Hello,
The attached image shows a plot generated from a design with one continuous variable (gender) and three continuous variables (called N, E, P), analyzed using the "old" method via command line rather than QDEC. For mri_glmfit, the DODS option was used and the contrast was 0 0 1 -1 0 0 0 0 (to compare the relation between N and thickness, in females versus males).
I am wondering why the fitted lines seem to be so far "off" from the actual data points. Could anyone please shed some light on this issue? Please let me know if you need any more information about our analysis.
Thanks, - Jerry
This is an issue with how that line is drawn when you have more than one continuous variable. One of the covariates is the x-axis. The data points are projected into the plane where the 2nd variable is 0. However, then the line is computed, it is not done so assuming that the 2nd variable is 0, so it does not look like it fits. This is a bug that's been on my plate to fix for a while.
doug
Jerry Yeou-Wei Chen wrote:
Hello,
The attached image shows a plot generated from a design with one continuous variable (gender) and three continuous variables (called N, E, P), analyzed using the "old" method via command line rather than QDEC. For mri_glmfit, the DODS option was used and the contrast was 0 0 1 -1 0 0 0 0 (to compare the relation between N and thickness, in females versus males).
I am wondering why the fitted lines seem to be so far "off" from the actual data points. Could anyone please shed some light on this issue? Please let me know if you need any more information about our analysis.
Thanks,
- Jerry
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Thanks Doug! So just to clarify: this is only a rendering bug? not an analysis processing bug?
Quoting Doug Greve greve@nmr.mgh.harvard.edu:
This is an issue with how that line is drawn when you have more than one continuous variable. One of the covariates is the x-axis. The data points are projected into the plane where the 2nd variable is 0. However, then the line is computed, it is not done so assuming that the 2nd variable is 0, so it does not look like it fits. This is a bug that's been on my plate to fix for a while.
doug
Jerry Yeou-Wei Chen wrote:
Hello,
The attached image shows a plot generated from a design with one continuous variable (gender) and three continuous variables (called N, E, P), analyzed using the "old" method via command line rather than QDEC. For mri_glmfit, the DODS option was used and the contrast was 0 0 1 -1 0 0 0 0 (to compare the relation between N and thickness, in females versus males).
I am wondering why the fitted lines seem to be so far "off" from the actual data points. Could anyone please shed some light on this issue? Please let me know if you need any more information about our analysis.
Thanks,
- Jerry
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
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
yes, it is just a display bug, not an analysis bug.
jyw.chen@utoronto.ca wrote:
Thanks Doug! So just to clarify: this is only a rendering bug? not an analysis processing bug?
Quoting Doug Greve greve@nmr.mgh.harvard.edu:
This is an issue with how that line is drawn when you have more than one continuous variable. One of the covariates is the x-axis. The data points are projected into the plane where the 2nd variable is 0. However, then the line is computed, it is not done so assuming that the 2nd variable is 0, so it does not look like it fits. This is a bug that's been on my plate to fix for a while.
doug
Jerry Yeou-Wei Chen wrote:
Hello,
The attached image shows a plot generated from a design with one continuous variable (gender) and three continuous variables (called N, E, P), analyzed using the "old" method via command line rather than QDEC. For mri_glmfit, the DODS option was used and the contrast was 0 0 1 -1 0 0 0 0 (to compare the relation between N and thickness, in females versus males).
I am wondering why the fitted lines seem to be so far "off" from the actual data points. Could anyone please shed some light on this issue? Please let me know if you need any more information about our analysis.
Thanks,
- Jerry
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
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
freesurfer@nmr.mgh.harvard.edu
-
Doug Greve -
Jerry Yeou-Wei Chen -
jyw.chen＠utoronto.ca