Values As Categories always picking NaN Color
This issue was created automatically from an original Mantis Issue. Further discussion may take place here.
The "Interpret Values As Categories" option in the color map editor is still of limited value because it tends to mark everything as NaN. This might be either because of floating-point inaccuracies or point value interpolation or both, make it practically impossible to create categorical values. Consider the following simple but useful use case of marking contours:
- Create a Wavelet source. Apply.
- Add a Contour filter.
- Remove the default isovalue and create a range of 5 isovalues with the defaults. (It will actually only create 3 contours. That's OK.)
- Turn on Compute Scalars.
- Apply.
- Color by RTData.
- Open the Color Map Editor.
- Check on "Interpret Values As Categories"
- Click the Add Active Values button.
The add active values button adds three categories corresponding to the isosurfaces extracted by the contour surface, just as we expected. But everything is still colored with the NaN color. Huh? Changing colors does not help.
I also tried using the calculator with the floor function to make sure that everything has the exact same value. Still no luck. I was able to get it almost working by running that through point to cell, but even so one contour still was colored as NaN, which is really weird.