[Thread Prev][Thread Next][Index]

Re: [ferret_users] : wrong region when shading a derived variable



Hi Martin,
Thank you for these reports.

I think this may be the same NetCDF bug reading netCDF4 deflated data from a thredds server that you reported earlier in the week. In that example, http://www.pmel.noaa.gov/maillists/tmap/ferret_users/fu_2012/msg00423.html, the data was returned with some enormous values inserted, but where there were reasonable-looking data, the were shifted relative to the correct values in the netcdf-3 file.

And to update the Ferret List on Martin's earlier report, we've made a report to Unidata and they're looking into it.

Ansley

On 8/17/2012 4:28 AM, Martin Schmidt wrote:
Hi,
with the latest release of ferret (6.8.2) I encounter the following error:

use "http://phy-51.io-warnemuende.de/thredds/dodsC/genus/joined_ocean_day_zoo31_dmv_ascat.nc";

( a data set aggregated over time with THREDDS, fileformat netcdf4, deflation level 3)

shade/t=13-aug-2009:31-dec-2009 nitrogen_in_flagellates[l=@ave,k=1:20@din,x=8:20]

gives a result, that should be correct. Repeating the command,

shade/t=13-aug-2009:31-dec-2009 nitrogen_in_flagellates[l=@ave,k=1:20@din,x=8:20]/100

but now with a scaling factor gives another figure, the output region is now global in zonal direction.

shade/t=13-aug-2009:31-dec-2009/x=8:20 nitrogen_in_flagellates[l=@ave,k=1:20@din]/100

gives a correct figure again.

This seems to be a bug.

Greetings,
Martin Schmidt




[Thread Prev][Thread Next][Index]
Contact Us
Dept of Commerce / NOAA / OAR / PMEL / Ferret

Privacy Policy | Disclaimer | Accessibility Statement