Re: [Marsbar-users] marsbar - estimate results

[Matthew B. <mat...@gm...>]

Hi,

Sorry for the late reply.

On Wed, Jul 13, 2011 at 6:23 PM, Edgar Chan <e....@uq...> wrote:
> I'm currently having problems with Marsbar while trying to extract %signal change. I'm trying to extract %signal change from a predefined ROI, and for 1 participant, when I do the "estimate results" step, I get the following error msg in Matlab:
>
>
> Error in ==> spdiags at 114
>         a((len(k)+1):len(k+1),:) = [i i+d(k)
>         B(i+(m>=n)*d(k),k)];
>
> Error in ==> pr_estimate at 128
>    s     = spdiags(1./sqrt(diag(s)),0,nScan,nScan);
>
> Error in ==> pr_estimate at 345
>    SPM = pr_estimate(SPM,marsY);
>
> Error in ==> mardo_2.estimate at 72
> SPM        = pr_estimate(SPM, marsY);
>
> Error in ==> marsbar at 1282
> marsRes = estimate(marsD, marsY, struct(...
>
> Error in ==> spm at 831
> evalin('base',CBs{v-1})
>
>
> For this same participant, extracting data from other ROI using the same procedure and same design does not result in this error. Everything seems to be fine except for this particular ROI.

I think you might have hit this problem:

http://marsbar.sourceforge.net/faq.html#i-get-errors-using-the-spm-reml-estimation-for-fmri-designs-can-i-try-something-else

Does the fix there help you?

Best,

Matthew