Closed lorin closed 6 years ago
Work around
>>> f = date_format("%H")
>>> f.formatter.tz = ts_pst.tz
>>> f([ts_pst])
['15']
The issue is, because we use Matplotlib internals the most sane solution would be to use the timezone of the first date. So if you pass the format functions times with different dates the result may be surprising.
Seems reasonable, and it would cover my use case.
If you were concerned about a user submitting different timezones, you could do a check that they are all the same and emit a warning if they aren't.
date_format doesn't respect the time zone of a
pandas.Timestamp
: