Menu
▾
▴
#819 support for polyMesh dataset scheme
A mesh of triangles is about the oddest form of data Autoplot should ever support, supporting 3-D modelling files, supporting .stl and .3dmf files. It would be good to write code to support this boundary, to explore what problems this introduces.
I've introduced Schemes.triangleMesh to QDataSet and TriangleMeshRenderer to visualize this.
This did show where there was code in Autoplot which did not support the "internal" render type, which is used to support render types unknown to Autoplot.
Discussion
This will be within the code for v2024a_1, but should not be used in production code.
I generalized this to support 4-point polygons as well, since often we have grid of points and triangulation is not unique.
I generalized the few codes to support four-point polygons as well. The codes are now polyMesh instead of triMesh. QDataSet rules say that a slice of any dataset must be a qube, so all the polygons should have the same length. That said, there are other rules this is breaking already.
Last edit: Jeremy Faden 2024-01-18
I was revisiting this to see if we could have general polygons in the mesh. The rule that a slice of a QDataSet is always a QUBE prevents this. This rule should be reconsidered. Since this is all experimental, I'm going to break the rule.
There's a lot of overlap between this an existing types. For example, the 1D case of this is similar to an events list. It would be nice, too, if labels and colors could be attached to regions, as some feature digitizers have created.