How Big Is Each Group?

A bike courier has marked her drop-off points on a city map. Each point is [x, y], both in kilometres, so the two columns already share the same units.

She wants to split the points into k delivery zones and see how many stops land in each zone.

Task: write cluster_sizes(points, k).

• Fit KMeans(n_clusters=k, random_state=0, n_init=10) on the points exactly as given. Don't scale them.
• Count how many points ended up in each cluster.
• Return the counts as a list, sorted from largest to smallest.

The cluster numbers k-means hands out are arbitrary, which is why you return sorted counts and not the labels themselves.