微软开源分布式高性能GB框架LightGBM MacOS下编译安装过程

2017-04-14 08:24:06来源:CSDN作者:testcs_dn人点击

LightGBM(Light Gradient Boosting Machine)是一个基于决策树算法的快速的、分布式的、高性能 gradient boosting(GBDT、GBRT、GBM 或 MART)框架,可被用于排行、分类以及其他许多机器学习任务中。
开源项目地址: https://github.com/Microsoft/LightGBM

开源|LightGBM:三天内收获GitHub 1000+ Star,表现超越已有boosting工具。

LightGBM

LightGBM 是一个梯度 boosting 框架,使用基于学习算法的决策树。它可以说是分布式的,高效的,它有以下优势:
更快的训练效率
低内存使用
更好的准确率
支持并行学习
可处理大规模数据

环境描述

操作系统:mac os Sierra 10.12

iMac (21.5-inch, Late 2015)

处理器:1.6 GHz Intel Core i5内存:8 GB 1867 MHz DDR3图形卡:Intel HD Graphics 6000 1536 MB

编译过程

lightgbm 依赖 OpenMP 进行编译,苹果的 Clang 不支持。请使用 GCC / G++ 代替。参考官方给出的安装指南

brew install cmakebrew install gcc --without-multilibgit clone --recursive https://github.com/Microsoft/LightGBM ; cd LightGBMmkdir build ; cd buildcmake -DCMAKE_CXX_COMPILER=g++-6 -DCMAKE_C_COMPILER=gcc-6 .. make -j 
以下为我逐条执行的输出:1、使用 brew 安装 cmake,CentOS下使用 yum install cmake,Ubuntu下使用apt-get install cmake

$ brew install cmakeUpdating Homebrew...==> Downloading https://homebrew.bintray.com/bottles/cmake-3.7.2.sierra.bottle.tar.gz######################################################################## 100.0%==> Pouring cmake-3.7.2.sierra.bottle.tar.gz==> CaveatsEmacs Lisp files have been installed to:  /usr/local/share/emacs/site-lisp/cmake==> Summary  /usr/local/Cellar/cmake/3.7.2: 2,143 files, 29MBavenMac:conf aven$ cmakeUsage  cmake [options] <path-to-source>  cmake [options] <path-to-existing-build>Specify a source directory to (re-)generate a build system for it in thecurrent working directory.  Specify an existing build directory tore-generate its build system.Run 'cmake --help' for more information.

