Recurrences
Recurrences and Running Time
• An equation or inequality that describes a function in
terms of its value on smaller inputs.
T(n) = T(n-1) + n
• Recurrences arise when an algorithm contains recursive
calls to itself
• What is the actual running time of the algorithm?
• Need to solve the recurrence
– Find an explicit formula of the expression
– Bound the recurrence by an expression that involves n
2
Example Recurrences
• T(n) = T(n-1) + n Θ(n2)
– Recursive algorithm that loops through the input to
eliminate one item
• T(n) = T(n/2) + c Θ(lgn)
– Recursive algorithm that halves the input in one step
• T(n) = T(n/2) + n Θ(n)
– Recursive algorithm that halves the input but must
examine every item in the input
• T(n) = 2T(n/2) + 1 Θ(n)
– Recursive algorithm that splits the input into 2 halves
and does a constant amount of other work
3
Recurrent Algorithms
BINARY-SEARCH
• for an ordered array A, finds if x is in the array A[lo…hi]
Alg.: BINARY-SEARCH (A, lo, hi, x)
1 2 3 4 5 6 7 8
if (lo > hi) 2 3 5 7 9 10 11 12
return FALSE
mid (lo+hi)/2 mid
lo hi
if x = A[mid]
return TRUE
if ( x < A[mid] )
BINARY-SEARCH (A, lo, mid-1, x)
if ( x > A[mid] )
BINARY-SEARCH (A, mid+1, hi, x)
4
Example
• A[8] = {1, 2, 3, 4, 5, 7, 9, 11}
– lo = 1 hi = 8 x = 7
1 2 3 4 5 6 7 8
1 2 3 4 5 7 9 11 mid = 4, lo = 5, hi = 8
5 6 7 8
1 2 3 4 5 7 9 11 mid = 6, A[mid] = x
Found!
5
Another Example
• A[8] = {1, 2, 3, 4, 5, 7, 9, 11}
– lo = 1 hi = 8 x=6
1 2 3 4 5 6 7 8
1 2 3 4 5 7 9 11 mid = 4, lo = 5, hi = 8
low high
1 2 3 4 5 7 9 11 mid = 6, A[6] = 7, lo = 5, hi = 5
low high
1 2 3 4 5 7 9 11 mid = 5, A[5] = 5, lo = 6, hi = 5
NOT FOUND!
1 2 3 4 5 7 9 11
high low
6
Analysis of BINARY-SEARCH
Alg.: BINARY-SEARCH (A, lo, hi, x)
if (lo > hi) constant time: c1
return FALSE
mid (lo+hi)/2 constant time: c2
if x = A[mid] constant time: c3
return TRUE
if ( x < A[mid] )
BINARY-SEARCH (A, lo, mid-1, x) same problem of size n/2
if ( x > A[mid] )
BINARY-SEARCH (A, mid+1, hi, x) same problem of size n/2
• T(n) = c + T(n/2)
– T(n) – running time for an array of size n
7
Methods for Solving Recurrences
• Iteration method
• Substitution method
• Recursion tree method
• Master method
8
The Iteration Method
• Convert the recurrence into a summation and try
to bound it using known series
– Iterate the recurrence until the initial condition is
reached.
– Use back-substitution to express the recurrence in
terms of n and the initial (boundary) condition.
