Scribd
Upload
18 views

Analyzing Algorithms Part III

This document generalizes the number of instructions executed for an array of length n. It shows that the outer loop executes 2n + 2 instructions and the inner loop executes 7 * (n(n-1)/2) instructions. This gives a total complexity of O(n^2). A table demonstrates how the number of instructions increases polynomially as the size of the input array grows.

Uploaded by

RonaldMartinez
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
18 views

Analyzing Algorithms Part III

This document generalizes the number of instructions executed for an array of length n. It shows that the outer loop executes 2n + 2 instructions and the inner loop executes 7 * (n(n-1)/2) instructions. This gives a total complexity of O(n^2). A table demonstrates how the number of instructions increases polynomially as the size of the input array grows.

Uploaded by

RonaldMartinez
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Analyzing Algorithms Part III

In this lesson, we will work out a generalization for the number of instructions executed for
an array of length n

Generalizing Instruction Count

Let's try to generalize the running time for an array of size n.

Code Statement Cost

1.    for (int i = 0; i <


executed (2 x n) + 2 times.
input.length; i++) {

2.        int key = input[i]; executed n times

3.        int j = i - 1; executed n times

We already calculated the cost of each


iteration of the inner loop. We sum up the
costs across all the iterations. Note that
the inner loop will be executed n times
and each execution will result in
4.        while (j >= 0 &&
(iterations x 7)+2 instructions being
input[j] > key) {
executed. And the number of iterations
will successively increase from 0, 1, 2 all
the way up to(n-1). See the dry run in the
previous lesson to convince yourself of
the validity of this result.
(/learn)
[(0 x 7) + 2] + [(1 x 7) + 2] + [(2 x 7) + 2] + ...
5.             Inner loop [( (n-1) x 7) + 2]
statments = 2n + 7 * [ 0 + 1 + 2 + ... (n-1) ]
= 2n + 7 [ n(n-1)⁄2 ]

11.        }  //inner loop ends

12.   }

If we use the above formulas and apply an array of size 5 to them, then the
cost should match with what we calculated earlier.

T otal Cost = Outer loop instructions + Inner loop instructions


n ∗ (n − 1)
= [2 ∗ (n + 1) + 2n] + [2n + 7[ ]]
2
5 ∗ (5 − 1)
= [2 ∗ (5 + 1) + 10] + [10 + 7[ ]]
2
5∗4
= [12 + 10] + [10 + 7[ ]]
2
= 22 + [10 + 7 ∗ 10]

= 22 + 10 + 70

= 102 instructions

Summation of Series
(/learn)
We glossed over how we converted the sum of the series 0 + 1 + 2 ...(n-1) to
( 1)
n(n-1)⁄ ? However, even though we promised to steer clear of complicated
2
mathematics as much as possible, this is one of the cardinal summations that

you must know. Without a formal proof, remember that the sum of the first
k natural numbers can be represented as:

1 + 2 + 3 + 4⋯k
k ∗ (k + 1)
=
2

If you add the first 5 natural numbers the sum is:

k ∗ (k + 1)
=
2
5 ∗ (5 + 1)
=
2
30
=
2
= 15

However, our summation sums up to n-1 and includes a zero. We can drop
the zero because adding zero to anything yields the same value. We know:

k ∗ (k + 1)
1 + 2 + 3 + 4 ⋯ (k − 2) + (k − 1) + k =
2

(/learn)
We subtract k on both sides of the equation to get:
k ∗ (k + 1)
1 + 2 + 3 + 4 ⋯ (k − 2) + (k − 1) =
2
k 2 + k − 2k
1 + 2 + 3 + 4 ⋯ (k − 2) + (k − 1) =
2
k2 − k
1 + 2 + 3 + 4 ⋯ (k − 2) + (k − 1) =
2
k ∗ (k − 1)
1 + 2 + 3 + 4 ⋯ (k − 2) + (k − 1) =
2

Coming across this summation is very common in algorithmic analysis and,


without getting too technical, if you can identify this series, then you know
how to apply a formula to sum it up.

Running Time as n Increases

As the size of the array increases, so does the number of instructions


executed. Below is a table that shows the instructions executed for the worst
case as the size of the array increases.

Size of Array Instructions Executed

5 102

10 377

25 2252

75
(/learn) 19877
100 35252

250 219377

500 876252

1000 3502502

Size of Array vs Number of Instructions Executed

The intention of doing the dry run and accounting for each line of code
executed is to instill in the reader that the instruction count varies with the
input size. Calculating down to this detail is often not necessary and with
experience one can correctly recognize the complexity of various code
snippets. As you'll learn in the next chapters, we reason about algorithm
complexity in a ballpark sense, and calculating exact instruction counts is
almost always unnecessary.
(/learn)
←    Back Next    →
(/courses/big- (/courses/big-
MARK AS COMPLETED

o- o-
notation- notation-
for- for-
Analyzing Algorithms Part II
interviews- Problem Set 1
interviews-
and- and-
beyond/m27yx8vJ0Dr) beyond/gxnn6KylLor)
Ask a Question
Report
(https://discuss.educative.io/c/big-o-for-coding-interviews-and-beyond-c-h-
an Issue
afzal/basics-analyzing-algorithms-part-iii)

You might also like