Save models from xgb.cv

[vykhand]

#793 Please note that this change have not been tested. I could not install xgboost from the most recent repository due to some compilation errors - missing logging.h.

[pommedeterresautee]

try this to update your repository including submodule, with the following commands:

git submodule init
git submodule update

BTW, great idea

[vykhand]

Hi! Thanks for kind words. I can't figure out what is wrong. I did
git submodule init git submodule update

before the pull request but after commit. and then again after I got the appveyor error message/.

How do I fix it now?

[pommedeterresautee]

Do you compile on your computer now ?
Have you seen that:
https://github.com/dmlc/xgboost/blob/5d9b80cd8bef05e6f3fd7d4b7624d9a17c255978/doc/build.md#trouble-shooting

[pommedeterresautee]

regarding AppVeyor, I don't think it is related to you.

[vykhand]

Ah, sorry I thought you were talking about AppVeyor issue, the compilation issue is different - I am getting the exact same message as

#797

I don't think it could be related to submodules. I am getting it when installing from github from my repo and as well from dmlc.

[vykhand]

Hi Guys,

I am fairly new to the process of contributing to XGB, or to any other SW to be honest.
I fail to re-install xgb from my own git to test my change, due to some problems with compilation.
Is there any other way to test?
Otherwise, may be someone would be so kind and test my change?

With respect, Andrey

[tqchen]

closing this for now, see #892 for new guideline in enhancing training API