9
The Iteration Method
T(n) = c + T(n/2)
T(n) = c + T(n/2) T(n/2) = c + T(n/4)
= c + c + T(n/4) T(n/4) = c + T(n/8)
= c + c + c + T(n/8)
Assume n = 2k
T(n) = c + c + … + c + T(1)
k times
= clgn + T(1)
= Θ(lgn)
10
Iteration Method – Example
T(n) = n + 2T(n/2) Assume: n = 2k
T(n) = n + 2T(n/2) T(n/2) = n/2 + 2T(n/4)
= n + 2(n/2 + 2T(n/4))
= n + n + 4T(n/4)
= n + n + 4(n/4 + 2T(n/8))
= n + n + n + 8T(n/8)
… = in + 2iT(n/2i)
= kn + 2kT(1)
= nlgn + nT(1) = Θ(nlgn)
11
The substitution method
1. Guess a solution
2. Use induction to prove that the
solution works
12
Substitution method
• Guess a solution
– T(n) = O(g(n))
– Induction goal: apply the definition of the asymptotic notation
• T(n) ≤ d g(n), for some d > 0 and n ≥ n0
– Induction hypothesis: T(k) ≤ d g(k) for all k < n (strong induction)
• Prove the induction goal
– Use the induction hypothesis to find some values of the
constants d and n0 for which the induction goal holds
13
Example: Binary Search
T(n) = c + T(n/2)
• Guess: T(n) = O(lgn)
– Induction goal: T(n) ≤ d lgn, for some d and n ≥ n0
– Induction hypothesis: T(n/2) ≤ d lg(n/2)
• Proof of induction goal:
T(n) = T(n/2) + c ≤ d lg(n/2) + c
= d lgn – d + c ≤ d lgn
if: – d + c ≤ 0, d ≥ c
• Base case? 14
Example 2
T(n) = T(n-1) + n
• Guess: T(n) = O(n2)
– Induction goal: T(n) ≤ c n2, for some c and n ≥ n0
– Induction hypothesis: T(n-1) ≤ c(n-1)2 for all k < n
• Proof of induction goal:
T(n) = T(n-1) + n ≤ c (n-1)2 + n
= cn2 – (2cn – c - n) ≤ cn2
if: 2cn – c – n ≥ 0 c ≥ n/(2n-1) c ≥ 1/(2 – 1/n)
– For n ≥ 1 2 – 1/n ≥ 1 any c ≥ 1 will work
15
Example 3
T(n) = 2T(n/2) + n
• Guess: T(n) = O(nlgn)
– Induction goal: T(n) ≤ cn lgn, for some c and n ≥ n0
– Induction hypothesis: T(n/2) ≤ cn/2 lg(n/2)
• Proof of induction goal:
T(n) = 2T(n/2) + n ≤ 2c (n/2)lg(n/2) + n
= cn lgn – cn + n ≤ cn lgn
if: - cn + n ≤ 0 c ≥ 1
• Base case? 16
Changing variables
T(n) = 2T( n ) + lgn
– Rename: m = lgn n = 2m
T (2m) = 2T(2m/2) + m
– Rename: S(m) = T(2m)
S(m) = 2S(m/2) + m S(m) = O(mlgm)
(demonstrated before)
T(n) = T(2m) = S(m) = O(mlgm)=O(lgnlglgn)
Idea: transform the recurrence to one that you
have seen before
17
The recursion-tree method
Convert the recurrence into a tree:
– Each node represents the cost incurred at various
levels of recursion
– Sum up the costs of all levels
Used to “guess” a solution for the recurrence
18
Example 1
W(n) = 2W(n/2) + n2
• Subproblem size at level i is: n/2i
• Subproblem size hits 1 when 1 = n/2i i = lgn
• Cost of the problem at level i = (n/2i)2 No. of nodes at level i = 2i
• Total cost: lg n 1 2
n lg n 1
1
i
1
i
1
W ( n) 2
i 0
lg n
i
2 W (1) n
2
2
2
i 0
nn
2
2
i 0
O(n) n
1 1
O ( n) 2 n 2
2
W(n) = O(n2)
19
Example 2
E.g.: T(n) = 3T(n/4) + cn2
• Subproblem size at level i is: n/4i
• Subproblem size hits 1 when 1 = n/4i i = log4n
• Cost of a node at level i = c(n/4i)2
• Number of nodes at level i = 3i last level has 3log4n = nlog43 nodes
