[log in to unmask]" type="cite">

Hi,

Hard to tell - but do you really want the slash in the front of your file names "/Dual_Reg"? Try ls'ing the files written this way.  You want to look at the files in the "scripts+logs" directory to determine if any errors are reported.

Cheers,

Eugene

--

On 5 May 2013 12:48, Mohamed Ali Bahri <[log in to unmask]> wrote:

Any suggestion or comment about the flowing dual_regression problem?

Dear fsl experts,

I run dual_regression (v0.5 beta on linux ubuntu) on two groups (19 vs. 30 subjects) but as results I did not get the cross-subject statistics files (e.g. dr_stage3_ic(#ICA)_tstat(#con).nii.gz).


The used contrast are the following (generated using Glm):
==============================
/ContrastName1     "group A > group B"
/ContrastName2     "group B > group A"
/ContrastName3     "group A mean"
/ContrastName4     "group B mean"
/NumWaves    2
/NumContrasts    4
/PPheights        1.000000e+00    1.000000e+00    1.000000e+00 1.000000e+00
/RequiredEffect        0.962    0.962    0.866    0.419

/Matrix
1.000000e+00 -1.000000e+00
-1.000000e+00 1.000000e+00
1.000000e+00 0.000000e+00
0.000000e+00 1.000000e+00
=============================

The used command line:

dual_regression /Dual_Reg.gica/groupmelodic.ica/melodic_IC 1 /Dual_Reg/Design/CTRLvsAD.mat /Dual_Reg/Design/CTRLvsAD.con 500 /Dual_Reg/dr_results `cat /Dual_Reg.gica/.filelist`


Many thanks in advance,

-- 
 

Dr. Mohamed Ali Bahri,
Cyclotron Research Centre,
University of Liège, 
Belgium,
[log in to unmask]

-- 
 

Dr. Mohamed Ali Bahri,
Cyclotron Research Centre,
University of Liège, 
Belgium,
[log in to unmask]