COMPUTER SCIENCE COMPUTER SCIENCE

S E D G E W I C K / W A Y N E S E D G E W I C K / W A Y N E
PA R T I : P R O G R A M M I N G I N J AVA PA R T I : P R O G R A M M I N G I N J AVA

3. Arrays
Computer Science

• Basic concepts

Computer 3. Arrays • Typical array-processing code

Science
An Interdisciplinary Approach
• Two-dimensional arrays

ROBERT SEDGEWICK
1.4 K E V I N WAY N E

http://introcs.cs.princeton.edu

CS.3.A.Arrays.Basics

Basic building blocks for programming Your first data structure

A data structure is an arrangement of data that enables efficient processing by a program.

any program you might want to write An array is an indexed sequence of values of the same type.

objects
Examples. index value

functions and modules • 52 playing cards in a deck. 0 2

• 100 thousand students in an online class. 1 6

graphics, sound, and image I/O
• 1 billion pixels in a digital image. 2 A
arrays • 4 billion nucleotides in a DNA strand. 3 A
Ability to store and process
conditionals and loops huge amounts of data • 73 billion Google queries per year. ...
• 86 billion neurons in the brain. 49 3
Math text I/O
• 50 trillion cells in the human body. 50 K
primitive data types assignment statements • 6.02 ⇥ 1023 particles in a mole.
51 4

Main purpose. Facilitate storage and manipulation of data.

3 4
Processing many values of the same type Memory representation of an array

10 values, without arrays An array is an indexed sequence of values of the same type.

double a0 = 0.0;
double a1 = 0.0;
A computer's memory is also an indexed sequence of memory locations. stay tuned for many details

double a2 = 0.0; 10 values, with an array 1 million values, with an array • Each primitive type value occupies a fixed number of locations.
double a3 = 0.0;
double a4 = 0.0; • Array values are stored in contiguous locations.
double[] a; double[] a;
double a5 = 0.0; a = new double[10]; a = new double[1000000];
double a6 = 0.0; ... ... a for simplicity in this lecture, think of a as the memory address of the first location
double a7 = 0.0; a[4] = 3.0; a[234567] = 3.0; the actual implementation in Java is just slightly more complicated.
double a8 = 0.0; ... ...
double a9 = 0.0; a[0] a[1] a[2] a[3] a[4] a[5] a[6] a[7] a[8] a[9]
a[8] = 8.0; a[876543] = 8.0;
... ... ...
a4 = 3.0; double x = a[4] + a[8]; double x = a[234567] + a[876543];
...
a8 = 8.0; Critical concepts
... • Indices start at 0.
double x = a4 + a8; scales to handle huge amounts of data
an easy alternative
• Given i, the operation of accessing the value a[i] is extremely efficient.
• The assignment b = a makes the names b and a refer to the same array.
it does not copy the array,
as with primitive types
tedious and error-prone code
(stay tuned for details)

5 6

Java language support for arrays Copying an array

Basic support operation typical code To copy an array, create a new array , then copy all the values.

Declare an array double[] a;

double[] b = new double[a.length];
Create an array of a given length a = new double[1000]; for (int i = 0; i < a.length; i++)
b[i] = a[i];
Refer to an array entry by index a[i] = b[j] + c[k];

a i b i
Refer to the length of an array a.length;
0.3 0.6 0.99 0.01 0.5 0.3 0.6 0.99 0.01 0.5
Initialization options
no need to use a loop like
for (int i = 0; i < 1000; i++) Important note: The code b = a does not copy an array (it makes b and a refer to the same array).
a[i] = 0.0;
operation typical code
double[] b = new double[a.length];
Default initialization to 0 for numeric types a = new double[1000];
b = a;
BUT cost of creating an
array is proportional to
its length.
Declare, create and initialize in one statement double[] a = new double[1000]; a b

0.3 0.6 0.99 0.01 0.5

Initialize to literal values double[] x = { 0.3, 0.6, 0.1 };
7 8
Programming with arrays: typical examples Pop quiz 1 on arrays

