Release 1.1.0

Co-authored-by: fis <jm.yuan@outlook.com>

Jenkinsfile

@@ -64,7 +64,6 @@ pipeline {
             'build-cpu': { BuildCPU() },
             'build-cpu-rabit-mock': { BuildCPUMock() },
             'build-cpu-non-omp': { BuildCPUNonOmp() },
-            'build-gpu-cuda9.0': { BuildCUDA(cuda_version: '9.0') },
             'build-gpu-cuda10.0': { BuildCUDA(cuda_version: '10.0') },
             'build-gpu-cuda10.1': { BuildCUDA(cuda_version: '10.1') },
             'build-jvm-packages': { BuildJVMPackages(spark_version: '2.4.3') },
@@ -251,10 +250,10 @@ def BuildCUDA(args) {
     ${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 ${commit_id} manylinux2010_x86_64
     """
-    // Stash wheel for CUDA 9.0 target
-    if (args.cuda_version == '9.0') {
+    // Stash wheel for CUDA 10.0 target
+    if (args.cuda_version == '10.0') {
       echo 'Stashing Python wheel...'
-      stash name: 'xgboost_whl_cuda9', includes: 'python-package/dist/*.whl'
+      stash name: 'xgboost_whl_cuda10', includes: 'python-package/dist/*.whl'
       path = ("${BRANCH_NAME}" == 'master') ? '' : "${BRANCH_NAME}/"
       s3Upload bucket: 'xgboost-nightly-builds', path: path, acl: 'PublicRead', workingDir: 'python-package/dist', includePathPattern:'**/*.whl'
       echo 'Stashing C++ test executable (testxgboost)...'
@@ -298,7 +297,7 @@ def BuildJVMDoc() {
 
 def TestPythonCPU() {
   node('linux && cpu') {
-    unstash name: 'xgboost_whl_cuda9'
+    unstash name: 'xgboost_whl_cuda10'
     unstash name: 'srcs'
     unstash name: 'xgboost_cli'
     echo "Test Python CPU"
@@ -315,7 +314,7 @@ def TestPythonCPU() {
 def TestPythonGPU(args) {
   nodeReq = (args.multi_gpu) ? 'linux && mgpu' : 'linux && gpu'
   node(nodeReq) {
-    unstash name: 'xgboost_whl_cuda9'
+    unstash name: 'xgboost_whl_cuda10'
     unstash name: 'srcs'
     echo "Test Python GPU: CUDA ${args.cuda_version}"
     def container_type = "gpu"

cmake/Python_version.in

@@ -1 +1 @@
-@xgboost_VERSION_MAJOR@.@xgboost_VERSION_MINOR@.@xgboost_VERSION_PATCH@rc2
+@xgboost_VERSION_MAJOR@.@xgboost_VERSION_MINOR@.@xgboost_VERSION_PATCH@

doc/dump.schema

@@ -0,0 +1,55 @@
+{
+    "$schema": "http://json-schema.org/draft-07/schema#",
+    "definitions": {
+        "split_node": {
+            "type": "object",
+            "properties": {
+                "nodeid": {
+                    "type": "number",
+                    "minimum": 0
+                },
+                "depth": {
+                    "type": "number",
+                    "minimum": 0
+                },
+                "yes": {
+                    "type": "number",
+                    "minimum": 0
+                },
+                "no": {
+                    "type": "number",
+                    "minimum": 0
+                },
+                "split": {
+                    "type": "string"
+                },
+                "children": {
+                    "type": "array",
+                    "items": {
+                        "oneOf": [
+                            {"$ref": "#/definitions/split_node"},
+                            {"$ref": "#/definitions/leaf_node"}
+                        ]
+                    },
+                    "maxItems": 2
+                }
+            },
+            "required": ["nodeid", "depth", "yes", "no", "split", "children"]
+        },
+        "leaf_node": {
+            "type": "object",
+            "properties": {
+                "nodeid": {
+                    "type": "number",
+                    "minimum": 0
+                },
+                "leaf": {
+                    "type": "number"
+                }
+            },
+            "required": ["nodeid", "leaf"]
+        }
+    },
+    "type": "object",
+    "$ref": "#/definitions/split_node"
+}

