Peer Review Feedback

[srfrew]

https://github.com/UBC-MDS/data-analysis-review-2023/issues/21

[srfrew]

Documentation

• [x] Community guidelines: Add mention in README about third parties wishing to 1) Contribute to the software 2) Report issues or problems with the software 3) Seek support. Refer to contributing.md

[srfrew]

Reviewer Recommendations

• [x] Language: simplify results and discussion and remove hyperparameter tuning table (+2)
• [x] move scripts from src to scripts to keep things less confusing (+3)
• [x] subdirectories for figures, tables, etc. in results folder (+2)
• [x] Clean up branches etc within the repository
• [x] Docker instructions didn't recommend navigating to root of directory, and launching docker
• [x] Include instructions for local dependencies
• [x] Rational for selecting SVC over Decision Tree

[srfrew]

My opinions on which to implement:

Reviewer Recommendations

• Language: simplify results and discussion and summarize CV results table in report (+2). I think this could be simply done, but would prefer to summarize some numerics from the CV results at a minimum to see how things converged
• move scripts from src to scripts to keep things less confusing (+3). I'd rather not do this as I find things more confusing, but in bigger projects this may be preferred
• subdirectories for figures, tables, etc. in results folder (+2) for a small project like this one this may add more complexity (e.g., folders with one file in them)
• Clean up branches etc within the repository We should definitely do this
• Docker instructions didn't recommend navigating to root of directory, and launching docker Definitely implement root navigation step
• Potentially include instructions for local dependencies this was removed in a previous release, can include in local dependencies
• Rational for selecting SVC over Decision Tree add a sentence explaining the decision to go for a lower-overfitting model

[meretelutz]

Assignments for this week:

• [ ] Change log - Jake
• [x] Makefile and Instructions on README- Waleed
• [x] Writing/Language in the report - Merete
• [x] File structure - Simon

[srfrew]

I've updated the results folder structure, moved code to a scripts folder, adjusted in-script references, and updated the makefile... This might cause significant merge conflicts when brought in, so idk where everything else is at

[meretelutz]

I've finished my work on the report's results and discussion. I added the rationale of why we chose SVM over Decision Tree, got rid of the hyperparameter optimization table, clarified the discussion around results/discussion/further improvements. I also deleted all my old branches! I think that's everything assigned to me, but let me know if there are any other issues!

51 #52

[WaleedMahmood1]

I've finished my part too, in updating the README.md file. I believe this should not have any conflicts with main. If someone could review and accept the pull request, that would be great!