Hi Doug,
I saw on a previous mailing earlier this week that someone has been having
a similar error and that your advice was to demean the ICV. I apologise if
this is a stupid question, but what does it mean to demean? And, how do i
do that?
Thanks in advance!
Laura.
On Wed, Feb 29, 2012 at
it means to subtract the mean of the icv values from each. that way the
values are close to zero, and closer in value to the other values in the
glm (which seems to make the glm happier).
n.
On Thu, 2012-03-01 at 11:04 -0500, Laura Tully wrote:
Hi Doug,
I saw on a previous mailing
In the terminal output, it says that it will exclude a subject because
there is a NaN data point. NaN = not a number. Check your data table
to make sure that everything is ok.
doug
Nick Schmansky wrote:
it means to subtract the mean of the icv values from each. that way the
values are close
Hi Nick and Doug,
Thanks for the help, I'll give the demeaning a try.
Re: the missing values - all my missing data points are denoted with NaN so
that I can include those subjects on the measures that they do have, but
exclude them on measures they don't have (I think this is the only way to
If it's working with demeaning, then that was the problem. The subjects
with NaNs are going to be excluded entirely as we have no way to
estimate missing fields.
doug
Laura Tully wrote:
Hi Nick and Doug,
Thanks for the help, I'll give the demeaning a try.
Re: the missing values - all my
The demeaning/mean centering worked a treat! Thank you freesurfer experts!
Laura.
On Thu, Mar 1, 2012 at 11:36 AM, Douglas N Greve
gr...@nmr.mgh.harvard.eduwrote:
If it's working with demeaning, then that was the problem. The subjects
with NaNs are going to be excluded entirely as we have no
btw, to tie-in and follow-up on a related posted question about missing
values, if NaN is used for the missing value (as you did), then qdec
will automatically exclude that subject in an analysis if that variable
is used. the rejected subject(s) will be printed to the terminal.
N.
On Thu,
Hi Doug,
Below is the output in the terminal window from running the analysis to the
crash. Attached are the qdec.fsgd and y.fsgd files from the qdec output.
I'm not actually sure what these file represent - could you enlighten me?
Thanks!
Laura.
Number of subjects: 55
Number of factors:
Hi,
I am trying to run a between group analysis in QDEC whilst controlling for
intrancranial volume. The design has one dichotomous variable (group2,
group2), one continuous (performance on behavioral task), and one nuisance
variable (intracranial volume). each time I run it qec crashes with the
Can you send the fsgd file that is created? Also, please send more text
output. It's hard to tell from a single line.
doug
Laura Tully wrote:
Hi,
I am trying to run a between group analysis in QDEC whilst controlling
for intrancranial volume. The design has one dichotomous variable
10 matches
Mail list logo
