[vtkusers] vtkThreshold question

[Randall Hand]

Well, I'll see if I can find a dataset to replicate the problem and drop it
to you guys.

In the meantime, just some more info:
1) The stack trace contained solely the vtkDataArray::Get line, no further
information... Guess Debug flags are disabled somewhere...
2) The code with a vtkPolyData:

    vtkCellArray *cells = vtkCellArray::New();
    cells->Allocate(nCells, 10);
    vtkIdList *ids = vtkIdList::New();
    ids->Allocate(3);

    for(i=0; i<nCells; i++) {
        fscanf(fptr, "%li %li %li %li %li", &junk, &count, &id1, &id2,
&id3);
        if (count != 3) {
            points->Delete();
            BAIL("ADCIRC Elements can only contain 3 points!");
        }
        if (id1 > nPoints) { BAIL("Cell contains a point that wasn't
loaded!"); }
        if (id2 > nPoints) { BAIL("Cell contains a point that wasn't
loaded!"); }
        if (id3 > nPoints) { BAIL("Cell contains a point that wasn't
loaded!"); }
        if (id1 <= 0) { BAIL("Cell contains a point <= 0!"); }
        if (id2 <= 0) { BAIL("Cell contains a point <= 0!"); }
        if (id3 <= 0) { BAIL("Cell contains a point <= 0!"); }
        ids->Reset();
        ids->InsertNextId(id1-1);
        ids->InsertNextId(id2-1);
        ids->InsertNextId(id3-1);
        cells->InsertNextCell(ids);
    }
    fclose(fptr);

     vtkPolyData *output = vtkPolyData::New();
     output->Initialize();
     output->Allocate();
     output->SetPoints(points);
     output->SetPolys(cells);

3) the modified code, with a vtkUnstructuredGrid
    vtkUnstructuredGrid *output = vtkUnstructuredGrid::New();
    output->Initialize();
    output->Allocate(nCells, 10);
    output->SetPoints(points);
    vtkIdList *ids = vtkIdList::New();
    ids->Allocate(3);

    for(i=0; i<nCells; i++) {
        fscanf(fptr, "%li %li %li %li %li", &junk, &count, &id1, &id2,
&id3);
        if (count != 3) {
            points->Delete();
            BAIL("ADCIRC Elements can only contain 3 points!");
        }
        if (id1 > nPoints) { BAIL("Cell contains a point that wasn't
loaded!"); }
        if (id2 > nPoints) { BAIL("Cell contains a point that wasn't
loaded!"); }
        if (id3 > nPoints) { BAIL("Cell contains a point that wasn't
loaded!"); }
        if (id1 <= 0) { BAIL("Cell contains a point <= 0!"); }
        if (id2 <= 0) { BAIL("Cell contains a point <= 0!"); }
        if (id3 <= 0) { BAIL("Cell contains a point <= 0!"); }
        ids->Reset();
        ids->InsertNextId(id1-1);
        ids->InsertNextId(id2-1);
        ids->InsertNextId(id3-1);
        output->InsertNextCell(VTK_TRIANGLE, ids);
    }
    fclose(fptr);

So for reference, 2 things changed between the two pieces of code:
1) went from vtkPolyData to vtkUnstructuredGrid
2) Went from assigned all the cells in 1 shot with a vtkCellArray *
SetPolys, to individually adding each cell to the output Dataset.



On 6/14/06, Amy Squillacote <amy.squillacote at kitware.com> wrote:
>
> Hi Randall,
>
> vtkThreshold should be able to handle vtkPolyData; it accepts any type
> of vtkDataSet as input, and there's no apparent limitation on the type
> of cells that can be contained in the dataset to be thresholded. Would
> you please post the stack trace? If you have a small dataset that can be
> used to reproduce the problem (and that you would be able to post), that
> would be very helpful as well.
>
> - Amy
>
> Randall Hand wrote:
> > I have some code to load ADCIRC data into memory as vtkPolyData.  It's
> > been working great until yesterday when I tried to pass it to a
> > vtkThreshold to eliminate some of the "masked" values, and it
> > segfaulted.  a backtrace showed it somewhere in a
> > vtkDataArray::GetValue call.
> >
> > I fought with it for a good while, before I just (on a whim) switched
> > from vtkPolyData to vtkUnstructuredGrid, with all VTK_TRIANGLE cells.
> > And low and behold, it works.
> >
> > Is there some reason why a vtkPolyData can't be thresholded ( VTK5.0
> > stable)?
> >
> > ----------------------------------------
> > Randall Hand
> > Visualization Scientist
> > ERDC MSRC-ITL
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > This is the private VTK discussion list.
> > Please keep messages on-topic. Check the FAQ at:
> http://www.vtk.org/Wiki/VTK_FAQ
> > Follow this link to subscribe/unsubscribe:
> > http://www.vtk.org/mailman/listinfo/vtkusers
> >
>
>


-- 
----------------------------------------
Randall Hand
Visualization Scientist
ERDC MSRC-ITL
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20060614/a89ac282/attachment.htm>