coldeyes Liao
2008-May-19 19:14 UTC
[R] How to get confidence interval and coefficient in Logic Regression
sorry to bother everyone.
i have question to get the coefficient and confidence interval in Logic
Regression with Logistic model. below i list the R code
X <- matrix(as.numeric(runif(400) < 0.5), 50,8)
colnames(X) <- paste("X", 1:ncol(X), sep="")
rownames(X) <- paste("case", 1:nrow(X), sep="")
# Define expected result: Y = (NOT X1) AND X5
Y <- as.numeric(!X[,1] & X[,5])
z<-as.numeric(runif(50))
model.dat<-cbind(X,z,Y)
set.seed(12345)
Annealing <- logreg.anneal.control(start = 4, end = -4, iter = 1000, update
= 50)
logicfit <- logreg(resp=model.dat[,10],
bin=model.dat[,1:8],sep=model.dat[,9],
type = 3,
select = 2,
ntrees=2,
nleaves=3,
anneal.control=Annealing)
use names(logicfit) , i didn't see coef class. how to i get the
coefficient of the model , especially for the additive effect.
is there command i can calculate the sd so that i can have confidence
interval.
model.dat2<-model.dat
secondly i try to use alltrees <- eval.logreg(logicfit , model.dat2) to
score a new data set model.dat2. but this doesn't work.
thanks for your help.
leo
[[alternative HTML version deleted]]
