Skip to content

Instantly share code, notes, and snippets.

@eustin

eustin/stat_sig_random_9.py

Created May 15, 2023 22:49
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save eustin/93e3abc81b0ad8ac514140d6af4a6c4f to your computer and use it in GitHub Desktop.
Save eustin/93e3abc81b0ad8ac514140d6af4a6c4f to your computer and use it in GitHub Desktop.
Download ZIP
stat_sig_random_9
Raw
stat_sig_random_9.py
NUM_SIMULATIONS = 10_000
simulated_diffs_in_rates = []
for _ in range(NUM_SIMULATIONS):
rng.shuffle(all_users)
control_conversion_rate = all_users[:NUM_CONTROL_USERS].mean()
variant_conversion_rate = all_users[NUM_CONTROL_USERS:].mean()
simulated_diffs_in_rates.append(variant_conversion_rate - control_conversion_rate)
simulated_diffs_in_rates = np.array(simulated_diffs_in_rates)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment