Running Time (1.

1)
w Most algorithms transform
best case average case worst case
120 100

Analysis of Algorithms

Input

Algorithm

Output

input objects into output objects. w The running time of an algorithm typically grows with the input size. w Average case time is often difficult to determine. w We focus on the worst case running time.
n n

Running Time

80 60 40 20 0

An algorithm is a step-by-step procedure for solving a problem in a finite amount of time.

Easier to analyze Crucial to applications such as games, finance and robotics

Analysis of Algorithms

1000

2000

3000

4000

Input Size

Experimental Studies ( 1.6)

w Write a program
9000 8000 7000

Limitations of Experiments
w It is necessary to implement the
algorithm, which may be difficult

implementing the algorithm w Run the program with inputs of varying size and composition w Use a method like System.currentTimeMillis() to get an accurate measure of the actual running time w Plot the results

Time (ms)

6000 5000 4000 3000 2000 1000 0 0 50 100

w Results may not be indicative of the

running time on other inputs not included in the experiment. w In order to compare two algorithms, the same hardware and software environments must be used
Input Size
3 Analysis of Algorithms 4

Analysis of Algorithms

Theoretical Analysis
w Uses a high-level description of the
algorithm instead of an implementation w Characterizes running time as a function of the input size, n. w Takes into account all possible inputs w Allows us to evaluate the speed of an algorithm independent of the hardware/software environment
Analysis of Algorithms 5

Pseudocode (1.1)
w High-level description w w w w
Example: find max element of an array of an algorithm More structured than Algorithm arrayMax(A, n) English prose Input array A of n integers Less detailed than a Output maximum element of A program currentMax A[0] Preferred notation for for i 1 to n 1 do describing algorithms if A[i] > currentMax then Hides program design currentMax A[i] issues return currentMax
Analysis of Algorithms 6

Pseudocode Details
w Control flow
n n n n n

The Random Access Machine (RAM) Model

w A CPU w An potentially unbounded

w Method call
var.method (arg [, arg])

if then [else ] while do repeat until for do Indentation replaces braces

w Return value
return expression

w Expressions
Assignment (like = in Java) = Equality testing (like == in Java) n 2 Superscripts and other mathematical formatting allowed
7

w Method declaration
Algorithm method (arg [, arg]) Input Output

bank of memory cells, each of which can hold an arbitrary number or character

w Memory cells are numbered and accessing

any cell in memory takes unit time.
Analysis of Algorithms 8

Analysis of Algorithms

Primitive Operations
w Basic computations w w w w
performed by an algorithm Identifiable in pseudocode Largely independent from the programming language Exact definition not important (we will see why later) Assumed to take a constant amount of time in the RAM model
Analysis of Algorithms

Counting Primitive Operations (1.1)

w By inspecting the pseudocode, we can determine the
n

w Examples:
n

n n

Evaluating an expression Assigning a value to a variable Indexing into an array Calling a method Returning from a method

maximum number of primitive operations executed by an algorithm, as a function of the input size Algorithm arrayMax(A, n) currentMax A[0] for i 1 to n 1 do if A[i] > currentMax then currentMax A[i] { increment counter i } return currentMax # operations 2 2 +n 2(n 1) 2(n 1) 2(n 1) 1 Total 7n 1
10

Analysis of Algorithms

Estimating Running Time

w Algorithm arrayMax executes 7n 1 primitive
operations in the worst case. Define:
a = Time taken by the fastest primitive operation b = Time taken by the slowest primitive operation

Growth Rate of Running Time

w Changing the hardware/ software
environment
n n

w Let T(n) be worst-case time of arrayMax. Then

a (7n 1) T(n) b(7n 1)