jvm-packages/pom.xml

@@ -6,7 +6,7 @@
 
     <groupId>ml.dmlc</groupId>
     <artifactId>xgboost-jvm_2.12</artifactId>
-    <version>1.1.0-RC2</version>
+    <version>1.1.0</version>
     <packaging>pom</packaging>
     <name>XGBoost JVM Package</name>
     <description>JVM Package for XGBoost</description>

jvm-packages/xgboost4j-example/pom.xml

@@ -6,10 +6,10 @@
     <parent>
         <groupId>ml.dmlc</groupId>
         <artifactId>xgboost-jvm_2.12</artifactId>
-        <version>1.1.0-RC2</version>
+        <version>1.1.0</version>
     </parent>
     <artifactId>xgboost4j-example_2.12</artifactId>
-    <version>1.1.0-RC2</version>
+    <version>1.1.0</version>
     <packaging>jar</packaging>
     <build>
         <plugins>
@@ -26,7 +26,7 @@
         <dependency>
             <groupId>ml.dmlc</groupId>
             <artifactId>xgboost4j-spark_${scala.binary.version}</artifactId>
-            <version>1.1.0-RC2</version>
+            <version>1.1.0</version>
         </dependency>
         <dependency>
             <groupId>org.apache.spark</groupId>
@@ -37,7 +37,7 @@
         <dependency>
             <groupId>ml.dmlc</groupId>
             <artifactId>xgboost4j-flink_${scala.binary.version}</artifactId>
-            <version>1.1.0-RC2</version>
+            <version>1.1.0</version>
         </dependency>
         <dependency>
             <groupId>org.apache.commons</groupId>

jvm-packages/xgboost4j-flink/pom.xml

@@ -6,10 +6,10 @@
     <parent>
         <groupId>ml.dmlc</groupId>
         <artifactId>xgboost-jvm_2.12</artifactId>
-        <version>1.1.0-RC2</version>
+        <version>1.1.0</version>
     </parent>
     <artifactId>xgboost4j-flink_2.12</artifactId>
-    <version>1.1.0-RC2</version>
+    <version>1.1.0</version>
     <build>
         <plugins>
             <plugin>
@@ -26,7 +26,7 @@
         <dependency>
             <groupId>ml.dmlc</groupId>
             <artifactId>xgboost4j_${scala.binary.version}</artifactId>
-            <version>1.1.0-RC2</version>
+            <version>1.1.0</version>
         </dependency>
         <dependency>
             <groupId>org.apache.commons</groupId>

jvm-packages/xgboost4j-spark/pom.xml

@@ -6,7 +6,7 @@
     <parent>
         <groupId>ml.dmlc</groupId>
         <artifactId>xgboost-jvm_2.12</artifactId>
-        <version>1.1.0-RC2</version>
+        <version>1.1.0</version>
     </parent>
     <artifactId>xgboost4j-spark_2.12</artifactId>
     <build>
@@ -24,7 +24,7 @@
         <dependency>
             <groupId>ml.dmlc</groupId>
             <artifactId>xgboost4j_${scala.binary.version}</artifactId>
-            <version>1.1.0-RC2</version>
+            <version>1.1.0</version>
         </dependency>
         <dependency>
             <groupId>org.apache.spark</groupId>

jvm-packages/xgboost4j/pom.xml

@@ -6,10 +6,10 @@
     <parent>
         <groupId>ml.dmlc</groupId>
         <artifactId>xgboost-jvm_2.12</artifactId>
-        <version>1.1.0-RC2</version>
+        <version>1.1.0</version>
     </parent>
     <artifactId>xgboost4j_2.12</artifactId>
-    <version>1.1.0-RC2</version>
+    <version>1.1.0</version>
     <packaging>jar</packaging>
 
