Back to the main page.
Bug 2778 - ft_regressconfound potentially copies unwanted fields over from input to output
| Status | CLOSED FIXED |
| Reported | 2014-12-12 10:36:00 +0100 |
| Modified | 2015-02-11 10:46:52 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P5 normal |
| Assigned to: | Robert Oostenveld |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Robert Oostenveld - 2014-12-12 10:36:10 +0100
It now is like this % output handling dataout = datain; if istimelock ... elseif isfreq ... elseif issource ... end where known fields get updated in the if-section. Safer would be to initially only copy the administrative fields (label/trialinfo etc) from detain to dataout and explicitly insert all fields that should be retained/recomputed.
Robert Oostenveld - 2014-12-12 17:06:50 +0100
mac011> svn commit ft_regressconfound.m test/test_example_incorporate_head_movements.m Sending ft_regressconfound.m Sending test/test_example_incorporate_head_movements.m Transmitting file data .. Committed revision 10040. I cleaned up the overall code. The test_ft_regresscounfound is doing fine, I was not able to check the other test script as I don't have the data here. That should be followed up on the dashboard.
Arjen Stolk - 2014-12-13 00:18:11 +0100
agreed, thanks. for ease of observation, this is the fix you implemented: % prepare the output, start with only the administrative fields dataout = keepfields(datain, {'label', 'time', 'freq', 'pos', 'dim', 'transform', 'inside', 'outside', 'trialinfo', 'sampleinfo'});
Robert Oostenveld - 2015-02-11 10:44:54 +0100
Closed several bugs that were recently resolved. If you don't like the resolution, please reopen and explain.