Resampling large gaps should result in NaNs
Closed this issue · 4 comments
fpavogt commented
Describe the change
When resampling profiles, large gaps should result in NaNs. We absolutely do not want to interpolate over large distances.
The maximum time gaps should be set to: 2s ?
@GonzagueRomanens, @g-martucci, @cfelix-ch: do you think this number is reasonable ?
GonzagueRomanens commented
@fpavogt 2s looks reasonable to me.
fpavogt commented
This issue was opened in the wrong repo.
g-martucci commented
@fpavogt<https://github.com/fpavogt> 2-second gap to be linearly interpolated makes sense and is safe. For temperature profiles even 5 seconds could be reasonable as the profile is largely linear, but for humidity 2 seconds should be the maximum size.
From: Gonzague Romanens ***@***.***>
Sent: Tuesday, 12 July, 2022 10:04
To: MeteoSwiss/ampycloud ***@***.***>
Cc: Martucci Giovanni ***@***.***>; Mention ***@***.***>
Subject: Re: [MeteoSwiss/ampycloud] Resampling large gaps should result in NaNs (Issue #81)
@fpavogt<https://github.com/fpavogt> 2s looks reasonable to me.
—
Reply to this email directly, view it on GitHub<#81>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/APP66D5XP6QFZ4OUXEDFQMTVTURHTANCNFSM53KCBXDA>.
You are receiving this because you were mentioned.[https://github.com/notifications/beacon/APP66D6PCW5PKKPWLNYIMJTVTURHTA5CNFSM53KCBXDKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOIZVXYZA.gif]Message ID: ***@***.******@***.***>>