Advanced R - Object oriented programming

# Advanced R - Object oriented programming
## Part 1 - Intro & Basics
### Hannes Oberreiter
### Cohort 5

---

## OOP - Intro

Three types to choose in R:

- **S3** (base)
  - base R, use it if possible! solving simple problems
  - *S4* more strict rewrite of S3
  - *RC* special case of S4, objects are mutable (can be modified in place)
- **R6**
  - model objects independently of R (web API)
  - similar to RC, will be discussed in R6 Section
- **S4**
  - planning a big project? go for S4
  - used by Bioconductor (https://www.bioconductor.org/developers/package-guidelines/)

---
class: left, top

### Two main paradigms

- **encapsulated OOP**
  - method belongs to objects or classes `object.method(arg1, arg2)`, eg. R6
- **functional OOP**
  - method belong to generic function `generic(object, arg2, arg3)`

*functional OOP* in my opinion should be called *multiple dispatch*:

> A method no longer “belongs” to a class. Methods belong primarily to the generic function for which they are defined.

<small>
How S4 Methods Work, John Chambers, August 30, 2006
https://developer.r-project.org/howMethodsWork.pdf
</small>

---
class: left, top

### Multiple Dispatch Example

```r
mde <- function(data, signature) {
  switch(signature,
    char = str_count(data),
    number = sum(data)
  )
}
```

```r
mde("test", "char")
```

```
## [1] 4
```

```r
mde(c(10, 20), "number")
```

```
## [1] 30
```

> The process of finding the correct method given a class is called **method dispatch**

> *functional OOP*: This is called functional because from the outside it looks like a regular function call, and internally the components are also functions..

---

## OOP - Glossary

- **polymorphism** - The Main Reason to use OOP
  - use the same function form for different types of input

```r
dplyr::glimpse(ggplot2::diamonds$cut)
```

```
##  Ord.factor w/ 5 levels "Fair"<"Good"<..: 5 4 2 4 2 3 3 3 1 3 ...
```

```r
base::summary(ggplot2::diamonds$cut)
```

```
##      Fair      Good Very Good   Premium     Ideal 
##      1610      4906     12082     13791     21551
```
--

```r
dplyr::glimpse(diamonds$carat)
```

```
##  num [1:53940] 0.23 0.21 0.23 0.29 0.31 0.24 0.24 0.26 0.22 0.23 ...
```

```r
base::summary(diamonds$carat)
```

```
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
##  0.2000  0.4000  0.7000  0.7979  1.0400  5.0100
```

---
class:center, middle

> Conceptually, a generic function extends the idea of a function in R by allowing different methods to be selected corresponding to the classes of the objects supplied as arguments in a call to the function.

---

## OOP - Glossary

> OOP allows us to extent to class to also add new summary functions for new types (which would not be possible for with if_else from the outside)

- **encapsulation**
  - user don't worry about the object details, they are hidden
--

- **class**
  - describes object
  - defined **fields** contained in every instance
- **method**
  - describes what the object of class can do

> Classes are hierarchy, if one method does not exist it will use the parent method. The child **inherits** its parents methods and traits. (method dispatch == process to find the correct method given a class)

---
class: center, middle

<div class="figure">
<img src="data:image/png;base64,#./img/inheritance.png" alt="By User:Crashed greek - User:Crashed greek, CC BY 4.0, https://commons.wikimedia.org/w/index.php?curid=64193508" width="" />
<p class="caption">By User:Crashed greek - User:Crashed greek, CC BY 4.0, https://commons.wikimedia.org/w/index.php?curid=64193508</p>
</div>

---
class: left, top

## Inheritance

Example:

- ordered factor *inherits* from a regular factor
- generalized linear model *inherits* from a linear model.

---
class: left, middle

## sloop  - sail the seas of OOP

```r
library(sloop)
sloop::otype(1:10)
```

```
## [1] "base"
```

```r
sloop::otype(mtcars)
```

```
## [1] "S3"
```

```r
mle_obj <- stats4::mle(function(x = 1) (x - 2) ^ 2)
sloop::otype(mle_obj)
```

```
## [1] "S4"
```

---
class: center, middle

## Base versus OO objects

![Base versus OO objects](data:image/png;base64,#https://d33wubrfki0l68.cloudfront.net/5d6b4926530f3da70f16b54dcdf1a55eb8fa3d71/703e0/diagrams/oo-venn.png)

---
class: left, middle

## Base versus OO objects

Base Objects do not have a class attribute to access

```r
attributes(1337L)$class
```

```
## NULL
```

```r
# class can be misleading with base objects
class(1337L)
```

```
## [1] "integer"
```

```r
# we can show class vector with sloop
# returns the implicit class that the S3 and S4 systems will use
sloop::s3_class(1337L)
```

```
## [1] "integer" "numeric"
```
> Note: `numeric` in S3 and S4 can mean either integer or double

---
class: left, middle

## Base versus OO objects 2

Only OO has a class attribute, but every object has a base type:

```r
typeof(1337L)
```

```
## [1] "integer"
```

```r
typeof(trees)
```

```
## [1] "list"
```

```r
attr(trees, "class")
```

```
## [1] "data.frame"
```

Base types are mostly written in C and use switch statements --> we cannot simply add new base types without manipulating source code.

---
class: left, middle

```r
# INTSXP, Vectors Chapter 3
typeof(1L)
```

```
## [1] "integer"
```

```r
# CLOSXP, Functions Chapter 6
typeof(mean)
```

```
## [1] "closure"
```

```r
# ENVSXP, Environments Chapter 7
typeof(globalenv())
```

```
## [1] "environment"
```

```r
# S4 Chapter 15 (don't inherit from any base type)
mle_obj <- stats4::mle(function(x = 1) (x - 2)^2)
typeof(mle_obj)
```

```
## [1] "S4"
```

---
class: left, top

## Primitive function

```r
getGeneric("+")
```

```
## standardGeneric for "+" defined from package "base"
##   belonging to group(s): Arith 
## 
## function (e1, e2) 
## standardGeneric("+", .Primitive("+"))
## <bytecode: 0x7fc522259800>
## <environment: 0x7fc522251bf0>
## Methods may be defined for arguments: e1, e2
## Use  showMethods("+")  for currently available ones.
```

```r
showMethods("+")
```

```
## Function: + (package base)
## e1="Date", e2="Duration"
## e1="Date", e2="Period"
## e1="difftime", e2="Duration"
## e1="Duration", e2="Date"
## e1="Duration", e2="difftime"
## e1="Duration", e2="Duration"
## e1="Duration", e2="numeric"
## e1="Duration", e2="POSIXct"
## e1="Duration", e2="POSIXlt"
## e1="numeric", e2="Duration"
## e1="numeric", e2="Period"
## e1="Period", e2="Date"
## e1="Period", e2="numeric"
## e1="Period", e2="Period"
## e1="Period", e2="POSIXct"
## e1="Period", e2="POSIXlt"
## e1="POSIXct", e2="Duration"
## e1="POSIXct", e2="Period"
## e1="POSIXlt", e2="Duration"
## e1="POSIXlt", e2="Period"
```