Dear FreeSurfer Team,
I processed in MATLAB for Linear Mixed Effects Models Analysis with this X matrix:
X matrix is 'Subjects N X 11', 11 columns are as follow,
1. Intercept (all '1') 2. time (from baseline time point) 3. time2 4. Patients group = 1, Control = 0 5. 4.X time 6. 4.X time2 7. Patients group = 0, Control = 1 8. 7. X time 9. 8. X time2 10. age 11. extra values for covariate
1. When I process this command, [lhTh01,lhRe01] = lme_mass_fit_EMinit(X,[1],Y,ni,lhcortex,3,4) The MATLAB windows print the warning alarm repeatedly. " Warning: Matrix is close to singular or badly scaled. Results may be inaccurate." Is this X matrix something wrong?
2. After complete 'lme_mass_fit_EMinit' process with the warning, I tried the next command, [lhRgs01,lhRgMeans01] = lme_mass_RgGrow(lhsphere, lhRe01, lhTh01,lhcortex,2,95); This command took a long time to process, and it caused the whole stop MATLAB works. Is it also because of the X matrix...?
Could you let me know the solution for the X matrix problems ?
Best Wishes,
Han
Dear Martin Reuter,
Thank you for your reply.
Our patients have several time points, but controls have only 1 time points, so when I coded the patients group =1, controls groups = 0 , the 2. time and 3. time2 column were same as 5. 4.X time 6. 4.X time2 column. and 8. 7. X time and 9. 8. X time2 were all zeros.
And then, how about this X matrix (Subjects N X 5)?
1. Intercept (all '1') 2. 4.X time (from baseline time point) 3. 4.X time2 4. age 5. extra values for covariate
is it a vallid matrix for test the effects of group X time2 ?
Thank you,
Han.
On Sat, May 21, 2016 at 3:13 PM, Martin Reuter mreuter@nmr.mgh.harvard.edu wrote:
Hi Han, Try to find a local statistician to help you with your analysis. About your matrix: rows need to be number of all time points from all subject. Time 2 should probably not be there. Also columns 7-9 need to be dropped (they are just the negative of the rows before).
Best Martin On May 21, 2016 3:32 PM, Hanbyul Cho hanbyul.h.cho@gmail.com wrote:
Dear FreeSurfer Team,
I processed in MATLAB for Linear Mixed Effects Models Analysis with this X matrix:
X matrix is 'Subjects N X 11', 11 columns are as follow,
Intercept (all '1')
time (from baseline time point)
time2
Patients group = 1, Control = 0
4.X time
4.X time2
Patients group = 0, Control = 1
- X time
- X time2
age
extra values for covariate
When I process this command, [lhTh01,lhRe01] = lme_mass_fit_EMinit(X,[1],Y,ni,lhcortex,3,4) The MATLAB windows print the warning alarm repeatedly. " Warning: Matrix is close to singular or badly scaled. Results may be inaccurate." Is this X matrix something wrong?
After complete 'lme_mass_fit_EMinit' process with the warning, I tried the next command, [lhRgs01,lhRgMeans01] = lme_mass_RgGrow(lhsphere, lhRe01, lhTh01,lhcortex,2,95); This command took a long time to process, and it caused the whole stop MATLAB works. Is it also because of the X matrix...?
Could you let me know the solution for the X matrix problems ?
Best Wishes,
Han
Freesurfer mailing list 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.
Dear Han,
if controls have only 1 time point, you cannot compare/analyse any measurement of change across the groups (such as atrophy rates). You can only do a cross sectional analysis at baseline. For patients you could separately look at atrophy rates (if they differ from zero, which should be true for any type of group - so that alone is not very exciting). Usually there is only one time column with time-from-baseline for example
me1 0 ... me2 1.2 ... me 3 1.8 ... you1 0 .. you2 0.9 ... he1 0 she1 0 she2 1.1 and so on (this could be in years). There can be differently many rows per subject, depending on the number of time points. Single time point subjects, can be mixed in and will help to estimate cross subject variances, but they won't really help much with estimates of longitudinal changes. Having a full group with only a single time point does not make sense in a longitudinal design.
Best, Martin
On 05/21/2016 06:42 PM, Hanbyul Cho wrote:
Dear Martin Reuter,
Thank you for your reply.
Our patients have several time points, but controls have only 1 time points, so when I coded the patients group =1, controls groups = 0 , the 2. time and 3. time2 column were same as 5. 4.X time 6. 4.X time2 column. and 8. 7. X time and 9. 8. X time2 were all zeros.
And then, how about this X matrix (Subjects N X 5)?
- Intercept (all '1')
- 4.X time (from baseline time point)
- 4.X time2
- age
- extra values for covariate
is it a vallid matrix for test the effects of group X time2 ?
Thank you,
Han.
