fetch_mldata Deprecated and Down Routinely

[ZaydH]

I was looking at your nnPU chainer implementation. If I am not mistaken, it seems to download data from mldata.org. This sklearn issue report discusses the instability of mldata.org and how it often goes down for extended periods of time. For example, mldata.org is down at the moment.

When you run the fetch_mldata command, you get the warning in sklearn:

DeprecationWarning: Function fetch_mldata is deprecated; fetch_mldata was deprecated in version 0.20 and will be removed in version 0.22

What are your thoughts on moving away from fetch_mldata to an alternative? Different alternatives include:

• Hosting the file on Github (probably in a different repository)
• Moving to another source for the data, e.g., torchvision

I would be happy to work with you on this if you are interested.

[ZaydH]

@kiyoshiiriemon makes a good proposed changed in nnPU's repository. The change he proposes in his issue report seems to be very clean and straightforward.

[t-sakai-kure]

Thank you for your comment, @ZaydH.
The function fetch_openml looks nice as long as people use sklearn v0.21.2>.
I will update the code in a few days.

[ZaydH]

@t-sakai-kure Tomoya you're right.

It may be worth adding a requirements.txt to the package to make it more portable and robust going forward since dependencies may obsolete commands.

[t-sakai-kure]

@ZaydH Zayd, thanks.
I changed the code based on this comment scikit-learn/scikit-learn#12504 (comment).