CS 498ABD: Algorithms for Big Data

Introduction to Randomized
Algorithms: QuickSort
Lecture 2
August 27, 2020

Chandra (UIUC) CS498ABD 1 Fall 2020 1 / 33

Outline

Today
Randomized Algorithms – Two types
Las Vegas
Monte Carlo
Randomized Quick Sort

Chandra (UIUC) CS498ABD 2 Fall 2020 2 / 33

 Part I

Introduction to Randomized
Algorithms

Chandra (UIUC) CS498ABD 3 Fall 2020 3 / 33

Randomized Algorithms

Input x Output y
Deterministic Algorithm

Chandra (UIUC) CS498ABD 4 Fall 2020 4 / 33

Randomized Algorithms

Input x Output y
Deterministic Algorithm

random bits r

Input x Output yr
Randomized Algorithm

Chandra (UIUC) CS498ABD 4 Fall 2020 4 / 33

Example: Randomized QuickSort

QuickSort ?
1 Pick a pivot element from array
2 Split array into 3 subarrays: those smaller than pivot, those
larger than pivot, and the pivot itself.
3 Recursively sort the subarrays, and concatenate them.

Randomized QuickSort
1 Pick a pivot element uniformly at random from the array
2 Split array into 3 subarrays: those smaller than pivot, those
larger than pivot, and the pivot itself.
3 Recursively sort the subarrays, and concatenate them.

Chandra (UIUC) CS498ABD 5 Fall 2020 5 / 33

Example: Randomized Quicksort
Recall: QuickSort can take ⌦(n2 ) time to sort array of size n.
=

R , 2, 3 , U , 5
,
6, -
. .

,
100
-

I
too
0 2=3
-

,
-

= I \
Of 2 3, too
-
. -

Tnk n + Tcu -17 n

Chandra (UIUC) CS498ABD 6 Fall 2020 6 / 33

Example: Randomized Quicksort
Recall: QuickSort can take ⌦(n2 ) time to sort array of size n.
Theorem
Randomized QuickSort sorts a given array of length n in
O
=
O(n log n) expected time.

witu¥dihG .

Chandra (UIUC) CS498ABD 6 Fall 2020 6 / 33

Example: Randomized Quicksort
Recall: QuickSort can take ⌦(n2 ) time to sort array of size n.
Theorem
Randomized QuickSort sorts a given array of length n in
O(n log n) expected time.

Note: On every input randomized QuickSort takes O(n log n) time

in expectation. On every input it may take ⌦(n2 ) time with some
small probability.

Chandra (UIUC) CS498ABD 6 Fall 2020 6 / 33

Example: Verifying Matrix Multiplication

Problem
Given three n ⇥ n matrices A, B, C is AB = C ?

Chandra (UIUC) CS498ABD 7 Fall 2020 7 / 33

Example: Verifying Matrix Multiplication

Problem
Given three n ⇥ n matrices A, B, C is AB = C ?

Deterministic algorithm:
1 Multiply A and B and check if equal to C .
2 Running time? O(n3 ) by straight forward approach. O(n2.37 )
with fast matrix multiplication (complicated and impractical).
' ' "

E'
"

a
I -

an:c .

Chandra (UIUC) CS498ABD 7 Fall 2020 7 / 33

Example: Verifying Matrix Multiplication
Problem
Given three n ⇥ n matrices A, B, C is AB = C ?

Randomized algorithm:
1 Pick a random n ⇥ 1 vector r .

-ux I
2 Return the answer of the equality ABr = Cr .
3 Running time?

.
It
Chandra (UIUC) CS498ABD 8 Fall 2020 8 / 33
Example: Verifying Matrix Multiplication
Problem
Given three n ⇥ n matrices A, B, C is AB = C ?

Randomized algorithm:
1 Pick a random n ⇥ 1 vector r .
2 Return the answer of the equality ABr = Cr .
3 Running time? O(n2 )!

Chandra (UIUC) CS498ABD 8 Fall 2020 8 / 33

Example: Verifying Matrix Multiplication
Problem
Given three n ⇥ n matrices A, B, C is AB = C ?

