All geek questions in one place

Cancel use with xgboost

I installed xgboost using whl from http://www.lfd.uci.edu/~gohlke/pythonlibs

When I tried:

 import xgboost

And I have the following message:

d: \ program Files \ python \ Lib \ site packages \ sklearn \ cross_validation.py: 44: Deprecated warning: this module is deprecated in version 0.18 in favor of the model_selection module, into which all refactored classes and functions are moved. Also note that the interface of the new CV iterators is different from the new modules. This module will be removed at 0.20. "This module will be removed at 0.20", DeprecationWarning)

If I print import sklearn before import xgboost , I do not get the message. I believe that this message will not affect the results, but how to avoid it? I also checked through pip if all the packages are updated.

+5
source share
2 answers

Firstly:

Yes, it does not affect the results at the moment.

Next, how to avoid this?

Well, the answer is not easy. Software is developing. Dependencies are inevitable. Resistance is futile. Package and configuration management policies are the only way to deal with it.

What is best suited for such needs?

  • Isolate your experiments, allocating VM is enough to rely on

  • Continue to use reliable package management - Travis OLIPHANT Anaconda is the way to go (+ 3)

  • Provide configuration management to prevent bleeding after the “new” packets fall under the sun. Anaconda allows you to “freeze” a controlled [environment] , where you determine which version / release number of the corresponding packages is stored (and Anaconda maintains cross-dependencies for things under control, so we can simply benefit from a clear and well-identified [environment] -s, in which the code used to run and continue to use)

The naive advice of always updating the “latest version” can simply ruin the currently working toys and smash things into chaos. It is better to explicitly define / configure / identify / apply fully controlled [environment] -s for work and keep it for many years.

If someone is in accordance with ISO / EN-9000 +, NATO-STANAG AQAP-130 +, etc., there are not much better ways to keep going.

0
source

I had the same problem. I used the following commands instead of pip install according to the xgboost installation guide, and the warning disappeared:

 git clone --recursive https://github.com/dmlc/xgboost cd xgboost make -j4 cd python-package python setup.py install

Be sure to remove the previously installed xgboost:

 pip uninstall xgboost
0
source

Source: https://habr.com/ru/post/1261514/

More articles:


All Articles