---

name: analysis-visualization description: "Routes PyPSA result analysis, statistics, plotting, NetworkCollection comparisons, and temporal or spatial clustering tasks." disable-model-invocation: true metadata: disco-role: operating license: MIT

Analysis and Visualization

Use this route when you need to inspect PyPSA networks, summarize results, compare scenarios, or present outputs as static or interactive charts and maps.

Start here

• references/statistics-reference.md — metrics, filters, custom groupers, solved-versus-unsolved behavior, and NetworkCollection statistics.
• references/plotting-reference.md — static and interactive charts, n.plot.map, n.plot.iplot, and n.plot.explore.
• references/clustering-collections.md — temporal resampling/downsampling/segmentation, spatial busmaps, and NetworkCollection comparison limits.
• references/troubleshooting.md — empty statistics, plot backend issues, optional dependency skips, and collection alignment problems.
• scripts/pypsa_analysis_smoke.py — tiny headless smoke for statistics, plots, and NetworkCollection comparison.
• scripts/pypsa_clustering_smoke.py — tiny smoke for temporal clustering plus optional spatial clustering.

Use this route for

• n.stats / n.statistics metrics, filters, groupers, and custom grouper registration.
• Static or interactive statistics plots.
• Static network maps with n.plot.map and interactive maps with n.plot.iplot or n.plot.explore.
• Temporal clustering via resample, downsample, or segment.
• Spatial clustering via busmaps and clustered-network aggregation.
• Scenario comparison with NetworkCollection.

Do not use this route for

• Building or fixing the underlying network structure and component tables.
• Importing or exporting network data.
• Running optimization or power-flow setup.

Quick rules of thumb

• If the network is unsolved, start with input-side metrics such as installed capacity; solve-dependent metrics can be empty until optimization or power-flow has run.
• For headless static plots, use the bundled smoke script or set Matplotlib to an off-screen backend.
• For missing optional dependencies, read the troubleshooting reference before trying to force-install extras.
• For collection comparisons, keep scenario names explicit and align snapshots or periods whenever you need direct time-series comparison.

Typical questions this route answers

• "Why are my statistics empty?"
• "How do I group by carrier, bus carrier, country, or a custom grouper?"
• "How do I draw a headless bar, area, or map plot?"
• "How do I compare two scenario networks?"
• "How do I resample, downsample, or segment snapshots?"
• "How do I cluster buses with or without scikit-learn?"