Access command-line args in system array For brevity, N is a.length and b.length in all this code.
Q. What does the following code print?
int stake = Integer.parseInt(args[0]);
int goal = Integer.parseInt(args[1]); Copy to another array
int trials = Integer.parseInt(args[2]); double[] b = new double[N]; public class PQarray1
for (int i = 0; i < N; i++) {
b[i] = a[i]; public static void main(String[] args)
{
int[] a = new int[6];
Create an array with N random values
int[] b = new int[a.length];
double[] a = new double[N]; Print array values, one per line
for (int i = 0; i < N; i++) for (int i = 0; i < N; i++) b = a;
a[i] = Math.random(); System.out.println(a[i]); for (int i = 1; i < b.length; i++)
b[i] = i;

for (int i = 0; i < a.length; i++)

System.out.print(a[i] + " ");
Compute the average of array values Find the maximum of array values System.out.println();
double sum = 0.0; double max = a[0];
for (int i = 0; i < N; i++) for (int i = 0; i < b.length; i++)
for (int i = 1; i < N; i++) System.out.print(b[i] + " ");
sum += a[i]; if (a[i] > max) max = a[i]; System.out.println();
double average = sum / N; }
}

9 10

Pop quiz 1 on arrays Programming with arrays: typical bugs

Q. What does the following code print?

Array index out of bounds double[] a = new double[10];
for (int i = 1; i <= 10; i++)
a[i] = Math.random();
public class PQarray1
{ No a[10] (and a[0] unused)
public static void main(String[] args)
{
int[] a = new int[6];
int[] b = new int[a.length]; Uninitialized array double[] a;
for (int i = 0; i < 9; i++)
b = a; After this, b and a refer to the same array
for (int i = 1; i < b.length; i++)
a[i] = Math.random();
b[i] = i;
Never created the array
for (int i = 0; i < a.length; i++)
System.out.print(a[i] + " ");
System.out.println();

for (int i = 0; i < b.length; i++) Undeclared variable a = new double[10];

System.out.print(b[i] + " "); A. for (int i = 0; i < 10; i++)
% java PQ4_1
System.out.println(); a[i] = Math.random();
0 1 2 3 4 5
}
} 0 1 2 3 4 5 What type of data does a refer to?

11 12
 COMPUTER SCIENCE COMPUTER SCIENCE
S E D G E W I C K / W A Y N E S E D G E W I C K / W A Y N E
PA R T I : P R O G R A M M I N G I N J AVA PA R T I : P R O G R A M M I N G I N J AVA

Image sources

http://commons.wikimedia.org/wiki/File:CERN_Server_03.jpg

3. Arrays
• Basic concepts
• Examples of array-processing code
• Two-dimensional arrays

CS.3.A.Arrays.Basics CS.3.B.Arrays.Examples

Example of array use: create a deck of cards Example of array use: create a deck of cards

Define three arrays String[] rank = {"2", "3", "4", "5", "6", "7", "8", "9", "10", "J", "Q", "K", "A" };
public class Deck
• Ranks. String[] suit = { " ", " ", " ", " " }; {
public static void main(String[] args)
• Suits. {
String[] deck = new String[52];
• Full deck. j String[] rank = {"2", "3", "4", "5", "6", "7", "8", "9", "10", "J", "Q", "K", "A" };
String[] suit = { " ", " ", " ", " " };

String[] deck = new String[52]; no color in Unicode;

for (int j = 0; j < 4; j++) artistic license for lecture
Use nested for loops to put all the cards in the deck.
for (int i = 0; i < 13; i++)
deck[i + 13*j] = rank[i] + suit[j];
better style to use rank.length and suit.length j
for (int j = 0; j < 4; j++) clearer in lecture to use 4 and 13 for (int i = 0; i < 52; i++)
0 1 2 3
for (int i = 0; i < 13; i++) System.out.print(deck[i] + " ");
suit
deck[i + 13*j] = rank[i] + suit[j]; System.out.println();
i } % java Deck
0 1 2 3 4 5 6 7 8 9 10 11 12 }
2 3 4 5 6 7 8 9 10 J Q K A
rank 2 3 4 5 6 7 8 9 10 J Q K A 2 3 4 5 6 7 8 9 10 J Q K A
2 3 4 5 6 7 8 9 10 J Q K A
2 3 4 5 6 7 8 9 10 J Q K A
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ...
%
deck 2 3 4 5 6 7 8 9 10 J Q K A 2 3 4 5 6 7 8 9 ...

