[roc-curve] roc-curve: ROC Curve with AUC

[MarkusNeusinger]

Plot Request

Receiver Operating Characteristic (ROC) curve for binary classification.

Use Cases

• Binary classifier evaluation
• Threshold selection
• Model comparison

Expected Features

• ROC curve (TPR vs FPR)
• Diagonal reference line (random classifier)
• AUC score displayed
• Optional: multiple curves for model comparison

[claude]

✅ Spec ID: roc-curve

Rationale: ROC curves are a fundamental classification evaluation tool showing True Positive Rate vs False Positive Rate, commonly used in ML and statistics.

Add label approved to begin generation.

[github-actions]

✅ Specification Ready: roc-curve

PR: #2278

---

specification.md

roc-curve: ROC Curve with AUC

Description

A Receiver Operating Characteristic (ROC) curve visualizes the performance of a binary classifier by plotting the True Positive Rate (TPR) against the False Positive Rate (FPR) at various classification thresholds. The Area Under the Curve (AUC) provides a single metric summarizing model performance, where 1.0 indicates perfect classification and 0.5 represents random guessing.

Applications

• Evaluating binary classification model performance in machine learning pipelines
• Comparing multiple models to select the best classifier for production
• Selecting optimal classification thresholds based on sensitivity/specificity trade-offs
• Assessing diagnostic test accuracy in medical research

Data

• fpr (numeric) - False Positive Rate values (0 to 1)
• tpr (numeric) - True Positive Rate values (0 to 1)
• auc (numeric) - Area Under the Curve score (0 to 1)
• Size: typically 100-1000 threshold points per curve
• Example: sklearn.metrics.roc_curve output from binary classification predictions

Notes

• Include a diagonal reference line (y=x) representing random classifier performance
• Display AUC score in legend or annotation
• Use distinct colors/styles when comparing multiple models
• Axes should range from 0 to 1 with equal aspect ratio preferred

---

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

---

🤖 spec-create workflow

[MarkusNeusinger]

✅ Closed with 9/9 implementations in repo (roc-curve)