File size: 1,986 Bytes
798602c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 | # stats/inference/ci_mean.py
import numpy as np
from scipy.stats import norm, t
from .estimators import estimate_mean, estimate_sigma
def ci_mean_analytic(
*,
data,
estimator,
alpha,
dist,
sigma_estimator,
trim_param=None,
winsor_limits=None,
weights=None,
):
"""
Analytic confidence interval for the mean.
- Mean is computed with the user-chosen mean estimator.
- σ is computed with the user-chosen deviation estimator.
"""
n = len(data)
mu_hat = estimate_mean(
data,
estimator,
trim_param=trim_param,
winsor_limits=winsor_limits,
weights=weights,
)
sigma_hat = estimate_sigma(
data=data,
estimator=sigma_estimator,
)
scale = sigma_hat / np.sqrt(n)
if dist == "t":
return (
t.ppf(alpha / 2, n - 1, loc=mu_hat, scale=scale),
t.ppf(1 - alpha / 2, n - 1, loc=mu_hat, scale=scale),
)
return (
norm.ppf(alpha / 2, loc=mu_hat, scale=scale),
norm.ppf(1 - alpha / 2, loc=mu_hat, scale=scale),
)
def ci_mean_bootstrap(
*,
data,
estimator,
alpha,
B,
trim_param=None,
winsor_limits=None,
weights=None,
):
"""
Bootstrap CI for the mean using the user-chosen mean estimator.
"""
data = np.asarray(data)
n = len(data)
boot_stats = np.empty(B)
for b in range(B):
idx = np.random.choice(n, size=n, replace=True)
boot_data = data[idx]
boot_weights = None
if weights is not None:
boot_weights = np.asarray(weights)[idx]
boot_stats[b] = estimate_mean(
boot_data,
estimator,
trim_param=trim_param,
winsor_limits=winsor_limits,
weights=boot_weights,
)
return np.quantile(boot_stats, [alpha / 2, 1 - alpha / 2])
|