Empty subtypes of Index return their type, rather than Index #10599
Conversation
jreback
added
Indexing
| @@ -938,6 +938,10 @@ def test_difference(self): | |||
| self.assertEqual(len(result), 0) | |||
| self.assertEqual(result.name, first.name) | |||
|
|
|||
| # empty difference for subtypes | |||
| result = self.periodIndex.difference(self.periodIndex) | |||
| self.assertIsInstance(result, pd.PeriodIndex) | |||
There was a problem hiding this comment.
use tm.assert_index_equal(...) here (and directly construct the expected)
| # GH 10596 - empty difference retains index's type | ||
|
|
||
| result = idx.difference(idx) | ||
| self.assertIsInstance(result, type(idx)) |
There was a problem hiding this comment.
so the way to tthis generically is define a create_empty method analagous to create_index for each subclass which has is an empty version of create_index.
|
@jreback I'm having some issues with comparing empty Indexes. For Timedelta, Datetime & Categorical, I get the equivalent of this: AssertionError: TimedeltaIndex([], dtype='timedelta64[ns]', freq='D') != TimedeltaIndex([], dtype='timedelta64[ns]', freq='D')I'd love to take a look at the wider problem in the future; in the meantime is there a way of getting this fix in? Is there a strong reason you don't like the type test rather than the direct comparison? |
|
no, use |
|
just asserting type papers over the meta data propogation, so its just creates a new bug, rather than fixing the existing one. |
max-sixty
force-pushed
the
master
branch
3 times, most recently
from
767f716 to
d1e478f
Compare
|
Great @jreback, that should be good to go. Let me know if I've missed anything. Cheers |
|
pls add a release note in whatsnew/v0.17.0 squash to a single commit ping when green. |
max-sixty
force-pushed
the
master
branch
3 times, most recently
from
d64c90b to
beab262
Compare
|
I have a PyTables test fail - does anyone have any idea what this might be? I've dug around, but not getting anywhere fast; and have zero PyTables experience / installation. ERROR: test_to_hdf_with_object_column_names (pandas.io.tests.test_pytables.TestHDFStore)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/travis/build/pydata/pandas/pandas/io/tests/test_pytables.py", line 4678, in test_to_hdf_with_object_column_names
df.to_hdf(path, 'df', format='table', data_columns=True)
File "/home/travis/build/pydata/pandas/pandas/core/generic.py", line 920, in to_hdf
return pytables.to_hdf(path_or_buf, key, self, **kwargs)
File "/home/travis/build/pydata/pandas/pandas/io/pytables.py", line 269, in to_hdf
f(store)
File "/home/travis/build/pydata/pandas/pandas/io/pytables.py", line 264, in <lambda>
f = lambda store: store.put(key, value, **kwargs)
File "/home/travis/build/pydata/pandas/pandas/io/pytables.py", line 826, in put
self._write_to_group(key, value, append=append, **kwargs)
File "/home/travis/build/pydata/pandas/pandas/io/pytables.py", line 1275, in _write_to_group
s.write(obj=value, append=append, complib=complib, **kwargs)
File "/home/travis/build/pydata/pandas/pandas/io/pytables.py", line 3798, in write
**kwargs)
File "/home/travis/build/pydata/pandas/pandas/io/pytables.py", line 3404, in create_axes
axis=axis
File "/home/travis/build/pydata/pandas/pandas/core/frame.py", line 2522, in reindex_axis
fill_value=fill_value)
File "/home/travis/build/pydata/pandas/pandas/core/generic.py", line 1852, in reindex_axis
limit=limit)
File "/home/travis/build/pydata/pandas/pandas/core/index.py", line 3062, in reindex
if not is_categorical_dtype(target) and not target.is_unique:
File "properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:39442)
File "/home/travis/build/pydata/pandas/pandas/core/index.py", line 744, in is_unique
return self._engine.is_unique
File "index.pyx", line 213, in pandas.index.IndexEngine.is_unique.__get__ (pandas/index.c:4392)
File "index.pyx", line 248, in pandas.index.IndexEngine._do_unique_check (pandas/index.c:4874)
File "index.pyx", line 261, in pandas.index.IndexEngine._ensure_mapping_populated (pandas/index.c:5047)
File "index.pyx", line 267, in pandas.index.IndexEngine.initialize (pandas/index.c:5135)
File "hashtable.pyx", line 703, in pandas.hashtable.PyObjectHashTable.map_locations (pandas/hashtable.c:11242)
ValueError: Buffer dtype mismatch, expected 'Python object' but got 'double' |
|
what are the columns on the df |
|
The columns are either strings or categoricals - that's the line this fails on. types_should_run = [ tm.makeStringIndex, tm.makeCategoricalIndex ]
...
for index in types_should_run:
df = DataFrame(np.random.randn(10, 2), columns=index(2))
with ensure_clean_path(self.path) as path:
df.to_hdf(path, 'df', format='table', data_columns=True)
result = pd.read_hdf(path, 'df', where="index = [{0}]".format(df.index[0]))
assert(len(result)) |
| @@ -344,7 +344,7 @@ Bug Fixes | |||
| - Bug in ``ExcelReader`` when worksheet is empty (:issue:`6403`) | |||
| - Bug in ``Table.select_column`` where name is not preserved (:issue:`10392`) | |||
| - Bug in ``offsets.generate_range`` where ``start`` and ``end`` have finer precision than ``offset`` (:issue:`9907`) | |||
|
|
|||
| - Bug in ``Subclasses of Index with no values returned Index objects rather than their own classes, in some cases`` (:issue:`10596`) | |||
There was a problem hiding this comment.
use the double-back ticks Index only here (otherwise you are quoting the entire string).
Resolves #10596