     <dependencies>

python-package/xgboost/VERSION

@@ -1 +1 @@
-1.1.0rc2
+1.1.0

src/common/column_matrix.h

@@ -82,14 +82,16 @@ template <typename BinIdxType>
 class DenseColumn: public Column<BinIdxType> {
  public:
   DenseColumn(ColumnType type, common::Span<const BinIdxType> index,
-              uint32_t index_base,
-              const std::vector<bool>::const_iterator missing_flags)
+              uint32_t index_base, const std::vector<bool>& missing_flags,
+              size_t feature_offset)
       : Column<BinIdxType>(type, index, index_base),
-        missing_flags_(missing_flags) {}
-  bool IsMissing(size_t idx) const { return missing_flags_[idx]; }
+        missing_flags_(missing_flags),
+        feature_offset_(feature_offset) {}
+  bool IsMissing(size_t idx) const { return missing_flags_[feature_offset_ + idx]; }
  private:
   /* flags for missing values in dense columns */
-  std::vector<bool>::const_iterator missing_flags_;
+  const std::vector<bool>& missing_flags_;
+  size_t feature_offset_;
 };
 
 /*! \brief a collection of columns, with support for construction from
@@ -208,10 +210,8 @@ class ColumnMatrix {
                                                  column_size };
     std::unique_ptr<const Column<BinIdxType> > res;
     if (type_[fid] == ColumnType::kDenseColumn) {
-      std::vector<bool>::const_iterator column_iterator = missing_flags_.begin();
-      advance(column_iterator, feature_offset);  // increment iterator to right position
       res.reset(new DenseColumn<BinIdxType>(type_[fid], bin_index, index_base_[fid],
-                                            column_iterator));
+                                             missing_flags_, feature_offset));
     } else {
       res.reset(new SparseColumn<BinIdxType>(type_[fid], bin_index, index_base_[fid],
                                        {&row_ind_[feature_offset], column_size}));

src/tree/tree_model.cc

@@ -68,20 +68,20 @@ class TreeGenerator {
     return result;
   }
 
-  virtual std::string Indicator(RegTree const& tree, int32_t nid, uint32_t depth) {
+  virtual std::string Indicator(RegTree const& tree, int32_t nid, uint32_t depth) const {
     return "";
   }
-  virtual std::string Integer(RegTree const& tree, int32_t nid, uint32_t depth) {
+  virtual std::string Integer(RegTree const& tree, int32_t nid, uint32_t depth) const {
     return "";
   }
-  virtual std::string Quantitive(RegTree const& tree, int32_t nid, uint32_t depth) {
+  virtual std::string Quantitive(RegTree const& tree, int32_t nid, uint32_t depth) const {
     return "";
   }
-  virtual std::string NodeStat(RegTree const& tree, int32_t nid) {
+  virtual std::string NodeStat(RegTree const& tree, int32_t nid) const {
     return "";
   }
 
-  virtual std::string PlainNode(RegTree const& tree, int32_t nid, uint32_t depth) = 0;
+  virtual std::string PlainNode(RegTree const& tree, int32_t nid, uint32_t depth) const = 0;
 
   virtual std::string SplitNode(RegTree const& tree, int32_t nid, uint32_t depth) {
     auto const split_index = tree[nid].SplitIndex();
@@ -110,7 +110,7 @@ class TreeGenerator {
     return result;
   }
 
-  virtual std::string LeafNode(RegTree const& tree, int32_t nid, uint32_t depth) = 0;
+  virtual std::string LeafNode(RegTree const& tree, int32_t nid, uint32_t depth) const = 0;
   virtual std::string BuildTree(RegTree const& tree, int32_t nid, uint32_t depth) = 0;
 
  public:
@@ -181,7 +181,7 @@ class TextGenerator : public TreeGenerator {
   TextGenerator(FeatureMap const& fmap, std::string const& attrs, bool with_stats) :
       TreeGenerator(fmap, with_stats) {}
 