Affects T(n) by a constant factor, but Does not alter the growth rate of T(n)

w The linear growth rate of the running

time T(n) is an intrinsic property of algorithm arrayMax
Analysis of Algorithms

w Hence, the running time T(n) is bounded by two

linear functions

Analysis of Algorithms

11

12

Growth Rates
w Growth rates of
functions:
n n n

Constant Factors
1E+30 1E+28 1E+26 1E+24 1E+22 1E+20 1E+18 1E+16 1E+14 1E+12 1E+10 1E+8 1E+6 1E+4 1E+2 1E+0 1E+0 Cubic Quadratic Linear

w The growth rate is

not affected by
n n

T (n )

w Examples
n n

w In a log-log chart,

the slope of the line corresponds to the growth rate of the function

102 n + 105 is a linear function 105 n2 + 108 n is a quadratic function

T ( n)

Linear n Quadratic n2 Cubic n 3

constant factors or lower-order terms

1E+26 1E+24 1E+22 1E+20 1E+18 1E+16 1E+14 1E+12 1E+10 1E+8 1E+6 1E+4 1E+2 1E+0 1E+0

Quadratic Quadratic Linear Linear

1E+2

1E+4 n

1E+6

1E+8

1E+10

1E+2

1E+4

1E+6

1E+8

1E+10

n
Analysis of Algorithms 13 Analysis of Algorithms

14

Big-Oh Notation (1.2)

w Given functions f(n) and
10,000
3n 2n+10 n

Big-Oh Example
1,000,000

g(n), we say that f(n) is 1,000 O(g(n)) if there are positive constants 100 c and n0 such that
10

w Example: the function

n2
n n n

n^2 100n 10n n

is not O(n)

100,000 10,000 1,000 100 10

f(n) cg(n) for n n0 w Example: 2n + 10 is O(n)

n n n n

2n + 10 cn (c 2) n 10 n 10/(c 2) Pick c = 3 and n 0 = 10

n 2 cn nc The above inequality cannot be satisfied since c must be a constant

1 1 10 100 1,000

1 1 10 100 1,000

n
15 Analysis of Algorithms 16

Analysis of Algorithms

More Big-Oh Examples

n

Big-Oh and Growth Rate

w The big-Oh notation gives an upper bound on the
growth rate of a function

7n-2
7n-2 is O(n) need c > 0 and n 0 1 such that 7n-2 cn for n n 0 this is true for c = 7 and n 0 = 1

w The statement f(n) is O(g(n)) means that the growth

rate of f(n) is no more than the growth rate of g(n) according to their growth rate Yes No Yes
Analysis of Algorithms

n 3n 3

+ 20n2 + 5

3n 3 + 20n2 + 5 is O(n3 ) need c > 0 and n 0 1 such that 3n3 + 20n 2 + 5 cn3 for n n 0 this is true for c = 4 and n 0 = 21

w We can use the big-Oh notation to rank functions

f(n) is O(g(n)) g(n) grows more f(n) grows more Same growth g(n) is O(f(n)) No Yes Yes
18

n3

log n + log log n

3 log n + log log n is O(log n) need c > 0 and n 0 1 such that 3 log n + log log n clog n for n n 0 this is true for c = 4 and n 0 = 2
Analysis of Algorithms 17

Big-Oh Rules
w If is f(n) a polynomial of degree d, then f(n) is
O(n d), i.e.,
n n

Asymptotic Algorithm Analysis

w The asymptotic analysis of an algorithm determines
the running time in big-Oh notation
n

w To perform the asymptotic analysis

n

Drop lower-order terms Drop constant factors Say 2n is O(n) instead of 2n is O(n2 )

We find the worst-case number of primitive operations executed as a function of the input size We express this function with big-Oh notation We determine that algorithm arrayMax executes at most 7n 1 primitive operations We say that algorithm arrayMax runs in O(n) time

w Use the smallest possible class of functions

n

w Example:
n n

w Use the simplest expression of the class

n

w Since constant factors and lower-order terms are

Say 3n + 5 is O(n) instead of + 5 is O(3n) 3n

Analysis of Algorithms 19

eventually dropped anyhow, we can disregard them when counting primitive operations
Analysis of Algorithms 20

