Re: [Freesurfer] matrix error

The right way to do it is to remove the column entirely. You can't do this inside of the FSGD unless you use DOSS. You can use DOSS if there is no difference between the groups in the age slope. In fact, it is a good idea to run this test first. So, create an FSGD with just the age. Create a contrast to test for a difference between the groups on age slope (and use DODS). If there are no differences, then re-run your original model using DOSS (you will need to change the contrast vector).

doug

On 05/07/2012 02:50 PM, Allie Rosen wrote:

Hi Doug,

Is there a different number I can substitute into the fifth column so the numbers are not all zeroes?

Thank you, Allie

On Mon, May 7, 2012 at 12:39 PM, Allie Rosen <rosen.allie@gmail.com mailto:rosen.allie@gmail.com> wrote:

Hi Doug,

Thanks for the help. The fifth column is all zeroes because the
covariate - length of disease - isn't applicable to control
subjects. In the end, I want to test the difference in thickness
between the two groups, regressing out the effect of both
covariates (age and length of disease). I think this means the
difference in intercepts between the two groups.

Thanks again,
Allie



On Mon, May 7, 2012 at 12:29 PM, Douglas N Greve
<greve@nmr.mgh.harvard.edu <mailto:greve@nmr.mgh.harvard.edu>> wrote:

    Hi Allie, the 5th column of your design matrix is all 0s,
    which causes
    the ill conditioning. I don't think this design fits into our
    standard
    framework. You can still analyze and test, we'll just have to
    figure out
    exactly what needs to be done. In the end, what is it that you
    want to
    test? Just a difference in intercepts between the two groups?
    doug

    On 05/07/2012 09:38 AM, Allie Rosen wrote:
    > Hello,
    >
    > I am analyzing cortical thickness in a patient and control
    group. I
    > have two covariates: age and length of disease. I am using
    the model from:
    >
    > http://surfer.nmr.mgh.harvard.edu/fswiki/Fsgdf2G2V
    >
    > However, for the control group I put length of disease as 0
    in my FSGD
    > file. When I run the GLM, I get the following error message:
    >
    > matrix is ill-conditioned or badly scaled, condno = 1e+08
    >
    > Can you please tell me what I am doing wrong? Below is the
    required
    > information.
    >
    > 1. Your command line:
    >
    >     mri_glmfit --y lh.disease_length.thickness.20.mgz --fsgd
    > fsgd_disease_length.txt dods --glmdir
    > lh.disease_length.thickness.20.glmdir --surf fsaverage lh --C
    > groupdiff_length_disease.mat
    >
    >   2. The FSGD file
    >
    > GroupDescriptorFile 1
    > Title Anorexia vs Controls
    > Class Control plus blue
    > Class Patient circle green
    > Variables Age Disease_Length
    >
    > Input C07 Control 37 0
    > Input C28 Control 28 0
    > Input C39 Control 37 0
    > Input C54 Control 40 0
    > Input C57 Control 32 0
    > Input C10 Control 54 0
    > Input C33 Control 31 0
    > Input C29 Control 22 0
    > Input A01 Patient 39 20
    > Input A02 Patient 24 11
    > Input A03 Patient 35 15
    > Input A04 Patient 40 10
    > Input A05 Patient 35 20
    > Input A06 Patient 57 20
    > Input A07 Patient 31 16
    > Input A08 Patient 20 8
    >
    >   3. And the design matrix
    >
    > Design matrix ------------------
    >  1.000   0.000   37.000   0.000   0.000   0.000;
    >  1.000   0.000   28.000   0.000   0.000   0.000;
    >  1.000   0.000   37.000   0.000   0.000   0.000;
    >  1.000   0.000   40.000   0.000   0.000   0.000;
    >  1.000   0.000   32.000   0.000   0.000   0.000;
    >  1.000   0.000   54.000   0.000   0.000   0.000;
    >  1.000   0.000   31.000   0.000   0.000   0.000;
    >  1.000   0.000   22.000   0.000   0.000   0.000;
    >  0.000   1.000   0.000   39.000   0.000   20.000;
    >  0.000   1.000   0.000   24.000   0.000   11.000;
    >  0.000   1.000   0.000   35.000   0.000   15.000;
    >  0.000   1.000   0.000   40.000   0.000   10.000;
    >  0.000   1.000   0.000   35.000   0.000   20.000;
    >  0.000   1.000   0.000   57.000   0.000   20.000;
    >  0.000   1.000   0.000   31.000   0.000   16.000;
    >  0.000   1.000   0.000   20.000   0.000   8.000;
    > --------------------------------
    >
    > Thank you,
    > Allie
    >
    >
    >
    >
    > _______________________________________________
    > Freesurfer mailing list
    > Freesurfer@nmr.mgh.harvard.edu
    <mailto:Freesurfer@nmr.mgh.harvard.edu>
    > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

    --
    Douglas N. Greve, Ph.D.
    MGH-NMR Center
    greve@nmr.mgh.harvard.edu <mailto:greve@nmr.mgh.harvard.edu>
    Phone Number: 617-724-2358 <tel:617-724-2358>
    Fax: 617-726-7422 <tel:617-726-7422>

    Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
    <http://surfer.nmr.mgh.harvard.edu/fswiki/BugReporting>
    FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
    <http://www.nmr.mgh.harvard.edu/facility/filedrop/index.html>

    _______________________________________________
    Freesurfer mailing list
    Freesurfer@nmr.mgh.harvard.edu
    <mailto: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.