Purpose
To code S4 Classes

> setClass("Distribution", representation(domain = "character"),
+     prototype = prototype(domain = "Statistics"))
[1] "Distribution"
> a <- new("Distribution")

Question
What happens when the prototype is a random number generator

> setClass("test", representation(val = "numeric"), prototype = prototype(val = rnorm(10)))
[1] "test"
> a <- new("test")
> b <- new("test")
> print(a)
An object of class "test"
Slot "val":
 [1]  1.3973270 -0.4279899  1.5370501 -0.2618448 -1.4394468  2.8641059
 [7]  1.9870864  0.5204481 -0.7454103 -0.6267789
> print(b)
An object of class "test"
Slot "val":
 [1]  1.3973270 -0.4279899  1.5370501 -0.2618448 -1.4394468  2.8641059
 [7]  1.9870864  0.5204481 -0.7454103 -0.6267789

Experimenting with superclass and subclass

> setClass("A_Vin", representation(s1 = "numeric"), prototype = prototype(s1 = 0))
[1] "A_Vin"
> setClass("W_Vin", representation(c1 = "character"))
[1] "W_Vin"
> setClass("WA_Vin", contains = (c("A_Vin", "W_Vin")))
[1] "WA_Vin"
> a <- new("A_Vin", s1 = 20)
> b <- new("W_Vin", c1 = "Vinay")
> c1 <- new("WA_Vin", b, a)
> print(c1)
An object of class "WA_Vin"
Slot "s1":
[1] 20


Slot "c1":
[1] "Vinay"

Experimenting with superclass and subclass

> setClass("A_Vin2", representation(s1 = "numeric"), prototype = prototype(s1 = 0))
[1] "A_Vin2"
> setClass("W_Vin2", representation(s1 = "character"))
[1] "W_Vin2"

Initialize method - same role as init constructor

> setClass("A_Vin3", representation(s1 = "numeric", s2 = "numeric"),
+     prototype = prototype(s1 = 0, s2 = 0))
[1] "A_Vin3"
> setMethod("initialize", "A_Vin3", function(.Object, ..., v1) {
+     callNextMethod(.Object, ..., s1 = v1, s2 = v1 * v1)
+ })
[1] "initialize"
> new("A_Vin3", v1 = 9)
An object of class "A_Vin3"
Slot "s1":
[1] 9


Slot "s2":
[1] 81

The above initialize method has a strange syntax. But I need to remember
anyways

Create Norm_V classes and Rand_Gen Classes

> setClass("Norm_V", representation = representation(mu = "numeric",
+     sigma = "numeric", rsample = "numeric"), prototype = prototype(mu = 0,
+     sigma = 1))
[1] "Norm_V"
> setClass("Rand_Gen", contains = "Norm_V", representation = representation(mu.r = "numeric",
+     sigma.r = "numeric", rsample.r = "numeric"))
[1] "Rand_Gen"

You can do all the jazz in the callNextMethod itself

> setMethod("initialize", "Norm_V", function(.Object, ..., mu,
+     sigma) {
+     callNextMethod(.Object, ..., mu = mu, sigma = sigma, rsample = rnorm(10,
+         mu, sigma))
+ })
[1] "initialize"
> new("Norm_V", mu = 0, sigma = 1)
An object of class "Norm_V"
Slot "mu":
[1] 0


Slot "sigma":
[1] 1


Slot "rsample":
 [1]  0.2159963 -0.3503482 -0.3815239  1.1360410  0.1242483 -1.9083046
 [7] -1.6307656  1.1471177 -0.6211488  2.0170386

You can use function inside the initialize

> setMethod("initialize", "Norm_V", function(.Object, ..., mu,
+     sigma) {
+     .Object <- callNextMethod()
+     .Object@mu <- mu
+     .Object@sigma <- sigma
+     .Object@rsample <- rnorm(10, mu, sigma)
+     .Object
+ })
[1] "initialize"
> new("Norm_V", mu = 0, sigma = 1)
An object of class "Norm_V"
Slot "mu":
[1] 0


Slot "sigma":
[1] 1


Slot "rsample":
 [1] -0.033404128 -1.231109141  0.351006319  0.690291731 -0.002945728
 [6] -1.122482248  0.260513590  0.208631694  0.197541945  0.724580051

Based on the the parent class, populate sample mean, sigma

> setMethod("initialize", "Rand_Gen", function(.Object, ...) {
+     .Object <- callNextMethod()
+     .Object@mu.r <- mean(.Object@rsample)
+     .Object@sigma.r <- sd(.Object@rsample)
+     .Object@rsample.r <- rnorm(10, .Object@mu.r, .Object@sigma.r)
+     .Object
+ })
[1] "initialize"
> x <- new("Rand_Gen", mu = 0, sigma = 1)
> print(x)
An object of class "Rand_Gen"
Slot "mu.r":
[1] 0.619733


Slot "sigma.r":
[1] 0.7967828


Slot "rsample.r":
 [1]  0.83961341 -0.07846612 -0.03680434  1.68289629  0.38536194  1.64090678
 [7] -0.76857785  0.85652711  1.49383104  1.45461954


Slot "mu":
[1] 0


Slot "sigma":
[1] 1


Slot "rsample":
 [1]  0.68504934  1.41442472  1.05487330  0.99589984 -0.65258381  0.07906395
 [7]  0.94952952  1.42991343  1.00729814 -0.76613796

The above is a code where you create a class which takes mu and sd It generates a random sample stores in the parent class The subclass then calculates the sample mean and sample sd and then generates a new sample based on sample estimate