Files and Directories

Results

The results from the initial major df-analyze run using only the LightGBM model are located in results/traffic_results_subset.

Downloading Data

Data should be downloaded from the Montgomery Count Website and saved in the project root as traffic_violations_complete.csv (see clean_data.py).

Code Files

Python Scripts

  • clean_data.py: performs cleaning of the raw data, and places outputs in traffic_data
  • summary.py: summarizes the results of the full job (see "Job Scripts" below) into various tables (some printed to stdout, some saved)

Job Scripts

Contains the actual scripts for running on Compute Canada. Will require setting up a container to run df-analyze.