hestia-earth-models 0.73.0__py3-none-any.whl → 0.73.2__py3-none-any.whl

tests/models/utils/test_stats.py

@@ -1,186 +0,0 @@
-from numpy import inf, sqrt
-from numpy.testing import assert_almost_equal
-from pytest import mark
-
-from hestia_earth.models.utils.stats import (
-    _calc_confidence_level, add_normal_distributions, calc_confidence_level_monte_carlo, calc_precision_monte_carlo,
-    calc_required_iterations_monte_carlo, calc_z_critical, lerp_normal_distributions, subtract_normal_distributions
-)
-
-
-# confidence_level, n_sided, z_critical
-CONFIDENCE_INTERVAL_PARAMS = [
-    # 1 sided
-    (0, 1, -inf),
-    (50, 1, 0),
-    (80, 1, 0.8416),
-    (90, 1, 1.2816),
-    (95, 1, 1.6449),
-    (99, 1, 2.3263),
-    (100, 1, inf),
-    # 2 sided
-    (0, 2, 0),
-    (50, 2, 0.6745),
-    (80, 2, 1.2816),
-    (90, 2, 1.6449),
-    (95, 2, 1.9600),
-    (99, 2, 2.5758),
-    (100, 2, inf)
-]
-
-
-@mark.parametrize(
-    "confidence_level, n_sided, z_critical",
-    CONFIDENCE_INTERVAL_PARAMS,
-    ids=[f"z={z}, n={n}" for _, n, z in CONFIDENCE_INTERVAL_PARAMS]
-)
-def test_calc_confidence_level(confidence_level, n_sided, z_critical):
-    result = _calc_confidence_level(z_critical, n_sided=n_sided)
-    assert_almost_equal(result, confidence_level, decimal=2)
-
-
-@mark.parametrize(
-    "confidence_level, n_sided, z_critical",
-    CONFIDENCE_INTERVAL_PARAMS,
-    ids=[f"conf={conf}, n={n}" for conf, n, _ in CONFIDENCE_INTERVAL_PARAMS]
-)
-def test_calc_z_critical(confidence_level, n_sided, z_critical):
-    result = calc_z_critical(confidence_level, n_sided=n_sided)
-    assert_almost_equal(result, z_critical, decimal=4)
-
-
-# confidence_level, n_iterations, precision, sd
-MONTE_CARLO_PARAMS = [
-    (95, 80767, 0.01, 1.45),
-    (95, 1110, 0.01, 0.17),
-    (99, 1917, 0.01, 0.17),
-    (50, 102, 100.18, 1500)
-]
-
-
-@mark.parametrize(
-    "confidence_level, n_iterations, precision, sd",
-    MONTE_CARLO_PARAMS,
-    ids=[f"n={n}, prec={prec}, sd={sd}" for _, n, prec, sd in MONTE_CARLO_PARAMS]
-)
-def test_calc_confidence_level_monte_carlo(confidence_level, n_iterations, precision, sd):
-    result = calc_confidence_level_monte_carlo(n_iterations, precision, sd,)
-    assert_almost_equal(result, confidence_level, decimal=2)
-
-
-@mark.parametrize(
-    "confidence_level, n_iterations, precision, sd",
-    MONTE_CARLO_PARAMS,
-    ids=[f"conf={conf}, prec={prec}, sd={sd}" for conf, _, prec, sd in MONTE_CARLO_PARAMS]
-)
-def test_calc_required_iterations_monte_carlo(confidence_level, n_iterations, precision, sd):
-    result = calc_required_iterations_monte_carlo(confidence_level, precision, sd)
-    assert result == n_iterations
-
-
-@mark.parametrize(
-    "confidence_level, n_iterations, precision, sd",
-    MONTE_CARLO_PARAMS,
-    ids=[f"conf={conf}, n={n}, sd={sd}" for conf, n, _, sd in MONTE_CARLO_PARAMS]
-)
-def test_calc_precision_monte_carlo(confidence_level, n_iterations, precision, sd):
-    result = calc_precision_monte_carlo(confidence_level, n_iterations, sd)
-    assert_almost_equal(result, precision, decimal=2)
-
-
-# mu_1, sigma_1, mu_2, sigma_2, rho, sum_mean, sum_sigma, diff_mean, diff_sigma
-PARAMS_NORMAL_DIST = [
-    # 2 standard normal distributions, perfectly negative correlation
-    (0, 1, 0, 1, -1, 0, 0, 0, 2),
-    # 2 standard normal distributions, negative correlation
-    (0, 1, 0, 1, -0.5, 0, 1, 0, sqrt(3)),
-    # 2 standard normal distributions, no correlation
-    (0, 1, 0, 1, 0, 0, sqrt(2), 0, sqrt(2)),
-    # 2 standard normal distributions, positive correlation
-    (0, 1, 0, 1, 0.5, 0, sqrt(3), 0, 1),
-    # 2 standard normal distributions, perfectly positive correlation
-    (0, 1, 0, 1, 1, 0, 2, 0, 0),
