Guides¶
Each guide shows one task, with code that runs and the output it gives. The guides use the bundled datasets, so you can paste the code into a Python session with no files and no network.
| Guide | Task |
|---|---|
| Load data | Build a frame from memory, from a file, from object storage, or from another library. |
| Compute graph metrics | Centrality, components, triangles, and communities, as columns. |
| Use node attributes | Join your own node table, filter on it, and aggregate over neighbours. |
| Traverse the graph | Hops, shortest paths, and random walks. |
| Weight the edges | Run a weighted algorithm with an expression over edge columns. |
| Work on a subgraph | Run algorithms over a filtered edge set or a traversal result. |
| Shape the result | Filter, sort, select, group, and join. |
| Summarize the graph | Density, path length, diameter, and describe. |
| Get the result out | Arrow, polars, dicts, Parquet, and CSV. |
Every guide starts with the same two lines: