there are no infinite values in that regressor - it's just that with a
TR of 32s (?) and such a block design, the sampling of the predicted
response nearly only consists of 0s and 1s.
You can have access to the time course of your regressors with:
>> load SPM.mat
Salem Boussida wrote:
> Dear SPMers,
> I have a question about the GLM modelisation.
> In the graphical window, when regressors are plotted against scans, there is
> an abnormal result (it seems like the number of the estimated regressors is
> infinite). To better understand, a picture (*.jpg) is joined to my message.
> Could you help me please.
> Thank you,
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG