Fix automl handling of positional initializer args

econml/automated_ml.py

@@ -41,11 +41,10 @@
 
 def setAutomatedMLWorkspace(create_workspace=False,
                             create_resource_group=False, workspace_region=None, *,
-                            subscription_id=None, resource_group=None, workspace_name=None, auth=None):
+                            auth=None, subscription_id, resource_group, workspace_name):
     """Set configuration file for AutomatedML actions with the EconML library. If
     ``create_workspace`` is set true, a new workspace is created
-    for the user. If ``create_workspace`` is set true, a new workspace is
-    created for the user.
+    for the user.
 
     Parameters
     ----------
@@ -68,8 +67,7 @@ def setAutomatedMLWorkspace(create_workspace=False,
         authentication portal in the browser.
 
     subscription_id: String, required
-       Definition of a class that will serve as the parent class of the
-       AutomatedMLMixin. This class must inherit from _BaseDML.
+       Azure subscription ID for the subscription under which to run the models
 
     resource_group: String, required
        Name of resource group of workspace to be created or set.
@@ -285,12 +283,12 @@ def __init__(self, *args, **kwargs):
         # Loop through the kwargs and args if any of them is an AutoMLConfig file, pass them
         # create model and pass model into final.
         new_args = ()
-        for var in args:
+        for idx, arg in enumerate(args):
             # If item is an automl config, get its corresponding
             # AutomatedML Model and add it to new_Args
-            if isinstance(var, EconAutoMLConfig):
-                var = self._get_automated_ml_model(kwarg, key)
-            new_args += (var,)
+            if isinstance(arg, EconAutoMLConfig):
+                arg = self._get_automated_ml_model(arg, f"arg{idx}")
+            new_args += (arg,)
 
         for key in kwargs:
             kwarg = kwargs[key]

econml/drlearner.py

@@ -1502,7 +1502,7 @@ def n_crossfit_splits(self):
     @n_crossfit_splits.setter
     def n_crossfit_splits(self, value):
         if value != 'raise':
-            warn("Deprecated by parameter `n_splits` and will be removed in next version.")
+            warn("Deprecated by parameter `cv` and will be removed in next version.")
         self.cv = value
 
     @property
@@ -1534,4 +1534,4 @@ def subsample_fr(self, value):
             warn("The parameter 'subsample_fr' has been deprecated and will be removed in the next version. "
                  "Use 'max_samples' instead, with the convention that "
                  "'subsample_fr=x' is equivalent to 'max_samples=x/2'.")
-            max_samples = .45 if value == 'auto' else value / 2
+            self.max_samples = .45 if value == 'auto' else value / 2

econml/tests/test_automated_ml.py

@@ -45,10 +45,10 @@
     AutomatedKernelDML = addAutomatedML(KernelDML)
     AutomatedNonParamDML = \
         addAutomatedML(NonParamDML)
-    AutomatedForestDML = addAutomatedML(ForestDML)
+    AutomatedCausalForestDML = addAutomatedML(CausalForestDML)
 
     AUTOML_SETTINGS_REG = {
-        'experiment_timeout_minutes': 1,
+        'experiment_timeout_minutes': 15,
         'enable_early_stopping': True,
         'iteration_timeout_minutes': 1,
         'max_cores_per_iteration': 1,
@@ -61,7 +61,7 @@
     }
 
     AUTOML_SETTINGS_CLF = {
-        'experiment_timeout_minutes': 1,
+        'experiment_timeout_minutes': 15,
         'enable_early_stopping': True,
         'iteration_timeout_minutes': 1,
         'max_cores_per_iteration': 1,
@@ -118,7 +118,7 @@ def automl_model_sample_weight_reg():
 
 
 @pytest.mark.automl
-class TestAutomatedDML(unittest.TestCase):
+class TestAutomatedML(unittest.TestCase):
 
     @classmethod
     def setUpClass(cls):
@@ -134,7 +134,6 @@ def setUpClass(cls):
 
     def test_nonparam(self):
         """Testing the completion of the fit and effect estimation of an automated Nonparametic DML"""
-        Y, T, X, _ = ihdp_surface_B()
         est = AutomatedNonParamDML(model_y=automl_model_reg(),
                                    model_t=automl_model_clf(),
                                    model_final=automl_model_sample_weight_reg(), featurizer=None,
@@ -144,7 +143,6 @@ def test_nonparam(self):
 
     def test_param(self):
         """Testing the completion of the fit and effect estimation of an automated Parametric DML"""
-        Y, T, X, _ = ihdp_surface_B()
         est = AutomatedLinearDML(model_y=automl_model_reg(),
                                  model_t=GradientBoostingClassifier(),
                                  featurizer=None,
@@ -154,28 +152,21 @@ def test_param(self):
 
     def test_forest_dml(self):
         """Testing the completion of the fit and effect estimation of an AutomatedForestDML"""
-
-        Y, T, X, _ = ihdp_surface_B()
-        est = AutomatedForestDML(model_y=automl_model_reg(),
-                                 model_t=GradientBoostingClassifier(),
-                                 discrete_treatment=True,
-                                 n_estimators=1000,
-                                 subsample_fr=.8,
-                                 min_samples_leaf=10,
-                                 min_impurity_decrease=0.001,
-                                 verbose=0, min_weight_fraction_leaf=.01)
+        est = AutomatedCausalForestDML(model_y=automl_model_reg(),
+                                       model_t=GradientBoostingClassifier(),
+                                       discrete_treatment=True,
+                                       n_estimators=1000,
+                                       max_samples=.4,
+                                       min_samples_leaf=10,
+                                       min_impurity_decrease=0.001,
+                                       verbose=0, min_weight_fraction_leaf=.01)
         est.fit(Y, T, X=X)
         _ = est.effect(X)
 
-
-@pytest.mark.automl
-class TestAutomatedMetalearners(unittest.TestCase):
-
     def test_TLearner(self):
         """Testing the completion of the fit and effect estimation of an AutomatedTLearner"""
         # TLearner test
         # Instantiate TLearner
-        Y, T, X, _ = ihdp_surface_B()
         est = AutomatedTLearner(models=automl_model_reg())
 
         # Test constant and heterogeneous treatment effect, single and multi output y
@@ -188,7 +179,6 @@ def test_SLearner(self):
         # Test constant treatment effect with multi output Y
         # Test heterogeneous treatment effect
         # Need interactions between T and features
-        Y, T, X, _ = ihdp_surface_B()
         est = AutomatedSLearner(overall_model=automl_model_reg())
 
         est.fit(Y, T, X=X)
@@ -206,3 +196,20 @@ def test_DALearner(self):
 
         est.fit(Y, T, X=X)
         _ = est.effect(X)
+
+    def test_positional(self):
+        """Test that positional arguments can be used with AutoML wrappers"""
+
+        class TestEstimator:
+            def __init__(self, model_x):
+                self.model_x = model_x
+
+            def fit(self, X, Y):
+                self.model_x.fit(X, Y)
+                return self
+
+            def predict(self, X):
+                return self.model_x.predict(X)
+
+        AutoMLTestEstimator = addAutomatedML(TestEstimator)
+        AutoMLTestEstimator(automl_model_reg()).fit(X, Y).predict(X)