pauc 0.1.7

Compute ROC AUC and confidence intervals using DeLong’s method

pAUC: A simple Python package to calculate ROC AUC confidence intervals using DeLong’s method

---

📌 Installation

pip install pauc

---

📌 Quick Usage

from pauc import roc_auc_ci_score
import numpy as np

y_true = np.array([0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0, 1])
y_pred = np.array([0.1, 0.35, 0.24, 0.8, 0.2, 0.85, 0.13, 0.85, 0.74, 0.58, 0.71, 0.25])

auc, (lb, ub) = roc_auc_ci_score(y_true, y_pred)
print(f'AUC: {auc:.3f}, 95% CI: ({lb:.3f}, {ub:.3f})')

AUC: 0.708
95% CI: (0.378, 1.000)

---

📌 Comparing Two Models

from pauc.roc_auc_ci import delong_roc_test
import numpy as np

y_true = np.array([0, 0, 1, 1, 0, 1, 0])
pred1 = np.array([0.1, 0.35, 0.4, 0.8, 0.2, 0.75, 0.1])
pred2 = np.array([0.5, 0.92, 0.1, 0.1, 0.8, 0.95, 0.9])

log_pval = delong_roc_test(y_true, pred1, pred2)
p_value = 10 ** log_pval
print(f"DeLong’s test p-value: {p_value}")

---

📌 Plot ROC Curve with Confidence Interval

from pauc import plot_roc_with_ci
plot_roc_with_ci(y_true, y_pred)

This displays:

• ✅ Mean ROC curve
• 📉 Shaded 95% CI band from bootstrapping
• 📈 AUC with TPR envelope AUC range
• ℹ️ TPR envelope range is not a formal statistical CI—it's the area under the lower/upper percentile ROC curves.

---

📌 Why DeLong’s Test?

DeLong’s method (DeLong et al. 1988, Sun and Xu 2014) is:

• 📈 Statistically robust and widely used
• ✅ Ideal for estimating AUC confidence intervals
• 🔁 Suitable for comparing correlated ROC curves

---

📌 Citation

If you use pAUC, please cite:

• DeLong et al., Biometrics, 1988:
  Comparing the areas under two or more correlated ROC curves: a nonparametric approach

• Sun & Xu, IEEE Signal Processing Letters, 2014:
  Fast Implementation of DeLong’s Algorithm for Comparing the Areas Under Correlated ROC Curves

---

📌 License

Released under the MIT License.

---

✨ Enjoy using pAUC for statistically sound AUC comparisons!