Easier just to use the coef() extractor with the approriate name for
subscripting:
> x <- glm(Result~Test*SwabSite,data=sub.dat,family=binomial)
> xs <- summary(x)
> coef(xs)
Value Std. Error t value
(Intercept) 2.2899864 0.3699669 6.1897064
Test -2.4512546 0.4279627 -5.7277291
SwabSite 0.3368489 0.5585084 0.6031223
Test:SwabSite 1.5902032 0.6695218 2.3751327
>
> coef(xs)["Test",]
Value Std. Error t value
-2.451255 0.4279627 -5.727729
>
> coef(xs)["Test:SwabSite",]
Value Std. Error t value
1.590203 0.6695218 2.375133
Brad Biggerstaff, Ph.D.
(970) 221-6473 ... BBiggerstaff@cdc.gov
-----Original Message-----
From: Chuck Cleland [mailto:ccleland@optonline.net]
Sent: Friday, May 23, 2003 9:12 AM
To: Steve Sullivan
Cc: s-news@lists.biostat.wustl.edu
Subject: Re: [S] extracting rows of $coefficients by variable name?
Steve Sullivan wrote:
> Is there a handy way to extract information from a summarized lm or glm
> object's $coefficients by variable name? I want to pull out only the
> coefficient estimate, SE, t-value and p-value for a particular variable
> (my treatment indicator).
Steve:
You could use use subscripting on the returned matrix with
grep() to isolate the row by name. Here is an example:
my.lm <- lm(Y ~ ., data = mydata)
summary(my.lm)$coefficients[grep("COND",
row.names(summary(my.lm)$coefficients)),]
where "COND" matches a particular coefficient name
hope this helps,
Chuck Cleland
--------------------------------------------------------------------
This message was distributed by s-news@lists.biostat.wustl.edu. To
unsubscribe send e-mail to s-news-request@lists.biostat.wustl.edu with
the BODY of the message: unsubscribe s-news
|