I don't know much about netcdf, but I'd guess that you are creating a field
containing the reallocations and a series group of ten fields containing
the pressure. Could you send the output of Print(..., "rd") on the
Imported object?
Greg
Matthew Bettencourt
<[EMAIL PROTECTED]> To:
[email protected]
Sent by: cc:
[EMAIL PROTECTED] Subject: Re: [opendx-users]
netcdf and doubles
son.ibm.com
08/05/2002 01:23 PM
Please respond to
opendx-users
I am trying to do that. I have created a field callled reallocations
dimensions:
time = UNLIMITED ; // (10 currently)
nx = 8 ;
ny = 5 ;
start = 2 ;
delta = 2 ;
variables:
double reallocations(ny, nx, start) ;
reallocations:field = "real_positions, vector" ;
double pressure(time, ny, nx) ;
pressure:field = "Pressure, scalar, series" ;
data:
....
I then can import this and select either one to see. Now, I can use
compute on real_positions to map the values of that field into a float
float(a)
then I can replace the positions of real_positions with the data and
that works right (I get a mapped grid). However, if I try to replace
the positions of pressure with the data of real_positions that blows up
with the error.
bad object: object hierarchies don't match
ANy idea on what I am doing wrong?? I am really trying to create a
restart file that I can viz with DX and the model requires doubles for
its grid and I don't wanna demote them for obvious reasons.
Thanks
Matt
P.S, for the replace command I am using as my src field real_positions,
pressure for my dest field. srcname is data, destname is positions.
Thanks
David Thompson wrote:
> You'd have to look and see if they are even getting imported as a
> component to the field. Use Print to do this. If they are getting
> imported, then you can convert them to float using the Compute module
> and put them into the positions component using Mark.
>
> David
>
>> Is there any way to convert those once in DX??
>> Matt
>>
>> David Thompson wrote:
>>
>>> Matt,
>>>
>>> OpenDX only supports "float" positions.
>>>
>>> David
>>>
>>>> Is there a way to have positions as doubles in a netcdf file?? I
>>>> can't seem to figure this out?? If I do
>>>>
>>>> variables:
>>>> double locations(start, delta) ;
>>>> float pressure(time, ny, nx) ;
>>>> pressure:field = "Pressure, scalar, series" ;
>>>> pressure:positions = "locations, regular" ;
>>>>
>>>> I get no plot when I import the data, however, if I use
>>>>
>>>> variables:
>>>> float locations(start, delta) ;
>>>> float pressure(time, ny, nx) ;
>>>> pressure:field = "Pressure, scalar, series" ;
>>>> pressure:positions = "locations, regular" ;
>>>>
>>>> it is fine?? This seems goofy?? Am I doing something wrong?
>>>> Thanks
>>>> Matt
>>>
>
>
