[confusion-matrix] confusion-matrix: Confusion Matrix Heatmap

[MarkusNeusinger]

Plot Request

Confusion matrix visualization for classification model evaluation.

Use Cases

• Binary and multi-class classification evaluation
• Model performance comparison
• Error analysis and debugging

Expected Features

• Annotated heatmap with counts/percentages
• Class labels on axes
• Color scale indicating frequency
• Optional normalization (by row/column/total)

[claude]

✅ Spec ID: confusion-matrix

Rationale: Confusion matrix is a distinct plot type in machine learning model evaluation, requiring specialized features (class labels, normalization options, true/predicted axes) beyond generic annotated heatmaps.

Add label approved to begin generation.

[github-actions]

✅ Specification Ready: confusion-matrix

PR: #2277

---

specification.md

confusion-matrix: Confusion Matrix Heatmap

Description

A specialized heatmap visualization for evaluating classification model performance, displaying the counts or proportions of predicted vs actual class labels. The confusion matrix reveals true positives, false positives, true negatives, and false negatives at a glance, making it essential for understanding model behavior, identifying class imbalances, and diagnosing specific misclassification patterns.

Applications

• Binary classification model evaluation showing sensitivity and specificity
• Multi-class classifier performance analysis with per-class accuracy breakdown
• Model comparison to identify which classes are most often confused
• Error analysis to understand systematic misclassification patterns

Data

• true_labels (categorical) - ground truth class labels for each sample
• predicted_labels (categorical) - model-predicted class labels for each sample
• class_names (string) - display names for each class on the axes
• Size: 2-20 classes (larger matrices may have readability issues)
• Example: Classification results from a trained model on test data

Notes

• Label axes clearly: "True Label" (y-axis) and "Predicted Label" (x-axis)
• Annotate cells with counts or percentages for precise interpretation
• Support normalization options: none (raw counts), by row (recall), by column (precision), or by total
• Use sequential colormap (e.g., Blues) for count data
• Include colorbar showing the value scale
• Consider highlighting diagonal (correct predictions) for visual clarity

---

Next: Add approved label to merge this specification to main.

---

🤖 spec-create workflow

[github-actions]

🚀 Specification Merged!

Specification: confusion-matrix
PR: #2277 (merged)

The specification is now in main and synced to PostgreSQL.

---

Next: Add generate:{library} labels to start generating implementations.

Available libraries: matplotlib, seaborn, plotly, bokeh, altair, plotnine, pygal, highcharts, letsplot

Or use workflow dispatch for bulk operations.

---

🤖 spec-create workflow

[MarkusNeusinger]

✅ Closed with 9/9 implementations in repo (confusion-matrix)