Paul Ullrich

Updated. ~ Paul On Aug 31, 2020, at 12:03 PM, Charlie Zender zender@users.sourceforge.net wrote: Please allow read permission on those files ncra on type byte (automatic type promotion) https://sourceforge.net/p/nco/discussion/9829/thread/adc5a6b224/?limit=25#5daa/f058 Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/nco/discussion/9829/ https://sourceforge.net/p/nco/discussion/9829/ To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/...

Sorry posted an example below but should've hit "reply". The issue is that I don't want it to downcast to NC_BYTE after, but rather keep it as a double or float.

On Cori: ncra /global/cscratch1/sd/paullric/MERRA2_blocking/out/1980/*.nc4 MERRA2_blocking_1980.nc4

Thanks to my efforts doing feature tracking I'm sitting with thousands of files that use NC_BYTE to store boolean data at each grid point. Is there an easy way to ncra these files in a manner that automatically promotes them to double before averaging? Otherwise essentially all values get cast down to zero after averaging.

Ah, nces. I hadn't even thought of that (nor the other ensemble commands). Thanks, works great!

Hi Charlie (et al.), So I've got 25 years of daily mean data (365 days per year, one year per file), each with times that go from 0001-01-01 to 0001-12-31 that I'd like to combine to make a single 365 day daily mean climatology (again with times from 0001-01-01 to 0001-12-31). None of the dimensions in the file are currently indicated as record dimensions. Is this doable with NCO? Or would I need to fall back on NCL/Python? ~ Paul