0% found this document useful (0 votes)
71 views

Lecture 2

This document discusses algorithm analysis and efficiency. It explains that algorithm analysis studies the computing resource requirements of algorithms, especially running time which is the most important resource. Empirical and theoretical analyses are introduced as two approaches to measure efficiency. Theoretical analysis determines resource needs mathematically and is machine-independent. Basic operations are defined and algorithms are analyzed by counting their basic operations. Examples show how to compute the time complexity of various algorithms as functions of input size. The concepts of best, worst, and average cases are also covered.

Uploaded by

miki
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
71 views

Lecture 2

This document discusses algorithm analysis and efficiency. It explains that algorithm analysis studies the computing resource requirements of algorithms, especially running time which is the most important resource. Empirical and theoretical analyses are introduced as two approaches to measure efficiency. Theoretical analysis determines resource needs mathematically and is machine-independent. Basic operations are defined and algorithms are analyzed by counting their basic operations. Examples show how to compute the time complexity of various algorithms as functions of input size. The concepts of best, worst, and average cases are also covered.

Uploaded by

miki
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 41

Lecture Two- Algorithm Analysis

Data Structure and Algorithm


Analysis
Algorithm
analysis
 Studies computing resource requirements of different
algorithms
 Computing Resources
 Running time (Most precious)
 Memory usage
 Communication bandwidth etc
 Why need algorithm analysis ?
 Writing a working program is not good enough
 The program may be inefficient!
 If the program is run on a large data set, then the running
time becomes an issue
 Goal is to pick up an efficient algorithm for the problem at
2 hand
Reasons to perform analyze
algorithms
 It enables us to:

 Predict performance of algorithms

 Compare algorithms.

 Provide guarantees on running time/space of algorithms

 Understand theoretical basis.

 Primary practical reason: avoid performance bugs.

 client gets poor performance because programmer did not

understand performance characteristics

3
How to Measure
Efficiency/performance?
 Two approaches to measure algorithms
efficiency/performance
Empirical
Implement the algorithms and
Trying them on different instances
of input
Use/plot actual clock time to pick
one
Theoretical/Asymptotic Analysis
Determine quantity of resource required
4 mathematically needed by each
Example-
Empirical Actual clock time
Input size

5
Drawbacks of empirical
methods
 It is difficult to use actual clock because clock time varies
based on
Specific processor speed
Current processor load
Specific data for a particular run of the program
Input size
Input properties
Programming language (C++, java, python …)
The programmer (You, Me, Billgate …)
Operating environment/platform (PC, sun, smartphone
etc)
6 Therefore, it is quite machine dependent
Machine independent
analysis
 Critical resources:
Time, Space (disk, RAM), Programmer’s effort, Ease
of use (user’s effort).
 Factors affecting running time:
 System dependent effects.
Hardware: CPU,
memory, cache, …
Software: compiler,
interpreter, garbage
collector, …
System: operating system,
7 Input data/ Problem
network, other apps, …
Machine independent
analysis…
 For most algorithms, running time depends on “size” of

the input.
Size is often the number of inputs processed

Example:- in searching problem, size is the no of items

to be sorted
 Running time is expressed as T(n) for some function T

on input size n.
8
Machine independent
analysis
 Efficiency of an algorithm is measured in terms of the number of basic
operations it performs.
 Not based on actual time-clock
 We assume that every basic operation takes constant time.
 Arbitrary time
 Examples of Basic Operations:
 Single Arithmetic Operation (Addition, Subtraction,
Multiplication)
 Assignment Operation
 Single Input/Output Operation
 Single Boolean Operation
 Function Return
 We do not distinguish between the basic operations.
 Examples of Non-basic Operations are
9
 Sorting, Searching.
Examples: Count of Basic Operations
T(n)
Sample Code
int
count()
{ Int k=0;
cout<< “Enter an
integer”; cin>>n;
for (i = 0;i < n;i+
+) k = k+1;
return 0;
}
10
Examples: Count of Basic Operations
T(n)
Sample Code
Count of Basic Operations (Time
int
Units)
count()
{  1 for the assignment statement: int k=0

Int k=0; “Enter an integer”; 1 for the output statement.


cout<<
 1 for the input statement.
cin>>n;

 In the for loop:
for (i = 0;i < n;i+
 1 assignment, n+1tests, and n
+) k = k+1; increments.
return 0;  n loops of 2 units for an assignment,
} and an addition.
 1 for the return statement.

