squash a couple of errors

Author: ResidentMario

pandas/core/generic.py

@@ -3446,8 +3446,9 @@ def fillna(self, value=None, method=None, axis=None, inplace=False,
                limit=None, downcast=None, errors=None):
         inplace = validate_bool_kwarg(inplace, 'inplace')
 
-        if value:
-            # fill values by column, not all at once, to respect dtypes.
+        # if a singular fill value is provided, validate it
+        if value is not None:
+            # fill values by column, not all at once, to respect dtypes
             if not isinstance(value, dict) and isinstance(self, ABCDataFrame):
                 value = {col: value for col in self.columns}
             try:

pandas/core/internals.py

@@ -366,6 +366,9 @@ def fillna(self, value, limit=None, inplace=False, downcast=None,
         """ fillna on the block with the value. If we fail, then convert to
         ObjectBlock and try again
         """
+        if not errors:
+            errors = 'coerce'
+
         inplace = validate_bool_kwarg(inplace, 'inplace')
 
         if not self._can_hold_na: