Due to the current gap in continued funding from the U.S. National Science Foundation (NSF), the NSF Unidata Program Center has temporarily paused most operations. See NSF Unidata Pause in Most Operations for details.
Hi Bill, On Fri, May 14, 2004 at 05:57:11AM -0500, Bill Hibbard wrote: > Hi Russell, > > Thanks for the test program - they always help a lot. > > I can see two problems: > > 1. I called visad.jmet.DumpType.dumpDataType() on the > Field you are displaying and it has 1927 missing values > among its 4402 samples. All these missing points will > cause trouble with contouring. The data was displayed as contours in the source application, so that appears not to be a problem. > 2. More relevant, the Gridded2DSet constructor call: > > Gridded2DSet newFieldDomainSet > new Gridded2DSet(RealTupleType.LatitudeLongitudeTuple, > index2LatLonField.getFloats(false), > index2LatLonField.getFloats(false)[0].length); > > constructs a Gridded2DSet with manifold dimension = 1. > That is, its not really a 2-D domain, but a 1-D domain > that zig-zags through 2-D. That was it. Changing to a 2-D domain with manifold dimension = 2 (as you described in a separate message, quoted below) fixed it, and I may even understand why. :) To get a true 2-D domain, call the Gridded2DSet constructor signature with 4 arguments, the last two being X and Y lengths (their product should be the single length you are passing now). Thank you, Russell -- Russell Steicke -- Fortune says: Yow! Is my fallout shelter termite proof?
visad
archives: