def test_error_p_oob(self, sdata):
sipw = StochasticIPTW(sdata.dropna(), treatment='art', outcome='dead')
with pytest.raises(ValueError):
sipw.fit(p=1.8)
with pytest.raises(ValueError):
sipw.fit(p=-0.8)
def test_error_conditional(self, sdata):
sipw = StochasticIPTW(sdata.dropna(), treatment='art', outcome='dead')
sipw.treatment_model(
model=
'male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0',
print_results=False)
with pytest.raises(ValueError):
sipw.fit(p=[0.8, 0.1, 0.1],
conditional=["df['male']==1", "df['male']==0"])
def test_cond_treatment_continuous(self, cdata):
r_pred = 1246.4285662061
sipw = StochasticIPTW(cdata.dropna(), treatment='art', outcome='cd4_wk45')
sipw.treatment_model(model='male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0',
print_results=False)
sipw.fit(p=[0.75, 0.90], conditional=["df['male']==1", "df['male']==0"])
r = sipw.marginal_outcome
npt.assert_allclose(r_pred, r, atol=1e-5)
def test_cond_treatment(self, sdata):
r_pred = 0.117340102
sipw = StochasticIPTW(sdata.dropna(), treatment='art', outcome='dead')
sipw.treatment_model(model='male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0',
print_results=False)
sipw.fit(p=[0.75, 0.90], conditional=["df['male']==1", "df['male']==0"])
r = sipw.marginal_outcome
npt.assert_allclose(r_pred, r, atol=1e-7)
def test_uncond_treatment_continuous(self, cdata):
r_pred = 1249.4477406809
sipw = StochasticIPTW(cdata.dropna(), treatment='art', outcome='cd4_wk45')
sipw.treatment_model(model='male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0',
print_results=False)
sipw.fit(p=0.8)
r = sipw.marginal_outcome
npt.assert_allclose(r_pred, r, atol=1e-5)
def test_uncond_treatment(self, sdata):
r_pred = 0.1165162207
sipw = StochasticIPTW(sdata.dropna(), treatment='art', outcome='dead')
sipw.treatment_model(model='male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0',
print_results=False)
sipw.fit(p=0.8)
r = sipw.marginal_outcome
npt.assert_allclose(r_pred, r, atol=1e-7)
def test_match_iptw_continuous(self, cdata):
model = 'male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0'
cdata = cdata.dropna().copy()
# Estimating Marginal Structural Model
ipt = IPTW(cdata, treatment='art', outcome='cd4_wk45')
ipt.treatment_model(model_denominator=model, stabilized=False, print_results=False)
ipt.marginal_structural_model('art')
ipt.fit()
# Estimating 'Stochastic Treatment'
sipw = StochasticIPTW(cdata, treatment='art', outcome='cd4_wk45')
sipw.treatment_model(model=model, print_results=False)
sipw.fit(p=1.0)
r_all = sipw.marginal_outcome
sipw.fit(p=0.0)
r_non = sipw.marginal_outcome
npt.assert_allclose(ipt.average_treatment_effect['ATE'][1], r_all - r_non, atol=1e-4)
def test_match_iptw(self, sdata):
model = 'male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0'
sdata = sdata.dropna().copy()
# Estimating Marginal Structural Model
ipt = IPTW(sdata, treatment='art', outcome='dead')
ipt.treatment_model(model_denominator=model, stabilized=False, print_results=False)
ipt.marginal_structural_model('art')
ipt.fit()
# Estimating 'Stochastic Treatment'
sipw = StochasticIPTW(sdata, treatment='art', outcome='dead')
sipw.treatment_model(model=model, print_results=False)
sipw.fit(p=1.0)
r_all = sipw.marginal_outcome
sipw.fit(p=0.0)
r_non = sipw.marginal_outcome
npt.assert_allclose(ipt.risk_difference['RD'][1], r_all - r_non, atol=1e-7)
def test_match_iptw_continuous(self, cdata):
model = 'male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0'
cdata = cdata.dropna().copy()
# Estimating Marginal Structural Model
ipt = IPTW(cdata, treatment='art', stabilized=False)
ipt.regression_models(model)
ipt.fit()
cdata['iptw'] = ipt.Weight
ind = sm.cov_struct.Independence()
f = sm.families.family.Gaussian()
linrisk = smf.gee('cd4_wk45 ~ art', cdata['id'], cdata, cov_struct=ind, family=f, weights=cdata['iptw']).fit()
# Estimating 'Stochastic Treatment'
sipw = StochasticIPTW(cdata, treatment='art', outcome='cd4_wk45')
sipw.treatment_model(model='male + age0 + age_rs1 + age_rs2 + cd40 + cd4_rs1 + cd4_rs2 + dvl0',
print_results=False)
sipw.fit(p=1.0)
r_all = sipw.marginal_outcome
sipw.fit(p=0.0)
r_non = sipw.marginal_outcome
npt.assert_allclose(linrisk.params[1], r_all - r_non, atol=1e-4)