Research Request - Transit Ridership Dashboard GTFS Refactor

[KatrinaMKaiser]

Research Question

Single sentence description: What will it take to migrate the transit ridership dashboard created Fall 2022 to warehouse v2?

Detailed description: Transit ridership dashboard was developed with GTFS routes and stop_times and other data using an older data model. Need to move away from looping over operators identified via calitp_id and use services/feeds for identification. This is the first of ~3-4 issues re: updating the dashboard.

How will this research be used?

This is the first step to update the dashboard. Next steps will require getting fresh training data (stop level ridership). Additional features, model additions come after that.

Stakeholders & End-Users

The main stakeholder is CARB, who uses this tool to support applicants for the AHSC grant which requires estimated ridership changes. There is potential to use this tool for other transit-related grants that involve the CARB GHG calculator.

Metrics

• To validate, check how many additional services/operators get added/dropped between the old and new versions of the underlying table

Data sources

• Cal-ITP data sources:

  • GTFS Schedule
  • NTD monthly ridership (currently using ad-hoc import into GCS but switch to the version in the warehouse in a future issue)
  • NTD annual ridership (not incorporated in current version, incorporate into new version in same future issue)

• External data sources:

  • stop-level ridership from LA Metro/SBMTD/MTS (old)
  • ACS data (old from public BQ, update in a future issue)

• Remaining data source questions:

Deliverables

new underlying data and structure for the dashboard

Timeline of deliverables

Estimated completion date: Targeting June 30. Start ahead of official #1122 in early June.

[shweta487]

Research Task - Update ACS data on Transit Ridership Dashboard (#1128)

[shweta487]

Research Task - Integrating NTD Data into Transit Ridership Dashboard(#1130)

[shweta487]

Ridership Dashboard Refactoring Updates as of 08/27/2024

• The initial integration of warehouse v2 data into the Transit Ridership Dashboard has been successfully completed. The create_stop_freq_refactor.py script now processes the updated data, generating a parquet file that includes detailed trip and route information for the dashboard's ridership analysis.
• Finished initial cleaning and integration of stop-level data from SBMTD, LA Metro, and Monterey Salinas Transit with ridership data. Used fuzzy match technique in process_sbmtd_refactor.ipynb to solve the join issue with the new stop_id. Additional steps needed, particularly considering stop_code match with stop_id. Also considering using this technique on process_mst_refactor.ipynb.
• join_analytical_file_refactor.ipynb : Incomplete as of now.

[shweta487]

#1246 (comment)