuning of the installation directories:
   --sysconfdir=DIR        read-only single-machine data [PREFIX/etc]
   --sharedstatedir=DIR    modifiable architecture-independent data [PREFIX/com]
   --localstatedir=DIR     modifiable single-machine data [PREFIX/var]
+  --runstatedir=DIR       modifiable per-process data [LOCALSTATEDIR/run]
   --libdir=DIR            object code libraries [EPREFIX/lib]
   --includedir=DIR        C header files [PREFIX/include]
   --oldincludedir=DIR     C header files for non-gcc [/usr/include]
@@ -2698,7 +2710,7 @@ fi
 if test `uname -s` = "Darwin"
 then
   OPENMP_CXXFLAGS='-Xclang -fopenmp'
-  OPENMP_LIB='/usr/local/lib/libomp.dylib'
+  OPENMP_LIB='-lomp'
   ac_pkg_openmp=no
   { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether OpenMP will work in a package" >&5
 $as_echo_n "checking whether OpenMP will work in a package... " >&6; }
@@ -2713,14 +2725,14 @@ main ()
   return 0;
 }
 _ACEOF
-  ${CC} -o conftest conftest.c /usr/local/lib/libomp.dylib -Xclang -fopenmp 2>/dev/null && ./conftest && ac_pkg_openmp=yes
+  ${CC} -o conftest conftest.c ${OPENMP_LIB} ${OPENMP_CXXFLAGS} 2>/dev/null && ./conftest && ac_pkg_openmp=yes
   { $as_echo "$as_me:${as_lineno-$LINENO}: result: ${ac_pkg_openmp}" >&5
 $as_echo "${ac_pkg_openmp}" >&6; }
   if test "${ac_pkg_openmp}" = no; then
     OPENMP_CXXFLAGS=''
     OPENMP_LIB=''
     echo '*****************************************************************************************'
-    echo 'WARNING: OpenMP is unavailable on this Mac OSX system. Training speed may be suboptimal.'
+    echo '         OpenMP is unavailable on this Mac OSX system. Training speed may be suboptimal.'
     echo '         To use all CPU cores for training jobs, you should install OpenMP by running\n'
     echo '             brew install libomp'
     echo '*****************************************************************************************'

R-package/configure.ac

@@ -1,6 +1,6 @@
 ### configure.ac					-*- Autoconf -*-
 
-AC_PREREQ(2.62)
+AC_PREREQ(2.69)
 
 AC_INIT([xgboost],[0.6-3],[],[xgboost],[])
 
@@ -29,17 +29,17 @@ fi
 if test `uname -s` = "Darwin"
 then
   OPENMP_CXXFLAGS='-Xclang -fopenmp'
-  OPENMP_LIB='/usr/local/lib/libomp.dylib'
+  OPENMP_LIB='-lomp'
   ac_pkg_openmp=no
   AC_MSG_CHECKING([whether OpenMP will work in a package])
   AC_LANG_CONFTEST([AC_LANG_PROGRAM([[#include <omp.h>]], [[ return (omp_get_max_threads() <= 1); ]])])
-  ${CC} -o conftest conftest.c /usr/local/lib/libomp.dylib -Xclang -fopenmp 2>/dev/null && ./conftest && ac_pkg_openmp=yes
+  ${CC} -o conftest conftest.c ${OPENMP_LIB} ${OPENMP_CXXFLAGS} 2>/dev/null && ./conftest && ac_pkg_openmp=yes
   AC_MSG_RESULT([${ac_pkg_openmp}])
   if test "${ac_pkg_openmp}" = no; then
     OPENMP_CXXFLAGS=''
     OPENMP_LIB=''
     echo '*****************************************************************************************'
-    echo 'WARNING: OpenMP is unavailable on this Mac OSX system. Training speed may be suboptimal.'
+    echo '         OpenMP is unavailable on this Mac OSX system. Training speed may be suboptimal.'
     echo '         To use all CPU cores for training jobs, you should install OpenMP by running\n'
     echo '             brew install libomp'
     echo '*****************************************************************************************'
@@ -52,4 +52,3 @@ AC_SUBST(ENDIAN_FLAG)
 AC_SUBST(BACKTRACE_LIB)
 AC_CONFIG_FILES([src/Makevars])
 AC_OUTPUT
-

R-package/demo/README.md

@@ -17,4 +17,4 @@ Benchmarks
 Notes
 ====
 * Contribution of examples, benchmarks is more than welcomed!
-* If you like to share how you use xgboost to solve your problem, send a pull request:)
+* If you like to share how you use xgboost to solve your problem, send a pull request :)

R-package/demo/basic_walkthrough.R

@@ -3,8 +3,8 @@ require(methods)
 
 # we load in the agaricus dataset
 # In this example, we are aiming to predict whether a mushroom is edible
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 train <- agaricus.train
 test <- agaricus.test
 # the loaded data is stored in sparseMatrix, and label is a numeric vector in {0,1}
@@ -26,7 +26,7 @@ bst <- xgboost(data = as.matrix(train$data), label = train$label, max_depth = 2,
 # you can also put in xgb.DMatrix object, which stores label, data and other meta datas needed for advanced features
 print("Training xgboost with xgb.DMatrix")
 dtrain <- xgb.DMatrix(data = train$data, label = train$label)
-bst <- xgboost(data = dtrain, max_depth = 2, eta = 1, nrounds = 2, nthread = 2, 
+bst <- xgboost(data = dtrain, max_depth = 2, eta = 1, nrounds = 2, nthread = 2,
                objective = "binary:logistic")
 
 # Verbose = 0,1,2
@@ -46,7 +46,7 @@ bst <- xgboost(data = dtrain, max_depth = 2, eta = 1, nrounds = 2,
 
 #--------------------basic prediction using xgboost--------------
 # you can do prediction using the following line
-# you can put in Matrix, sparseMatrix, or xgb.DMatrix 
+# you can put in Matrix, sparseMatrix, or xgb.DMatrix
 pred <- predict(bst, test$data)
 err <- mean(as.numeric(pred > 0.5) != test$label)
 print(paste("test-error=", err))
@@ -58,31 +58,31 @@ xgb.save(bst, "xgboost.model")
 bst2 <- xgb.load("xgboost.model")
 pred2 <- predict(bst2, test$data)
 # pred2 should be identical to pred
-print(paste("sum(abs(pred2-pred))=", sum(abs(pred2-pred))))
+print(paste("sum(abs(pred2-pred))=", sum(abs(pred2 - pred))))
 
 # save model to R's raw vector
-raw = xgb.save.raw(bst)
+raw <- xgb.save.raw(bst)
 # load binary model to R
 bst3 <- xgb.load(raw)
 pred3 <- predict(bst3, test$data)
 # pred3 should be identical to pred
-print(paste("sum(abs(pred3-pred))=", sum(abs(pred3-pred))))
+print(paste("sum(abs(pred3-pred))=", sum(abs(pred3 - pred))))
 
 #----------------Advanced features --------------
 # to use advanced features, we need to put data in xgb.DMatrix
-dtrain <- xgb.DMatrix(data = train$data, label=train$label)
-dtest <- xgb.DMatrix(data = test$data, label=test$label)
+dtrain <- xgb.DMatrix(data = train$data, label = train$label)
+dtest <- xgb.DMatrix(data = test$data, label = test$label)
 #---------------Using watchlist----------------
 # watchlist is a list of xgb.DMatrix, each of them is tagged with name
-watchlist <- list(train=dtrain, test=dtest)
+watchlist <- list(train = dtrain, test = dtest)
 # to train with watchlist, use xgb.train, which contains more advanced features
-# watchlist allows us to monitor the evaluation result on all data in the list 
+# watchlist allows us to monitor the evaluation result on all data in the list
 print("Train xgboost using xgb.train with watchlist")
-bst <- xgb.train(data=dtrain, max_depth=2, eta=1, nrounds=2, watchlist=watchlist,
+bst <- xgb.train(data = dtrain, max_depth = 2, eta = 1, nrounds = 2, watchlist = watchlist,
                  nthread = 2, objective = "binary:logistic")
 # we can change evaluation metrics, or use multiple evaluation metrics
 print("train xgboost using xgb.train with watchlist, watch logloss and error")
-bst <- xgb.train(data=dtrain, max_depth=2, eta=1, nrounds=2, watchlist=watchlist,
+bst <- xgb.train(data = dtrain, max_depth = 2, eta = 1, nrounds = 2, watchlist = watchlist,
                  eval_metric = "error", eval_metric = "logloss",
                  nthread = 2, objective = "binary:logistic")
 
@@ -90,17 +90,17 @@ bst <- xgb.train(data=dtrain, max_depth=2, eta=1, nrounds=2, watchlist=watchlist
 xgb.DMatrix.save(dtrain, "dtrain.buffer")
 # to load it in, simply call xgb.DMatrix
 dtrain2 <- xgb.DMatrix("dtrain.buffer")
-bst <- xgb.train(data=dtrain2, max_depth=2, eta=1, nrounds=2, watchlist=watchlist,
+bst <- xgb.train(data = dtrain2, max_depth = 2, eta = 1, nrounds = 2, watchlist = watchlist,
                  nthread = 2, objective = "binary:logistic")
 # information can be extracted from xgb.DMatrix using getinfo
-label = getinfo(dtest, "label")
+label <- getinfo(dtest, "label")
 pred <- predict(bst, dtest)
-err <- as.numeric(sum(as.integer(pred > 0.5) != label))/length(label)
+err <- as.numeric(sum(as.integer(pred > 0.5) != label)) / length(label)
 print(paste("test-error=", err))
 
 # You can dump the tree you learned using xgb.dump into a text file
-dump_path = file.path(tempdir(), 'dump.raw.txt')
-xgb.dump(bst, dump_path, with_stats = T)
+dump_path <- file.path(tempdir(), 'dump.raw.txt')
+xgb.dump(bst, dump_path, with_stats = TRUE)
 
 # Finally, you can check which features are the most important.
 print("Most important features (look at column Gain):")

R-package/demo/boost_from_prediction.R

@@ -1,7 +1,7 @@
 require(xgboost)
 # load in the agaricus dataset
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 
@@ -11,12 +11,12 @@ watchlist <- list(eval = dtest, train = dtrain)
 #
 print('start running example to start from a initial prediction')
 # train xgboost for 1 round
-param <- list(max_depth=2, eta=1, nthread = 2, silent=1, objective='binary:logistic')
+param <- list(max_depth = 2, eta = 1, nthread = 2, objective = 'binary:logistic')
 bst <- xgb.train(param, dtrain, 1, watchlist)
 # Note: we need the margin value instead of transformed prediction in set_base_margin
 # do predict with output_margin=TRUE, will always give you margin values before logistic transformation
-ptrain <- predict(bst, dtrain, outputmargin=TRUE)
-ptest  <- predict(bst, dtest, outputmargin=TRUE)
+ptrain <- predict(bst, dtrain, outputmargin = TRUE)
+ptest  <- predict(bst, dtest, outputmargin = TRUE)
 # set the base_margin property of dtrain and dtest
 # base margin is the base prediction we will boost from
 setinfo(dtrain, "base_margin", ptrain)

R-package/demo/caret_wrapper.R

@@ -1,5 +1,5 @@
 # install development version of caret library that contains xgboost models
-devtools::install_github("topepo/caret/pkg/caret") 
+devtools::install_github("topepo/caret/pkg/caret")
 require(caret)
 require(xgboost)
 require(data.table)
@@ -9,17 +9,17 @@ require(e1071)
 # Load Arthritis dataset in memory.
 data(Arthritis)
 # Create a copy of the dataset with data.table package (data.table is 100% compliant with R dataframe but its syntax is a lot more consistent and its performance are really good).
-df <- data.table(Arthritis, keep.rownames = F)
+df <- data.table(Arthritis, keep.rownames = FALSE)
 
 # Let's add some new categorical features to see if it helps. Of course these feature are highly correlated to the Age feature. Usually it's not a good thing in ML, but Tree algorithms (including boosted trees) are able to select the best features, even in case of highly correlated features.
 # For the first feature we create groups of age by rounding the real age. Note that we transform it to factor (categorical data) so the algorithm treat them as independant values.
-df[,AgeDiscret:= as.factor(round(Age/10,0))]
+df[, AgeDiscret := as.factor(round(Age / 10, 0))]
 
 # Here is an even stronger simplification of the real age with an arbitrary split at 30 years old. I choose this value based on nothing. We will see later if simplifying the information based on arbitrary values is a good strategy (I am sure you already have an idea of how well it will work!).
-df[,AgeCat:= as.factor(ifelse(Age > 30, "Old", "Young"))]
+df[, AgeCat := as.factor(ifelse(Age > 30, "Old", "Young"))]
 
 # We remove ID as there is nothing to learn from this feature (it will just add some noise as the dataset is small).
-df[,ID:=NULL]
+df[, ID := NULL]
 
 #-------------Basic Training using XGBoost in caret Library-----------------
 # Set up control parameters for caret::train

R-package/demo/create_sparse_matrix.R

@@ -6,10 +6,10 @@ if (!require(vcd)) {
   require(vcd)
 }
 # According to its documentation, Xgboost works only on numbers.
-# Sometimes the dataset we have to work on have categorical data. 
+# Sometimes the dataset we have to work on have categorical data.
 # A categorical variable is one which have a fixed number of values. By example, if for each observation a variable called "Colour" can have only "red", "blue" or "green" as value, it is a categorical variable.
 #
-# In R, categorical variable is called Factor. 
+# In R, categorical variable is called Factor.
 # Type ?factor in console for more information.
 #
 # In this demo we will see how to transform a dense dataframe with categorical variables to a sparse matrix before analyzing it in Xgboost.
@@ -19,7 +19,7 @@ if (!require(vcd)) {
 data(Arthritis)
 
 # create a copy of the dataset with data.table package (data.table is 100% compliant with R dataframe but its syntax is a lot more consistent and its performance are really good).
-df <- data.table(Arthritis, keep.rownames = F)
+df <- data.table(Arthritis, keep.rownames = FALSE)
 
 # Let's have a look to the data.table
 cat("Print the dataset\n")
@@ -32,17 +32,17 @@ str(df)
 # Let's add some new categorical features to see if it helps. Of course these feature are highly correlated to the Age feature. Usually it's not a good thing in ML, but Tree algorithms (including boosted trees) are able to select the best features, even in case of highly correlated features.
 
 # For the first feature we create groups of age by rounding the real age. Note that we transform it to factor (categorical data) so the algorithm treat them as independant values.
-df[,AgeDiscret:= as.factor(round(Age/10,0))]
+df[, AgeDiscret := as.factor(round(Age / 10, 0))]
 
 # Here is an even stronger simplification of the real age with an arbitrary split at 30 years old. I choose this value based on nothing. We will see later if simplifying the information based on arbitrary values is a good strategy (I am sure you already have an idea of how well it will work!).
-df[,AgeCat:= as.factor(ifelse(Age > 30, "Old", "Young"))]
+df[, AgeCat := as.factor(ifelse(Age > 30, "Old", "Young"))]
 
 # We remove ID as there is nothing to learn from this feature (it will just add some noise as the dataset is small).
-df[,ID:=NULL]
+df[, ID := NULL]
 
 # List the different values for the column Treatment: Placebo, Treated.
 cat("Values of the categorical feature Treatment\n")
-print(levels(df[,Treatment]))
+print(levels(df[, Treatment]))
 
 # Next step, we will transform the categorical data to dummy variables.
 # This method is also called one hot encoding.
@@ -52,16 +52,16 @@ print(levels(df[,Treatment]))
 #
 # Formulae Improved~.-1 used below means transform all categorical features but column Improved to binary values.
 # Column Improved is excluded because it will be our output column, the one we want to predict.
-sparse_matrix = sparse.model.matrix(Improved~.-1, data = df)
+sparse_matrix <- sparse.model.matrix(Improved ~ . - 1, data = df)
 
 cat("Encoding of the sparse Matrix\n")
 print(sparse_matrix)
 
 # Create the output vector (not sparse)
-# 1. Set, for all rows, field in Y column to 0; 
-# 2. set Y to 1 when Improved == Marked; 
+# 1. Set, for all rows, field in Y column to 0;
+# 2. set Y to 1 when Improved == Marked;
 # 3. Return Y column
-output_vector = df[,Y:=0][Improved == "Marked",Y:=1][,Y]
+output_vector <- df[, Y := 0][Improved == "Marked", Y := 1][, Y]
 
 # Following is the same process as other demo
 cat("Learning...\n")

R-package/demo/cross_validation.R

@@ -1,25 +1,25 @@
 require(xgboost)
 # load in the agaricus dataset
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 
 nrounds <- 2
-param <- list(max_depth=2, eta=1, silent=1, nthread=2, objective='binary:logistic')
+param <- list(max_depth = 2, eta = 1, nthread = 2, objective = 'binary:logistic')
 
 cat('running cross validation\n')
 # do cross validation, this will print result out as
 # [iteration]  metric_name:mean_value+std_value
 # std_value is standard deviation of the metric
-xgb.cv(param, dtrain, nrounds, nfold=5, metrics={'error'})
+xgb.cv(param, dtrain, nrounds, nfold = 5, metrics = {'error'})
 
 cat('running cross validation, disable standard deviation display\n')
 # do cross validation, this will print result out as
 # [iteration]  metric_name:mean_value+std_value
 # std_value is standard deviation of the metric
-xgb.cv(param, dtrain, nrounds, nfold=5,
-       metrics='error', showsd = FALSE)
+xgb.cv(param, dtrain, nrounds, nfold = 5,
+       metrics = 'error', showsd = FALSE)
 
 ###
 # you can also do cross validation with cutomized loss function
@@ -29,18 +29,18 @@ print ('running cross validation, with cutomsized loss function')
 
 logregobj <- function(preds, dtrain) {
   labels <- getinfo(dtrain, "label")
-  preds <- 1/(1 + exp(-preds))
+  preds <- 1 / (1 + exp(-preds))
   grad <- preds - labels
   hess <- preds * (1 - preds)
   return(list(grad = grad, hess = hess))
 }
 evalerror <- function(preds, dtrain) {
   labels <- getinfo(dtrain, "label")
-  err <- as.numeric(sum(labels != (preds > 0)))/length(labels)
+  err <- as.numeric(sum(labels != (preds > 0))) / length(labels)
   return(list(metric = "error", value = err))
 }
 
-param <- list(max_depth=2, eta=1, silent=1,
+param <- list(max_depth = 2, eta = 1,
               objective = logregobj, eval_metric = evalerror)
 # train with customized objective
 xgb.cv(params = param, data = dtrain, nrounds = nrounds, nfold = 5)

R-package/demo/custom_objective.R

@@ -1,7 +1,7 @@
 require(xgboost)
 # load in the agaricus dataset
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 
@@ -15,7 +15,7 @@ num_round <- 2
 # this is loglikelihood loss
 logregobj <- function(preds, dtrain) {
   labels <- getinfo(dtrain, "label")
-  preds <- 1/(1 + exp(-preds))
+  preds <- 1 / (1 + exp(-preds))
   grad <- preds - labels
   hess <- preds * (1 - preds)
   return(list(grad = grad, hess = hess))
@@ -29,36 +29,36 @@ logregobj <- function(preds, dtrain) {
 # Take this in mind when you use the customization, and maybe you need write customized evaluation function
 evalerror <- function(preds, dtrain) {
   labels <- getinfo(dtrain, "label")
-  err <- as.numeric(sum(labels != (preds > 0)))/length(labels)
+  err <- as.numeric(sum(labels != (preds > 0))) / length(labels)
   return(list(metric = "error", value = err))
 }
 
-param <- list(max_depth=2, eta=1, nthread = 2, verbosity=0, 
-              objective=logregobj, eval_metric=evalerror)
+param <- list(max_depth = 2, eta = 1, nthread  =  2, verbosity = 0,
+              objective = logregobj, eval_metric = evalerror)
 print ('start training with user customized objective')
 # training with customized objective, we can also do step by step training
 # simply look at xgboost.py's implementation of train
 bst <- xgb.train(param, dtrain, num_round, watchlist)
 
 #
-# there can be cases where you want additional information 
+# there can be cases where you want additional information
 # being considered besides the property of DMatrix you can get by getinfo
 # you can set additional information as attributes if DMatrix
 
-# set label attribute of dtrain to be label, we use label as an example, it can be anything 
+# set label attribute of dtrain to be label, we use label as an example, it can be anything
 attr(dtrain, 'label') <- getinfo(dtrain, 'label')
 # this is new customized objective, where you can access things you set
 # same thing applies to customized evaluation function
 logregobjattr <- function(preds, dtrain) {
   # now you can access the attribute in customized function
   labels <- attr(dtrain, 'label')
-  preds <- 1/(1 + exp(-preds))
+  preds <- 1 / (1 + exp(-preds))
   grad <- preds - labels
   hess <- preds * (1 - preds)
   return(list(grad = grad, hess = hess))
 }
-param <- list(max_depth=2, eta=1, nthread = 2, verbosity=0, 
-              objective=logregobjattr, eval_metric=evalerror)
+param <- list(max_depth = 2, eta = 1, nthread  =  2, verbosity = 0,
+              objective = logregobjattr, eval_metric = evalerror)
 print ('start training with user customized objective, with additional attributes in DMatrix')
 # training with customized objective, we can also do step by step training
 # simply look at xgboost.py's implementation of train

R-package/demo/early_stopping.R

@@ -1,20 +1,20 @@
 require(xgboost)
 # load in the agaricus dataset
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 # note: for customized objective function, we leave objective as default
 # note: what we are getting is margin value in prediction
 # you must know what you are doing
-param <- list(max_depth=2, eta=1, nthread=2, verbosity=0)
+param <- list(max_depth = 2, eta = 1, nthread = 2, verbosity = 0)
 watchlist <- list(eval = dtest)
 num_round <- 20
 # user define objective function, given prediction, return gradient and second order gradient
 # this is loglikelihood loss
 logregobj <- function(preds, dtrain) {
   labels <- getinfo(dtrain, "label")
-  preds <- 1/(1 + exp(-preds))
+  preds <- 1 / (1 + exp(-preds))
   grad <- preds - labels
   hess <- preds * (1 - preds)
   return(list(grad = grad, hess = hess))
@@ -27,7 +27,7 @@ logregobj <- function(preds, dtrain) {
 # Take this in mind when you use the customization, and maybe you need write customized evaluation function
 evalerror <- function(preds, dtrain) {
   labels <- getinfo(dtrain, "label")
-  err <- as.numeric(sum(labels != (preds > 0)))/length(labels)
+  err <- as.numeric(sum(labels != (preds > 0))) / length(labels)
   return(list(metric = "error", value = err))
 }
 print ('start training with early Stopping setting')

R-package/demo/generalized_linear_model.R

@@ -1,7 +1,7 @@
 require(xgboost)
 # load in the agaricus dataset
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 ##
@@ -11,14 +11,14 @@ dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 ##
 
