Comparison between pytz and pandas timezones fails

Question

I want to compare the timezone of a pd.DatetimeIndex with a pytz.timezone to see if the DatetimeIndex has the expected timezone. But the comparison fails, possibly because using the tzinfo argument does not work as explained in this answer.

import pandas as pd
import pytz
import unittest

tzstr = 'Europe/Vienna'
manual_tz = pytz.timezone(tzstr)

timestamp = pd.to_datetime('2020-03-02 07:00:00+01:00').tz_convert(tzstr)

tc = unittest.TestCase()
tc.assertEqual(timestamp.tz, manual_tz)

AssertionError: <DstTzInfo 'Europe/Vienna' CET+1:00:00 STD> != <DstTzInfo 'Europe/Vienna' LMT+1:05:00 STD>

How can I check that the timezone of timestamp is as expected? It should be the same as pytz.timezone in my opinion, but the behavior of pytz and/or pandas makes them somehow different.

Alternative question formulation if it is not possible to make a comparison like this: What do I have to look out for, to not come across this problem any more? It did happen to me more than once already. I found a question similar to this, but having to use multiple dates to compare if the offset is always the same does not seem like the best way to.

Answer 1

I think your matter is linked to this answer

The default zone name and offset delivered when pytz creates a timezone object are the earliest ones available for that zone, and sometimes they can seem kind of strange.

This is what I have when I run your code:

AssertionError: <DstTzInfo 'Europe/Vienna' CET+1:00:00 STD> != <DstTzInfo 'Europe/Vienna' LMT+1:05:00 STD>

What you can do is this:

import pandas as pd
import pytz
import unittest
import datetime

tzstr = 'Europe/Vienna'
manual_tz = pytz.timezone(tzstr)
pytz_localize = manual_tz.localize(datetime.datetime(2020, 3, 2, 7, 0, 0, 0))


timestamp = pd.to_datetime('2020-03-02 07:00:00+01:00').tz_convert(tzstr)

tc = unittest.TestCase()
tc.assertEqual(timestamp.tz, pytz_localize.tzinfo)

But IHMO, it's weird to have to create two dates for comparing the timezone information.

What do you mean by compare the timezone? What do you want to check?

Edit
About testing timezone and conversions. What I would do is find test cases you want to check. Something I would check is that I correctly handle the DST change, for this I would do something like this:

import pandas as pd

# DST in Austria in 2021.
all_dates = ['2021-03-28 01:59:00+0100', '2021-03-28 03:01:00+0200']
utc_dates = pd.to_datetime(['2021-03-28 00:59:00+00:00', '2021-03-28 01:01:00+00:00']).tz_localize('UTC')

timezoned_timestamps = pd.to_datetime(all_dates, utc=True).tz_convert("Europe/Vienna")

# then we check if it's equal after UTC conversion
pd.testing.assert_index_equal(utc_dates, timezoned_timestamps.tz_convert('UTC'))

The parameter utc=True in the first to_datetime call is mandatory:

However, timezone-aware inputs with mixed time offsets (for example issued from a timezone with daylight savings, such as Europe/Paris) are not successfully converted to a DatetimeIndex. Instead a simple Index containing datetime.datetime objects is returned

From the documentation.