Racial Disparities in the Enforcement of Marijuana Violations in the US
Installation
Data
To download the existing datasets, reports, and maps, you must install DVC:
pip install dvc[gs]
Followed by:
dvc pull
This should automatically download all the required datasets.
Python
To install the required python libraries, please install a fresh python environment (using anaconda or such like), then run the command:
pip install -r requirements.txt
pointing to the requirements.txt file in the NIBRS directory.
Reproducing the Results
Enforcement Ratios
In order to produce the enforcement ratios you must either download the pre-requisite datasets from DVC, using dvc pull, or you may produce all results yourself using the following steps:
-
Download the appropriate raw NIBRS files for each state. This can be achieved by running the python script:
python data_downloading/download_and_extract.py. This currently defaults to downloading years 2010-2019 and over all states. This can be changed from within the script. -
Produce the appropriate NIBRS dataset. Using the query_nibrs.py python script
python scripts/python/data_processing/query_nibrs.py. The script has a number of arguments that can be explored with:python query_nibrs.py -h. -
Produce the enforcement ratios. Run the selection_ratio.py script
scripts/python/data_processing/selection_bias.pywith appropriate arguments. Runscripts/python/data_processing/selection_bias.py -hfor help, or consult the image below:
Paper Figures
There are many scripts used to create the figures and tables in the paper, please consult the table below to find which script corresponds to which figure:
Figures
| Figure | Script |
|---|---|
| 1 | R/generate_plots_4paper.R |
| 2 | python/enforcement_ratio_model_plots.py |
| 3 | R/generate_plots_4paper.R |
| S1 | python/nsduh_usage_plot.py |
| S2 | R/generate_plots_4paper.R |
| S3 | python/enforcement_ratio_location_plot.py |
| S4 | python/legalized_states_agency_reporting_plot.py |
| S5 | python/enforcement_rate_by_demographic.py |
| S6 | python/enforcement_rate_by_demographic.py |
| S7 | R/generate_plots_4paper.R |
| S8 | python/enforcement_ratio_model_plots.py |
| S9 | R/generate_plots_4paper.R |
| S10 | python/enforcement_ratio_model_plots.py |
| S11 | R/generate_plots_4paper.R |
| S12 | python/enforcement_ratio_model_plots.py |
| S13 | R/generate_plots_4paper.R |
| S14 | python/enforcement_ratio_model_plots.py |
| S15 | R/generate_plots_4paper.R |
| S16 | Doesn't exist in paper? Bug in latex maybe? |
| S17 | python/enforcement_ratio_model_plots.py |
| S18 | R/generate_plots_4paper.R |
| S19 | python/time_distribution_plot.py + python/enforcement_ratio_model_plots.py |
| S20 | R/generate_plots_4paper.R |
Tables
| Table | Script |
|---|---|
| 1 | N/A |
| 2 | python/enforcement_ratio_n_counties_table.py |
| 3 | R/process_usage_data_nsduh.R |
| 4 | R/get_stats_on_nibrs.R |
| 5 | python/opportunity_atlas_regression_table.py + python/time_regression_table.py |
| S1 | R/process_usage_data_nsduh.R |
| S2 | R/get_stats_on_nibrs.R |
| S3 | R/get_stats_on_nibrs.R |
| S4 | R/get_stats_on_nibrs.R |
| S5 | python/opportunity_atlas_regression_table.py + python/time_regression_table.py |