Back to the main page.
Bug 684 - channelposition returns incorrect results on grad-structure in attached file
| Status | CLOSED WORKSFORME |
| Reported | 2011-05-24 14:01:00 +0200 |
| Modified | 2011-06-22 14:18:44 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P1 normal |
| Assigned to: | Jan-Mathijs Schoffelen |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Jan-Mathijs Schoffelen - 2011-05-24 14:01:08 +0200
Created attachment 57 gradiometer structure the pnt-matrix returned contains duplicate positions for channel 1 and 257 check what's wrong and fix it.
Jan-Mathijs Schoffelen - 2011-05-24 14:40:36 +0200
Created attachment 58 gradiometer structure 2
Jan-Mathijs Schoffelen - 2011-05-24 15:08:32 +0200
made a temporary fix by changing the value of 0.5 in line 63 in channelposition into 0.4 This is very nasty one, and we should discuss in the fieldtrip meeting how to make the detection more robust.
Jan-Mathijs Schoffelen - 2011-06-02 11:55:10 +0200
I created a tentative fix in channelposition @ stephen & diego: can you guys check whether ft_megplanar still gives nans in some channels? Cheers, JM
Diego Lozano Soldevilla - 2011-06-02 12:44:22 +0200
(In reply to comment #3) > I created a tentative fix in channelposition > > @ stephen & diego: can you guys check whether ft_megplanar still gives nans in > some channels? > > Cheers, > > JM Hi Jan-Mathijs, There aren't no NaNs in my montage :) Best, Diego
Diego Lozano Soldevilla - 2011-06-08 11:01:50 +0200
(In reply to comment #4) > (In reply to comment #3) > > I created a tentative fix in channelposition > > > > @ stephen & diego: can you guys check whether ft_megplanar still gives nans in > > some channels? > > > > Cheers, > > > > JM > > Hi Jan-Mathijs, > > There aren't no NaNs in my montage :) > > Best, > > Diego Hi again Jan-Mathijs, I don't know for what reason but I'm getting again NaNs in data.grad.tra structure but only in one subject and not in others. I'm using the same analysis pipeline... Importantly, I'm only having problems for topoploting ( ft_topoplotTFR) freq info but not for multiploting (ft_multiplotTFR). For ft_multiplotTFR in interactive mode I still having topoplot problems. I attach you a data set with and without topoplot problems Best Diego
Diego Lozano Soldevilla - 2011-06-08 13:46:53 +0200
(In reply to comment #5) > (In reply to comment #4) > > (In reply to comment #3) > > > I created a tentative fix in channelposition > > > > > > @ stephen & diego: can you guys check whether ft_megplanar still gives nans in > > > some channels? > > > > > > Cheers, > > > > > > JM > > > > Hi Jan-Mathijs, > > > > There aren't no NaNs in my montage :) > > > > Best, > > > > Diego > > Hi again Jan-Mathijs, > > I don't know for what reason but I'm getting again NaNs in data.grad.tra > structure but only in one subject and not in others. I'm using the same > analysis pipeline... Importantly, I'm only having problems for topoploting ( > ft_topoplotTFR) freq info but not for multiploting (ft_multiplotTFR). For > ft_multiplotTFR in interactive mode I still having topoplot problems. > > I attach you a data set with and without topoplot problems > > > Best > > Diego The attachment is in the common file H:\common\temporary\4JanMathijs. If you need more information, just let me know Best, Diego
Jan-Mathijs Schoffelen - 2011-06-08 14:46:16 +0200
can you upload as attachments the data_bad.grad and the data_good.grad from BEFORE you do the planar transformation, i.e. after ft_rejectcomponent?
Diego Lozano Soldevilla - 2011-06-08 15:20:13 +0200
(In reply to comment #7) > can you upload as attachments the data_bad.grad and the data_good.grad from > BEFORE you do the planar transformation, i.e. after ft_rejectcomponent? It's done. you'll find 4 files in the common temporary folder. For example data_good_after_IC_before_planar.mat is the data with no problems in ploting ('good_') after IC rejection ('after_IC') and before planar calculation ('before_planar'). I did that because I'm suspecting that IC rejection step is doing something rare with the channel configuration => see this bug http://bugzilla.fcdonders.nl/show_bug.cgi?id=718 Diego
Jan-Mathijs Schoffelen - 2011-06-08 16:36:31 +0200
Created attachment 72 gradiometer structures grad1 works grad2 works (before IC) grad3 doesn't work grad4 doesn"t work (before IC)
Jan-Mathijs Schoffelen - 2011-06-09 17:18:37 +0200
I cannot reproduce the problem. I work with the 4 grad structures in the attachment #72. These are obtained from the 4 data structures posted by Diego. I do the following: cfg = []; cfg.neighbourdist = 4; cd ~/matlab/fieldtrip/private grad1p = megplanar(cfg, grad1); % etc for the other 3 any(isnan(grad1p.tra(:))) %returns 0 in all 4 cases. My diagnosis is that the problem is not caused by the issue in channelposition.