Pandas timezone - how to check if it is not defined for a whole column? - pandas

Question

To check if a timezone is not defined for the first row of a "timestamp" column in a pandas Series I can query .tz for a single element with:
import pandas as pd
dates = pd.Series(pd.date_range('2/2/2002', periods=10, freq='M'))
assert dates.iloc[0].tz is None
Do I have a way to check if there are elements where the timezone is defined, or even better, a way to list all the timezones in the whole series, without looping through its elements, such as:
dates.iloc[5] = dates.iloc[5].tz_localize('Africa/Abidjan')
dates.iloc[7] = dates.iloc[7].tz_localize('Africa/Banjul')
zones = []
for k in range(dates.shape[0]):
zones.append(dates.iloc[k].tz)
print(set(zones))
?

Answer 1

You can get the time zone setting of a datetime Series using the dt accessor, i.e. S.dt.tz. This will raise ValueError if you have multiple time zones since the datetime objects will then be stored in an object array, as opposed to a datetime64 array if you have only one time zone or None. You can make use of this to get a solution that is a bit more efficient than looping every time:
import pandas as pd
# tzinfo is None:
dates0 = pd.Series(pd.date_range('2/2/2002', periods=10, freq='M'))
# one timezone:
dates1 = dates0.dt.tz_localize('Africa/Abidjan')
# mixed timezones:
dates2 = dates0.copy()
dates2.iloc[5] = dates2.iloc[5].tz_localize('Africa/Abidjan')
dates2.iloc[7] = dates2.iloc[7].tz_localize('Africa/Banjul')
for ds in [dates0, dates1, dates2]:
try:
zones = ds.dt.tz
except ValueError:
zones = set(t.tz for t in ds.values)
print(zones)
# prints
None
Africa/Abidjan
{None, <DstTzInfo 'Africa/Banjul' GMT0:00:00 STD>, <DstTzInfo 'Africa/Abidjan' GMT0:00:00 STD>}