Back to the main page.
Bug 1932 - prepare_mesh_headshape may create NaN pnt values
| Status | CLOSED FIXED |
| Reported | 2013-01-16 16:10:00 +0100 |
| Modified | 2013-06-06 15:11:49 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P3 normal |
| Assigned to: | Johanna |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Johanna - 2013-01-16 16:10:53 +0100
Created attachment 406 test_fairsurface.m I ran into this problem by calling ft_prepare_mesh with a headshape of 21285 .pnt but no .tri and with a cfg.numvertices 3000. See attached script test_fairsurface.m. On line 241 of prepare_mesh_headshape (which is actually a line within the subfunction fairsurface), the difference between desired and existing points is computed, which if 0, then the weighting param w_i is NaN and the new points created into XYZmm_o are NaN rather than adjusted.
Johanna - 2013-01-17 10:18:06 +0100
suggested change: if sum(d_i)==0 w_i = zeros(size(d_i)); else w_i = d_i/sum(d_i); end both for line 241 and 249 of prepare_mesh_headshape.m. Is this ok with the intended output? I tried finding the original file on which fairsurface was based in SPM8, to see if that file had any updates, but coudln't find it.