Randomized algorithm:
1 Pick a random n ⇥ 1 vector r .
2 Return the answer of the equality ABr = Cr .
3 Running time? O(n2 )!
Theorem
If AB = C then the algorithm will always say YES. If AB 6= C then
the algorithm will say YES with probability at most 1/2. Can repeat
the algorithm 100 times independently to reduce the probability of a
false positive to 1/2100 .
Chandra (UIUC) CS498ABD 8 Fall 2020 8 / 33
Why randomized algorithms?
1 Many many applications in algorithms, data structures and
computer science!
2 In some cases only known algorithms are randomized or
randomness is provably necessary.
3 Often randomized algorithms are (much) simpler and/or more
efficient.
4 Several deep connections to mathematics, physics etc.
5 ...
6 Lots of fun!

Chandra (UIUC) CS498ABD 9 Fall 2020 9 / 33

Average case analysis vs Randomized
algorithms
Average case analysis:
1 Fix a deterministic algorithm.
2 Assume inputs comes from a probability distribution.
3 Analyze the algorithm’s average performance over the
distribution over inputs.

Randomized algorithms:
1 Algorithm uses random bits in addition to input.
2 Analyze algorithms average performance over the given input
where the average is over the random bits that the algorithm
uses.
3 On each input behaviour of algorithm is random. Analyze
worst-case over all inputs of the (average) performance.
Chandra (UIUC) CS498ABD 10 Fall 2020 10 / 33
Types of Randomized Algorithms
Typically one encounters the following types:
1 Las Vegas randomized algorithms: for a given input x
output of algorithm is always correct but the running time is a
random variable. In this case we are interested in analyzing the
expected running time.

Chandra (UIUC) CS498ABD 11 Fall 2020 11 / 33

Types of Randomized Algorithms
Typically one encounters the following types:
1 Las Vegas randomized algorithms: for a given input x
output of algorithm is always correct but the running time is a
random variable. In this case we are interested in analyzing the
expected running time.
2 Monte Carlo randomized algorithms: for a given input x the
running time is deterministic but the output is random; correct
with some probability. In this case we are interested in analyzing
the probability of the correct output (and also the running time).
3 Algorithms whose running time and output may both be random.

Chandra (UIUC) CS498ABD 11 Fall 2020 11 / 33

Analyzing Las Vegas Algorithms
Deterministic algorithm Q for a problem ⇧:
1 Let Q(x) be the time for Q to run on input x of length |x|.
2 Worst-case analysis: run time on worst input for a given size n.
Twc (n) = max Q(x).
x:|x|=n
=
it
Olntosn )

Chandra (UIUC) CS498ABD 12 Fall 2020 12 / 33

Analyzing Las Vegas Algorithms
Deterministic algorithm Q for a problem ⇧:
1 Let Q(x) be the time for Q to run on input x of length |x|.
2 Worst-case analysis: run time on worst input for a given size n.
Twc (n) = max Q(x).
x:|x|=n

Randomized algorithm R for a problem ⇧:

1 Let R(x) be the time for Q to run on input x of length |x|.
=
2 R(x) is a random variable: depends on random bits used by R.
3
E[R(x)] is the expected running time for R on x
4 Worst-case analysis: expected time on worst input of size n
Trand wc (n) = max E[R(x)] .
x:|x|=n
= I

Chandra (UIUC) CS498ABD 12 Fall 2020 12 / 33

Analyzing Monte Carlo Algorithms
Randomized algorithm M for a problem ⇧:
1 Let M(x) be the time for M to run on input x of length |x|.
For Monte Carlo, assumption is that run time is deterministic.
2 Let Pr[x] be the probability that M is correct on x.
3 Pr[x] is a random variable: depends on random bits used by M.
4 Worst-case analysis: success probability on worst input

Prand wc (n) = min Pr[x] .

x:|x|=n

Chandra (UIUC) CS498ABD 13 Fall 2020 13 / 33

 Part II

Randomized Quick Sort

Chandra (UIUC) CS498ABD 14 Fall 2020 14 / 33

Randomized QuickSort

Randomized QuickSort
1 Pick a pivot element uniformly at random from the array.
2 Split array into 3 subarrays: those smaller than pivot, those
larger than pivot, and the pivot itself.
3 Recursively sort the subarrays, and concatenate them.

1 array: 16, 12, 14, 20, 5, 3, 18, 19, 1 I 1,3, 5, 12,14, 16, 18,19
, 20

rank of an element as its

position in

the sorted aerag .

Chandra (UIUC) CS498ABD 15 Fall 2020 15 / 33

Analysis

What events to count?

Number of Comparisions.

