Vancouver data projectWhich Vancouver neighbourhoods have the most parks with a playground?
The City of Vancouver publishes its parks and the facilities in each park as two open data files. Our older children, about 10 to 12, join them in Python. The first surprise comes before any counting. The figures are our counts, not the City’s.
The trap. Both files separate columns with semicolons, not commas. Open one with Python’s csv module in the usual way and it finds a single column whose name is the entire header line, parkid;name;official;..., so every later step fails. This is the four-line check that reveals it:import csv
with open("parks.csv", encoding="utf-8-sig") as f:
reader = csv.DictReader(f)
print(reader.fieldnames)Adding delimiter=";" fixes it. Children who meet this once never forget to look at a file before trusting it.
Parks with at least one Playgrounds row, by neighbourhood: our count from the City of Vancouver Parks and Parks facilities files, downloaded 22 October 2026 (top seven of 21)| Neighbourhood (as named in the file) | Parks with a playground |
|---|
| Renfrew-Collingwood | 12 |
| Kensington-Cedar Cottage | 11 |
| Hastings-Sunrise | 9 |
| Grandview-Woodland | 8 |
| Mount Pleasant | 8 |
| Downtown | 7 |
| Marpole | 7 |
Our script joins the two files on parkid. It first collects the IDs of parks with a Playgrounds row, then walks the parks file and counts each matching park once by neighbourhood:import csv
playground_parks = set()
with open("parks-facilities.csv", encoding="utf-8-sig") as f:
for row in csv.DictReader(f, delimiter=";"):
if row["facilitytype"] == "Playgrounds":
playground_parks.add(row["parkid"])
by_area = {}
with_washroom = 0
with open("parks.csv", encoding="utf-8-sig") as f:
for row in csv.DictReader(f, delimiter=";"):
if row["parkid"] in playground_parks:
area = row["neighbourhoodname"]
by_area[area] = by_area.get(area, 0) + 1
if row["washrooms"] == "Y":
with_washroom += 1
print(len(playground_parks), "parks list a playground")
print(with_washroom, "of them list a washroom")
for area in sorted(by_area, key=by_area.get, reverse=True)[:5]:
print(area, by_area[area])It reports 114 parks with a playground, 52 of which also list a washroom, and the five neighbourhoods at the top of the table above.
Parks, not swings
The facilities file has a facilitycount column, and a few parks list two, three or four playgrounds. We count parks, not playgrounds, so a park is counted once however many it lists, and we never add the counts into a city total.
One neighbourhood with none
The parks file uses 22 neighbourhood names. 21 have at least one park with a playground row; Shaughnessy has none in this file. That is a fact about the file, not a claim about where children play.
A useful second question
Parents of young children often care about washrooms. Only 52 of the 114 playground parks list one, so a filter on two columns at once answers a real family question.
Two numbers that disagree
The City’s own description says the Park Board maintains over 220 parks, while the parks file we downloaded has 218 rows. Both can be true: a description is written once, and the facilities extract is updated weekly. Students learn to report the number they computed and the date they downloaded it.
Data source: City of Vancouver Open Data, Parks and Parks facilities. Contains information licensed under the Open Government Licence - Vancouver (licence). The City of Vancouver does not endorse this page, our counts or our classes.
