Matplotlib Data Visualization Guide

Problem

Matplotlib is widely used for Python visualization, but real plotting work often runs into implicit plt.plot state, messy multi-axis layouts, overlapping labels, inconsistent color maps, and poor export settings. This skill addresses those practical engineering issues and helps turn quick exploratory charts into maintainable, export-ready visualization code.

Core Capabilities and Workflow

• Interface choice: Prefer the object-oriented API such as fig, ax = plt.subplots() to avoid pyplot statefulness; use plt shortcuts only for quick exploration.
• Object model: Organize code around Figure, Axes, Artist, and Axis so multi-panel plots, axes, legends, and annotations remain easier to maintain.
• Plot coverage: Supports common statistical and scientific charts, including line plots, scatter plots, bar charts, histograms, heatmaps, contour plots, and box plots.
• Styling and customization: Covers colors, style sheets, rcParams, annotations, and color map selection, reducing ad-hoc per-chart parameter tuning.
• Export and performance: Provides practical savefig parameters such as dpi, bbox_inches='tight', and transparent=True, and suggests rasterized=True or downsampling for large datasets.
• 3D and interactivity: Includes guidance for 3D visualization, animation, Jupyter workflows, and GUI integration.

Boundaries and Caveats

This skill is best for writing and reviewing Python data-visualization code, not for choosing business chart designs. If the project has strict GUI backend, font, publishing, or rendering constraints, verify the setup in the target environment. For complex dashboards or real-time visualization, Matplotlib can serve as a low-level plotting component, but it may not replace more specialized visualization frameworks.