Back to the main page.
Bug 2939 - conflict in cfg.headshape.headshape & cfg.individual.headshape in ft_volumerealign & ft_interactiverealign respectively
| Status | CLOSED FIXED |
| Reported | 2015-07-28 21:55:00 +0200 |
| Modified | 2019-08-10 12:31:16 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Linux |
| Importance: | P5 normal |
| Assigned to: | |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Stephen Whitmarsh - 2015-07-28 21:55:52 +0200
Hi there buggers, Found my password... Anyway, here goes again: ft_volumerealign now expects the headshape struct in cfg.individual.headshape (or breaks on line 68). However, in my case its also calls ft_interactiverealign which expects cfg.headshape.headshape (line 509) which isn't passed through properly. At that point it is in cfg.headshape again. Its a bugger, no? All the best from the Swedish "summer". Hope its better wherever you are. Cheers, Stephen p.s. working on SVN rev 10561
Stephen Whitmarsh - 2015-07-28 22:27:37 +0200
NB: To be clear: ft_volumerealign calls ft_interactiveralign
Jan-Mathijs Schoffelen - 2015-07-29 08:22:12 +0200
Line 68 in my version of ft_volumerealign is in the middle of the documentation section of the function...
Stephen Whitmarsh - 2015-07-29 09:31:58 +0200
Ah sorry - other way around. Line 68(67) in ft_interactiverealign, 509 in ft_volumerealign.
Jan-Mathijs Schoffelen - 2015-07-29 09:35:10 +0200
Aha, that explains it. Looks like Robert's recent changes to accommodate the new naming convention (from 'vol' to 'headmodel') causes some problems. Do you happen to have some data and code to quickly reproduce this?
Stephen Whitmarsh - 2015-07-29 09:57:06 +0200
Thanks! Coming up - but server just crashed...
Stephen Whitmarsh - 2015-07-29 10:18:28 +0200
Created attachment 726 headshape
Stephen Whitmarsh - 2015-07-29 10:20:08 +0200
It crashes at the second call to ft_volumerealign. I attached the headshape. The MRI is too large to upload. Shall I send it to you (JM) by mail? Else you should be able to use any MRI, but then need to run through the first part. cfg = []; cfg.method = 'interactive'; cfg.coordsys = 'neuromag'; [mri_realigned] = ft_volumerealign(cfg, mri_orig); cfg = []; cfg.method = 'headshape'; cfg.headshape.headshape = shape; [mri_realigned] = ft_volumerealign(cfg, mri_realigned);
Jan-Mathijs Schoffelen - 2015-07-29 11:05:01 +0200
[jansch@mentat002 fieldtrip]$ svn commit -m "bugfix - fixed some incompatibilities after making the change in naming from vol to headmodel" ft_interactiverealign.m Sending ft_interactiverealign.m Transmitting file data . Committed revision 10562. [jansch@mentat002 fieldtrip]$ Hi Stephen, The calls to ft_checkconfig at the beginning in ft_interactiverealign caused the problem. I think I fixed it now.
Stephen Whitmarsh - 2015-07-29 11:17:40 +0200
It works! Held!