On Sat, May 21, 2016 at 3:13 PM, Martin Reuter <mreuter@nmr.mgh.harvard.edu mailto:mreuter@nmr.mgh.harvard.edu> wrote:
Hi Han, Try to find a local statistician to help you with your analysis. About your matrix: rows need to be number of all time points from all subject. Time 2 should probably not be there. Also columns 7-9 need to be dropped (they are just the negative of the rows before). Best Martin On May 21, 2016 3:32 PM, Hanbyul Cho <hanbyul.h.cho@gmail.com <mailto:hanbyul.h.cho@gmail.com>> wrote: Dear FreeSurfer Team, I processed in MATLAB for Linear Mixed Effects Models Analysis with this X matrix: X matrix is 'Subjects N X 11', 11 columns are as follow, 1. Intercept (all '1') 2. time (from baseline time point) 3. time2 4. Patients group = 1, Control = 0 5. 4.X time 6. 4.X time2 7. Patients group = 0, Control = 1 8. 7. X time 9. 8. X time2 10. age 11. extra values for covariate 1. When I process this command, [lhTh01,lhRe01] = lme_mass_fit_EMinit(X,[1],Y,ni,lhcortex,3,4) The MATLAB windows print the warning alarm repeatedly. " Warning: Matrix is close to singular or badly scaled. Results may be inaccurate." Is this X matrix something wrong? 2. After complete 'lme_mass_fit_EMinit' process with the warning, I tried the next command, [lhRgs01,lhRgMeans01] = lme_mass_RgGrow(lhsphere, lhRe01, lhTh01,lhcortex,2,95); This command took a long time to process, and it caused the whole stop MATLAB works. Is it also because of the X matrix...? Could you let me know the solution for the X matrix problems ? Best Wishes, Han _______________________________________________ Freesurfer mailing list 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.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Dear Martin Reuter,
Thank you for your clear explanation.
Best wishes,
Han.
On Tue, May 24, 2016 at 12:41 PM, Martin Reuter <mreuter@nmr.mgh.harvard.edu
wrote:
Dear Han,
if controls have only 1 time point, you cannot compare/analyse any measurement of change across the groups (such as atrophy rates). You can only do a cross sectional analysis at baseline. For patients you could separately look at atrophy rates (if they differ from zero, which should be true for any type of group - so that alone is not very exciting). Usually there is only one time column with time-from-baseline for example
me1 0 ... me2 1.2 ... me 3 1.8 ... you1 0 .. you2 0.9 ... he1 0 she1 0 she2 1.1 and so on (this could be in years). There can be differently many rows per subject, depending on the number of time points. Single time point subjects, can be mixed in and will help to estimate cross subject variances, but they won't really help much with estimates of longitudinal changes. Having a full group with only a single time point does not make sense in a longitudinal design.
Best, Martin
On 05/21/2016 06:42 PM, Hanbyul Cho wrote:
Dear Martin Reuter,
Thank you for your reply.
Our patients have several time points, but controls have only 1 time points, so when I coded the patients group =1, controls groups = 0 , the 2. time and 3. time2 column were same as 5. 4.X time 6. 4.X time2 column. and 8. 7. X time and 9. 8. X time2 were all zeros.
And then, how about this X matrix (Subjects N X 5)?
- Intercept (all '1')
- 4.X time (from baseline time point)
- 4.X time2
- age
- extra values for covariate
is it a vallid matrix for test the effects of group X time2 ?
Thank you,
Han.
On Sat, May 21, 2016 at 3:13 PM, Martin Reuter < mreuter@nmr.mgh.harvard.edu> wrote:
Hi Han, Try to find a local statistician to help you with your analysis. About your matrix: rows need to be number of all time points from all subject. Time 2 should probably not be there. Also columns 7-9 need to be dropped (they are just the negative of the rows before).
Best Martin On May 21, 2016 3:32 PM, Hanbyul Cho hanbyul.h.cho@gmail.com wrote:
Dear FreeSurfer Team,
I processed in MATLAB for Linear Mixed Effects Models Analysis with this X matrix:
X matrix is 'Subjects N X 11', 11 columns are as follow,
Intercept (all '1')
time (from baseline time point)
time2
Patients group = 1, Control = 0
4.X time
4.X time2
Patients group = 0, Control = 1
- X time
- X time2
age
extra values for covariate
When I process this command, [lhTh01,lhRe01] = lme_mass_fit_EMinit(X,[1],Y,ni,lhcortex,3,4) The MATLAB windows print the warning alarm repeatedly. " Warning: Matrix is close to singular or badly scaled. Results may be inaccurate." Is this X matrix something wrong?
After complete 'lme_mass_fit_EMinit' process with the warning, I tried the next command, [lhRgs01,lhRgMeans01] = lme_mass_RgGrow(lhsphere, lhRe01, lhTh01,lhcortex,2,95); This command took a long time to process, and it caused the whole stop MATLAB works. Is it also because of the X matrix...?
Could you let me know the solution for the X matrix problems ?
Best Wishes,
Han
Freesurfer mailing list 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.
Freesurfer mailing listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Martin Reuter, PhD Assistant Professor of Radiology, Harvard Medical School Assistant Professor of Neurology, Harvard Medical School A.A.Martinos Center for Biomedical Imaging Massachusetts General Hospital Research Affiliate, CSAIL, MIT Phone: +1-617-724-5652 Web : http://reuter.mit.edu
Freesurfer mailing list 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.
freesurfer@nmr.mgh.harvard.edu
-
Hanbyul Cho -
Martin Reuter