Re: combining data

From: Sylvia Murphy (murphys AT cgd.ucar.edu)
Date: Fri Oct 08 2004 - 14:55:02 MDT


jim,

how are your files structured before you bring them into NCL.

if they are slice in time then I would use the netCDF ncrcat operator to
concatinate in time. If they are sliced in space or level, there is not a
very efficient way of combining them.

sm

On Fri, 8 Oct 2004, Jim Edwards wrote:

> Hi,
>
> I have a number of netcdf files (one per processor of a parallel job)
> that I need to combine
> to visualize. This has proven to be a painfully slow process and I
> wonder if any of you have any suggestions on speeding it up.
>
> Currently I am using the following ncl:
>
>
> fnames = systemfunc("ls movies/*.nc")
> f = addfiles(fnames,"r") ; Open netCDF files.
> fcnt = dimsizes(fnames)
>
> vcnt=new((/fcnt,3/),"integer")
> do i=0,fcnt-1
> vcnt(i,:) = dimsizes(f[i]->Q)
> end do
> ntimes = vcnt(0,0)
> nlev = vcnt(0,1)
> nxy = dim_sum(vcnt(:,2))
>
> var = new((/ntimes,nlev,nxy/),"double")
> do n=1,ntimes
> do k=1,nlev
> var(n,k,:)=f[:]->Q(n,k,:)
> end do
> end do
>
>
> --
> Jim Edwards jedwards@ucar.edu
> IBM Applications Analyst
> NCAR SCD
> BOULDER CO 303-497-1842
>
>
> _______________________________________________
> ncl-talk mailing list
> ncl-talk@ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>

******************************************************
Sylvia Murphy EML: murphys@ucar.edu
NCAR CGD/CSEG PHN: 303-497-1720
1850 Table Mesa Drive FAX: 303-497-1333
Boulder CO 80305

WEB: http://www.cgd.ucar.edu/csm/support/
     http://www.cgd.ucar.edu/csm/support/CSM_Graphics/
******************************************************

_______________________________________________
ncl-talk mailing list
ncl-talk@ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk



This archive was generated by hypermail 2b29 : Tue Oct 12 2004 - 17:06:33 MDT