Back to the main page.

Bug 2308 - ft_clusterplot does not like ANOVAs (which can only return positive clusters)

Reported 2013-10-03 11:00:00 +0200
Modified 2014-01-29 13:28:33 +0100
Product: FieldTrip
Component: core
Version: unspecified
Hardware: PC
Operating System: Windows
Importance: P3 normal
Assigned to: Eelke Spaak
Depends on:
Blocks: 2058
See also:

- 2013-10-03 11:00:16 +0200

Hi guys, I was just trying to plot the results of a call to ft_freqstatistics where I used a 'depsamplesF' test. Obviously the F value returned cannot be negative, so I was not surprised to see 'posclusters' and 'posclusterslabelmat' in my data structure, but no 'negclusters' or 'negclusterslabelmat'. However, when I then call ft_clusterplot to plot the data, I get the following error... >> ft_clusterplot(cfg,stat) reading layout from file CTF275.lay the call to "ft_prepare_layout" took 0 seconds and required the additional allocation of an estimated 0 MB ??? Reference to non-existent field 'negclusterslabelmat'. Error in ==> ft_clusterplot at 179 negCLM = squeeze(stat.negclusterslabelmat); Adding a fake - empty - field makes it happy again... stat.negclusterslabelmat=[]; % this now plots ...but I guess it would be neater if clusterplot were prepared for the output from freqstatistics. Best, Tom

Eelke Spaak - 2013-10-09 12:04:14 +0200

Should be fixed in rev 8589 (and on /home/common/matlab/fieldtrip/ in about 15 minutes). If it isn't, please reopen, Tom :)

Eelke Spaak - 2014-01-29 13:28:33 +0100

changing lots of bugs from resolved to closed.