Red Deer data projectHow many buses run in Red Deer on a given day? A Python project on the city's transit feed
The City of Red Deer publishes its transit schedules as a GTFS feed, a zip of linked text files, on its open data portal. Our students answer one plain question with it and learn how dates really work in code. All counts are ours, from the zip downloaded on 22 October 2026.
Our counts of scheduled trips in the Red Deer Transit feed for sample dates, from the script below. A trip is one run of one bus along a route| Date | Day | Service in force | Trips scheduled |
|---|
| 22 October 2026 | Thursday | Weekday | 724 |
| 24 October 2026 | Saturday | Saturday | 623 |
| 25 October 2026 | Sunday | Sunday | 318 |
| 11 November 2026 | Wednesday | Sunday timetable (exception in the file) | 318 |
| 27 November 2026 | Friday | Special Friday service (exception in the file) | 658 |
| 6 December 2026 | Sunday | None: the feed has ended | 0 |
This is the script that prints those numbers. It uses only Python's standard library and reads the files straight out of the zip, so nothing needs unpacking first:import csv
import datetime
import io
import zipfile
feed = zipfile.ZipFile("gtfs.zip")
def table(name):
with feed.open(name) as f:
return list(csv.DictReader(io.TextIOWrapper(f, encoding="utf-8-sig")))
calendar = table("calendar.txt")
exceptions = table("calendar_dates.txt")
trips = table("trips.txt")
days = ["monday", "tuesday", "wednesday", "thursday", "friday", "saturday", "sunday"]
def services_on(day):
stamp = day.strftime("%Y%m%d")
active = set()
for row in calendar:
if row["start_date"] <= stamp <= row["end_date"] and row[days[day.weekday()]] == "1":
active.add(row["service_id"])
for row in exceptions:
if row["date"] == stamp:
if row["exception_type"] == "1":
active.add(row["service_id"])
else:
active.discard(row["service_id"])
return active
def trips_on(day):
active = services_on(day)
return sum(1 for trip in trips if trip["service_id"] in active)
for text in ["2026-10-22", "2026-10-24", "2026-10-25", "2026-11-11", "2026-11-27", "2026-12-06"]:
day = datetime.date.fromisoformat(text)
print(text, day.strftime("%A"), trips_on(day))
1. Rules first
calendar.txt gives four services, each a set of weekdays between two dates: service 1 runs Monday to Friday, 2 only on Fridays, 3 on Sundays and 4 on Saturdays, all from 30 August to 5 December 2026. The function checks the date range and the weekday column.
2. Then the exceptions
calendar_dates.txt overrides the rules. It switches weekday service off on 7 September, 30 September, 12 October, 11 November and 27 November 2026, and switches Sunday service on for the first four of those dates. Students look the dates up and work out why.
3. Then the trap
Service 2 says "Friday" in the calendar, yet the exceptions remove it on 13 of the 14 Fridays in the feed and add it back only on 27 November. Read the calendar alone and you would count 658 extra trips every Friday that never run.
Zero does not mean no buses
Ask the script about 6 December 2026 and it prints 0. That is not a service cut; the feed we downloaded simply ends on 5 December 2026, and the next one had not been published. A student who reports "Red Deer has no buses in December" has made the most common data mistake there is: treating missing data as a fact about the world.
Other things students find: the feed lists 613 stops and 2,323 trips across 42 routes, of which 40 actually have trips; the earliest weekday departure is at 04:57 and the latest at 23:54; and the stop with the most weekday stop times, 541 of them, is "49 AV @ 48 ST SORENSEN STN". The agency file declares its time zone as America/Edmonton. Contains information licensed under the Open Government Licence - The City of Red Deer (licence, dataset page). Nothing here is endorsed by the City.
