BUG: names on union and intersection for Index were inconsistent (#19849)

Closes gh-9862.

xref gh-9943.

Author: Dr-Irv

doc/source/whatsnew/v0.24.0.txt

@@ -1220,6 +1220,8 @@ Indexing
 - Bug in `scalar in Index` if scalar is a float while the ``Index`` is of integer dtype (:issue:`22085`)
 - Bug in `MultiIndex.set_levels` when levels value is not subscriptable (:issue:`23273`)
 - Bug where setting a timedelta column by ``Index`` causes it to be casted to double, and therefore lose precision (:issue:`23511`)
+- Bug in :func:`Index.union` and :func:`Index.intersection` where name of the ``Index`` of the result was not computed correctly for certain cases (:issue:`9943`, :issue:`9862`)
+
 
 Missing
 ^^^^^^^

pandas/core/indexes/base.py

@@ -61,7 +61,7 @@
 import pandas.core.sorting as sorting
 from pandas.io.formats.printing import (
     pprint_thing, default_pprint, format_object_summary, format_object_attrs)
-from pandas.core.ops import make_invalid_op
+from pandas.core.ops import make_invalid_op, get_op_result_name
 from pandas.core.strings import StringMethods
 
 __all__ = ['Index']
@@ -1253,7 +1253,7 @@ 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):
@@ -2745,19 +2745,15 @@ def __or__(self, other):
     def __xor__(self, other):
         return self.symmetric_difference(other)
 
-    def _get_consensus_name(self, other):
+    def _get_reconciled_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
+        If the result of a set operation will be self,
+        return self, unless the name changes, in which
+        case make a shallow copy of self.
         """
-        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):
@@ -2785,10 +2781,10 @@ 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_reconciled_name_object(other)
 
         if len(self) == 0:
-            return other._get_consensus_name(self)
+            return other._get_reconciled_name_object(self)
 
         # TODO: is_dtype_union_equal is a hack around
         # 1. buggy set ops with duplicates (GH #13432)
@@ -2851,11 +2847,10 @@ 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._constructor(result, name=get_op_result_name(self, other))
 
     def intersection(self, other):
         """
@@ -2885,7 +2880,7 @@ def intersection(self, other):
         other = ensure_index(other)
 
         if self.equals(other):
-            return self._get_consensus_name(other)
+            return self._get_reconciled_name_object(other)
 
         if not is_dtype_equal(self.dtype, other.dtype):
             this = self.astype('O')
@@ -2905,7 +2900,7 @@ 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
 
@@ -4175,7 +4170,7 @@ 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/category.py

@@ -25,6 +25,7 @@
 import pandas.core.common as com
 import pandas.core.missing as missing
 import pandas.core.indexes.base as ibase
+from pandas.core.ops import get_op_result_name
 from pandas.core.arrays.categorical import Categorical, contains
 
 _index_doc_kwargs = dict(ibase._index_doc_kwargs)
@@ -324,6 +325,10 @@ def itemsize(self):
         # Size of the items in categories, not codes.
         return self.values.itemsize
 
+    def _wrap_setop_result(self, other, result):
+        name = get_op_result_name(self, other)
+        return self._shallow_copy(result, name=name)
+
     def get_values(self):
         """ return the underlying data as an ndarray """
         return self._data.get_values()

pandas/core/indexes/datetimes.py

@@ -34,6 +34,7 @@
 
 from pandas.core.indexes.base import Index, _index_shared_docs
 from pandas.core.indexes.numeric import Int64Index
+from pandas.core.ops import get_op_result_name
 import pandas.compat as compat
 from pandas.tseries.frequencies import to_offset, Resolution
 from pandas.core.indexes.datetimelike import (
@@ -592,6 +593,10 @@ def union(self, other):
         y : Index or DatetimeIndex
         """
         self._assert_can_do_setop(other)
+
+        if len(other) == 0 or self.equals(other) or len(self) == 0:
+            return super(DatetimeIndex, self).union(other)
+
         if not isinstance(other, DatetimeIndex):
             try:
                 other = DatetimeIndex(other)
@@ -674,7 +679,7 @@ 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.freq == other.freq and
                 self._can_fast_union(other)):
@@ -745,11 +750,11 @@ 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)
         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)
+        return self._shallow_copy(result, name=name, freq=None, tz=self.tz)
 
     def intersection(self, other):
         """
@@ -765,6 +770,10 @@ def intersection(self, other):
         y : Index or DatetimeIndex
         """
         self._assert_can_do_setop(other)
+
+        if self.equals(other):
+            return self._get_reconciled_name_object(other)
+
         if not isinstance(other, DatetimeIndex):
             try:
                 other = DatetimeIndex(other)

pandas/core/indexes/interval.py

@@ -25,6 +25,7 @@
 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 (
@@ -1048,7 +1049,7 @@ 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

@@ -21,7 +21,7 @@
 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()
 
@@ -215,7 +215,7 @@ 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
@@ -288,7 +288,7 @@ 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

