BUG: Appending Series with dateutil timezone raises TypeError: unhashable type: 'tzutc'

[semigodking]

Code Sample, a copy-pastable example if possible

# Your code here
import pandas as pd
from dateutil.parser import parse as dt_parse

date = dt_parse('2018-10-24 07:30:00+00:00')
ser = pd.Series({'date': date})
ser2 = pd.Series({'date': date, 'a': 1.0, 'b': 2.0})
df = pd.DataFrame(columns=['c', 'd'])
# No exception
df2 = df.append(ser, ignore_index=True)
# Exception raised from next line
df2 = df.append(ser2, ignore_index=True)

Problem description

The last line of code above causes an exception.

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-5-c167f10681a7> in <module>()
      1 ser2 = pd.Series({'date': date})
----> 2 df2 = df.append(ser, ignore_index=True)

~/vp/venv/lib/python3.5/site-packages/pandas/core/frame.py in append(self, other, ignore_index, verify_integrity)
   5155                               index=index,
   5156                               columns=combined_columns)
-> 5157             other = other._convert(datetime=True, timedelta=True)
   5158             if not self.columns.equals(combined_columns):
   5159                 self = self.reindex(columns=combined_columns)

~/vp/venv/lib/python3.5/site-packages/pandas/core/generic.py in _convert(self, datetime, numeric, timedelta, coerce, copy)
   4064             self._data.convert(datetime=datetime, numeric=numeric,
   4065                                timedelta=timedelta, coerce=coerce,
-> 4066                                copy=copy)).__finalize__(self)
   4067 
   4068     # TODO: Remove in 0.18 or 2017, which ever is sooner

~/vp/venv/lib/python3.5/site-packages/pandas/core/internals.py in convert(self, **kwargs)
   3458 
   3459     def convert(self, **kwargs):
-> 3460         return self.apply('convert', **kwargs)
   3461 
   3462     def replace(self, **kwargs):

~/vp/venv/lib/python3.5/site-packages/pandas/core/internals.py in apply(self, f, axes, filter, do_integrity_check, consolidate, **kwargs)
   3322 
   3323             kwargs['mgr'] = self
-> 3324             applied = getattr(b, f)(**kwargs)
   3325             result_blocks = _extend_blocks(applied, result_blocks)
   3326 

~/vp/venv/lib/python3.5/site-packages/pandas/core/internals.py in convert(self, *args, **kwargs)
   2128 
   2129         if by_item and not self._is_single_block:
-> 2130             blocks = self.split_and_operate(None, f, False)
   2131         else:
   2132             values = f(None, self.values.ravel(), None)

~/vp/venv/lib/python3.5/site-packages/pandas/core/internals.py in split_and_operate(self, mask, f, inplace)
    476             # need a new block
    477             if m.any():
--> 478                 nv = f(m, v, i)
    479             else:
    480                 nv = v if inplace else v.copy()

~/vp/venv/lib/python3.5/site-packages/pandas/core/internals.py in f(m, v, i)
   2118         def f(m, v, i):
   2119             shape = v.shape
-> 2120             values = fn(v.ravel(), **fn_kwargs)
   2121             try:
   2122                 values = values.reshape(shape)

~/vp/venv/lib/python3.5/site-packages/pandas/core/dtypes/cast.py in soft_convert_objects(values, datetime, numeric, timedelta, coerce, copy)
    805     # Soft conversions
    806     if datetime:
--> 807         values = lib.maybe_convert_objects(values, convert_datetime=datetime)
    808 
    809     if timedelta and is_object_dtype(values.dtype):

pandas/_libs/src/inference.pyx in pandas._libs.lib.maybe_convert_objects()

TypeError: unhashable type: 'tzutc'

Expected Output

No exception.

