CODEWITHSUNDEEP
rmodelglm
liguang
liguang

Reputation: 161

Exclude a variable from glm()

I have multiple datasets where the response variable is always in the last column of the data frame. I want to run a GLM (logistic regression) and automate it. I call glm() by position but this method always includes the last variable.

data(iris)
head(iris) 
train<- iris 
logit <- glm(train[,length(train)]~ . , 
             data = train, family = "binomial")
summary(logit)

I tried writting train[,length(train)]~ . -train[,length(train)] but it doesn't work.

Upvotes: 1

Views: 282

Answers (1)

Haezer
Haezer

Reputation: 476

Quite verbose but I think that should work : 

logit <- glm(formula(paste0(names(train)[length(train)], '~.')), 
             data = train, 
             family = "binomial")

or using tail : 

logit <- glm(formula(paste0(tail(names(train), 1), '~.')), 
             data = train, 
             family = "binomial")

Upvotes: 1

Related Questions