---

name: figure-grids description: "Use seaborn grid objects and figure-level plot returns for facets, pair/joint layouts, legend movement, axes access, and layout customization." disable-model-invocation: true metadata: disco-role: operating license: BSD 3-Clause

Figure Grids

Use this sub-skill when a task involves FacetGrid, PairGrid, JointGrid, pairplot, jointplot, or customizing a grid returned by relplot, displot, catplot, or lmplot.

Route Here For

• Mapping functions over facets with FacetGrid.map() or FacetGrid.map_dataframe().
• Pairwise variable grids with PairGrid, pairplot, and lower/upper/diagonal mappings.
• Joint and marginal plots with JointGrid and jointplot.
• Accessing g.figure, g.axes, g.axes_dict, g.ax, or g.legend.
• Resizing figure-level outputs, setting axis labels/titles, adding reference lines, and moving legends.
• Debugging ax= misuse with figure-level functions.

Use Another Sub-skill For

• Choosing statistical plot function parameters: ../function-interface/SKILL.md.
• Data shape/variable validation: ../data-utilities/SKILL.md.
• Theme/palette decisions: ../themes-palettes/SKILL.md.

Start With

1. If the user names a figure-level function, capture its return (g = sns.relplot(...)) and customize g, not a preexisting ax.
2. Use height and aspect for seaborn figure-level sizing; use g.figure.set_size_inches(...) when exact final size matters.
3. Use g.axes_dict when facets have semantic keys and g.axes.flat for bulk axis iteration.
4. Use sns.move_legend(g_or_ax, ...) to recreate legends safely.
5. Save through g.figure.savefig(...) or g.savefig(...) when available.

References

• Grid class and convenience API map: references/api-reference.md.
• Layout and customization recipes: references/workflows.md.
• Failure recovery: references/troubleshooting.md.

Quick Smoke Check

python sub-skills/figure-grids/scripts/grid_smoke.py --output-dir /tmp/seaborn-grid-smoke