Output of pd.show_versions()

[paste the output of pd.show_versions() here below this line]
INSTALLED VERSIONS

commit: None
python: 3.5.2.final.0
python-bits: 64
OS: Linux
OS-release: 4.4.0-130-generic
machine: x86_64
processor: x86_64
byteorder: little
LC_ALL: None
LANG: en_HK.UTF-8
LOCALE: en_HK.UTF-8

pandas: 0.21.0
pytest: None
pip: 9.0.1
setuptools: 36.2.7
Cython: None
numpy: 1.13.3
scipy: None
pyarrow: None
xarray: None
IPython: 6.2.1
sphinx: None
patsy: None
dateutil: 2.6.1
pytz: 2017.2
blosc: None
bottleneck: None
tables: None
numexpr: None
feather: None
matplotlib: 2.2.2
openpyxl: None
xlrd: None
xlwt: None
xlsxwriter: None
lxml: None
bs4: None
html5lib: 1.0b10
sqlalchemy: 1.1.13
pymysql: None
psycopg2: 2.7.1 (dt dec pq3 ext lo64)
jinja2: 2.9.6
s3fs: None
fastparquet: None
pandas_gbq: None
pandas_datareader: None

[jbrockmendel]

The traceback references a file that hasn’t existed for a while. Can you try this on master?

[mroeschke]

I'm getting the same error on master but different traceback. I had posted an issue about hashability of dateutil timezones, but it appears there's good reason why they aren't yet. dateutil/dateutil#792

In [6]: pd.__version__
Out[6]: '0.24.0.dev0+997.ga197837c4'

In [7]: df2 = df.append(ser2, ignore_index=True)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-7-d913f5e96179> in <module>()
----> 1 df2 = df.append(ser2, ignore_index=True)

~/pandas-mroeschke/pandas/core/frame.py in append(self, other, ignore_index, verify_integrity, sort)
   6497                               index=index,
   6498                               columns=combined_columns)
-> 6499             other = other._convert(datetime=True, timedelta=True)
   6500             if not self.columns.equals(combined_columns):
   6501                 self = self.reindex(columns=combined_columns)

~/pandas-mroeschke/pandas/core/generic.py in _convert(self, datetime, numeric, timedelta, coerce, copy)
   5509             self._data.convert(datetime=datetime, numeric=numeric,
   5510                                timedelta=timedelta, coerce=coerce,
-> 5511                                copy=copy)).__finalize__(self)
   5512
   5513     def convert_objects(self, convert_dates=True, convert_numeric=False,

~/pandas-mroeschke/pandas/core/internals/managers.py in convert(self, **kwargs)
    535
    536     def convert(self, **kwargs):
--> 537         return self.apply('convert', **kwargs)
    538
    539     def replace(self, **kwargs):

~/pandas-mroeschke/pandas/core/internals/managers.py in apply(self, f, axes, filter, do_integrity_check, consolidate, **kwargs)
    403                                             copy=align_copy)
    404
--> 405             applied = getattr(b, f)(**kwargs)
    406             result_blocks = _extend_blocks(applied, result_blocks)
    407

~/pandas-mroeschke/pandas/core/internals/blocks.py in convert(self, *args, **kwargs)
   2350
   2351         if by_item and not self._is_single_block:
-> 2352             blocks = self.split_and_operate(None, f, False)
   2353         else:
   2354             values = f(None, self.values.ravel(), None)

~/pandas-mroeschke/pandas/core/internals/blocks.py in split_and_operate(self, mask, f, inplace)
    500             # need a new block
    501             if m.any():
--> 502                 nv = f(m, v, i)
    503             else:
    504                 nv = v if inplace else v.copy()

~/pandas-mroeschke/pandas/core/internals/blocks.py in f(m, v, i)
   2340         def f(m, v, i):
   2341             shape = v.shape
-> 2342             values = fn(v.ravel(), **fn_kwargs)
   2343             try:
   2344                 values = values.reshape(shape)

~/pandas-mroeschke/pandas/core/dtypes/cast.py in soft_convert_objects(values, datetime, numeric, timedelta, coerce, copy)
    854         try:
    855             values = lib.maybe_convert_objects(values,
--> 856                                                convert_datetime=datetime)
    857         except OutOfBoundsDatetime:
    858             pass

~/pandas-mroeschke/pandas/_libs/lib.pyx in pandas._libs.lib.maybe_convert_objects()
   2046     # we try to coerce datetime w/tz but must all have the same tz
   2047     if seen.datetimetz_:
-> 2048         if len({getattr(val, 'tzinfo', None) for val in objects}) == 1:
   2049             from pandas import DatetimeIndex
   2050             return DatetimeIndex(objects)

TypeError: unhashable type: 'tzutc'