Chapter 16 Trade-Offs

Can we make a class using S3, S4, and R6 so we can easily directly compare the three, and their generics and methods?

16.0.1 Using our example from the S4 Chapter:

setClass("Person", 
  slots = c(
    name = "character", 
    age = "numeric"
  )
)

setGeneric("age", function(x) standardGeneric("age"))
setGeneric("age<-", function(x, value) standardGeneric("age<-"))

setMethod("age", "Person", function(x) x@age)
setMethod("age<-", "Person", function(x, value) {
  x@age <- value
  x
})

john <- new("Person", name = "John Smith", age = 12)
age(john)

[1] 12

A generic is a function that will dispatch differently based on the class it is applied to. Above we apply the age generic to Person using setMethod but we could create an age method that behaves differently for another class, like dogs (which will add 7 to the supplied age)

setClass("Person", slots = c(name = "character",age = "numeric"))

setGeneric("age", function(x) standardGeneric("age"))
setGeneric("age<-", function(x, value) standardGeneric("age<-"))

setMethod("age", "Person", function(x) x@age)
setMethod("age<-", "Person", function(x, value) {
  x@age <- value
  x
})

setClass("Dog", slots = c(name = "character",  age = "numeric"))

# You want the age generic to automatically return 7x the slot in human years
setMethod("age", "Dog", function(x) {x@age*7} ) 
setMethod("age<-", "Dog", function(x, value) {
  x@age <- value
  x
})


spot <- new("Dog", name = "Spot Smith", age = 12)
age(spot)

We can do this with S3:

johnny <- structure(
  list(name = "Johnny Smith",
       age = 12),
  class = "s3_human")

stripe <- structure(
  list(name = "Stripe Smith",
       age = 12),
  class = "s3_dog")

s3_age <- function(x){UseMethod("s3_age")}
s3_age.s3_dog <- function(x){x$age*7}
s3_age.s3_human <- function(x){x$age}

s3_age(johnny)

## [1] 12

s3_age(stripe)

## [1] 84

And R6

Person <- R6Class("Person", list(
  name = NULL,
  age = NULL,
  initialize = function(name, age = NA) {
    stopifnot(is.character(name), length(name) == 1)
    stopifnot(is.numeric(age), length(age) == 1)
    self$name <- name
    self$age <- age
  }
))

john <- Person$new(name = "John Smith", age = 12)
john$age

[1] 12

Dog <- R6Class(
  "Dog",
  inherit = Person,
  public = list(
  initialize = function(name, age = NA) {
    super$initialize(name,age)
    self$age <- age * 7
  }
))
spot <- Dog$new(name = "Spot Smith", age = 12) 
spot$age

[1] 84