1. Recently intend to use Xgboost to do Titanic exercises, toss for a while to do well. Below, the detailed process of their own problem-solving is recorded for reference.
First tried the next anaconda from the Anaconda prompt to install xgboost, but unfortunately so far, this method has not been able to implement the Xgboost installation.
Then, on the Internet to find a lot of information about the installation of xgboost under Python, most of the first to provide the GitHub link, first create a C + + version of Xgboost, and then converted to the Python version. However, GitHub does not now support downloading xgboost-windows files and downloading them from other locations. Links are as follows:
Https://pan.baidu.com/s/1cDzufs
2. After downloading good xgboost files, You need to open the Xgboost.sln project in the Windows directory in the Xgboost folder (preferably with the new VS, the old version vs., I use the VS2015, when I open sln will be prompted whether to upgrade the project file, select Yes)
At this point you can see that there are four main sections under the Xgboost project, as shown in the following illustration:
Right-click the project name (xgboost), select Configuration Manager, change the solution configuration from debug to release in the upper-left corner of the configuration Manager, and choose Win32 or Win64 depending on your situation.
Then right-click to regenerate the solution, as shown in the following illustration: When you see all the Xgboost, the success of the 4-word, the C + + version of the installation is successful.
3. The next step is to generate xgboost for the Python version.
(1) If it is directly under the Python idle installation, then open cmd, into the Xgboost python-package directory, such as: U:\Python\Package\xgboost-master\python-package ,
Then enter the Python setup.py install, carriage return, and if Xgboost needs a dependency pack (such as scipy, etc.), the system will automatically install the Python-package package in the Xgboost package.
(2) In addition, can also be installed under Anaconda. First Open Anaconda Prompt, also switch to the Xgboost python-package directory, enter the Python setup.py install, the installation process as shown in the following figure:
If the words finished processing dependencies for xgboost = 0.4 are displayed at the bottom, the Xgboost has been successfully installed. If there are problems during the installation process, most of the problems are missing xgboost dependencies, you can download it by yourself, or you can download it by anaconda the Conda way.
4. Finally, determine if the Xgboost was successfully installed.