 # change booster to gblinear, so that we are fitting a linear model
-# alpha is the L1 regularizer 
+# alpha is the L1 regularizer
 # lambda is the L2 regularizer
 # you can also set lambda_bias which is L2 regularizer on the bias term
 param <- list(objective = "binary:logistic", booster = "gblinear",
               nthread = 2, alpha = 0.0001, lambda = 1)
 
 # normally, you do not need to set eta (step_size)
-# XGBoost uses a parallel coordinate descent algorithm (shotgun), 
+# XGBoost uses a parallel coordinate descent algorithm (shotgun),
 # there could be affection on convergence with parallelization on certain cases
 # setting eta to be smaller value, e.g 0.5 can make the optimization more stable
 
@@ -30,5 +30,4 @@ num_round <- 2
 bst <- xgb.train(param, dtrain, num_round, watchlist)
 ypred <- predict(bst, dtest)
 labels <- getinfo(dtest, 'label')
-cat('error of preds=', mean(as.numeric(ypred>0.5)!=labels),'\n')
-
+cat('error of preds=', mean(as.numeric(ypred > 0.5) != labels), '\n')

R-package/demo/gpu_accelerated.R

@@ -1,9 +1,9 @@
 # An example of using GPU-accelerated tree building algorithms
-# 
-# NOTE: it can only run if you have a CUDA-enable GPU and the package was 
+#
+# NOTE: it can only run if you have a CUDA-enable GPU and the package was
 #       specially compiled with GPU support.
 #
-# For the current functionality, see 
+# For the current functionality, see
 # https://xgboost.readthedocs.io/en/latest/gpu/index.html
 #
 
@@ -21,8 +21,8 @@ m <- X[, sel] %*% betas - 1 + rnorm(N)
 y <- rbinom(N, 1, plogis(m))
 
 tr <- sample.int(N, N * 0.75)
-dtrain <- xgb.DMatrix(X[tr,], label = y[tr])
-dtest <- xgb.DMatrix(X[-tr,], label = y[-tr])
+dtrain <- xgb.DMatrix(X[tr, ], label = y[tr])
+dtest <- xgb.DMatrix(X[-tr, ], label = y[-tr])
 wl <- list(train = dtrain, test = dtest)
 
 # An example of running 'gpu_hist' algorithm

R-package/demo/interaction_constraints.R

@@ -4,34 +4,39 @@ library(data.table)
 set.seed(1024)
 
 # Function to obtain a list of interactions fitted in trees, requires input of maximum depth
-treeInteractions <- function(input_tree, input_max_depth){
-  trees <- copy(input_tree)  # copy tree input to prevent overwriting
+treeInteractions <- function(input_tree, input_max_depth) {
+  ID_merge <- i.id <- i.feature <- NULL  # Suppress warning "no visible binding for global variable"
+
+  trees <- data.table::copy(input_tree)  # copy tree input to prevent overwriting
   if (input_max_depth < 2) return(list())  # no interactions if max depth < 2
   if (nrow(input_tree) == 1) return(list())
 
   # Attach parent nodes
-  for (i in 2:input_max_depth){
-    if (i == 2) trees[, ID_merge:=ID] else trees[, ID_merge:=get(paste0('parent_',i-2))]
-    parents_left <- trees[!is.na(Split), list(i.id=ID, i.feature=Feature, ID_merge=Yes)]
-    parents_right <- trees[!is.na(Split), list(i.id=ID, i.feature=Feature, ID_merge=No)]
+  for (i in 2:input_max_depth) {
+    if (i == 2) trees[, ID_merge := ID] else trees[, ID_merge := get(paste0('parent_', i - 2))]
+    parents_left <- trees[!is.na(Split), list(i.id = ID, i.feature = Feature, ID_merge = Yes)]
+    parents_right <- trees[!is.na(Split), list(i.id = ID, i.feature = Feature, ID_merge = No)]
 
-    setorderv(trees, 'ID_merge')
-    setorderv(parents_left, 'ID_merge')
-    setorderv(parents_right, 'ID_merge')
+    data.table::setorderv(trees, 'ID_merge')
+    data.table::setorderv(parents_left, 'ID_merge')
+    data.table::setorderv(parents_right, 'ID_merge')
 
-    trees <- merge(trees, parents_left, by='ID_merge', all.x=T)
-    trees[!is.na(i.id), c(paste0('parent_', i-1), paste0('parent_feat_', i-1)):=list(i.id, i.feature)]
-    trees[, c('i.id','i.feature'):=NULL]
+    trees <- merge(trees, parents_left, by = 'ID_merge', all.x = TRUE)
+    trees[!is.na(i.id), c(paste0('parent_', i - 1), paste0('parent_feat_', i - 1))
+          := list(i.id, i.feature)]
+    trees[, c('i.id', 'i.feature') := NULL]
 
-    trees <- merge(trees, parents_right, by='ID_merge', all.x=T)
-    trees[!is.na(i.id), c(paste0('parent_', i-1), paste0('parent_feat_', i-1)):=list(i.id, i.feature)]
-    trees[, c('i.id','i.feature'):=NULL]
+    trees <- merge(trees, parents_right, by = 'ID_merge', all.x = TRUE)
+    trees[!is.na(i.id), c(paste0('parent_', i - 1), paste0('parent_feat_', i - 1))
+          := list(i.id, i.feature)]
+    trees[, c('i.id', 'i.feature') := NULL]
   }
 
   # Extract nodes with interactions
-  interaction_trees <- trees[!is.na(Split) & !is.na(parent_1), 
-                             c('Feature',paste0('parent_feat_',1:(input_max_depth-1))), with=F]
-  interaction_trees_split <- split(interaction_trees, 1:nrow(interaction_trees))
+  interaction_trees <- trees[!is.na(Split) & !is.na(parent_1),
+                             c('Feature', paste0('parent_feat_', 1:(input_max_depth - 1))),
+                             with = FALSE]
+  interaction_trees_split <- split(interaction_trees, seq_len(nrow(interaction_trees)))
   interaction_list <- lapply(interaction_trees_split, as.character)
 
   # Remove NAs (no parent interaction)
@@ -47,59 +52,62 @@ treeInteractions <- function(input_tree, input_max_depth){
 
 # Generate sample data
 x <- list()
-for (i in 1:10){
-  x[[i]] = i*rnorm(1000, 10)
+for (i in 1:10) {
+  x[[i]] <- i * rnorm(1000, 10)
 }
 x <- as.data.table(x)
 
-y = -1*x[, rowSums(.SD)] + x[['V1']]*x[['V2']] + x[['V3']]*x[['V4']]*x[['V5']] + rnorm(1000, 0.001) + 3*sin(x[['V7']])
+y <- -1 * x[, rowSums(.SD)] + x[['V1']] * x[['V2']] + x[['V3']] * x[['V4']] * x[['V5']]
+     + rnorm(1000, 0.001) + 3 * sin(x[['V7']])
 
-train = as.matrix(x)
+train <- as.matrix(x)
 
 # Interaction constraint list (column names form)
-interaction_list <- list(c('V1','V2'),c('V3','V4','V5'))
+interaction_list <- list(c('V1', 'V2'), c('V3', 'V4', 'V5'))
 
 # Convert interaction constraint list into feature index form
 cols2ids <- function(object, col_names) {
   LUT <- seq_along(col_names) - 1
   names(LUT) <- col_names
-  rapply(object, function(x) LUT[x], classes="character", how="replace")
+  rapply(object, function(x) LUT[x], classes = "character", how = "replace")
 }
-interaction_list_fid = cols2ids(interaction_list, colnames(train))
+interaction_list_fid <- cols2ids(interaction_list, colnames(train))
 
 # Fit model with interaction constraints
-bst = xgboost(data = train, label = y, max_depth = 4,
-              eta = 0.1, nthread = 2, nrounds = 1000,
-              interaction_constraints = interaction_list_fid)
+bst <- xgboost(data = train, label = y, max_depth = 4,
+               eta = 0.1, nthread = 2, nrounds = 1000,
+               interaction_constraints = interaction_list_fid)
 
 bst_tree <- xgb.model.dt.tree(colnames(train), bst)
-bst_interactions <- treeInteractions(bst_tree, 4)  # interactions constrained to combinations of V1*V2 and V3*V4*V5
+bst_interactions <- treeInteractions(bst_tree, 4)
+  # interactions constrained to combinations of V1*V2 and V3*V4*V5
 
 # Fit model without interaction constraints
-bst2 = xgboost(data = train, label = y, max_depth = 4,
-               eta = 0.1, nthread = 2, nrounds = 1000)
+bst2 <- xgboost(data = train, label = y, max_depth = 4,
+                eta = 0.1, nthread = 2, nrounds = 1000)
 
 bst2_tree <- xgb.model.dt.tree(colnames(train), bst2)
 bst2_interactions <- treeInteractions(bst2_tree, 4)  # much more interactions
 
 # Fit model with both interaction and monotonicity constraints
-bst3 = xgboost(data = train, label = y, max_depth = 4,
-               eta = 0.1, nthread = 2, nrounds = 1000,
-               interaction_constraints = interaction_list_fid,
-               monotone_constraints = c(-1,0,0,0,0,0,0,0,0,0))
+bst3 <- xgboost(data = train, label = y, max_depth = 4,
+                eta = 0.1, nthread = 2, nrounds = 1000,
+                interaction_constraints = interaction_list_fid,
+                monotone_constraints = c(-1, 0, 0, 0, 0, 0, 0, 0, 0, 0))
 
 bst3_tree <- xgb.model.dt.tree(colnames(train), bst3)
-bst3_interactions <- treeInteractions(bst3_tree, 4)  # interactions still constrained to combinations of V1*V2 and V3*V4*V5
+bst3_interactions <- treeInteractions(bst3_tree, 4)
+  # interactions still constrained to combinations of V1*V2 and V3*V4*V5
 
 # Show monotonic constraints still apply by checking scores after incrementing V1
 x1 <- sort(unique(x[['V1']]))
-for (i in 1:length(x1)){
-  testdata <- copy(x[, -c('V1')])
+for (i in seq_along(x1)){
+  testdata <- copy(x[, - ('V1')])
   testdata[['V1']] <- x1[i]
-  testdata <- testdata[, paste0('V',1:10), with=F]
+  testdata <- testdata[, paste0('V', 1:10), with = FALSE]
   pred <- predict(bst3, as.matrix(testdata))
-  
+
   # Should not print out anything due to monotonic constraints
   if (i > 1) if (any(pred > prev_pred)) print(i)
-  prev_pred <- pred 
+  prev_pred <- pred
 }

R-package/demo/poisson_regression.R

@@ -1,7 +1,6 @@
 data(mtcars)
 head(mtcars)
-bst = xgboost(data=as.matrix(mtcars[,-11]),label=mtcars[,11],
-              objective='count:poisson',nrounds=5)
-pred = predict(bst,as.matrix(mtcars[,-11]))
-sqrt(mean((pred-mtcars[,11])^2))
-
+bst <- xgboost(data = as.matrix(mtcars[, -11]), label = mtcars[, 11],
+               objective = 'count:poisson', nrounds = 5)
+pred <- predict(bst, as.matrix(mtcars[, -11]))
+sqrt(mean((pred - mtcars[, 11]) ^ 2))

R-package/demo/predict_first_ntree.R

@@ -1,23 +1,23 @@
 require(xgboost)
 # load in the agaricus dataset
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 
-param <- list(max_depth=2, eta=1, silent=1, objective='binary:logistic')
+param <- list(max_depth = 2, eta = 1, objective = 'binary:logistic')
 watchlist <- list(eval = dtest, train = dtrain)
-nrounds = 2
+nrounds <- 2
 
 # training the model for two rounds
-bst = xgb.train(param, dtrain, nrounds, nthread = 2, watchlist)
+bst <- xgb.train(param, dtrain, nrounds, nthread = 2, watchlist)
 cat('start testing prediction from first n trees\n')
-labels <- getinfo(dtest,'label')
+labels <- getinfo(dtest, 'label')
 
 ### predict using first 1 tree
-ypred1 = predict(bst, dtest, ntreelimit=1)
+ypred1 <- predict(bst, dtest, ntreelimit = 1)
 # by default, we predict using all the trees
-ypred2 = predict(bst, dtest)
+ypred2 <- predict(bst, dtest)
 
-cat('error of ypred1=', mean(as.numeric(ypred1>0.5)!=labels),'\n')
-cat('error of ypred2=', mean(as.numeric(ypred2>0.5)!=labels),'\n')
+cat('error of ypred1=', mean(as.numeric(ypred1 > 0.5) != labels), '\n')
+cat('error of ypred2=', mean(as.numeric(ypred2 > 0.5) != labels), '\n')

R-package/demo/predict_leaf_indices.R

@@ -5,34 +5,34 @@ require(Matrix)
 set.seed(1982)
 
 # load in the agaricus dataset
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 dtrain <- xgb.DMatrix(data = agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(data = agaricus.test$data, label = agaricus.test$label)
 
-param <- list(max_depth=2, eta=1, silent=1, objective='binary:logistic')
-nrounds = 4
+param <- list(max_depth = 2, eta = 1, objective = 'binary:logistic')
+nrounds <- 4
 
 # training the model for two rounds
-bst = xgb.train(params = param, data = dtrain, nrounds = nrounds, nthread = 2)
+bst <- xgb.train(params = param, data = dtrain, nrounds = nrounds, nthread = 2)
 
 # Model accuracy without new features
-accuracy.before <- sum((predict(bst, agaricus.test$data) >= 0.5) == agaricus.test$label) / length(agaricus.test$label)
+accuracy.before <- (sum((predict(bst, agaricus.test$data) >= 0.5) == agaricus.test$label)
+                    / length(agaricus.test$label))
 
 # by default, we predict using all the trees
-
-pred_with_leaf = predict(bst, dtest, predleaf = TRUE)
+pred_with_leaf <- predict(bst, dtest, predleaf = TRUE)
 head(pred_with_leaf)
 
 create.new.tree.features <- function(model, original.features){
   pred_with_leaf <- predict(model, original.features, predleaf = TRUE)
   cols <- list()
-  for(i in 1:model$niter){
+  for (i in 1:model$niter) {
     # max is not the real max but it s not important for the purpose of adding features
-    leaf.id <- sort(unique(pred_with_leaf[,i]))
-    cols[[i]] <- factor(x = pred_with_leaf[,i], level = leaf.id)
+    leaf.id <- sort(unique(pred_with_leaf[, i]))
+    cols[[i]] <- factor(x = pred_with_leaf[, i], level = leaf.id)
   }
-  cbind(original.features, sparse.model.matrix( ~ . -1, as.data.frame(cols)))
+  cbind(original.features, sparse.model.matrix(~ . - 1, as.data.frame(cols)))
 }
 
 # Convert previous features to one hot encoding
@@ -47,7 +47,9 @@ watchlist <- list(train = new.dtrain)
 bst <- xgb.train(params = param, data = new.dtrain, nrounds = nrounds, nthread = 2)
 
 # Model accuracy with new features
-accuracy.after <- sum((predict(bst, new.dtest) >= 0.5) == agaricus.test$label) / length(agaricus.test$label)
+accuracy.after <- (sum((predict(bst, new.dtest) >= 0.5) == agaricus.test$label)
+                   / length(agaricus.test$label))
 
 # Here the accuracy was already good and is now perfect.
-cat(paste("The accuracy was", accuracy.before, "before adding leaf features and it is now", accuracy.after, "!\n"))
+cat(paste("The accuracy was", accuracy.before, "before adding leaf features and it is now",
+          accuracy.after, "!\n"))

R-package/demo/runall.R

@@ -1,14 +1,14 @@
 # running all scripts in demo folder
-demo(basic_walkthrough)
-demo(custom_objective)
-demo(boost_from_prediction)
-demo(predict_first_ntree)
-demo(generalized_linear_model)
-demo(cross_validation)
-demo(create_sparse_matrix)
-demo(predict_leaf_indices)
-demo(early_stopping)
-demo(poisson_regression)
-demo(caret_wrapper)
-demo(tweedie_regression)
-#demo(gpu_accelerated) # can only run when built with GPU support
+demo(basic_walkthrough, package = 'xgboost')
+demo(custom_objective, package = 'xgboost')
+demo(boost_from_prediction, package = 'xgboost')
+demo(predict_first_ntree, package = 'xgboost')
+demo(generalized_linear_model, package = 'xgboost')
+demo(cross_validation, package = 'xgboost')
+demo(create_sparse_matrix, package = 'xgboost')
+demo(predict_leaf_indices, package = 'xgboost')
+demo(early_stopping, package = 'xgboost')
+demo(poisson_regression, package = 'xgboost')
+demo(caret_wrapper, package = 'xgboost')
+demo(tweedie_regression, package = 'xgboost')
+#demo(gpu_accelerated, package = 'xgboost')  # can only run when built with GPU support

R-package/demo/tweedie_regression.R

@@ -8,12 +8,12 @@ data(AutoClaim)
 dt <- data.table(AutoClaim)
 
 # exclude these columns from the model matrix
-exclude <-  c('POLICYNO', 'PLCYDATE', 'CLM_FREQ5', 'CLM_AMT5', 'CLM_FLAG', 'IN_YY')
+exclude <- c('POLICYNO', 'PLCYDATE', 'CLM_FREQ5', 'CLM_AMT5', 'CLM_FLAG', 'IN_YY')
 
 # retains the missing values
 # NOTE: this dataset is comes ready out of the box
 options(na.action = 'na.pass')
-x <- sparse.model.matrix(~ . - 1, data = dt[, -exclude, with = F])
+x <- sparse.model.matrix(~ . - 1, data = dt[, -exclude, with = FALSE])
 options(na.action = 'na.omit')
 
 # response
@@ -21,29 +21,29 @@ y <- dt[, CLM_AMT5]
 
 d_train <- xgb.DMatrix(data = x, label = y, missing = NA)
 
-# the tweedie_variance_power parameter determines the shape of 
+# the tweedie_variance_power parameter determines the shape of
 # distribution
 # - closer to 1 is more poisson like and the mass
-#   is more concentrated near zero 
-# - closer to 2 is more gamma like and the mass spreads to the 
+#   is more concentrated near zero
+# - closer to 2 is more gamma like and the mass spreads to the
 #   the right with less concentration near zero
 
 params <- list(
   objective = 'reg:tweedie',
-  eval_metric = 'rmse', 
+  eval_metric = 'rmse',
   tweedie_variance_power = 1.4,
   max_depth = 6,
   eta = 1)
 
 bst <- xgb.train(
-  data = d_train, 
-  params = params, 
+  data = d_train,
+  params = params,
   maximize = FALSE,
-  watchlist = list(train = d_train), 
+  watchlist = list(train = d_train),
   nrounds = 20)
 
 var_imp <- xgb.importance(attr(x, 'Dimnames')[[2]], model = bst)
 
 preds <- predict(bst, d_train)
 
-rmse <- sqrt(sum(mean((y - preds)^2)))
+rmse <- sqrt(sum(mean((y - preds) ^ 2)))

R-package/inst/make-r-def.R

@@ -0,0 +1,96 @@
+# [description]
+#     Create a definition file (.def) from a .dll file, using objdump. This
+#     is used by FindLibR.cmake when building the R package with MSVC.
+#
+# [usage]
+#
+#     Rscript make-r-def.R something.dll something.def
+#
+# [references]
+#    * https://www.cs.colorado.edu/~main/cs1300/doc/mingwfaq.html
+
+args <- commandArgs(trailingOnly = TRUE)
+
+IN_DLL_FILE <- args[[1L]]
+OUT_DEF_FILE <- args[[2L]]
+DLL_BASE_NAME <- basename(IN_DLL_FILE)
+
+message(sprintf("Creating '%s' from '%s'", OUT_DEF_FILE, IN_DLL_FILE))
+
+# system() will not raise an R exception if the process called
+# fails. Wrapping it here to get that behavior.
+#
+# system() introduces a lot of overhead, at least on Windows,
+# so trying processx if it is available
+.pipe_shell_command_to_stdout <- function(command, args, out_file) {
+    has_processx <- suppressMessages({
+        suppressWarnings({
+            require("processx")  # nolint
+        })
+    })
+    if (has_processx) {
+        p <- processx::process$new(
+            command = command
+            , args = args
+            , stdout = out_file
+            , windows_verbatim_args = FALSE
+        )
+        invisible(p$wait())
+    } else {
+        message(paste0(
+            "Using system2() to run shell commands. Installing "
+            , "'processx' with install.packages('processx') might "
+            , "make this faster."
+        ))
+        exit_code <- system2(
+            command = command
+            , args = shQuote(args)
+            , stdout = out_file
+        )
+        if (exit_code != 0L) {
+            stop(paste0("Command failed with exit code: ", exit_code))
+        }
+    }
+    return(invisible(NULL))
+}
+
+# use objdump to dump all the symbols
+OBJDUMP_FILE <- "objdump-out.txt"
+.pipe_shell_command_to_stdout(
+    command = "objdump"
+    , args = c("-p", IN_DLL_FILE)
+    , out_file = OBJDUMP_FILE
+)
+
+objdump_results <- readLines(OBJDUMP_FILE)
+result <- file.remove(OBJDUMP_FILE)
+
+# Only one table in the objdump results matters for our purposes,
+# see https://www.cs.colorado.edu/~main/cs1300/doc/mingwfaq.html
+start_index <- which(
+    grepl(
+        pattern = "[Ordinal/Name Pointer] Table"
+        , x = objdump_results
+        , fixed = TRUE
+    )
+)
+empty_lines <- which(objdump_results == "")
+end_of_table <- empty_lines[empty_lines > start_index][1L]
+
+# Read the contents of the table
+exported_symbols <- objdump_results[(start_index + 1L):end_of_table]
+exported_symbols <- gsub("\t", "", exported_symbols)
+exported_symbols <- gsub(".*\\] ", "", exported_symbols)
+exported_symbols <- gsub(" ", "", exported_symbols)
+
+# Write R.def file
+writeLines(
+    text = c(
+        paste0("LIBRARY \"", DLL_BASE_NAME, "\"")
+        , "EXPORTS"
+        , exported_symbols
+    )
+    , con = OUT_DEF_FILE
+    , sep = "\n"
+)
+message(sprintf("Successfully created '%s'", OUT_DEF_FILE))