Chandra (UIUC) CS498ABD 16 Fall 2020 16 / 33

Analysis

What events to count?

Number of Comparisions.

What is the probability space?

All the coin tosses at all levels and parts of recursion.

Chandra (UIUC) CS498ABD 16 Fall 2020 16 / 33

Analysis

What events to count?

Number of Comparisions.

What is the probability space?

All the coin tosses at all levels and parts of recursion.

Too Big!!

Chandra (UIUC) CS498ABD 16 Fall 2020 16 / 33

Analysis

What events to count?

Number of Comparisions.

What is the probability space?

All the coin tosses at all levels and parts of recursion.

Too Big!!

What random variables to define?

What are the events of the algorithm?

Chandra (UIUC) CS498ABD 16 Fall 2020 16 / 33

Analysis via Recurrence
1 Given array A of size n, let Q(A) be number of comparisons of
randomized QuickSort on A.
2 Note that Q(A) is a random variable.
3 Let Aileft and Airight be the left and right arrays obtained if rank i
element chosen as pivot.

Let Xi be indicator random variable, which is set to 1 if pivot is

t.iq#I
of rank i in A, else zero.
n
X ⇣ ⌘
Q(A) = n + Xi · Q(Aileft ) + Q(Airight ) .
i =1

T
E
Chandra (UIUC) CS498ABD
random variables
17 Fall 2020 17 / 33
Analysis via Recurrence
1 Given array A of size n, let Q(A) be number of comparisons of
randomized QuickSort on A.
2 Note that Q(A) is a random variable.
3 Let Aileft and Airight be the left and right arrays obtained if rank i
element chosen as pivot.

Let Xi be indicator random variable, which is set to 1 if pivot is

of rank i in A, else zero.
n
X ⇣ ⌘
Q(A) = n + Xi · Q(Aileft ) + Q(Airight ) .
i =1

Since each element of A has probability exactly of 1/n of being

chosen:
E[Xi ] = Pr[pivot has rank i ] = 1/n.
Chandra (UIUC) CS498ABD 17 Fall 2020 17 / 33
Independence of Random Variables

Lemma
Random variables Xi is independent of random variables Q(Aileft ) as
well as Q(Airight ), i.e.
⇥ i
⇤ ⇥ i
⇤
h Xi · Q(Aleft )i = E[Xi ] EhQ(Aleft ) i
E
E Xi · Q(Airight ) = E[Xi ] E Q(Airight )

Proof.
This is because the algorithm, while recursing on Q(Aileft ) and
Q(Airight ) uses new random coin tosses that are independent of the
coin tosses used to decide the first pivot. Only the latter decides
value of Xi .

Chandra (UIUC) CS498ABD 18 Fall 2020 18 / 33

Analysis via Recurrence
Let T (n) = maxA:|A|=n E[Q(A)] be the worst-case expected
running time of randomized QuickSort on arrays of size n.

Chandra (UIUC) CS498ABD 19 Fall 2020 19 / 33

Analysis via Recurrence
Let T (n) = maxA:|A|=n E[Q(A)] be the worst-case expected
running time of randomized QuickSort on arrays of size n.

We have, for any A:

n
X ⇣ ⌘
Q(A) = n + Xi Q(Aileft ) + Q(Airight )
i =1

Chandra (UIUC) CS498ABD 19 Fall 2020 19 / 33

Analysis via Recurrence
Let T (n) = maxA:|A|=n E[Q(A)] be the worst-case expected
running time of randomized QuickSort on arrays of size n.

We have, for any A:

n
X ⇣ ⌘
Q(A) = n + Xi Q(Aileft ) + Q(Airight )
i =1
By linearity of expectation, and independence random variables:
h i X n ⇣ h i h i⌘
i i
E Q(A) = n + E[Xi ] E Q(Aleft ) + E Q(Aright ) .
i =1 =
M
'
I

t
y
laid !
" n - i⇐

. ,

Chandra (UIUC) CS498ABD 19 Fall 2020 19 / 33

Analysis via Recurrence
Let T (n) = maxA:|A|=n E[Q(A)] be the worst-case expected
running time of randomized QuickSort on arrays of size n.

We have, for any A:

n
X ⇣ ⌘
Q(A) = n + Xi Q(Aileft ) + Q(Airight )
i =1
By linearity of expectation, and independence random variables:
h i X n ⇣ h i h i⌘
E Q(A) O
i i
=n+ E[Xi ] E Q(Aleft ) + E Q(Aright ) .
i =1

h
vi n
X 1
i t
)
¥0
E Q(A)  n +
i =1
n
(T (i
=
1) + T (n i )) .

