Re: LinearLatLonSet or other?

• To: Ben Podoll <benjamin.podoll@xxxxxxxxxxxxx>
• Subject: Re: LinearLatLonSet or other?
• From: Bill Hibbard <billh@xxxxxxxxxxxxx>
• Date: Mon, 17 Feb 2003 03:08:38 -0600 (CST)

Hi Ben,

If you have explicit latitudes and longitudes, use them to
construct a Gridded2DSet (if they are organized in a
topologically rectangular grid - which may be a curved grid),
or an Irregular2DSet (if they are not organized in a
topologically rectangular grid).

You dont have to use LinearLatLonSet for latitudes and
longitudes. It only exists to allow interpolation between
opposite edges of a linear lat-lon grid that meet (that
is, a grid that fully circles the globe).

Good luck,
Bill

On Sun, 16 Feb 2003, Ben Podoll wrote:

> If I have a file that contains a bunch of lat/lon points that record
> where samples where taken and in separate files are the respective
> samples.
> (For this question assume everything is 2D)
>
> I initially used VisAD to do something like this:
>
> ----------------------------
>
> RealyType latitude,longitude,variable;
> domain_tuple = new RealTupleType(latitude, longitude);
>
> domain_set = new LinearLatLonSet
>       (domain_tuple, latMin, latMax, Ny, lonMin, lonMax, Nx);
>
> float[][] flat_samples = <read in my samples>
> func_domain_range = new FunctionType(domain_tuple, variable);
> vals_ff = new FlatField(func_domain_range, domain_set);
> vals_ff.setSamples(flat_samples, false);
>
> ---------------------------
>
> My question is should I be using something other than LinearLatLonSet
> and actually read in all my points instead in an float[1][Nx*Ny]?
> (To be more accurate?)
>
> Thanks,
> Ben
>
>
>

• Follow-Ups:
  • ProjectionControlJ3D and Zoom
    • From: Ben Podoll

• References:
  • LinearLatLonSet or other?
    • From: Ben Podoll