Research Request - Roll-up single day speeds to weekday vs weekend + peak vs offpeak average speeds

[tiffanychu90]

Complete the below when receiving a research request, and continue to add to this issue as you receive additional details and produce deliverables. Be sure to also add the appropriate project-level label to this issue (eg gtfs-rt, DLA).

Research Question

Single sentence description: The speeds dataset that is saved out in the public GCS bucket needs to be aggregated up a bit more. Use April 2023 dates since we have 4 dates downloaded....we can add 3 more dates for a full week, and roughly sketch out how speeds by stop segments would be reported if use shape_id but report out 4 rows for each segment:

• weekday-peak, weekday-offpeak
• weekend-peak weekend-offpeak
• some kind of more aggregated roll-up Research Request - Cache each month's single day speeds data into public GCS bucket  #923

Detailed description: Since shape_id seems to be changing for some operators over the monthly time horizon, let's leave the shape variation for another time and still use shape_id, but use a tighter time horizon over which we can do some roll-ups. Within a week, shape_id is not likely to change, so let's aggregate to this way and leave some granularity before we decide.

In the long term, we do have to move to route-direction, but as soon as we do, we risk the same shape_id-stop_sequence combination not referring to the same physical segment. This script also finds the longest shape for a route-direction.

References:

• https://github.com/kotrc/GTFS-route-shapes/tree/master

How will this research be used?

We want to save some version of this roll-up to the public GCS bucket

Metrics

• Same p20_mph, p50_mph, p80_mph columns, but sliced differently

Data sources

• Cal-ITP data sources: speeds_stop_segments_{analysis_date} for April dates

Deliverables

Adapt current scripts to concatenate single-day segment speeds before averaging