 T (n) = 1+1+1+(1+n+1+n)+2n+1
11 =
4n+6
Examples: Count of Basic Operations
T(n)
int total(int n)
{
Int sum=0;
for (int i=1;i<=n;i++)
sum=sum+i;
return sum;
}

12
Examples: Count of Basic Operations
T(n)
Count of Basic Operations
Sample Code Units
(Time
)
int total(int n)
{
 1 for the assignment statement: int sum=0
Int sum=0;
 In the for loop:
for (inti=1;i<=n;i+  1 assignment, n+1tests, and n increments.
+)  n loops of 2 units for an assignment, and
sum=sum+i an addition.
 1 for the return statement.
; return sum;
}
 T (n) = 1+ (1+n+1+n)+2n+1 =
4n+4

13
Examples: Count of Basic Operations
T(n) void func()
{
Int
x=0;
Int i=0;
Int
j=1;
cout<<
“Enter
an
Integer
value”;
cin>>n;
while
(i<n){
x++;
14 } i++;
Examples: Count of Basic Operations
T(n)
Sample Code Count of Basic Operations (Time
void Units)
func()
{
Int x=0;  1 for the first assignment statement: x=0;
Int i=0;  1 for the second assignment statement: i=0;
Int j=1;  1 for the third assignment statement: j=1;
cout<< “Enter an Integer value”;  1 for the output statement.
cin>>n;  1 for the input statement.
while (i<n){  In the first while loop:
x++;  n+1tests
i++;  n loops of 2 units for the two increment (addition)
operations
}
while (j<n)
 In the second while loop:
 n tests
{  n-1 increments
j++;
 T (n) = 1+1+1+1+1+n+1+2n+n+n-1 = 5n+5
}
}

15
Examples: Count of Basic Operations
T(n)
Sample Code
int sum (int
n)
{ int partial_sum= 0;
for (int i = 1; i < = n; i++)
partial_sum= partial_sum+ (i * i *
i);
return partial_sum;
}

16
Examples: Count of Basic Operations
T(n)
Sample code Count of Basic Operations (Time
Units)
int sum (int n)
{
 1 for the assignment.
int partial_sum= 0;
for (int i = 1; i < = n; i++)  1 assignment, n+1tests, and n increments.

partial_sum= partial_sum+ (i * i  n loops of 4 units for an assignment, an


addition, and two multiplications.
* i);
 1 for the return statement.
return partial_sum;
}
 T (n) = 1+(1+n+1+n)+4n+1 =
6n+4

17
Example: Computation of Run-
time
Suppose we have hardware capable of executing 106
instructions per second. How long would it take to
execute an algorithm whose complexity function was
T
(n) = 2n2 on an input size of n =108?

18
Example: Computation of Run-
time
 Suppose we have hardware capable of executing 10 6 instructions

per second. How long would it take to execute an algorithm


whose complexity function was T (n) = 2n2 on an input size of
n
=108?

The total number of operations to be performed would be


T(108):
T(108) = 2*(108)2 =2*1016
The required number of seconds would be given
by T(108)/106 so:
Running time = 2*1016/106 = 2*1010
The number of seconds per day is 86,400 so this is about
231,480 days (634 years).
19
Types of Algorithm complexity
analysis
 Best case.
 Lower bound on cost.
 Determined by “easiest” input.
 Provides a goal for all inputs.
 Worst case.
 Upper bound on cost.
 Determined by “most difficult” input.
 Provides a guarantee for all inputs.
 Average case. Expected cost for random
input.
 Need a model for “random” input.
 Provides a way to predict performance.

