[Thread Prev][Thread Next][Index]
Re: Saving 4D data slicewise
Hi Nils,
Easy solution. /APPEND can either extent a file along the time axis or add
new variables, OR ORVER_WRITE EXISTING PARTS OF A FILE. The qualifiers
/ILIMITS, /JLIMITS, etc. are convenient tools in your problem. You
need only (for example):
let var=u*v ! on some grid where u,v are 4 dimensional: u(i,j,k,l)
save/ilimits=1:10 var[i=1]
repeat/i=2:10 (save/file=var.cdf/append var[i=`i`]) ! not /l=`l`
The command "save/ilimits=1:10 var[i=1]" inserts missing data flags into
I=2:10.
- steve
| NOAA/PMEL | ph. (206) 526-6080
Steve Hankin | 7600 Sand Point Way NE | FAX (206) 526-6744
| Seattle, WA 98115-0070 | hankin@pmel.noaa.gov
On Wed, 14 Jan 1998, nils rix wrote:
> Hello Ferreteers,
> I'm trying to do a rather memory intensive calculation and to be
> able to do it at all I would like to save intermediate results slicewise
> to a file.
> Unfortunately the nature of the problem is such that a slice must be defined
> along one of the first two axes (i or j-index say!!!). Does anyone know if
> it is possible to in principle do something like
>
> let var=u*v ! on some grid where u,v are 4 dimensional: u(i,j,k,l)
> repeat/i=1:10 (save/file=var.cdf/append/i=`i` var[i=`i`]) ! not /l=`l`
>
> I know it works for the time axis, but that's not what I need.
> Does netcdf allow to increase any dimension of a variable in an existing
> file?
>
> Cheers Nils
>
> address:
> ---------------------------------------------------------------------------
> Nils H. Rix
> Dept. Theor. Oceanogr. phone: (nat.) 0-431-597-3986
> Inst. f. Meereskunde : (int.) ++49-431-597-3986
> Duesternbrookerweg 20 fax: ++49-431-565-876
> 24105 Kiel, FRG e-mail: nrix@ifm.uni-kiel.de
> ---------------------------------------------------------------------------
>
>
>
[Thread Prev][Thread Next][Index]
Dept of Commerce /
NOAA /
OAR /
PMEL /
TMAP
Contact Us | Privacy Policy | Disclaimer | Accessibility Statement