zygmuntz / goodbooks-10k Goto Github PK

I maintain a bunch of recommender packages, and this dataset could be super useful for my users!

Would you be willing to integrate this into LightFM or Spotlight? (As in, make it easily downloadable and so on.)

Please help I get error:

current_user <- "7329"
str(ratingmat)
rated_items <- which(!is.na((as.data.frame(ratingmat[current_user, ]))))
selected_users <- names(which(apply(!is.na(ratingmat[ ,rated_items]), 1, sum) >= 2))
head(selected_users, 40)

Error in ratingmat[current_user, ] : subscript out of bounds

Is the books_xml.tar.gz corrupt - I get a data stream error with gzip
Tried redownloading same error :-(
Thanks
David.

Could you publish the scripts you ran to obtain the dataset, so we can reproduce an up-to-date version?

the data seems to be backdated. There are more update available. The data should be regenerated.
Please Update the dataset.

9.78043902348e+12 is not an ISBN :)

I'd like to cite this dataset in my paper. Do you have a preferred citation style? This is what I've come up with:

@article{goodbooks2017,
  author = {Zajac, Zygmunt},
  title = {Goodbooks-10K dataset},
  year = {2017},
  publisher = {GitHub},
  journal = {GitHub},
  howpublished = {\url{https://github.com/zygmuntz/goodbooks-10k}},
}

Is there any way to get the rating/review timestamps?
(I know the data is ordered by time/row index, but I want explicit times, for each user/rating).

• Awesome dataset! I'm having a lot of fun with it!

try to search "0679781587" in BX-BOOK-Rating.csv and you can find rating record, but "0679781587" doesn't appear in BX-BOOK.csv