126
JAVA NOTES
DATA STRUCTURES AND ALGORITHMS
Terry Marris August 2001
15 RECURSION
15.1 LEARNING OUTCOMES
By the end of this lesson the student should be able to
• explain the principles of recursion
• trace a sequence of recursive method calls
• write and test simple recursive methods
15.2 PRE-REQUISITES
The student should be comfortable with using arrays and the linear search, binary search and
selection sort methods.
15.3 INTRODUCTION
We explore principles of recursion through worked examples.
127
15.4 POWER
We have probably worked out 25 something like this:
We started with the simplest case: 20 = 1
Then we multiplied the previous result by 2: 21 = 2 × 20 = 2
Then we multiplied the previous result by 2: 22 = 2 × 21 = 4
Then we multiplied the previous result by 2: 23 = 2 × 22 = 8
Then we multiplied the previous result by 2: 24 = 2 × 23 = 16
Then we multiplied the previous result by 2: 25 = 2 × 24 = 32
In general, 2n = 2 × 2n-1
provided n is an integer >= 0.
Now we can write the Java method to return the value of two raised to some power, n.
public static int powerOfTwo(int n)
/* returns two raised to some power n.
Requires n >= 0.
*/
{
if (n == 0) // the simple case
return 1;
else if (n > 0) // the recursive step,
return 2 * powerOfTwo(n-1); // one step closer to the
// simple case
return -1; // error, n cannot be negative
}
System.out.println(powerOfTwo(3)); results in 8 being displayed.
A recursive method is one that calls itself. The powerOfTwo(int n) method described above
makes a call to itself in powerOfTwo(n-1).
In the simplest form of recursion, we always have
(a) the simple case that ends the sequence of recursive calls and
(b) the recursive step, that is always one step closer to the simple case
Just to emphasise the principles, let us trace the sequence of method calls resulting from
System.out.println(powerOfTwo(3));.
128
powerOfTwo(3)
n=3
int powerOfTwo(int n)
{
if (n == 0)
return 1;
else if (n > 0) frame 1
return 2 * powerOfTwo(n-1);
}
n=2
int powerOfTwo(int n)
{
if (n == 0)
return 1;
else if (n > 0) frame 2
return 2 * powerOfTwo(n-1);
}
n=1
int powerOfTwo(int n)
{
if (n == 0)
return 1;
else if (n > 0)
frame 3 return 2 * powerOfTwo(n-1);
}
n=0
int powerOfTwo(int n)
{
if (n == 0)
return 1;
frame 4 else if (n > 0)
return 2 * powerOfTwo(n-1);
}
129
powerOfTwo(3) makes the first call to int powerofTwo(int) in frame 1.
In frame 1, n = 3. Since n > 0 the recursive call powerOfTwo(2) is made.
In frame 2, n = 2. Since n > 0 the recursive call powerOfTwo(1) is made.
In frame 3, n = 1. Since n > 0 the recursive call powerOfTwo(0) is made.
In frame 4, n = 0. So 1 is returned to frame 3.
In frame 3, powerOfTwo(0) is 1. So 2 * 1 = 2 is returned to frame 2.
In frame 2, powerOfTwo(1) is 2. So 2 * 2 = 4 is returned to frame 1.
In frame 1, powerOfTwo(2) is 4. So 2 * 4 = 8 is returned.
In general, calculation and return is suspended until the result from the next recursive call is
obtained.
Recursion is a way of thinking.
130
15.5 LINEAR SEARCH
The method contains() shown below performs a recursive linear search on an array of ints for
a given int, target. It searches through the array from the given fromIndex up to the given
toIndex inclusive.
The simple case occurs when fromIndex exceeds toIndex; then target is not in the array.
