The formula for calculating the confidence interval P/R for given sample parameters.˜

Question

Algorithms

Good day.
Could someone suggest an analytical quick-calculation formula for determining the confidence interval of the measured Precision / Recall values ​​of a binary classifier based on the size of the validating sample and the percentage distribution of classes in it? I didn’t find a ready-made analytical solution on the Internet, but what I wrote right away works through the sum of the series and has an error in machine rounding.