Introduction
This chapter lays the groundwork for the material covered in the course and sets out the expectations for the
readers.
The Interview Room
You're in an interview room, and you've just completed white-boarding the
interviewer's question. Inside, you feel ecstatic; It's your dream company and
you nailed the question. However, just when you think the coast is clear, the
interviewer follows up with a question on the time or space complexity of
your solution. The blank drawn across your face gives away your ineptness at
analyzing algorithm complexity in space or time. In this common interview
situation, most computer science graduates are still able to muster up a
reasonable answer. For candidates lacking the formal education, however,
reasoning about algorithms in terms of the big Oh notation becomes a big uh
oh!.
The intended readership of this course includes folks who are graduates of
boot-camps, career-switchers into programming jobs or just established
industry veterans who want a quick revision on the topic of complexity
theory. The course uses layman's terms to describe the concepts commonly
used in the industry while intentionally avoiding the intricate mathematical
gymnastics often involved in analyzing algorithms so that the subject material
is comprehensible for readers at all skill levels. At the same time, this course
may not deliver great value to individuals who already have a firm grasp of
algorithms and data-structures or to those that hold a degree in computer
science.
Why does complexity matter?
If you are working on a class assignment or a customer application with a few
dozen users, you can take the liberty to ignore the space and time complexity
of your solution. However, the moment you step into the professional world
and write software with strict SLAs (service level agreements) or for millions
of users, your choice of algorithm and data-structures starts to matter. A well
designed and thought out software will scale elegantly and set itself apart
from poorly written competitors.
Usually, the 101 example used to teach complexity is sorting of integers. Below
are three example programs that sort integers using different sorting
algorithms.
Bubble Sort
import java.util.Random;
class Demonstration {
static int SIZE = 10000;
static Random random = new Random(System.currentTimeMillis());
static int[] input = new int[SIZE];
public static void main( String args[] ) {
createTestData();
long start = System.currentTimeMillis();
bubbleSort(input);
long end = System.currentTimeMillis();
System.out.println("Time taken = " + (end - start));
}
static void bubbleSort(int[] input) {
for (int i = 0; i < SIZE; i++) {
for (int j = 0; j < SIZE - 1; j++) {
if (input[j] > input[j + 1]) {
int tmp = input[j];
input[j] = input[j + 1];
input[j + 1] = tmp;
}
}
}
}
static void createTestData() {
for (int i = 0; i < SIZE; i++) {
input[i] = random.nextInt(10000);
}
}
}
Merge Sort
import java.util.Random;
class Demonstration {
static int SIZE = 5000;
static Random random = new Random(System.currentTimeMillis());
static int[] input = new int[SIZE];
static int[] scratch = new int[SIZE];
public static void main( String args[] ) {
createTestData();
long start = System.currentTimeMillis();
mergeSort(0, input.length - 1, input);
long end = System.currentTimeMillis();
System.out.println("Time taken = " + (end - start));
}
static void mergeSort(int start, int end, int[] input) {
if (start == end) {
return;
}
int mid = (start + end) / 2;
// sort first half
mergeSort(start, mid, input);
// sort second half
mergeSort(mid + 1, end, input);
// merge the two sorted arrays
int i = start;
int j = mid + 1;
int k;
for (k = start; k <= end; k++) {
scratch[k] = input[k];
}
k = start;
while (k <= end) {
if (i <= mid && j <= end) {
input[k] = Math.min(scratch[i], scratch[j]);
if (input[k] == scratch[i]) {
i++;
} else {
j++;
}
} else if (i <= mid && j > end) {
input[k] = scratch[i];
i++;
} else {
input[k] = scratch[j];
j++;
}
k++;
}
}
static void createTestData() {
for (int i = 0; i < SIZE; i++) {
input[i] = random.nextInt(1000);
}
}
}
Merge Sort
Dual Pivot Quicksort
import java.util.Random;
import java.util.Arrays;
class Demonstration {
static int SIZE = 5000;
static Random random = new Random(System.currentTimeMillis());
static int[] input = new int[SIZE];
public static void main( String args[] ) {
createTestData();
long start = System.currentTimeMillis();
Arrays.sort(input);
long end = System.currentTimeMillis();
System.out.println("Time taken = " + (end - start));
}
static void createTestData() {
for (int i = 0; i < SIZE; i++) {
input[i] = random.nextInt(10000);
}
}
}
Java's Dual Pivot Quicksort
The code implementations above run on datasets of 5000 elements. Below are
some crude tests (run on a Macbook) for the three algorithms. All times are in
milliseconds
Java's
Size Bubble Sort Merge Sort
util.Arrays.sort
5000 90 9 4
10000 299 15 17
100000 24719 26 33
1000000 2474482 150 113
The results tabulated above should provide plenty of motivation for software
developers to increase their understanding of algorithmic performance.
Bubble sort progressively degrades in performance as the input size increases,
whereas the other two algorithms perform roughly the same. We'll have more
to say about their performances in the next chapter.
Time & Space
The astute reader would notice the use of an additional array in the merge
sort code, named scratch array. This is the classic trade-off in the computer
science realm. Throwing more space at a problem allows us to bring down the
execution time, and vice versa. Generally, time and space have an inversely
proportional relationship with each other; if space increases then execution
time decreases, and if execution time increases then space decreases.
