Back to the main page.
Bug 3153 - ft_statfun_pooledT fails on 2015b (Field assignment to a non-structure array object), works on 2015a
| Status | CLOSED FIXED |
| Reported | 2016-06-22 13:26:00 +0200 |
| Modified | 2017-01-17 11:29:41 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P5 normal |
| Assigned to: | Robert Oostenveld |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Robert Oostenveld - 2016-06-22 13:26:08 +0200
This was detected in a test script. ------------------------------------------------ Field assignment to a non-structure array object. Error in ft_statfun_pooledT (line 76) s.stat = s; Error in ft_statistics_montecarlo (line 267) [statobs, cfg] = statfun(cfg, dat, design); Error in ft_freqstatistics (line 193) [stat, cfg] = statmethod(cfg, dat, design); Error in test_bug798 (line 46) stat = ft_freqstatistics(cfg,t2_subj1,t2_subj2,t2_subj1_null,t2_subj2_null);
Robert Oostenveld - 2016-06-22 13:29:28 +0200
this is simply like s = 1 s.stat = s; at the end of the statfun. I remember that a particular matlab version got strict about this, but that later matlab versions relaxed it again.
Robert Oostenveld - 2016-06-22 13:29:52 +0200
I suspect multiple of the older statfuns to have this pattern. It can easily be fixed.
Robert Oostenveld - 2016-06-22 13:34:43 +0200
(In reply to Robert Oostenveld from comment #1) I am actually not sure about the "later versions". Better forget that.
Robert Oostenveld - 2016-06-22 13:35:41 +0200
roboos@mentat001> git commit -a [master 57a863f] ENH - improve compatibility with 2015b and prevent "Field assignment to a non-structure array object". See http://bugzilla.fieldtriptoolbox.org/show_bug.cgi?id=3153 6 files changed, 12 insertions(+), 12 deletions(-)