Hi David,
Thanks for the report. I do see what look to be correct time bounds
when I try this, but the latitude bounds are just missing values. Do a
sp ncdump sst.nc | more
and look on down the file to see the bounds variables. I'm looking at
another issue with writing out bounds on files right now, so I'll check
into this as well. It looks like we need to add attributes on the
coordinate variables, and obviously get the correct values written out!
Ansley
David Wang wrote:
Hi Ferreters,
I lost all the bounds attributes in the output netcdf file with the
following sample script. It's not a big issue but I'm simply curious
because the manual it's said "all irregular axes are saved with a
bounds attribute" and both Y and T axes are irregular. This is in
Ferret v6.1 on a 32-bit Linux box.
Thanks,
David
\can mode verify
set mem/size=512
can data/a
can region
use "http://esgcet.llnl.gov/cgi-bin/dap-cgi.py/ipcc4/20c3m/gfdl_cm2_0/pcmdi.ipcc4.gfdl_cm2_0.20c3m.run2.ocn.mo.xml"
list/nohead
(`tos,return=xaxis`).bounds
list/nohead
(`tos,return=yaxis`).bounds
list/nohead
(`tos,return=taxis`).bounds
let sst =
tos[x=120:280,y=20s:20n,l=1:12]-273.15
save/clobber/file=sst.nc sst
sp ncdump -h sst.nc
set mode/last verify
|