For loop

Print the numbers 1 to 50 on the console.

for (nb in (1:50)){
  cat(nb,", ", sep = "")
}

Assume you have several machine learning algorithms rpart, knn, neuralnet, randomForest, lm and xgbTree. Iterate through these algorithms.

algos = c('rpart', 'knn', 'neuralnet', 'randomForest', 'lm','xgbTree')
for (alg in algos){
  cat(alg,", ", sep = "")
}

Now display the iteration number as well.

algos = c('rpart', 'knn', 'neuralnet', 'randomForest', 'lm','xgbTree')
nb = 1
for (alg in algos){
  cat(nb,". ", alg,"\n", sep = "")
  nb <- nb+1
}

This will come in “handy” when comparing algorithms.

Loops are often used for aggregation purposes. For instance, you can compute the sum or product of multiple numbers.

Write a for loop that computes the cumulative product using a for loop. That means, multiply the number \(e \in \{2, 4, \dots, 10\}\) and output as variable \(s\)

s = 1
for (e in 2*(1:5)){
  s = s*e
}
s

In order to decipher a code you will need seven Fibonacci numbers, where the first one is greater than two.

\[ F_1=1,~F_2=1,~F_n=F_{n-1}+F_{n-2} \]

F[1]=1; F[2]=1;
for (n in 3:10){
  F[n] = F[n-1] + F[n-2]
}
F[F>2]

While loop

energy = 10
while (energy > 5){
  cat('Energy level ', energy, '\n')
  energy = energy - 2
} 

Introduction

There are many built in functions already. What are typical base functions? For instance, sum(), mean(), min(),max() (run library(help = "base") for a comprehensive list).

Let us re-implement sum as a script using a for loop.

x = c(1,2,3,4) # given vector
s = 0 # our summation variable
for (k in 1:length(x)){
  s = s + x[k] # add element to sum
}
s  # display result

Now let us use use the above as function and name it: mySum()

# define a function
mySum <- function(x){
  s = 0
  for (k in 1:length(x)){
    s = s + x[k]
  }
  return(s)  
}

# use the function
x = c(1,2,3,4) # given vector
mySum(x) # call the function

Let us reflect on the above function. mySum is the function name. Then we assign <- a function block function(x){}, where x is an input variable. The function return() returns the results from within the function’s body to the environment (workspace) in which it was called.

myMin <- function(x){
  m = Inf # special means m is at infinity
  for (k in 1:length(x)){
    if (x[k]<m) m = x[k];
  }
  return(m)  
}

# use function
x=c(65,70,24,26,36,65,83,34,42,34)
myMin(x)

Several input/output variables

What if we have several input variables? Let us have a function that multiplies three numbers.

multi3 <- function(a,b,c){ return(a*b*c)}
multi3 (2,3,4)

What if we have several output variables? Let us return the variables a,b and c.

ret3 <- function(){
  L = null; # initialise empty list
  L$a = 1; L$b = 2; L$c = 3;
  return(L)
}
ret3()

Practicalities

Acknowledgment

This tutorial was created using RStudio, R, rmarkdown, and many other tools and libraries. The packages learnr and gradethis were particularly useful. I’m very grateful to Prof. Andy Field for sharing his disovr package, which allowed me to improve the style of this tutorial and get more familiar with learnr. Allison Horst wrote a very instructive blog “Teach R with learnr: a powerful tool for remote teaching”, which encouraged me to continue with learnr. By the way, I find her statistic illustrations amazing.