[R] Error: (subscript) logical subscript too long

IOANNA ii54250 at msn.com
Mon Apr 2 17:52:48 CEST 2012


Hello, 

I am trying to perform a logistic regression using counts. For example:

cedegren <-
read.table("http://www.cloudstat.org/index.php?do=/attachment/download/id_95
/", header=T)
attach(cedegren)
ced.del <- cbind(sDel, sNoDel)
ced.logr <- glm(ced.del ~ cat + follows + factor(class),
family=binomial("logit"))

This works. However, if I change the family to Gaussian:

ced.logr <- glm(ced.del ~ cat + follows + factor(class), family=gaussian)

I get the error:
Error: (subscript) logical subscript too long

I would like to use the probit function. Is this possible?

Best regards, 
Ioanna



More information about the R-help mailing list