History of AFNI updates  

|
October 30, 2009 04:35PM
Hi Alena,

Randomizing that seems very challenging. The programs are not
written to have a variable list of durations (e.g. C can last
2 or 3 TRs, but no other duration). The other aspects are more
doable, chosing which sequence and how they are spaced.

You essentially have 4 event sets that last at least 7 TRs each.
Note that make_random_timing.py can do TR-locked timing (or not),
along with minimum and maximum ISI, if you want.

But how to choose from a list of durations is not clear (given
the current code). How necessary are those duration lists?

By the way, does this mean you will use duration modulation
with 3dDeconvolve? i.e. -stim_times_AM1 K XXX dmBLOCK(0)
If event A can last 1-3 TRs, you would really need to use
dmBLOCK(0) in particular (no other dmBLOCK would be quite
appropriate).

- rick

Subject Author Posted

design matrix generation

Alena October 29, 2009 06:28PM

Re: design matrix generation

rick reynolds October 29, 2009 08:09PM

Re: design matrix generation

Alena October 30, 2009 01:46PM

Re: design matrix generation

rick reynolds October 30, 2009 04:35PM

Re: design matrix generation

Alena November 02, 2009 11:16AM

Re: design matrix generation

rick reynolds November 02, 2009 12:23PM