s-news
[Top] [All Lists]

estimated risk by covariates

To: s-plus news <s-news@lists.biostat.wustl.edu>
Subject: estimated risk by covariates
From: jose Bartolomei <surfprjab@hotmail.com>
Date: Mon, 29 Jun 2009 15:11:04 +0000
Importance: Normal

Dear S-users,

I have significant interaction terms in my logistic and Poisson models.

Example of a Poisson model:

 

glm(A ~ B : C + offset(logPop) , family=poisson(link=log), data="">

B and C are two level (dichotomous) variables, but I have up to 8 levels in some cases.

 

I would like to calculate and plot the estimated risk by levels of the third factor (C).

 

I am nothing like a programmer, so ….

 

There is an already built function or script to guide me in this task?

 

Thanks in advance,

Jose



Hotmail® has ever-growing storage! Don’t worry about storage limits. Check it out.
<Prev in Thread] Current Thread [Next in Thread>
  • estimated risk by covariates, jose Bartolomei <=