Back to the main page.

Bug 3211 - ft_sourceinterpolate: failure to reshape funparameter data into cubic shape + retainment of pos

Reported 2016-11-22 13:38:00 +0100
Modified 2019-08-10 12:33:03 +0200
Product: FieldTrip
Component: core
Version: unspecified
Hardware: PC
Operating System: Linux
Importance: P5 normal
Assigned to:
Depends on:
See also:

Jan-Mathijs Schoffelen - 2016-11-22 13:38:58 +0100

report from Marseille workshop: in a standard approach, one of the participants ran into the thing that has been mentioned in the title. At first sight, all seems to have been executed OK, so it's not clear to me what is happening. - try and quickly replicate this - if the problem is indeed persistent: fix it we are using fieldtrip20161014

Jan-Mathijs Schoffelen - 2016-11-24 09:11:21 +0100

my first attempt at reproduction of this bug failed surprisingly. I did it on my Mac (OS el capitan), matlab 2015b + latest version of fieldtrip, interpolating a randn(npos,1) 'avg.pow' field of the standard sourcemodel 10 mm onto the single_subj_T1.nii No reshaping problems etc.

Hesham - 2016-11-24 13:03:42 +0100

Created attachment 817 source and source_conn

Hesham - 2016-11-24 13:06:02 +0100

Reminder: The problem arose when only few positions were selected to compute source activity. To compute source connectivity cfg.method = 'coh' and cfg.complex = 'absimg'. The coherence spectrum is not well-reshaped and the positions are duplicated. I have uploaded the 'source' and 'source_conn' structure

Jan-Mathijs Schoffelen - 2016-11-24 20:03:53 +0100

(In reply to Hesham from comment #3) Ah, I see. Then it probably translates to the fact that if the pos-list is not reshapable into a 3D, this strange behavior happens. Can you upload the original (full) pos-list, and the dim-vector (i.e. number of x/y/z voxels)?

Jan-Mathijs Schoffelen - 2016-11-24 20:05:12 +0100

Oh, and actually you uploaded the stuff in the wrong bug :o), should have been 3212

Jan-Mathijs Schoffelen - 2016-12-08 08:56:54 +0100

It has not been possible to reproduce this. Since I did not receive feedback when asked, there's no use to keep this open.

Robert Oostenveld - 2019-08-10 12:33:03 +0200

This closes a whole series of bugs that have been resolved (either FIXED/WONTFIX/INVALID) for quite some time. If you disagree, please file a new issue on