Integration with recommender packages about goodbooks-10k HOT 7 CLOSED

Closing upon successful completion (maciejkula/spotlight#62).

from goodbooks-10k.

Sure thing! What do you need from me?

from goodbooks-10k.

I would write all the code on my end, but I was wondering if you'd be up to do some more data packaging.

For bandwidth and speed of loading, I use h5py files for my datasets. Would you be willing to re-package the data in that format, and maybe include them as Github release files? (Something along the lines of the downloads here).

from goodbooks-10k.

That's a good idea, I will make a release. Please send me some example code for packing CSV into HDF5.

from goodbooks-10k.

Awesome!

I have some code here (warning, not end-user code, might be fairly horrible). It basically boils down to stuffing numpy arrays into the file with the appropriate names.

from goodbooks-10k.

I made a release: https://github.com/zygmuntz/goodbooks-10k/releases/tag/v1.0

As regards HDF5, I created a file with integer-only data: ratings, to_read and book_tags. Books and tags contain numbers mixed with strings so I skipped them.

from goodbooks-10k.

Great, I'll have a look soon and let you know how I get on.

from goodbooks-10k.