I would like to save a df with the following index to an Excel file:

```dtype='datetime64[ns, Atlantic/Reykjavik]', name=u'Dates', length=4360, freq='H')


it shows the following error:

TypeError: Timestamp subtraction must have the same timezones or no timezones

Comment From: jorisvandenbossche

@dacoex Thanks for the report! I suppose this is a duplicate of https://github.com/pydata/pandas/issues/7056

This is not really implemented at the moment (excel also does not support timezones, so an implementation would something be like a removal of the timezone + warning). For now, you can convert it to naive timestamps yourself before exporting to excel.

Comment From: dacoex

Just as reference for anyone: use

df.tz_localize(None)

http://pandas-docs.github.io/pandas-docs-travis/timeseries.html#time-zone-handling