The Big R-Book

♣E♣Answers to Selected Questions

Question 1 on page 14

A suggestion could be the distribution of IQ (as in intelligence quotient). This seems to follow more or less a normal distribution with a standard deviation of 15 points.

Question 2 on page 31

All operations in R are by default done on objects element by element. Hence, the following code is sufficient to create an object nottemC that contains all temperatures in degrees Celsius.

nottemC <- 5/9 * nottem - 32

While this concept makes such calculations very easy, it must be noted that this will – by default – not produce a matrix or vector product as you might expect.

Question 3 on page 34

There are many ways to tackle this problem. Here is one.

dotproduct <- function (m1, m2) {
   if(!is.matrix(m1) || !is.matrix(m2)) { 
   print(“ERROR 1: m1 and m2 must be matrices.”); 
   return(-1);
}
if (ncol(m1) != nrow(m2)) { 
    print(“ERROR 2: ncol(m1) must match nrow(m2).”) 
    return(-2);
}
# Note that we do not check for values NA or Inf. In R they work
# as one would expect.
m <- matrix(nrow=nrow(m1), ncol=ncol(m2))
for (k in 1:nrow(m1)) { 
   for (l in 1:ncol(m1)){ 
       m[k,l] <- 0 
       for (n in 1:ncol(m1)) {  
            m[k,l] <- m[k,l] ...