R-package/man/a-compatibility-note-for-saveRDS-save.Rd

@@ -0,0 +1,64 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/utils.R
+\name{a-compatibility-note-for-saveRDS-save}
+\alias{a-compatibility-note-for-saveRDS-save}
+\title{Do not use \code{\link[base]{saveRDS}} or \code{\link[base]{save}} for long-term archival of
+models. Instead, use \code{\link{xgb.save}} or \code{\link{xgb.save.raw}}.}
+\description{
+It is a common practice to use the built-in \code{\link[base]{saveRDS}} function (or
+\code{\link[base]{save}}) to persist R objects to the disk. While it is possible to persist
+\code{xgb.Booster} objects using \code{\link[base]{saveRDS}}, it is not advisable to do so if
+the model is to be accessed in the future. If you train a model with the current version of
+XGBoost and persist it with \code{\link[base]{saveRDS}}, the model is not guaranteed to be
+accessible in later releases of XGBoost. To ensure that your model can be accessed in future
+releases of XGBoost, use \code{\link{xgb.save}} or \code{\link{xgb.save.raw}} instead.
+}
+\details{
+Use \code{\link{xgb.save}} to save the XGBoost model as a stand-alone file. You may opt into
+the JSON format by specifying the JSON extension. To read the model back, use
+\code{\link{xgb.load}}.
+
+Use \code{\link{xgb.save.raw}} to save the XGBoost model as a sequence (vector) of raw bytes
+in a future-proof manner. Future releases of XGBoost will be able to read the raw bytes and
+re-construct the corresponding model. To read the model back, use \code{\link{xgb.load.raw}}.
+The \code{\link{xgb.save.raw}} function is useful if you'd like to persist the XGBoost model
+as part of another R object.
+
+Note: Do not use \code{\link{xgb.serialize}} to store models long-term. It persists not only the
+model but also internal configurations and parameters, and its format is not stable across
+multiple XGBoost versions. Use \code{\link{xgb.serialize}} only for checkpointing.
+
+For more details and explanation about model persistence and archival, consult the page
+\url{https://xgboost.readthedocs.io/en/latest/tutorials/saving_model.html}.
+}
+\examples{
+data(agaricus.train, package='xgboost')
+bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2,
+               eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
+
+# Save as a stand-alone file; load it with xgb.load()
+xgb.save(bst, 'xgb.model')
+bst2 <- xgb.load('xgb.model')
+
+# Save as a stand-alone file (JSON); load it with xgb.load()
+xgb.save(bst, 'xgb.model.json')
+bst2 <- xgb.load('xgb.model.json')
+if (file.exists('xgb.model.json')) file.remove('xgb.model.json')
+
+# Save as a raw byte vector; load it with xgb.load.raw()
+xgb_bytes <- xgb.save.raw(bst)
+bst2 <- xgb.load.raw(xgb_bytes)
+
+# Persist XGBoost model as part of another R object
+obj <- list(xgb_model_bytes = xgb.save.raw(bst), description = "My first XGBoost model")
+# Persist the R object. Here, saveRDS() is okay, since it doesn't persist
+# xgb.Booster directly. What's being persisted is the future-proof byte representation
+# as given by xgb.save.raw().
+saveRDS(obj, 'my_object.rds')
+# Read back the R object
+obj2 <- readRDS('my_object.rds')
+# Re-construct xgb.Booster object from the bytes
+bst2 <- xgb.load.raw(obj2$xgb_model_bytes)
+if (file.exists('my_object.rds')) file.remove('my_object.rds')
+
+}

R-package/man/normalize.Rd

@@ -0,0 +1,18 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/xgb.ggplot.R
+\name{normalize}
+\alias{normalize}
+\title{Scale feature value to have mean 0, standard deviation 1}
+\usage{
+normalize(x)
+}
+\arguments{
+\item{x}{Numeric vector}
+}
+\value{
+Numeric vector with mean 0 and sd 1.
+}
+\description{
+This is used to compare multiple features on the same plot.
+Internal utility function
+}

R-package/man/prepare.ggplot.shap.data.Rd

@@ -0,0 +1,27 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/xgb.ggplot.R
+\name{prepare.ggplot.shap.data}
+\alias{prepare.ggplot.shap.data}
+\title{Combine and melt feature values and SHAP contributions for sample
+observations.}
+\usage{
+prepare.ggplot.shap.data(data_list, normalize = FALSE)
+}
+\arguments{
+\item{data_list}{List containing 'data' and 'shap_contrib' returned by
+\code{xgb.shap.data()}.}
+
+\item{normalize}{Whether to standardize feature values to have mean 0 and
+standard deviation 1 (useful for comparing multiple features on the same
+plot). Default \code{FALSE}.}
+}
+\value{
+A data.table containing the observation ID, the feature name, the
+  feature value (normalized if specified), and the SHAP contribution value.
+}
+\description{
+Conforms to data format required for ggplot functions.
+}
+\details{
+Internal utility function.
+}

R-package/man/xgb.Booster.complete.Rd

@@ -38,6 +38,8 @@ bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_dep
                eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
 saveRDS(bst, "xgb.model.rds")
 
+# Warning: The resulting RDS file is only compatible with the current XGBoost version.
+# Refer to the section titled "a-compatibility-note-for-saveRDS-save".
 bst1 <- readRDS("xgb.model.rds")
 if (file.exists("xgb.model.rds")) file.remove("xgb.model.rds")
 # the handle is invalid:

R-package/man/xgb.create.features.Rd

@@ -24,9 +24,9 @@ This is the function inspired from the paragraph 3.1 of the paper:
 
 \strong{Practical Lessons from Predicting Clicks on Ads at Facebook}
 
-\emph{(Xinran He, Junfeng Pan, Ou Jin, Tianbing Xu, Bo Liu, Tao Xu, Yan, xin Shi, Antoine Atallah, Ralf Herbrich, Stuart Bowers, 
+\emph{(Xinran He, Junfeng Pan, Ou Jin, Tianbing Xu, Bo Liu, Tao Xu, Yan, xin Shi, Antoine Atallah, Ralf Herbrich, Stuart Bowers,
 Joaquin Quinonero Candela)}
- 
+
 International Workshop on Data Mining for Online Advertising (ADKDD) - August 24, 2014
 
 \url{https://research.fb.com/publications/practical-lessons-from-predicting-clicks-on-ads-at-facebook/}.
@@ -37,10 +37,10 @@ Extract explaining the method:
 convenient way to implement non-linear and tuple transformations
 of the kind we just described. We treat each individual
 tree as a categorical feature that takes as value the
-index of the leaf an instance ends up falling in. We use 
-1-of-K coding of this type of features. 
+index of the leaf an instance ends up falling in. We use
+1-of-K coding of this type of features.
 
-For example, consider the boosted tree model in Figure 1 with 2 subtrees, 
+For example, consider the boosted tree model in Figure 1 with 2 subtrees,
 where the first subtree has 3 leafs and the second 2 leafs. If an
 instance ends up in leaf 2 in the first subtree and leaf 1 in
 second subtree, the overall input to the linear classifier will

R-package/man/xgb.cv.Rd

@@ -28,12 +28,15 @@ xgb.cv(
 )
 }
 \arguments{
-\item{params}{the list of parameters. Commonly used ones are:
+\item{params}{the list of parameters. The complete list of parameters is
+  available in the \href{http://xgboost.readthedocs.io/en/latest/parameter.html}{online documentation}. Below
+  is a shorter summary:
 \itemize{
   \item \code{objective} objective function, common ones are
   \itemize{
-    \item \code{reg:squarederror} Regression with squared loss
-    \item \code{binary:logistic} logistic regression for classification
+    \item \code{reg:squarederror} Regression with squared loss.
+    \item \code{binary:logistic} logistic regression for classification.
+    \item See \code{\link[=xgb.train]{xgb.train}()} for complete list of objectives.
   }
   \item \code{eta} step size of each boosting step
   \item \code{max_depth} maximum depth of the tree
@@ -67,6 +70,8 @@ from each CV model. This parameter engages the \code{\link{cb.cv.predict}} callb
   \item \code{error} binary classification error rate
   \item \code{rmse} Rooted mean square error
   \item \code{logloss} negative log-likelihood function
+  \item \code{mae} Mean absolute error
+  \item \code{mape} Mean absolute percentage error
   \item \code{auc} Area under curve
   \item \code{aucpr} Area under PR curve
   \item \code{merror} Exact matching error, used to evaluate multi-class classification
@@ -151,7 +156,7 @@ The cross-validation process is then repeated \code{nrounds} times, with each of
 
 All observations are used for both training and validation.
 
-Adapted from \url{http://en.wikipedia.org/wiki/Cross-validation_\%28statistics\%29#k-fold_cross-validation}
+Adapted from \url{https://en.wikipedia.org/wiki/Cross-validation_\%28statistics\%29}
 }
 \examples{
 data(agaricus.train, package='xgboost')

R-package/man/xgb.dump.Rd

@@ -16,14 +16,14 @@ xgb.dump(
 \arguments{
 \item{model}{the model object.}
 
-\item{fname}{the name of the text file where to save the model text dump. 
+\item{fname}{the name of the text file where to save the model text dump.
 If not provided or set to \code{NULL}, the model is returned as a \code{character} vector.}
 
 \item{fmap}{feature map file representing feature types.
-Detailed description could be found at 
+Detailed description could be found at
 \url{https://github.com/dmlc/xgboost/wiki/Binary-Classification#dump-model}.
 See demo/ for walkthrough example in R, and
-\url{https://github.com/dmlc/xgboost/blob/master/demo/data/featmap.txt} 
+\url{https://github.com/dmlc/xgboost/blob/master/demo/data/featmap.txt}
 for example Format.}
 
 \item{with_stats}{whether to dump some additional statistics about the splits.
@@ -47,7 +47,7 @@ data(agaricus.train, package='xgboost')
 data(agaricus.test, package='xgboost')
 train <- agaricus.train
 test <- agaricus.test
-bst <- xgboost(data = train$data, label = train$label, max_depth = 2, 
+bst <- xgboost(data = train$data, label = train$label, max_depth = 2,
                eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
 # save the model in file 'xgb.model.dump'
 dump_path = file.path(tempdir(), 'model.dump')

R-package/man/xgb.importance.Rd

@@ -22,7 +22,7 @@ Non-null \code{feature_names} could be provided to override those in the model.}
 
 \item{trees}{(only for the gbtree booster) an integer vector of tree indices that should be included
 into the importance calculation. If set to \code{NULL}, all trees of the model are parsed.
-It could be useful, e.g., in multiclass classification to get feature importances 
+It could be useful, e.g., in multiclass classification to get feature importances
 for each class separately. IMPORTANT: the tree index in xgboost models
 is zero-based (e.g., use \code{trees = 0:4} for first 5 trees).}
 
@@ -37,7 +37,7 @@ For a tree model, a \code{data.table} with the following columns:
 \itemize{
   \item \code{Features} names of the features used in the model;
   \item \code{Gain} represents fractional contribution of each feature to the model based on
-       the total gain of this feature's splits. Higher percentage means a more important 
+       the total gain of this feature's splits. Higher percentage means a more important
        predictive feature.
   \item \code{Cover} metric of the number of observation related to this feature;
   \item \code{Frequency} percentage representing the relative number of times
@@ -51,7 +51,7 @@ A linear model's importance \code{data.table} has the following columns:
   \item \code{Class} (only for multiclass models) class label.
 }
 
-If \code{feature_names} is not provided and \code{model} doesn't have \code{feature_names}, 
+If \code{feature_names} is not provided and \code{model} doesn't have \code{feature_names},
 index of the features will be used instead. Because the index is extracted from the model dump
 (based on C++ code), it starts at 0 (as in C/C++ or Python) instead of 1 (usual in R).
 }
@@ -61,21 +61,21 @@ Creates a \code{data.table} of feature importances in a model.
 \details{
 This function works for both linear and tree models.
 
-For linear models, the importance is the absolute magnitude of linear coefficients. 
-For that reason, in order to obtain a meaningful ranking by importance for a linear model, 
-the features need to be on the same scale (which you also would want to do when using either 
+For linear models, the importance is the absolute magnitude of linear coefficients.
+For that reason, in order to obtain a meaningful ranking by importance for a linear model,
+the features need to be on the same scale (which you also would want to do when using either
 L1 or L2 regularization).
 }
 \examples{
 
 # binomial classification using gbtree:
 data(agaricus.train, package='xgboost')
-bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2, 
+bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2,
                eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
 xgb.importance(model = bst)
 
 # binomial classification using gblinear:
-bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, booster = "gblinear", 
+bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, booster = "gblinear",
                eta = 0.3, nthread = 1, nrounds = 20, objective = "binary:logistic")
 xgb.importance(model = bst)
 

R-package/man/xgb.load.Rd

@@ -17,8 +17,8 @@ Load xgboost model from the binary model file.
 }
 \details{
 The input file is expected to contain a model saved in an xgboost-internal binary format
-using either \code{\link{xgb.save}} or \code{\link{cb.save.model}} in R, or using some 
-appropriate methods from other xgboost interfaces. E.g., a model trained in Python and 
+using either \code{\link{xgb.save}} or \code{\link{cb.save.model}} in R, or using some
+appropriate methods from other xgboost interfaces. E.g., a model trained in Python and
 saved from there in xgboost format, could be loaded from R.
 
 Note: a model saved as an R-object, has to be loaded using corresponding R-methods,
@@ -29,7 +29,7 @@ data(agaricus.train, package='xgboost')
 data(agaricus.test, package='xgboost')
 train <- agaricus.train
 test <- agaricus.test
-bst <- xgboost(data = train$data, label = train$label, max_depth = 2, 
+bst <- xgboost(data = train$data, label = train$label, max_depth = 2,
                eta = 1, nthread = 2, nrounds = 2,objective = "binary:logistic")
 xgb.save(bst, 'xgb.model')
 bst <- xgb.load('xgb.model')

R-package/man/xgb.model.dt.tree.Rd

@@ -20,7 +20,7 @@ Non-null \code{feature_names} could be provided to override those in the model.}
 
 \item{model}{object of class \code{xgb.Booster}}
 
-\item{text}{\code{character} vector previously generated by the \code{xgb.dump} 
+\item{text}{\code{character} vector previously generated by the \code{xgb.dump}
 function  (where parameter \code{with_stats = TRUE} should have been set).
 \code{text} takes precedence over \code{model}.}
 
@@ -53,10 +53,10 @@ The columns of the \code{data.table} are:
  \item \code{Quality}: either the split gain (change in loss) or the leaf value
  \item \code{Cover}: metric related to the number of observation either seen by a split
                      or collected by a leaf during training.
-} 
+}
 
 When \code{use_int_id=FALSE}, columns "Yes", "No", and "Missing" point to model-wide node identifiers
-in the "ID" column. When \code{use_int_id=TRUE}, those columns point to node identifiers from 
+in the "ID" column. When \code{use_int_id=TRUE}, those columns point to node identifiers from
 the corresponding trees in the "Node" column.
 }
 \description{
@@ -67,17 +67,17 @@ Parse a boosted tree model text dump into a \code{data.table} structure.
 
 data(agaricus.train, package='xgboost')
 
-bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2, 
+bst <- xgboost(data = agaricus.train$data, label = agaricus.train$label, max_depth = 2,
                eta = 1, nthread = 2, nrounds = 2,objective = "binary:logistic")
 
 (dt <- xgb.model.dt.tree(colnames(agaricus.train$data), bst))
 
-# This bst model already has feature_names stored with it, so those would be used when 
+# This bst model already has feature_names stored with it, so those would be used when
 # feature_names is not set:
 (dt <- xgb.model.dt.tree(model = bst))
 
 # How to match feature names of splits that are following a current 'Yes' branch:
 
 merge(dt, dt[, .(ID, Y.Feature=Feature)], by.x='Yes', by.y='ID', all.x=TRUE)[order(Tree,Node)]
- 
+
 }

R-package/man/xgb.plot.deepness.Rd

@@ -23,7 +23,7 @@ or a data.table result of the \code{xgb.model.dt.tree} function.}
 
 \item{which}{which distribution to plot (see details).}
 
-\item{plot}{(base R barplot) whether a barplot should be produced. 
+\item{plot}{(base R barplot) whether a barplot should be produced.
 If FALSE, only a data.table is returned.}
 
 \item{...}{other parameters passed to \code{barplot} or \code{plot}.}
@@ -45,10 +45,10 @@ When \code{which="2x1"}, two distributions with respect to the leaf depth
 are plotted on top of each other:
 \itemize{
  \item the distribution of the number of leafs in a tree model at a certain depth;
- \item the distribution of average weighted number of observations ("cover") 
+ \item the distribution of average weighted number of observations ("cover")
        ending up in leafs at certain depth.
 }
-Those could be helpful in determining sensible ranges of the \code{max_depth} 
+Those could be helpful in determining sensible ranges of the \code{max_depth}
 and \code{min_child_weight} parameters.
 
 When \code{which="max.depth"} or \code{which="med.depth"}, plots of either maximum or median depth

R-package/man/xgb.plot.shap.Rd

@@ -131,6 +131,7 @@ bst <- xgboost(agaricus.train$data, agaricus.train$label, nrounds = 50,
 xgb.plot.shap(agaricus.test$data, model = bst, features = "odor=none")
 contr <- predict(bst, agaricus.test$data, predcontrib = TRUE)
 xgb.plot.shap(agaricus.test$data, contr, model = bst, top_n = 12, n_col = 3)
+xgb.ggplot.shap.summary(agaricus.test$data, contr, model = bst, top_n = 12)  # Summary plot
 
 # multiclass example - plots for each class separately:
 nclass <- 3
@@ -149,6 +150,7 @@ xgb.plot.shap(x, model = mbst, trees = trees0 + 1, target_class = 1, top_n = 4,
               n_col = 2, col = col, pch = 16, pch_NA = 17)
 xgb.plot.shap(x, model = mbst, trees = trees0 + 2, target_class = 2, top_n = 4,
               n_col = 2, col = col, pch = 16, pch_NA = 17)
+xgb.ggplot.shap.summary(x, model = mbst, target_class = 0, top_n = 4)  # Summary plot
 
 }
 \references{

R-package/man/xgb.plot.shap.summary.Rd

@@ -0,0 +1,78 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/xgb.ggplot.R, R/xgb.plot.shap.R
+\name{xgb.ggplot.shap.summary}
+\alias{xgb.ggplot.shap.summary}
+\alias{xgb.plot.shap.summary}
+\title{SHAP contribution dependency summary plot}
+\usage{
+xgb.ggplot.shap.summary(
+  data,
+  shap_contrib = NULL,
+  features = NULL,
+  top_n = 10,
+  model = NULL,
+  trees = NULL,
+  target_class = NULL,
+  approxcontrib = FALSE,
+  subsample = NULL
+)
+
+xgb.plot.shap.summary(
+  data,
+  shap_contrib = NULL,
+  features = NULL,
+  top_n = 10,
+  model = NULL,
+  trees = NULL,
+  target_class = NULL,
+  approxcontrib = FALSE,
+  subsample = NULL
+)
+}
+\arguments{
+\item{data}{data as a \code{matrix} or \code{dgCMatrix}.}
+
+\item{shap_contrib}{a matrix of SHAP contributions that was computed earlier for the above
+\code{data}. When it is NULL, it is computed internally using \code{model} and \code{data}.}
+
+\item{features}{a vector of either column indices or of feature names to plot. When it is NULL,
+feature importance is calculated, and \code{top_n} high ranked features are taken.}
+
+\item{top_n}{when \code{features} is NULL, top_n [1, 100] most important features in a model are taken.}
+
+\item{model}{an \code{xgb.Booster} model. It has to be provided when either \code{shap_contrib}
+or \code{features} is missing.}
+
+\item{trees}{passed to \code{\link{xgb.importance}} when \code{features = NULL}.}
+
+\item{target_class}{is only relevant for multiclass models. When it is set to a 0-based class index,
+only SHAP contributions for that specific class are used.
+If it is not set, SHAP importances are averaged over all classes.}
+
+\item{approxcontrib}{passed to \code{\link{predict.xgb.Booster}} when \code{shap_contrib = NULL}.}
+
+\item{subsample}{a random fraction of data points to use for plotting. When it is NULL,
+it is set so that up to 100K data points are used.}
+}
+\value{
+A \code{ggplot2} object.
+}
+\description{
+Compare SHAP contributions of different features.
+}
+\details{
+A point plot (each point representing one sample from \code{data}) is
+produced for each feature, with the points plotted on the SHAP value axis.
+Each point (observation) is coloured based on its feature value. The plot
+hence allows us to see which features have a negative / positive contribution
+on the model prediction, and whether the contribution is different for larger
+or smaller values of the feature. We effectively try to replicate the
+\code{summary_plot} function from https://github.com/slundberg/shap.
+}
+\examples{
+# See \code{\link{xgb.plot.shap}}.
+}
+\seealso{
+\code{\link{xgb.plot.shap}}, \code{\link{xgb.ggplot.shap.summary}},
+  \url{https://github.com/slundberg/shap}
+}

R-package/man/xgb.plot.tree.Rd

@@ -60,7 +60,7 @@ The content of each node is organised that way:
  \item \code{Gain} (for split nodes): the information gain metric of a split
        (corresponds to the importance of the node in the model).
  \item \code{Value} (for leafs): the margin value that the leaf may contribute to prediction.
-} 
+}
 The tree root nodes also indicate the Tree index (0-based).
 
 The "Yes" branches are marked by the "< split_value" label.
@@ -80,7 +80,7 @@ xgb.plot.tree(model = bst)
 xgb.plot.tree(model = bst, trees = 0, show_node_id = TRUE)
 
 \dontrun{
-# Below is an example of how to save this plot to a file. 
+# Below is an example of how to save this plot to a file.
 # Note that for `export_graph` to work, the DiagrammeRsvg and rsvg packages must also be installed.
 library(DiagrammeR)
 gr <- xgb.plot.tree(model=bst, trees=0:1, render=FALSE)

R-package/man/xgb.save.Rd

@@ -15,21 +15,25 @@ xgb.save(model, fname)
 Save xgboost model to a file in binary format.
 }
 \details{
-This methods allows to save a model in an xgboost-internal binary format which is universal 
+This methods allows to save a model in an xgboost-internal binary format which is universal
 among the various xgboost interfaces. In R, the saved model file could be read-in later
-using either the \code{\link{xgb.load}} function or the \code{xgb_model} parameter 
+using either the \code{\link{xgb.load}} function or the \code{xgb_model} parameter
 of \code{\link{xgb.train}}.
 
