Hi - I am trying to do a GLM from the command line with fsl_glm
A call without arguments tells me that
>
> Usage:
> fsl_glm -i <input> -d <design> -o <output> [options]
>
> Compulsory arguments (You MUST set one or more of):
> -i,--in input file name (text matrix or 3D/4D image file)
>
> Optional arguments (You may optionally specify one or more of):
> etc...
>
So I passed this file as an input (or something similar)
glm.txt:
/tmp/test/subjA001.nii.gz 0 65.3
/tmp/test/subjA002.nii.gz 0 54.8
/tmp/test/subjA003.nii.gz 0 75.1
/tmp/test/subjA004.nii.gz 0 60.2
/tmp/test/subjB001.nii.gz 1 69.1
/tmp/test/subjB002.nii.gz 1 81.4
/tmp/test/subjB003.nii.gz 1 72.7
/tmp/test/subjB004.nii.gz 1 55.8
and then:
$ fsl_glm -i glm.txt
But that gives the following output:
> ols_dof: Error in determining trace, resorting to basic calculation
> Segmentation fault
But the design seems fine enough -- when I try in matlab:
>> a=[0 65.3
0 54.8
0 75.1
0 60.2
1 69.1
1 81.4
1 72.7
1 55.8
];
>> pinv(a)
that just gives a proper pseudo-inverse.
Does anyone know what's wrong with this syntax?
thanks
Alle Meije
|