pltrees / abcboost Goto Github PK

If I have trained a model and saved it, how can I view the ranking of parameters and feature importance of the model

#5 (comment)

您的python/windows/目录下有abcboost.pyd文件，我直接成功下载下来在相同目录下建立了一个test.py，再import abcboost后，发现找不到abcboost指定的模块，请问一下这是什么原因造成的的呢？

G:\abcboost-main\abcboost-main\build>cmake ..
-- Selecting Windows SDK version to target Windows 10.0.22000.
CMake Error at CMakeLists.txt:2 (project):
Failed to run MSBuild command:

C:/Program Files (x86)/MSBuild/14.0/bin/MSBuild.exe

to get the value of VCTargetsPath:

Microsoft (R) 生成引擎版本 14.0.23107.0
版权所有(C) Microsoft Corporation。保留所有权利。

生成启动时间为 2022/8/9 14:40:16。
节点 1 上的项目“G:\abcboost-main\abcboost-main\build\CMakeFiles\3.24.0\VCTargetsPath.vcxproj”(默认目标)。
InitializeBuildStatus:
  正在创建“Debug\VCTargetsPath.tlog\unsuccessfulbuild”，因为已指定“AlwaysCreate”。
PostBuildEvent:
  echo VCTargetsPath=
  :VCEnd
  VCTargetsPath=
FinalizeBuildStatus:
  正在删除文件“Debug\VCTargetsPath.tlog\unsuccessfulbuild”。
  正在对“Debug\VCTargetsPath.tlog\VCTargetsPath.lastbuildstate”执行 Touch 任务。
已完成生成项目“G:\abcboost-main\abcboost-main\build\CMakeFiles\3.24.0\VCTargetsPath.vcxproj”(默认目标)的操作。

已成功生成。
    0 个警告
    0 个错误

已用时间 00:00:00.26

-- Configuring incomplete, errors occurred!
See also "G:/abcboost-main/abcboost-main/build/CMakeFiles/CMakeOutput.log".

Only the accuracy rate cannot impress industrial application users.

My data is about 300k*1000, params: max_bin=255, max leaves=255, learning_rate=0.1, n_threads=16.
Compare to LGB, the training speed is about 1/3~1/4, memory taken is about 5x data original size (5x than numpy format).
We could improve the engineering implementation.