plotly_skill

Overview

This skill provides ready-to-use Plotly functionality for creating interactive, publication-quality visualizations in Python. It covers Plotly Express for fast, high-level plotting and Graph Objects for fine-grained control, plus layouts, interactivity, and export options. Use it to build dashboards, exploratory figures, and web-embeddable charts with hover, zoom, and selection features.

How this skill works

The skill exposes common Plotly patterns: quick plots via plotly.express, detailed figures via plotly.graph_objects, subplots, templates, and interactive features like hover templates, rangesliders, and animations. It also includes instructions for exporting interactive HTML and static images (kaleido). Examples show scientific, statistical, time-series, and multi-plot dashboard workflows that can be adapted to DataFrame or array data.

When to use it

• You need interactive charts with hover text, zoom/pan, or selection tools.
• Building dashboards or web-embeddable figures (Dash or standalone HTML).
• Exploratory data analysis where dynamic filtering or animations help reveal patterns.
• Combining multiple traces, custom layouts, or specialized chart types (3D, maps, financial).
• Preparing interactive figures for presentations or collaborative review.

Best practices

• Start with Plotly Express for common plots; switch to Graph Objects when you need precise control or custom traces.
• Keep hover templates concise and use <extra></extra> to hide redundant trace info for clarity.
• Use templates to maintain consistent styling across many figures and dashboards.
• For reproducible static outputs, export images with kaleido; for shareable interactive files, use fig.write_html with include_plotlyjs='cdn'.
• When building dashboards, limit trace count per plot or use aggregation to preserve responsiveness.

Example use cases

• Interactive scatter matrix for genomics or omics exploratory analysis with brushing and selection.
• Time-series dashboards with rangesliders and linked subplots for clinical or financial data.
• 3D surface and mesh visualizations for materials-science or structural biology data.
• Heatmaps and annotated images for correlation matrices or spatial omics.
• Multi-panel figures combining scatter, histogram, and box plots for a publication-ready exploratory report.