2  Introduction

As a statistician and data scientist, the objects I interact with most often in R are data frames and regression models. Actually R’s numbers and vectors are “objects” of a sort, but not in the sense that they need an object system. You can see the difference here:

# you can check whether something is an object with is.object()
is.object( c("a", "b", "c") )

[1] FALSE

is.object( 13 )

[1] FALSE

is.object( data.frame() )

[1] TRUE

# to be an object means to have a class:
attr( c("a", "b", "c"), "class")

NULL

attr( 13, "class")

NULL

attr( data.frame(), "class")

[1] "data.frame"

Objects are the nouns of a computer language. They provide two main functions: encapsulation and polymorphism.

• Encapsulation: many different parts can be bundled together into a single object.
• Polymorphism: you can do many different things using functions that share the same name (e.g., plot()).