Adding tests wherever code coverage dips

[shreyapandit]

Added test for running initialize_city, and one for running pipeline before viz. Now, based on codecov feedback, I am working on testing the standardize_concerns and standardize_crashes scripts under data_standardization.

[terryf82]

That's great @shreyapandit thanks for taking the lead on a not-so-glamorous job!

I wrote the original versions of both standardize_concerns and standardize_crashes. Based on feedback from Jenny I then rewrote sections of standardize_crashes to remove city-specific switching, but haven't similarly reviewed concerns. Also I've completed an online python course since then, so my coding abilities have probably improved quite a bit!

If you have any concerns about them let me know, there is probably quite a bit of scope for improvement.

Thanks.

[j-t-t]

Shreya, great, thank you! Can I suggest that you do a PR for what you've already done? That way if people make changes to files you're testing, there won't be as much chance for conflicts.

Speaking of which, regarding testing for standardize_crashes/concerns - on my (badly named) branch more_analysis, I've added another standardization script, and as a result, refactored the directory a little to add a util file, and added tests for those new functions. So I'm going to try to get that finished today and put in a PR so if you're writing tests on that, we won't conflict.

[shreyapandit]

Thanks @j-t-t and @terryf82 ! I will look into refactoring the concerns script as well, and will make a pull request soon

[shreyapandit]

The pull request can be found here: https://github.com/Data4Democracy/crash-model/pull/152

[terryf82]

Hey Shreya do you need a hand with this? It looks like the PR has some failing tests, though I'm not sure if these are real or just timeouts..

[bpben]

Codecoverage is available!