BUG: names on union and intersection for Index were inconsistent (GH9943 GH9862)

Author: Dr-Irv

doc/source/whatsnew/v0.23.0.txt

@@ -837,6 +837,7 @@ Indexing
 - Bug in :func:`IntervalIndex.symmetric_difference` where the symmetric difference with a non-``IntervalIndex`` did not raise (:issue:`18475`)
 - Bug in :class:`IntervalIndex` where set operations that returned an empty ``IntervalIndex`` had the wrong dtype (:issue:`19101`)
 - Bug in :meth:`DataFrame.drop_duplicates` where no ``KeyError`` is raised when passing in columns that don't exist on the ``DataFrame`` (issue:`19726`)
+- Bug in :func:`Index.union` where resulting names were not computed correctly for certain cases (:issue:`9943`, :issue:`9862`)
 
 
 MultiIndex

pandas/core/indexes/base.py

@@ -57,7 +57,7 @@
 import pandas.core.algorithms as algos
 import pandas.core.sorting as sorting
 from pandas.io.formats.printing import pprint_thing
-from pandas.core.ops import make_invalid_op
+from pandas.core.ops import make_invalid_op, get_op_result_name
 from pandas.core.config import get_option
 from pandas.core.strings import StringMethods
 
@@ -1191,7 +1191,7 @@ def _convert_can_do_setop(self, other):
             other = Index(other, name=self.name)
             result_name = self.name
         else:
-            result_name = self.name if self.name == other.name else None
+            result_name = get_op_result_name(self, other)
         return other, result_name
 
     def _convert_for_op(self, value):
@@ -2263,19 +2263,15 @@ def __or__(self, other):
     def __xor__(self, other):
         return self.symmetric_difference(other)
 
-    def _get_consensus_name(self, other):
+    def _get_setop_name_object(self, other):
         """
-        Given 2 indexes, give a consensus name meaning
-        we take the not None one, or None if the names differ.
-        Return a new object if we are resetting the name
+        Given 2 indexes, give a setop name and object, meaning
+        we use get_op_result_name to return the name, and then
+        return a new object if we are resetting the name
         """
-        if self.name != other.name:
-            if self.name is None or other.name is None:
-                name = self.name or other.name
-            else:
-                name = None
-            if self.name != name:
-                return self._shallow_copy(name=name)
+        name = get_op_result_name(self, other)
+        if self.name != name:
+            return self._shallow_copy(name=name)
         return self
 
     def union(self, other):
@@ -2303,10 +2299,10 @@ def union(self, other):
         other = _ensure_index(other)
 
         if len(other) == 0 or self.equals(other):
-            return self._get_consensus_name(other)
+            return self._get_setop_name_object(other)
 
         if len(self) == 0:
-            return other._get_consensus_name(self)
+            return other._get_setop_name_object(self)
 
         # TODO: is_dtype_union_equal is a hack around
         # 1. buggy set ops with duplicates (GH #13432)
@@ -2369,11 +2365,10 @@ def union(self, other):
                                   stacklevel=3)
 
         # for subclasses
-        return self._wrap_union_result(other, result)
+        return self._wrap_setop_result(other, result)
 
-    def _wrap_union_result(self, other, result):
-        name = self.name if self.name == other.name else None
-        return self.__class__(result, name=name)
+    def _wrap_setop_result(self, other, result):
+        return self.__class__(result, name=get_op_result_name(self, other))
 
     def intersection(self, other):
         """
@@ -2403,7 +2398,7 @@ def intersection(self, other):
         other = _ensure_index(other)
 
         if self.equals(other):
-            return self._get_consensus_name(other)
+            return self._get_setop_name_object(other)
 
         if not is_dtype_equal(self.dtype, other.dtype):
             this = self.astype('O')
@@ -2423,7 +2418,7 @@ def intersection(self, other):
         if self.is_monotonic and other.is_monotonic:
             try:
                 result = self._inner_indexer(lvals, rvals)[0]
-                return self._wrap_union_result(other, result)
+                return self._wrap_setop_result(other, result)
             except TypeError:
                 pass
 