Computing Prefix Averages

w We further illustrate
asymptotic analysis with two algorithms for prefix averages w The i-th prefix average of an array X is average of the first (i + 1) elements of X: A[i] = (X[0] + X[1] + + X[i])/(i+1)
35 30 25 20 15 10 5 0 1 2 3 4 5 6 7

Prefix Averages (Quadratic)

w The following algorithm computes prefix averages in
quadratic time by applying the definition Algorithm prefixAverages1(X, n) Input array X of n integers Output array A of prefix averages of X #operations A new array of n integers n for i 0 to n 1 do n s X[0] n for j 1 to i do 1 + 2 + + (n 1) s s + X[j] 1 + 2 + + (n 1) A[i] s / (i + 1) n return A 1
Analysis of Algorithms 22

X A

w Computing the array A of

prefix averages of another array X has applications to financial analysis

Analysis of Algorithms

21

Arithmetic Progression
w The running time of
7 6 5 4 3 2 1 0 1 2 3 4 5 6
23

Prefix Averages (Linear)

w The following algorithm computes prefix averages in
linear time by keeping a running sum Algorithm prefixAverages2(X, n) Input array X of n integers Output array A of prefix averages of X A new array of n integers s0 for i 0 to n 1 do s s + X[i] A[i] s / (i + 1) return A

prefixAverages1 is O(1 + 2 + + n) w The sum of the first n integers is n(n + 1) / 2

n

#operations

There is a simple visual proof of this fact

w Thus, algorithm

prefixAverages1 runs in O(n 2) time

n 1 n n n 1
24

w Algorithm prefixAverages2 runs in O(n) time

Analysis of Algorithms

Analysis of Algorithms

Math you need to Review

w Summations (Sec. 1.3.1) w Logarithms and Exponents (Sec. 1.3.2)
w properties of logarithms:
log b(xy) = logbx + logby log b (x/y) = log bx - log by log bxa = alogbx log ba = logx a/log x b w properties of exponentials: a(b+c) = aba c abc = (ab)c ab /ac = a(b-c) Proof techniques (Sec. 1.3.3) b = a loga b b c = a c*log a b Basic probability (Sec. 1.3.4)
Analysis of Algorithms 25

Relatives of Big-Oh
w big-Omega
f(n) is (g(n)) if there is a constant c > 0 and an integer constant n0 1 such that f(n) cg(n) for n n0 w big-Theta n f(n) is (g(n)) if there are constants c > 0 and c > 0 and an integer constant n 0 1 such that cg(n) f(n) cg(n) for n n0 w little-oh n f(n) is o(g(n)) if, for any constant c > 0, there is an integer constant n 0 0 such that f(n) cg(n) for n n 0 w little-omega n f(n) is (g(n)) if, for any constant c > 0, there is an integer constant n 0 0 such that f(n) cg(n) for n n 0
n

w w

Analysis of Algorithms

26

Intuition for Asymptotic Notation

Big-Oh n f(n) is O(g(n)) if f(n) is asymptotically less than or equal to g(n) big-Omega n f(n) is (g(n)) if f(n) is asymptotically greater than or equal to g(n) big-Theta n f(n) is (g(n)) if f(n) is asymptotically equal to g(n) little-oh n f(n) is o(g(n)) if f(n) is asymptotically strictly less than g(n) little-omega n f(n) is (g(n)) if is asymptotically strictly greater than g(n)

Example Uses of the Relatives of Big-Oh

n

5n 2 is (n2 ) f(n) is (g(n)) if there is a constant c > 0 and an integer constant n 0 1 such that f(n) cg(n) for n n0 let c = 5 and n0 = 1 5n 2 is (n) f(n) is (g(n)) if there is a constant c > 0 and an integer constant n 0 1 such that f(n) cg(n) for n n0 let c = 1 and n0 = 1 2 is (n) 5n f(n) is (g(n)) if, for any constant c > 0, there is an integer constant n0 0 such that f(n) cg(n) for n n0 need 5n02 cn0 given c, the n0 that satifies this is n0 c/5 0
Analysis of Algorithms 28

Analysis of Algorithms

27