Online Python classes in Würzburg

import csv
from collections import Counter

counts = {}
with open("wuerzburg_parking.csv", encoding="utf-8") as f:
    for row in csv.DictReader(f):
        counts.setdefault(row["car_park"], Counter())[row["status"]] += 1

def share(c, labels):
    return 100 * sum(c[l] for l in labels) / sum(c.values())

ranked = sorted(counts, key=lambda p: share(counts[p], ["belegt"]), reverse=True)
print("Car parks:", len(counts))
for p in ranked[:3]:
    print(p, round(share(counts[p], ["belegt"]), 1), "% belegt")

Kranengarage wins clearly: in 7.3 per cent of readings it had two or fewer spaces left. But "two spaces left" is a much easier bar to reach in a small car park than a big one. Our own arithmetic shows why: with 50 spaces, two free means 96 per cent full; with 1,000 spaces, two free means 99.8 per cent full. Ranking by "belegt" quietly favours small car parks; the file gives no sizes, so we cannot say which car parks are small.

So Version 2 counts a reading as nearly full if it is either belegt or fast voll, which brings in the percentage rule as well. Then it lists any car park that never reached even fast voll.

nearly = ["belegt", "fast voll"]
ranked = sorted(counts, key=lambda p: share(counts[p], nearly), reverse=True)
for p in ranked[:3]:
    print(p, round(share(counts[p], nearly), 1), "% belegt or fast voll")
never = [p for p in counts if share(counts[p], nearly) == 0]
print("Never even fast voll:", sorted(never))

Now Juliusspital comes first. Neither ranking is "the truth"; they answer different questions. "Where am I most likely to find no space at all?" and "Where is it most often busy?" are both reasonable, and the program must say which it answers. Splitting by day of the week, converted from the file's UTC times to Würzburg time, adds a second surprise: the car parks are not busy on the same days.

Data: Parkplatzdaten Würzburg, Smarte Region Würzburg, data from WVV/SVG, licence CC BY 4.0, exported 3 November 2026; the status definitions are quoted from the dataset description. Rankings and shares are our own. The data does not give the number of spaces in each car park, so we do not guess sizes; the 50 and 1,000 space examples are ours.

1. Ages 12 to 14. Without code: if a car park has 40 spaces, how many must be taken for it to be fast voll, and how many for belegt? Repeat for 400.
2. Ages 15 to 17. Run both versions, then write Version 3 that ranks by fast voll alone. Predict the top car park first.
3. Oberstufe practice. On paper, trace share() for a Counter of 6 frei, 3 fast voll and 1 belegt with each label list. Bavaria's Informatik Abitur rewards exactly this kind of tracing.
4. Adults. Convert the times to Würzburg local time with zoneinfo and find the hour when Juliusspital is most often nearly full. Explain why converting matters across 25 October 2026.