2、安装gcc,CentOS下使用 yum install gcc,Ubuntu下使用apt-get install gcc,可能还需要安装g++

$ brew install gcc --without-multilibUpdating Homebrew...==> Auto-updated Homebrew!Updated 1 tap (homebrew/core).==> Updated Formulaeaws-sdk-cpp              docker                   go                       pari                     qjson                    tarsnap-guiawscli                   dspdfviewer              godep                    pdf2htmlex               qscintilla2              tomcat@7certbot                  flow                     knot-resolver            percona-server           qtads                    transcryptchapel                   gammaray                 kobalt                   pngquant                 quazip                   urhcockatrice               gearboy                  liblastfm                pushpin                  qwt                      yarncodequery                gearsystem               logtalk                  qbs                      qwtpolar                 zurlcolor-code               git-cola                 mono                     qca                      qxmppdbus                     gitlab-ci-multi-runner   nano                     qcachegrind              rabbitmqdjview4                  gnupg                    paket                    qjackctl                 riemann==> Renamed Formulaepyqt5 -> pyqt                                                               qt5 -> qt==> Installing dependencies for gcc: gmp, mpfr, libmpc, isl==> Installing gcc dependency: gmp==> Downloading https://homebrew.bintray.com/bottles/gmp-6.1.2.sierra.bottle.1.tar.gz######################################################################## 100.0%==> Pouring gmp-6.1.2.sierra.bottle.1.tar.gz  /usr/local/Cellar/gmp/6.1.2: 18 files, 3.1MB==> Installing gcc dependency: mpfr==> Downloading https://homebrew.bintray.com/bottles/mpfr-3.1.5.sierra.bottle.tar.gz######################################################################## 100.0%==> Pouring mpfr-3.1.5.sierra.bottle.tar.gz  /usr/local/Cellar/mpfr/3.1.5: 25 files, 3.6MB==> Installing gcc dependency: libmpc==> Downloading https://homebrew.bintray.com/bottles/libmpc-1.0.3.sierra.bottle.tar.gz######################################################################## 100.0%==> Pouring libmpc-1.0.3.sierra.bottle.tar.gz  /usr/local/Cellar/libmpc/1.0.3: 11 files, 345.6KB==> Installing gcc dependency: isl==> Downloading https://homebrew.bintray.com/bottles/isl-0.18.sierra.bottle.tar.gz######################################################################## 100.0%==> Pouring isl-0.18.sierra.bottle.tar.gz  /usr/local/Cellar/isl/0.18: 80 files, 3.8MB==> Installing gcc --without-multilib==> Using the sandbox==> Downloading https://ftpmirror.gnu.org/gcc/gcc-6.3.0/gcc-6.3.0.tar.bz2==> Downloading from http://mirrors.ustc.edu.cn/gnu/gcc/gcc-6.3.0/gcc-6.3.0.tar.bz2######################################################################## 100.0%==> Downloading https://raw.githubusercontent.com/Homebrew/formula-patches/e9e0ee09389a54cc4c8fe1c24ebca3cd765ed0ba/gcc/6.1.0-jit.patch######################################################################## 100.0%==> Patching==> Applying 6.1.0-jit.patchpatching file gcc/jit/Make-lang.in==> ../configure --build=x86_64-apple-darwin16.4.0 --prefix=/usr/local/Cellar/gcc/6.3.0_1 --libdir=/usr/local/Cellar/gcc/6.3.0_1/lib/gcc/6 --enable-la==> make bootstrap==> make install  /usr/local/Cellar/gcc/6.3.0_1: 1,358 files, 237.8MB, built in 63 minutes 14 seconds
3、克隆项目代码

$ git clone --recursive https://github.com/Microsoft/LightGBM ; cd LightGBMCloning into 'LightGBM'...remote: Counting objects: 5315, done.remote: Compressing objects: 100% (31/31), done.remote: Total 5315 (delta 5), reused 2 (delta 2), pack-reused 5282Receiving objects: 100% (5315/5315), 4.06 MiB | 21.00 KiB/s, done.Resolving deltas: 100% (3644/3644), done.Submodule 'include/boost/compute' (https://github.com/boostorg/compute) registered for path 'compute'Cloning into '/Users/aven/software/LightGBM/compute'...Submodule path 'compute': checked out '1380a04582080bbe2364352b336270bc4bfa3025'avenMac:LightGBM aven$ pwd/Users/aven/software/LightGBM
4、创建构建目录