We check the first item in the array:
array[fromIndex] == target
before going on to search the rest of the array with the recursive call contains(array,
fromIndex+1, toIndex, target). The recursive call brings you one step closer to the simple
case. Since each call adds 1 to fromIndex there must come a time when fromIndex exceeds
toIndex.
public static boolean contains(int[] array,
int fromIndex, int toIndex,
int target)
{
if (fromIndex > toIndex)
return false;
return array[fromIndex] == target ||
contains(array, fromIndex+1, toIndex, target);
}
Given the array
int[] array = { 3, 5, 7, 11, 13 };
the call to contains(array, 0, 4, 7) results in
false || false || true || false || false being evaluated. The end result, of course, is true.
131
15.6 BINARY SEARCH
The binarySearch() method shown below is implemented using recursion. The method sorts
the given array of ints between the given indices, lo and hi.
There are two simple cases that terminate the sequence of recursive calls.
If lo exceeds hi then there is nothing left to search and the target, the int we are looking for, is
not in the array.
If the target is in the middle location of the array, then we have found what we were looking
for.
There are two possible recursive calls, each one a step closer to a simple case.
binarySearch(array, lo, middle-1, target) searches the left-hand part of the array.
binarySearch(array, middle+1, hi, target) searches the right hand part of the array.
public static boolean binarySearch(
int[] array, int lo, int hi, int target)
{
if (lo > hi)
return false;
int middle = (lo + hi) / 2;
if (array[middle] == target)
return true;
else if (target < array[middle])
return binarySearch(array, lo, middle-1, target);
return binarySearch(array, middle+1, hi, target);
}
132
15.7 SELECTION SORT
The selection sort method shown below is implemented using recursion.
We have two helper methods. swap() exchanges the values in the array at two given indices,
i and j. minLocation() returns the location of the least value in the give array between the two
indices from and to inclusive.
The simple case occurs when from exceeds to; there is no more array to sort.
Each recursive call brings you one step closer to the simple case. For each recursive call
made, from is advanced by one so there is successively smaller and smaller portions of the
array from which to select the least value and place it in its correct position.
public static void swap(int[] array, int i, int j)
{
int t = array[i];
array[i] = array[j];
array[j] = t;
}
public static int minLocation(int[] array, int from, int to)
{
int min = Integer.MAX_VALUE;
int minLoc = -1;
for (int i = from; i <= to; i++) {
if (array[i] < min) {
min = array[i];
minLoc = i;
}
}
return minLoc;
}
public static void selectionSort(
int[] array, int from, int to)
{
if (from > to)
return;
int minLoc = minLocation(array, from, to);
swap(array, from, minLoc);
selectionSort(array, from+1, to);
}
133
15.8 REVIEW
15.9 FURTHER READING
ARNOW D & WEISS G Introduction to Programming Using Java pp 461
LEWIS & LOFTUS Java Software Solution pp 468
In the next lesson we use recursion to implement binary search tree operations.
15.10 EXERCISES
1 Trace the sequence of recursive method calls for method Q shown below when the first
call to method Q is System.out.println(Q(5));
public static int Q(int n)
{
if (n == 1)
return 1;
return 13 + Q(n - 1);
134
2 Trace the sequence of recursive method calls for method X shown below when the first
call to method X is System.out.println(sum(2, 3)); Hence (or otherwise) give a descriptive
name to X.
public static int X(int m, int n)
{
if (n == 0)
return m;
if (n == 1)
return m + 1;
return X(m+1, n-1);
}
3 Trace the sequence of recursive method calls for method Y shown below for each of
(a) System.out.println(Y(13, 7));
(b) System.out.println(Y(14, 7));
(c) System.out.println(Y(15, 7));
Hence (or otherwise) give a descriptive name to Y.
public static int Y(int m, int n)
{
if (m < n)
return m;
return Y(m-n, n);
}
4 Develop and test a method that uses recursion to determine the result of mn where m and n
are both integers >= 0.
5 Test the recursive linear search method defined in §15.5 above. Remember to carefully
check boundaries.
6 Thoroughly test the recursive binary search method defined in §15.6 above (see §5.5).
Which implementation, the iterative one described in §5 .4 or the recursive one, do you find
easier to understand? Explain why.
7 Test the recursive selection sort method described in §15.7 above.
