Binary Classifier Evaluation Suite

In machine learning, training a model is only half the battle—evaluating its performance is equally critical for understanding whether the model is suitable for deployment. For binary classification problems, a comprehensive set of evaluation metrics provides different perspectives on model behavior, especially when dealing with imbalanced datasets or when different types of errors carry different costs.

This problem requires you to implement a complete binary classifier evaluation suite that computes multiple interconnected metrics from the fundamental building blocks of the confusion matrix.

Understanding the Confusion Matrix

The confusion matrix is a 2×2 table that summarizes the prediction results for a binary classification problem:

	Predicted Positive (1)	Predicted Negative (0)
Actual Positive (1)	True Positives (TP)	False Negatives (FN)
Actual Negative (0)	False Positives (FP)	True Negatives (TN)

• True Positives (TP): Correctly predicted positive cases
• False Negatives (FN): Positive cases incorrectly predicted as negative (Type II error)
• False Positives (FP): Negative cases incorrectly predicted as positive (Type I error)
• True Negatives (TN): Correctly predicted negative cases

Metrics to Compute

1. Accuracy

The proportion of all predictions that are correct: $$\text{Accuracy} = \frac{TP + TN}{TP + TN + FP + FN}$$

2. F1 Score

The harmonic mean of precision and recall, providing a balanced measure: $$\text{Precision} = \frac{TP}{TP + FP}$$ $$\text{Recall} = \frac{TP}{TP + FN}$$ $$\text{F1 Score} = 2 \times \frac{\text{Precision} \times \text{Recall}}{\text{Precision} + \text{Recall}}$$

If both precision and recall are zero, the F1 Score is defined as 0.

3. Specificity (True Negative Rate)

The proportion of actual negatives that are correctly identified: $$\text{Specificity} = \frac{TN}{TN + FP}$$

If there are no actual negatives (TN + FP = 0), specificity is undefined and should return 0.

4. Negative Predictive Value (NPV)

The proportion of negative predictions that are correct: $$\text{NPV} = \frac{TN}{TN + FN}$$

If there are no negative predictions (TN + FN = 0), NPV is undefined and should return 0.

Your Task

Implement a function that takes two lists—ground truth labels and model predictions—and returns a tuple containing the confusion matrix and all four metrics. Round all float values to 3 decimal places.