Back to the main page.
Bug 1054 - channelposition crashes for sens from layout neuromag306all.lay because of missing.tra
| Status | CLOSED WORKSFORME |
| Reported | 2011-10-18 16:58:00 +0200 |
| Modified | 2012-02-06 15:37:13 +0100 |
| 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: |
Jörn M. Horschig - 2011-10-18 16:58:35 +0200
??? Reference to non-existent field 'tra'. Error in ==> channelposition at 240 meanpnt = [meanpnt; mean(sens.coilpos(abs(sens.tra(sel1(j),:))>0.5,:), 1)]; Error in ==> fixsens at 61 chanpos = channelposition(sens, 'channel', 'all'); Error in ==> ft_prepare_neighbours at 147 sens = fixsens(sens);
Jan-Mathijs Schoffelen - 2011-10-19 15:24:13 +0200
using ft_datatype_sens rather than fixsens
Jörn M. Horschig - 2011-10-20 15:30:44 +0200
different day, same... problem ;) Using the 2-D layout to determine the neighbours reading layout from file NM122.lay the call to "ft_layoutplot" took 0 seconds and an estimated NaN MB ??? Reference to non-existent field 'tra'. Error in ==> channelposition at 189 meanpnt1 = mean(sens.pnt(abs(sens.tra(sel(1),:))>0.5,:), 1); Error in ==> ft_datatype_sens at 103 [chanpos, chanori, chanlab] = channelposition(sens, 'channel', 'all'); Error in ==> ft_prepare_neighbours at 149 sens = ft_datatype_sens(sens);
Jörn M. Horschig - 2011-10-20 15:33:34 +0200
related, see bug 1055
Jörn M. Horschig - 2011-10-20 15:39:35 +0200
the problem seems to occur on every MEG layout
Jan-Mathijs Schoffelen - 2011-10-20 15:46:49 +0200
why is a call to channelposition necessary to begin with when the input is a layout-structure?
Jörn M. Horschig - 2011-10-20 16:01:38 +0200
probably it is not, instead, as Robert suggested in bug 1055, it might be better to deal with the sens structure in ft_read_sens,ft_prepare_layout directly. not sure if this is being done. let's say, I will look into this and resolve this one and bug 1055
Jörn M. Horschig - 2011-11-23 12:19:46 +0100
seems to work now
Boris Reuderink - 2012-02-03 22:11:09 +0100
Jörn, any clue why it suddenly started working? It should not be this magical..
Jörn M. Horschig - 2012-02-06 13:53:36 +0100
there was quite some work done on the whole sens structure definition, e.g. fixsens is now called in low-level functions. By this we ensure that the new sens-structure is being used in all high-level functions. I suppose that this and maybe some updating on the high-level code made this bug go away.