Chandra (UIUC) CS498ABD 19 Fall 2020 19 / 33

Analysis via Recurrence
0
Let T (n) = maxA:|A|=n E[Q(A)] be the worst-case expected
running time of randomized QuickSort on arrays of size n.

Chandra (UIUC) CS498ABD 20 Fall 2020 20 / 33

Analysis via Recurrence
Let T (n) = maxA:|A|=n E[Q(A)] be the worst-case expected
running time of randomized QuickSort on arrays of size n.
We derived:
h i n
X 1
E Q(A)  n + (T (i 1) + T (n i )) .
i =1
n

Note that above holds for any A of size n. Therefore

n
X 1
max E[Q(A)] = T (n)  n + (T (i 1) + T (n i )) .
A:|A|=n
= i =1
n

Chandra (UIUC) CS498ABD 20 Fall 2020 20 / 33

Solving the Recurrence

EE
n
X 1
T (n)  n + (T (i 1) + T (n i ))
i =1
n
with base case T (1) = 0.
id
IE ht (Th Dt Tcn
'

TIME
- -

n t
,

Chandra (UIUC) CS498ABD 21 Fall 2020 21 / 33

Solving the Recurrence

n
X 1
T (n)  n + (T (i 1) + T (n i ))
i =1
n = =
=

with base case T (1) = 0.

Lemma
T (n) = O(n log n).

Chandra (UIUC) CS498ABD 21 Fall 2020 21 / 33

Solving the Recurrence

n
X 1
T (n)  n + (T (i 1) + T (n i ))
i =1
n
with base case T (1) = 0.
Lemma
T (n) = O(n log n).

Proof.
(Guess and) Verify by induction.

Chandra (UIUC) CS498ABD 21 Fall 2020 21 / 33

 Part III

Slick analysis of QuickSort

Chandra (UIUC) CS498ABD 22 Fall 2020 22 / 33

A Slick Analysis of QuickSort
Let Q(A) be number of comparisons done on input array A:
1 For 1  i < j \ < n let Rij be the event that rank i element is
=

compared with rank j element.

2 Xij is the indicator random variable for Rij . That is, Xij = 1 if
=
rank i is compared with rank j element, otherwise 0.

Chandra (UIUC) CS498ABD 23 Fall 2020 23 / 33

A Slick Analysis of QuickSort
Let Q(A) be number of comparisons done on input array A:
1 For 1  i < j < n let Rij be the event that rank i element is
compared with rank j element.
2 Xij is the indicator random variable for Rij . That is, Xij = 1 if
rank i is compared with rank j element, otherwise 0.
X
Q(A) = Xij
1i <j n

and hence by linearity of expectation,

h i X h i X h i
E Q(A) = E Xij = Pr Rij .
1i <j n 1i <j n

Chandra (UIUC) CS498ABD 23 Fall 2020 23 / 33

A Slick Analysis of QuickSort
Rij = rank i element is compared with rank j element.
Question: What is Pr[Rij ]?

Chandra (UIUC) CS498ABD 24 Fall 2020 24 / 33

A Slick Analysis of QuickSort
Rij = rank i element is compared with rank j element.
Question: What is Pr[Rij ]?
7 5 9 1 3 4 8 6
With ranks: 6 4 8 1 2 3 7 5
x =

Chandra (UIUC) CS498ABD 24 Fall 2020 24 / 33

A Slick Analysis of QuickSort
Rij = rank i element is compared with rank j element.
Question: What is Pr[Rij ]?
7 5 9 1 3 4 8 6
With ranks: 6 4 8 1 2 3 7 5
-
-

As such, probability of comparing 5 to 8 is Pr[R4,7 ].

=

Chandra (UIUC) CS498ABD 24 Fall 2020 24 / 33

A Slick Analysis of QuickSort
Rij = rank i element is compared with rank j element.
Question: What is Pr[Rij ]?
7 5 9 1 3 4 8 6
With ranks: 6 4 8 1 2 3 7 5
1 If pivot too small (say 3 [rank 2]). Partition and call recursively:

7 5 9 1 3 4 8 6
=) 1 3 7 5 9 4 8 6
Decision if to compare 5 to 8 is moved to subproblem.

