Hi Adrian,
Is your post related to this more recent one, as you are both from the
same university?
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=spm;76172077.1707
I am aware that the eigs() function has by default its starting vector
randomly generated so this might explain what you are observing.
This aside, it would be useful to have more details about the data and
model you are using so that we can look into this more specifically.
Best regards,
Guillaume.
On 02/07/17 15:35, Adrian O wrote:
> Hi !
> I am facing the same problem. When I invert model with the same configuration, I get similar but different results for the same dataset and same A,B,C matrices.
>
> Why it is like that ?
>
> I suspect it is due to the fact that EM algorithm has random element inside... am I right ??
>
> Adrian
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|