There was a similar effort here: https://groups.google.com/forum/#!searchin/keras-users/datasets/keras-users/n6jE9eFcaYI/Roo-rWK6CQAJ - where someone wrote a small library to abstract the loading of open source datasets. While having extra dependencies is something that should probably be avoided, I don't think it makes a lot of sense to build a lot of code into scikit-learn to load and fetch datasets (except the truly common ones like newsgroups/iris, etc).
I think with the proliferation of new deep learning libraries that are cropping up, it would be good to agree to a common format to load/store different datasets without duplicating the effort in many places. On Wed, 9 Dec 2015 at 19:58 Andreas Mueller <t3k...@gmail.com> wrote: > > > On 12/09/2015 01:48 PM, Gael Varoquaux wrote: > > On Wed, Dec 09, 2015 at 12:33:55PM -0500, Andreas Mueller wrote: > >> I guess we use the matlab data with is not required by mldata. > >> We could add code that tries to fetch the matlab, and if that doesn't > >> work uses the hdf5, > > I'd rather not. I'd rather we just have a good error message. > > > Never mind, I don't think this was the problem here anyhow. > > > ------------------------------------------------------------------------------ > _______________________________________________ > Scikit-learn-general mailing list > Scikit-learn-general@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general >
------------------------------------------------------------------------------
_______________________________________________ Scikit-learn-general mailing list Scikit-learn-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
