- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Hi everyone, Please bear with me.
I'm confused to finish this model alone.
There is a discrete independent variable and a binary dependent variable (percentage).
I can use proc glimmix to generate the figure below. But how can i generate a regression curve by Legendre polynomials like in the second figure?
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Hello,
I think you can fit a spline if you just want to know about the overall "trend".
If you want Legendre (orthogonal) polynomials, they are in SAS/IML ORPOL Function.
Good luck,
Koen
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
See this note that (in addition to estimating odds ratios) shows fitting a quadratic or spline model and producing a plot.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
If you fit a curve, then you are assuming that the independent variable (group) is continuous. But you said it is discrete. So which is it?