15 16
Pop quiz 2 on arrays Pop quiz 2 on arrays

Q. What happens if the order of the for loops in Deck is switched? Q. What happens if the order of the for loops in Deck is switched?

for (int j = 0; j < 4; j++) for (int i = 0; i < 13; i++) for (int j = 0; j < 4; j++) for (int i = 0; i < 13; i++)
for (int i = 0; i < 13; i++) for (int j = 0; j < 4; j++) for (int i = 0; i < 13; i++) for (int j = 0; j < 4; j++)
deck[i + 13*j] = rank[i] + suit[j]; deck[i + 13*j] = rank[i] + suit[j]; deck[i + 13*j] = rank[i] + suit[j]; deck[i + 13*j] = rank[i] + suit[j];

A. The array is filled in a different order, but the output is the same. j
0 1 2 3
suit

i
0 1 2 3 4 5 6 7 8 9 10 11 12
rank 2 3 4 5 6 7 8 9 10 J Q K A

0 1 2 ... 12 13 14 15 ... 25 26 27 28 ... 38 39 40 41 ... 51

deck 2 3 4 ... A 2 3 4 ... A 2 3 4 ... A 2 3 4 ... A

NOTE: Error on page 92 in 3rd printing of text (see errata list on booksite).

17 18

Pop quiz 3 on arrays Pop quiz 3 on arrays

Q. Change Deck to put the cards in rank order in the array. Q. Change Deck to put the cards in rank order in the array.

% java Deck % java Deck

2 2 2 2 3 3 3 3 4 4 4 4 5 5 5 5 6 6 6 6 7 7 7 7 8 8 8 8 9 9 9 9 2 2 2 2 3 3 3 3 4 4 4 4 5 5 5 5 6 6 6 6 7 7 7 7 8 8 8 8 9 9 9 9
10 10 10 10 J J J J Q Q Q Q K K K K A A A A 10 10 10 10 J J J J Q Q Q Q K K K K A A A A
% %

A. for (int i = 0; i < 13; i++)

for (int j = 0; j < 4; j++)
deck[4*i + j] = rank[i] + suit[j]; j
0 1 2 3
suit

i
0 1 2 3 4 5 6 7 8 9 10 11 12
rank 2 3 4 5 6 7 8 9 10 J Q K A

0 1 2 3 4 5 6 7 8 9 10 11 12 ...
deck 2 2 2 2 3 3 3 3 4 4 4 4 5 ...

19 20
 Array application: take a card, any card

Problem: Print a random sequence of N cards.

Algorithm
Take N from the command line and do the following N times
• Calculate a random index r between 0 and 51.
• Print deck[r].

Implementation: Add this code instead of printing deck in Deck.

for (int i = 0; i < N; i++) each value between 0 and 51 equally likely
{
int r = (int) (Math.random() * 52);
System.out.println(deck[r]);
}

Note: Same method is effective for printing a random sequence from any data collection.
21 22

Array application: random sequence of cards Array application: shuffle and deal from a deck of cards

public class DrawCards Problem: Print N random cards from a deck.

{
public static void main(String[] args)
{
int N = Integer.parseInt(args[0]);
Algorithm: Shuffle the deck, then deal.

String[] rank = {"2", "3", "4", "5", "6", "7", "8", "9","10", "J", "Q", "K", "A" };
• Consider each card index i from 0 to 51.
String[] suit = { " ", " ", " ", " " }; • Calculate a random index r between i and 51.
String[] deck = new String[52]; • Exchange deck[i] with deck[r]
for (int i = 0; i < 13; i++) % java DrawCards 10
for (int j = 0; j < 4; j++) • Print the first N cards in the deck.
6 K 10 8 9 9 6 10 3 5
deck[i + 13*j] = rank[i] + suit[j];