@@ -3552,7 +3547,7 @@ def _join_monotonic(self, other, how='left', return_indexers=False):
             return join_index
 
     def _wrap_joined_index(self, joined, other):
-        name = self.name if self.name == other.name else None
+        name = get_op_result_name(self, other)
         return Index(joined, name=name)
 
     def _get_string_slice(self, key, use_lhs=True, use_rhs=True):

pandas/core/indexes/datetimes.py

@@ -38,6 +38,7 @@
 
 from pandas.core.indexes.base import Index, _index_shared_docs
 from pandas.core.indexes.numeric import Int64Index, Float64Index
+from pandas.core.ops import get_op_result_name
 import pandas.compat as compat
 from pandas.tseries.frequencies import to_offset, get_period_alias, Resolution
 from pandas.core.indexes.datetimelike import (
@@ -1237,7 +1238,7 @@ def _maybe_utc_convert(self, other):
         return this, other
 
     def _wrap_joined_index(self, joined, other):
-        name = self.name if self.name == other.name else None
+        name = get_op_result_name(self, other)
         if (isinstance(other, DatetimeIndex) and
                 self.offset == other.offset and
                 self._can_fast_union(other)):
@@ -1333,8 +1334,8 @@ def __iter__(self):
                                                  box="timestamp")
             return iter(converted)
 
-    def _wrap_union_result(self, other, result):
-        name = self.name if self.name == other.name else None
+    def _wrap_setop_result(self, other, result):
+        name = get_op_result_name(self, other)
         if not timezones.tz_compare(self.tz, other.tz):
             raise ValueError('Passed item and index have different timezone')
         return self._simple_new(result, name=name, freq=None, tz=self.tz)

pandas/core/indexes/interval.py

@@ -26,6 +26,7 @@
 from pandas.core.indexes.base import (
     Index, _ensure_index,
     default_pprint, _index_shared_docs)
+from pandas.core.ops import get_op_result_name
 
 from pandas._libs import Timestamp, Timedelta
 from pandas._libs.interval import (
@@ -1351,7 +1352,7 @@ def func(self, other):
                 raise TypeError(msg.format(op=op_name))
 
             result = getattr(self._multiindex, op_name)(other._multiindex)
-            result_name = self.name if self.name == other.name else None
+            result_name = get_op_result_name(self, other)
 
             # GH 19101: ensure empty results have correct dtype
             if result.empty:

pandas/core/indexes/numeric.py

@@ -18,7 +18,7 @@
 from pandas.util._decorators import Appender, cache_readonly
 import pandas.core.dtypes.concat as _concat
 import pandas.core.indexes.base as ibase
-
+from pandas.core.ops import get_op_result_name
 
 _num_index_shared_docs = dict()
 
@@ -187,7 +187,7 @@ def _convert_scalar_indexer(self, key, kind=None):
                 ._convert_scalar_indexer(key, kind=kind))
 
     def _wrap_joined_index(self, joined, other):
-        name = self.name if self.name == other.name else None
+        name = get_op_result_name(self, other)
         return Int64Index(joined, name=name)
 
     @classmethod
@@ -264,7 +264,7 @@ def _convert_index_indexer(self, keyarr):
         return keyarr
 
     def _wrap_joined_index(self, joined, other):
-        name = self.name if self.name == other.name else None
+        name = get_op_result_name(self, other)
         return UInt64Index(joined, name=name)
 
     @classmethod

pandas/core/indexes/period.py

@@ -19,6 +19,7 @@
     _ensure_object)
 from pandas.core.dtypes.dtypes import PeriodDtype
 from pandas.core.dtypes.generic import ABCSeries
+from pandas.core.ops import get_op_result_name
 
 import pandas.tseries.frequencies as frequencies
 from pandas.tseries.frequencies import get_freq_code as _gfc
@@ -996,8 +997,8 @@ def _assert_can_do_setop(self, other):
             msg = _DIFFERENT_FREQ_INDEX.format(self.freqstr, other.freqstr)
             raise IncompatibleFrequency(msg)
 
