NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
Hi Gunnar, A very hard problem to diagnose based on the two JPEGs. I can think of four possibilities: 1. There is a bug in Irregular3Det.valueToInterp(), in which case we need a lot more information to reproduce it. 2. The difference between our resample methods is in how we handle points that fall on the outside boundary of the convex hull of the irregular points. 3. If you have any missing values, how our resample methods differ in handing that. 4. Perhaps the topology you are passing to DelaunayCustom does not include all the tetrahedra that you are using in your resample method. 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 Wed, 17 Apr 2002, Gunnar Sletta wrote: > > Hello > > I've been having some problems with resampling datasets based on the > Irregular3DSet class. I've created the irregular set from a DelaunayCustom > object which received the tri and elm array as input. The other arrays were > left to the delaunay object to complete. The irregular data was resampled on > a Linear3DSet of size 64x64x64 using weighted average and no error. The > result can be seen in image > http://www.ifi.uio.no/~gunnarsl/stuff/resample.jpg. It looks a bit sparse. > > The image http://www.ifi.uio.no/~gunnarsl/stuff/myresample.jpg is the result > of the same irregular dataset but I used my irregular set resample method, > which uses the tri and elm arrays from the delaunay to resample the field > values directly. The volume is dense. > > Also, if I use the IsoContour display map, the field from VisAD's resample() > method has a lot of holes, while the field from my own method is smooth. > > Has you experienced anything like this, or any clue as to what might be > causing it? > > Thanks > > Gunnar > > > > --------------------------------- > Do You Yahoo!? > Yahoo! Tax Center - online filing with TurboTax
visad archives: