more removing get_backdoor_variables

Signed-off-by: Nicholas Parente <[email protected]>

dowhy/causal_refuters/assess_overlap.py

@@ -41,6 +41,7 @@ def __init__(self, *args, **kwargs):
         """
         super().__init__(*args, **kwargs)
         # TODO: Check that the target estimand has backdoor variables?
+        # TODO: Is this algorithm compatible with other adjustment criterions, besides backdoor?
         self._backdoor_vars = self._target_estimand.get_backdoor_variables()
         self._cat_feats = kwargs.pop("cat_feats", [])
         self._support_config = kwargs.pop("support_config", None)

dowhy/do_sampler.py

@@ -87,9 +87,9 @@ def __init__(
         self.dep_type = [self._variable_types[var] for var in self._outcome_names]
 
         self.indep_type = [
-            self._variable_types[var] for var in self._treatment_names + self._target_estimand.get_backdoor_variables()
+            self._variable_types[var] for var in self._treatment_names + self._target_estimand.get_adjustment_set()
         ]
-        self.density_types = [self._variable_types[var] for var in self._target_estimand.get_backdoor_variables()]
+        self.density_types = [self._variable_types[var] for var in self._target_estimand.get_adjustment_set()]
 
         self.outcome_lower_support = self._data[self._outcome_names].min().values
         self.outcome_upper_support = self._data[self._outcome_names].max().values
@@ -98,9 +98,9 @@ def __init__(
 
     def _sample_point(self, x_z):
         """
-        OVerride this if your sampling method only allows sampling a point at a time.
+        Override this if your sampling method only allows sampling a point at a time.
         :param : numpy.array: x_z is a numpy array containing the values of x and z in the order of the list given by
-        self._treatment_names + self._target_estimand.get_backdoor_variables()
+        self._treatment_names + self._target_estimand.get_adjustment_set()
         :return: numpy.array:  a sampled outcome point
         """
         raise NotImplementedError
@@ -132,7 +132,7 @@ def disrupt_causes(self):
 
     def point_sample(self):
         if self.num_cores == 1:
-            sampled_outcomes = self._df[self._treatment_names + self._target_estimand.get_backdoor_variables()].apply(
+            sampled_outcomes = self._df[self._treatment_names + self._target_estimand.get_adjustment_set()].apply(
                 self._sample_point, axis=1
             )
         else:
@@ -142,7 +142,7 @@ def point_sample(self):
             sampled_outcomes = np.array(
                 p.map(
                     self.sampler.sample_point,
-                    self._df[self._treatment_names + self._target_estimand.get_backdoor_variables()].values,
+                    self._df[self._treatment_names + self._target_estimand.get_adjustment_set()].values,
                 )
             )
             sampled_outcomes = pd.DataFrame(sampled_outcomes, columns=self._outcome_names)
@@ -155,7 +155,7 @@ def sample(self):
         :return:
         """
         sampled_outcomes = self.sampler.sample(
-            self._df[self._treatment_names + self._target_estimand.get_backdoor_variables()].values
+            self._df[self._treatment_names + self._target_estimand.get_adjustment_set()].values
         )
         sampled_outcomes = pd.DataFrame(sampled_outcomes, columns=self._outcome_names)
         self._df[self._outcome_names] = sampled_outcomes

dowhy/do_samplers/kernel_density_sampler.py

@@ -13,8 +13,8 @@ def __init__(self, *args, **kwargs):
         if (
             len(self._data) > 300
             or max(
-                len(self._treatment_names + self._target_estimand.get_backdoor_variables()),
-                len(self._outcome_names + self._target_estimand.get_backdoor_variables()),
+                len(self._treatment_names + self._target_estimand.get_adjustment_set()),
+                len(self._outcome_names + self._target_estimand.get_adjustment_set()),
             )
             >= 3
         ):
@@ -32,7 +32,7 @@ def __init__(self, *args, **kwargs):
     def _fit_conditional(self):
         self.conditional_density = KDEMultivariateConditional(
             endog=self._data[self._outcome_names],
-            exog=self._data[self._treatment_names + self._target_estimand.get_backdoor_variables()],
+            exog=self._data[self._treatment_names + self._target_estimand.get_adjustment_set()],
             dep_type="".join(self.dep_type),
             indep_type="".join(self.indep_type),
             bw=self.bw,
@@ -51,7 +51,7 @@ def _construct_sampler(self):
             self.outcome_lower_support,
             self._outcome_names,
             self._treatment_names,
-            self._target_estimand.get_backdoor_variables(),
+            self._target_estimand.get_adjustment_set(),
             self._data,
             self.dep_type,
             self.indep_type,

dowhy/do_samplers/multivariate_weighting_sampler.py

@@ -53,7 +53,7 @@ def make_treatment_effective(self, x):
     def disrupt_causes(self):
         self._df["state_propensity"] = state_propensity_score(
             self._data,
-            self._target_estimand.get_backdoor_variables(),
+            self._target_estimand.get_adjustment_set(),
             self._treatment_names,
             variable_types=self._variable_types,
         )

dowhy/do_samplers/weighting_sampler.py

@@ -52,7 +52,7 @@ def make_treatment_effective(self, x):
     def disrupt_causes(self):
         self._df["propensity_score"] = state_propensity_score(
             self._data,
-            self._target_estimand.get_backdoor_variables(),
+            self._target_estimand.get_adjustment_set(),
             self._treatment_names,
             variable_types=self._variable_types,
         )