Re: [ferret_users] Problem with excess variables

[Ansley Manke <ansley.b.manke@xxxxxxxx>]

Hi,
No, there is not a way to specify only accessing a subset of the variables in files you open.

Here is one idea: Are the files a set of time steps of the same variables? If so you could create a multi-netcdf file, or descriptor file for them, and then they would appear to Ferret as a single dataset.  Please see "descriptor file" in the Users Guide for more on this and instructions on creating a descriptor file.

If the data aren't the time steps for the variables, write back and we'll see if we can come up with another idea.

Ansley

Samrat Rao wrote:

Hello,
       I am trying to execute the following commands via a GO file:

use dset1.nc,dset2.nc,......,dset20.nc

let var1 = var_a[d=dset1.nc] + var_a[d=dset2.nc] + ..... +var_a[d=dset20.nc]
let var2 = var_b[d=dset1.nc] + var_b[d=dset2.nc] + ..... +var_b[d=dset20.nc]

fill/l=1 var1 + var2

When i try run the GO file, i get the error:
   **TMAP ERR: limit on number of variables has been reached
             MAX=        2000
             Data set: ./dset20.nc

The GO file at this stage has not yet reached the line 'let var1 = .....'

So what do i do?

Is it possible that for netCDF files one can specify the variables that need to be used, thus limiting the problem of excess variables?

Any help is appreciated.