def test_doubleml_exception_confint():
dml_plr_confint = DoubleMLPLR(dml_data, ml_g, ml_m)
msg = 'joint must be True or False. Got 1.'
with pytest.raises(TypeError, match=msg):
dml_plr_confint.confint(joint=1)
msg = "The confidence level must be of float type. 5% of type <class 'str'> was passed."
with pytest.raises(TypeError, match=msg):
dml_plr_confint.confint(level='5%')
msg = r'The confidence level must be in \(0,1\). 0.0 was passed.'
with pytest.raises(ValueError, match=msg):
dml_plr_confint.confint(level=0.)
msg = r'Apply fit\(\) before confint\(\).'
with pytest.raises(ValueError, match=msg):
dml_plr_confint.confint()
msg = r'Apply fit\(\) & bootstrap\(\) before confint\(joint=True\).'
with pytest.raises(ValueError, match=msg):
dml_plr_confint.confint(joint=True)
dml_plr_confint.fit() # error message should still appear till bootstrap was applied as well
with pytest.raises(ValueError, match=msg):
dml_plr_confint.confint(joint=True)
dml_plr_confint.bootstrap()
df_ci = dml_plr_confint.confint(joint=True)
assert isinstance(df_ci, pd.DataFrame)
def test_doubleml_exception_p_adjust():
dml_plr_p_adjust = DoubleMLPLR(dml_data, ml_g, ml_m)
msg = r'Apply fit\(\) before p_adjust\(\).'
with pytest.raises(ValueError, match=msg):
dml_plr_p_adjust.p_adjust()
dml_plr_p_adjust.fit()
msg = r'Apply fit\(\) & bootstrap\(\) before p_adjust'
with pytest.raises(ValueError, match=msg):
dml_plr_p_adjust.p_adjust(method='romano-wolf')
dml_plr_p_adjust.bootstrap()
p_val = dml_plr_p_adjust.p_adjust(method='romano-wolf')
assert isinstance(p_val, pd.DataFrame)
msg = "The p_adjust method must be of str type. 0.05 of type <class 'float'> was passed."
with pytest.raises(TypeError, match=msg):
dml_plr_p_adjust.p_adjust(method=0.05)
def test_doubleml_exception_bootstrap():
dml_plr_boot = DoubleMLPLR(dml_data, ml_g, ml_m)
msg = r'Apply fit\(\) before bootstrap\(\).'
with pytest.raises(ValueError, match=msg):
dml_plr_boot.bootstrap()
dml_plr_boot.fit()
msg = 'Method must be "Bayes", "normal" or "wild". Got Gaussian.'
with pytest.raises(ValueError, match=msg):
dml_plr_boot.bootstrap(method='Gaussian')
msg = "The number of bootstrap replications must be of int type. 500 of type <class 'str'> was passed."
with pytest.raises(TypeError, match=msg):
dml_plr_boot.bootstrap(n_rep_boot='500')
msg = 'The number of bootstrap replications must be positive. 0 was passed.'
with pytest.raises(ValueError, match=msg):
dml_plr_boot.bootstrap(n_rep_boot=0)
n_treat = 1
n_rep = 2
n_folds = 3
n_obs = 100
n_rep_boot = 314
plr_dml1 = DoubleMLPLR(dml_data_plr,
Lasso(),
Lasso(),
dml_procedure='dml1',
n_rep=n_rep,
n_folds=n_folds)
plr_dml1.fit()
plr_dml1.bootstrap(n_rep_boot=n_rep_boot)
pliv_dml1 = DoubleMLPLIV(dml_data_pliv,
Lasso(),
Lasso(),
Lasso(),
dml_procedure='dml1',
n_rep=n_rep,
n_folds=n_folds)
pliv_dml1.fit()
pliv_dml1.bootstrap(n_rep_boot=n_rep_boot)
irm_dml1 = DoubleMLIRM(dml_data_irm,
Lasso(),
LogisticRegression(),
dml_procedure='dml1',