-Note: a model can also be saved as an R-object (e.g., by using \code{\link[base]{readRDS}} 
-or \code{\link[base]{save}}). However, it would then only be compatible with R, and 
-corresponding R-methods would need to be used to load it.
+Note: a model can also be saved as an R-object (e.g., by using \code{\link[base]{readRDS}}
+or \code{\link[base]{save}}). However, it would then only be compatible with R, and
+corresponding R-methods would need to be used to load it. Moreover, persisting the model with
+\code{\link[base]{readRDS}} or \code{\link[base]{save}}) will cause compatibility problems in
+future versions of XGBoost. Consult \code{\link{a-compatibility-note-for-saveRDS-save}} to learn
+how to persist models in a future-proof way, i.e. to make the model accessible in future
+releases of XGBoost.
 }
 \examples{
 data(agaricus.train, package='xgboost')
 data(agaricus.test, package='xgboost')
 train <- agaricus.train
 test <- agaricus.test
-bst <- xgboost(data = train$data, label = train$label, max_depth = 2, 
+bst <- xgboost(data = train$data, label = train$label, max_depth = 2,
                eta = 1, nthread = 2, nrounds = 2,objective = "binary:logistic")
 xgb.save(bst, 'xgb.model')
 bst <- xgb.load('xgb.model')

R-package/man/xgb.shap.data.Rd

@@ -0,0 +1,55 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/xgb.plot.shap.R
+\name{xgb.shap.data}
+\alias{xgb.shap.data}
+\title{Prepare data for SHAP plots. To be used in xgb.plot.shap, xgb.plot.shap.summary, etc.
+Internal utility function.}
+\usage{
+xgb.shap.data(
+  data,
+  shap_contrib = NULL,
+  features = NULL,
+  top_n = 1,
+  model = NULL,
+  trees = NULL,
+  target_class = NULL,
+  approxcontrib = FALSE,
+  subsample = NULL,
+  max_observations = 1e+05
+)
+}
+\arguments{
+\item{data}{data as a \code{matrix} or \code{dgCMatrix}.}
+
+\item{shap_contrib}{a matrix of SHAP contributions that was computed earlier for the above
+\code{data}. When it is NULL, it is computed internally using \code{model} and \code{data}.}
+
+\item{features}{a vector of either column indices or of feature names to plot. When it is NULL,
+feature importance is calculated, and \code{top_n} high ranked features are taken.}
+
+\item{top_n}{when \code{features} is NULL, top_n [1, 100] most important features in a model are taken.}
+
+\item{model}{an \code{xgb.Booster} model. It has to be provided when either \code{shap_contrib}
+or \code{features} is missing.}
+
+\item{trees}{passed to \code{\link{xgb.importance}} when \code{features = NULL}.}
+
+\item{target_class}{is only relevant for multiclass models. When it is set to a 0-based class index,
+only SHAP contributions for that specific class are used.
+If it is not set, SHAP importances are averaged over all classes.}
+
+\item{approxcontrib}{passed to \code{\link{predict.xgb.Booster}} when \code{shap_contrib = NULL}.}
+
+\item{subsample}{a random fraction of data points to use for plotting. When it is NULL,
+it is set so that up to 100K data points are used.}
+}
+\value{
+A list containing: 'data', a matrix containing sample observations
+  and their feature values; 'shap_contrib', a matrix containing the SHAP contribution
+  values for these observations.
+}
+\description{
+Prepare data for SHAP plots. To be used in xgb.plot.shap, xgb.plot.shap.summary, etc.
+Internal utility function.
+}
+\keyword{internal}

R-package/man/xgb.train.Rd

@@ -42,9 +42,9 @@ xgboost(
 )
 }
 \arguments{
-\item{params}{the list of parameters.
-       The complete list of parameters is available at \url{http://xgboost.readthedocs.io/en/latest/parameter.html}.
-       Below is a shorter summary:
+\item{params}{the list of parameters. The complete list of parameters is
+  available in the \href{http://xgboost.readthedocs.io/en/latest/parameter.html}{online documentation}. Below
+  is a shorter summary:
 
 1. General Parameters
 
@@ -82,13 +82,23 @@ xgboost(
 \item \code{objective} specify the learning task and the corresponding learning objective, users can pass a self-defined function to it. The default objective options are below:
   \itemize{
     \item \code{reg:squarederror} Regression with squared loss (Default).
+    \item \code{reg:squaredlogerror}: regression with squared log loss \eqn{1/2 * (log(pred + 1) - log(label + 1))^2}. All inputs are required to be greater than -1. Also, see metric rmsle for possible issue with this objective.
     \item \code{reg:logistic} logistic regression.
+    \item \code{reg:pseudohubererror}: regression with Pseudo Huber loss, a twice differentiable alternative to absolute loss.
     \item \code{binary:logistic} logistic regression for binary classification. Output probability.
     \item \code{binary:logitraw} logistic regression for binary classification, output score before logistic transformation.
-    \item \code{num_class} set the number of classes. To use only with multiclass objectives.
+    \item \code{binary:hinge}: hinge loss for binary classification. This makes predictions of 0 or 1, rather than producing probabilities.
+    \item \code{count:poisson}: poisson regression for count data, output mean of poisson distribution. \code{max_delta_step} is set to 0.7 by default in poisson regression (used to safeguard optimization).
+    \item \code{survival:cox}: Cox regression for right censored survival time data (negative values are considered right censored). Note that predictions are returned on the hazard ratio scale (i.e., as HR = exp(marginal_prediction) in the proportional hazard function \code{h(t) = h0(t) * HR)}.
+    \item \code{survival:aft}: Accelerated failure time model for censored survival time data. See \href{https://xgboost.readthedocs.io/en/latest/tutorials/aft_survival_analysis.html}{Survival Analysis with Accelerated Failure Time} for details.
+    \item \code{aft_loss_distribution}: Probabilty Density Function used by \code{survival:aft} and \code{aft-nloglik} metric.
     \item \code{multi:softmax} set xgboost to do multiclass classification using the softmax objective. Class is represented by a number and should be from 0 to \code{num_class - 1}.
     \item \code{multi:softprob} same as softmax, but prediction outputs a vector of ndata * nclass elements, which can be further reshaped to ndata, nclass matrix. The result contains predicted probabilities of each data point belonging to each class.
     \item \code{rank:pairwise} set xgboost to do ranking task by minimizing the pairwise loss.
+    \item \code{rank:ndcg}: Use LambdaMART to perform list-wise ranking where \href{https://en.wikipedia.org/wiki/Discounted_cumulative_gain}{Normalized Discounted Cumulative Gain (NDCG)} is maximized.
+    \item \code{rank:map}: Use LambdaMART to perform list-wise ranking where \href{https://en.wikipedia.org/wiki/Evaluation_measures_(information_retrieval)#Mean_average_precision}{Mean Average Precision (MAP)} is maximized.
+    \item \code{reg:gamma}: gamma regression with log-link. Output is a mean of gamma distribution. It might be useful, e.g., for modeling insurance claims severity, or for any outcome that might be \href{https://en.wikipedia.org/wiki/Gamma_distribution#Applications}{gamma-distributed}.
+    \item \code{reg:tweedie}: Tweedie regression with log-link. It might be useful, e.g., for modeling total loss in insurance, or for any outcome that might be \href{https://en.wikipedia.org/wiki/Tweedie_distribution#Applications}{Tweedie-distributed}.
   }
   \item \code{base_score} the initial prediction score of all instances, global bias. Default: 0.5
   \item \code{eval_metric} evaluation metrics for validation data. Users can pass a self-defined function to it. Default: metric will be assigned according to objective(rmse for regression, and error for classification, mean average precision for ranking). List is provided in detail section.
@@ -205,16 +215,18 @@ User may set one or several \code{eval_metric} parameters.
 Note that when using a customized metric, only this single metric can be used.
 The following is the list of built-in metrics for which Xgboost provides optimized implementation:
   \itemize{
-     \item \code{rmse} root mean square error. \url{http://en.wikipedia.org/wiki/Root_mean_square_error}
-     \item \code{logloss} negative log-likelihood. \url{http://en.wikipedia.org/wiki/Log-likelihood}
-     \item \code{mlogloss} multiclass logloss. \url{http://wiki.fast.ai/index.php/Log_Loss}
+     \item \code{rmse} root mean square error. \url{https://en.wikipedia.org/wiki/Root_mean_square_error}
+     \item \code{logloss} negative log-likelihood. \url{https://en.wikipedia.org/wiki/Log-likelihood}
+     \item \code{mlogloss} multiclass logloss. \url{https://scikit-learn.org/stable/modules/generated/sklearn.metrics.log_loss.html}
      \item \code{error} Binary classification error rate. It is calculated as \code{(# wrong cases) / (# all cases)}.
            By default, it uses the 0.5 threshold for predicted values to define negative and positive instances.
            Different threshold (e.g., 0.) could be specified as "error@0."
      \item \code{merror} Multiclass classification error rate. It is calculated as \code{(# wrong cases) / (# all cases)}.
-     \item \code{auc} Area under the curve. \url{http://en.wikipedia.org/wiki/Receiver_operating_characteristic#'Area_under_curve} for ranking evaluation.
+     \item \code{mae} Mean absolute error
+     \item \code{mape} Mean absolute percentage error
+     \item \code{auc} Area under the curve. \url{https://en.wikipedia.org/wiki/Receiver_operating_characteristic#'Area_under_curve} for ranking evaluation.
      \item \code{aucpr} Area under the PR curve. \url{https://en.wikipedia.org/wiki/Precision_and_recall} for ranking evaluation.
-     \item \code{ndcg} Normalized Discounted Cumulative Gain (for ranking task). \url{http://en.wikipedia.org/wiki/NDCG}
+     \item \code{ndcg} Normalized Discounted Cumulative Gain (for ranking task). \url{https://en.wikipedia.org/wiki/NDCG}
   }
 
 The following callbacks are automatically created when certain parameters are set:

R-package/man/xgbConfig.Rd

@@ -0,0 +1,39 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/xgb.config.R
+\name{xgb.set.config, xgb.get.config}
+\alias{xgb.set.config, xgb.get.config}
+\alias{xgb.set.config}
+\alias{xgb.get.config}
+\title{Set and get global configuration}
+\usage{
+xgb.set.config(...)
+
+xgb.get.config()
+}
+\arguments{
+\item{...}{List of parameters to be set, as keyword arguments}
+}
+\value{
+\code{xgb.set.config} returns \code{TRUE} to signal success. \code{xgb.get.config} returns
+a list containing all global-scope parameters and their values.
+}
+\description{
+Global configuration consists of a collection of parameters that can be applied in the global
+scope. See \url{https://xgboost.readthedocs.io/en/stable/parameter.html} for the full list of
+parameters supported in the global configuration. Use \code{xgb.set.config} to update the
+values of one or more global-scope parameters. Use \code{xgb.get.config} to fetch the current
+values of all global-scope parameters (listed in
+\url{https://xgboost.readthedocs.io/en/stable/parameter.html}).
+}
+\examples{
+# Set verbosity level to silent (0)
+xgb.set.config(verbosity = 0)
+# Now global verbosity level is 0
+config <- xgb.get.config()
+print(config$verbosity)
+# Set verbosity level to warning (1)
+xgb.set.config(verbosity = 1)
+# Now global verbosity level is 1
+config <- xgb.get.config()
+print(config$verbosity)
+}

R-package/src/Makevars.in

@@ -3,12 +3,12 @@ PKGROOT=../../
 ENABLE_STD_THREAD=1
 # _*_ mode: Makefile; _*_
 
-CXX_STD = CXX11
+CXX_STD = CXX14
 
 XGB_RFLAGS = -DXGBOOST_STRICT_R_MODE=1 -DDMLC_LOG_BEFORE_THROW=0\
            -DDMLC_ENABLE_STD_THREAD=$(ENABLE_STD_THREAD) -DDMLC_DISABLE_STDIN=1\
            -DDMLC_LOG_CUSTOMIZE=1 -DXGBOOST_CUSTOMIZE_LOGGER=1\
-           -DRABIT_CUSTOMIZE_MSG_ -DRABIT_STRICT_CXX98_
+           -DRABIT_CUSTOMIZE_MSG_
 
 # disable the use of thread_local for 32 bit windows:
 ifeq ($(R_OSTYPE)$(WIN),windows)
@@ -19,6 +19,7 @@ $(foreach v, $(XGB_RFLAGS), $(warning $(v)))
 PKG_CPPFLAGS=  -I$(PKGROOT)/include -I$(PKGROOT)/dmlc-core/include -I$(PKGROOT)/rabit/include -I$(PKGROOT) $(XGB_RFLAGS)
 PKG_CXXFLAGS= @OPENMP_CXXFLAGS@ @ENDIAN_FLAG@ -pthread
 PKG_LIBS = @OPENMP_CXXFLAGS@ @OPENMP_LIB@ @ENDIAN_FLAG@ @BACKTRACE_LIB@ -pthread
-OBJECTS= ./xgboost_R.o ./xgboost_custom.o ./xgboost_assert.o ./init.o\
-         $(PKGROOT)/amalgamation/xgboost-all0.o $(PKGROOT)/amalgamation/dmlc-minimum0.o\
-         $(PKGROOT)/rabit/src/engine_empty.o $(PKGROOT)/rabit/src/c_api.o
+OBJECTS= ./xgboost_R.o ./xgboost_custom.o ./xgboost_assert.o ./init.o \
+         $(PKGROOT)/amalgamation/xgboost-all0.o $(PKGROOT)/amalgamation/dmlc-minimum0.o \
+         $(PKGROOT)/rabit/src/engine.o $(PKGROOT)/rabit/src/c_api.o \
+         $(PKGROOT)/rabit/src/allreduce_base.o

R-package/src/Makevars.win

@@ -15,12 +15,12 @@ xgblib:
 	cp -r ../../include .
 	cp -r ../../amalgamation .
 
-CXX_STD = CXX11
+CXX_STD = CXX14
 
 XGB_RFLAGS = -DXGBOOST_STRICT_R_MODE=1 -DDMLC_LOG_BEFORE_THROW=0\
            -DDMLC_ENABLE_STD_THREAD=$(ENABLE_STD_THREAD) -DDMLC_DISABLE_STDIN=1\
            -DDMLC_LOG_CUSTOMIZE=1 -DXGBOOST_CUSTOMIZE_LOGGER=1\
-           -DRABIT_CUSTOMIZE_MSG_ -DRABIT_STRICT_CXX98_
+           -DRABIT_CUSTOMIZE_MSG_
 
 # disable the use of thread_local for 32 bit windows:
 ifeq ($(R_OSTYPE)$(WIN),windows)
@@ -31,8 +31,9 @@ $(foreach v, $(XGB_RFLAGS), $(warning $(v)))
 PKG_CPPFLAGS=  -I$(PKGROOT)/include -I$(PKGROOT)/dmlc-core/include -I$(PKGROOT)/rabit/include -I$(PKGROOT) $(XGB_RFLAGS)
 PKG_CXXFLAGS= $(SHLIB_OPENMP_CXXFLAGS) $(SHLIB_PTHREAD_FLAGS)
 PKG_LIBS = $(SHLIB_OPENMP_CXXFLAGS) $(SHLIB_PTHREAD_FLAGS)
-OBJECTS= ./xgboost_R.o ./xgboost_custom.o ./xgboost_assert.o ./init.o\
-         $(PKGROOT)/amalgamation/xgboost-all0.o $(PKGROOT)/amalgamation/dmlc-minimum0.o\
-         $(PKGROOT)/rabit/src/engine_empty.o $(PKGROOT)/rabit/src/c_api.o
+OBJECTS= ./xgboost_R.o ./xgboost_custom.o ./xgboost_assert.o ./init.o \
+         $(PKGROOT)/amalgamation/xgboost-all0.o $(PKGROOT)/amalgamation/dmlc-minimum0.o \
+         $(PKGROOT)/rabit/src/engine.o $(PKGROOT)/rabit/src/c_api.o \
+         $(PKGROOT)/rabit/src/allreduce_base.o
 
 $(OBJECTS) : xgblib

R-package/src/init.c

@@ -43,6 +43,8 @@ extern SEXP XGDMatrixNumRow_R(SEXP);
 extern SEXP XGDMatrixSaveBinary_R(SEXP, SEXP, SEXP);
 extern SEXP XGDMatrixSetInfo_R(SEXP, SEXP, SEXP);
 extern SEXP XGDMatrixSliceDMatrix_R(SEXP, SEXP);
+extern SEXP XGBSetGlobalConfig_R(SEXP);
+extern SEXP XGBGetGlobalConfig_R();
 
 static const R_CallMethodDef CallEntries[] = {
   {"XGBoosterBoostOneIter_R",     (DL_FUNC) &XGBoosterBoostOneIter_R,     4},
@@ -73,6 +75,8 @@ static const R_CallMethodDef CallEntries[] = {
   {"XGDMatrixSaveBinary_R",       (DL_FUNC) &XGDMatrixSaveBinary_R,       3},
   {"XGDMatrixSetInfo_R",          (DL_FUNC) &XGDMatrixSetInfo_R,          3},
   {"XGDMatrixSliceDMatrix_R",     (DL_FUNC) &XGDMatrixSliceDMatrix_R,     2},
+  {"XGBSetGlobalConfig_R",        (DL_FUNC) &XGBSetGlobalConfig_R,        1},
+  {"XGBGetGlobalConfig_R",        (DL_FUNC) &XGBGetGlobalConfig_R,        0},
   {NULL, NULL, 0}
 };
 

R-package/src/xgboost_R.cc

@@ -1,6 +1,7 @@
 // Copyright (c) 2014 by Contributors
 #include <dmlc/logging.h>
 #include <dmlc/omp.h>
+#include <dmlc/common.h>
 #include <xgboost/c_api.h>
 #include <vector>
 #include <string>
@@ -49,6 +50,21 @@ void _DMatrixFinalizer(SEXP ext) {
   R_API_END();
 }
 
+SEXP XGBSetGlobalConfig_R(SEXP json_str) {
+  R_API_BEGIN();
+  CHECK_CALL(XGBSetGlobalConfig(CHAR(asChar(json_str))));
+  R_API_END();
+  return R_NilValue;
+}
+
+SEXP XGBGetGlobalConfig_R() {
+  const char* json_str;
+  R_API_BEGIN();
+  CHECK_CALL(XGBGetGlobalConfig(&json_str));
+  R_API_END();
+  return mkString(json_str);
+}
+
 SEXP XGDMatrixCreateFromFile_R(SEXP fname, SEXP silent) {
   SEXP ret;
   R_API_BEGIN();
@@ -77,12 +93,16 @@ SEXP XGDMatrixCreateFromMat_R(SEXP mat,
     din = REAL(mat);
   }
   std::vector<float> data(nrow * ncol);
+  dmlc::OMPException exc;
   #pragma omp parallel for schedule(static)
   for (omp_ulong i = 0; i < nrow; ++i) {
-    for (size_t j = 0; j < ncol; ++j) {
-      data[i * ncol +j] = is_int ? static_cast<float>(iin[i + nrow * j]) : din[i + nrow * j];
-    }
+    exc.Run([&]() {
+      for (size_t j = 0; j < ncol; ++j) {
+        data[i * ncol +j] = is_int ? static_cast<float>(iin[i + nrow * j]) : din[i + nrow * j];
+      }
+    });
   }
+  exc.Rethrow();
   DMatrixHandle handle;
   CHECK_CALL(XGDMatrixCreateFromMat(BeginPtr(data), nrow, ncol, asReal(missing), &handle));
   ret = PROTECT(R_MakeExternalPtr(handle, R_NilValue, R_NilValue));
@@ -111,11 +131,15 @@ SEXP XGDMatrixCreateFromCSC_R(SEXP indptr,
   for (size_t i = 0; i < nindptr; ++i) {
     col_ptr_[i] = static_cast<size_t>(p_indptr[i]);
   }
+  dmlc::OMPException exc;
   #pragma omp parallel for schedule(static)
   for (int64_t i = 0; i < static_cast<int64_t>(ndata); ++i) {
-    indices_[i] = static_cast<unsigned>(p_indices[i]);
-    data_[i] = static_cast<float>(p_data[i]);
+    exc.Run([&]() {
+      indices_[i] = static_cast<unsigned>(p_indices[i]);
+      data_[i] = static_cast<float>(p_data[i]);
+    });
   }
+  exc.Rethrow();
   DMatrixHandle handle;
   CHECK_CALL(XGDMatrixCreateFromCSCEx(BeginPtr(col_ptr_), BeginPtr(indices_),
                                       BeginPtr(data_), nindptr, ndata,
@@ -160,12 +184,16 @@ SEXP XGDMatrixSetInfo_R(SEXP handle, SEXP field, SEXP array) {
   R_API_BEGIN();
   int len = length(array);
   const char *name = CHAR(asChar(field));
+  dmlc::OMPException exc;
   if (!strcmp("group", name)) {
     std::vector<unsigned> vec(len);
     #pragma omp parallel for schedule(static)
     for (int i = 0; i < len; ++i) {
-      vec[i] = static_cast<unsigned>(INTEGER(array)[i]);
+      exc.Run([&]() {
+        vec[i] = static_cast<unsigned>(INTEGER(array)[i]);
+      });
     }
+    exc.Rethrow();
     CHECK_CALL(XGDMatrixSetUIntInfo(R_ExternalPtrAddr(handle),
                                     CHAR(asChar(field)),
                                     BeginPtr(vec), len));
@@ -173,8 +201,11 @@ SEXP XGDMatrixSetInfo_R(SEXP handle, SEXP field, SEXP array) {
     std::vector<float> vec(len);
     #pragma omp parallel for schedule(static)
     for (int i = 0; i < len; ++i) {
-      vec[i] = REAL(array)[i];
+      exc.Run([&]() {
+        vec[i] = REAL(array)[i];
+      });
     }
+    exc.Rethrow();
     CHECK_CALL(XGDMatrixSetFloatInfo(R_ExternalPtrAddr(handle),
                                      CHAR(asChar(field)),
                                      BeginPtr(vec), len));
@@ -265,11 +296,15 @@ SEXP XGBoosterBoostOneIter_R(SEXP handle, SEXP dtrain, SEXP grad, SEXP hess) {
       << "gradient and hess must have same length";
   int len = length(grad);
   std::vector<float> tgrad(len), thess(len);
+  dmlc::OMPException exc;
   #pragma omp parallel for schedule(static)
   for (int j = 0; j < len; ++j) {
-    tgrad[j] = REAL(grad)[j];
-    thess[j] = REAL(hess)[j];
+    exc.Run([&]() {
+      tgrad[j] = REAL(grad)[j];
+      thess[j] = REAL(hess)[j];
+    });
   }
+  exc.Rethrow();
   CHECK_CALL(XGBoosterBoostOneIter(R_ExternalPtrAddr(handle),
                                  R_ExternalPtrAddr(dtrain),
                                  BeginPtr(tgrad), BeginPtr(thess),
@@ -375,7 +410,7 @@ SEXP XGBoosterSaveJsonConfig_R(SEXP handle) {
 
 SEXP XGBoosterLoadJsonConfig_R(SEXP handle, SEXP value) {
   R_API_BEGIN();
-  XGBoosterLoadJsonConfig(R_ExternalPtrAddr(handle), CHAR(asChar(value)));
+  CHECK_CALL(XGBoosterLoadJsonConfig(R_ExternalPtrAddr(handle), CHAR(asChar(value))));
   R_API_END();
   return R_NilValue;
 }
@@ -397,9 +432,9 @@ SEXP XGBoosterSerializeToBuffer_R(SEXP handle) {
 
 SEXP XGBoosterUnserializeFromBuffer_R(SEXP handle, SEXP raw) {
   R_API_BEGIN();
-  XGBoosterUnserializeFromBuffer(R_ExternalPtrAddr(handle),
+  CHECK_CALL(XGBoosterUnserializeFromBuffer(R_ExternalPtrAddr(handle),
                                  RAW(raw),
-                                 length(raw));
+                                 length(raw)));
   R_API_END();
   return R_NilValue;
 }

