Odd ghost cell behavior of spcth data
This issue was created automatically from an original Mantis Issue. Further discussion may take place here.
While using the plot over line filter with some spcth data, I noticed some spikes in the plot that looked incorrect. After experimenting some more, I think that the issue is with errant behavior with ghost cell reading (or handling). To replicate, do the following:
- Open the cth-small example dataset from David K. (Kitware should have this but we can provide a copy if not.) Load in all cell arrays and go to the last time step. Apply.
- Add the Threshold filter (don't worry, the amount of cells we extract will be very small). Select Temperature (eV) as the scalar field and set the lower and upper thresholds as .04 and .06 respectively. Apply. Verify that the output is empty (that is, there are no cells with a temperature in this range).
- Add the Plot Over Line to the Original SPCTH data. Set the two points to be (0.5,-0.1,-3) and (0.5,-0.1,1). Apply.
- In the resulting plot, turn off all variables except for Temperature (eV). Notice that the plot has a strange spike that goes to about 0.05 (which should not be there since there are no values in this range).
- Go back to the 3D view and observer the output of the Threshold filter (in my experiments that filter was re-run, but if not you can force a re-run by toggling the All Scalars flag). You will see that many cells are now extracted. It looks like these cells are on the borders of blocks and that leads me to believe they are ghost cells with the wrong values.
Attached is a screenshot after the weird ghost cells are extracted.