sorry, I mean to do it for both the FSGD file and the matlab script. I just tried it with your data and it works. doug
soft.join Huang wrote:
Hi, doug.
Thanks for your suggestion, but it does not seem to work. After having divided the ICV by 1e6, I reculated the betas, and got the similar results as before in Matlab, just the beta4 -- correspondent to ICV -- multiplied a 10^6. And then I computed the rank of matrix X that the value is 4, indicating that each 2 variables in design matrix do not highly correlated. Also I have finished an regress analysis including one class variable and one continuous variable using mri_glmfit, got the same result in Matlab. A wired situation... Is it any additional constrains in mri_glmfit when doing regression analysis?
Lijie Huang
On Mon, May 2, 2011 at 9:09 PM, Douglas N Greve <greve@nmr.mgh.harvard.edu mailto:greve@nmr.mgh.harvard.edu> wrote:
I think the problem may be that your data are "badly scaled" meaning that one column is much greater than another column. Try dividing the ICV by 1e6 and see if you get more similar results. doug soft.join Huang wrote: Doug, Thanks for your reply. I recalculated the beta value and got the same results as the Matlab does, but be different from the mri_glmfit's. I can not figure out where the problem is. Details following: I use the command mri_glmfit --table lh.volume.txt --fsgd volume.gender.fsgd doss --C gender.diff.mtx --glmdir lh.vol.glmdir to finish regression analysis. and got the beta value ( beta1 = 1.9781e+03, beta2 = 2.0639e+03, beta3 = 110.9220, beta4 = 5.1964e-04 ) from beta.mgh file. Using X.mat file and the same dependent variable Y.mat as mri_glmfit's, I got totally different results in Matlab( beta1 = 5.19e+02, beta2 = 4.44e+02, beta3 = 90.32, beta4 = 0.0015 ). I don't know how it happened... Hope for your reply. The data I used are all in the attachment. Best Regrads, Lijie Huang On Fri, Apr 29, 2011 at 9:45 PM, <greve@nmr.mgh.harvard.edu <mailto:greve@nmr.mgh.harvard.edu> <mailto:greve@nmr.mgh.harvard.edu <mailto:greve@nmr.mgh.harvard.edu>>> wrote: Lijie, I don't know what the matlab regress() function does. mri_glmfit is just solving the GLM equation, which you can do easily in matlab: beta = inv(X'*X)*X'*y; try that and see if you get the same as mri_glmfit. doug > Hi all, > > I'm working on a group analysis using mri_glmfit. > In analysis directory, I got a X.mat file which containing the matrix X ( > size 50x4 ) used to regress analysis, also the beta value from beta.mgh > file. > Strangely, I loaded the X.mat in Matlab and recalculated the beta using > function regress in matlab, but got a different result from that of > mri_glmfit. > I don't know where the problem is, is there any difference in algorithm > between mri_glmfit and matlab function regress( regress(Y,X) )? or I made > some mistakes during the procedure. > The X.mat file and the dependent variable Y.mat are in the attachment. > > Thanks in advance > > Lijie Huang > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu <mailto:Freesurfer@nmr.mgh.harvard.edu> <mailto: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. -- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu <mailto:greve@nmr.mgh.harvard.edu> Phone Number: 617-724-2358 Fax: 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>