# Create Statistical Visualizations with Seaborn

Choosing appropriate statistical plots and settings can be difficult when datasets contain multiple variable types and groups. This skill guides Seaborn plot selection, composition, styling, and troubleshooting.

## Install

```bash
npx skillstore add davila7/seaborn
```

## Metadata

- Status: approved
- Slug: davila7-seaborn
- Skillstore revision: r2
- Version status: missing
- Tree hash: 113b3d9006e8055436e1aa85a6d7c3091b4ff54b177411ed4b23aac9dfee5b15
- Author: davila7
- GitHub username: davila7
- License: MIT
- Repository: https://github.com/davila7/claude-code-templates/tree/main/cli-tool/components/skills/scientific/seaborn
- Ref: c43861a65bb95efcae259cd161c9d6f4dc7eec6f
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 76
- Public page: https://skillstore.pages.dev/skills/davila7-seaborn
- Manifest: https://skillstore.pages.dev/api/skills/davila7-seaborn/manifest

## Capabilities

- Recommends Seaborn plot types for relational, distributional, categorical, regression, and matrix data.
- Explains axes-level, figure-level, and seaborn.objects interfaces.
- Guides semantic mappings, faceting, palettes, themes, and statistical estimators.
- Describes long-form and wide-form data preparation.
- Provides patterns for exploratory, publication, multi-panel, and time-series figures.
- Diagnoses common issues involving legends, labels, figure size, colors, and KDE bandwidth.

## Use Cases

- Explore a New Dataset: Select relationship, distribution, and correlation plots that reveal structure, outliers, and group differences.
- Prepare Publication Figures: Design clear multi-panel figures with controlled themes, labels, uncertainty, palettes, and export settings.
- Build Reusable Plotting Workflows: Choose compatible Seaborn interfaces and Matplotlib integration patterns for repeatable visualization functions.

## Prompt Templates

### Choose a Basic Plot

```
Create a Seaborn plot for [goal] using DataFrame [name]. Columns: [columns]. Recommend the plot type and explain each mapping.
```

### Compare Groups

```
Compare [metric] across [category] and [group]. Use an accessible palette, display uncertainty appropriately, and include clear labels.
```

### Design a Faceted Analysis

```
Build a faceted analysis of [dataset] using [row] and [column] groups. Control ordering, scales, legends, and statistical estimation.
```

### Produce a Publication Figure

```
Design a publication figure for [research question] with Seaborn objects or Matplotlib integration. Specify transformations, layers, layout, export settings, and validation checks.
```

## Limitations

- Requires Python with Seaborn, Matplotlib, pandas, and any analysis-specific dependencies installed.
- Does not inspect datasets automatically; users must provide schemas, samples, or relevant column details.
- Generated plots still require validation for statistical correctness, accessibility, and publication requirements.
- Version-specific Seaborn APIs may differ from the reference material.

## Best Practices

- Use tidy, long-form data with meaningful column names whenever possible.
- Match plot type and statistical estimator to the data structure and analytical question.
- Validate labels, uncertainty, accessibility, and export quality before sharing a figure.

## Anti Patterns

- Do not use color as the only encoding when accessibility matters.
- Do not accept automatic aggregation or confidence intervals without checking their statistical meaning.
- Do not overload a figure with excessive categories, facets, or semantic mappings.

## Security Audit

- Audited at: 2026-07-23T15:14:27.867\+00:00
- Summary: All 201 static findings are false positives caused by Markdown backticks and ordinary Seaborn plotting terms. The skill contains documentation and Python visualization examples, with no shell execution, reconnaissance, prompt injection, credential access, or exfiltration intent. The quick-start sample may download Seaborn's public tips dataset, but it does not send user data.

## Stats

- Views: 976
- Downloads: 9
- Favorites: 0
- Popularity score: 0