for (int i = 0; i < N; i++) % java DrawCards 10

{ 2 A 5 A 10 Q K K A A Implementation for (int i = 0; i < 52; i++)
int r = (int) (Math.random() * 52); { each value
System.out.print(deck[r] + " "); int r = i + (int) (Math.random() * (52-i)); between i and 51
} % java DrawCards 10 equally likely
String t = deck[r];
System.out.println(); 6 10 4 A K Q K 7 5 Q deck[r] = deck[i];
} deck[i] = t;
}
% java DrawCards 10 }
A J 5 K Q 5 9 9 6 K for (int i = 0; i < N; i++)
System.out.print(deck[i]);
System.out.println();
Note: Sample is with replacement (same card can appear multiple times). appears twice
23 24
Array application: shuffle a deck of 10 cards (trace) Array application: shuffle and deal from a deck of cards
deck
i r
for (int i = 0; i < 10; i++) 0 1 2 3 4 5 6 7 8 9 public class DealCards
{ {
int r = i + (int) (Math.random() * (10-i)); 2 3 4 5 6 7 8 9 10 J public static void main(String[] args)
{
String t = deck[r]; int N = Integer.parseInt(args[0]);
0 7 9 3 4 5 6 7 8 2 10 J
deck[r] = deck[i];
deck[i] = t; 1 3 String[] rank = {"2", "3", "4", "5", "6", "7", "8", "9", "10", "J", "Q", "K", "A" };
9 5 4 3 6 7 8 2 10 J
} String[] suit = { " ", " ", " ", " " };
2 9 9 5 J 3 6 7 8 2 10 4 String[] deck = new String[52];
for (int i = 0; i < 13; i++)
Q. Why does this method work? 3 9 9 5 J 4 6 7 8 2 10 3 for (int j = 0; j < 4; j++)
deck[i + 13*j] = rank[i] + suit[j];
• Uses only exchanges, so the deck after 4 6 9 5 J 4 8 7 6 2 10 3
for (int i = 0; i < 52; i++)
the shuffle has the same cards as before. {
5 9 9 5 J 4 8 3 6 2 10 7
int r = i + (int) (Math.random() * (52-i));
• N i equally likely values for deck[i]. String t = deck[r];
6 8 9 5 J 4 8 3 10 2 6 7 deck[r] = deck[i];
• Therefore N ⇥(N 1)⇥(N 1)... ⇥2⇥1 = N ! deck[i] = t; random poker hand
equally likely values for deck[]. 7 9 9 5 J 4 8 3 10 7 6 2 }

8 8 9 5 J 4 8 3 10 7 6 2 for (int i = 0; i < N; i++) % java DealCards 5

Initial order is immaterial.
System.out.print(deck[i]); 9 Q 6 4 2 random bridge hand
9 9 9 5 J 4 8 3 10 7 6 2 System.out.println();
}
} % java DealCards 13
Note: Same method is effective for randomly rearranging any type of data. 3 4 10 6 6 2 9 8 A 3 9 5 Q
25 26

Coupon collector Array application: coupon collector

Coupon collector problem

• M different types of coupons. Coupon collector simulation public class Coupon
{
• Collector acquires random coupons, one at a time, each type equally likely. • Generate random int values public static void main(String[] args)
{
Q. What is the expected number of coupons needed to acquire a full collection? between 0 and M 1. int M = Integer.parseInt(args[0]);
int cards = 0; // number of cards collected
• Count number used to generate int distinct = 0; // number of distinct cards
each value at least once.
Example: Collect all ranks in a random sequence of cards (M = boolean[] found = new boolean[M];
while (distinct < M)
{
Sequence Key to the implementation int r = (int) (Math.random() * M);
cards++;
9 5 8 10 2 A 10 Q 3 9 5 9 7 2 8 6 Q K 10 A 4 J
• Create a boolean array of length M. if (!found[r])
{
(Initially all false by default.) distinct++; % java Coupon 13
found[r] = true; 46
Collection • When r generated, check the r th } % java Coupon 13
2 3 4 5 6 7 8 9 10 J Q K A
value in the array. } 22

