replace latlong during entity creation

featuretools/entityset/entity.py

@@ -11,7 +11,8 @@
     convert_all_variable_data,
     convert_variable_data,
     get_linked_vars,
-    infer_variable_types
+    infer_variable_types,
+    replace_latlong_nan
 )
 from featuretools.utils.wrangle import (
     _check_time_type,
@@ -82,6 +83,11 @@ def __init__(self, id, df, entityset, variable_types=None,
 
         self.set_secondary_time_index(secondary_time_index)
 
+        # Fill in single `NaN` values in LatLong variables with a tuple
+        latlongs = [k for k, v in self.variable_types.items() if v == vtypes.LatLong]
+        for latlong in latlongs:
+            self.df[latlong] = replace_latlong_nan(self.df[latlong])
+
     def __repr__(self):
         repr_out = u"Entity: {}\n".format(self.id)
         repr_out += u"  Variables:"

featuretools/primitives/standard/transform_primitive.py

@@ -5,6 +5,7 @@
     TransformPrimitive
 )
 from featuretools.utils import convert_time_units
+from featuretools.utils.entity_utils import replace_latlong_nan
 from featuretools.variable_types import (
     Boolean,
     DateOfBirth,
@@ -511,7 +512,7 @@ class Latitude(TransformPrimitive):
     def get_function(self):
         def latitude(latlong):
             if latlong.hasnans:
-                latlong = _replace_latlong_nan(latlong)
+                latlong = replace_latlong_nan(latlong)
             return pd.Series(x[0] for x in latlong)
         return latitude
 
@@ -534,7 +535,7 @@ class Longitude(TransformPrimitive):
     def get_function(self):
         def longitude(latlong):
             if latlong.hasnans:
-                latlong = _replace_latlong_nan(latlong)
+                latlong = replace_latlong_nan(latlong)
             return pd.Series(x[1] for x in latlong)
         return longitude
 
@@ -577,9 +578,9 @@ def __init__(self, unit='miles'):
     def get_function(self):
         def haversine(latlong1, latlong2):
             if latlong1.hasnans:
-                latlong1 = _replace_latlong_nan(latlong1)
+                latlong1 = replace_latlong_nan(latlong1)
             if latlong2.hasnans:
-                latlong2 = _replace_latlong_nan(latlong2)
+                latlong2 = replace_latlong_nan(latlong2)
             lat_1s = np.array([x[0] for x in latlong1])
             lon_1s = np.array([x[1] for x in latlong1])
             lat_2s = np.array([x[0] for x in latlong2])
@@ -635,8 +636,3 @@ def get_function(self):
         def age(x, time=None):
             return (time - x).dt.days / 365
         return age
-
-
-def _replace_latlong_nan(values):
-    """replace a single `NaN` value with a tuple: `(np.nan, np.nan)`"""
-    return np.where(values.isnull(), pd.Series([(np.nan, np.nan)] * len(values)), values)

featuretools/tests/entityset_tests/test_entity.py

@@ -223,3 +223,11 @@ def test_passing_strings_to_variable_types_dfs():
     features = ft.dfs(entities, relationships, target_entity="teams", features_only=True)
     name_class = features[0].entity['name'].__class__
     assert name_class == variable_types['text']
+
+
+def test_replace_latlong_nan_during_entity_creation(es):
+    nan_es = ft.EntitySet("latlong_nan")
+    df = es['log'].df.copy()
+    df['latlong'][0] = np.nan
+    entity = ft.Entity(id="nan_latlong_entity", df=df, entityset=nan_es, variable_types=es['log'].variable_types)
+    assert entity.df['latlong'][0] == (np.nan, np.nan)

featuretools/tests/primitive_tests/test_transform_features.py

@@ -622,6 +622,13 @@ def test_haversine_with_nan(es):
     assert np.allclose(values, real, atol=0.0001, equal_nan=True)
 
 
+def test_hav(es):
+    df = es['log'].df
+    df['latlong'] = np.nan
+    es['log'].update_data(df)
+    fm, features = ft.dfs(entityset=es, target_entity='sessions', trans_primitives=['latitude', 'longitude', 'haversine'], agg_primitives=['last', 'first'])
+
+
 def test_text_primitives(es):
     words = ft.Feature(es['log']['comments'], primitive=NumWords)
     chars = ft.Feature(es['log']['comments'], primitive=NumCharacters)

featuretools/tests/utils_tests/test_entity_utils.py

@@ -1,3 +1,4 @@
+import numpy as np
 import pandas as pd
 
 import featuretools as ft
@@ -6,7 +7,8 @@
     convert_all_variable_data,
     convert_variable_data,
     get_linked_vars,
-    infer_variable_types
+    infer_variable_types,
+    replace_latlong_nan
 )
 
 
@@ -163,3 +165,11 @@ def test_get_linked_vars():
 
     customers_linked_vars = get_linked_vars(es['customers'])
     assert customers_linked_vars == ['customer_id']
+
+
+def test_replace_latlong_nan():
+    values = pd.Series([(np.nan, np.nan), np.nan, (10, 5)])
+    result = replace_latlong_nan(values)
+    assert result[0] == values[0]
+    assert result[1] == (np.nan, np.nan)
+    assert result[2] == values[2]

featuretools/utils/entity_utils.py

@@ -191,3 +191,8 @@ def col_is_datetime(col):
                 return True
 
     return False
+
+
+def replace_latlong_nan(values):
+    """replace a single `NaN` value with a tuple: `(np.nan, np.nan)`"""
+    return np.where(values.isnull(), pd.Series([(np.nan, np.nan)] * len(values)), values)