mkdir build ; cd build
5、执行 cmake 

$ cmake -DCMAKE_CXX_COMPILER=g++-6 -DCMAKE_C_COMPILER=gcc-6 .. -- The C compiler identification is GNU 6.3.0-- The CXX compiler identification is GNU 6.3.0-- Checking whether C compiler has -isysroot-- Checking whether C compiler has -isysroot - yes-- Checking whether C compiler supports OSX deployment target flag-- Checking whether C compiler supports OSX deployment target flag - yes-- Check for working C compiler: /usr/local/bin/gcc-6-- Check for working C compiler: /usr/local/bin/gcc-6 -- works-- Detecting C compiler ABI info-- Detecting C compiler ABI info - done-- Detecting C compile features-- Detecting C compile features - done-- Checking whether CXX compiler has -isysroot-- Checking whether CXX compiler has -isysroot - yes-- Checking whether CXX compiler supports OSX deployment target flag-- Checking whether CXX compiler supports OSX deployment target flag - yes-- Check for working CXX compiler: /usr/local/bin/g++-6-- Check for working CXX compiler: /usr/local/bin/g++-6 -- works-- Detecting CXX compiler ABI info-- Detecting CXX compiler ABI info - done-- Detecting CXX compile features-- Detecting CXX compile features - done-- Try OpenMP C flag = [-fopenmp]-- Performing Test OpenMP_FLAG_DETECTED-- Performing Test OpenMP_FLAG_DETECTED - Success-- Try OpenMP CXX flag = [-fopenmp]-- Performing Test OpenMP_FLAG_DETECTED-- Performing Test OpenMP_FLAG_DETECTED - Success-- Found OpenMP: -fopenmp  -- Configuring doneCMake Warning (dev):  Policy CMP0042 is not set: MACOSX_RPATH is enabled by default.  Run "cmake  --help-policy CMP0042" for policy details.  Use the cmake_policy command to  set the policy and suppress this warning.  MACOSX_RPATH is not specified for the following targets:   _lightgbmThis warning is for project developers.  Use -Wno-dev to suppress it.-- Generating done-- Build files have been written to: /Users/aven/software/LightGBM/build
6、构建