2 3 4 5 6 7 8 9 10 J Q K A 22 cards needed • If true, ignore it (not new). System.out.println(cards); % java Coupon 13

to complete } 54
2 5 8 9 10 Q A collection • If false, count it as new distinct } % java Coupon 13
value (and set r th entry to true) 27
9 10
27 28
Array application: coupon collector (trace for M = 6) Simulation, randomness, and analysis (revisited)

found
Coupon collector problem
r distinct cards
0 1 2 3 4 5 • M different types of coupons.
F F F F F F 0 0 • Collector acquires random coupons, one at a time, each type equally likely.
boolean[] found = new boolean[M]; 2 F F T F F F 1 1 Q. What is the expected number of coupons needed to acquire a full collection?
Pierre-Simon Laplace
while (distinct < M)
1749-1827
{ 0 T F T F F F 2 2
int r = (int) (Math.random() * M);
A. (known via mathematical analysis for centuries) About M ln M + .57721M .
cards++; 4 T F T F T F 3 3
if (!found[r]) type M expected wait
0 T F T F T F 3 4 % java Coupon 4
{ 11
distinct++; playing card suits 4 8
1 T T T F T F 4 5 % java Coupon 13
found[r] = true; playing card ranks 13 41 38
} 2 T T T F T F 4 6
} baseball cards 1200 9201 % java Coupon 1200
5 T T T F T T 5 7 8789
Magic™ cards 12534 125508
% java Coupon 12534
0 T T T F T T 5 8 125671
Remarks
1 T T T F T T 5 9
• Computer simulation can help validate mathematical analysis.
3 T T T T T T 6 10 Example: Is Math.random()
• Computer simulation can also validate software behavior. simulating randomness?
29 30

Simulation, randomness, and analysis (revisited) Simulation, randomness, and analysis (revisited)

Coupon collector problem

Once simulation is debugged, experimental evidence is easy to obtain.
• M different types of coupons.
Analogous code for coupon collector, this lecture
Gambler's ruin simulation, previous lecture • Collector acquires random coupons, one at a time, each type equally likely.
public class CouponCollector
{ Q. What is the expected number of coupons needed to acquire a full collection?
public class Gambler
{ public static void main(String[] args)
public static void main(String[] args) {
{ int M = Integer.parseInt(args[0]);
int stake = Integer.parseInt(args[0]); int trials = Integer.parseInt(args[1]); Predicted by mathematical analysis Observed by computer simulation
int goal = Integer.parseInt(args[1]); int cards = 0;
int trials = Integer.parseInt(args[2]); boolean[] found;
type M M ln M + .57721M
int wins = 0; for (int i = 0; i < trials; i++) % java CouponCollector 4 1000000
for (int i = 0; i < trials; i++) { playing card suits 4 8 8
{ int distinct = 0;
int t = stake; found = new boolean[M]; % java CouponCollector 13 1000000
while (t > 0 && t < goal) while (distinct < M) playing card ranks 13 41 41
{ {
if (Math.random() < 0.5) t++; int r = (int) (Math.random() * M); % java CouponCollector 52 100000
cards++; playing cards 52 236
else t--; 236
} if (!found[r])
if (t == goal) wins++; { baseball cards 1200 9201 % java CouponCollector 1200 10000
} distinct++; 9176
System.out.println(wins + " wins of " + trials); found[r] = true;
} } magic cards 12534 125508 % java CouponCollector 12534 1000
} } 125920
}
System.out.println(cards/trials);
} Hypothesis. Centuries-old analysis is correct and Math.random() simulates randomness.
}
31 32
 COMPUTER SCIENCE COMPUTER SCIENCE
S E D G E W I C K / W A Y N E S E D G E W I C K / W A Y N E
PA R T I : P R O G R A M M I N G I N J AVA PA R T I : P R O G R A M M I N G I N J AVA

Image sources

http://www.vis.gr.jp/~nazoya/cgi-bin/catalog/img/CARDSBIC809_red.jpg
http://www.alegriphotos.com/Shuffling_cards_in_casino-photo-deae1081e5ebc6631d6871f8b320b808.html
http://iveypoker.com/wp-content/uploads/2013/09/Dealing.jpg
http://upload.wikimedia.org/wikipedia/commons/b/bf/Pierre-Simon,_marquis_de_Laplace_(1745-1827)_-_Guérin.jpg
3. Arrays
• Basic concepts
• Examples of array-processing code
• Two-dimensional arrays

CS.3.B.Arrays.Examples CS.3.C.Arrays.2D

Two-dimensional arrays Java language support for two-dimensional arrays (basic support)