-    # different normal distributions, perfectly negative correlation
-    (50000, 3000, 45000, 9000, -1, 95000, 6000, 5000, 12000),
-    # different normal distributions, no correlation
-    (50000, 3000, 45000, 9000, 0, 95000, sqrt(90000000), 5000, sqrt(90000000)),
-    # different normal distributions, perfectly positive correlation
-    (50000, 3000, 45000, 9000, 1, 95000, 12000, 5000, 6000)
-]
-IDS_ADD_NORMAL_DIST = [
-    f"N({mu_1}, {sigma_1}^2) + N({mu_2}, {sigma_2}^2), rho: {rho}"
-    for mu_1, sigma_1, mu_2, sigma_2, rho, *_ in PARAMS_NORMAL_DIST
-]
-IDS_SUBTRACT_DIST = [
-    f"N({mu_1}, {sigma_1}^2) - N({mu_2}, {sigma_2}^2), rho: {rho}"
-    for mu_1, sigma_1, mu_2, sigma_2, rho, *_ in PARAMS_NORMAL_DIST
-]
-
-
-@mark.parametrize(
-    "mu_1, sigma_1, mu_2, sigma_2, rho, sum_mean, sum_sigma, _diff_mean, _diff_sigma",
-    PARAMS_NORMAL_DIST,
-    ids=IDS_ADD_NORMAL_DIST
-)
-def test_add_normal_distributions(mu_1, sigma_1, mu_2, sigma_2, rho, sum_mean, sum_sigma, _diff_mean, _diff_sigma):
-    result = add_normal_distributions(mu_1, sigma_1, mu_2, sigma_2, rho)
-    assert result == (sum_mean, sum_sigma)
-
-
-@mark.parametrize(
-    "mu_1, sigma_1, mu_2, sigma_2, rho, _sum_mean, _sum_sigma, diff_mean, diff_sigma",
-    PARAMS_NORMAL_DIST,
-    ids=IDS_SUBTRACT_DIST
-)
-def test_subtract_normal_distributions(mu_1, sigma_1, mu_2, sigma_2, rho, _sum_mean, _sum_sigma, diff_mean, diff_sigma):
-    result = subtract_normal_distributions(mu_1, sigma_1, mu_2, sigma_2, rho)
-    assert result == (diff_mean, diff_sigma)
-
-
-# mu_1, sigma_1, mu_2, sigma_2, alpha, rho, Z_mean, Z_sigma
-PARAMS_LERP_NORMAL_DIST = [
-    # 2 standard normal distributions, perfectly negative correlation
-    (0, 1, 0, 1, 0, -1, 0, 1),
-    (0, 1, 0, 1, 0.5, -1, 0, 0),
-    (0, 1, 0, 1, 1, -1, 0, 1),
-    # 2 standard normal distributions, no correlation
-    (0, 1, 0, 1, 0, 0, 0, 1),
-    (0, 1, 0, 1, 0.5, 0, 0, sqrt(0.5)),
-    (0, 1, 0, 1, 1, 0, 0, 1),
-    # 2 standard normal distributions, perfectly positive correlation
-    (0, 1, 0, 1, 0, 1, 0, 1),
-    (0, 1, 0, 1, 0.5, 1, 0, 1),
-    (0, 1, 0, 1, 1, 1, 0, 1),
-    # different normal distributions, perfectly negative correlation
-    (10000, 3000, 5000, 2500, -0.5, -1, 12500, 5750),
-    (10000, 3000, 5000, 2500, 0, -1, 10000, 3000),
-    (10000, 3000, 5000, 2500, 0.5, -1, 7500, 250),
-    (10000, 3000, 5000, 2500, 1, -1, 5000, 2500),
-    (10000, 3000, 5000, 2500, 1.5, -1, 2500, 5250),
-    # different normal distributions, no correlation
-    (10000, 3000, 5000, 2500, -0.5, 0, 12500, sqrt(21812500)),
-    (10000, 3000, 5000, 2500, 0, 0, 10000, 3000),
-    (10000, 3000, 5000, 2500, 0.5, 0, 7500, sqrt(3812500)),
-    (10000, 3000, 5000, 2500, 1, 0, 5000, 2500),
-    (10000, 3000, 5000, 2500, 1.5, 0, 2500, sqrt(16312500)),
-    # different normal distributions, perfectly positive correlation
-    (10000, 3000, 5000, 2500, -0.5, 1, 12500, 3250),
-    (10000, 3000, 5000, 2500, 0, 1, 10000, 3000),
-    (10000, 3000, 5000, 2500, 0.5, 1, 7500, 2750.0),
-    (10000, 3000, 5000, 2500, 1, 1, 5000, 2500),
-    (10000, 3000, 5000, 2500, 1.5, 1, 2500, 2250)
-]
-IDS_LERP_NORMAL_DIST = [
-    f"N({mu_1}, {sigma_1}^2) - N({mu_2}, {sigma_2}^2), alpha: {alpha}, rho: {rho}"
-    for mu_1, sigma_1, mu_2, sigma_2, alpha, rho, *_ in PARAMS_LERP_NORMAL_DIST
-]
-
-
-@mark.parametrize(
-    "mu_1, sigma_1, mu_2, sigma_2, alpha, rho, Z_mean, Z_sigma",
-    PARAMS_LERP_NORMAL_DIST,
-    ids=IDS_LERP_NORMAL_DIST
-)
-def test_lerp_normal_distributions(mu_1, sigma_1, mu_2, sigma_2, alpha, rho, Z_mean, Z_sigma):
-    result = lerp_normal_distributions(mu_1, sigma_1, mu_2, sigma_2, alpha, rho)
-    assert result == (Z_mean, Z_sigma)