Re: MetApps ShapefileAdapter

Hi James,

There is 'invisible' logic inside VisAD that is supposed
to detect longitude seems and even irregular seems in
arbitrary map projections, and reorganize sampling topologies
to correctly display data. It works in many cases, but also
still has some mysterious failures. My advice is to try
doing what you want in a straightforward way, and let us
know if you run into problems.

Cheers,
Bill
----------------------------------------------------------
Bill Hibbard, SSEC, 1225 W. Dayton St., Madison, WI  53706
hibbard@xxxxxxxxxxxxxxxxx  608-263-4427  fax: 608-263-6738
http://www.ssec.wisc.edu/~billh/vis.html


On Sun, 26 May 2002, James Kelly wrote:

> Hi Don, Bill et al,
>
> On the topic of projections/coordinate systems, I have a question (or
> two!) about a coordinate system we would like to use.
>
> The problem is basically how to handle the "dateline problem".
>
> Currently we have an application which represents the entire world by
> longitudes 0 to 360, using setRange(0.0, 360.0) on the scalar map of
> Longitude. This is exactly the range that we want, but do not want to
> use longitudes of 180 to 360 to represent the western hemishphere.
>
> So effectively we want a range which goes from 0 to 180, then from -180
> to 0 on one continuous display (ie dateline in the middle of the
> display). So both the data and the display (ie labels) reflect this.
>
> My impressions from the current discussion are that to achieve this we
> would need:
> 1) to define a coordinate system which mapped the native data of (0 to
> 180; and -180 to 0) onto the (0 to 360) used by the display
>
> 2) and define a display side CS consistent with this
>
> 3) maybe we might need to fiddle with other display characteristics to
> ensure labels were consistent?
>
> Am I on the right track? If so, I'd be grateful for an indication of how
> much effort this might involve (ie day(s), week(s) or month(s)).
>
> Thanks,
>
> James
>
>
> Don Murray wrote:
> >
> > Luke-
> >
> > "Luke A. Catania (TEC)" wrote:
> > >
> > > Is there a way to use a map projection with VisAD?   If so  there example
> > > code?? After using the DisplayImplJ3D, I was able to get the map of the
> > > world shape file displayed, but it does not look correct because of the 
> > > lack
> > > of a map projection.
> >
> > As Bill mentioned, we use a Display side CS in MetApps.
> > See ucar.unidata.view.geoloc.MapProjectionDisplay.  This
> > will take either a visad.georef.MapProjection or a
> > ucar.unidata.geoloc.Projection as input to set the background
> > projection.  For a version of MapProjectionDisplay that
> > uses vanilla VisAD (i.e. no Unidata classes), see:
> >
> > ftp://ftp.unidata.ucar.edu/pub/dmurray/MapProjectionDisplay.java
> >
> > This does not allow you to change projections, though whereas
> > ucar.unidata.view.geoloc.MapProjectionDisplay does allow changes
> > by creating a new DisplayRealTupleType with each change.
> >
> > Don
> > *************************************************************
> > Don Murray                               UCAR Unidata Program
> > dmurray@xxxxxxxxxxxxxxxx                        P.O. Box 3000
> > (303) 497-8628                              Boulder, CO 80307
> > http://www.unidata.ucar.edu/staff/donm
> > *************************************************************
>
> --
> James Kelly                                          Regional Computing
> Phone: 61-3-9669-4724 Fax: 61-3-9669-4128     Email: J.Kelly@xxxxxxxxxx
>