Handle return type when creating features from Id variables

featuretools/primitives/primitive_base.py

@@ -12,8 +12,10 @@
     _check_timedelta
 )
 from featuretools.variable_types import (
+    Categorical,
     Datetime,
     DatetimeTimeIndex,
+    Id,
     Numeric,
     NumericTimeIndex,
     Variable
@@ -100,12 +102,14 @@ def entity(self):
     # P TODO: this should get refactored to return_type
     @property
     def variable_type(self):
+        from . import direct_feature
         feature = self
         return_type = self.return_type
 
         while return_type is None:
-            feature = feature.base_features[0]
-            return_type = feature.return_type
+            # get return type of first base feature
+            base_feature = feature.base_features[0]
+            return_type = base_feature.return_type
 
             # only the original time index should exist
             # so make this feature's return type just a Datetime
@@ -114,6 +118,13 @@ def variable_type(self):
             elif return_type == NumericTimeIndex:
                 return_type = Numeric
 
+            # direct features should keep the Id return type, but all other features should get
+            # converted to Categorical
+            if not isinstance(feature, direct_feature.DirectFeature) and return_type == Id:
+                return_type = Categorical
+
+            feature = base_feature
+
         return return_type
 
     @property

featuretools/tests/primitive_tests/test_primitive_base.py

@@ -4,7 +4,7 @@
 from ..testing_utils import make_ecommerce_entityset
 
 from featuretools.primitives import Feature, IdentityFeature, Last, Mode, Sum
-from featuretools.variable_types import Datetime, Numeric
+from featuretools.variable_types import Categorical, Datetime, Id, Numeric
 
 
 @pytest.fixture(scope='module')
@@ -106,3 +106,17 @@ def test_return_type_inference_datetime_time_index(es):
 def test_return_type_inference_numeric_time_index(es_numeric):
     last = Last(es_numeric["log"]["datetime"], es_numeric["customers"])
     assert last.variable_type == Numeric
+
+
+def test_return_type_inference_id(es):
+    # direct features should keep Id variable type
+    direct_id_feature = Feature(es["sessions"]["customer_id"], es["log"])
+    assert direct_id_feature.variable_type == Id
+
+    # aggregations of Id variable types should get converted
+    mode = Mode(es["log"]["session_id"], es["customers"])
+    assert mode.variable_type == Categorical
+
+    # also test direct feature of aggregation
+    mode_direct = Feature(mode, es["sessions"])
+    assert mode_direct.variable_type == Categorical