Minor feedback - Githubissues

ksedivyhaley commented 4 years ago

EDA

Warning: best practice to split test data as early as possible - before viewing df.tail().
"So we might need to drop some columns when fitting the model or choose the model that won't be affected by the collinearity of predictors." - good observation! I'd put it in the conclusions (your conclusion states more vaguely that you'll "consider" it but it's worth being specific).
Do you know why there are so many 0s in the repayment status?

Other

Proposal should also suggest likely tables/figures for end report
README in data folder with data citation is nice! It's good to have the citation in the main README as well, and you should cite also the repository.

ksedivyhaley commented 4 years ago

Clarification: I think I mentioned this in class, but the only one of these items that I'd consider a priority to fix is the data citation. If you address the others, great, but focus on the download script #22 and making sure you clearly communicate your analysis in the final report (keeping in mind the questions in #20 ).

zoepan00 commented 4 years ago

Thank you Kate!

UBC-MDS / DSCI_522_group-314

Minor feedback #23