Test case failing - test_timezones.py:TestTimeZones.test_normalize_tz

[ravinimmi]

Code Sample, a copy-pastable example if possible

nosetests pandas/tseries/tests/test_timezones.py:TestTimeZones.test_normalize_tz

F
======================================================================
FAIL: test_normalize_tz (pandas.tseries.tests.test_timezones.TestTimeZones)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/ravikumarnimmi/work/pandas/pandas/tseries/tests/test_timezones.py", line 1398, in test_normalize_tz
    self.assertTrue(result.is_normalized)
AssertionError: False is not true

----------------------------------------------------------------------
Ran 1 test in 0.026s

FAILED (failures=1)

Expected Output

output of pd.show_versions()

INSTALLED VERSIONS
------------------
commit: f7528866e07048f051f46f621ca76947730c7d32
python: 3.5.1.final.0
python-bits: 64
OS: Darwin
OS-release: 15.5.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: en_US.UTF-8
LANG: en_US.UTF-8

pandas: 0.18.1+129.gf752886
nose: 1.3.7
pip: 8.1.2
setuptools: 23.0.0
Cython: 0.24
numpy: 1.11.0
scipy: None
statsmodels: None
xarray: None
IPython: 4.2.0
sphinx: None
patsy: None
dateutil: 2.5.3
pytz: 2016.4
blosc: None
bottleneck: None
tables: None
numexpr: None
matplotlib: None
openpyxl: None
xlrd: None
xlwt: None
xlsxwriter: None
lxml: None
bs4: None
html5lib: None
httplib2: None
apiclient: None
sqlalchemy: None
pymysql: None
psycopg2: None
jinja2: None
boto: None
pandas_datareader: None

[jreback]

this passes on travis & on 15.4. maybe there is something peculiar on your setup. If you can debug great.

[ravinimmi]

@jreback
This code in ipython shell

from pandas.tseries.index import date_range
from dateutil.tz import tzlocal
rng = date_range('1/1/2000 9:30', periods=1, freq='D', tz=tzlocal())
result = rng.normalize()
print(result)
print(result.is_normalized)

gives the following output

DatetimeIndex(['2000-01-01 00:00:00+05:30'], dtype='datetime64[ns, tzlocal()]', freq=None)
False

I printed value of dts after this line which gives

{'year': 1999, 'month': 12, 'day': 31, 'sec': 0, 'as': 0, 'ps': 0, 'hour': 18, 'us': 0, 'min': 30}

My local timezone is 'Asia/Kolkata'

[jreback]

you are referring to a line which is not triggered. This is a local timezone (from whatever tz you are in).

In [1]: from dateutil.tz import tzlocal

In [2]: tzlocal
Out[2]: dateutil.tz.tz.tzlocal

In [3]: tzlocal()
Out[3]: tzlocal()

In [4]: rng = date_range('1/1/2000 9:30', periods=1, freq='D', tz=tzlocal())

In [5]: rng
Out[5]: DatetimeIndex(['2000-01-01 09:30:00-05:00'], dtype='datetime64[ns, tzlocal()]', freq='D')

In [6]: rng.normalize()
Out[6]: DatetimeIndex(['2000-01-01 00:00:00-05:00'], dtype='datetime64[ns, tzlocal()]', freq=None)

In [7]: rng.normalize().is_normalized
Out[7]: True

In [8]: rng = date_range('1/1/2000 9:30', periods=1, freq='D', tz='Asia/Kolkata')

In [9]: rng
Out[9]: DatetimeIndex(['2000-01-01 09:30:00+05:30'], dtype='datetime64[ns, Asia/Kolkata]', freq='D')

In [10]: rng.normalize().is_normalized
Out[10]: True

In [11]: rng.normalize()
Out[11]: DatetimeIndex(['2000-01-01 00:00:00+05:30'], dtype='datetime64[ns, Asia/Kolkata]', freq=None)

[ravinimmi]

@jreback Sorry I meant after this line
Could you please check once by changing your timezone?

import os
import time

from pandas.tseries.index import date_range
from dateutil.tz import tzlocal
os.environ['TZ'] = 'Asia/Kolkata'
time.tzset()
date_range('1/1/2000 9:30', periods=1, freq='D', tz=tzlocal()).normalize().is_normalized

[jreback]

that doesn't make sense
you would have to do that before dateutil imports

[ravinimmi]

yes

import os
import time
os.environ['TZ'] = 'Asia/Kolkata'
time.tzset()

from pandas.tseries.index import date_range
from dateutil.tz import tzlocal
date_range('1/1/2000 9:30', periods=1, freq='D', tz=tzlocal()).normalize().is_normalized

But what does it print?

[jreback]

So it looks normalized, but possibly is_normalized has a bug in it. the tzlocal has a different path (and might not be right).

So use that as a test and see what you can do about fixing.

You will have to capture the current tz and reset it after. We do something like this in io/tests/test_pytables.py (near the end)