Chandra (UIUC) CS498ABD 24 Fall 2020 24 / 33

A Slick Analysis of QuickSort
Rij = rank i element is compared with rank j element.
Question: What is Pr[Rij ]?
7 5 9 1 3 4 8 6
With ranks: 6 4 8 1 2 3 7 5
1 If pivot too small (say 3 [rank 2]). Partition and call recursively:

7 5 9 1 3 4 8 6
=) 1 3 7 5 9 4 8 6
Decision if to compare 5 to 8 is moved to subproblem.
2 If pivot too large (say 9 [rank 8]):

7 5 9 1 3 4 8 6 7 5 1 3 4 8 6 9
=)
Decision if to compare 5 to 8 moved to subproblem.

Chandra (UIUC) CS498ABD 24 Fall 2020 24 / 33

A Slick Analysis of QuickSort
Question: What is Pr[Ri,j ]?

7 5 9 1 3 4 8 6 As such, probability of com-

6 4 8 1 2 3 7 5 paring 5 to 8 is Pr[R4,7 ].
1 If pivot is 5 (rank 4). Bingo!

7 5 9 1 3 4 8 6 1 3 4 5 7 9 8 6
=)

Chandra (UIUC) CS498ABD 25 Fall 2020 25 / 33

A Slick Analysis of QuickSort
Question: What is Pr[Ri,j ]?

7 5 9 1 3 4 8 6 As such, probability of com-

6 4 8 1 2 3 7 5 paring 5 to 8 is Pr[R4,7 ].
1 If pivot is 5 (rank 4). Bingo!

7 5 9 1 3 4 8 6 1 3 4 5 7 9 8 6
=)
2 If pivot is 8 (rank 7). Bingo!

7 5 9 1 3 4 8 6 7 5 1 3 4 6 8 9
=)

Chandra (UIUC) CS498ABD 25 Fall 2020 25 / 33

A Slick Analysis of QuickSort
Question: What is Pr[Ri,j ]?

7 5 9 1 3 4 8 6 As such, probability of com-

6 4 8 1 2 3 7 5 paring 5 to 8 is Pr[R4,7 ].
1 If pivot is 5 (rank 4). Bingo!

7 5 9 1 3 4 8 6 1 3 4 5 7 9 8 6
=)
2 If pivot is 8 (rank 7). Bingo!

7 5 9 1 3 4 8 6
=) 7 5 1 3 4 6 8 9
3 If pivot in between the two numbers (say 6 [rank 5]):

7 5 9 1 3 4 8 6
=) 5 1 3 4 6 7 8 9
5 and 8 will never be compared to each other.

Chandra (UIUC) CS498ABD 25 Fall 2020 25 / 33

A Slick Analysis of QuickSort
Question: What is Pr[Ri,j ]?

Conclusion:
Ri ,j happens if and only if:
i th or j th ranked element is the first pivot out of
i th to j th ranked elements.

Chandra (UIUC) CS498ABD 26 Fall 2020 26 / 33

Digression
Consider the following experiment:
Every day John decides whether to wear a tie by tossing a biased
coin that comes up heads with probability p > 0 (and tails
otherwise). He wears a tie if it comes up heads.
If the coin is heads he tosses an unbiased coin to decide whether
to wear a red tie or a blue tie.

Chandra (UIUC) CS498ABD 27 Fall 2020 27 / 33

Digression
Consider the following experiment:
Every day John decides whether to wear a tie by tossing a biased
coin that comes up heads with probability p > 0 (and tails
otherwise). He wears a tie if it comes up heads.
If the coin is heads he tosses an unbiased coin to decide whether
to wear a red tie or a blue tie.

Question: What is the probability that John wore a red tie on the
first day he wore a tie?

Chandra (UIUC) CS498ABD 27 Fall 2020 27 / 33

A Slick Analysis of QuickSort
Question: What is Pr[Rij ]?

Chandra (UIUC) CS498ABD 28 Fall 2020 28 / 33

A Slick Analysis of QuickSort
Question: What is Pr[Rij ]?
Lemma
h i
Pr Rij O
= j 2
i +1
.
→
1,2 , 3, 4,5,
by ,
- - r
,
100
⑤
j - i -
- I
E Ps[Rig]
znltnicicjen
ulan
=

Chandra (UIUC) CS498ABD 28 Fall 2020 28 / 33

A Slick Analysis of QuickSort
Question: What is Pr[Rij ]?
Lemma
h i
2
Pr Rij = j i +1
.