R-package/src/xgboost_R.h

@@ -21,6 +21,19 @@
  */
 XGB_DLL SEXP XGCheckNullPtr_R(SEXP handle);
 
+/*!
+ * \brief Set global configuration
+ * \param json_str a JSON string representing the list of key-value pairs
+ * \return R_NilValue
+ */
+XGB_DLL SEXP XGBSetGlobalConfig_R(SEXP json_str);
+
+/*!
+ * \brief Get global configuration
+ * \return JSON string
+ */
+XGB_DLL SEXP XGBGetGlobalConfig_R();
+
 /*!
  * \brief load a data matrix
  * \param fname name of the content

R-package/src/xgboost_custom.cc

@@ -13,27 +13,10 @@ void CustomLogMessage::Log(const std::string& msg) {
 }
 }  // namespace dmlc
 
-// implements rabit error handling.
-extern "C" {
-  void XGBoostAssert_R(int exp, const char *fmt, ...);
-  void XGBoostCheck_R(int exp, const char *fmt, ...);
-}
-
-namespace rabit {
-namespace utils {
-extern "C" {
-  void (*Printf)(const char *fmt, ...) = Rprintf;
-  void (*Assert)(int exp, const char *fmt, ...) = XGBoostAssert_R;
-  void (*Check)(int exp, const char *fmt, ...) = XGBoostCheck_R;
-  void (*Error)(const char *fmt, ...) = error;
-}
-}
-}
-
 namespace xgboost {
 ConsoleLogger::~ConsoleLogger() {
   if (cur_verbosity_ == LogVerbosity::kIgnore ||
-      cur_verbosity_ <= global_verbosity_) {
+      cur_verbosity_ <= GlobalVerbosity()) {
     dmlc::CustomLogMessage::Log(log_stream_.str());
   }
 }

R-package/tests/helper_scripts/generate_models.R

@@ -0,0 +1,105 @@
+# Script to generate reference models. The reference models are used to test backward compatibility
+# of saved model files from XGBoost version 0.90 and 1.0.x.
+library(xgboost)
+library(Matrix)
+
+set.seed(0)
+metadata <- list(
+  kRounds = 2,
+  kRows = 1000,
+  kCols = 4,
+  kForests = 2,
+  kMaxDepth = 2,
+  kClasses = 3
+)
+X <- Matrix(data = rnorm(metadata$kRows * metadata$kCols), nrow = metadata$kRows,
+            ncol = metadata$kCols, sparse = TRUE)
+w <- runif(metadata$kRows)
+
+version <- packageVersion('xgboost')
+target_dir <- 'models'
+
+save_booster <- function (booster, model_name) {
+  booster_bin <- function (model_name) {
+    return (file.path(target_dir, paste('xgboost-', version, '.', model_name, '.bin', sep = '')))
+  }
+  booster_json <- function (model_name) {
+    return (file.path(target_dir, paste('xgboost-', version, '.', model_name, '.json', sep = '')))
+  }
+  booster_rds <- function (model_name) {
+    return (file.path(target_dir, paste('xgboost-', version, '.', model_name, '.rds', sep = '')))
+  }
+  xgb.save(booster, booster_bin(model_name))
+  saveRDS(booster, booster_rds(model_name))
+  if (version >= '1.0.0') {
+    xgb.save(booster, booster_json(model_name))
+  }
+}
+
+generate_regression_model <- function () {
+  print('Regression')
+  y <- rnorm(metadata$kRows)
+
+  data <- xgb.DMatrix(X, label = y)
+  params <- list(tree_method = 'hist', num_parallel_tree = metadata$kForests,
+                 max_depth = metadata$kMaxDepth)
+  booster <- xgb.train(params, data, nrounds = metadata$kRounds)
+  save_booster(booster, 'reg')
+}
+
+generate_logistic_model <- function () {
+  print('Binary classification with logistic loss')
+  y <- sample(0:1, size = metadata$kRows, replace = TRUE)
+  stopifnot(max(y) == 1, min(y) == 0)
+
+  objective <- c('binary:logistic', 'binary:logitraw')
+  name <- c('logit', 'logitraw')
+
+  for (i in seq_len(length(objective))) {
+    data <- xgb.DMatrix(X, label = y, weight = w)
+    params <- list(tree_method = 'hist', num_parallel_tree = metadata$kForests,
+                   max_depth = metadata$kMaxDepth, objective = objective[i])
+    booster <- xgb.train(params, data, nrounds = metadata$kRounds)
+    save_booster(booster, name[i])
+  }
+}
+
+generate_classification_model <- function () {
+  print('Multi-class classification')
+  y <- sample(0:(metadata$kClasses - 1), size = metadata$kRows, replace = TRUE)
+  stopifnot(max(y) == metadata$kClasses - 1, min(y) == 0)
+
+  data <- xgb.DMatrix(X, label = y, weight = w)
+  params <- list(num_class = metadata$kClasses, tree_method = 'hist',
+                 num_parallel_tree = metadata$kForests, max_depth = metadata$kMaxDepth,
+                 objective = 'multi:softmax')
+  booster <- xgb.train(params, data, nrounds = metadata$kRounds)
+  save_booster(booster, 'cls')
+}
+
+generate_ranking_model <- function () {
+  print('Learning to rank')
+  y <- sample(0:4, size = metadata$kRows, replace = TRUE)
+  stopifnot(max(y) == 4, min(y) == 0)
+  kGroups <- 20
+  w <- runif(kGroups)
+  g <- rep(50, times = kGroups)
+
+  data <- xgb.DMatrix(X, label = y, group = g)
+  # setinfo(data, 'weight', w)
+  # ^^^ does not work in version <= 1.1.0; see https://github.com/dmlc/xgboost/issues/5942
+  # So call low-level function XGDMatrixSetInfo_R directly. Since this function is not an exported
+  # symbol, use the triple-colon operator.
+  .Call(xgboost:::XGDMatrixSetInfo_R, data, 'weight', as.numeric(w))
+  params <- list(objective = 'rank:ndcg', num_parallel_tree = metadata$kForests,
+                 tree_method = 'hist', max_depth = metadata$kMaxDepth)
+  booster <- xgb.train(params, data, nrounds = metadata$kRounds)
+  save_booster(booster, 'ltr')
+}
+
+dir.create(target_dir)
+
+invisible(generate_regression_model())
+invisible(generate_logistic_model())
+invisible(generate_classification_model())
+invisible(generate_ranking_model())

R-package/tests/helper_scripts/run_lint.R

@@ -0,0 +1,71 @@
+library(lintr)
+library(crayon)
+
+my_linters <- list(
+  absolute_path_linter = lintr::absolute_path_linter,
+  assignment_linter = lintr::assignment_linter,
+  closed_curly_linter = lintr::closed_curly_linter,
+  commas_linter = lintr::commas_linter,
+  equals_na = lintr::equals_na_linter,
+  infix_spaces_linter = lintr::infix_spaces_linter,
+  line_length_linter = lintr::line_length_linter,
+  no_tab_linter = lintr::no_tab_linter,
+  object_usage_linter = lintr::object_usage_linter,
+  object_length_linter = lintr::object_length_linter,
+  open_curly_linter = lintr::open_curly_linter,
+  semicolon = lintr::semicolon_terminator_linter,
+  seq = lintr::seq_linter,
+  spaces_inside_linter = lintr::spaces_inside_linter,
+  spaces_left_parentheses_linter = lintr::spaces_left_parentheses_linter,
+  trailing_blank_lines_linter = lintr::trailing_blank_lines_linter,
+  trailing_whitespace_linter = lintr::trailing_whitespace_linter,
+  true_false = lintr::T_and_F_symbol_linter,
+  unneeded_concatenation = lintr::unneeded_concatenation_linter
+)
+
+results <- lapply(
+  list.files(path = '.', pattern = '\\.[Rr]$', recursive = TRUE),
+  function (r_file) {
+    cat(sprintf("Processing %s ...\n", r_file))
+    list(r_file = r_file,
+         output = lintr::lint(filename = r_file, linters = my_linters))
+  })
+num_issue <- Reduce(sum, lapply(results, function (e) length(e$output)))
+
+lint2str <- function(lint_entry) {
+  color <- function(type) {
+    switch(type,
+      "warning" = crayon::magenta,
+      "error" = crayon::red,
+      "style" = crayon::blue,
+      crayon::bold
+    )
+  }
+
+  paste0(
+    lapply(lint_entry$output,
+      function (lint_line) {
+        paste0(
+          crayon::bold(lint_entry$r_file, ":",
+          as.character(lint_line$line_number), ":",
+          as.character(lint_line$column_number), ": ", sep = ""),
+          color(lint_line$type)(lint_line$type, ": ", sep = ""),
+          crayon::bold(lint_line$message), "\n",
+          lint_line$line, "\n",
+          lintr:::highlight_string(lint_line$message, lint_line$column_number, lint_line$ranges),
+          "\n",
+          collapse = "")
+      }),
+    collapse = "")
+}
+
+if (num_issue > 0) {
+  cat(sprintf('R linters found %d issues:\n', num_issue))
+  for (entry in results) {
+    if (length(entry$output)) {
+      cat(paste0('**** ', crayon::bold(entry$r_file), '\n'))
+      cat(paste0(lint2str(entry), collapse = ''))
+    }
+  }
+  quit(save = 'no', status = 1)  # Signal error to parent shell
+}

R-package/tests/testthat.R

@@ -1,4 +1,4 @@
 library(testthat)
 library(xgboost)
 
-test_check("xgboost")
+test_check("xgboost", reporter = ProgressReporter)

R-package/tests/testthat/test_basic.R

@@ -2,22 +2,23 @@ require(xgboost)
 
 context("basic functions")
 
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 train <- agaricus.train
 test <- agaricus.test
 set.seed(1994)
 
 # disable some tests for Win32
-windows_flag = .Platform$OS.type == "windows" &&
+windows_flag <- .Platform$OS.type == "windows" &&
                .Machine$sizeof.pointer != 8
-solaris_flag = (Sys.info()['sysname'] == "SunOS")
+solaris_flag <- (Sys.info()['sysname'] == "SunOS")
 
 test_that("train and predict binary classification", {
-  nrounds = 2
+  nrounds <- 2
   expect_output(
     bst <- xgboost(data = train$data, label = train$label, max_depth = 2,
-                  eta = 1, nthread = 2, nrounds = nrounds, objective = "binary:logistic")
+                  eta = 1, nthread = 2, nrounds = nrounds, objective = "binary:logistic",
+                  eval_metric = "error")
   , "train-error")
   expect_equal(class(bst), "xgb.Booster")
   expect_equal(bst$niter, nrounds)
@@ -30,24 +31,24 @@ test_that("train and predict binary classification", {
 
   pred1 <- predict(bst, train$data, ntreelimit = 1)
   expect_length(pred1, 6513)
-  err_pred1 <- sum((pred1 > 0.5) != train$label)/length(train$label)
+  err_pred1 <- sum((pred1 > 0.5) != train$label) / length(train$label)
   err_log <- bst$evaluation_log[1, train_error]
   expect_lt(abs(err_pred1 - err_log), 10e-6)
 })
 
 test_that("parameter validation works", {
   p <- list(foo = "bar")
-  nrounds = 1
+  nrounds <- 1
   set.seed(1994)
 
   d <- cbind(
     x1 = rnorm(10),
     x2 = rnorm(10),
     x3 = rnorm(10))
-  y <- d[,"x1"] + d[,"x2"]^2 +
-    ifelse(d[,"x3"] > .5, d[,"x3"]^2, 2^d[,"x3"]) +
+  y <- d[, "x1"] + d[, "x2"]^2 +
+    ifelse(d[, "x3"] > .5, d[, "x3"]^2, 2^d[, "x3"]) +
     rnorm(10)
-  dtrain <- xgb.DMatrix(data=d, info = list(label=y))
+  dtrain <- xgb.DMatrix(data = d, info = list(label = y))
 
   correct <- function() {
     params <- list(max_depth = 2, booster = "dart",
@@ -65,20 +66,20 @@ test_that("parameter validation works", {
       xgb.train(params = params, data = dtrain, nrounds = nrounds))
     print(output)
   }
-  expect_output(incorrect(), "bar, foo")
+  expect_output(incorrect(), '\\\\"bar\\\\", \\\\"foo\\\\"')
 })
 
 
 test_that("dart prediction works", {
-  nrounds = 32
+  nrounds <- 32
   set.seed(1994)
 
   d <- cbind(
     x1 = rnorm(100),
     x2 = rnorm(100),
     x3 = rnorm(100))
-  y <- d[,"x1"] + d[,"x2"]^2 +
-    ifelse(d[,"x3"] > .5, d[,"x3"]^2, 2^d[,"x3"]) +
+  y <- d[, "x1"] + d[, "x2"]^2 +
+    ifelse(d[, "x3"] > .5, d[, "x3"]^2, 2^d[, "x3"]) +
     rnorm(100)
 
   set.seed(1994)
@@ -87,23 +88,23 @@ test_that("dart prediction works", {
                                 eta = 1, nthread = 2, nrounds = nrounds, objective = "reg:squarederror")
   pred_by_xgboost_0 <- predict(booster_by_xgboost, newdata = d, ntreelimit = 0)
   pred_by_xgboost_1 <- predict(booster_by_xgboost, newdata = d, ntreelimit = nrounds)
-  expect_true(all(matrix(pred_by_xgboost_0, byrow=TRUE) == matrix(pred_by_xgboost_1, byrow=TRUE)))
+  expect_true(all(matrix(pred_by_xgboost_0, byrow = TRUE) == matrix(pred_by_xgboost_1, byrow = TRUE)))
 
   pred_by_xgboost_2 <- predict(booster_by_xgboost, newdata = d, training = TRUE)
-  expect_false(all(matrix(pred_by_xgboost_0, byrow=TRUE) == matrix(pred_by_xgboost_2, byrow=TRUE)))
+  expect_false(all(matrix(pred_by_xgboost_0, byrow = TRUE) == matrix(pred_by_xgboost_2, byrow = TRUE)))
 
   set.seed(1994)
-  dtrain <- xgb.DMatrix(data=d, info = list(label=y))
-  booster_by_train <- xgb.train( params = list(
-                                   booster = "dart",
-                                   max_depth = 2,
-                                   eta = 1,
-                                   rate_drop = 0.5,
-                                   one_drop = TRUE,
-                                   nthread = 1,
-                                   tree_method= "exact",
-                                   objective = "reg:squarederror"
-                                 ),
+  dtrain <- xgb.DMatrix(data = d, info = list(label = y))
+  booster_by_train <- xgb.train(params = list(
+                                    booster = "dart",
+                                    max_depth = 2,
+                                    eta = 1,
+                                    rate_drop = 0.5,
+                                    one_drop = TRUE,
+                                    nthread = 1,
+                                    tree_method = "exact",
+                                    objective = "reg:squarederror"
+                                ),
                                 data = dtrain,
                                 nrounds = nrounds
                                 )
@@ -111,9 +112,9 @@ test_that("dart prediction works", {
   pred_by_train_1 <- predict(booster_by_train, newdata = dtrain, ntreelimit = nrounds)
   pred_by_train_2 <- predict(booster_by_train, newdata = dtrain, training = TRUE)
 
-  expect_true(all(matrix(pred_by_train_0, byrow=TRUE) == matrix(pred_by_xgboost_0, byrow=TRUE)))
-  expect_true(all(matrix(pred_by_train_1, byrow=TRUE) == matrix(pred_by_xgboost_1, byrow=TRUE)))
-  expect_true(all(matrix(pred_by_train_2, byrow=TRUE) == matrix(pred_by_xgboost_2, byrow=TRUE)))
+  expect_true(all(matrix(pred_by_train_0, byrow = TRUE) == matrix(pred_by_xgboost_0, byrow = TRUE)))
+  expect_true(all(matrix(pred_by_train_1, byrow = TRUE) == matrix(pred_by_xgboost_1, byrow = TRUE)))
+  expect_true(all(matrix(pred_by_train_2, byrow = TRUE) == matrix(pred_by_xgboost_2, byrow = TRUE)))
 })
 
 test_that("train and predict softprob", {
@@ -122,7 +123,7 @@ test_that("train and predict softprob", {
   expect_output(
     bst <- xgboost(data = as.matrix(iris[, -5]), label = lb,
                    max_depth = 3, eta = 0.5, nthread = 2, nrounds = 5,
-                   objective = "multi:softprob", num_class=3)
+                   objective = "multi:softprob", num_class = 3, eval_metric = "merror")
   , "train-merror")
   expect_false(is.null(bst$evaluation_log))
   expect_lt(bst$evaluation_log[, min(train_merror)], 0.025)
@@ -130,17 +131,17 @@ test_that("train and predict softprob", {
   pred <- predict(bst, as.matrix(iris[, -5]))
   expect_length(pred, nrow(iris) * 3)
   # row sums add up to total probability of 1:
-  expect_equal(rowSums(matrix(pred, ncol=3, byrow=TRUE)), rep(1, nrow(iris)), tolerance = 1e-7)
+  expect_equal(rowSums(matrix(pred, ncol = 3, byrow = TRUE)), rep(1, nrow(iris)), tolerance = 1e-7)
   # manually calculate error at the last iteration:
   mpred <- predict(bst, as.matrix(iris[, -5]), reshape = TRUE)
   expect_equal(as.numeric(t(mpred)), pred)
   pred_labels <- max.col(mpred) - 1
-  err <- sum(pred_labels != lb)/length(lb)
+  err <- sum(pred_labels != lb) / length(lb)
   expect_equal(bst$evaluation_log[5, train_merror], err, tolerance = 5e-6)
   # manually calculate error at the 1st iteration:
   mpred <- predict(bst, as.matrix(iris[, -5]), reshape = TRUE, ntreelimit = 1)
   pred_labels <- max.col(mpred) - 1
-  err <- sum(pred_labels != lb)/length(lb)
+  err <- sum(pred_labels != lb) / length(lb)
   expect_equal(bst$evaluation_log[1, train_merror], err, tolerance = 5e-6)
 })
 
@@ -150,7 +151,7 @@ test_that("train and predict softmax", {
   expect_output(
     bst <- xgboost(data = as.matrix(iris[, -5]), label = lb,
                    max_depth = 3, eta = 0.5, nthread = 2, nrounds = 5,
-                   objective = "multi:softmax", num_class=3)
+                   objective = "multi:softmax", num_class = 3, eval_metric = "merror")
   , "train-merror")
   expect_false(is.null(bst$evaluation_log))
   expect_lt(bst$evaluation_log[, min(train_merror)], 0.025)
@@ -158,7 +159,7 @@ test_that("train and predict softmax", {
 
   pred <- predict(bst, as.matrix(iris[, -5]))
   expect_length(pred, nrow(iris))
-  err <- sum(pred != lb)/length(lb)
+  err <- sum(pred != lb) / length(lb)
   expect_equal(bst$evaluation_log[5, train_merror], err, tolerance = 5e-6)
 })
 
@@ -167,18 +168,18 @@ test_that("train and predict RF", {
   lb <- train$label
   # single iteration
   bst <- xgboost(data = train$data, label = lb, max_depth = 5,
-                 nthread = 2, nrounds = 1, objective = "binary:logistic",
+                 nthread = 2, nrounds = 1, objective = "binary:logistic", eval_metric = "error",
                  num_parallel_tree = 20, subsample = 0.6, colsample_bytree = 0.1)
   expect_equal(bst$niter, 1)
   expect_equal(xgb.ntree(bst), 20)
 
   pred <- predict(bst, train$data)
-  pred_err <- sum((pred > 0.5) != lb)/length(lb)
+  pred_err <- sum((pred > 0.5) != lb) / length(lb)
   expect_lt(abs(bst$evaluation_log[1, train_error] - pred_err), 10e-6)
   #expect_lt(pred_err, 0.03)
 
   pred <- predict(bst, train$data, ntreelimit = 20)
-  pred_err_20 <- sum((pred > 0.5) != lb)/length(lb)
+  pred_err_20 <- sum((pred > 0.5) != lb) / length(lb)
   expect_equal(pred_err_20, pred_err)
 
   #pred <- predict(bst, train$data, ntreelimit = 1)
@@ -193,19 +194,20 @@ test_that("train and predict RF with softprob", {
   set.seed(11)
   bst <- xgboost(data = as.matrix(iris[, -5]), label = lb,
                  max_depth = 3, eta = 0.9, nthread = 2, nrounds = nrounds,
-                 objective = "multi:softprob", num_class=3, verbose = 0,
+                 objective = "multi:softprob", eval_metric = "merror",
+                 num_class = 3, verbose = 0,
                  num_parallel_tree = 4, subsample = 0.5, colsample_bytree = 0.5)
   expect_equal(bst$niter, 15)
-  expect_equal(xgb.ntree(bst), 15*3*4)
+  expect_equal(xgb.ntree(bst), 15 * 3 * 4)
   # predict for all iterations:
-  pred <- predict(bst, as.matrix(iris[, -5]), reshape=TRUE)
+  pred <- predict(bst, as.matrix(iris[, -5]), reshape = TRUE)
   expect_equal(dim(pred), c(nrow(iris), 3))
   pred_labels <- max.col(pred) - 1
-  err <- sum(pred_labels != lb)/length(lb)
+  err <- sum(pred_labels != lb) / length(lb)
   expect_equal(bst$evaluation_log[nrounds, train_merror], err, tolerance = 5e-6)
   # predict for 7 iterations and adjust for 4 parallel trees per iteration
-  pred <- predict(bst, as.matrix(iris[, -5]), reshape=TRUE, ntreelimit = 7 * 4)
-  err <- sum((max.col(pred) - 1) != lb)/length(lb)
+  pred <- predict(bst, as.matrix(iris[, -5]), reshape = TRUE, ntreelimit = 7 * 4)
+  err <- sum((max.col(pred) - 1) != lb) / length(lb)
   expect_equal(bst$evaluation_log[7, train_merror], err, tolerance = 5e-6)
 })
 
@@ -223,7 +225,7 @@ test_that("use of multiple eval metrics works", {
 
 test_that("training continuation works", {
   dtrain <- xgb.DMatrix(train$data, label = train$label)
-  watchlist = list(train=dtrain)
+  watchlist <- list(train = dtrain)
   param <- list(objective = "binary:logistic", max_depth = 2, eta = 1, nthread = 2)
 
