Hi Xiang-Zhen,

Are you sure that you need to assume different variances for each group? Try instead assigning all subjects to the same variance group (that is, put the value "1" for subjects in the file design.grp).

All the best,

Anderson


On 7 November 2014 09:59, Xiang-zhen Kong <[log in to unmask]> wrote:
Hi, all,
I tried to do a repeated ANOVA analysis (a factor with 6 levels, 5 subjects).
Attached is the related file prepared for flamoe.
 
Here is the cmd I used.
flameo --cope=merged.nii.gz --mask=../mask.nii.gz --dm=./behavior.mat --tc=./behavior.con --cs=./behavior.grp --fc=./behavior.fts --runmode=ols
 
But I got the following outputs:
------------------------------------------
Log directory is: logdir
Setting up:
ntptsing=6.000000
6.000000
6.000000
6.000000
6.000000
 
nonseparable design matrix and variance groups
 
An exception has been thrown
nonseparable design matrix and variance groupsTrace: Design::setup; Gsmanager::setup.
---------------------------------------------
 
I have checked these files, but got nothing.
Anyone came across this error?  Please help.
 
Thanks.
 
 
Best,
Xiang-zhen
 
 
 

¡i¤p¤Õ¦¨¹³¡jhttp://www.conxz.net