Back to the main page.
Bug 2875 - ft_sourceanalysis with method 'dics' fails due to non existing field 'label'
| Status | CLOSED FIXED |
| Reported | 2015-03-31 14:22:00 +0200 |
| Modified | 2019-08-10 12:30:46 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P5 normal |
| Assigned to: | Tzvetan Popov |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Tzvetan Popov - 2015-03-31 14:22:04 +0200
Testing the test_beamformer.m script returns an error during the call to ft_sourceanalysis. This is because in ft_checkdata on line 1197 the label field is explicitly removed: data = rmfield(data, 'label'); The test_script runs successfully after commenting the line out. Can somebody comment on whether this line is needed?
Robert Oostenveld - 2015-04-01 14:59:37 +0200
removing the label field does make sense regarding the content: after this step there should only be crsspctrm and no powspctrm, whereas label refers to powspctrm. But it is uncommon (inpossible?) to obtain such a data structure by calling ft_freqanalysis, so FT functions in general would not expect label to be missing.
Robert Oostenveld - 2015-04-08 11:56:12 +0200
I looked at the code, it was due to a bug I introduced as an additional check on data consistency. roboos@mentat001> svn commit utilities/ft_datatype_freq.m Sending utilities/ft_datatype_freq.m Transmitting file data . Committed revision 10324.