scientific-visualization

What it does

Conveys step-by-step instructions for creating publication-ready figures for scientific manuscripts using matplotlib, seaborn, and plotly. It emphasizes multi-panel layouts, error bars, significance markers, colorblind-safe palettes, and exporting figures in vector formats (PDF, EPS, TIFF) or raster formats (TIFF, PNG) per publication guidelines. It includes examples for basic line plots, seaborn statistical plots, heatmaps, multi-panel figures, and journal-specific styling, with explicit export calls like save_publication_figure and save_for_journal.

How it works

The skill instructs the agent to:

• Apply a publication-style preset via from style_presets import apply_publication_style and apply_publication_style('default').
• Create figures with appropriate sizes (e.g., single-column 3.5 inches, or journal-specific dimensions) and configure fonts (sans-serif like Arial/Helvetica) and axis labeling with units.
• Use colorblind-safe palettes (e.g., Okabe-Ito) and suggest perceptually uniform colormaps for heatmaps (e.g., viridis, cividis; avoid jet/rainbow).
• Add error bars (SD/SEM/CI) and include statistical markers in figures; include individual data points when possible.
• Build multi-panel figures using GridSpec, ensuring consistent styling and proper panel labeling (A, B, C, ...).
• Prepare figures for specific journals by configuring journal styles and exporting with journal-specific utilities like save_for_journal.
• Save figures in required formats and DPI by invoking save_publication_figure(fig, 'name', formats=[...], dpi=...) or save_for_journal.

When to use it

Use when:

• Creating plots for scientific manuscripts and journal submissions (Nature, Science, Cell, PLOS, etc.).
• Ensuring color accessibility and publication-ready typography.
• Building multi-panel figures with consistent styling and correct sizing.
• Exporting figures in PDF/EPS/TIFF (vector preferred) at appropriate DPI.
• Preparing figures that should work in color and grayscale and conform to specific journal guidelines.

What it can touch

The skill references tools and scripts by name, including:

• style_presets (apply_publication_style, configure_for_journal)
• figure_export (save_publication_figure, save_for_journal)
• color_palettes (OKABE_ITO_LIST, apply_palette)

Caveats

• Relies on external scripts and style files located under assets/ and references/ (e.g., publication_guidelines, journal_requirements).
• Emphasizes colorblind-safe palettes and grayscale testing; specific outcomes depend on the plotted data.
• No claims of automatic compliance with every journal; requires following provided workflow and checks.