Hi Tom, Thanks a million for your reply. Actually, I found out the problem. I had renamed Stattalk.m script, which I downloaded some years ago, to g.m. I also had modified the script so that I could select multiple subjects (peak voxel values at certain MNI coordinates). Indeed, as you say, I realized that there was a confusion a few days ago. So, I deleted the script and the problem is gone now. YOU ARE QUITE RIGHT! Now, SnPM works marvelously. Again, I appreciate your reply. Best regards, Shary At 02:45 PM 09/05/2005 -0400, Thomas E. Nichols wrote: >Shary, > > > I am very interested in SNPM2 (Permutation) toolbox in SPM2 > > which you have developed. In order for me to understand how the > > toolbox works, I started with the example provided on the web page > > http://www.sph.umich.edu/ni-stat/ SnPM/ex.html#exchngNperms and > > the related sample images. I followed exactly (several times) the > > instructions; however, Matlab issued the following messages: >[...] > >Sorry you've had trouble. We've tried to reproduce your problem >and haven't been able to. I do have a few ideas, though: > > 1. If you haven't already, please start with a 'clean > slate'. Start with an empty directory and make > sure you're using a fresh downloaded copy of SnPM. > When starting, make sure there are no errors current; > (you can make sure of this by hitting Ctrl-C several > times). > > 2. Check for other programs in your path. In one of your > emails I saw the error > > ??? Attempt to execute SCRIPT g as a function. > > Error in ==> c:\spm\spm2\snpm2\snpm_cp.m > On line 183 ==> if size(CONT,2) ~= size([H C B G],2) > > this suggests that you have a script g.m in your path > somewhere, and Matlab is getting confused with it and > the variable name G. > > >Please let me know how it goes. > >-Tom > > > -- Thomas Nichols -------------------- Department of Biostatistics > http://www.sph.umich.edu/~nichols University of Michigan > [log in to unmask] 1420 Washington Heights > -------------------------------------- Ann Arbor, MI 48109-2029