No - I meant you don't need BOTH - you need one or the other.

Cheers.

On 31 Jul 2012, at 15:36, sandra petris wrote:

ah thanks a lot Steve! I'm gonna delete the -D option and the 1st colums, let'see what happens

ciao! Sandra

2012/7/31 Stephen Smith <[log in to unmask]>

Hi - the answer is clearly in the error message  :-)

You don't need the -D option *and* the EV full of 1s.

Steve.

On 31 Jul 2012, at 12:34, sandra petris wrote:

Hi Stephen,

when I run randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d design_nico.mat -t design_nico.con -n 10000 -D --T2 -V I got the following error message:

Loading Data: 

Data loaded

Warning: You have demeaned your data, but at least one design column has non-zero mean

ERROR: Program failed

An exception has been thrown

Runtime error:- detected by Newmat: process fails to converge

MatrixType = Rect   # Rows = 15; # Cols = 4

Trace: SVD; pinv.

thanks in advance for any help

Sandra

2012/7/31 Stephen Smith <[log in to unmask]>

Hi - these look - what was the error message?

On 30 Jul 2012, at 16:14, Sandra Petris wrote:

Dear fsl experts,


I have a question regarding TBSS analysis. I would like to compute the correlation (or regression) between FA and behavioral scores.
I have one group of subjects (15 subjects) and behavioral scores (BS) taken from 3 tasks plus 1 regressors of no interest, age that I have demeaned (is it correct?)
I assume that the design matrix should look like this:



/NumWaves 5
/NumPoints 15
/PPheights 1 1 1 1 1
/Matrix
1 1,0434 1,1634 1,9836 -4,7333
1 0,3511 0,6998 0,6660 -6,7333
1 0,3733 0,1620 1,3627 4,2667
1 -0,7298 1,1634 0,9047 -3,7333
1 -0,3459 -0,1429 0,6176 -1,7333
1 -0,0511 0,1898 -0,1555 -3,7333
1 -0,0954 -2,1662 0,0961 -4,7333
1 -0,7031 -1,0768 -0,6896 -5,7333
1 0,3342 0,4163 -0,7304 9,2667
1 0,8755 -0,1218 -0,4432  -2,7333
1 -1,3132 -1,0332 0,5826  6,2667
1 0,4781 -0,9344 0,2008 -0,7333
1 -0,0692 0,1785 -1,1205 18,2667
1 0,6799 1,1634 -1,0984  1,2667
1 -0,8272    0,3381 -2,1765 -4,7333


Is the following contrast matrix correct?


/NumWaves 5
/NumContrasts 6
/PPheights 1 1 1 1 1
/Matrix
0  1 0 0 0
0 -1 0 0 0
0 0 1 0 0
0 0 -1 0 0
0 0 0 1 0
0 0 0 -1 0


When I run the randomize command I receive an error message.....are the NumWaves/NumConntrast an PPWheights correct?

Any help will be greatly appreciated


Best, Sandra

---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------

---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------

---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------