Online coding and Python classes in Coatbridge

The learner downloads OpenStreetMap data for a rectangle over Coatbridge and keeps the public roads: 159.2 km in 7,195 short segments. For each segment Python works out its bearing from north with atan2, counts the road both ways (a street running north also runs south), and adds its length to one of 36 slices of ten degrees. The result is a compass chart of the town. Two numbers summarise it. Orientation entropy measures how evenly the length is spread: it is 3.584 if every slice is equal and 1.386 for a perfect grid using only four directions. The order score rescales that from 0, no preferred directions, to 1, a perfect grid.

Coatbridge scores 0.085, close to having no preferred direction at all. A third of its road length lies near its four main directions, which is exactly what you would expect if streets pointed every way equally, since those directions cover a third of the compass. Glasgow city centre scores 0.870: 92.7% of its road length runs along four directions, with the two strongest slices, centred about 10 and 100 degrees east of north, holding 18.7% and 21.4% of all the length. The compass chart shows a cross for Glasgow and something much closer to a circle for Coatbridge.

The score describes streets, not quality, and it depends on choices: the size of each slice, which roads count, and where the rectangle is drawn. Part of the exercise is changing those choices and seeing how much the answer moves.

Questioning how a number was made lets Coatbridge teenagers use AI analysis without being taken in by it, a sound reason to learn Python in 2026. The longer argument is in why teenagers should still learn to code in 2026.