Hi Russ,
the --tr setting is only used for plotting purposes, no need to do
anything, it should give valid decompositions. Note, however, that in
cases where the 4th dim is subjects rather than time, melodic will
likely pick up quite a few components related to structural
differences depending on what you've done to get all the data into the
same space.
best regards
Christian
On 9 Nov 2007, at 22:55, Russell Poldrack wrote:
> hi - I would like to perform ICA on a dataset where the 4th
> dimension is different subjects rather than different timepoints in
> a timeseries. I have set the TR to 100 secs in order to prevent any
> confusion about temporal autocorrelation; is that reasonable, and is
> there anything else that I need to do to make sure that the
> decomposition is valid?
>
> thanks
> russ
>
> ---
> Russell A. Poldrack, Ph.d.
> Associate Professor
> Wendell Jeffrey and Bernice Wenzel Term Chair in Behavioral
> Neuroscience
> UCLA Department of Psychology
> Franz Hall, Box 951563
> Los Angeles, CA 90095-1563
>
> phone: 310-794-1224
> fax: 310-206-5895
> email: [log in to unmask]
> web: www.poldracklab.org
>
>
>
|