change toolkit to rabitlearn

README.md

@@ -8,29 +8,19 @@ rabit is a light weight library that provides a fault tolerant interface of Allr
 
 Features
 ====
-* Portable library
+All these features comes from the facts about small rabbit:)
-  - Rabit is a library instead of a framework, a program only needs to link the library to run.
+* Portable: rabit is light weight and runs everywhere
-* Flexibility in programming
+  - Rabit is a library instead of a framework, a program only needs to link the library to run
+  - Rabit only replies on a mechanism to start program, which was provided by most framework
+  - You can port rabit programs easily to many frameworks, including Hadoop, MPI without changing your code
+* Scalable and Flexible: rabit runs fast
+  * Rabit program use Allreduce to communicate, and do not suffer the cost between iterations of MapReduce abstraction.
   - Programs can call rabit functions in any order, as opposed to frameworks where callbacks are offered and called by the framework, i.e. inversion of control principle.
   - Programs persist over all the iterations, unless they fail and recover.
-* Fault tolerance 
+* Fault Tolerant: rabit dig burrows to avoid disasters
   - Rabit programs can recover the model and results using synchronous function calls.
-* MPI compatible
-  - Code that uses the rabit interface also compiles with existing MPI compilers
-  - Users can use MPI Allreduce with no code modification
 
 Use Rabit
 ====
 * Type make in the root folder will compile the rabit library in lib folder
 * Add lib to the library path and include to the include path of compiler
-
-Design Notes
-====
-* Rabit is designed for algorithms that replicate the same global model across nodes, while each node operates on a local partition of the data.
-* The collection of global statistics is done using Allreduce
-
-Design Goals
-====
-* rabit should run fast
-* rabit should be light weight
-* rabit should safely dig burrows to avoid disasters

rabit-learn/README.md

@@ -0,0 +1,17 @@
+Rabit-Learn
+====
+This folder contains implementation of distributed machine learning algorithm using rabit.
+It also contain links to the Machine Learning packages that uses rabit.
+
+* Contribution of toolkits, examples, benchmarks is more than welcomed!
+
+Toolkits
+====
+* [KMeans Clustering](kmeans)
+* [XGBoost: eXtreme Gradient Boosting](https://github.com/tqchen/xgboost/tree/unity/multi-node)
+  - xgboost is a very fast boosted tree(also known as GBDT) library, that can run more than
+    10 times faster than existing packages
+  - Rabit carries xgboost to distributed enviroment, inheritating all the benefits of xgboost
+    single node version, and scale it to even larger problems
+
+

rabit-learn/common.mk

@@ -1,29 +1,19 @@
+# this is the common build script for rabit programs
+# you do not have to use it 
 export CC  = gcc
 export CXX = g++
 export MPICXX = mpicxx
-export LDFLAGS= -pthread -lm -L../lib
+export LDFLAGS= -pthread -lm -L../../lib
-export CFLAGS = -Wall -O3 -msse2  -Wno-unknown-pragmas -fPIC -I../include 
+export CFLAGS = -Wall -O3 -msse2  -Wno-unknown-pragmas -fPIC -I../../include -I../common
 
-# specify tensor path
+.PHONY: clean all lib mpi
-BIN = kmeans.rabit 
+all: $(BIN) $(MOCKBIN)
-MOCKBIN= kmeans.mock
+mpi: $(MPIBIN)
-MPIBIN = kmeans.mpi
-# objectives that makes up rabit library
-OBJ = kmeans.o
-.PHONY: clean all lib 
-all: $(BIN)
 
 lib:
-	cd ..;make lib/librabit.a lib/librabit_mock.a; cd -
+	cd ../..;make lib/librabit.a lib/librabit_mock.a; cd -
 libmpi:
-	cd ..;make lib/librabit_mpi.a;cd -
+	cd ../..;make lib/librabit_mpi.a;cd -
-
-kmeans.o: kmeans.cc ../src/*.h
-
-# we can link against MPI version to get use MPI
-kmeans.rabit: kmeans.o lib
-kmeans.mock: kmeans.o lib
-kmeans.mpi: kmeans.o libmpi
 
 $(BIN) : 
 	$(CXX) $(CFLAGS) -o $@ $(filter %.cpp %.o %.c %.cc,  $^) $(LDFLAGS) -lrabit

rabit-learn/kmeans/Makefile

@@ -0,0 +1,16 @@
+# specify tensor path
+BIN = kmeans.rabit 
+MOCKBIN= kmeans.mock
+MPIBIN = kmeans.mpi
+# objectives that makes up rabit library
+OBJ = kmeans.o
+
+# common build script for programs
+include ../common.mk
+
+# dependenies here
+kmeans.rabit: kmeans.o lib
+kmeans.mock: kmeans.o lib
+kmeans.mpi: kmeans.o libmpi
+kmeans.o: kmeans.cc ../../src/*.h
+

toolkit/.gitignore

@@ -1,2 +0,0 @@
-kmeans
-*.mpi