@@ -12,9 +12,9 @@
     is_integer_dtype,
     is_datetime64_any_dtype,
     is_bool_dtype,
-    pandas_dtype,
+    pandas_dtype
 )
-
+from pandas.core.ops import get_op_result_name
 from pandas.core.accessor import PandasDelegate, delegate_names
 from pandas.core.indexes.datetimes import DatetimeIndex, Int64Index, Index
 from pandas.core.indexes.datetimelike import (
@@ -848,8 +848,8 @@ 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/range.py

@@ -5,7 +5,7 @@
 
 import numpy as np
 
-from pandas._libs import index as libindex
+from pandas._libs import index as libindex, lib
 import pandas.compat as compat
 from pandas.compat import get_range_parameters, lrange, range
 from pandas.compat.numpy import function as nv
@@ -263,8 +263,9 @@ def tolist(self):
     @Appender(_index_shared_docs['_shallow_copy'])
     def _shallow_copy(self, values=None, **kwargs):
         if values is None:
+            name = kwargs.get("name", self.name)
             return RangeIndex._simple_new(
-                name=self.name, **dict(self._get_data_as_items()))
+                name=name, **dict(self._get_data_as_items()))
         else:
             kwargs.setdefault('name', self.name)
             return self._int64index._shallow_copy(values, **kwargs)
@@ -344,6 +345,10 @@ def intersection(self, other):
         -------
         intersection : Index
         """
+
+        if self.equals(other):
+            return self._get_reconciled_name_object(other)
+
         if not isinstance(other, RangeIndex):
             return super(RangeIndex, self).intersection(other)
 
@@ -424,10 +429,9 @@ def union(self, other):
         union : Index
         """
         self._assert_can_do_setop(other)
-        if len(other) == 0 or self.equals(other):
-            return self
-        if len(self) == 0:
-            return other
+        if len(other) == 0 or self.equals(other) or len(self) == 0:
+            return super(RangeIndex, self).union(other)
+
         if isinstance(other, RangeIndex):
             start_s, step_s = self._start, self._step
             end_s = self._start + self._step * (len(self) - 1)
@@ -498,7 +502,12 @@ def __getitem__(self, key):
         super_getitem = super(RangeIndex, self).__getitem__
 
         if is_scalar(key):
-            n = int(key)
+            if not lib.is_integer(key):
+                raise IndexError("only integers, slices (`:`), "
+                                 "ellipsis (`...`), numpy.newaxis (`None`) "
+                                 "and integer or boolean "
+                                 "arrays are valid indices")
+            n = com.cast_scalar_indexer(key)
             if n != key:
                 return super_getitem(key)
             if n < 0:
@@ -649,7 +658,8 @@ def _evaluate_numeric_binop(self, other):
                     return op(self._int64index, other)
                     # TODO: Do attrs get handled reliably?
 
-            return _evaluate_numeric_binop
+            name = '__{name}__'.format(name=op.__name__)
+            return compat.set_function_name(_evaluate_numeric_binop, name, cls)
 
         cls.__add__ = _make_evaluate_binop(operator.add)
         cls.__radd__ = _make_evaluate_binop(ops.radd)

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
 from pandas.core.indexes.datetimelike import (
@@ -281,6 +282,10 @@ def union(self, other):
         y : Index or TimedeltaIndex
         """
         self._assert_can_do_setop(other)
+
+        if len(other) == 0 or self.equals(other) or len(self) == 0:
+            return super(TimedeltaIndex, self).union(other)
+
         if not isinstance(other, TimedeltaIndex):
             try:
                 other = TimedeltaIndex(other)
@@ -313,7 +318,7 @@ 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)
@@ -373,10 +378,6 @@ 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
-        return self._simple_new(result, name=name, freq=None)
-
     def intersection(self, other):
         """
         Specialized intersection for TimedeltaIndex objects. May be much faster
@@ -391,6 +392,10 @@ def intersection(self, other):
         y : Index or TimedeltaIndex
         """
         self._assert_can_do_setop(other)
+
+        if self.equals(other):
+            return self._get_reconciled_name_object(other)
+
         if not isinstance(other, TimedeltaIndex):
             try:
                 other = TimedeltaIndex(other)

pandas/tests/indexes/common.py

@@ -360,10 +360,10 @@ def test_has_duplicates(self, indices):
     def test_duplicated(self, indices, keep):
         if type(indices) is not self._holder:
             pytest.skip('Can only check if we know the index type')
-        if not len(indices) or isinstance(indices, MultiIndex):
+        if not len(indices) or isinstance(indices, (MultiIndex, RangeIndex)):
             # MultiIndex tested separately in:
             # tests/indexes/multi/test_unique_and_duplicates
-            pytest.skip('Skip check for empty Index and MultiIndex')
+            pytest.skip('Skip check for empty Index, MultiIndex, RangeIndex')
 
         idx = self._holder(indices)
         if idx.has_duplicates:

pandas/tests/indexes/conftest.py

@@ -15,6 +15,7 @@
                         tm.makeTimedeltaIndex(100),
                         tm.makeIntIndex(100),
                         tm.makeUIntIndex(100),
+                        tm.makeRangeIndex(100),
                         tm.makeFloatIndex(100),
                         Index([True, False]),
                         tm.makeCategoricalIndex(100),