Print

Print


Hi,

We find that most people use the GUI or have a fixed design that they want to script, so we haven't dealt with this request before if memory serves.  It is obviously possible to have the software work differently, and the pseudo-code I sent you should let you achieve a more flexible script-based approach.  There is no theoretical reason that the software works precisely in this way. Our philosophy was that the design file for FEAT separates off the processing done by the GUI from the raw stats that need to be calculated by the subsequent scripts.  So that is why the design file works in the most fundamental specification (based on the real EVs).  All you need to get what you want, as far as I understand, is the small bit of script that converts to the real EV format, that replicates what is done inside the GUI.

I hope that helps.
All the best,
	Mark



> On 17 Aug 2017, at 10:36, Wiktor Olszowy <[log in to unmask]> wrote:
> 
> Dear Mark,
> 
> Thanks a lot for your reply! I am actually wondering what is the rationale of specifying the contrasts in the FEAT design file? Based on the HRF model, the experimental design and other information that is included in the FEAT design file, the contrasts could be automatically generated by FEAT after it reads the FEAT design file. If I understand it correctly, SPM does it so. When specifying the HRF model:
> 
>      jobs{1}.stats{1}.fmri_spec.bases.fir/fourier/gamma.length = par1
>      jobs{1}.stats{1}.fmri_spec.bases.fir/fourier/gamma.order = par2,
> 
> SPM.mat includes the contrasts.
> 
> Best,
> 
> Wiktor Olszowy