# Don't change this cell; just run it. import numpy as np import pandas as pd # Safe settings for Pandas. pd.set_option('mode.chained_assignment', 'raise') import matplotlib.pyplot as plt %matplotlib inline # The OKpy testing system. from client.api.notebook import Notebook ok = Notebook('nobler_monkeys.ok')

nobler = pd.read_csv('nobler_monkeys.csv') nobler

rna_copies_k = np.array(nobler['RNA copies']) / 1000

# Plot the two distributions side by side. fig, axes = plt.subplots(1, 2) axes[0].hist(rna_copies_k[:8]); axes[0].set_title('Modvax group') axes[1].hist(rna_copies_k[8:]); axes[1].set_title('Placebo group');

modvax_mean = np.mean(rna_copies_k[:8]) placebo_mean = np.mean(rna_copies_k[8:]) print('Modvax mean', modvax_mean) print('Placebo mean', placebo_mean)

plac_modvax_diff = placebo_mean - modvax_mean print('Placebo - modvax mean difference', plac_modvax_diff)

#- You may want to simulate a single trial here. modvax= rna_copies_k[:8] placebo= rna_copies_k[8:] pooled= np.append(modvax, placebo) shuffled= np.random.permutation(pooled) fake_modvax= shuffled[: 8] fake_placebo= shuffled[8 :] fake_difference= np.mean(fake_modvax)- np.mean(fake_placebo) fake_difference

n=10000 counts= np.zeros(n) for i in np.arange(n): trials= np.random.randint(0, 100, size=17) cured= trials<=90 count= np.count_nonzero(cured) counts[i]= count

# Build up the sampling distribution from the ideal (null) world. n_iters = 10000 fake_diffs = np.zeros (n_iters) for i in np.arange(n_iters): modvax= rna_copies_k[:8] placebo= rna_copies_k[8:] pooled = np.append(modvax, placebo) shuffled= np.random.permutation(pooled) fake_modvax= shuffled [: 8] fake_placebo= shuffled [8 :] fake_difference= np.mean(fake_modvax)- np.mean(fake_placebo) fake_diffs[i]= fake_difference # Show the first 10 values. fake_diffs[:10]

_ = ok.grade('q_fake_diffs')

#- Consider a histogram of the sampling distribution here. plt.hist(fake_diffs)

prop_ge = sum(fake_diffs>=plac_modvax_diff) /32 # Show the proportion. prop_ge

_ = ok.grade('q_prop_ge')

# For your convenience, you can run this cell to run all the tests at once! import os _ = [ok.grade(q[:-3]) for q in os.listdir("tests") if q.startswith('q')]