def test_BayesProportionsEstimation_with_a_bad_proportion_returns_ValueError(
make_a_bad_proportion, make_b_list):
with pytest.raises(ValueError) as e:
BayesProportionsEstimation(a=make_a_bad_proportion, b=make_b_list)
assert (str(
e.value
) == "the count of successes for a and/or b exceeds the number of trials")
def test_BayesProportionsEstimation_with_make_invalid_seed_returns_ValueError(
make_a_list, make_b_list, make_invalid_seed):
with pytest.raises(ValueError) as e:
BayesProportionsEstimation(a=make_a_list,
b=make_b_list,
seed=make_invalid_seed)
assert str(e.value) == "seed must be a positive integer or None"
def test_plot_posterior_with_error(make_a_list, make_b_list,
make_explicit_seed):
try:
BayesProportionsEstimation(a=make_a_list,
b=make_b_list,
seed=make_explicit_seed).posterior_plot()
except:
raise pytest.fail()
def test_infer_delta_bayes_factor_returns_correct_values(
make_a_list, make_b_list, make_explicit_seed,
make_infer_delta_bayes_factor_result):
bf, i = BayesProportionsEstimation(
a=make_a_list, b=make_b_list,
seed=make_explicit_seed).infer_delta_bayes_factor()
assert np.isclose(bf, make_infer_delta_bayes_factor_result[0])
assert i == make_infer_delta_bayes_factor_result[1]
def test_infer_delta_probability_returns_correct_values(
make_a_list, make_b_list, make_explicit_seed,
make_infer_delta_probability_result):
p, i = BayesProportionsEstimation(
a=make_a_list, b=make_b_list,
seed=make_explicit_seed).infer_delta_probability()
assert np.isclose(p, make_infer_delta_probability_result[0])
assert i == make_infer_delta_probability_result[1]
def test_BayesProportionsEstimation_quantile_summary_returns_correct_results(
make_a_list, make_b_list, make_explicit_seed,
make_quantile_summary_results):
test = np.array(
BayesProportionsEstimation(
a=make_a_list, b=make_b_list,
seed=make_explicit_seed).quantile_summary())[:, 0:3]
assert np.array_equal(test, make_quantile_summary_results)
def test_BayesProportionsEstimation_with_make_invalid_prior_returns_ValueError(
make_a_list, make_b_list, make_invalid_prior):
with pytest.raises(ValueError) as e:
BayesProportionsEstimation(a=make_a_list,
b=make_b_list,
prior_alpha=make_invalid_prior)
assert str(
e.value) == "the prior_alpha and/or prior_beta parameters must be > 0"
def test_BayesProportionsEstimation_with_a_str_returns_ValueError(
make_a_str, make_b_list):
with pytest.raises(ValueError) as e:
BayesProportionsEstimation(a=make_a_str, b=make_b_list)
assert (
str(e.value) ==
"type(a).__name__ and/or type(b).__name__ must be 'list', 'ndarray' or 'DataFrame'"
)
def test_BayesProportionsEstimation_with_mmake_explicit_seed_initialises(
make_a_list, make_b_list, make_explicit_seed):
try:
BayesProportionsEstimation(a=make_a_list,
b=make_b_list,
seed=make_explicit_seed)
except:
raise pytest.fail()
def test_BayesProportionsEstimation_with_make_explicit_prior_alpha_and_make_explicit_prior_beta_intialises(
make_a_list, make_b_list, make_explicit_prior):
try:
BayesProportionsEstimation(a=make_a_list,
b=make_b_list,
prior_alpha=make_explicit_prior)
except:
raise pytest.fail()
def test_BayesProportionsEstimation_get_posteriors_returns_correct_results(
make_a_list,
make_b_list,
make_explicit_n,
make_explicit_seed,
make_get_posterior_results,
):
test = BayesProportionsEstimation(
make_a_list, make_b_list, n=make_explicit_n,
seed=make_explicit_seed).get_posteriors()
test = np.array(pd.DataFrame(test))
assert np.allclose(test, make_get_posterior_results)
def test_BayesProportionsEstimation_with_make_a_series_and_make_b_series_initialises(
make_a_series, make_b_series):
try:
BayesProportionsEstimation(a=make_a_series, b=make_b_series)
except:
raise pytest.fail()
def test_BayesProportionsEstimation_with_make_a_numpy_and_make_b_numpy_initialises(
make_a_numpy, make_b_numpy):
try:
BayesProportionsEstimation(a=make_a_numpy, b=make_b_numpy)
except:
raise pytest.fail()
def test_BayesProportionsEstimation_with_make_a_list_and_make_b_list_initialises(
make_a_list, make_b_list):
try:
BayesProportionsEstimation(a=make_a_list, b=make_b_list)
except:
raise pytest.fail()