finding certain variable name from glm model summary in R

Question

I have a glm model and I want to select the name of the variable whose coefficient has the highest p-value. I know how to find the highest p-value, and I know how to get the number of the variable (in the order in which it appears in the model), but I don't know how to actually get the variable name. The reason I would like to do this is that I want to create a loop that on each iteration removes the variable with the least significant coefficient and reruns the model. I would do that manually, but I just have way too many variables.

Answer 1

The following sample code could be of help. This code outputs the column name corresponding to Maximum P value (also, ignores the intercept)

    mydata <- read.csv("http://www.ats.ucla.edu/stat/data/binary.csv")
    mydata$rank <- factor(mydata$rank)
    mylogit <- glm(admit ~ gre + gpa + rank, data = mydata, family = "binomial")
    coefficients <- coef(summary(mylogit))
    maxPColumn <- rownames(coefficients)[2:nrow(coefficients)][which.max(coefficients[2:nrow(coefficients),4])]
    maxPColumn