Dear Veera,
the problem that the ncdx-modified file does not work has a very simple
reason. In your original file, the so-called 'coordinate variables' are
missing! Coordinate variables are netCDF variables of dimension 1,
with the same name as the corresponding dimension.
For instance:
dimensions:
tlat = 118 ;
variables:
float tlat(tlat) ;
tlat:longname = ""
...
data:
tlat = .,.,., ;
etc.
If you have then e.g. the variable
long landt(tlat, tlon) ;
landt:long_name = "land/sea mask at tracer points" ;
landt:options = "0: land, 1: sea" ;
landt:field = "Tmask, scalar" ;
landt:positions = "tgrid2" ;
the underlying coordinates are 'tlat' and 'tlon'.
However, there are no coordinate values specified for those ...
ncdx then adds correctly the (OpenDX specific) attributes
landt:field = "landt, scalar" ;
landt:positions = "tlat_002_002, product;tlon_002_001,
product;";
... assuming that the respective coordinate variables (netCDF variables 'tlat'
and 'tlon') are present in the netCDF file.
Since this is not the case, ncdx cannot write the
additional (OpenDX specific) arrays 'tlat_00???' ... which are used for the
positions (in this case a DX-product-array).
So, if you add the required coordinate variables to your netCDF file,
ncdx should work ...
Hope this helps!
Yours,
Patrick Jöckel
On Wednesday 02 April 2003 16:59, Rachel V. Pollard wrote:
> i am trying to make netcdf files that i have for another visualization
> system
> work in opendx. i don't seem to be understanding the positions aspect
> in terms of needing these extra attributes. i am using ncdx to help make my
> files opendx compliant, but none seem to work so far in opendx.
>
> i have read the user's guide and other letters, but still am not sure of
> what to do.
> i have run my original file through opendx, the file after ncdx and a
> version that
> i have modified according to what i thought should be the positions--none
> of these three
> work.
>
> the various errors are: connection to server broken, signal 11 caught and
> the last
> one (from the modified version) is "ERROR: Import: Internal error: netCDF
> library errror getting origins"
>
> it seems the problem is in the postions components somewhere. Notice how
> ncdx put new variables and
> calls it a product--this new variable is not below in data so this will
> not work.
>
> My colleague suggested that the data is regular, thus i put this in my mod
> version. But if it were
> regular i am assuming it would have worked without any positions
> attributes added.
>
> Does anyone have any thoughts on this? Or pointers to relevant info on
> the subject?
>
> Here are the files: http://puma.wellesley.edu/~rpollard
> Original file: has _orig in it
> Post ncdx file: has _ncdx
> My modifications: has _mod
>
> Thanks in advance for any thoughts,
> veera&())
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> http://puma.wellesley.edu/~rpollard/
--
---------------------------------------------------------------
Dr. Patrick Joeckel
Max Planck Institute for Chemistry
Joh.-J. Becher Weg 27
55128 Mainz phone:++49-6131-305452
Germany fax :++49-6131-305436
e-mail:[EMAIL PROTECTED]
http://www.mpch-mainz.mpg.de/~joeckel
---------------------------------------------------------------
Never trust an operating system you don't have sources for.
