Precision and recall are sometimes rolled together into a composite score known as an F1 score, and we would like this score to be as large as possible. If F1 is high, then both recall, and precision are also likely to be high. You won't find a situation where a model has good precision but bad recall, or vice versa.
So, we can compute the F1 score by using the f1_score() function, as follows:
The first output is the F1 score for survivorship, and the second one is the F1 score for those who died.