A two-dimensional array is a doubly-indexed sequence of values of the same type. operation typical code

grade Declare a two-dimensional array double[][] a;

Examples 0 1 2 3 4 5 ... Create a two-dimensional array of a given length a = new double[1000][1000];

0 A A C B A C
• Matrices in math calculations. Refer to an array entry by index a[i][j] = b[i][j] * c[j][k];
1 B B B B A A
student ID

• Grades for students in an online class. 2 C D D B C A

Refer to the number of rows a.length;
• Outcomes of scientific experiments. 3 A A A A A A
can be different
4 C C B C B B Refer to the number of columns a[i].length;
• Transactions for bank customers. for each row
5 A A A B A A
• Pixels in a digital image. ... Refer to row i a[i] no way to refer
to column j
• Geographic data
a[][]
• ...
y-coordinate

a[0][0] a[0][1] a[0][2] a[0][3] a[0][4] a[0][5] a[0][6] a[0][7] a[0][8] a[0][9]

a[1] a[1][0] a[1][1] a[1][2] a[1][3] a[1][4] a[1][5] a[1][6] a[1][7] a[1][8] a[1][9]

a[2][0] a[2][1] a[2][2] a[2][3] a[2][4] a[2][5] a[2][6] a[2][7] a[2][8] a[2][9]

a 3-by-10 array
Main purpose. Facilitate storage and manipulation of data.
x-coordinate 35 36
Java language support for two-dimensional arrays (initialization) Application of arrays: vector and matrix calculations

Mathematical abstraction: vector Mathematical abstraction: matrix

no need to use nested loops like
Java implementation: 1D array Java implementation: 2D array
for (int i = 0; i < 1000; i++)
operation typical code for (int j = 0; j < 1000; j++)
a[i][j] = 0.0;
Default initialization to 0
a = new double[1000][1000]; Vector addition Matrix addition
for numeric types BUT cost of creating an
array is proportional to
its size. double[] c = new double[N]; double[][] c = new double[N][N];
Declare, create and initialize for (int i = 0; i < N; i++) for (int i = 0; i < N; i++)
double[][] a = new double[1000][1000]; c[i] = a[i] + b[i]; for (int j = 0; j < N; j++)
in a single statement
c[i][j] = a[i][j] + b[i][j];

double[][] p =
{
.70 .20 .10 .80 .30 .50 1.5 .50 .60
{ .92, .02, .02, .02, .02 },
{ .02, .92, .32, .32, .32 }, .30 .60 .10 + .50 .10 .40 = .80 .70 .50 .30 .60 .10 + .10 .40 .10 = .40 1.0 .20
Initialize to literal values
{ .02, .02, .02, .92, .02 }, .50 .10 .40 .10 .30 .40 .60 .40 .80
{ .92, .02, .02, .02, .02 },
{ .47, .02, .47, .02, .02 },
};

37 38

Application of arrays: vector and matrix calculations Pop quiz 4 on arrays

Mathematical abstraction: vector Mathematical abstraction: matrix Q. How many multiplications to multiply two N-by-N matrices?
Java implementation: 1D array Java implementation: 2D array

double[][] c = new double[N][N];

for (int i = 0; i < N; i++)
Vector dot product Matrix multiplication
for (int j = 0; j < N; j++)
double sum = 0.0; double[][] c = new double[N][N]; for (int k = 0; k < N; k++)
for (int i = 0; i < N; i++) for (int i = 0; i < N; i++)
c[i][j] += a[i][k] * b[k][j];
sum += a[i]*b[i]; for (int j = 0; j < N; j++)
for (int k = 0; k < N; k++)
c[i][j] += a[i][k] * b[k][j];
1. N
.30 .60 .10 · .50 .10 .40 = .25

