drop keyword in ds.rolling(time=7, drop=True).mean()?

[ahuang11]

Should rolling have a drop keyword, similar to squeeze(drop=True)?

import xarray as xr
air = xr.tutorial.open_dataset('air_temperature')
air = air.rolling(time=7, drop=True).mean()

Equivalent:

import xarray as xr
air = xr.tutorial.open_dataset('air_temperature')
air = air.rolling(time=7).mean()
air = air.isel(time=slice(6, None))

Actual implementation will require considering min_period / center=True too

[mathause]

This is not something I would expect or miss so knowing your use-case would be helpful. Just as a word of caution, the implementation of rolling is a bit convoluted, so this may take a bit of effort.

[dcherian]

Do you want to avoid the padding?

[ahuang11]

Yes I want to drop the padding (remove the nans from the beginning)

…

On Thu, Feb 18, 2021, 3:44 PM Deepak Cherian ***@***.***> wrote: Do you want to avoid the padding? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#4588 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADU7FFVBUJGELDFGWQ6TARDS7WC23ANCNFSM4TX2CAEQ> .

[dcherian]

Thanks. Closing as duplicate of #2007. I have a solution in progress.