External Email - Use Caution
For example, if my Class CaseFemaleNonWhite has no subject, should I cancel
the contrast between CaseFemaleNonWhite vs ControlFemaleNonWhite?
However, that will leave my ControlFemaleNonWhite subjects unrecognized? Or
it is better to cancel the race dis
External Email - Use Caution
got you, thanks
Best,
Zeng
On Thu, Nov 11, 2021 at 4:06 PM Douglas N. Greve
wrote:
> For DODS you would add 16 zeros (8 classes times 2 continuous variables)
>
> On 11/11/2021 4:00 PM, Zeng, Qi wrote:
>
> External Email - Use Caution
> So
For DODS you would add 16 zeros (8 classes times 2 continuous variables)
On 11/11/2021 4:00 PM, Zeng, Qi wrote:
External Email - Use Caution
So for the DODS case, is "1 -1 1 -1 1 -1 1 -1 0 0 0 0" for the
following FSGD:
GroupDescriptorFile 1
Title MyStudy
Class CaseMaleWhite
Class C
External Email - Use Caution
So for the DODS case, is "1 -1 1 -1 1 -1 1 -1 0 0 0 0" for the following
FSGD:
GroupDescriptorFile 1
Title MyStudy
Class CaseMaleWhite
Class ControlMaleWhite
Class CaseMaleNonWhite
Class ControlMaleNonWhite
Class CaseFemaleWhite
Class ControlFemaleWhit
For the DOSS case, yes.
On 11/11/2021 3:23 PM, Zeng, Qi wrote:
External Email - Use Caution
Hi Douglas,
If to set up the FSGD as below, is my design matrix for case and
control comparison as "1 -1 1 -1 1 -1 1 -1 0 0" ?
GroupDescriptorFile 1
Title MyStudy
Class CaseMaleWhite
Class
External Email - Use Caution
Hi Douglas,
If to set up the FSGD as below, is my design matrix for case and control
comparison as "1 -1 1 -1 1 -1 1 -1 0 0" ?
GroupDescriptorFile 1
Title MyStudy
Class CaseMaleWhite
Class ControlMaleWhite
Class CaseMaleNonWhite
Class ControlMaleNon
I would advice setting up the FSGD files with 8 classes based on your 3
discrete, 2-level variables, eg, MaleControlWhite then use two
continuous variables Age and eTIV.
You should definitely normalize the eTIV (and probably age too)
On 11/10/2021 7:07 PM, Zeng, Qi wrote:
External Ema
External Email - Use Caution
Hi Douglas,
Just to be clear, should I recode discrete variables as numeric as sample1)
or change discrete variables into strings 2) or even change the comparison
groups into strings too 3)?
1)
Variables Age Gender Race eTIV
Input Sub001_base Control 8
Did you read the error message? It says:
Variable 2 has character string Female
Variables should be continuous numbers
You have listed Gender and Race as strings, they have to be numbers (eg,
1, 2)
You probably want to have these be discrete variables instead of continuous
On 11/10/2021 6:20 PM
External Email - Use Caution
Hi Douglas,
Here is the command line:
mri_glmfit \
--y rh.thickness-rate.MyStudy.10.mgh \
--fsgd FSGD/MyStudy.fsgd dods\
--C Contrasts/Case-Control_age_gender_race_etiv.mtx\
--surf fsaverage rh \
--cortex \
--glmdir rh.thickness-rate.MyStud
please send your command line and full terminal output as well as the
fsgd file
On 11/10/2021 5:12 PM, Zeng, Qi wrote:
External Email - Use Caution
Hi Freesurfer experts,
Here, I want to design a simple group comparison (two levels: Case vs.
Control) and also count for 4 covariates
11 matches
Mail list logo
