added support for metalearners from econml

dowhy/causal_estimators/econml_cate_estimator.py

@@ -20,9 +20,6 @@ class EconmlCateEstimator(CausalEstimator):
             self._observed_common_causes = pd.get_dummies(self._observed_common_causes, drop_first=True)
         else:
             self._observed_common_causes= None
-            error_msg ="No common causes/confounders present."
-            self.logger.error(error_msg)
-            raise Exception(error_msg)
 
         # Instrumental variables names, if present
         self._instrumental_variable_names = self._target_estimand.instrumental_variables
@@ -64,8 +61,13 @@ class EconmlCateEstimator(CausalEstimator):
             Z = np.reshape(np.array(self._instrumental_variables), (n_samples, len(self._instrumental_variable_names)))
 
         # Calling the econml estimator's fit method
+        (module_name, _, class_name) = self._econml_methodname.rpartition(".")
         if self.identifier_method == "backdoor":
-            self.estimator.fit(Y, T, X, W, **self.method_params["fit_params"])
+            if module_name == "econml.metalearners":
+                # Meta learners  only need X (e.g., data is from a randomized experiment)
+                self.estimator.fit(Y, T, X, **self.method_params["fit_params"])
+            else:
+                self.estimator.fit(Y, T, X, W, **self.method_params["fit_params"])
         else:
             self.estimator.fit(Y, T, X, Z, **self.method_params["fit_params"])
 
@@ -87,6 +89,7 @@ class EconmlCateEstimator(CausalEstimator):
             self._treatment_value = [self._treatment_value]
         T0_test = np.repeat([self._control_value], n_target_units, axis=0)
         T1_test = np.repeat([self._treatment_value], n_target_units, axis=0)
+
         est = self.estimator.effect(X_test, T0 = T0_test, T1 = T1_test)
         ate = np.mean(est)