External Email - Use Caution
Hello FreeSurfer Developers, my fsgd is this： GroupDescriptorFile 1 Title post compane Class d Class m Class h
RescaleFlag 1 DemeanFlag 1 Variables gender Age Edu Input bd1 d 0 15 8 Input bd2 d 1 19 13 Input bd7 d 0 17 11 Input bd8 d 1 19 13 Input bd6 m 1 22 12 Input bd9 m 1 26 7 Input bd10 m 0 53 15 Input bd11 m 1 56 4 Input hc3 h 0 45 7 Input hc4 h 1 18 13 Input hc1 h 0 52 7 Input hc2 h 1 24 10
When I run Three Groups (One Factor/Three Levels), three Covariates,my MTX is
1 -1 0 0 0 0 0 0 0 0 0 0 1 0 -1 0 0 0 0 0 0 0 0 0 It’s wrong,What should I do -------------------------------- ERROR: matrix is ill-conditioned or badly scaled, condno = 1e+08
-------------------------------- -------- ERROR: READ THIS ----------------- --vvvvvvvvvvvvvvvvvvvvvvvvvvvvvv----- Possible problem with experimental design: Check for duplicate entries and/or lack of range of continuous variables within a class. If you seek help with this problem, make sure to send: 1. Your command line: mri_glmfit --y rh_teach66.mgh --fsgd glmteach66.fsgd --C glmteach66.mtx --glmdir group66.area_66sm.rh --fwhm 10 --surface fsaverage rh --eres-save 2. The terminal output of this program (ie, everything it prints to the screen) 3. The FSGD file (glmteach66.fsgd) 4. The design matrix group66.area_66sm.rh/Xg.dat Attempting to diagnose further WARNING: Class d has 4 members. With 3 variables and using DODS, This is the bare minimum which may cause problems with the design matrix. WARNING: Class m has 4 members. With 3 variables and using DODS, This is the bare minimum which may cause problems with the design matrix. WARNING: Class h has 4 members. With 3 variables and using DODS, This is the bare minimum which may cause problems with the design matrix. SumSq: Min=1.491914 (col 10), Max=2.676059 (col 11) ... could not determine the cause of the problem --^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^----
ERROR: cannot find residual group66.area_66sm.lh/eres needed with --perm-resid when running mri_glmfit, make sure to include --eres-save ERROR: cannot find residual group66.area_66sm.rh/eres needed with --perm-resid when running mri_glmfit, make sure to include --eres-save glm66.sh: line 14: --perm-resid: command not found -----------------------
You don't have enough members in the classes to fit the model that you have. Eg, class d has 4 members with which you are fitting an offset and 3 slopes leaving you with 0 degrees of freedom. You need to get more subjects or change your model. btw, gender is usually modeled as a discrete rather than continuous covariate.
On 4/11/2022 8:06 AM, David wrote:
External Email - Use Caution
Hello FreeSurfer Developers, my fsgd is this： GroupDescriptorFile 1 Title post compane Class d Class m Class h
RescaleFlag 1 DemeanFlag 1 Variables gender Age Edu Input bd1 d 0 15 8 Input bd2 d 1 19 13 Input bd7 d 0 17 11 Input bd8 d 1 19 13 Input bd6 m 1 22 12 Input bd9 m 1 26 7 Input bd10 m 0 53 15 Input bd11 m 1 56 4 Input hc3 h 0 45 7 Input hc4 h 1 18 13 Input hc1 h 0 52 7 Input hc2 h 1 24 10
When I run Three Groups (One Factor/Three Levels), three Covariates,my MTX is
1 -1 0 0 0 0 0 0 0 0 0 0 1 0 -1 0 0 0 0 0 0 0 0 0 It’s wrong,What should I do
ERROR: matrix is ill-conditioned or badly scaled, condno = 1e+08
-------- ERROR: READ THIS ----------------- --vvvvvvvvvvvvvvvvvvvvvvvvvvvvvv----- Possible problem with experimental design: Check for duplicate entries and/or lack of range of continuous variables within a class. If you seek help with this problem, make sure to send: 1. Your command line: mri_glmfit --y rh_teach66.mgh --fsgd glmteach66.fsgd --C glmteach66.mtx --glmdir group66.area_66sm.rh --fwhm 10 --surface fsaverage rh --eres-save 2. The terminal output of this program (ie, everything it prints to the screen) 3. The FSGD file (glmteach66.fsgd) 4. The design matrix group66.area_66sm.rh/Xg.dat Attempting to diagnose further WARNING: Class d has 4 members. With 3 variables and using DODS, This is the bare minimum which may cause problems with the design matrix. WARNING: Class m has 4 members. With 3 variables and using DODS, This is the bare minimum which may cause problems with the design matrix. WARNING: Class h has 4 members. With 3 variables and using DODS, This is the bare minimum which may cause problems with the design matrix. SumSq: Min=1.491914 (col 10), Max=2.676059 (col 11) ... could not determine the cause of the problem --^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^----
ERROR: cannot find residual group66.area_66sm.lh/eres needed with --perm-resid when running mri_glmfit, make sure to include --eres-save ERROR: cannot find residual group66.area_66sm.rh/eres needed with --perm-resid when running mri_glmfit, make sure to include --eres-save glm66.sh: line 14: --perm-resid: command not found
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
David -
Douglas N. Greve