Code Sample, a copy-pastable example if possible
import pandas as pd
dates = pd.date_range('2017-01-01', '2017-01-31', tz='UTC')
deltas = pd.TimedeltaIndex(pd.np.arange(0, len(dates)), unit='m')
dates + deltas
# Removing the tz='UTC' (making dates tz naive) works
throws:
TypeError Traceback (most recent call last)
<ipython-input-28-708df3869f57> in <module>()
3 dates = pd.date_range('2017-01-01', '2017-01-31', tz='UTC')
4 deltas = pd.TimedeltaIndex(pd.np.arange(0, len(dates)), unit='m')
----> 5 dates + deltas
~/miniconda/lib/python3.6/site-packages/pandas/core/indexes/datetimelike.py in __add__(self, other)
643 from pandas.tseries.offsets import DateOffset
644 if isinstance(other, TimedeltaIndex):
--> 645 return self._add_delta(other)
646 elif isinstance(self, TimedeltaIndex) and isinstance(other, Index):
647 if hasattr(other, '_add_delta'):
~/miniconda/lib/python3.6/site-packages/pandas/core/indexes/datetimes.py in _add_delta(self, delta)
814 new_values = self._add_delta_td(delta)
815 elif isinstance(delta, TimedeltaIndex):
--> 816 new_values = self._add_delta_tdi(delta)
817 # update name when delta is Index
818 name = com._maybe_match_name(self, delta)
~/miniconda/lib/python3.6/site-packages/pandas/core/indexes/datetimelike.py in _add_delta_tdi(self, other)
727 mask = (self._isnan) | (other._isnan)
728 new_values[mask] = iNaT
--> 729 return new_values.view(self.dtype)
730
731 def isin(self, values):
TypeError: data type not understood
Problem description
Adding TimedeltaIndex to timezone-aware DatetimeIndex throws a TypeError. The error goes away if the DatetimeIndex does not have a timezone.
(I vaguely remember seeing an issue on this, but can't seem to find it now when I search.)
Output of pd.show_versions()
INSTALLED VERSIONS
------------------
commit: None
python: 3.6.2.final.0
python-bits: 64
OS: Linux
OS-release: 4.10.0-33-generic
machine: x86_64
processor: x86_64
byteorder: little
LC_ALL: C.UTF-8
LANG: C.UTF-8
LOCALE: en_US.UTF-8
pandas: 0.21.0
pytest: None
pip: 9.0.1
setuptools: 36.5.0.post20170921
Cython: None
numpy: 1.13.3
scipy: 0.19.1
pyarrow: None
xarray: None
IPython: 6.2.1
sphinx: None
patsy: None
dateutil: 2.6.1
pytz: 2017.3
blosc: None
bottleneck: None
tables: 3.4.2
numexpr: 2.6.4
feather: None
matplotlib: 2.1.0
openpyxl: None
xlrd: 1.0.0
xlwt: None
xlsxwriter: None
lxml: None
bs4: 4.6.0
html5lib: 0.9999999
sqlalchemy: None
pymysql: None
psycopg2: None
jinja2: 2.10
s3fs: None
fastparquet: None
pandas_gbq: None
pandas_datareader: None
Comment From: jschendel
This should be fixed on master via #18653
In [1]: import pandas as pd
...:
...: dates = pd.date_range('2017-01-01', '2017-01-31', tz='UTC')
...: deltas = pd.TimedeltaIndex(pd.np.arange(0, len(dates)), unit='m')
...: dates + deltas
...:
Out[1]:
DatetimeIndex(['2017-01-01 00:00:00+00:00', '2017-01-02 00:01:00+00:00',
'2017-01-03 00:02:00+00:00', '2017-01-04 00:03:00+00:00',
'2017-01-05 00:04:00+00:00', '2017-01-06 00:05:00+00:00',
'2017-01-07 00:06:00+00:00', '2017-01-08 00:07:00+00:00',
'2017-01-09 00:08:00+00:00', '2017-01-10 00:09:00+00:00',
'2017-01-11 00:10:00+00:00', '2017-01-12 00:11:00+00:00',
'2017-01-13 00:12:00+00:00', '2017-01-14 00:13:00+00:00',
'2017-01-15 00:14:00+00:00', '2017-01-16 00:15:00+00:00',
'2017-01-17 00:16:00+00:00', '2017-01-18 00:17:00+00:00',
'2017-01-19 00:18:00+00:00', '2017-01-20 00:19:00+00:00',
'2017-01-21 00:20:00+00:00', '2017-01-22 00:21:00+00:00',
'2017-01-23 00:22:00+00:00', '2017-01-24 00:23:00+00:00',
'2017-01-25 00:24:00+00:00', '2017-01-26 00:25:00+00:00',
'2017-01-27 00:26:00+00:00', '2017-01-28 00:27:00+00:00',
'2017-01-29 00:28:00+00:00', '2017-01-30 00:29:00+00:00',
'2017-01-31 00:30:00+00:00'],
dtype='datetime64[ns, UTC]', freq='1441T')
In [2]: pd.__version__
Out[2]: '0.22.0.dev0+427.gee9c7e9'
Comment From: jreback
yep
closing as duplicate of https://github.com/pandas-dev/pandas/issues/17558
