Function for building a model with a variable variable and other constants

Question

It’s simple, but I don’t remember what this procedure is called, so I could not find this function. I want to study the effects and gradients of a simple lm () model by building the response of one variable at a time, the rest are kept constant.

Can someone tell me which function to use for this? I seem to remember that this is a function that generates multiple graphs, or something like that. It might be something like a sensitivity analysis ... Sorry for the initial question.

Thank you in advance!

Answer 1

car . + ( ).

library(car)   # for avPlots(...)
fit <- lm(mpg~wt+hp+disp, mtcars)
crPlots(fit)

, termplot(...) .