Fix some lints

[etiennebacher]

Hi @vincentarelbundock, I've made a new package to find lints and automatically fix them (flint). I used marginaleffects as one of my test cases and found a few lints so I thought you might want to remove them. In particular, I think it might have fixed a bug in R/methods_nnet.R.

FYI those fixes were automatically generated with flint::fix_dir("R", exclude_linters = "equal_assignment"). End of the self-advertising :)

[vincentarelbundock]

Thanks @etiennebacher. The new tool looks great.

Yes, this is definitely a bug. Good catch.

Unfortunately, it also looks like flint introduced a new bug—or at least a departure from intended behavior.

model <- glm(mpg ~ hp, data = mtcars)

# same behavior
sup <- c("glm", "lm")
isTRUE(all(sup %in% class(model)))
> [1] TRUE
isTRUE(all(inherits(model, sup)))
> [1] TRUE

# different behavior
sup <- c("glm", "junk")
isTRUE(all(sup %in% class(model)))
> [1] FALSE
isTRUE(all(inherits(model, sup)))
> [1] TRUE

[etiennebacher]

Thanks, nice catch, I reverted the change here and added a fix in flint