---

name: plotting-diagnostics description: "Diagnose optional umap.plot static, datashaded, connectivity, diagnostic, hover/search, and nearest-neighbour workflows." disable-model-invocation: true metadata: disco-role: operating license: BSD 3-Clause

Plotting Diagnostics

Use this sub-skill when you already have a fitted UMAP-family mapper and need to inspect or recover plotting workflows in umap.plot.

Covers

• points, connectivity, diagnostic, interactive, nearest_neighbour_distribution, and show.
• Optional umap.plot import failures, dependency checks, and install advice.
• Static matplotlib, datashaded large-data, connectivity graph, Bokeh interactive, hover/search, and nearest-neighbour histogram paths.

Route Elsewhere

• Fit, transform, update, inverse-transform, nearest-neighbour graph preparation, or base mapper configuration: core embedding.
• Supervised label training, target metrics, densMAP semantics, or density interpretation: supervised density.
• Parametric UMAP model/loss-history semantics: parametric UMAP. Plotting mechanics for an already fitted 2D mapper can route back here.

First Checks

1. Confirm the mapper is fitted and exposes embedding_ or embedding.
2. Confirm the plotted embedding is 2D.
3. Confirm labels, values, subset_points, and hover_data align to plotted rows.
4. Treat plot extras as optional. Base UMAP does not require umap.plot.
5. Run scripts/check_plotting_stack.py before debugging optional plotting failures.

Recovery Path

1. If import umap.plot fails, install umap-learn[plot] or the conda package set in plotting reference.
2. If plot extras are unavailable, fall back to mapper.embedding_ and plain matplotlib or export the embedding for plotting elsewhere.
3. For static and datashaded plots, read plotting reference for API contracts and row-alignment rules.
4. For errors, display confusion, hover/search mismatches, or performance issues, read troubleshooting.

Display Rules

• Use umap.plot.output_notebook() before interactive notebook display.
• Use umap.plot.output_file("plot.html") before interactive file output.
• Use umap.plot.show(plot_to_show) on the object returned by interactive or on a matplotlib axis when you want the package display helper.