Thanks Doug. Any suggestions for a more appropriate design? I guess this is a common situation where two or more parameters only apply to one group (in this case patients) and are zero or constant in the other.
Best, Anders 2015-03-02 21:21 GMT+01:00 Douglas N Greve <gr...@nmr.mgh.harvard.edu>: > > It looks like the 4th column is always -0.9 and the 6th col is always > -9.138 which makes them redundant with column 2 and causes the error. > doug > > On 03/02/2015 03:16 PM, Anders Hougaard wrote: >> Dear Doug, >> >> Thanks for the answer. I deliberately chose not to demean. >> With demeaning I get the same error (see below). Any suggestions? >> >> Best, >> Anders >> >> Design matrix ------------------ >> 1.000 0.000 -0.600 0.000 17.137 0.000; >> 1.000 0.000 -0.800 0.000 11.962 0.000; >> 1.000 0.000 7.100 0.000 11.190 0.000; >> 1.000 0.000 0.700 0.000 30.736 0.000; >> 1.000 0.000 3.100 0.000 23.735 0.000; >> 1.000 0.000 0.100 0.000 21.121 0.000; >> 1.000 0.000 1.100 0.000 12.395 0.000; >> 1.000 0.000 2.100 0.000 24.912 0.000; >> 1.000 0.000 0.100 0.000 -8.021 0.000; >> 1.000 0.000 -0.700 0.000 8.877 0.000; >> 1.000 0.000 0.100 0.000 0.893 0.000; >> 1.000 0.000 7.100 0.000 29.849 0.000; >> 1.000 0.000 -0.400 0.000 -1.825 0.000; >> 1.000 0.000 2.100 0.000 14.361 0.000; >> 1.000 0.000 2.100 0.000 7.267 0.000; >> 1.000 0.000 0.100 0.000 34.169 0.000; >> 1.000 0.000 0.100 0.000 30.500 0.000; >> 1.000 0.000 1.100 0.000 9.219 0.000; >> 1.000 0.000 0.100 0.000 26.270 0.000; >> 1.000 0.000 1.100 0.000 2.320 0.000; >> 1.000 0.000 0.100 0.000 22.325 0.000; >> 1.000 0.000 0.100 0.000 13.323 0.000; >> 1.000 0.000 1.100 0.000 -2.628 0.000; >> 1.000 0.000 7.100 0.000 12.379 0.000; >> 1.000 0.000 1.100 0.000 -2.581 0.000; >> 1.000 0.000 0.100 0.000 -4.492 0.000; >> 1.000 0.000 1.100 0.000 11.549 0.000; >> 1.000 0.000 0.100 0.000 -8.413 0.000; >> 1.000 0.000 1.100 0.000 9.663 0.000; >> 1.000 0.000 1.100 0.000 8.655 0.000; >> 1.000 0.000 1.100 0.000 25.778 0.000; >> 1.000 0.000 3.100 0.000 2.739 0.000; >> 1.000 0.000 0.100 0.000 21.745 0.000; >> 1.000 0.000 -0.400 0.000 -5.762 0.000; >> 1.000 0.000 0.100 0.000 14.969 0.000; >> 1.000 0.000 0.100 0.000 3.535 0.000; >> 1.000 0.000 0.100 0.000 -0.287 0.000; >> 1.000 0.000 1.100 0.000 -2.283 0.000; >> 1.000 0.000 0.100 0.000 -0.257 0.000; >> 1.000 0.000 0.100 0.000 -1.207 0.000; >> 1.000 0.000 0.100 0.000 1.140 0.000; >> 1.000 0.000 0.600 0.000 -4.829 0.000; >> 1.000 0.000 1.100 0.000 2.213 0.000; >> 1.000 0.000 0.100 0.000 2.503 0.000; >> 1.000 0.000 0.600 0.000 3.193 0.000; >> 1.000 0.000 0.100 0.000 10.361 0.000; >> 1.000 0.000 0.100 0.000 3.546 0.000; >> 1.000 0.000 6.600 0.000 12.532 0.000; >> 1.000 0.000 0.100 0.000 -1.286 0.000; >> 1.000 0.000 0.100 0.000 -5.428 0.000; >> 1.000 0.000 -0.400 0.000 -1.661 0.000; >> 1.000 0.000 -0.800 0.000 21.444 0.000; >> 1.000 0.000 0.100 0.000 21.517 0.000; >> 1.000 0.000 -0.600 0.000 8.515 0.000; >> 1.000 0.000 0.100 0.000 19.530 0.000; >> 1.000 0.000 1.100 0.000 1.824 0.000; >> 1.000 0.000 0.100 0.000 15.795 0.000; >> 1.000 0.000 0.600 0.000 -6.140 0.000; >> 1.000 0.000 2.600 0.000 0.803 0.000; >> 1.000 0.000 0.100 0.000 16.904 0.000; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> 0.000 1.000 0.000 -0.900 0.000 -9.138; >> -------------------------------- >> ERROR: matrix is ill-conditioned or badly scaled, condno = 1e+08 >> -------------------------------- >> Possible problem with experimental design: >> Check for duplicate entries and/or lack of range of >> continuous variables within a class. >> >> 2015-03-01 19:41 GMT+01:00 Douglas Greve <gr...@nmr.mgh.harvard.edu>: >>> Try demeaning your covariates. By demeaning, I mean to compute the mean >>> across all subjects, then subtract the mean from all values (for each >>> covariate). >>> doug >>> >>> >>> >>> On 3/1/15 2:40 AM, Anders Hougaard wrote: >>> >>> Dear all, >>> >>> I'm comparing cortical thickness of a group of patients to a group of >>> controls. >>> I want to regress out the effects of gender, age, disease severity and >>> disease duration. >>> For the controls, the last two parameters equals zero. >>> >>> When running mri_glmfit, i get ERROR: matrix is ill-conditioned or >>> badly scaled, condno = 1e+08 >>> See the design matrix attached. My contrasts are set as >>> 1 -1 0 0 0 0 0 0 0 0 >>> and >>> -1 1 0 0 0 0 0 0 0 0 >>> >>> What should I do to improve this? >>> >>> All the best, >>> Anders >>> >>> >>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer@nmr.mgh.harvard.edu >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> >>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer@nmr.mgh.harvard.edu >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> >>> The information in this e-mail is intended only for the person to whom it is >>> addressed. If you believe this e-mail was sent to you in error and the >>> e-mail >>> contains patient information, please contact the Partners Compliance >>> HelpLine at >>> http://www.partners.org/complianceline . If the e-mail was sent to you in >>> error >>> but does not contain patient information, please contact the sender and >>> properly >>> dispose of the e-mail. >>> >> _______________________________________________ >> Freesurfer mailing list >> Freesurfer@nmr.mgh.harvard.edu >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> > > -- > Douglas N. Greve, Ph.D. > MGH-NMR Center > gr...@nmr.mgh.harvard.edu > Phone Number: 617-724-2358 > Fax: 617-726-7422 > > Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting > FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 > www.nmr.mgh.harvard.edu/facility/filedrop/index.html > Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/ > > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
