[Thread Prev][Thread Next][Index]

Re: [ferret_users]:capricious reshape function



Thank you Ansley and Dr J Kaduk

I still don't understand exactly what was my problem but i solved it reducing the number of file. I check all them one by one and they are all OK. But removing few first file it work. May be a problem with the descriptor , while it surprising that for 2 variables it works, and all variable have the same time axis.

What i wanted to do with the climatological axes, is:
let Vbar=var3[gt=month_reg@asn]
Do i really need to cancel climatological_axes.cdf before use it or is it just more convenient?

Thanks a lot
Xavier

Ansley Manke a écrit :

Hi Xavier,
If you are just defining a new time axis with the same number of points as the existing axis, you can use a regridding operation to put your data onto that time axis. The @ASN regridding operation just assigns the data to the new axis.

  define AXIS/T/T0="01-jan-0000"/unit="days" taxis=a
  let Var3=vbar[gt=taxis@ASN]

I'm not sure what you are doing with the climatological_axes, but generally when you are going to use it for a regridding operation, you want to make sure to CANCEL that data set after opening it. All you are doing is defining the axes it contains so you can use them later in your script. That could possibly be why you are getting a "Variable not found" error, if you have not canceled the climatological_axes.cdf dataset and it is defined as the default data set. Though here, you seem to be specifying [d=1] everywhere so that may not be what's happening.


Xavier Couvelard wrote:

Hi all
I'am regriding my model output to obtain climatological axes
i have a lots of file that leaded me to use a descriptor file .
and i also use the climatological axes

SET MEM/SIZE=500

set data avg.0.mc

use climatological_axes.cdf

i first define a new time axes and reshape my variable
Define AXIS/T/T0="01-jan-0000"/unit="days" taxis=a

let Var3=reshape(vbar[d=1],t[gt=taxis]+x[gx=vbar,d=1]+y[gy=vbar,d=1])
.............
.............
So, while it works very well with my two first variable (zeta and ubar) with vbar (2 and the two others variable) it don't

for exemple
sha var3[l=1]
** netCDF error: Variable not found
while sha var2[l=1] who is ubar make a good plot and vbar[d=1,l=1] also make a good plot.

is someone has an idea?

Xavier



--
*********************************
Xavier Couvelard
Phd Student (LEGOS)
IRD Noumea
BPA05 noumea cedex
New Caledonia
(687) 260728
*********************************



[Thread Prev][Thread Next][Index]

Contact Us
Dept of Commerce / NOAA / OAR / PMEL / TMAP

Privacy Policy | Disclaimer | Accessibility Statement