Triangles per Vertex produces many single triangles

[kkiesling]

The statistic for triangles per vertex is producing many vertices that only have one triangle connected to it. This is theoretically not possible for a 3D geometry for a vertex to have 1 triangle. I think the minimum number of triangles has to be 3. The method for calculating this value in the code should be reviewed.

[kkiesling]

@yqin43 discovered while we were working on the roughness stat (#59) that there are random free-floating vertices in the geometries that are not connected to any triangles. This is probably picking up these "free-floaters." We should probably take these out of the stat calculation and just print out how many verts have 0 triangles (eg, "X vertices were found to have no adjacent triangles.").