$ make -j Scanning dependencies of target _lightgbmScanning dependencies of target lightgbm[  2%] Building CXX object CMakeFiles/_lightgbm.dir/src/c_api.cpp.o[  6%] Building CXX object CMakeFiles/_lightgbm.dir/src/application/application.cpp.o[  8%] Building CXX object CMakeFiles/lightgbm.dir/src/main.cpp.o[  8%] Building CXX object CMakeFiles/_lightgbm.dir/src/boosting/boosting.cpp.o[ 10%] Building CXX object CMakeFiles/lightgbm.dir/src/boosting/boosting.cpp.o[ 14%] Building CXX object CMakeFiles/lightgbm.dir/src/application/application.cpp.o[ 16%] Building CXX object CMakeFiles/_lightgbm.dir/src/boosting/gbdt.cpp.o[ 18%] Building CXX object CMakeFiles/lightgbm.dir/src/boosting/gbdt.cpp.o[ 20%] Building CXX object CMakeFiles/lightgbm.dir/src/io/dataset.cpp.o[ 20%] Building CXX object CMakeFiles/lightgbm.dir/src/io/bin.cpp.o[ 22%] Building CXX object CMakeFiles/_lightgbm.dir/src/io/bin.cpp.o[ 24%] Building CXX object CMakeFiles/_lightgbm.dir/src/io/config.cpp.o[ 30%] Building CXX object CMakeFiles/_lightgbm.dir/src/io/dataset.cpp.o[ 30%] Building CXX object CMakeFiles/lightgbm.dir/src/io/config.cpp.o[ 30%] Building CXX object CMakeFiles/lightgbm.dir/src/io/dataset_loader.cpp.o[ 32%] Building CXX object CMakeFiles/_lightgbm.dir/src/io/dataset_loader.cpp.o[ 38%] Building CXX object CMakeFiles/_lightgbm.dir/src/io/metadata.cpp.o[ 40%] Building CXX object CMakeFiles/_lightgbm.dir/src/metric/dcg_calculator.cpp.o[ 40%] Building CXX object CMakeFiles/_lightgbm.dir/src/io/parser.cpp.o[ 40%] Building CXX object CMakeFiles/_lightgbm.dir/src/io/tree.cpp.o[ 44%] Building CXX object CMakeFiles/lightgbm.dir/src/io/tree.cpp.o[ 46%] Building CXX object CMakeFiles/_lightgbm.dir/src/network/linkers_mpi.cpp.o[ 72%] Building CXX object CMakeFiles/lightgbm.dir/src/metric/metric.cpp.o[ 58%] Building CXX object CMakeFiles/_lightgbm.dir/src/network/linkers_socket.cpp.o[ 72%] Building CXX object CMakeFiles/lightgbm.dir/src/network/network.cpp.o[ 62%] Building CXX object CMakeFiles/lightgbm.dir/src/io/parser.cpp.o[ 72%] Building CXX object CMakeFiles/_lightgbm.dir/src/metric/metric.cpp.o[ 72%] Building CXX object CMakeFiles/_lightgbm.dir/src/network/network.cpp.o[ 72%] Building CXX object CMakeFiles/_lightgbm.dir/src/objective/objective_function.cpp.o[ 72%] Building CXX object CMakeFiles/lightgbm.dir/src/network/linkers_mpi.cpp.o[ 72%] Building CXX object CMakeFiles/lightgbm.dir/src/objective/objective_function.cpp.o[ 72%] Building CXX object CMakeFiles/lightgbm.dir/src/metric/dcg_calculator.cpp.o[ 72%] Building CXX object CMakeFiles/_lightgbm.dir/src/network/linker_topo.cpp.o[ 72%] Building CXX object CMakeFiles/lightgbm.dir/src/network/linker_topo.cpp.o[ 72%] Building CXX object CMakeFiles/lightgbm.dir/src/io/metadata.cpp.o[ 72%] Building CXX object CMakeFiles/lightgbm.dir/src/network/linkers_socket.cpp.o[ 74%] Building CXX object CMakeFiles/_lightgbm.dir/src/treelearner/data_parallel_tree_learner.cpp.o[ 76%] Building CXX object CMakeFiles/_lightgbm.dir/src/treelearner/serial_tree_learner.cpp.o[ 78%] Building CXX object CMakeFiles/_lightgbm.dir/src/treelearner/feature_parallel_tree_learner.cpp.o[ 80%] Building CXX object CMakeFiles/_lightgbm.dir/src/treelearner/gpu_tree_learner.cpp.o[ 82%] Building CXX object CMakeFiles/lightgbm.dir/src/treelearner/voting_parallel_tree_learner.cpp.o[ 84%] Building CXX object CMakeFiles/_lightgbm.dir/src/treelearner/tree_learner.cpp.o[ 86%] Building CXX object CMakeFiles/lightgbm.dir/src/treelearner/serial_tree_learner.cpp.o[ 88%] Building CXX object CMakeFiles/lightgbm.dir/src/treelearner/tree_learner.cpp.o[ 94%] Building CXX object CMakeFiles/lightgbm.dir/src/treelearner/gpu_tree_learner.cpp.o[ 94%] Building CXX object CMakeFiles/lightgbm.dir/src/treelearner/data_parallel_tree_learner.cpp.o[ 94%] Building CXX object CMakeFiles/lightgbm.dir/src/treelearner/feature_parallel_tree_learner.cpp.o[ 96%] Building CXX object CMakeFiles/_lightgbm.dir/src/treelearner/voting_parallel_tree_learner.cpp.o[100%] Linking CXX shared library ../lib_lightgbm.so[100%] Linking CXX executable ../lightgbm[100%] Built target lightgbm[100%] Built target _lightgbm
执行到 96% 的时候会卡一会儿,如果不希望执行“make -j”命令时对电脑影响太大的话,执行“make -j1”,使用CPU的一核,不会占满CPU。

构建成功后会在上层目录生成两个文件,lib_lightgbm.so  lightgbm,一个库文件,一个可执行文件。


相关阅读:

微软开源分布式高性能GB框架LightGBM安装使用

微软开源分布式高性能GB框架LightGBM安装使用——Python

====================文档信息=======================

版权声明:非商用自由转载-保持署名-注明出处

署名(BY) :testcs_dn(微wx笑)

文章出处:[无知人生,记录点滴](http://blog.csdn.NET/testcs_dn)

==============欢迎关注我的个人微信订阅号(微wx笑)==========

最新文章

123

最新摄影

闪念基因

微信扫一扫

第七城市微信公众平台