Truth be told, you won’t be using object-oriented programming in most day-to-day R programming. Most analyses you do in R involve the transformation of data, typically implemented as some sort of data flow, which is best captured by functional programming. When you write such pipelines, you will probably be using polymorphic functions, but you rarely need to create your own classes. If you need to implement a new statistical model, however, you usually do want to create a new class.
A lot of data analysis requires that you infer parameters of interest or you build a model to predict ...
