Below I show how to measure an experiment using quantile bootstrapping in Python.
(Fake) Example Data
First, let’s generate some (fake) example data with a hard-coded difference between the treatment outcome versus the control.
from typing import List
import numpy as npdef gen_fake_dists(
n: float
) -> List[int]:
n = round(n)
mu1, sigma = 10.0, 1.0 # mean and standard deviation
control = np.random.normal(mu1, sigma, n)
mu2…