• Total cost:
log4 n 1 i
i
3 2 3 1
T ( n) cn n
16
log4 3
cn 2 n log4 3
i 0 16
3
cn 2 n log4 3 O(n 2 )
i 0
1
16
T(n) = O(n ) 2 20
Example 2 - Substitution
T(n) = 3T(n/4) + cn2
• Guess: T(n) = O(n2)
– Induction goal: T(n) ≤ dn2, for some d and n ≥ n0
– Induction hypothesis: T(n/4) ≤ d (n/4)2
• Proof of induction goal:
T(n) = 3T(n/4) + cn2
≤ 3d (n/4)2 + cn2
= (3/16) d n2 + cn2
≤ d n2 if: d ≥ (16/13)c
• Therefore: T(n) = O(n2)
21
Example 3 (simpler proof)
W(n) = W(n/3) + W(2n/3) + n
• The longest path from the root to
a leaf is:
n (2/3)n (2/3)2 n … 1
• Subproblem size hits 1 when
1 = (2/3)in i=log3/2n
• Cost of the problem at level i = n
• Total cost:
lg n
W (n) n n ... n(log 3/ 2 n) n O(n lg n)
3
lg
2
W(n) = O(nlgn) 22
Example 3
W(n) = W(n/3) + W(2n/3) + n
• The longest path from the root to
a leaf is:
n (2/3)n (2/3)2 n … 1
• Subproblem size hits 1 when
1 = (2/3)in i=log3/2n
• Cost of the problem at level i = n
• Total cost:
(log3 / 2 n ) 1
W (n) n n ... i 0
n 2(log3 / 2 n )W (1)
log3 / 2 n
lg n 1
n
i 0
1 nlog3 / 2 2 n log3/ 2 n O(n) n
lg 3/ 2
O(n)
lg 3/ 2
n lg n O(n)
W(n) = O(nlgn) 23
Example 3 - Substitution
W(n) = W(n/3) + W(2n/3) + O(n)
• Guess: W(n) = O(nlgn)
– Induction goal: W(n) ≤ dnlgn, for some d and n ≥ n0
– Induction hypothesis: W(k) ≤ d klgk for any K < n
(n/3, 2n/3)
• Proof of induction goal:
Try it out as an exercise!!
• T(n) = O(nlgn)
24
Master’s method
• “Cookbook” for solving recurrences of the form:
n
T (n) aT f (n)
b
where, a ≥ 1, b > 1, and f(n) > 0
Idea: compare f(n) with nlogba
• f(n) is asymptotically smaller or larger than nlogba by
a polynomial factor n
• f(n) is asymptotically equal with nlogba
25
Master’s method
• “Cookbook” for solving recurrences of the form:
n
T (n) aT f (n)
b
where, a ≥ 1, b > 1, and f(n) > 0
Case 1: if f(n) = O(nlogba -) for some > 0, then: T(n) = (nlogba)
Case 2: if f(n) = (nlogba), then: T(n) = (nlogba lgn)
Case 3: if f(n) = (nlogba +) for some > 0, and if
af(n/b) ≤ cf(n) for some c < 1 and all sufficiently large n, then:
T(n) = (f(n))
regularity condition
26
Examples
T(n) = 2T(n/2) + n
a = 2, b = 2, log22 = 1
Compare nlog22 with f(n) = n
f(n) = (n) Case 2
T(n) = (nlgn)
28
Examples
T(n) = 2T(n/2) + n2
a = 2, b = 2, log22 = 1
Compare n with f(n) = n2
f(n) = (n1+) Case 3 verify regularity cond.
a f(n/b) ≤ c f(n)
2 n2/4 ≤ c n2 c = ½ is a solution (c<1)
T(n) = (n2)
29
Examples (cont.)
T(n) = 2T(n/2) + n
a = 2, b = 2, log22 = 1
Compare n with f(n) = n1/2
f(n) = O(n1-) Case 1
T(n) = (n)
30
Examples
T(n) = 3T(n/4) + nlgn
a = 3, b = 4, log43 = 0.793
Compare n0.793 with f(n) = nlgn
f(n) = (nlog43+) Case 3
Check regularity condition:
3(n/4)lg(n/4) ≤ (3/4)nlgn = c f(n), c=3/4
T(n) = (nlgn)
31
Examples
T(n) = 2T(n/2) + nlgn
a = 2, b = 2, log22 = 1
• Compare n with f(n) = nlgn
– seems like case 3 should apply
• f(n) must be polynomially larger by a factor of n
• In this case it is only larger by a factor of lgn
32
