New analysis for VisionEval input files using NTD data #260
Description
This will be an analysis in the analysis tab.
The output will be two csv files meant to be used for running VisionEval. VisionEval is a software for modeling outcomes of policy and investment changes in large areas. Dan Cotey, a WSDOT colleague, is the main contact for WSDOT’s use of this software and previously worked with Public Transportation Division’s data team to gather the necessary input data for it. Some of the input data can be pulled from the National Transit Database (NTD).
Pulling the Data
NTD data is available online and the specific table that contains the data we need is the Annual Data – Metrics table. The table can be downloaded as a csv file, or accessed via an API endpoint (documentation here).
Columns needed in this analysis:
- Agency
- State
- NTD ID
- Report Year
- Primary UZA Name
- Mode
- Total Operating Expenses
- Vehicle Revenue Miles
By default, the analysis will pull the most recent year of data and rows where State = "WA" and Primary UZA Name doesn't contain "Non-UZA". The user should be able to change these selections if they want, but only one year and State will be chosen at a time. Data for the previous calendar year is made available in mid-November; for example, data for Report Year 2024 was available in November 2025.
Input parameters that the user will select are year and state.
Output
There will be two .csv files as output. Words in [brackets] reference columns from the NTD data.
Use the table below when matching NTD [Mode]s to VisionEval modes for the column specifications.
| VisionEval Mode | Applicable NTD Mode(s) |
|---|---|
| DR | DR |
| VP | VP |
| MB | MB |
| RB | CB |
| MG | MG |
| SR | SR, TB |
| HR | HR, LR |
| CR | CR |
Note: Any NTD [Mode]s not listed in the table above are currently not included in VisionEval modeling and can be ignored.
marea_transit_service.csv
The first output file should match the format specified in VisionEval’s user manual for the marea_transit_service.csv file. As the “Geo” column below suggests, [Vehicle Revenue Miles] will be grouped and summed based on [Report Year] and [Primary UZA Name]. In other words, there will be one row for each [Report Year] and [Primary UZA Name] combination.
Column specification:
- Geo: [Primary UZA Name]
- Year: [Report Year]
- DRRevMi: Sum of [Vehicle Revenue Miles] for [Mode]s specified in table above
- VPRevMi: Sum of [Vehicle Revenue Miles] for [Mode]s specified in table above
- MBRevMi: Sum of [Vehicle Revenue Miles] for [Mode]s specified in table above
- RBRevMi: Sum of [Vehicle Revenue Miles] for [Mode]s specified in table above
- MGRevMi: Sum of [Vehicle Revenue Miles] for [Mode]s specified in table above
- SRRevMi: Sum of [Vehicle Revenue Miles] for [Mode]s specified in table above
- HRRevMi: Sum of [Vehicle Revenue Miles] for [Mode]s specified in table above
- CRRevMi: Sum of [Vehicle Revenue Miles] for [Mode]s specified in table above
cost_per_revenue_mile.csv
The second output file is not a direct input file for VisionEval, but will be used by Dan for related purposes. There will be one row for each [Report Year] and Mode combination.
Column specification:
- YearOfDollars: [Report Year]
- Mode: Applicable [Mode]s as specified in table above
- CostPerRevenueMile: [Total Operating Expenses] divided by [Vehicle Revenue Miles]