Creating Heatmaps

Heatmaps use color to represent values in a matrix, making patterns and relationships immediately visible. They're essential for correlation analysis, confusion matrices, and many other applications.

Basic Heatmap

Adding Value Annotations

Diverging Colormaps

For data with a meaningful center (e.g., change from baseline):

Calendar Heatmap

Custom Color Normalization

Practice: Create a Performance Heatmap

Key Takeaways

• Heatmaps visualize matrix data using color intensity
• Use imshow() for basic heatmaps
• Add annotations with ax.text() for each cell
• Choose text color based on background for readability
• Diverging colormaps for data with meaningful center
• Use percentile normalization to handle outliers
• Always add a colorbar to explain the color scale

In the next lesson, we'll explore correlation matrices and their visualization.