Hi - the easy way to find out how to create the appropriate design.fsf is to setup the kind of model (including contrasts) that you want from the GUI, and save the design.fsf - and then just compare that against what you're producing from your scripts, to see what they need to do differently.

Cheers.



On 27 May 2010, at 19:16, Magno R. Guillen B. wrote:

AUTOMATING THE OPERATION OF FSL VIA SCRIPTS.

 

 

Hi,

 

I am a postdoc student at Florida International University; I have been using fsl for 3 years by now. I have tried to automate the operation of fsl via scripting. I have been succesful in automating the processing of 1 contrast model, but I have not been able to automate 2 or more contrasts.
I  have written the script to mimic the structure of the desing.fsf used in the operation of fsl so I can automate the processing of several datsets with one single contrast, but I have failed to aautomate 2 or more contarst design in fsl (modifying designTemplateFEAT.fsf ).

 

I have executed manually fsl to process a RH-LH finger tapping paradigm  to compare to the file generated by my script. On runing my script with multi contrasts, fsl performs the prestat part but it  stops after a while generating the error message ERRORS OCURED DURING THE ANALYSIS.

 

I have noticed that there is a file called: design.com

 

Which  has:
/ContrastName1    rh
/ContrastName2    lh
/NumWaves   2
/NumContrasts     2
/PPheights        1.279210e+00      1.279210e+00
/RequiredEffect         1.285 1.285

 

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

 

The same file is generated when I run my script whit the following content:
/ContrastName1   
/ContrastName2   

 

 

I would appreciate any reference or hint to write a script that generates a model with more than 1 contrast. I thought it was only needed to write these lines on the design.fsf file but I was wrong.

 

# Number of contrasts
set fmri(ncon_orig) 2
set fmri(ncon_real) 2

 

Any help is deeply appreciated.

 

 

 

 



-- 
Magno R. Guillen, PhD
Office: 2220 ECE Building
Phone: (305) 348 - 4106
College of Engineering
Center of Advanced Education & Technology  - CATE
Florida International University - FIU
URL--> http://web.eng.fiu.edu/~mguil004/

This message, including any attachments hereto, may contain privileged and/or confidential information and is intended solely for the attention and use of the intended addressee(s).  If you are not the intended addressee, you may neither use, copy, nor deliver to anyone this message or any of its attachments.  In such case, you should immediately destroy this message and its attachments and kindly notify the sender by reply mail.  Unless made by a person with actual authority conferred by Dr. Guillen.  You are urged to verify any information that is confusing and report any errors/concerns to me in writing.

Este mensaje, incluyendo cualquier documento adjunto, puede contener información confidencial y/o privilegiada y está dirigida solamente para la atención y uso del o de los destinatarios. Si usted no es el destinatario, usted no puede usar, copiar o entregar a nadie este mensaje o cualquiera de los documentos adjuntos. Si éste es el caso, usted debe destruir inmediatamente este mensaje y sus adjuntos y le solicitamos que por favor notifique por la misma vía al remitente. Salvo que el mensaje provenga de una persona con autoridad real conferida por el Dr. Guillen. Se le exhorta que verifique cualquier información confusa y me informe cualquier error o preocupación de forma escrita.




---------------------------------------------------------------------------
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
---------------------------------------------------------------------------