   # for the reference, use 4 iterations at once:
@@ -245,17 +247,18 @@ test_that("training continuation works", {
     expect_equal(bst$raw, bst2$raw)
   expect_equal(dim(bst2$evaluation_log), c(2, 2))
   # test continuing from a model in file
-  xgb.save(bst1, "xgboost.model")
-  bst2 <- xgb.train(param, dtrain, nrounds = 2, watchlist, verbose = 0, xgb_model = "xgboost.model")
+  xgb.save(bst1, "xgboost.json")
+  bst2 <- xgb.train(param, dtrain, nrounds = 2, watchlist, verbose = 0, xgb_model = "xgboost.json")
   if (!windows_flag && !solaris_flag)
     expect_equal(bst$raw, bst2$raw)
   expect_equal(dim(bst2$evaluation_log), c(2, 2))
+  file.remove("xgboost.json")
 })
 
 test_that("model serialization works", {
   out_path <- "model_serialization"
   dtrain <- xgb.DMatrix(train$data, label = train$label)
-  watchlist = list(train=dtrain)
+  watchlist <- list(train = dtrain)
   param <- list(objective = "binary:logistic")
   booster <- xgb.train(param, dtrain, nrounds = 4, watchlist)
   raw <- xgb.serialize(booster)
@@ -273,7 +276,7 @@ test_that("xgb.cv works", {
   expect_output(
     cv <- xgb.cv(data = train$data, label = train$label, max_depth = 2, nfold = 5,
                  eta = 1., nthread = 2, nrounds = 2, objective = "binary:logistic",
-                 verbose=TRUE)
+                 eval_metric = "error", verbose = TRUE)
   , "train-error:")
   expect_is(cv, 'xgb.cv.synchronous')
   expect_false(is.null(cv$evaluation_log))
@@ -292,13 +295,13 @@ test_that("xgb.cv works with stratified folds", {
   set.seed(314159)
   cv <- xgb.cv(data = dtrain, max_depth = 2, nfold = 5,
                eta = 1., nthread = 2, nrounds = 2, objective = "binary:logistic",
-               verbose=TRUE, stratified = FALSE)
+               verbose = TRUE, stratified = FALSE)
   set.seed(314159)
   cv2 <- xgb.cv(data = dtrain, max_depth = 2, nfold = 5,
                 eta = 1., nthread = 2, nrounds = 2, objective = "binary:logistic",
-                verbose=TRUE, stratified = TRUE)
+                verbose = TRUE, stratified = TRUE)
   # Stratified folds should result in a different evaluation logs
-  expect_true(all(cv$evaluation_log[, test_error_mean] != cv2$evaluation_log[, test_error_mean]))
+  expect_true(all(cv$evaluation_log[, test_logloss_mean] != cv2$evaluation_log[, test_logloss_mean]))
 })
 
 test_that("train and predict with non-strict classes", {
@@ -319,7 +322,7 @@ test_that("train and predict with non-strict classes", {
   expect_equal(pr0, pr)
 
   # dense matrix-like input of non-matrix class with some inheritance
-  class(train_dense) <- c('pphmatrix','shmatrix')
+  class(train_dense) <- c('pphmatrix', 'shmatrix')
   expect_true(is.matrix(train_dense))
   expect_error(
     bst <- xgboost(data = train_dense, label = train$label, max_depth = 2,
@@ -337,15 +340,15 @@ test_that("train and predict with non-strict classes", {
 test_that("max_delta_step works", {
   dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
   watchlist <- list(train = dtrain)
-  param <- list(objective = "binary:logistic", eval_metric="logloss", max_depth = 2, nthread = 2, eta = 0.5)
-  nrounds = 5
+  param <- list(objective = "binary:logistic", eval_metric = "logloss", max_depth = 2, nthread = 2, eta = 0.5)
+  nrounds <- 5
   # model with no restriction on max_delta_step
   bst1 <- xgb.train(param, dtrain, nrounds, watchlist, verbose = 1)
   # model with restricted max_delta_step
   bst2 <- xgb.train(param, dtrain, nrounds, watchlist, verbose = 1, max_delta_step = 1)
   # the no-restriction model is expected to have consistently lower loss during the initial interations
   expect_true(all(bst1$evaluation_log$train_logloss < bst2$evaluation_log$train_logloss))
-  expect_lt(mean(bst1$evaluation_log$train_logloss)/mean(bst2$evaluation_log$train_logloss), 0.8)
+  expect_lt(mean(bst1$evaluation_log$train_logloss) / mean(bst2$evaluation_log$train_logloss), 0.8)
 })
 
 test_that("colsample_bytree works", {

R-package/tests/testthat/test_callbacks.R

@@ -2,11 +2,12 @@
 
 require(xgboost)
 require(data.table)
+require(titanic)
 
 context("callbacks")
 
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 train <- agaricus.train
 test <- agaricus.test
 
@@ -21,24 +22,25 @@ ltrain <- add.noise(train$label, 0.2)
 ltest <- add.noise(test$label, 0.2)
 dtrain <- xgb.DMatrix(train$data, label = ltrain)
 dtest <- xgb.DMatrix(test$data, label = ltest)
-watchlist = list(train=dtrain, test=dtest)
+watchlist <- list(train = dtrain, test = dtest)
 
 
-err <- function(label, pr) sum((pr > 0.5) != label)/length(label)
+err <- function(label, pr) sum((pr > 0.5) != label) / length(label)
 
-param <- list(objective = "binary:logistic", max_depth = 2, nthread = 2)
+param <- list(objective = "binary:logistic", eval_metric = "error",
+              max_depth = 2, nthread = 2)
 
 
 test_that("cb.print.evaluation works as expected", {
 
-  bst_evaluation <- c('train-auc'=0.9, 'test-auc'=0.8)
+  bst_evaluation <- c('train-auc' = 0.9, 'test-auc' = 0.8)
   bst_evaluation_err <- NULL
   begin_iteration <- 1
   end_iteration <- 7
 
-  f0 <- cb.print.evaluation(period=0)
-  f1 <- cb.print.evaluation(period=1)
-  f5 <- cb.print.evaluation(period=5)
+  f0 <- cb.print.evaluation(period = 0)
+  f1 <- cb.print.evaluation(period = 1)
+  f5 <- cb.print.evaluation(period = 5)
 
   expect_false(is.null(attr(f1, 'call')))
   expect_equal(attr(f1, 'name'), 'cb.print.evaluation')
@@ -57,13 +59,13 @@ test_that("cb.print.evaluation works as expected", {
   expect_output(f1(), "\\[7\\]\ttrain-auc:0.900000\ttest-auc:0.800000")
   expect_output(f5(), "\\[7\\]\ttrain-auc:0.900000\ttest-auc:0.800000")
 
-  bst_evaluation_err  <- c('train-auc'=0.1, 'test-auc'=0.2)
+  bst_evaluation_err  <- c('train-auc' = 0.1, 'test-auc' = 0.2)
   expect_output(f1(), "\\[7\\]\ttrain-auc:0.900000\\+0.100000\ttest-auc:0.800000\\+0.200000")
 })
 
 test_that("cb.evaluation.log works as expected", {
 
-  bst_evaluation <- c('train-auc'=0.9, 'test-auc'=0.8)
+  bst_evaluation <- c('train-auc' = 0.9, 'test-auc' = 0.8)
   bst_evaluation_err <- NULL
 
   evaluation_log <- list()
@@ -75,37 +77,38 @@ test_that("cb.evaluation.log works as expected", {
   iteration <- 1
   expect_silent(f())
   expect_equal(evaluation_log,
-               list(c(iter=1, bst_evaluation)))
+               list(c(iter = 1, bst_evaluation)))
   iteration <- 2
   expect_silent(f())
   expect_equal(evaluation_log,
-               list(c(iter=1, bst_evaluation), c(iter=2, bst_evaluation)))
+               list(c(iter = 1, bst_evaluation), c(iter = 2, bst_evaluation)))
   expect_silent(f(finalize = TRUE))
   expect_equal(evaluation_log,
-               data.table(iter=1:2, train_auc=c(0.9,0.9), test_auc=c(0.8,0.8)))
+               data.table(iter = 1:2, train_auc = c(0.9, 0.9), test_auc = c(0.8, 0.8)))
 
-  bst_evaluation_err  <- c('train-auc'=0.1, 'test-auc'=0.2)
+  bst_evaluation_err  <- c('train-auc' = 0.1, 'test-auc' = 0.2)
   evaluation_log <- list()
   f <- cb.evaluation.log()
 
   iteration <- 1
   expect_silent(f())
   expect_equal(evaluation_log,
-               list(c(iter=1, c(bst_evaluation, bst_evaluation_err))))
+               list(c(iter = 1, c(bst_evaluation, bst_evaluation_err))))
   iteration <- 2
   expect_silent(f())
   expect_equal(evaluation_log,
-               list(c(iter=1, c(bst_evaluation, bst_evaluation_err)),
-                    c(iter=2, c(bst_evaluation, bst_evaluation_err))))
+               list(c(iter = 1, c(bst_evaluation, bst_evaluation_err)),
+                    c(iter = 2, c(bst_evaluation, bst_evaluation_err))))
   expect_silent(f(finalize = TRUE))
   expect_equal(evaluation_log,
-               data.table(iter=1:2,
-                          train_auc_mean=c(0.9,0.9), train_auc_std=c(0.1,0.1),
-                          test_auc_mean=c(0.8,0.8), test_auc_std=c(0.2,0.2)))
+               data.table(iter = 1:2,
+                          train_auc_mean = c(0.9, 0.9), train_auc_std = c(0.1, 0.1),
+                          test_auc_mean = c(0.8, 0.8), test_auc_std = c(0.2, 0.2)))
 })
 
 
-param <- list(objective = "binary:logistic", max_depth = 4, nthread = 2)
+param <- list(objective = "binary:logistic", eval_metric = "error",
+              max_depth = 4, nthread = 2)
 
 test_that("can store evaluation_log without printing", {
   expect_silent(
@@ -173,16 +176,16 @@ test_that("cb.reset.parameters works as expected", {
 })
 
 test_that("cb.save.model works as expected", {
-  files <- c('xgboost_01.model', 'xgboost_02.model', 'xgboost.model')
+  files <- c('xgboost_01.json', 'xgboost_02.json', 'xgboost.json')
   for (f in files) if (file.exists(f)) file.remove(f)
 
   bst <- xgb.train(param, dtrain, nrounds = 2, watchlist, eta = 1, verbose = 0,
-                   save_period = 1, save_name = "xgboost_%02d.model")
-  expect_true(file.exists('xgboost_01.model'))
-  expect_true(file.exists('xgboost_02.model'))
-  b1 <- xgb.load('xgboost_01.model')
+                   save_period = 1, save_name = "xgboost_%02d.json")
+  expect_true(file.exists('xgboost_01.json'))
+  expect_true(file.exists('xgboost_02.json'))
+  b1 <- xgb.load('xgboost_01.json')
   expect_equal(xgb.ntree(b1), 1)
-  b2 <- xgb.load('xgboost_02.model')
+  b2 <- xgb.load('xgboost_02.json')
   expect_equal(xgb.ntree(b2), 2)
 
   xgb.config(b2) <- xgb.config(bst)
@@ -191,9 +194,9 @@ test_that("cb.save.model works as expected", {
 
   # save_period = 0 saves the last iteration's model
   bst <- xgb.train(param, dtrain, nrounds = 2, watchlist, eta = 1, verbose = 0,
-                   save_period = 0)
-  expect_true(file.exists('xgboost.model'))
-  b2 <- xgb.load('xgboost.model')
+                   save_period = 0, save_name = 'xgboost.json')
+  expect_true(file.exists('xgboost.json'))
+  b2 <- xgb.load('xgboost.json')
   xgb.config(b2) <- xgb.config(bst)
   expect_equal(bst$raw, b2$raw)
 
@@ -236,8 +239,8 @@ test_that("early stopping xgb.train works", {
 test_that("early stopping using a specific metric works", {
   set.seed(11)
   expect_output(
-    bst <- xgb.train(param, dtrain, nrounds = 20, watchlist, eta = 0.6,
-                     eval_metric="logloss", eval_metric="auc",
+    bst <- xgb.train(param[-2], dtrain, nrounds = 20, watchlist, eta = 0.6,
+                     eval_metric = "logloss", eval_metric = "auc",
                      callbacks = list(cb.early.stop(stopping_rounds = 3, maximize = FALSE,
                                                     metric_name = 'test_logloss')))
   , "Stopping. Best iteration")
@@ -252,6 +255,26 @@ test_that("early stopping using a specific metric works", {
   expect_equal(logloss_log, logloss_pred, tolerance = 1e-5)
 })
 
+test_that("early stopping works with titanic", {
+  # This test was inspired by https://github.com/dmlc/xgboost/issues/5935
+  # It catches possible issues on noLD R
+  titanic <- titanic::titanic_train
+  titanic$Pclass <-  as.factor(titanic$Pclass)
+  dtx <- model.matrix(~ 0 + ., data = titanic[, c("Pclass", "Sex")])
+  dty <- titanic$Survived
+
+  xgboost::xgboost(
+    data = dtx,
+    label = dty,
+    objective = "binary:logistic",
+    eval_metric = "auc",
+    nrounds = 100,
+    early_stopping_rounds = 3
+  )
+
+  expect_true(TRUE)  # should not crash
+})
+
 test_that("early stopping xgb.cv works", {
   set.seed(11)
   expect_output(
@@ -267,12 +290,12 @@ test_that("early stopping xgb.cv works", {
 
 test_that("prediction in xgb.cv works", {
   set.seed(11)
-  nrounds = 4
+  nrounds <- 4
   cv <- xgb.cv(param, dtrain, nfold = 5, eta = 0.5, nrounds = nrounds, prediction = TRUE, verbose = 0)
   expect_false(is.null(cv$evaluation_log))
   expect_false(is.null(cv$pred))
   expect_length(cv$pred, nrow(train$data))
-  err_pred <- mean( sapply(cv$folds, function(f) mean(err(ltrain[f], cv$pred[f]))) )
+  err_pred <- mean(sapply(cv$folds, function(f) mean(err(ltrain[f], cv$pred[f]))))
   err_log <- cv$evaluation_log[nrounds, test_error_mean]
   expect_equal(err_pred, err_log, tolerance = 1e-6)
 
@@ -308,7 +331,7 @@ test_that("prediction in early-stopping xgb.cv works", {
   expect_false(is.null(cv$pred))
   expect_length(cv$pred, nrow(train$data))
 
-  err_pred <- mean( sapply(cv$folds, function(f) mean(err(ltrain[f], cv$pred[f]))) )
+  err_pred <- mean(sapply(cv$folds, function(f) mean(err(ltrain[f], cv$pred[f]))))
   err_log <- cv$evaluation_log[cv$best_iteration, test_error_mean]
   expect_equal(err_pred, err_log, tolerance = 1e-6)
   err_log_last <- cv$evaluation_log[cv$niter, test_error_mean]

R-package/tests/testthat/test_config.R

@@ -0,0 +1,21 @@
+context('Test global configuration')
+
+test_that('Global configuration works with verbosity', {
+  old_verbosity <- xgb.get.config()$verbosity
+  for (v in c(0, 1, 2, 3)) {
+    xgb.set.config(verbosity = v)
+    expect_equal(xgb.get.config()$verbosity, v)
+  }
+  xgb.set.config(verbosity = old_verbosity)
+  expect_equal(xgb.get.config()$verbosity, old_verbosity)
+})
+
+test_that('Global configuration works with use_rmm flag', {
+  old_use_rmm_flag <- xgb.get.config()$use_rmm
+  for (v in c(TRUE, FALSE)) {
+    xgb.set.config(use_rmm = v)
+    expect_equal(xgb.get.config()$use_rmm, v)
+  }
+  xgb.set.config(use_rmm = old_use_rmm_flag)
+  expect_equal(xgb.get.config()$use_rmm, old_use_rmm_flag)
+})

R-package/tests/testthat/test_custom_objective.R

@@ -4,8 +4,8 @@ require(xgboost)
 
 set.seed(1994)
 
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 watchlist <- list(eval = dtest, train = dtrain)
@@ -20,12 +20,12 @@ logregobj <- function(preds, dtrain) {
 
 evalerror <- function(preds, dtrain) {
   labels <- getinfo(dtrain, "label")
-  err <- as.numeric(sum(labels != (preds > 0))) / length(labels)
+  err <- as.numeric(sum(labels != (preds > 0.5))) / length(labels)
   return(list(metric = "error", value = err))
 }
 
-param <- list(max_depth=2, eta=1, nthread = 2,
-              objective=logregobj, eval_metric=evalerror)
+param <- list(max_depth = 2, eta = 1, nthread = 2,
+              objective = logregobj, eval_metric = evalerror)
 num_round <- 2
 
 test_that("custom objective works", {
@@ -37,12 +37,19 @@ test_that("custom objective works", {
 })
 
 test_that("custom objective in CV works", {
-  cv <- xgb.cv(param, dtrain, num_round, nfold=10, verbose=FALSE)
+  cv <- xgb.cv(param, dtrain, num_round, nfold = 10, verbose = FALSE)
   expect_false(is.null(cv$evaluation_log))
   expect_equal(dim(cv$evaluation_log), c(2, 5))
   expect_lt(cv$evaluation_log[num_round, test_error_mean], 0.03)
 })
 
+test_that("custom objective with early stop works", {
+  bst <- xgb.train(param, dtrain, 10, watchlist)
+  expect_equal(class(bst), "xgb.Booster")
+  train_log <- bst$evaluation_log$train_error
+  expect_true(all(diff(train_log) <= 0))
+})
+
 test_that("custom objective using DMatrix attr works", {
 
   attr(dtrain, 'label') <- getinfo(dtrain, 'label')
@@ -54,14 +61,14 @@ test_that("custom objective using DMatrix attr works", {
     hess <- preds * (1 - preds)
     return(list(grad = grad, hess = hess))
   }
-  param$objective = logregobjattr
+  param$objective <- logregobjattr
   bst <- xgb.train(param, dtrain, num_round, watchlist)
   expect_equal(class(bst), "xgb.Booster")
 })
 
 test_that("custom objective with multi-class works", {
-  data = as.matrix(iris[, -5])
-  label =  as.numeric(iris$Species) - 1
+  data <- as.matrix(iris[, -5])
+  label <-  as.numeric(iris$Species) - 1
   dtrain <- xgb.DMatrix(data = data, label = label)
   nclasses <- 3
 
@@ -72,6 +79,10 @@ test_that("custom objective with multi-class works", {
     hess <- rnorm(dim(as.matrix(preds))[1])
     return (list(grad = grad, hess = hess))
   }
-  param$objective = fake_softprob
-  bst <- xgb.train(param, dtrain, 1, num_class=nclasses)
+  fake_merror <- function(preds, dtrain) {
+    expect_equal(dim(data)[1] * nclasses, dim(as.matrix(preds))[1])
+  }
+  param$objective <- fake_softprob
+  param$eval_metric <- fake_merror
+  bst <- xgb.train(param, dtrain, 1, num_class = nclasses)
 })

R-package/tests/testthat/test_dmatrix.R

@@ -3,29 +3,29 @@ require(Matrix)
 
 context("testing xgb.DMatrix functionality")
 
-data(agaricus.test, package='xgboost')
-test_data <- agaricus.test$data[1:100,]
+data(agaricus.test, package = 'xgboost')
+test_data <- agaricus.test$data[1:100, ]
 test_label <- agaricus.test$label[1:100]
 
 test_that("xgb.DMatrix: basic construction", {
   # from sparse matrix
-  dtest1 <- xgb.DMatrix(test_data, label=test_label)
+  dtest1 <- xgb.DMatrix(test_data, label = test_label)
 
   # from dense matrix
-  dtest2 <- xgb.DMatrix(as.matrix(test_data), label=test_label)
+  dtest2 <- xgb.DMatrix(as.matrix(test_data), label = test_label)
   expect_equal(getinfo(dtest1, 'label'), getinfo(dtest2, 'label'))
   expect_equal(dim(dtest1), dim(dtest2))
 
   #from dense integer matrix
   int_data <- as.matrix(test_data)
   storage.mode(int_data) <- "integer"
-  dtest3 <- xgb.DMatrix(int_data, label=test_label)
+  dtest3 <- xgb.DMatrix(int_data, label = test_label)
   expect_equal(dim(dtest1), dim(dtest3))
 })
 
 test_that("xgb.DMatrix: saving, loading", {
   # save to a local file
-  dtest1 <- xgb.DMatrix(test_data, label=test_label)
+  dtest1 <- xgb.DMatrix(test_data, label = test_label)
   tmp_file <- tempfile('xgb.DMatrix_')
   expect_true(xgb.DMatrix.save(dtest1, tmp_file))
   # read from a local file
@@ -35,12 +35,12 @@ test_that("xgb.DMatrix: saving, loading", {
   expect_equal(getinfo(dtest1, 'label'), getinfo(dtest3, 'label'))
 
   # from a libsvm text file
-  tmp <- c("0 1:1 2:1","1 3:1","0 1:1")
+  tmp <- c("0 1:1 2:1", "1 3:1", "0 1:1")
   tmp_file <- 'tmp.libsvm'
   writeLines(tmp, tmp_file)
   dtest4 <- xgb.DMatrix(tmp_file, silent = TRUE)
   expect_equal(dim(dtest4), c(3, 4))
-  expect_equal(getinfo(dtest4, 'label'), c(0,1,0))
+  expect_equal(getinfo(dtest4, 'label'), c(0, 1, 0))
   unlink(tmp_file)
 })
 
@@ -61,46 +61,46 @@ test_that("xgb.DMatrix: getinfo & setinfo", {
 
   expect_true(setinfo(dtest, 'weight', test_label))
   expect_true(setinfo(dtest, 'base_margin', test_label))
-  expect_true(setinfo(dtest, 'group', c(50,50)))
+  expect_true(setinfo(dtest, 'group', c(50, 50)))
   expect_error(setinfo(dtest, 'group', test_label))
 
-  # providing character values will give a warning
-  expect_warning(setinfo(dtest, 'weight', rep('a', nrow(test_data))))
+  # providing character values will give an error
+  expect_error(setinfo(dtest, 'weight', rep('a', nrow(test_data))))
 
   # any other label should error
   expect_error(setinfo(dtest, 'asdf', test_label))
 })
 
 test_that("xgb.DMatrix: slice, dim", {
-  dtest <- xgb.DMatrix(test_data, label=test_label)
+  dtest <- xgb.DMatrix(test_data, label = test_label)
   expect_equal(dim(dtest), dim(test_data))
   dsub1 <- slice(dtest, 1:42)
   expect_equal(nrow(dsub1), 42)
   expect_equal(ncol(dsub1), ncol(test_data))
 
