Closed
Description
This "worked" before becuase we used pytest.raises(..., msg=), but msg isn't an argument for pytest, so it wasn't validated.
Skipping for now
pytest pandas/tests/io/test_pytables.py -k test_to_hdf_with_object_column_names -x --pdb
==================================================================== test session starts =====================================================================
platform darwin -- Python 2.7.14, pytest-3.5.1, py-1.5.3, pluggy-0.6.0
rootdir: /Users/taugspurger/sandbox/pandas-27, inifile: setup.cfg
plugins: xdist-1.22.2, forked-0.2
collected 177 items / 176 deselected
pandas/tests/io/test_pytables.py F
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> traceback >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
self = <pandas.tests.io.test_pytables.TestHDFStore object at 0x1f1f92bed0>
def test_to_hdf_with_object_column_names(self):
# GH9057
# Writing HDF5 table format should only work for string-like
# column types
types_should_fail = [tm.makeIntIndex, tm.makeFloatIndex,
tm.makeDateIndex, tm.makeTimedeltaIndex,
tm.makePeriodIndex]
types_should_run = [tm.makeStringIndex, tm.makeCategoricalIndex]
if compat.PY3:
types_should_run.append(tm.makeUnicodeIndex)
else:
types_should_fail.append(tm.makeUnicodeIndex)
for index in types_should_fail:
df = DataFrame(np.random.randn(10, 2), columns=index(2))
with ensure_clean_path(self.path) as path:
with catch_warnings(record=True):
with tm.assert_raises_regex(
ValueError, ("cannot have non-object label "
"DataIndexableCol")):
df.to_hdf(path, 'df', format='table',
> data_columns=True)
pandas/tests/io/test_pytables.py:4890:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
pandas/util/testing.py:2393: in __exit__
return self.exception_matches(exc_type, exc_value, trace_back)
pandas/util/testing.py:2429: in exception_matches
raise_with_traceback(e, trace_back)
pandas/tests/io/test_pytables.py:4890: in test_to_hdf_with_object_column_names
data_columns=True)
pandas/core/generic.py:1990: in to_hdf
return pytables.to_hdf(path_or_buf, key, self, **kwargs)
pandas/io/pytables.py:279: in to_hdf
f(store)
pandas/io/pytables.py:273: in <lambda>
f = lambda store: store.put(key, value, **kwargs)
pandas/io/pytables.py:886: in put
self._write_to_group(key, value, append=append, **kwargs)
pandas/io/pytables.py:1363: in _write_to_group
s.write(obj=value, append=append, complib=complib, **kwargs)
pandas/io/pytables.py:3945: in write
self._handle.create_table(self.group, **options)
../../miniconda3/envs/travis-27/lib/python2.7/site-packages/tables/file.py:1055: in create_table
chunkshape=chunkshape, byteorder=byteorder)
../../miniconda3/envs/travis-27/lib/python2.7/site-packages/tables/table.py:773: in __init__
self.description = Description(description)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
self = <[UnicodeEncodeError("'ascii' codec can't encode character u'\u05db' in position 51: ordinal not in range(128)") raised in repr()] SafeRepr object at 0x1f216c58c0>
classdict = {'35כזכקל01ט': Float64Col(shape=(), dflt=0.0, pos=2), 'index': Int64Col(shape=(), dflt=0, pos=0), 'כ9כחגההי86': Float64Col(shape=(), dflt=0.0, pos=1)}
nestedlvl = -1, validate = True
def __init__(self, classdict, nestedlvl=-1, validate=True):
if not classdict:
raise ValueError("cannot create an empty data type")
# Do a shallow copy of classdict just in case this is going to
# be shared by other instances
newdict = self.__dict__
newdict["_v_name"] = "/" # The name for root descriptor
newdict["_v_names"] = []
newdict["_v_dtypes"] = {}
newdict["_v_types"] = {}
newdict["_v_dflts"] = {}
newdict["_v_colobjects"] = {}
newdict["_v_is_nested"] = False
nestedFormats = []
nestedDType = []
if not hasattr(newdict, "_v_nestedlvl"):
newdict["_v_nestedlvl"] = nestedlvl + 1
cols_with_pos = [] # colum (position, name) pairs
cols_no_pos = [] # just column names
# Check for special variables and convert column descriptions
for (name, descr) in six.iteritems(classdict):
if name.startswith('_v_'):
if name in newdict:
# print("Warning!")
# special methods &c: copy to newdict, warn about conflicts
warnings.warn("Can't set attr %r in description class %r"
% (name, self))
else:
# print("Special variable!-->", name, classdict[name])
newdict[name] = descr
continue # This variable is not needed anymore
columns = None
if (type(descr) == type(IsDescription) and
issubclass(descr, IsDescription)):
# print("Nested object (type I)-->", name)
columns = descr().columns
elif (type(descr.__class__) == type(IsDescription) and
issubclass(descr.__class__, IsDescription)):
# print("Nested object (type II)-->", name)
columns = descr.columns
elif isinstance(descr, dict):
# print("Nested object (type III)-->", name)
columns = descr
else:
# print("Nested object (type IV)-->", name)
descr = copy.copy(descr)
# The copies above and below ensure that the structures
# provided by the user will remain unchanged even if we
# tamper with the values of ``_v_pos`` here.
if columns is not None:
descr = Description(copy.copy(columns), self._v_nestedlvl)
classdict[name] = descr
pos = getattr(descr, '_v_pos', None)
if pos is None:
cols_no_pos.append(name)
else:
cols_with_pos.append((pos, name))
# Sort field names:
#
# 1. Fields with explicit positions, according to their
# positions (and their names if coincident).
# 2. Fields with no position, in alfabetical order.
cols_with_pos.sort()
cols_no_pos.sort()
keys = [name for (pos, name) in cols_with_pos] + cols_no_pos
pos = 0
# Get properties for compound types
for k in keys:
if validate:
# Check for key name validity
check_name_validity(k)
# Class variables
object = classdict[k]
newdict[k] = object # To allow natural naming
if not (isinstance(object, Col) or
isinstance(object, Description)):
raise TypeError('Passing an incorrect value to a table column.'
' Expected a Col (or subclass) instance and '
'got: "%s". Please make use of the Col(), or '
'descendant, constructor to properly '
'initialize columns.' % object)
object._v_pos = pos # Set the position of this object
object._v_parent = self # The parent description
pos += 1
newdict['_v_colobjects'][k] = object
newdict['_v_names'].append(k)
object.__dict__['_v_name'] = k
if not isinstance(k, str):
# numpy only accepts "str" for field names
if sys.version_info[0] < 3:
# Python 2.x: unicode --> str
> kk = k.encode() # use the default encoding
E AssertionError: "cannot have non-object label DataIndexableCol" does not match "'ascii' codec can't encode character u'\u05db' in position 0: ordinal not in range(128)"
Working around this for now.