Precision measures the proportion of observations that the algorithm predicts to be positive that actually are positive labels. In general, precision is calculated by:
Precision = True Positives / (True Positives + False Positives)
Using an example:
Precision = 150 / (150 + 60) = .714