[R] what actually a logistic regression fits

Bin Yue leffgh at 163.com
Fri Dec 7 08:17:52 CET 2007


Dear all:
    I did the following because I was not sure what a logistic regression
fits ."small.glm" is a glm fit , setting the family to be binomial.

>fitted(small.glm)->p
>log(p/(1-p))->left   
> for(i in 1:length(coef(small.glm))){
+  coef(small.glm)[i]*model.matrix(small.glm)[,i]->res[,i]
+ }
> apply(res,1,sum)->right
> plot(left,right)  #I got a straight line whose slope was about 1
Therefore ,I am almost sure that what a logistic regression fits is
log(p/(1-p))~b0+b1*x.
Is that right? 
Regards,

-----
Best regards,
Bin Yue

*************
student for a Master program in South Botanical Garden , CAS

-- 
View this message in context: http://www.nabble.com/what-actually-a-logistic-regression-fits-tf4960618.html#a14207920
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list