Dear Joe,
You can specify the argument type="response" to predict. See the help page
for predict.glm for details.
I hope that this helps,
John
At 11:41 AM 2/13/2003 -0800, you wrote:
I regressed a count variable on 5 ratio-scale
predictors using glm(...,family=poisson)in S Plus 2000
(Windows 2000). The fit was good (1-res/null deviance
= 0.75, all variables significant, errors distributed
neatly across observed values, etc.). I then
extrapolated predictions across the very large dataset
from which the training dataset was sampled by
applying the summary() coefficients to their
respective variables.
The predicted values, however, are not on the same
scale as the response, and I have been unable to
transform the predictions to the appropriate scale.
WHAT EQUATION DOES S PLUS 2000 USE TO TRANSFORM THE
LINEAR PREDICTIONS OF GLM(...,FAMILY=POISSON) TO THE
RESPONSE SCALE?
Thank you,
Joe Sexton
=====
Landscape Ecology: Modeling and Analysis Center
& Department of Forest, Range, and Wildlife Science
Utah State University
"Go deeper."
-Zarathustra
__________________________________________________
Do you Yahoo!?
Yahoo! Shopping - Send Flowers for Valentine's Day
http://shopping.yahoo.com
--------------------------------------------------------------------
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
-----------------------------------------------------
John Fox
Department of Sociology
McMaster University
Hamilton, Ontario, Canada L8S 4M4
email: jfox@mcmaster.ca
phone: 905-525-9140x23604
web: www.socsci.mcmaster.ca/jfox
-----------------------------------------------------
|