MARTA Developer Resources

This is my attempt to catalog information related to the MARTA's developer feeds – both for my own use and for the benefit of other programmers.

My personal opinion is that the community would be better served if this were hosted on the MARTA website, so I will happily turn over these files – at no charge, of course – if they give MARTA a head start to hosting a more detailed set of developer pages. (But I would appreciate attribution for my work.)

Contact

Please be aware that I am not affiliated with MARTA. Contact information for MARTA can be found on the agency's website at:
http://www.itsmarta.com/Contact-us.aspx

If you would like to contact me about this resource (and please feel welcome to do so, especially if you find any errors I need to correct), my e-mail address is joe@joederose.us.

Real-Time Data

MARTA has a real-time bus feed that provides data in XML (via a SOAP call) and JSON (both via SOAP and via a RESTful call). MARTA wants to be aware of any developers using the real-time feed, so I am not providing links to it here; instead, you will need to submit a request on the MARTA developer page at http://www.itsmarta.com/brd-request-form.aspx. After you submit that form, you will immediately be taken to a page with instructions on using the system. Since those instructions are a bit overwhelming on the first visit, here's what it boils down to for access to the (easier and more complete) RESTful feed:

• Click on the link immediately below the heading, "MARTA Bus Real-time Data RESTful Web Service" (near the top of the page).
• On the next page, click on both ("GetAllBus" and "GetBusByRoute/{Route}") of the "GET" links in the "Method" column of the table.
• Take note of the URL near the top of the respective pages for those options.

Once you have the URL, you can retrieve information as follows:

This will return an array of JSON objects (one JSON object per bus), that looks like this:

[{"ADHERENCE":"0","BLOCKID":"2","BLOCK_ABBR":"1-2","DIRECTION":"Southbound","LATITUDE":"33.7562339","LONGITUDE":"-84.3918858","MSGTIME":"4\/7\/2013 12:42:43 PM","ROUTE":"1","STOPID":"904680","TIMEPOINT":"Luckie St & North Ave.","TRIPID":"3698436","VEHICLE":"2981"},{"ADHERENCE":"-3","BLOCKID":"1","BLOCK_ABBR":"1-1","DIRECTION":"Northbound","LATITUDE":"33.7889023","LONGITUDE":"-84.4262254","MSGTIME":"4\/7\/2013 12:51:21 PM","ROUTE":"1","STOPID":"901556","TIMEPOINT":"Huff Rd & Howell Mill","TRIPID":"3713428","VEHICLE":"2853"}]

Following, as best I understand them, are the meanings for the field names:

GTFS (General Transit Feed Specification)

Source Information

The MARTA GTFS feed is available on the MARTA website at:
http://www.itsmarta.com/marta-developer-resources.aspx

This feed was designed for Google Transit, and meets Google's specifications as outlined at
https://developers.google.com/transit/gtfs/reference

UML

Following is a UML diagram of the table relationships in MARTA's GTFS feed. I included calendar.txt and calendar_dates.txt on this example to show the field relationship, but I think a query that incorporates these tables would be needlessly complex. For my own work, I plan to set up separate function calls to get the appropriate service_id from those tables, and simply pass in the resultant value as a parameter in the WHERE clause of the query.

Tables (.txt Files)

Following are the tables provided in the MARTA GTFS feeds. Please note the following:

• Hyperlinks are to the Google table specifications.
• Data type information is not specified by MARTA nor (for most fields) by Google. The types listed below reflect my best assumption regarding the most useful data type to apply.
• The parenthetical numbers after "text" in the "Data Type" column reflect the longest line in the current dataset – not a prescribed maximum length. For my own database design, I intend to pad those numbers somewhat to allow for longer text in the future, and I recommend that others do the same.
• Notes are my own observations about the data (which I believe to be correct, of course – but which may reflect misunderstandings on my part).

Table: agency.txt

This table provides information about the transit agency (in our case, MARTA).

Data Example

Table: stops.txt

This table provides information about specific transit stops and stations.

Data Example

Table: routes.txt

This table provides information about specific transit stops and stations.

Data Example

Table: trips.txt

This table provides information about specific transit trips (meaning from one end-point of a route to the other).

Data Example

Table: stop_times.txt

This table provides information about stop times.

Data Example

Parsing the arrival_time and departure_time fields into hour, minute, and second values is very straightforward. Most languages have a mechanism for manipulating strings based on regular expressions, and the VBA (VisualBasic for Applications) examples below illustrate a pretty common framework for parsing the information in languages without support for regular expressions. These techniques will return a text value; likely you'll want to convert it to a numerical value using the appropriate function in the language you're using – usually some variation of val(), value(), Integer() or parseInt().

Table: calendar.txt

This table provides information about the routine schedule.

Data Example

Table: calendar_dates.txt

This table provides information about interruptions in the schedule (usually for holiday service).

Data Example

Table: shapes.txt

This table provides information for drawing a map of a route.

Data Example