Proof.
Let a1 , . . . , ai , . . . , a←
j , . . . , an be elements of A in sorted order.
Let S = {ai , ai +1 , . . . , aj }
Observation: If pivot is chosen outside S then all of S either in left
array or right array.
Observation: ai and aj separated when a pivot is chosen from S for
the first time. Once separated no comparison.
Observation: ai is compared with aj if and only if either ai or aj is
chosen as a pivot from S at separation...
Chandra (UIUC) CS498ABD 28 Fall 2020 28 / 33
A Slick Analysis of QuickSort
Continued...

Lemma
h i

0
2
Pr Rij = j i +1
.

Proof.
Let a1 , . . . , ai , . . . , aj , . . . , an be sort of A. Let
S = {ai , ai +1 , . . . , aj }
Observation: ai is compared with aj if and only if either ai or aj is
chosen as a pivot from S at separation.
Observation: Given that pivot is chosen from S the probability that
it is ai or aj is exactly 2/|S| = 2/(j i + 1) since the pivot is
chosen uniformly at random from the array.

Chandra (UIUC) CS498ABD 29 Fall 2020 29 / 33

How much is this?
P
Hn = ni=1 1i is the n’th harmonic number
(A) Hn = ⇥(1). + →
An

I
(B) Hn = ⇥(log log n). Aqa , Ae ,
,
- - r

p N l
(C) Hn = ⇥( log n).
-

E Rhi
(D) Hn = ⇥(log n).
(E) Hn = ⇥(log2 n).
:÷
in
n
I
E ZE .

e-= ' e

2 An =L
2h Hn E 2n inn
=

Chandra (UIUC) CS498ABD 30 Fall 2020 30 / 33

And how much is this?
n 1X
X n i
1
Tn =
i =1 j =1
j
is equal to
(A) Tn = ⇥(n).
(B) Tn = ⇥(n log n).
(C) Tn = ⇥(n log2 n).
(D) Tn = ⇥(n2 ).
(E) Tn = ⇥(n3 ).

Chandra (UIUC) CS498ABD 31 Fall 2020 31 / 33

A Slick Analysis of QuickSort
Continued...

h i X X
E Q(A) = E[Xij ] = Pr[Rij ] .
1i <j n 1i <j n

Lemma
2
Pr[Rij ] = j i +1
.

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i X h i X 2
E Q(A) = Pr Rij =
1i <j n 1i <j n
j i +1

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i X 2
E Q(A) =
1i <j n
j i +1

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i X 2
E Q(A) =
1i <j n
j i +1
n
X1 n
X 2
=
i =1 j =i +1
j i +1

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i n n
X1 X 2
E Q(A) =
i =1 j =i +1
j i +1

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i n n
X1 X 1
E Q(A) = 2
i =1 i <j
j i +1

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i n n
X1 X 1
E Q(A) = 2
i =1 i <j
j i +1

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i n n
X1 X n
X1 nX
i +1
1 1
E Q(A) = 2 2
i =1 i <j
j i +1 i =1 =2

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i n n
X1 X n
X1 nX
i +1
1 1
E Q(A) = 2 2
i =1 i <j
j i +1 i =1 =2
n
X1 X
2 (Hn i +1 1)  2 Hn
i =1 1i <n

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

A Slick Analysis of QuickSort
Continued...

Lemma
2
Pr[Rij ] = j i +1
.

h i n n
X1 X n
X1 nX
i +1
1 1
E Q(A) = 2 2
i =1 i <j
j i +1 i =1 =2
n
X1 X
2 (Hn i +1 1)  2 Hn
i =1 1i <n

 2nHn = O(n log n)

Chandra (UIUC) CS498ABD 32 Fall 2020 32 / 33

Where do I get random bits?
Question: Are true random bits available in practice?
1 Buy them!
2 CPUs use physical phenomena to generate random bits.
3 Can use pseudo-random bits or semi-random bits from nature.
Several fundamental unresolved questions in complexity theory
on this topic. Beyond the scope of this course.
4 In practice pseudo-random generators work quite well in many
applications.
5 The model is interesting to think in the abstract and is very
useful even as a theoretical construct. One can derandomize
randomized algorithms to obtain deterministic algorithms.

Chandra (UIUC) CS498ABD 33 Fall 2020 33 / 33