-  dsub2 <- dtest[1:42,]
+  dsub2 <- dtest[1:42, ]
   expect_equal(dim(dtest), dim(test_data))
   expect_equal(getinfo(dsub1, 'label'), getinfo(dsub2, 'label'))
 })
 
 test_that("xgb.DMatrix: slice, trailing empty rows", {
-  data(agaricus.train, package='xgboost')
+  data(agaricus.train, package = 'xgboost')
   train_data <- agaricus.train$data
   train_label <- agaricus.train$label
-  dtrain <- xgb.DMatrix(data=train_data, label=train_label)
+  dtrain <- xgb.DMatrix(data = train_data, label = train_label)
   slice(dtrain, 6513L)
   train_data[6513, ] <- 0
-  dtrain <- xgb.DMatrix(data=train_data, label=train_label)
+  dtrain <- xgb.DMatrix(data = train_data, label = train_label)
   slice(dtrain, 6513L)
   expect_equal(nrow(dtrain), 6513)
 })
 
 test_that("xgb.DMatrix: colnames", {
-  dtest <- xgb.DMatrix(test_data, label=test_label)
+  dtest <- xgb.DMatrix(test_data, label = test_label)
   expect_equal(colnames(dtest), colnames(test_data))
-  expect_error( colnames(dtest) <- 'asdf')
-  new_names <- make.names(1:ncol(test_data))
-  expect_silent( colnames(dtest) <- new_names)
+  expect_error(colnames(dtest) <- 'asdf')
+  new_names <- make.names(seq_len(ncol(test_data)))
+  expect_silent(colnames(dtest) <- new_names)
   expect_equal(colnames(dtest), new_names)
   expect_silent(colnames(dtest) <- NULL)
   expect_null(colnames(dtest))
@@ -109,7 +109,7 @@ test_that("xgb.DMatrix: colnames", {
 test_that("xgb.DMatrix: nrow is correct for a very sparse matrix", {
   set.seed(123)
   nr <- 1000
-  x <- rsparsematrix(nr, 100, density=0.0005)
+  x <- rsparsematrix(nr, 100, density = 0.0005)
   # we want it very sparse, so that last rows are empty
   expect_lt(max(x@i), nr)
   dtest <- xgb.DMatrix(x)

R-package/tests/testthat/test_gc_safety.R

@@ -3,13 +3,14 @@ require(xgboost)
 context("Garbage Collection Safety Check")
 
 test_that("train and prediction when gctorture is on", {
-  data(agaricus.train, package='xgboost')
-  data(agaricus.test, package='xgboost')
+  data(agaricus.train, package = 'xgboost')
+  data(agaricus.test, package = 'xgboost')
   train <- agaricus.train
   test <- agaricus.test
   gctorture(TRUE)
   bst <- xgboost(data = train$data, label = train$label, max.depth = 2,
-  eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
+                 eta = 1, nthread = 2, nrounds = 2, objective = "binary:logistic")
   pred <- predict(bst, test$data)
   gctorture(FALSE)
+  expect_length(pred, length(test$label))
 })

R-package/tests/testthat/test_glm.R

@@ -3,12 +3,12 @@ context('Test generalized linear models')
 require(xgboost)
 
 test_that("gblinear works", {
-  data(agaricus.train, package='xgboost')
-  data(agaricus.test, package='xgboost')
+  data(agaricus.train, package = 'xgboost')
+  data(agaricus.test, package = 'xgboost')
   dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
   dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 
-  param <- list(objective = "binary:logistic", booster = "gblinear",
+  param <- list(objective = "binary:logistic", eval_metric = "error", booster = "gblinear",
                 nthread = 2, eta = 0.8, alpha = 0.0001, lambda = 0.0001)
   watchlist <- list(eval = dtest, train = dtrain)
 
@@ -16,7 +16,7 @@ test_that("gblinear works", {
   ERR_UL <- 0.005 # upper limit for the test set error
   VERB <- 0      # chatterbox switch
 
-  param$updater = 'shotgun'
+  param$updater <- 'shotgun'
   bst <- xgb.train(param, dtrain, n, watchlist, verbose = VERB, feature_selector = 'shuffle')
   ypred <- predict(bst, dtest)
   expect_equal(length(getinfo(dtest, 'label')), 1611)
@@ -29,7 +29,7 @@ test_that("gblinear works", {
   expect_equal(dim(h), c(n, ncol(dtrain) + 1))
   expect_is(h, "matrix")
 
-  param$updater = 'coord_descent'
+  param$updater <- 'coord_descent'
   bst <- xgb.train(param, dtrain, n, watchlist, verbose = VERB, feature_selector = 'cyclic')
   expect_lt(bst$evaluation_log$eval_error[n], ERR_UL)
 

R-package/tests/testthat/test_helpers.R

@@ -5,18 +5,18 @@ require(data.table)
 require(Matrix)
 require(vcd, quietly = TRUE)
 
-float_tolerance = 5e-6
+float_tolerance <- 5e-6
 
 # disable some tests for 32-bit environment
-flag_32bit = .Machine$sizeof.pointer != 8
+flag_32bit <- .Machine$sizeof.pointer != 8
 
 set.seed(1982)
 data(Arthritis)
-df <- data.table(Arthritis, keep.rownames = F)
-df[,AgeDiscret := as.factor(round(Age / 10,0))]
-df[,AgeCat := as.factor(ifelse(Age > 30, "Old", "Young"))]
-df[,ID := NULL]
-sparse_matrix <- sparse.model.matrix(Improved~.-1, data = df)
+df <- data.table(Arthritis, keep.rownames = FALSE)
+df[, AgeDiscret := as.factor(round(Age / 10, 0))]
+df[, AgeCat := as.factor(ifelse(Age > 30, "Old", "Young"))]
+df[, ID := NULL]
+sparse_matrix <- sparse.model.matrix(Improved~.-1, data = df) # nolint
 label <- df[, ifelse(Improved == "Marked", 1, 0)]
 
 # binary
@@ -46,8 +46,8 @@ mbst.GLM <- xgboost(data = as.matrix(iris[, -5]), label = mlabel, verbose = 0,
 test_that("xgb.dump works", {
   if (!flag_32bit)
     expect_length(xgb.dump(bst.Tree), 200)
-  dump_file = file.path(tempdir(), 'xgb.model.dump')
-  expect_true(xgb.dump(bst.Tree, dump_file, with_stats = T))
+  dump_file <- file.path(tempdir(), 'xgb.model.dump')
+  expect_true(xgb.dump(bst.Tree, dump_file, with_stats = TRUE))
   expect_true(file.exists(dump_file))
   expect_gt(file.size(dump_file), 8000)
 
@@ -63,7 +63,7 @@ test_that("xgb.dump works for gblinear", {
   # also make sure that it works properly for a sparse model where some coefficients
   # are 0 from setting large L1 regularization:
   bst.GLM.sp <- xgboost(data = sparse_matrix, label = label, eta = 1, nthread = 2, nrounds = 1,
-                        alpha=2, objective = "binary:logistic", booster = "gblinear")
+                        alpha = 2, objective = "binary:logistic", booster = "gblinear")
   d.sp <- xgb.dump(bst.GLM.sp)
   expect_length(d.sp, 14)
   expect_gt(sum(d.sp == "0"), 0)
@@ -110,9 +110,9 @@ test_that("predict feature contributions works", {
   pred <- predict(bst.GLM, sparse_matrix, outputmargin = TRUE)
   expect_lt(max(abs(rowSums(pred_contr) - pred)), 1e-5)
   # manual calculation of linear terms
-  coefs <- xgb.dump(bst.GLM)[-c(1,2,4)] %>% as.numeric
+  coefs <- xgb.dump(bst.GLM)[-c(1, 2, 4)] %>% as.numeric
   coefs <- c(coefs[-1], coefs[1]) # intercept must be the last
-  pred_contr_manual <- sweep(cbind(sparse_matrix, 1), 2, coefs, FUN="*")
+  pred_contr_manual <- sweep(cbind(sparse_matrix, 1), 2, coefs, FUN = "*")
   expect_equal(as.numeric(pred_contr), as.numeric(pred_contr_manual),
                tolerance = float_tolerance)
 
@@ -130,13 +130,13 @@ test_that("predict feature contributions works", {
   pred <- predict(mbst.GLM, as.matrix(iris[, -5]), outputmargin = TRUE, reshape = TRUE)
   pred_contr <- predict(mbst.GLM, as.matrix(iris[, -5]), predcontrib = TRUE)
   expect_length(pred_contr, 3)
-  coefs_all <- xgb.dump(mbst.GLM)[-c(1,2,6)] %>% as.numeric %>% matrix(ncol = 3, byrow = TRUE)
+  coefs_all <- xgb.dump(mbst.GLM)[-c(1, 2, 6)] %>% as.numeric %>% matrix(ncol = 3, byrow = TRUE)
   for (g in seq_along(pred_contr)) {
     expect_equal(colnames(pred_contr[[g]]), c(colnames(iris[, -5]), "BIAS"))
     expect_lt(max(abs(rowSums(pred_contr[[g]]) - pred[, g])), float_tolerance)
     # manual calculation of linear terms
     coefs <- c(coefs_all[-1, g], coefs_all[1, g]) # intercept needs to be the last
-    pred_contr_manual <- sweep(as.matrix(cbind(iris[,-5], 1)), 2, coefs, FUN="*")
+    pred_contr_manual <- sweep(as.matrix(cbind(iris[, -5], 1)), 2, coefs, FUN = "*")
     expect_equal(as.numeric(pred_contr[[g]]), as.numeric(pred_contr_manual),
                  tolerance = float_tolerance)
   }
@@ -147,8 +147,8 @@ test_that("SHAPs sum to predictions, with or without DART", {
     x1 = rnorm(100),
     x2 = rnorm(100),
     x3 = rnorm(100))
-  y <- d[,"x1"] + d[,"x2"]^2 +
-    ifelse(d[,"x3"] > .5, d[,"x3"]^2, 2^d[,"x3"]) +
+  y <- d[, "x1"] + d[, "x2"]^2 +
+    ifelse(d[, "x3"] > .5, d[, "x3"]^2, 2^d[, "x3"]) +
     rnorm(100)
   nrounds <- 30
 
@@ -160,7 +160,7 @@ test_that("SHAPs sum to predictions, with or without DART", {
           objective = "reg:squarederror",
           eval_metric = "rmse"),
         if (booster == "dart")
-          list(rate_drop = .01, one_drop = T)),
+          list(rate_drop = .01, one_drop = TRUE)),
       data = d,
       label = y,
       nrounds = nrounds)
@@ -168,21 +168,21 @@ test_that("SHAPs sum to predictions, with or without DART", {
     pr <- function(...)
       predict(fit, newdata = d, ...)
     pred <- pr()
-    shap <- pr(predcontrib = T)
-    shapi <- pr(predinteraction = T)
-    tol = 1e-5
+    shap <- pr(predcontrib = TRUE)
+    shapi <- pr(predinteraction = TRUE)
+    tol <- 1e-5
 
     expect_equal(rowSums(shap), pred, tol = tol)
     expect_equal(apply(shapi, 1, sum), pred, tol = tol)
-    for (i in 1 : nrow(d))
+    for (i in seq_len(nrow(d)))
       for (f in list(rowSums, colSums))
-        expect_equal(f(shapi[i,,]), shap[i,], tol = tol)
+        expect_equal(f(shapi[i, , ]), shap[i, ], tol = tol)
   }
 })
 
 test_that("xgb-attribute functionality", {
   val <- "my attribute value"
-  list.val <- list(my_attr=val, a=123, b='ok')
+  list.val <- list(my_attr = val, a = 123, b = 'ok')
   list.ch <- list.val[order(names(list.val))]
   list.ch <- lapply(list.ch, as.character)
   # note: iter is 0-index in xgb attributes
@@ -208,9 +208,9 @@ test_that("xgb-attribute functionality", {
   xgb.attr(bst, "my_attr") <- NULL
   expect_null(xgb.attr(bst, "my_attr"))
   expect_equal(xgb.attributes(bst), list.ch[c("a", "b", "niter")])
-  xgb.attributes(bst) <- list(a=NULL, b=NULL)
+  xgb.attributes(bst) <- list(a = NULL, b = NULL)
   expect_equal(xgb.attributes(bst), list.default)
-  xgb.attributes(bst) <- list(niter=NULL)
+  xgb.attributes(bst) <- list(niter = NULL)
   expect_null(xgb.attributes(bst))
 })
 
@@ -268,7 +268,7 @@ test_that("xgb.model.dt.tree works with and without feature names", {
   bst.Tree.x$feature_names <- NULL
   dt.tree.x <- xgb.model.dt.tree(model = bst.Tree.x)
   expect_output(str(dt.tree.x), 'Feature.*\\"3\\"')
-  expect_equal(dt.tree[, -4, with=FALSE], dt.tree.x[, -4, with=FALSE])
+  expect_equal(dt.tree[, -4, with = FALSE], dt.tree.x[, -4, with = FALSE])
 
   # using integer node ID instead of character
   dt.tree.int <- xgb.model.dt.tree(model = bst.Tree, use_int_id = TRUE)
@@ -295,7 +295,7 @@ test_that("xgb.importance works with and without feature names", {
   bst.Tree.x <- bst.Tree
   bst.Tree.x$feature_names <- NULL
   importance.Tree.x <- xgb.importance(model = bst.Tree)
-  expect_equal(importance.Tree[, -1, with=FALSE], importance.Tree.x[, -1, with=FALSE],
+  expect_equal(importance.Tree[, -1, with = FALSE], importance.Tree.x[, -1, with = FALSE],
                tolerance = float_tolerance)
 
   imp2plot <- xgb.plot.importance(importance_matrix = importance.Tree)
@@ -305,7 +305,7 @@ test_that("xgb.importance works with and without feature names", {
   # for multiclass
   imp.Tree <- xgb.importance(model = mbst.Tree)
   expect_equal(dim(imp.Tree), c(4, 4))
-  xgb.importance(model = mbst.Tree, trees = seq(from=0, by=nclass, length.out=nrounds))
+  xgb.importance(model = mbst.Tree, trees = seq(from = 0, by = nclass, length.out = nrounds))
 })
 
 test_that("xgb.importance works with GLM model", {
@@ -320,7 +320,7 @@ test_that("xgb.importance works with GLM model", {
   # for multiclass
   imp.GLM <- xgb.importance(model = mbst.GLM)
   expect_equal(dim(imp.GLM), c(12, 3))
-  expect_equal(imp.GLM$Class, rep(0:2, each=4))
+  expect_equal(imp.GLM$Class, rep(0:2, each = 4))
 })
 
 test_that("xgb.model.dt.tree and xgb.importance work with a single split model", {
@@ -335,8 +335,8 @@ test_that("xgb.model.dt.tree and xgb.importance work with a single split model",
 })
 
 test_that("xgb.plot.tree works with and without feature names", {
-  xgb.plot.tree(feature_names = feature.names, model = bst.Tree)
-  xgb.plot.tree(model = bst.Tree)
+  expect_silent(xgb.plot.tree(feature_names = feature.names, model = bst.Tree))
+  expect_silent(xgb.plot.tree(model = bst.Tree))
 })
 
 test_that("xgb.plot.multi.trees works with and without feature names", {
@@ -351,11 +351,47 @@ test_that("xgb.plot.deepness works", {
   xgb.ggplot.deepness(model = bst.Tree)
 })
 
+test_that("xgb.shap.data works when top_n is provided", {
+  data_list <- xgb.shap.data(data = sparse_matrix, model = bst.Tree, top_n = 2)
+  expect_equal(names(data_list), c("data", "shap_contrib"))
+  expect_equal(NCOL(data_list$data), 2)
+  expect_equal(NCOL(data_list$shap_contrib), 2)
+  expect_equal(NROW(data_list$data), NROW(data_list$shap_contrib))
+  expect_gt(length(colnames(data_list$data)), 0)
+  expect_gt(length(colnames(data_list$shap_contrib)), 0)
+
+  # for multiclass without target class provided
+  data_list <- xgb.shap.data(data = as.matrix(iris[, -5]), model = mbst.Tree, top_n = 2)
+  expect_equal(dim(data_list$shap_contrib), c(nrow(iris), 2))
+  # for multiclass with target class provided
+  data_list <- xgb.shap.data(data = as.matrix(iris[, -5]), model = mbst.Tree, top_n = 2, target_class = 0)
+  expect_equal(dim(data_list$shap_contrib), c(nrow(iris), 2))
+})
+
+test_that("xgb.shap.data works with subsampling", {
+  data_list <- xgb.shap.data(data = sparse_matrix, model = bst.Tree, top_n = 2, subsample = 0.8)
+  expect_equal(NROW(data_list$data), as.integer(0.8 * nrow(sparse_matrix)))
+  expect_equal(NROW(data_list$data), NROW(data_list$shap_contrib))
+})
+
+test_that("prepare.ggplot.shap.data works", {
+  data_list <- xgb.shap.data(data = sparse_matrix, model = bst.Tree, top_n = 2)
+  plot_data <- prepare.ggplot.shap.data(data_list, normalize = TRUE)
+  expect_s3_class(plot_data, "data.frame")
+  expect_equal(names(plot_data), c("id", "feature", "feature_value", "shap_value"))
+  expect_s3_class(plot_data$feature, "factor")
+  # Each observation should have 1 row for each feature
+  expect_equal(nrow(plot_data), nrow(sparse_matrix) * 2)
+})
+
 test_that("xgb.plot.shap works", {
   sh <- xgb.plot.shap(data = sparse_matrix, model = bst.Tree, top_n = 2, col = 4)
   expect_equal(names(sh), c("data", "shap_contrib"))
-  expect_equal(NCOL(sh$data), 2)
-  expect_equal(NCOL(sh$shap_contrib), 2)
+})
+
+test_that("xgb.plot.shap.summary works", {
+  expect_silent(xgb.plot.shap.summary(data = sparse_matrix, model = bst.Tree, top_n = 2))
+  expect_silent(xgb.ggplot.shap.summary(data = sparse_matrix, model = bst.Tree, top_n = 2))
 })
 
 test_that("check.deprecation works", {
@@ -374,3 +410,26 @@ test_that("check.deprecation works", {
   , "\'dumm\' was partially matched to \'dummy\'")
   expect_equal(res, list(a = 1, DUMMY = 22))
 })
+
+test_that('convert.labels works', {
+  y <- c(0, 1, 0, 0, 1)
+  for (objective in c('binary:logistic', 'binary:logitraw', 'binary:hinge')) {
+    res <- xgboost:::convert.labels(y, objective_name = objective)
+    expect_s3_class(res, 'factor')
+    expect_equal(res, factor(res))
+  }
+  y <- c(0, 1, 3, 2, 1, 4)
+  for (objective in c('multi:softmax', 'multi:softprob', 'rank:pairwise', 'rank:ndcg',
+                      'rank:map')) {
+    res <- xgboost:::convert.labels(y, objective_name = objective)
+    expect_s3_class(res, 'factor')
+    expect_equal(res, factor(res))
+  }
+  y <- c(1.2, 3.0, -1.0, 10.0)
+  for (objective in c('reg:squarederror', 'reg:squaredlogerror', 'reg:logistic',
+                      'reg:pseudohubererror', 'count:poisson', 'survival:cox', 'survival:aft',
+                      'reg:gamma', 'reg:tweedie')) {
+    res <- xgboost:::convert.labels(y, objective_name = objective)
+    expect_equal(class(res), 'numeric')
+  }
+})

R-package/tests/testthat/test_interaction_constraints.R

@@ -5,20 +5,20 @@ context("interaction constraints")
 set.seed(1024)
 x1 <- rnorm(1000, 1)
 x2 <- rnorm(1000, 1)
-x3 <- sample(c(1,2,3), size=1000, replace=TRUE)
-y <- x1 + x2 + x3 + x1*x2*x3 + rnorm(1000, 0.001) + 3*sin(x1)
-train <- matrix(c(x1,x2,x3), ncol = 3)
+x3 <- sample(c(1, 2, 3), size = 1000, replace = TRUE)
+y <- x1 + x2 + x3 + x1 * x2 * x3 + rnorm(1000, 0.001) + 3 * sin(x1)
+train <- matrix(c(x1, x2, x3), ncol = 3)
 
 test_that("interaction constraints for regression", {
   # Fit a model that only allows interaction between x1 and x2
   bst <- xgboost(data = train, label = y, max_depth = 3,
                  eta = 0.1, nthread = 2, nrounds = 100, verbose = 0,
-                 interaction_constraints = list(c(0,1)))
+                 interaction_constraints = list(c(0, 1)))
 
   # Set all observations to have the same x3 values then increment
   #  by the same amount
-  preds <- lapply(c(1,2,3), function(x){
-    tmat <- matrix(c(x1,x2,rep(x,1000)), ncol=3)
+  preds <- lapply(c(1, 2, 3), function(x){
+    tmat <- matrix(c(x1, x2, rep(x, 1000)), ncol = 3)
     return(predict(bst, tmat))
   })
 
@@ -40,16 +40,16 @@ test_that("interaction constraints scientific representation", {
   rows <- 10
   ## When number exceeds 1e5, R paste function uses scientific representation.
   ## See: https://github.com/dmlc/xgboost/issues/5179
-  cols <- 1e5+10
+  cols <- 1e5 + 10
 
-  d <- matrix(rexp(rows, rate=.1), nrow=rows, ncol=cols)
+  d <- matrix(rexp(rows, rate = .1), nrow = rows, ncol = cols)
   y <- rnorm(rows)
 
-  dtrain <- xgb.DMatrix(data=d, info = list(label=y))
+  dtrain <- xgb.DMatrix(data = d, info = list(label = y))
   inc <- list(c(seq.int(from = 0, to = cols, by = 1)))
 
-  with_inc <- xgb.train(data=dtrain, tree_method='hist',
-                        interaction_constraints=inc, nrounds=10)
-  without_inc <- xgb.train(data=dtrain, tree_method='hist', nrounds=10)
+  with_inc <- xgb.train(data = dtrain, tree_method = 'hist',
+                        interaction_constraints = inc, nrounds = 10)
+  without_inc <- xgb.train(data = dtrain, tree_method = 'hist', nrounds = 10)
   expect_equal(xgb.save.raw(with_inc), xgb.save.raw(without_inc))
 })

R-package/tests/testthat/test_interactions.R

@@ -9,9 +9,9 @@ test_that("predict feature interactions works", {
   # simulate some binary data and a linear outcome with an interaction term
   N <- 1000
   P <- 5
-  X <- matrix(rbinom(N * P, 1, 0.5), ncol=P, dimnames = list(NULL, letters[1:P]))
+  X <- matrix(rbinom(N * P, 1, 0.5), ncol = P, dimnames = list(NULL, letters[1:P]))
   # center the data (as contributions are computed WRT feature means)
-  X <- scale(X, scale=FALSE)
+  X <- scale(X, scale = FALSE)
 
   # outcome without any interactions, without any noise:
   f <- function(x) 2 * x[, 1] - 3 * x[, 2]
@@ -23,14 +23,14 @@ test_that("predict feature interactions works", {
   y <- f_int(X)
 
   dm <- xgb.DMatrix(X, label = y)
-  param <- list(eta=0.1, max_depth=4, base_score=mean(y), lambda=0, nthread=2)
+  param <- list(eta = 0.1, max_depth = 4, base_score = mean(y), lambda = 0, nthread = 2)
   b <- xgb.train(param, dm, 100)
-  
-  pred = predict(b, dm, outputmargin=TRUE)
+
+  pred <- predict(b, dm, outputmargin = TRUE)
 
