Errors in `vtkCellQuality` Verdict reference manual
The Verdict reference manual https://public.kitware.com/Wiki/images/6/6b/VerdictManual-revA.pdf has a number of errors in it when compared to actual computed values returned by `vtkCellQuality`. ### Quadrilateral skew and warpage The manual shows: ![quad_skew](/uploads/2971f36dc9856cc3117d3d39afb8226b/quad_skew.png) ![quad_warpage](/uploads/082983c21ccf8e5ac9284e27c9324443/quad_warpage.png) But, skew for a unit square is computed as `0` (not `1`) and warpage for a unit square is computed as `1` (not `0`). The tables make sense if we swap the values for "Acceptable Range" and "q for unit square". The other ranges may also need to be swapped ? ### Tetrahedral distortion The manual shows: ![tet_distortion](/uploads/df426ab52c75092e793618ecddfdbfc5/tet_distortion.png) But "q for unit equilateral tetrahedron" is computed as `1`, not `0` ### Tetrahedral scaled jacobian The manual shows: ![tet_scaled_jacobian](/uploads/f2162de4f227072c4cfc82d27947d684/tet_scaled_jacobian.png) But the acceptable and normal ranges both _exclude_ the value `1` which is the value of "q for unit equilateral tetrahedron". Shouldn't the ideal value of `1` be included in these ranges? Otherwise it suggests that a unit tetrahedron has unacceptable quality and is degenerate (which it is not). ### Hexahedral dimension The manual shows: ![hex_dimension](/uploads/c0e48d6aaa83be531fbafde9b99cba23/hex_dimension.png) But "q for unit cube" is computed as `sqrt(3)/3=0.577`, not `1` ### Hexahedral volume The manual shows: ![hex_volume](/uploads/2c2aaa4bc7a488f2603123dfcaf1ad77/hex_volume.png) But "q for unit cube" is computed as `1`, not `2` ### Naming differences There are also many differences between measure names in the manual and the measure name used by `vtkCellQuality`. Many of the differences are easy to resolve, e.g. "triangle maximum angle" is used in the manual but `vtkCellQuality` uses the method `SetQualityMeasureToMaxAngle` (i.e., "max", not "maximum"). And for tetrahedron, the manual uses the term "minimum dihedral angle", but with `vtkCellQuality` there is only `SetQualityMeasureToMinAngle`, (i.e. "min", not "minimum dihedral"), so here we have to guess that "dihedral" is omitted and assume both are referring to the same metric. In other cases, things are less clear. We have "mean aspect frobenius" in the manual, but `vtkCellQuality` uses `SetQualityMeasureToMedAspectFrobenius`, (i.e., "med", not "mean"). But "med" implies median, not mean, so it's not clear if the metric in the manual and the one computed by VTK are the same.
issue