The following method creates model formula in an automated fashion. In this program, df is a name of dataframe. 'admit' is a target variable. And we are building logistic regression.

names(df) <- make.names(names(df))

y <- "admit"

x <- names(df)[!names(df) %in% y]

mymodel <- as.formula(paste(y, paste(x, collapse="+"), sep="~"))

glm(mymodel, data=df, family = binomial)

**Suppose you want to include all the predictors except one.**

glm(admit~.-rank, data = df, family = binomial)

Here, we are telling R to ignore variable rank while building model.

**Drop more than one variable in formula**

glm(admit~.-rank -gre, data = df, family = binomial)

## Post a Comment