-    def _wrap_union_result(self, other, result):
-        name = self.name if self.name == other.name else None
+    def _wrap_setop_result(self, other, result):
+        name = get_op_result_name(self, other)
         result = self._apply_meta(result)
         result.name = name
         return result

pandas/core/indexes/timedeltas.py

@@ -26,6 +26,7 @@
 from pandas.core.base import _shared_docs
 from pandas.core.indexes.base import _index_shared_docs
 import pandas.core.common as com
+from pandas.core.ops import get_op_result_name
 import pandas.core.dtypes.concat as _concat
 from pandas.util._decorators import Appender, Substitution, deprecate_kwarg
 from pandas.core.indexes.datetimelike import TimelikeOps, DatetimeIndexOpsMixin
@@ -577,7 +578,7 @@ def join(self, other, how='left', level=None, return_indexers=False,
                           sort=sort)
 
     def _wrap_joined_index(self, joined, other):
-        name = self.name if self.name == other.name else None
+        name = get_op_result_name(self, other)
         if (isinstance(other, TimedeltaIndex) and self.freq == other.freq and
                 self._can_fast_union(other)):
             joined = self._shallow_copy(joined, name=name)
@@ -637,8 +638,8 @@ def _fast_union(self, other):
         else:
             return left
 
-    def _wrap_union_result(self, other, result):
-        name = self.name if self.name == other.name else None
+    def _wrap_setop_result(self, other, result):
+        name = get_op_result_name(self, other)
         return self._simple_new(result, name=name, freq=None)
 
     def intersection(self, other):

pandas/tests/indexes/test_base.py

@@ -730,7 +730,15 @@ def test_union(self):
         union = Index([]).union(first)
         assert union is first
 
-        # preserve names
+        # preserve names only when they are the same
+        # GH 9943 9862
+
+        first = Index(list('ab'), name='A')
+        second = Index(list('abc'), name='A')
+        union = first.union(second)
+        expected = Index(list('abc'), name='A')
+        tm.assert_index_equal(union, expected)
+
         first = Index(list('ab'), name='A')
         second = Index(list('ab'), name='B')
         union = first.union(second)
@@ -752,37 +760,66 @@ def test_union(self):
         first = Index(list('ab'))
         second = Index(list('ab'), name='B')
         union = first.union(second)
-        expected = Index(list('ab'), name='B')
+        expected = Index(list('ab'), name=None)
+        tm.assert_index_equal(union, expected)
+
+        # GH 9943 9862
+        first = Index(list('abc'))
+        second = Index(list('ab'), name='B')
+        union = first.union(second)
+        expected = Index(list('abc'), name=None)
         tm.assert_index_equal(union, expected)
 
         first = Index([])
         second = Index(list('ab'), name='B')
         union = first.union(second)
-        expected = Index(list('ab'), name='B')
+        expected = Index(list('ab'), name=None)
         tm.assert_index_equal(union, expected)
 
         first = Index(list('ab'))
         second = Index([], name='B')
         union = first.union(second)
-        expected = Index(list('ab'), name='B')
+        expected = Index(list('ab'), name=None)
         tm.assert_index_equal(union, expected)
 
         first = Index(list('ab'), name='A')
         second = Index(list('ab'))
         union = first.union(second)
-        expected = Index(list('ab'), name='A')
+        expected = Index(list('ab'), name=None)
+        tm.assert_index_equal(union, expected)
+
+        # GH 9943 9862
+        first = Index(list('ab'), name='A')
+        second = Index(list('abc'))
+        union = first.union(second)
+        expected = Index(list('abc'), name=None)
         tm.assert_index_equal(union, expected)
 
         first = Index(list('ab'), name='A')
         second = Index([])
         union = first.union(second)
-        expected = Index(list('ab'), name='A')
+        expected = Index(list('ab'), name=None)
         tm.assert_index_equal(union, expected)
 
         first = Index([], name='A')
         second = Index(list('ab'))
         union = first.union(second)
-        expected = Index(list('ab'), name='A')
+        expected = Index(list('ab'), name=None)
+        tm.assert_index_equal(union, expected)
+
+        # Chained unions handles names correctly
+        i1 = Index([1, 2], name='i1')
+        i2 = Index([3, 4], name='i2')
+        i3 = Index([5, 6], name='i3')
+        union = i1.union(i2.union(i3))
+        expected = i1.union(i2).union(i3)
+        tm.assert_index_equal(union, expected)
+
+        j1 = Index([1, 2], name='j1')
+        j2 = Index([], name='j2')
+        j3 = Index([], name='j3')
+        union = j1.union(j2.union(j3))
+        expected = j1.union(j2).union(j3)
         tm.assert_index_equal(union, expected)
 
         with tm.assert_produces_warning(RuntimeWarning):