-  std::string LeafNode(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string LeafNode(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     static std::string kLeafTemplate = "{tabs}{nid}:leaf={leaf}{stats}";
     static std::string kStatTemplate = ",cover={cover}";
     std::string result = SuperT::Match(
@@ -195,7 +195,7 @@ class TextGenerator : public TreeGenerator {
     return result;
   }
 
-  std::string Indicator(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string Indicator(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     static std::string const kIndicatorTemplate = "{nid}:[{fname}] yes={yes},no={no}";
     int32_t nyes = tree[nid].DefaultLeft() ?
                    tree[nid].RightChild() : tree[nid].LeftChild();
@@ -211,7 +211,7 @@ class TextGenerator : public TreeGenerator {
 
   std::string SplitNodeImpl(
       RegTree const& tree, int32_t nid, std::string const& template_str,
-      std::string cond, uint32_t depth) {
+      std::string cond, uint32_t depth) const {
     auto split_index = tree[nid].SplitIndex();
     std::string const result = SuperT::Match(
         template_str,
@@ -226,7 +226,7 @@ class TextGenerator : public TreeGenerator {
     return result;
   }
 
-  std::string Integer(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string Integer(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     static std::string const kIntegerTemplate =
         "{tabs}{nid}:[{fname}<{cond}] yes={left},no={right},missing={missing}";
     auto cond = tree[nid].SplitCond();
@@ -238,21 +238,21 @@ class TextGenerator : public TreeGenerator {
                          std::to_string(integer_threshold), depth);
   }
 
-  std::string Quantitive(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string Quantitive(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     static std::string const kQuantitiveTemplate =
         "{tabs}{nid}:[{fname}<{cond}] yes={left},no={right},missing={missing}";
     auto cond = tree[nid].SplitCond();
     return SplitNodeImpl(tree, nid, kQuantitiveTemplate, SuperT::ToStr(cond), depth);
   }
 
-  std::string PlainNode(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string PlainNode(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     auto cond = tree[nid].SplitCond();
     static std::string const kNodeTemplate =
         "{tabs}{nid}:[f{fname}<{cond}] yes={left},no={right},missing={missing}";
     return SplitNodeImpl(tree, nid, kNodeTemplate, SuperT::ToStr(cond), depth);
   }
 
-  std::string NodeStat(RegTree const& tree, int32_t nid) override {
+  std::string NodeStat(RegTree const& tree, int32_t nid) const override {
     static std::string const kStatTemplate = ",gain={loss_chg},cover={sum_hess}";
     std::string const result = SuperT::Match(
         kStatTemplate,
@@ -297,7 +297,7 @@ class JsonGenerator : public TreeGenerator {
   JsonGenerator(FeatureMap const& fmap, std::string attrs, bool with_stats) :
       TreeGenerator(fmap, with_stats) {}
 
-  std::string Indent(uint32_t depth) {
+  std::string Indent(uint32_t depth) const {
     std::string result;
     for (uint32_t i = 0; i < depth + 1; ++i) {
       result += "  ";
@@ -305,7 +305,7 @@ class JsonGenerator : public TreeGenerator {
     return result;
   }
 
-  std::string LeafNode(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string LeafNode(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     static std::string const kLeafTemplate =
         R"L({ "nodeid": {nid}, "leaf": {leaf} {stat}})L";
     static std::string const kStatTemplate =
@@ -321,11 +321,11 @@ class JsonGenerator : public TreeGenerator {
     return result;
   }
 
-  std::string Indicator(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string Indicator(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     int32_t nyes = tree[nid].DefaultLeft() ?
                    tree[nid].RightChild() : tree[nid].LeftChild();
     static std::string const kIndicatorTemplate =
-        R"ID( "nodeid": {nid}, "depth": {depth}, "split": "{fname}", "yes": {yes}, "no": {no}})ID";
+        R"ID( "nodeid": {nid}, "depth": {depth}, "split": "{fname}", "yes": {yes}, "no": {no})ID";
     auto split_index = tree[nid].SplitIndex();
     auto result = SuperT::Match(
         kIndicatorTemplate,
@@ -337,8 +337,9 @@ class JsonGenerator : public TreeGenerator {
     return result;
   }
 
