Back to the main page.
Bug 858 - Problem with megplanar on CTF 275
| Status | CLOSED FIXED |
| Reported | 2011-08-02 15:12:00 +0200 |
| Modified | 2011-09-06 14:37:03 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P1 normal |
| Assigned to: | Jörn M. Horschig |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Vladimir Litvak - 2011-08-02 15:12:08 +0200
Created attachment 112 Input data When I try to run ft_megplanar on my CTF 275 data I get in addition to the planar channels also MLT37 and later I find that the data in MLT37_dH and MLT37_dV is zero. I could track the problem to meglpanar_sincos (the same for orig and fitplane) that for some reason puts zeros in the rows and columns for this sensor. Here is an example code to reproduce the problem. I attach the input. montage = megplanar_sincos(cfg, grad); cind = strmatch('MLT37', montage.labelorg); rind = strmatch('MLT37', montage.labelnew); figure; plot(montage.tra(:, cind)); figure; plot(montage.tra(rind, :));
Jörn M. Horschig - 2011-08-02 15:37:15 +0200
Hi Vladimir, MLT37 is not in your neighbour-structure, thus no planar gradient cannot be computed on this channel. Can you let me know how you computed the neighbours? Best, Jörn
Vladimir Litvak - 2011-08-02 15:38:32 +0200
(In reply to comment #1) > Hi Vladimir, > > MLT37 is not in your neighbour-structure, thus no planar gradient cannot be > computed on this channel. Can you let me know how you computed the neighbours? > > Best, > Jörn I used the template
Vladimir Litvak - 2011-08-02 15:44:31 +0200
MLT37 is indeed missing there so the bug is in the template and not in the code.
Jörn M. Horschig - 2011-08-02 16:22:47 +0200
thanks for reporting this, Then, the problem is caused by my ignorance that in our CTF system, MLT37 is deactivated, and I constructed the template from the channels available in my data. I'm fixed the template using test data on our server with 275 channels. By the way, note that the 'triangulation' method gives similar results. Actually, the CTF template is currently exactly the same, and it will stay like this until I find time to improve the templates (but that's very tedious work)