Analyze the data generated by the pbf-scraping project for the Philly Bail Fund

This repository contains code and auxilliary data for analyzing data associated with bail hearings in Philadelphia, in partnership with the Philadelphia Bail Fund. The goal of this project is to provide PBF with flexible tools for analyzing their efforts and the general bail system in Philadelphia, and generate findings and visualizations to inform the general public and potentially influence decision-makers.

The current version of the dashboard is on Streamlit. For further analyses, view the analysis notebook(s) on nbviewer.

List of questions for analysis can be found in questions.md

If you are joining this project (and have joined the Code for Philly Slack workspace) and need access to the data, contact @irishryoon or @notchia for details.

• Fork this repository and clone your fork to work locally; ideally, work on branch off master within your own repository.
• Regularly update your master branch by merging with the upstream repo (this one) to help avoid merge conflicts.
• When you are ready to merge your changes, create a pull request. If there are merge conflicts, you can still make a pull request.
• If there are conflicts, check the reviewnb link that is automatically generated as a comment to your pull request. This allows you to see diffs in notebooks more clearly than the Github diff viewer, and makes it easier to figure out how you can update your PR to resolve the conflicts.

To run the Streamlit dashboard app, navigate to the dashboard folder and run the command streamlit run app.py. Note: dashboard.ipynb can be run all the way through from the top, but you may need up to ~16 GB of RAM. You can also run section 0 and then whichever subsection you are working on.

• Target audience: the general public. Could your grandmother understand this data and these visualizations?
• Contents
  1. Aggregate bail information: How many cases were seen? How often is monetary bail set?
  2. Visualizations on magistrate information: Is there a difference in bail type and bail amount set by different magistrates?
  3. Which neighborhoods are heavily impacted by monetary bail?
  4. Breakdown by race and gender: Is there a difference in bail type and bail amount set for defendants of different races?
  5. Snappy visualization of how much Philadelphians paid in bail (inspired by the US debt clock, but more doably, by countUp.js)
• To do & timeline
  • By December 14, 2020
    • Create content & preliminary visualizations for dashboard
    • Generate ideas for final dashboard content/visualization
    • Group check-in the week of December 14
    • Post 'help-wanted': someone who can convert python visualizations into a dashboard
  • By Jan 11, 2020
    • Finalize dashboard visualizations
    • Port over visualizations to web-viewable fromat

• Ideally by March or April 2021
• List of questions can be found in questions.md

• Open Data Philly
• Data.world
• How Bail Works, Billy Penn
• Effects of ROR policies