data-visualization

Generate publication-quality charts from GPU-accelerated dataframe results by applying headless matplotlib and seaborn patterns to visualize cuDF and cuML outputs, saving high-resolution PNG files directly to your workspace.

Create multi-panel figures for data distributions and feature importance by composing 1 to 4 panel layouts using matplotlib and seaborn, combining histograms with KDE, scatter diagnostics, and model summaries into a single PNG.

Matplotlib fails to render in a headless GPU sandbox because the default backend requires a display. Initialize the Agg backend using matplotlib.use('Agg') to enable headless rendering and save visual outputs as PNG files.

Yes, you can use seaborn for correlation heatmaps and histograms, with optional SciPy support for KDE calculations. The skill provides ready patterns for these chart types alongside bar, line, scatter, and box plots.

The best way to save time series visualizations for automated pipelines is using headless matplotlib rendering to generate high-resolution PNG files, allowing downstream embedding into reports without manual figure tuning.