DG MCLAREN wrote:
> Dear All,
>
> I'm trying to set up SPM5 to do an FIR analysis on a single subject. However, when I look at the design matrix, I was expecting values of 0s everyone except where the timepoint was at a specific post-stimulus time. This would be equivalent to an AFNI matrix and consistant with other FIR matrices I've seen. My matrix however, looks nothing like that. It has the general look to it, but almost every cell is non-zero. Does anyone have any suggestions?
>
> Thanks.
Not sure but I think this may be the answer...
It depends what you mean by "design matrix".
SPM implements a generalised linear model of the form:
K*W*Y = K*W*X*B + K*W*e
where K and W are matrices used for filtering and whitening the data
(and X is your design matrix).
If you look at the design matrix through the SPM interface then I think
it actually shows you K*X rather than just X. In this case you will
expect to see non-zero values in other cells as an effect of the
high-pass filtering.
If you want to look at just X it is stored in SPM.xX.X
Hope this helps,
Rich.
|