Hi Doug,
selxavg-sess -analysis group7_ISIXTR_BERT_ERFIRsm5pf5tpefsub_18cat -sf sessid7 -df sesspar -noomnibus
exits with the error:
... Saving offset to group7_ISIXTR_BERT_ERFIRsm5pf5tpefsub_18cat/h-offset_000.bfloat Slice 1, 77.2438 -------------- ??? Error using ==> minus Matrix dimensions must agree.
Error in ==> fast_selxavg at 439 extreg = extreg - repmat(mean(extreg), [ntrs 1]);
??? Undefined function or variable 'r'.
quiting matlab
------------------------------------------ ERROR: fast_selxavg() failed\n ERROR (/autofs/space/cuzco_014/users/raij/avml_fmri_mctogether/avml10_session_BRISI04): selxavg failed
I suspect this may have to do that, for each of the seven subjects, there are three runs with an unequal number of TRs.
Oddly enough, the same analysis worked in March 07. Is there a way to make this analysis work again? I am only interested in BOLD% signal change (FIR waveforms) so comparing runs with different numbers of TRs should be legal. What I would like to do is to calculate a single fixed effects group analysis where each of the three run types is output as its own category (the paradigm files are coded to allow that). Only runs that have the same number of TRs are collapsed across subjects.
(Running three separate analyses, one for each run type, would not help, as we need the output to be in a single block with multiple categories, just as in our earlier analysis)
Or maybe the error has to do with something else... my matlab setup.m should be fine to my knowledge.
machine: cuzco nmr-dev-env (same error with the standard edition) cd /space/cuzco/14/users/raij/avml_fmri_mctogether setenv SUBJECTS_DIR /space/cognito/5/users/raij/subjects_mri setenv SUBJECT fsaverage
Thanks!!!
-Tommi
--- Tommi Raij, M.D., Ph.D. MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging Bldg 149, 13th St Charlestown, MA 02129 U.S.A.