20
Best, Worst and Average
Cases
 Not all inputs of a given size take the same time.
 Sequential search for K in an array of n integers:
 Begin at first element in array and look at each element in turn
until K is found.
 Best Case: [Find at first position: 1 compare]
 Worst Case: [Find at last position: n compares]
 Average Case: [(n + 1)/2 compares]
 While average time seems to be the fairest measure, it may be difficult
to determine.
 Depends on distribution. Assumption for above analysis: Equally
likely at any position.
 When is worst case time important?
25  algorithms for time-critical systems
Order of Growth and Asymptotic
Analysis
 Suppose an algorithm for processing a retail store’s inventory takes:
 10,000 milliseconds to read the initial inventory from disk, and then
 10 milliseconds to process each transaction (items acquired or sold).
 Processing n transactions takes (10,000 + 10 n) milliseconds.
 Even though 10,000 > > 10, the "10 n" term will be more important
if the number of transactions is very large.
 We also know that these coefficients will change if we buy a
faster computer or disk drive, or use a different language or
compiler.
 we want to ignore constant factors (which get smaller and smaller
as technology improves)
 In fact, we will not worry about the exact values, but will look at
“broad classes" of values.
22
Growth rates
 The growth rate for an algorithm is the rate at which the cost of
the algorithm grows as the size of its input grows.

23
Rate of
Growth
 Consider the example of buying elephants and
goldfish: Cost: cost_of_elephants +
cost_of_goldfish Cost ~ cost_of_elephants
(approximation)
since the cost of the gold fish is insignificant when compared with cost
of elephants
 Similarly, the low order terms in a function are relatively insignificant
for n4 + 100n2 + 10n + 50 ~ n4
large
i.e., n that n4 + 100n2 + 10n + 50 and n4 have the same rate
we say
of growth
More Examples: fB(n)=n2+1 ~
n2

fA(n)=30n+8 ~
24
n
Visualizing Orders of Growth
 On a graph, as you go to the right, a faster
growing function eventually becomes larger...
Value of function 

fA(n)=30n+8

fB(n)=n2+1

Increasing n 
29
Asymptotic
analysis
 Refers to the study of an algorithm as the input size "gets big"
or reaches a limit.
 To compare two algorithms with running times f(n) and g(n), we need a
rough measure that characterizes how fast each function grows-
growth rate.
 Ignore constants [especially when input size very large]
 But constants may have impact on small input size
 Several notations are used to describe the running-time equation for an
algorithm.
 Big-Oh (O), Little-Oh (o)
 Big-Omega (Ω), Little-Omega(w)
 Theta Notation()

26
Big-Oh
Notation
 Definition
For f(n) a non-negatively valued function, f(n) is
in set O(g(n)) if there exist two positive
constants c and n0 such that f(n)≤cg(n)for all n>n0 .
 Usage: The algorithm is in O(n2) in [best ,average,
worst] case.
 Meaning: For all data sets big enough (i.e., n > n0), the
algorithm always executes in less than cg (n) steps [in
best, average or worst case].

27
Big-Oh Notation -
Visually

28
Big-
O Demonstrating that a function f(n) is in big-O of a function
g(n) requires that we find specific constants c and no for
which the inequality holds.

 The following points are facts that you can use for Big-Oh
problems:
1<= n for all n >= 1
n <= n2 for all n >= 1
2n < = n! for all n >= 4
log2n <= n for all n >= 2
n <= nlog2n for all n >= 2
34
Examples
f(n) = 10n + 5 and g(n) = n. Show that f(n) is in
O(g(n)).
To show that f(n) is O(g(n)) we must show constants
c and no such that
f(n) < = c.g(n) for all n > = no
10n + 5 < = c.n for all n > = no
Try c = 15.Then we need to show that 10n + 5
<= 15n
Solving for n we get: 5 < 5n or 1 < = n.
So f(n) =10n + 5 < = 15.g(n) for all n > = 1.
35  (c = 15, no = 1).
Examples
2n2 = O(n3): 2n2 ≤ cn3  2 ≤ cn  c = 1 and n0= 2

 n2 = O(n2): n2 ≤ cn2  c ≥ 1  c = 1 and n0= 1

 1000n2+1000n = O(n2):

1000n2+1000n ≤ 1000n2+ n2 =1001n2 c=1001 and n0 = 1000

 n = O(n2): n ≤ cn2  cn ≥ 1  c = 1 and n0= 1

31
More
Examples
Show that 30n+8 is O(n).
Show c,n0: 30n+8  cn,
n>n0.

Let c=31, n0=8.


