Easy

Precision Score

Easy

~12 min

code completion

Precision

When your classifier predicts "positive", how often is it right?

precision = \frac{T P}{T P + F P}

where:

TP (True Positive): predicted positive, actually positive

FP (False Positive): predicted positive, actually negative

High precision means: "When I say yes, I'm usually right."

In NumPy:

tp = np.sum((y_pred == 1) & (y_true == 1))
fp = np.sum((y_pred == 1) & (y_true == 0))
precision = tp / (tp + fp)

Your task:

Implement precision(y_true, y_pred). Assume at least one positive prediction exists.

Example Tests

2 TP, 1 FP: precision = 0.667

Input: {"y_pred":[1,1,1,0,0],"y_true":[1,0,1,0,1]}

Expected: 0.66667

No false positives: precision = 1.0

Input: {"y_pred":[1,1,0,0],"y_true":[1,1,1,0]}

Expected: 1

1 TP, 2 FP: precision = 0.333

Input: {"y_pred":[1,1,1,0],"y_true":[0,0,1,1]}

Expected: 0.33333

You can read the full problem statement above. Create a free account to run code in the browser, submit solutions, and track your progress.