The easiest way to solve this problem is to assign one or two events to be the missing conditions. This should not change any of the results. Make sure that you do not include these subjects if you want to do tests on the missing conditions.
doug
On 12/06/2013 04:24 PM, Eryilmaz, Huseyin Hamdi wrote:
Hi all,
I'm trying to run a simple event-related analysis, in which the conditions are determined based on subject's performance. It is possible that some of the predicted conditions do not appear in the paradigm file (mostly because the subject performs well). In these cases, do I need to renumber the conditions (I don't prefer this since I want the condition codes to be consistent across subjects) ? If I run the design with some conditions missing I get the following error. I checked the created Xtmp.mat, and XtX matrix contains NaNs in the columns corresponding to the conditions that don't appear in the paradigm file. Does anyone know a robust solution to this problem?
Many thanks! -Hamdi
/Creating Design Matrix ... creation time = 0.006 sec DoMCFit = 1 ntptot = 235, nX = 22, DOF = 213 Saving X matrix to /autofs/cluster/roffman/users/Stable5_PerRun/MTH20067AA/bold/ER_SIRP_TEST2/Xtmp.mat Error using svd Input to SVD must not contain NaN or Inf.
Error in cond (line 39) s = svd(A);
Error in fast_selxavg3 (line 256) XCond = cond(XtX);/
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer