Input: a CSV start,end with one row per trip. Aggregate trips into directed links (weight = number of trips), ignoring round trips (start = end) for the graph. Using networkx, print:
stations: 5, links: 9, density: 0.45
busiest stations (trips in + out): C 7, A 6, B 5
highest betweenness: C (0.83)
top flows: A->B 3, A->C 1, B->A 1
round trips (same station): 1
suggested idiom: adjacency matrix (dense network)
Strength = weighted in-degree + out-degree (top 3, ties alphabetical); betweenness on the undirected version (nx.betweenness_centrality, ties → alphabetically first); top 3 flows by weight, ties by start then end; suggest an adjacency matrix if density > 0.3, otherwise a node-link diagram (our rule of thumb).