insight-lane / crash-model

Build a crash prediction modeling application that leverages multiple data sources to generate a set of dynamic predictions we can use to identify potential trouble spots and direct timely safety interventions.
https://insightlane.org
MIT License
113 stars 40 forks source link

Waze #213

Closed j-t-t closed 5 years ago

j-t-t commented 5 years ago

Initial pass of doing something with waze data, just to get something out there.

Adds a standardization step and generation step to incorporate waze data

One other thing that this pull request contains is writing the features that we're using for this city to file. I did this because we were hard coding which features were being used in several places, and it was complicated enough to add new features that there was a high risk that it would be done wrong. Now features are generated by make_dataset's make_feature_list, drawing from a hardcoded list and from the city's config file

codecov[bot] commented 5 years ago

Codecov Report

Merging #213 into master will increase coverage by 1.91%. The diff coverage is 64.01%.

@@            Coverage Diff             @@
##           master     #213      +/-   ##
==========================================
+ Coverage   52.21%   54.12%   +1.91%     
==========================================
  Files          29       32       +3     
  Lines        2959     3161     +202     
==========================================
+ Hits         1545     1711     +166     
- Misses       1414     1450      +36
terryf82 commented 5 years ago

It's probably not worth holding this up, but is there any value in us defining waze-schema and waze-instance files as we have for other data sources?

j-t-t commented 5 years ago

Yes, I think we should, but since I didn't feel like I'd finalized the schema yet, I thought it was worth waiting to do that.