External Email - Use Caution        

Thanks. When I run mri-glmfit I get Error: matrix is ill-conditioned or badly scaled. 

I have normalised the continuous variable Age.

Kind regards,

Marie

---

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> on behalf of Greve, Douglas N.,Ph.D. <DGREVE@mgh.harvard.edu>
Sent: 03 July 2019 15:24
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] Qdec Message to Mailing List

 

yes, qdec can only have 2 factors with 2 levels. both qdec and mri_glmfit are linear methods (in fact, qdec just calls mri_glmfit).

On 7/3/2019 9:39 AM, Marie Hill wrote:

        External Email - Use Caution        

Dear Sir/Madam, 

Could I post the following message to the mailing list:

I am running a regression testing for significant difference in cortical volume, thickness etc with 3 factors: diagnosis, genotype and gender, with age as a nuisance factor. 

I am using qdec but receiving Error: factor 2 must have 2 levels.  

Is it true that Qdec only supports factors with 2 levels? There is a page for qdec levels stating to create a file diagnosis.levels and state the three levels. I did this however the same error occurs. 

I was going to use glm_fit instead but I think my study isn't a linear regression so isn't suitable (as I am not interested in the change with age, just between groups).

Kind regards,

Marie

_______________________________________________
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