API: Treat 1D vectors as columns in some cases for arith/cmp ops

pandas/core/frame.py

@@ -4960,7 +4960,7 @@ def _combine_match_index(self, other, func, level=None):
                                  copy=False)
         assert left.index.equals(right.index)
 
-        if left._is_mixed_type or right._is_mixed_type:
+        if ops.should_series_dispatch(left, right, func):
             # operate column-wise; avoid costly object-casting in `.values`
             return ops.dispatch_to_series(left, right, func)
         else:

pandas/core/ops.py

@@ -928,12 +928,14 @@ def should_series_dispatch(left, right, op):
     if left._is_mixed_type or right._is_mixed_type:
         return True
 
-    if not len(left.columns) or not len(right.columns):
+    if not len(left.columns) or (isinstance(right, ABCDataFrame) and
+                                 not len(right.columns)):
         # ensure obj.dtypes[0] exists for each obj
         return False
 
     ldtype = left.dtypes.iloc[0]
-    rdtype = right.dtypes.iloc[0]
+    rdtype = (right.dtypes.iloc[0]
+              if isinstance(right, ABCDataFrame) else right.dtype)
 
     if ((is_timedelta64_dtype(ldtype) and is_integer_dtype(rdtype)) or
             (is_timedelta64_dtype(rdtype) and is_integer_dtype(ldtype))):
@@ -1767,6 +1769,23 @@ def _combine_series_frame(self, other, func, fill_value=None, axis=None,
         return self._combine_match_columns(other, func, level=level)
 
 
+def _treat_as_column(left, right, axis=None):
+    # Only if it is _impossible_ to treat `right` as row-like and a precise
+    # fit to treat it as column-like do we treat it as a column.
+    if axis is not None:
+        return False
+
+    if np.ndim(right) != 1:
+        return False
+
+    if isinstance(right, ABCSeries):
+        # require an exact match
+        return (right.index.equals(left.index) and
+                not left.index.equals(left.columns))
+
+    return len(right) == len(left) != len(left.columns)
+
+
 def _align_method_FRAME(left, right, axis):
     """ convert rhs to meet lhs dims if input is list, tuple or np.ndarray """
 
@@ -1850,7 +1869,12 @@ def na_op(x, y):
         doc = _arith_doc_FRAME % op_name
 
     @Appender(doc)
-    def f(self, other, axis=default_axis, level=None, fill_value=None):
+    def f(self, other, axis=None, level=None, fill_value=None):
+
+        if _treat_as_column(self, other, axis):
+            axis = "index"
+        else:
+            axis = axis if axis is not None else default_axis
 
         other = _align_method_FRAME(self, other, axis)
 
@@ -1861,7 +1885,7 @@ def f(self, other, axis=default_axis, level=None, fill_value=None):
         elif isinstance(other, ABCSeries):
             # For these values of `axis`, we end up dispatching to Series op,
             # so do not want the masked op.
-            pass_op = op if axis in [0, "columns", None] else na_op
+            pass_op = op if axis in [0, "columns", "index", None] else na_op
             return _combine_series_frame(self, other, pass_op,
                                          fill_value=fill_value, axis=axis,
                                          level=level)
@@ -1923,7 +1947,11 @@ def _comp_method_FRAME(cls, func, special):
     @Appender('Wrapper for comparison method {name}'.format(name=op_name))
     def f(self, other):
 
-        other = _align_method_FRAME(self, other, axis=None)
+        axis = None
+        if _treat_as_column(self, other, axis=axis):
+            axis = "index"
+
+        other = _align_method_FRAME(self, other, axis=axis)
 
         if isinstance(other, ABCDataFrame):
             # Another DataFrame
@@ -1934,7 +1962,7 @@ def f(self, other):
 
         elif isinstance(other, ABCSeries):
             return _combine_series_frame(self, other, func,
-                                         fill_value=None, axis=None,
+                                         fill_value=None, axis=axis,
                                          level=None)
         else:
 

pandas/tests/arithmetic/conftest.py

@@ -153,19 +153,3 @@ def box_df_fail(request):
     Fixture equivalent to `box` fixture but xfailing the DataFrame case.
     """
     return request.param
-
-
-@pytest.fixture(params=[
-    pd.Index,
-    pd.Series,
-    pytest.param(pd.DataFrame,
-                 marks=pytest.mark.xfail(reason="Tries to broadcast "
-                                                "incorrectly",
-                                         strict=True, raises=ValueError))
-], ids=lambda x: x.__name__)
-def box_df_broadcast_failure(request):
-    """
-    Fixture equivalent to `box` but with the common failing case where
-    the DataFrame operation tries to broadcast incorrectly.
-    """
-    return request.param

pandas/tests/arithmetic/test_datetime64.py

@@ -1365,10 +1365,8 @@ def test_sub_period(self, freq, box):
                                     operator.sub, ops.rsub])
     @pytest.mark.parametrize('pi_freq', ['D', 'W', 'Q', 'H'])
     @pytest.mark.parametrize('dti_freq', [None, 'D'])
-    def test_dti_sub_pi(self, dti_freq, pi_freq, op, box_df_broadcast_failure):
+    def test_dti_sub_pi(self, dti_freq, pi_freq, op, box):
         # GH#20049 subtracting PeriodIndex should raise TypeError
-        box = box_df_broadcast_failure
-
         dti = pd.DatetimeIndex(['2011-01-01', '2011-01-02'], freq=dti_freq)
         pi = dti.to_period(pi_freq)
 

pandas/tests/arithmetic/test_period.py

@@ -344,9 +344,7 @@ class TestPeriodIndexArithmetic(object):
     # PeriodIndex - other is defined for integers, timedelta-like others,
     #   and PeriodIndex (with matching freq)
 
-    def test_parr_add_iadd_parr_raises(self, box_df_broadcast_failure):
-        box = box_df_broadcast_failure
-
+    def test_parr_add_iadd_parr_raises(self, box):
         rng = pd.period_range('1/1/2000', freq='D', periods=5)
         other = pd.period_range('1/6/2000', freq='D', periods=5)
         # TODO: parametrize over boxes for other?
@@ -388,9 +386,7 @@ def test_pi_sub_pi_with_nat(self):
         expected = pd.Index([pd.NaT, 0 * off, 0 * off, 0 * off, 0 * off])
         tm.assert_index_equal(result, expected)
 
-    def test_parr_sub_pi_mismatched_freq(self, box_df_broadcast_failure):
-        box = box_df_broadcast_failure
-
+    def test_parr_sub_pi_mismatched_freq(self, box):
         rng = pd.period_range('1/1/2000', freq='D', periods=5)
         other = pd.period_range('1/6/2000', freq='H', periods=5)
         # TODO: parametrize over boxes for other?