DMLIV/DRIV

README.md

@@ -326,8 +326,8 @@ from econml.iv.dr import LinearIntentToTreatDRIV
 from sklearn.ensemble import GradientBoostingRegressor, GradientBoostingClassifier
 from sklearn.linear_model import LinearRegression
 
-est = LinearIntentToTreatDRIV(model_Y_X=GradientBoostingRegressor(),
-                              model_T_XZ=GradientBoostingClassifier(),
+est = LinearIntentToTreatDRIV(model_y_xw=GradientBoostingRegressor(),
+                              model_t_xwz=GradientBoostingClassifier(),
                               flexible_model_effect=GradientBoostingRegressor())
 est.fit(Y, T, Z=Z, X=X) # OLS inference by default
 treatment_effects = est.effect(X_test)

doc/reference.rst

@@ -66,8 +66,7 @@ Double Machine Learning (DML) IV
 .. autosummary::
     :toctree: _autosummary
 
-    econml.iv.dml.DMLATEIV
-    econml.iv.dml.ProjectedDMLATEIV
+    econml.iv.dml.OrthoIV
     econml.iv.dml.DMLIV
     econml.iv.dml.NonParamDMLIV
 
@@ -79,6 +78,10 @@ Doubly Robust (DR) IV
 .. autosummary::
     :toctree: _autosummary
 
+    econml.iv.dr.DRIV
+    econml.iv.dr.LinearDRIV
+    econml.iv.dr.SparseLinearDRIV
+    econml.iv.dr.ForestDRIV
     econml.iv.dr.IntentToTreatDRIV
     econml.iv.dr.LinearIntentToTreatDRIV
 

econml/dml/_rlearner.py

@@ -372,7 +372,7 @@ def fit(self, Y, T, X=None, W=None, *, sample_weight=None, freq_weight=None, sam
                            cache_values=cache_values,
                            inference=inference)
 
-    def score(self, Y, T, X=None, W=None):
+    def score(self, Y, T, X=None, W=None, sample_weight=None):
         """
         Score the fitted CATE model on a new data set. Generates nuisance parameters
         for the new data set based on the fitted residual nuisance models created at fit time.
@@ -391,14 +391,17 @@ def score(self, Y, T, X=None, W=None):
             Features for each sample
         W: optional(n, d_w) matrix or None (Default=None)
             Controls for each sample
+        sample_weight: optional(n,) vector or None (Default=None)
+            Weights for each samples
+
 
         Returns
         -------
         score: float
             The MSE of the final CATE model on the new data.
         """
         # Replacing score from _OrthoLearner, to enforce Z=None and improve the docstring
-        return super().score(Y, T, X=X, W=W)
+        return super().score(Y, T, X=X, W=W, sample_weight=sample_weight)
 
     @property
     def rlearner_model_final_(self):

econml/dml/dml.py

@@ -460,7 +460,7 @@ def _gen_model_final(self):
     def _gen_rlearner_model_final(self):
         return _FinalWrapper(self._gen_model_final(), self.fit_cate_intercept, self._gen_featurizer(), False)
 
-    # override only so that we can update the docstring to indicate support for `StatsModelsInference`
+    # override only so that we can update the docstring to indicate support for `LinearModelFinalInference`
     @_deprecate_positional("X and W should be passed by keyword only. In a future release "
                            "we will disallow passing X and W by position.", ['X', 'W'])
     def fit(self, Y, T, X=None, W=None, *, sample_weight=None, freq_weight=None, sample_var=None, groups=None,

econml/dr/_drlearner.py

@@ -504,7 +504,7 @@ def refit_final(self, *, inference='auto'):
         return super().refit_final(inference=inference)
     refit_final.__doc__ = _OrthoLearner.refit_final.__doc__
 
-    def score(self, Y, T, X=None, W=None):
+    def score(self, Y, T, X=None, W=None, sample_weight=None):
         """
         Score the fitted CATE model on a new data set. Generates nuisance parameters
         for the new data set based on the fitted residual nuisance models created at fit time.
@@ -523,14 +523,16 @@ def score(self, Y, T, X=None, W=None):
             Features for each sample
         W: optional(n, d_w) matrix or None (Default=None)
             Controls for each sample
+        sample_weight: optional(n,) vector or None (Default=None)
+            Weights for each samples
 
         Returns
         -------
         score: float
             The MSE of the final CATE model on the new data.
         """
         # Replacing score from _OrthoLearner, to enforce Z=None and improve the docstring
-        return super().score(Y, T, X=X, W=W)
+        return super().score(Y, T, X=X, W=W, sample_weight=sample_weight)
 
     @property
     def multitask_model_cate(self):
@@ -693,7 +695,7 @@ class LinearDRLearner(StatsModelsCateEstimatorDiscreteMixin, DRLearner):
         \\theta_t(X) = \\left\\langle \\theta_t, \\phi(X) \\right\\rangle + \\beta_t
 
     where :math:`\\phi(X)` is the outcome features of the featurizers, or `X` if featurizer is None. :math:`\\beta_t`
-    is a an intercept of the CATE, which is included if ``fit_cate_intercept=True`` (Default). It fits this by
+    is an intercept of the CATE, which is included if ``fit_cate_intercept=True`` (Default). It fits this by
     running a standard ordinary linear regression (OLS), regressing the doubly robust outcome differences on X:
 
     .. math ::
@@ -1472,6 +1474,9 @@ def fit(self, Y, T, X=None, W=None, *, sample_weight=None, groups=None,
         -------
         self
         """
+        if X is None:
+            raise ValueError("This estimator does not support X=None!")
+
         return super().fit(Y, T, X=X, W=W,
                            sample_weight=sample_weight, groups=groups,
                            cache_values=cache_values, inference=inference)

econml/iv/dml/__init__.py

@@ -13,9 +13,10 @@
 
 """
 
-from ._dml import DMLATEIV, ProjectedDMLATEIV, DMLIV, NonParamDMLIV
+from ._dml import OrthoIV, DMLIV, NonParamDMLIV, DMLATEIV, ProjectedDMLATEIV
 
-__all__ = ["DMLATEIV",
-           "ProjectedDMLATEIV",
+__all__ = ["OrthoIV",
            "DMLIV",
-           "NonParamDMLIV"]
+           "NonParamDMLIV",
+           "DMLATEIV",
+           "ProjectedDMLATEIV"]