-  std::string SplitNodeImpl(RegTree const& tree, int32_t nid,
-                            std::string const& template_str, std::string cond, uint32_t depth) {
+  std::string SplitNodeImpl(RegTree const &tree, int32_t nid,
+                            std::string const &template_str, std::string cond,
+                            uint32_t depth) const {
     auto split_index = tree[nid].SplitIndex();
     std::string const result = SuperT::Match(
         template_str,
@@ -353,7 +354,7 @@ class JsonGenerator : public TreeGenerator {
     return result;
   }
 
-  std::string Integer(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string Integer(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     auto cond = tree[nid].SplitCond();
     const bst_float floored = std::floor(cond);
     const int32_t integer_threshold
@@ -367,7 +368,7 @@ class JsonGenerator : public TreeGenerator {
                          std::to_string(integer_threshold), depth);
   }
 
-  std::string Quantitive(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string Quantitive(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     static std::string const kQuantitiveTemplate =
         R"I( "nodeid": {nid}, "depth": {depth}, "split": "{fname}", )I"
         R"I("split_condition": {cond}, "yes": {left}, "no": {right}, )I"
@@ -376,7 +377,7 @@ class JsonGenerator : public TreeGenerator {
     return SplitNodeImpl(tree, nid, kQuantitiveTemplate, SuperT::ToStr(cond), depth);
   }
 
-  std::string PlainNode(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string PlainNode(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     auto cond = tree[nid].SplitCond();
     static std::string const kNodeTemplate =
         R"I( "nodeid": {nid}, "depth": {depth}, "split": {fname}, )I"
@@ -385,7 +386,7 @@ class JsonGenerator : public TreeGenerator {
     return SplitNodeImpl(tree, nid, kNodeTemplate, SuperT::ToStr(cond), depth);
   }
 
-  std::string NodeStat(RegTree const& tree, int32_t nid) override {
+  std::string NodeStat(RegTree const& tree, int32_t nid) const override {
     static std::string kStatTemplate =
         R"S(, "gain": {loss_chg}, "cover": {sum_hess})S";
     auto result = SuperT::Match(
@@ -529,7 +530,7 @@ class GraphvizGenerator : public TreeGenerator {
  protected:
   // Only indicator is different, so we combine all different node types into this
   // function.
-  std::string PlainNode(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string PlainNode(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     auto split = tree[nid].SplitIndex();
     auto cond = tree[nid].SplitCond();
     static std::string const kNodeTemplate =
@@ -563,7 +564,7 @@ class GraphvizGenerator : public TreeGenerator {
     return result;
   };
 
-  std::string LeafNode(RegTree const& tree, int32_t nid, uint32_t depth) override {
+  std::string LeafNode(RegTree const& tree, int32_t nid, uint32_t depth) const override {
     static std::string const kLeafTemplate =
         "    {nid} [ label=\"leaf={leaf-value}\" {params}]\n";
     auto result = SuperT::Match(kLeafTemplate, {

tests/cpp/tree/test_tree_model.cc

@@ -151,6 +151,10 @@ TEST(Tree, DumpJson) {
 
   str = tree.DumpModel(fmap, false, "json");
   ASSERT_EQ(str.find("cover"), std::string::npos);
+
+
+  auto j_tree = Json::Load({str.c_str(), str.size()});
+  ASSERT_EQ(get<Array>(j_tree["children"]).size(), 2);
 }
 
 TEST(Tree, DumpText) {

tests/python-gpu/test_gpu_basic_models.py

@@ -12,25 +12,15 @@ rng = np.random.RandomState(1994)
 class TestGPUBasicModels(unittest.TestCase):
     cputest = test_bm.TestModels()
 