Assume n>n0=8. Then
 cn = 31n = 30n + n >
30n+8,
So 30n+8 < cn.

32
No
Uniqueness
 There is no unique set of values for n 0 and c in proving the asymptotic
bounds
 Prove that 100n + 5 = O(n2)
100n + 5 ≤ 100n + n = 101n ≤ 101n2
for all n ≥ 5
n0 = 5 and c = 101 is a solution
100n + 5 ≤ 100n + 5n = 105n ≤ 105n2
for all n ≥ 1
n0 = 1 and c = 105 is also a solution
Must find SOME constants c and n0 that satisfy the asymptotic notation
relation
33
Order of common
functions
Notation Name Example

O(1) Constant Adding two numbers, c=a+b

O(log n) Logarithmic Finding an item in a sorted array with a binary search or a search
tree (best case)

O(n) Linear Finding an item in an unsorted list or a malformed tree (worst


case); adding two n-digit numbers

O(nlogn) Linearithmic Performing a Fast Fourier transform; heap sort, quick sort (best
case), or merge sort

O(n2) Quadratic Multiplying two n-digit numbers by a simple algorithm; adding


two n×n matrices; bubble sort (worst case or naive
implementation), shell sort, quick sort (worst case), or insertion
sort

34
Some properties of Big-
O
 Constant factors are may be ignored
For all k>0, kf is O(f)
 The growth rate of a sum of terms is the growth rate of
its fastest growing term.
Ex, an3 + bn2 is O(n3 )
 The growth rate of a polynomial is given by the
growth rate of its leading term
If f is a polynomial of degree d, then f is O(nd)

35
Implication of Big-Oh
notation
 We use Big-Oh notation to say how slowly code might run
as its input grows.
 Suppose we know that our algorithm uses at most O(f(n))
basic steps for any n inputs, and n is sufficiently large,
then we know that our algorithm will terminate after
executing at most constant times f(n) basic steps.
 We know that a basic step takes a constant time in a
machine.
 Hence, our algorithm will terminate in a constant times
f(n) units of time, for all large n.
36
2. Big – Omega
Definition : The function T(n) is Ω(F(n)) if there exist
Constants c and N such that
T(n) ≥ c.F(n) for all n ≥ N.
As n increases T(n) grows no slower than F(n) or in
the long run (for large n) T grows at least as fast as F
It computes the tight lower bound of T(n).
Describes the best case analysis
Examples
Find F(n) such that T(n) = Ω(F(n))
for T(n) = 3n+5
F(n) = n as 3n+5>=c.n for c=1 and n=0
lets check f(n)=3n+2 is in Ω(g(n)) for n>=k nd k =1
and c=1
3. Big Theta Notation
Definition: The function T(n) is Θ(F(n)) if there
exist constants c1, c2 and N such that c1.F(n) ≤
T(n) ≤ c2.F(n) for all n ≥ N.
As n increases, T(n) grows as fast as F(n)
It computes the tight optimal bound of T(n).
Describes the average case analysis.
Example:
Find F(n) such that T(n) = Θ(F(n)) for T(n)=2n2.
Solution: c1n2 ≤ 2n2 ≤ c2n2
F(n)=n2 because for c1=1 and c2=3,
n2 ≤ 2n2 ≤ 3n2 for all n.
Find f(n) such that T(n)= Θ(F(n)) for T(n)=5n+6
Big O Analysis
.
End of Lecture 2

Next Lecture:-Simple Sorting and Searching


Algorithms

41

You might also like