3.f) Bootstrapped estimates of the ATE Continuing in this train of thought, one way to account for this uncertainty in our analysis is to create an interval that will, with 95% confidence, contain the true average treatment effect. To do this, we can employ a strategy that we explored earlier in the course: the bootstrap! Fill in the skeleton code below to create a bootstrap algorithm. We'll use this function to compute the 95% confidence interval on the estimates of . Hint: To pass the autograder, make sure you sample the data using pandas (see the documentation here) and pass the random_state argument to the function call