Thank you Doug. Is the one that doesn't fail okay?
Best,
Fred
On Sep 5, 2017, at 16:57, Douglas N Greve greve@nmr.mgh.harvard.edu wrote:
in the one that fails you are including the binary gender as a continuous variable, which will fail in FSGD. All catagorical variables need to be coded as classes. see https://surfer.nmr.mgh.harvard.edu/fswiki/FsgdExamples
On 09/05/2017 03:07 PM, Uquillas, Federico D'Oleire wrote: Dear Doug,
Sending the FSGD described in our original email as failing due to it being "ill-conditioned or badly scaled" with DOSS [0 1 0 0 0]. It is titled "FAILS_N47_edited_BL_CT_PiB_centAge_fsgd.txt".
The other one is the one that runs and gives expected results with a DOSS contrast [0 1 0 0].
Thanks so much.
Best regards,
Fred
From: freesurfer-bounces@nmr.mgh.harvard.edu [freesurfer-bounces@nmr.mgh.harvard.edu] on behalf of Douglas N Greve [greve@nmr.mgh.harvard.edu] Sent: Tuesday, September 05, 2017 12:01 PM To: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] controlling for Sex in a model using two continuous variables onto CTh
can you send the fsgd file for the analysis that is failing?
On 08/29/2017 02:29 PM, Uquillas, Federico D'Oleire wrote: Dear FreeSurfer experts,
We have run a vertex-wise analysis regressing a continuous variable onto cortical thickness on the surface, and would like to verify that our contrast that uses Age and Sex as covariates was defined correctly, especially since there is probably more than one way to control for a binary variable like Sex.
On the website (https://surfer.nmr.mgh.harvard.edu/fswiki/DodsDoss) it suggests to make two regressors, one for Males, and one for Females, where for the former a 1 is indicative of the Male category, and a 1 on the latter is indicative of the Female category. However, using that FSGD file and running a DOSS contrast as [0 1 0 0 0], we get the error: matrix is ill-condition or badly scaled, condno=2.01889e+07. We believe this may be due to the fact that Male and Female categories are autocorrelated and Freesurfer likes variables to be de-meaned. Running a FSGD file with only Age and our variable of interest via a DOSS contrast of [0 1 0] works without any errors.
We thus created an FSGD file where we have Sex as our third variable (in addition to our continuous variable of interest and one demeaned continuous covariate), coding 1s for Males and 0s for Females, and we used a DOSS contrast of [0 1 0 0]. This provided us with a map that makes sense with our expectations.
We would like to verify if this is doing what we believe it is doing; that is, looking at the relationship of our continuous variable or interest after accounting for age and sex. Any thoughts would be greatly appreciated.
Thank you so much.
Best,
Fred
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 greve@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
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@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