nyctrains API

A FastAPI-based backend and Python package for working with the MTA's real-time subway and LIRR GTFS-RT data feeds. This project fetches, parses, and exposes real-time feeds as human-readable JSON, including stop names and (for LIRR) route names. You can use it as an HTTP API or as a Python library in your own projects.

Supported Feeds

• ace (A, C, E)
• bdfm (B, D, F, M)
• g (G)
• jz (J, Z)
• nqrw (N, Q, R, W)
• l (L)
• si (Staten Island Railway)
• 1234567 (1, 2, 3, 4, 5, 6, 7, S)
• lirr (Long Island Rail Road)

Installation

Install using pip:

pip install nyctrains

Static GTFS Data

This package relies on static GTFS data files being present in the resources/ directory for full functionality (e.g., adding stop names, route details).

Required files:

• resources/stops.txt (NYC Subway + LIRR stops)
• resources/routes.txt (NYC Subway + LIRR routes)
• resources/trips.txt
• resources/stop_times.txt

Note: The simple mappings for LIRR stop names (stops-lirr.txt) and route names (routes-lirr.txt) used in earlier versions have been replaced by the requirement for the full static GTFS files. You can typically download these files from the MTA developer resources page.

The application loads these files into pandas DataFrames at startup using the nyctrains.static_gtfs module.

Usage

This package provides Python tools and a FastAPI backend for working with MTA GTFS-RT subway and LIRR data. No API key is required to use the package or access the feeds.

Running the API Server

# Make sure you have the static GTFS files in ./resources/
uvicorn nyctrains.main:app --reload

Access the API documentation at http://127.0.0.1:8000/docs.

Make requests to the feed endpoint, e.g.: http://127.0.0.1:8000/subway/ace/json

Using as a Library

Example: Fetching a GTFS Feed

from nyctrains.mta_client import MTAClient
import asyncio

client = MTAClient()
feed_path = "nyct%2Fgtfs-ace"  # Example feed

data = asyncio.run(client.get_gtfs_feed(feed_path))
# data contains the raw protobuf bytes
print(f"Feed data length: {len(data)} bytes")

Example: Loading Static GTFS Data

from nyctrains import static_gtfs

# Load static data (caches results)
stops_df = static_gtfs.get_stops()
routes_df = static_gtfs.get_routes()

if stops_df is not None:
    print(f"Loaded {len(stops_df)} stops.")
if routes_df is not None:
    print(f"Loaded {len(routes_df)} routes.")

Example Output

{
  "header": {
    "gtfs_realtime_version": "2.0",
    "timestamp": "2025-04-15T21:04:02+00:00"
  },
  "entity": [
    {
      "id": "GO304_25_809_T",
      "trip_update": {
        "trip": {
          "trip_id": "GO304_25_809",
          "start_date": "20250415",
          "schedule_relationship": 0,
          "route_id": "6",
          "route_long_name": "Long Beach Branch",
          "direction_id": 1
        },
        "stop_time_update": [
          {
            "stop_id": "LBG",
            "stop_name": "Long Beach"
          }
        ]
      }
    }
  ]
}