UPDATE: Looks like there's a problem with NaNs (mean zero divided by mean zero). Will report back once I've corrected & confirmed.

stephen

stephen v. shepherd   phd
The Rockefeller University / 1230 York Avenue / New York NY 10065-6307 USA // 212.327.7699

On Sat, Jun 4, 2016 at 2:14 AM, S.V.Shepherd [work] <stephen.v.shepherd@gmail.com> wrote:
Hi,

I decided I should standardize my fMRI data across sessions by rescaling each voxel to have a run-mean of 100. Normalized data looked okay under manual inspection in FreeView. When I tried to analyze this data, however, I consistently get an error partway through:

"""
Index exceeds matrix dimensions.

Error in fast_selxavg3 (line 762)
      acfrun = nrho1segmn(nthseg).^([0:flac.ntp-1]');
 
>> ------------------------------------------
ERROR: fast_selxavg3() failed\n
a_100_Av_s_task_fmov_bgrnd failed
"""

Any guess what might be causing this, or how to troubleshoot?  Multiple analyses that worked on non-normalized data are mostly (but somehow not universally) failing in this same way.

Thanks much.

stephen


stephen v. shepherd   phd
The Rockefeller University / 1230 York Avenue / New York NY 10065-6307 USA // 212.327.7699