Sounds good. I tried this approach with Sphere with radius of number of events, but I wasn't sure if it only uses all events forward in time from first event or also goes back (which shouldn't happen for most cases anyway).
Thanks, Swaroop On Sun, May 19, 2013 at 9:17 PM, Yaroslav Halchenko <[email protected]>wrote: > > On Sun, 19 May 2013, Swaroop Guntupalli wrote: > > There is an example script on how to perform spatio-temporal > searchlight > > on event-related dataset, but is there a way to run only spatial > > searchlight without searching over time dimension? > > we should address this issue (which I guess you ran into if you didn't > provide "temporal" Sphere, right?) > > ATM to do it you would need to specify large enough (to get all) > "fake" Sphere for temporal dimension and run searchlight with roi_ids > pointing to first X features which must correspond all to the first > volume within each event (i.e. all having the same event_offsetidx). > > > -- > Yaroslav O. Halchenko, Ph.D. > http://neuro.debian.net http://www.pymvpa.org http://www.fail2ban.org > Senior Research Associate, Psychological and Brain Sciences Dept. > Dartmouth College, 419 Moore Hall, Hinman Box 6207, Hanover, NH 03755 > Phone: +1 (603) 646-9834 Fax: +1 (603) 646-1419 > WWW: http://www.linkedin.com/in/yarik > > _______________________________________________ > Pkg-ExpPsy-PyMVPA mailing list > [email protected] > http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa >
_______________________________________________ Pkg-ExpPsy-PyMVPA mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