   # SHAP contributions:
-  cont <- predict(b, dm, predcontrib=TRUE)
-  expect_equal(dim(cont), c(N, P+1))
+  cont <- predict(b, dm, predcontrib = TRUE)
+  expect_equal(dim(cont), c(N, P + 1))
   # make sure for each row they add up to marginal predictions
   max(abs(rowSums(cont) - pred)) %>% expect_lt(0.001)
   # Hand-construct the 'ground truth' feature contributions:
@@ -39,43 +39,43 @@ test_that("predict feature interactions works", {
      -3. * X[, 2] + 1. * X[, 2] * X[, 3], # attribute a HALF of the interaction term to feature #2
       1. * X[, 2] * X[, 3]               # and another HALF of the interaction term to feature #3
      )
-  gt_cont <- cbind(gt_cont, matrix(0, nrow=N, ncol=P + 1 - 3))
+  gt_cont <- cbind(gt_cont, matrix(0, nrow = N, ncol = P + 1 - 3))
   # These should be relatively close:
   expect_lt(max(abs(cont - gt_cont)), 0.05)
 
 
   # SHAP interaction contributions:
-  intr <- predict(b, dm, predinteraction=TRUE)
-  expect_equal(dim(intr), c(N, P+1, P+1))
+  intr <- predict(b, dm, predinteraction = TRUE)
+  expect_equal(dim(intr), c(N, P + 1, P + 1))
   # check assigned colnames
   cn <- c(letters[1:P], "BIAS")
   expect_equal(dimnames(intr), list(NULL, cn, cn))
 
   # check the symmetry
-  max(abs(aperm(intr, c(1,3,2)) - intr)) %>% expect_lt(0.00001)
+  max(abs(aperm(intr, c(1, 3, 2)) - intr)) %>% expect_lt(0.00001)
 
   # sums WRT columns must be close to feature contributions
-  max(abs(apply(intr, c(1,2), sum) - cont)) %>% expect_lt(0.00001)
+  max(abs(apply(intr, c(1, 2), sum) - cont)) %>% expect_lt(0.00001)
 
   # diagonal terms for features 3,4,5 must be close to zero
   Reduce(max, sapply(3:P, function(i) max(abs(intr[, i, i])))) %>% expect_lt(0.05)
 
   # BIAS must have no interactions
-  max(abs(intr[, 1:P, P+1])) %>% expect_lt(0.00001)
+  max(abs(intr[, 1:P, P + 1])) %>% expect_lt(0.00001)
 
   # interactions other than 2 x 3 must be close to zero
   intr23 <- intr
-  intr23[,2,3] <- 0
-  Reduce(max, sapply(1:P, function(i) max(abs(intr23[, i, (i+1):(P+1)])))) %>% expect_lt(0.05)
+  intr23[, 2, 3] <- 0
+  Reduce(max, sapply(1:P, function(i) max(abs(intr23[, i, (i + 1):(P + 1)])))) %>% expect_lt(0.05)
 
   # Construct the 'ground truth' contributions of interactions directly from the linear terms:
-  gt_intr <- array(0, c(N, P+1, P+1))
-  gt_intr[,2,3] <- 1. * X[, 2] * X[, 3] # attribute a HALF of the interaction term to each symmetric element
-  gt_intr[,3,2] <- gt_intr[, 2, 3]
+  gt_intr <- array(0, c(N, P + 1, P + 1))
+  gt_intr[, 2, 3] <- 1. * X[, 2] * X[, 3] # attribute a HALF of the interaction term to each symmetric element
+  gt_intr[, 3, 2] <- gt_intr[, 2, 3]
   # merge-in the diagonal based on 'ground truth' feature contributions
-  intr_diag = gt_cont - apply(gt_intr, c(1,2), sum)
-  for(j in seq_len(P)) {
-    gt_intr[,j,j] = intr_diag[,j]
+  intr_diag <- gt_cont - apply(gt_intr, c(1, 2), sum)
+  for (j in seq_len(P)) {
+    gt_intr[, j, j] <- intr_diag[, j]
   }
   # These should be relatively close:
   expect_lt(max(abs(intr - gt_intr)), 0.1)
@@ -107,7 +107,7 @@ test_that("SHAP contribution values are not NAN", {
 
   shaps <- as.data.frame(predict(fit,
     newdata = as.matrix(subset(d, fold == 1)[, ivs]),
-    predcontrib = T))
+    predcontrib = TRUE))
   result <- cbind(shaps, sum = rowSums(shaps), pred = predict(fit,
       newdata = as.matrix(subset(d, fold == 1)[, ivs])))
 
@@ -116,26 +116,26 @@ test_that("SHAP contribution values are not NAN", {
 
 
 test_that("multiclass feature interactions work", {
-  dm <- xgb.DMatrix(as.matrix(iris[,-5]), label=as.numeric(iris$Species)-1)
-  param <- list(eta=0.1, max_depth=4, objective='multi:softprob', num_class=3)
+  dm <- xgb.DMatrix(as.matrix(iris[, -5]), label = as.numeric(iris$Species) - 1)
+  param <- list(eta = 0.1, max_depth = 4, objective = 'multi:softprob', num_class = 3)
   b <- xgb.train(param, dm, 40)
-  pred = predict(b, dm, outputmargin=TRUE) %>% array(c(3, 150)) %>% t
+  pred <- predict(b, dm, outputmargin = TRUE) %>% array(c(3, 150)) %>% t
 
   # SHAP contributions:
-  cont <- predict(b, dm, predcontrib=TRUE)
+  cont <- predict(b, dm, predcontrib = TRUE)
   expect_length(cont, 3)
   # rewrap them as a 3d array
   cont <- unlist(cont) %>% array(c(150, 5, 3))
   # make sure for each row they add up to marginal predictions
-  max(abs(apply(cont, c(1,3), sum) - pred)) %>% expect_lt(0.001)
+  max(abs(apply(cont, c(1, 3), sum) - pred)) %>% expect_lt(0.001)
 
   # SHAP interaction contributions:
-  intr <- predict(b, dm, predinteraction=TRUE)
+  intr <- predict(b, dm, predinteraction = TRUE)
   expect_length(intr, 3)
   # rewrap them as a 4d array
   intr <- unlist(intr) %>% array(c(150, 5, 5, 3)) %>% aperm(c(4, 1, 2, 3)) # [grp, row, col, col]
   # check the symmetry
-  max(abs(aperm(intr, c(1,2,4,3)) - intr)) %>% expect_lt(0.00001)
+  max(abs(aperm(intr, c(1, 2, 4, 3)) - intr)) %>% expect_lt(0.00001)
   # sums WRT columns must be close to feature contributions
-  max(abs(apply(intr, c(1,2,3), sum) - aperm(cont, c(3,1,2)))) %>% expect_lt(0.00001)
+  max(abs(apply(intr, c(1, 2, 3), sum) - aperm(cont, c(3, 1, 2)))) %>% expect_lt(0.00001)
 })

R-package/tests/testthat/test_lint.R

@@ -1,27 +0,0 @@
-context("Code is of high quality and lint free")
-test_that("Code Lint", {
-  skip_on_cran()
-  skip_on_travis()
-  skip_if_not_installed("lintr")
-  my_linters <- list(
-    absolute_paths_linter=lintr::absolute_paths_linter,
-    assignment_linter=lintr::assignment_linter,
-    closed_curly_linter=lintr::closed_curly_linter,
-    commas_linter=lintr::commas_linter,
-    # commented_code_linter=lintr::commented_code_linter,
-    infix_spaces_linter=lintr::infix_spaces_linter,
-    line_length_linter=lintr::line_length_linter,
-    no_tab_linter=lintr::no_tab_linter,
-    object_usage_linter=lintr::object_usage_linter,
-    # snake_case_linter=lintr::snake_case_linter,
-    # multiple_dots_linter=lintr::multiple_dots_linter,
-    object_length_linter=lintr::object_length_linter,
-    open_curly_linter=lintr::open_curly_linter,
-    # single_quotes_linter=lintr::single_quotes_linter,
-    spaces_inside_linter=lintr::spaces_inside_linter,
-    spaces_left_parentheses_linter=lintr::spaces_left_parentheses_linter,
-    trailing_blank_lines_linter=lintr::trailing_blank_lines_linter,
-    trailing_whitespace_linter=lintr::trailing_whitespace_linter
-  )
-  # lintr::expect_lint_free(linters=my_linters) # uncomment this if you want to check code quality
-})

R-package/tests/testthat/test_model_compatibility.R

@@ -0,0 +1,109 @@
+require(xgboost)
+require(jsonlite)
+
+context("Models from previous versions of XGBoost can be loaded")
+
+metadata <- list(
+  kRounds = 2,
+  kRows = 1000,
+  kCols = 4,
+  kForests = 2,
+  kMaxDepth = 2,
+  kClasses = 3
+)
+
+run_model_param_check <- function (config) {
+  testthat::expect_equal(config$learner$learner_model_param$num_feature, '4')
+  testthat::expect_equal(config$learner$learner_train_param$booster, 'gbtree')
+}
+
+get_num_tree <- function (booster) {
+  dump <- xgb.dump(booster)
+  m <- regexec('booster\\[[0-9]+\\]', dump, perl = TRUE)
+  m <- regmatches(dump, m)
+  num_tree <- Reduce('+', lapply(m, length))
+  return (num_tree)
+}
+
+run_booster_check <- function (booster, name) {
+  # If given a handle, we need to call xgb.Booster.complete() prior to using xgb.config().
+  if (inherits(booster, "xgb.Booster") && xgboost:::is.null.handle(booster$handle)) {
+    booster <- xgb.Booster.complete(booster)
+  }
+  config <- jsonlite::fromJSON(xgb.config(booster))
+  run_model_param_check(config)
+  if (name == 'cls') {
+    testthat::expect_equal(get_num_tree(booster),
+                           metadata$kForests * metadata$kRounds * metadata$kClasses)
+    testthat::expect_equal(as.numeric(config$learner$learner_model_param$base_score), 0.5)
+    testthat::expect_equal(config$learner$learner_train_param$objective, 'multi:softmax')
+    testthat::expect_equal(as.numeric(config$learner$learner_model_param$num_class),
+                           metadata$kClasses)
+  } else if (name == 'logitraw') {
+    testthat::expect_equal(get_num_tree(booster), metadata$kForests * metadata$kRounds)
+    testthat::expect_equal(as.numeric(config$learner$learner_model_param$num_class), 0)
+    testthat::expect_equal(config$learner$learner_train_param$objective, 'binary:logitraw')
+  } else if (name == 'logit') {
+    testthat::expect_equal(get_num_tree(booster), metadata$kForests * metadata$kRounds)
+    testthat::expect_equal(as.numeric(config$learner$learner_model_param$num_class), 0)
+    testthat::expect_equal(config$learner$learner_train_param$objective, 'binary:logistic')
+  } else if (name == 'ltr') {
+    testthat::expect_equal(get_num_tree(booster), metadata$kForests * metadata$kRounds)
+    testthat::expect_equal(config$learner$learner_train_param$objective, 'rank:ndcg')
+  } else {
+    testthat::expect_equal(name, 'reg')
+    testthat::expect_equal(get_num_tree(booster), metadata$kForests * metadata$kRounds)
+    testthat::expect_equal(as.numeric(config$learner$learner_model_param$base_score), 0.5)
+    testthat::expect_equal(config$learner$learner_train_param$objective, 'reg:squarederror')
+  }
+}
+
+test_that("Models from previous versions of XGBoost can be loaded", {
+  bucket <- 'xgboost-ci-jenkins-artifacts'
+  region <- 'us-west-2'
+  file_name <- 'xgboost_r_model_compatibility_test.zip'
+  zipfile <- file.path(getwd(), file_name)
+  model_dir <- file.path(getwd(), 'models')
+  download.file(paste('https://', bucket, '.s3-', region, '.amazonaws.com/', file_name, sep = ''),
+                destfile = zipfile, mode = 'wb', quiet = TRUE)
+  unzip(zipfile, overwrite = TRUE)
+
+  pred_data <- xgb.DMatrix(matrix(c(0, 0, 0, 0), nrow = 1, ncol = 4))
+
+  lapply(list.files(model_dir), function (x) {
+    model_file <- file.path(model_dir, x)
+    m <- regexec("xgboost-([0-9\\.]+)\\.([a-z]+)\\.[a-z]+", model_file, perl = TRUE)
+    m <- regmatches(model_file, m)[[1]]
+    model_xgb_ver <- m[2]
+    name <- m[3]
+    is_rds <- endsWith(model_file, '.rds')
+
+    cpp_warning <- capture.output({
+      # Expect an R warning when a model is loaded from RDS and it was generated by version < 1.1.x
+      if (is_rds && compareVersion(model_xgb_ver, '1.1.1.1') < 0) {
+        booster <- readRDS(model_file)
+        expect_warning(predict(booster, newdata = pred_data))
+        expect_warning(run_booster_check(booster, name))
+      } else {
+        if (is_rds) {
+          booster <- readRDS(model_file)
+        } else {
+          booster <- xgb.load(model_file)
+        }
+        predict(booster, newdata = pred_data)
+        run_booster_check(booster, name)
+      }
+    })
+    if (compareVersion(model_xgb_ver, '1.0.0.0') < 0) {
+      # Expect a C++ warning when a model was generated in version < 1.0.x
+      m <- grepl(paste0('.*Loading model from XGBoost < 1\\.0\\.0, consider saving it again for ',
+                        'improved compatibility.*'), cpp_warning, perl = TRUE)
+      expect_true(length(m) > 0 && all(m))
+    } else if (is_rds && model_xgb_ver == '1.1.1.1') {
+      # Expect a C++ warning when a model is loaded from RDS and it was generated by version 1.1.x
+      m <- grepl(paste0('.*Attempted to load internal configuration for a model file that was ',
+                        'generated by a previous version of XGBoost.*'), cpp_warning, perl = TRUE)
+      expect_true(length(m) > 0 && all(m))
+    }
+  })
+})

R-package/tests/testthat/test_monotone.R

@@ -3,22 +3,21 @@ require(xgboost)
 context("monotone constraints")
 
 set.seed(1024)
-x = rnorm(1000, 10)
-y = -1*x + rnorm(1000, 0.001) + 3*sin(x)
-train = matrix(x, ncol = 1)
+x <- rnorm(1000, 10)
+y <- -1 * x + rnorm(1000, 0.001) + 3 * sin(x)
+train <- matrix(x, ncol = 1)
 
 
 test_that("monotone constraints for regression", {
-  bst = xgboost(data = train, label = y, max_depth = 2,
-                eta = 0.1, nthread = 2, nrounds = 100, verbose = 0,
-                monotone_constraints = -1)
-  
-  pred = predict(bst, train)
-  
-  ind = order(train[,1])
-  pred.ord = pred[ind]
-  expect_true({
-    !any(diff(pred.ord) > 0)
-  }, "Monotone Contraint Satisfied")
-  
+    bst <- xgboost(data = train, label = y, max_depth = 2,
+                   eta = 0.1, nthread = 2, nrounds = 100, verbose = 0,
+                   monotone_constraints = -1)
+
+    pred <- predict(bst, train)
+
+    ind <- order(train[, 1])
+    pred.ord <- pred[ind]
+    expect_true({
+        !any(diff(pred.ord) > 0)
+    }, "Monotone Contraint Satisfied")
 })

R-package/tests/testthat/test_parameter_exposure.R

@@ -2,8 +2,8 @@ context('Test model params and call are exposed to R')
 
 require(xgboost)
 
-data(agaricus.train, package='xgboost')
-data(agaricus.test, package='xgboost')
+data(agaricus.train, package = 'xgboost')
+data(agaricus.test, package = 'xgboost')
 
 dtrain <- xgb.DMatrix(agaricus.train$data, label = agaricus.train$label)
 dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)

R-package/tests/testthat/test_poisson_regression.R

@@ -5,10 +5,10 @@ set.seed(1994)
 
 test_that("poisson regression works", {
   data(mtcars)
-  bst <- xgboost(data = as.matrix(mtcars[,-11]), label = mtcars[,11],
-                objective = 'count:poisson', nrounds=10, verbose=0)
+  bst <- xgboost(data = as.matrix(mtcars[, -11]), label = mtcars[, 11],
+                objective = 'count:poisson', nrounds = 10, verbose = 0)
   expect_equal(class(bst), "xgb.Booster")
   pred <- predict(bst, as.matrix(mtcars[, -11]))
   expect_equal(length(pred), 32)
-  expect_lt(sqrt(mean( (pred - mtcars[,11])^2 )), 1.2)
+  expect_lt(sqrt(mean((pred - mtcars[, 11])^2)), 1.2)
 })

R-package/tests/testthat/test_ranking.R

@@ -0,0 +1,51 @@
+require(xgboost)
+require(Matrix)
+
+context('Learning to rank')
+
+test_that('Test ranking with unweighted data', {
+  X <- sparseMatrix(i = c(2, 3, 7, 9, 12, 15, 17, 18),
+                    j = c(1, 1, 2, 2,  3,  3,  4,  4),
+                    x = rep(1.0, 8), dims = c(20, 4))
+  y <- c(0, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0)
+  group <- c(5, 5, 5, 5)
+  dtrain <- xgb.DMatrix(X, label = y, group = group)
+
+  params <- list(eta = 1, tree_method = 'exact', objective = 'rank:pairwise', max_depth = 1,
+                 eval_metric = 'auc', eval_metric = 'aucpr')
+  bst <- xgb.train(params, dtrain, nrounds = 10, watchlist = list(train = dtrain))
+  # Check if the metric is monotone increasing
+  expect_true(all(diff(bst$evaluation_log$train_auc) >= 0))
+  expect_true(all(diff(bst$evaluation_log$train_aucpr) >= 0))
+})
+
+test_that('Test ranking with weighted data', {
+  X <- sparseMatrix(i = c(2, 3, 7, 9, 12, 15, 17, 18),
+                    j = c(1, 1, 2, 2,  3,  3,  4,  4),
+                    x = rep(1.0, 8), dims = c(20, 4))
+  y <- c(0, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0)
+  group <- c(5, 5, 5, 5)
+  weight <- c(1.0, 2.0, 3.0, 4.0)
+  dtrain <- xgb.DMatrix(X, label = y, group = group, weight = weight)
+
+  params <- list(eta = 1, tree_method = 'exact', objective = 'rank:pairwise', max_depth = 1,
+                 eval_metric = 'auc', eval_metric = 'aucpr')
+  bst <- xgb.train(params, dtrain, nrounds = 10, watchlist = list(train = dtrain))
+  # Check if the metric is monotone increasing
+  expect_true(all(diff(bst$evaluation_log$train_auc) >= 0))
+  expect_true(all(diff(bst$evaluation_log$train_aucpr) >= 0))
+  for (i in 1:10) {
+    pred <- predict(bst, newdata = dtrain, ntreelimit = i)
+    # is_sorted[i]: is i-th group correctly sorted by the ranking predictor?
+    is_sorted <- lapply(seq(1, 20, by = 5),
+      function (k) {
+        ind <- order(-pred[k:(k + 4)])
+        z <- y[ind + (k - 1)]
+        all(diff(z) <= 0)  # Check if z is monotone decreasing
+      })
+    # Since we give weights 1, 2, 3, 4 to the four query groups,
+    # the ranking predictor will first try to correctly sort the last query group
+    # before correctly sorting other groups.
+    expect_true(all(diff(as.numeric(is_sorted)) >= 0))
+  }
+})

R-package/tests/testthat/test_update.R

@@ -9,23 +9,23 @@ dtest <- xgb.DMatrix(agaricus.test$data, label = agaricus.test$label)
 
 # Disable flaky tests for 32-bit Windows.
 # See https://github.com/dmlc/xgboost/issues/3720
-win32_flag = .Platform$OS.type == "windows" && .Machine$sizeof.pointer != 8
+win32_flag <- .Platform$OS.type == "windows" && .Machine$sizeof.pointer != 8
 
 test_that("updating the model works", {
-  watchlist = list(train = dtrain, test = dtest)
+  watchlist <- list(train = dtrain, test = dtest)
 
   # no-subsampling
   p1 <- list(objective = "binary:logistic", max_depth = 2, eta = 0.05, nthread = 2)
   set.seed(11)
   bst1 <- xgb.train(p1, dtrain, nrounds = 10, watchlist, verbose = 0)
   tr1 <- xgb.model.dt.tree(model = bst1)
-  
+
   # with subsampling
   p2 <- modifyList(p1, list(subsample = 0.1))
   set.seed(11)
   bst2 <- xgb.train(p2, dtrain, nrounds = 10, watchlist, verbose = 0)
   tr2 <- xgb.model.dt.tree(model = bst2)
-  
+
   # the same no-subsampling boosting with an extra 'refresh' updater:
   p1r <- modifyList(p1, list(updater = 'grow_colmaker,prune,refresh', refresh_leaf = FALSE))
   set.seed(11)
@@ -57,7 +57,7 @@ test_that("updating the model works", {
   # all should be the same when no subsampling
   expect_equal(bst1$evaluation_log, bst1u$evaluation_log)
   expect_equal(tr1, tr1u, tolerance = 0.00001, check.attributes = FALSE)
-  
+
   # process type 'update' for model with subsampling, refreshing only the tree stats from training data:
   p2u <- modifyList(p2, list(process_type = 'update', updater = 'refresh', refresh_leaf = FALSE))
   bst2u <- xgb.train(p2u, dtrain, nrounds = 10, watchlist, verbose = 0, xgb_model = bst2)
@@ -72,7 +72,7 @@ test_that("updating the model works", {
   if (!win32_flag) {
     expect_equal(tr2r, tr2u, tolerance = 0.00001, check.attributes = FALSE)
   }
-  
+
   # process type 'update' for no-subsampling model, refreshing only the tree stats from TEST data:
   p1ut <- modifyList(p1, list(process_type = 'update', updater = 'refresh', refresh_leaf = FALSE))
   bst1ut <- xgb.train(p1ut, dtest, nrounds = 10, watchlist, verbose = 0, xgb_model = bst1)
@@ -93,12 +93,12 @@ test_that("updating works for multiclass & multitree", {
   set.seed(121)
   bst0 <- xgb.train(p0, dtr, 5, watchlist, verbose = 0)
   tr0 <- xgb.model.dt.tree(model = bst0)
-  
+
   # run update process for an original model with subsampling
-  p0u <- modifyList(p0, list(process_type='update', updater='refresh', refresh_leaf=FALSE))
+  p0u <- modifyList(p0, list(process_type = 'update', updater = 'refresh', refresh_leaf = FALSE))
   bst0u <- xgb.train(p0u, dtr, nrounds = bst0$niter, watchlist, xgb_model = bst0, verbose = 0)
   tr0u <- xgb.model.dt.tree(model = bst0u)
-  
+
   # should be the same evaluation but different gains and larger cover
   expect_equal(bst0$evaluation_log, bst0u$evaluation_log)
   expect_equal(tr0[Feature == 'Leaf']$Quality, tr0u[Feature == 'Leaf']$Quality)