How to extract a small area from a big GTFS feed?

How to Carve Out Your Own Little Transit World from a Giant GTFS Feed

So, you’re diving into the world of public transit data, huh? You’ve probably stumbled upon GTFS – the General Transit Feed Specification. Think of it as the universal language that transit agencies use to share their schedules, routes, and all that good stuff. It’s what powers those handy trip planners we all use i. But here’s the thing: sometimes, these GTFS feeds are HUGE, especially for big cities. Dealing with all that data can feel like trying to drink from a firehose. That’s where extracting a smaller area comes in handy. It’s like creating your own little transit sandbox to play in. Let’s explore how to do it.

Cracking the GTFS Code

A GTFS feed is basically a bunch of text files – CSVs, to be exact – all zipped up together i. Each file tells a different part of the transit story:

• agency.txt: Who runs the show? This file lists the transit agencies i.
• stops.txt: Where do you get on and off? This defines the locations of all the stops and stations i.
• routes.txt: What paths do the buses and trains take? This describes the routes i.
• trips.txt: When do they run? This maps routes to specific trips and service dates i.
• stop_times.txt: The nitty-gritty details – arrival and departure times for every stop on every trip i.
• calendar.txt & calendar_dates.txt: When is service available? This defines the service periods i.
• shapes.txt: (Optional) Where exactly do they go? This gives the geographic path the vehicle travels i.

Now, imagine a GTFS feed for, say, the entire New York metropolitan area. It’s massive! Trying to analyze that whole thing at once can bog down your computer and make your analysis take forever. Extracting a smaller area? That’s like hitting the fast-forward button.

Slicing and Dicing Your GTFS Feed: A Few Approaches

Alright, so how do we actually do this extraction thing? Here are a few ways to skin this cat:

Your Toolkit for GTFS Surgery

Okay, so you know what to do. Now, how do you do it? Luckily, there are tools for the job:

• gtfs_kit (Python): This is a personal favorite. It’s super easy to use, especially the feed.restrict_to_area(bbox) method. Bang, you’ve got your spatial filter i!
• gtfstools (R): If you’re an R aficionado, this package has you covered. Functions like filter_by_agency_id() and filter_by_sf() are your friends i.
• OneBusAway GTFS Transformer (Java): A solid option if you’re in the Java world. You can use JSON config files to get pretty specific with your filtering i.
• Transitland-lib: Another good tool for extracting various aspects of a GTFS data file i.
• gtfs-utils (Node.js): A library for processing GTFS datasets, designed to work efficiently even with large files and limited memory i.
• FME: A commercial option, but it’s powerful and can handle just about anything you throw at it i.
• ArcGIS Pro: If you’re already using ArcGIS Pro, the “Transit Feed (GTFS) toolset” is worth checking out i.
• GTFS Builder: A free Microsoft Excel-based web application for creating GTFS feeds, particularly useful for smaller transit agencies i.
• Podaris: A transport network planning platform that allows importing and exporting GTFS feeds, including the ability to create smaller extracts based on custom polygons or shapefiles i.

A Little Code to Get You Started

I won’t bore you with a full-blown tutorial, but here’s a taste of how you might use gtfs_kit and gtfstools:

Python (gtfs_kit):

python