-    def test_eta_decay_gpu_hist(self):
-        self.cputest.run_eta_decay('gpu_hist')
-
-    def test_deterministic_gpu_hist(self):
-        kRows = 1000
-        kCols = 64
-        kClasses = 4
-        # Create large values to force rounding.
-        X = np.random.randn(kRows, kCols) * 1e4
-        y = np.random.randint(0, kClasses, size=kRows)
-
+    def run_cls(self, X, y, deterministic):
         cls = xgb.XGBClassifier(tree_method='gpu_hist',
-                                deterministic_histogram=True,
+                                deterministic_histogram=deterministic,
                                 single_precision_histogram=True)
         cls.fit(X, y)
         cls.get_booster().save_model('test_deterministic_gpu_hist-0.json')
 
         cls = xgb.XGBClassifier(tree_method='gpu_hist',
-                                deterministic_histogram=True,
+                                deterministic_histogram=deterministic,
                                 single_precision_histogram=True)
         cls.fit(X, y)
         cls.get_booster().save_model('test_deterministic_gpu_hist-1.json')
@@ -40,7 +30,24 @@ class TestGPUBasicModels(unittest.TestCase):
         with open('test_deterministic_gpu_hist-1.json', 'r') as fd:
             model_1 = fd.read()
 
-        assert hash(model_0) == hash(model_1)
-
         os.remove('test_deterministic_gpu_hist-0.json')
         os.remove('test_deterministic_gpu_hist-1.json')
+
+        return hash(model_0), hash(model_1)
+
+    def test_eta_decay_gpu_hist(self):
+        self.cputest.run_eta_decay('gpu_hist')
+
+    def test_deterministic_gpu_hist(self):
+        kRows = 1000
+        kCols = 64
+        kClasses = 4
+        # Create large values to force rounding.
+        X = np.random.randn(kRows, kCols) * 1e4
+        y = np.random.randint(0, kClasses, size=kRows) * 1e4
+
+        model_0, model_1 = self.run_cls(X, y, True)
+        assert model_0 == model_1
+
+        model_0, model_1 = self.run_cls(X, y, False)
+        assert model_0 != model_1

tests/python/test_basic_models.py

@@ -325,7 +325,7 @@ class TestModels(unittest.TestCase):
         assert locale.getpreferredencoding(False) == loc
 
     @pytest.mark.skipif(**tm.no_json_schema())
-    def test_json_schema(self):
+    def test_json_io_schema(self):
         import jsonschema
         model_path = 'test_json_schema.json'
         path = os.path.dirname(
@@ -342,3 +342,35 @@ class TestModels(unittest.TestCase):
         jsonschema.validate(instance=json_model(model_path, parameters),
                             schema=schema)
         os.remove(model_path)
+
+    @pytest.mark.skipif(**tm.no_json_schema())
+    def test_json_dump_schema(self):
+        import jsonschema
+
+        def validate_model(parameters):
+            X = np.random.random((100, 30))
+            y = np.random.randint(0, 4, size=(100,))
+
+            parameters['num_class'] = 4
+            m = xgb.DMatrix(X, y)
+
+            booster = xgb.train(parameters, m)
+            dump = booster.get_dump(dump_format='json')
+
+            for i in range(len(dump)):
+                jsonschema.validate(instance=json.loads(dump[i]),
+                                    schema=schema)
+
+        path = os.path.dirname(
+            os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+        doc = os.path.join(path, 'doc', 'dump.schema')
+        with open(doc, 'r') as fd:
+            schema = json.load(fd)
+
+        parameters = {'tree_method': 'hist', 'booster': 'gbtree',
+                      'objective': 'multi:softmax'}
+        validate_model(parameters)
+
+        parameters = {'tree_method': 'hist', 'booster': 'dart',
+                      'objective': 'multi:softmax'}
+        validate_model(parameters)