Add support for column specific fillna to viz (apache#3066)

Fix apache#3029

superset/connectors/base/models.py

@@ -64,6 +64,10 @@ def uid(self):
     def column_names(self):
         return sorted([c.column_name for c in self.columns])
 
+    @property
+    def columns_types(self):
+        return {c.column_name: c.type for c in self.columns}
+
     @property
     def main_dttm_col(self):
         return "timestamp"

superset/viz.py

@@ -43,6 +43,7 @@ class BaseViz(object):
     verbose_name = "Base Viz"
     credits = ""
     is_timeseries = False
+    default_fillna = 0
 
     def __init__(self, datasource, form_data):
         if not datasource:
@@ -61,6 +62,21 @@ def __init__(self, datasource, form_data):
         self.status = None
         self.error_message = None
 
+    def get_fillna_for_type(self, col_type):
+        """Returns the value for use as filler for a specific Column.type"""
+        if col_type:
+            if col_type == 'TEXT' or col_type.startswith('VARCHAR'):
+                return ' NULL'
+        return self.default_fillna
+
+    def get_fillna_for_columns(self, columns=None):
+        """Returns a dict or scalar that can be passed to DataFrame.fillna"""
+        if columns is None:
+            return self.default_fillna
+        columns_types = self.datasource.columns_types
+        fillna = {c: self.get_fillna_for_type(columns_types.get(c)) for c in columns}
+        return fillna
+
     def get_df(self, query_obj=None):
         """Returns a pandas dataframe based on the query object"""
         if not query_obj:
@@ -102,7 +118,8 @@ def get_df(self, query_obj=None):
                 if self.datasource.offset:
                     df[DTTM_ALIAS] += timedelta(hours=self.datasource.offset)
             df.replace([np.inf, -np.inf], np.nan)
-            df = df.fillna(0)
+            fillna = self.get_fillna_for_columns(df.columns)
+            df = df.fillna(fillna)
         return df
 
     def get_extra_filters(self):

tests/core_tests.py

@@ -767,5 +767,16 @@ def test_slice_query_endpoint(self):
         assert 'language' in resp
         self.logout();
 
+    def test_viz_get_fillna_for_columns(self):
+        slc = self.get_slice("Girls", db.session)
+        q = slc.viz.query_obj()
+        results = slc.viz.datasource.query(q)
+        fillna_columns = slc.viz.get_fillna_for_columns(results.df.columns)
+        self.assertDictEqual(
+            fillna_columns,
+            {'name': ' NULL', 'sum__num': 0}
+        )
+
+
 if __name__ == '__main__':
     unittest.main()