Back to the main page.
Bug 1940 - A crash when scaling EEG sensor array
| Status | CLOSED FIXED |
| Reported | 2013-01-21 15:10:00 +0100 |
| Modified | 2013-06-05 12:15:11 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P3 major |
| Assigned to: | Roemer van der Meij |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Vladimir Litvak - 2013-01-21 15:10:41 +0100
Hi, I found a new bug when scaling Biosemi EEG sensor array. The error is: ??? Error using ==> ft_transform_geometry at 57 only a rigid body transformation without rescaling is allowed Error in ==> ft_transform_sens at 33 sens = ft_transform_geometry(transform, sens); I added a test for it to test_SPM_ft_integration (fails at the moment). The problem is in line 43 of ft_transform_geometry where the ft_voltype(input) returns something different from 'unknown' with elec as input. This in turn comes down to the fact that when there is a 'type' field in the input ft_voltype just returns in without looking any further. A possible fix is: checkrotation = (ft_datatype(input, 'volume') && ~strcmp(ft_voltype(input), 'unknown')) || ft_senstype(input, 'meg'); But perhaps the fix should be elsewhere. Vladimir
Roemer van der Meij - 2013-01-23 15:20:58 +0100
Hi Vladimir, Thank for the report, I made ft_voltype a bit more robust in detecting volumes, so the rotation check is no longer performed when transforming a sens-structure. The test script also runs through again now.