2. N 2
i x[i] y[i] x[i]*y[i] sum

0 0.3 0.5 0.15 0.15 3. N 3

.70 .20 .10 .80 .30 .50 .59 .32 .41
1 0.6 0.1 0.06 0.21 4. N 4
.30 .60 .10 * .10 .40 .10 = .31 .36 .25
2 0.1 0.4 0.04 0.25 .50 .10 .40 .10 .30 .40 .45 .31 .42

end-of-loop trace
39 40
Pop quiz 4 on arrays Self-avoiding random walks

Q. How many multiplications to multiply two N-by-N matrices? A dog walks around at
random in a city, never
revisiting any intersection.
double[][] c = new double[N][N];
for (int i = 0; i < N; i++)
for (int j = 0; j < N; j++) Q. Does the dog escape?
for (int k = 0; k < N; k++)
escape
c[i][j] += a[i][k] * b[k][j]; Model: a random process in an N-by-N lattice
• Start in the middle.
1. N • Move to a random neighboring intersection
but do not revisit any intersection.
2. N 2 • Outcome 1 (escape): reach edge of lattice.
Nested for loops: N ⇥ N ⇥ N • Outcome 2 (dead end): no unvisited neighbors.
3. N 3

4. N 4 Q. What are the chances of reaching a dead end? dead end

Approach: Use Monte Carlo simulation, recording visited positions in an N-by-N array.
41 42

Self-avoiding random walks Application of 2D arrays: self-avoiding random walks

% java SelfAvoidingWalker 10 100000

public class SelfAvoidingWalker 5% dead ends
{
% java SelfAvoidingWalker 20 100000
public static void main(String[] args) 32% dead ends
{
% java SelfAvoidingWalker 30 100000
int N = Integer.parseInt(args[0]); 58% dead ends
int trials = Integer.parseInt(args[1]);
% java SelfAvoidingWalker 40 100000
int deadEnds = 0; 77% dead ends
for (int t = 0; t < trials; t++)
% java SelfAvoidingWalker 50 100000
{ 87% dead ends
boolean[][] a = new boolean[N][N];
int x = N/2, y = N/2; % java SelfAvoidingWalker 60 100000
93% dead ends

while (x > 0 && x < N-1 && y > 0 && y < N-1) % java SelfAvoidingWalker 70 100000
96% dead ends
{
if (a[x-1][y] && a[x+1][y] && a[x][y-1] && a[x][y+1]) % java SelfAvoidingWalker 80 100000
98% dead ends
{ deadEnds++; break; }
% java SelfAvoidingWalker 90 100000
99% dead ends
a[x][y] = true;
double r = Math.random(); % java SelfAvoidingWalker 100 100000
if (r < 0.25) { if (!a[x+1][y]) x++; } 99% dead ends
else if (r < 0.50) { if (!a[x-1][y]) x--; }
100%
else if (r < 0.75) { if (!a[x][y+1]) y++; }
else if (r < 1.00) { if (!a[x][y-1]) y--; }
75%
}
}
System.out.println(100*deadEnds/trials + "% dead ends"); 50%

}
} 25%

0%
10 20 30 40 50 60 70 80 90 100

43 44
 Simulation, randomness, and analysis (revisited again) Your first data structure

Self-avoiding walk in an N-by-N lattice Arrays: A basic building block in programming

• Start in the middle. • They enable storage of large amounts of data (values all of the same type).
• Move to a random neighboring intersection (do not revisit any intersection). • With an index, a program can instantly access a given value.
• Efficiency derives from low-level computer hardware organization (stay tuned).
Applications
• Model the behavior of solvents and polymers.
Some applications in this course:
• Model the physics of magnetic materials.
• (many other physical phenomena)
^ digital images N-body simulation
Paul Flory
1910-1985 LFSR
Q. What is the probability of reaching a dead end? Nobel Prize 1974

Mathematicians and
A. Nobody knows (despite decades of study). physics researchers
cannot solve the problem.
Computational models play
digital audio
A. 99+% for N >100 (clear from simulations). YOU can! an essential role in modern
scientific research.

Remark: Computer simulation is often the only effective way to study a scientific phenomenon.
45 46

COMPUTER SCIENCE COMPUTER SCIENCE

S E D G E W I C K / W A Y N E S E D G E W I C K / W A Y N E
PA R T I : P R O G R A M M I N G I N J AVA PA R T I : P R O G R A M M I N G I N J AVA

Image sources

http://en.wikipedia.org/wiki/Airedale_Terrier#mediaviewer/File:Airedale_Terrier.jpg
http://www.nobelprize.org/nobel_prizes/chemistry/laureates/1974/flory_postcard.jpg
Computer Science

Computer 3. Arrays
Science An Interdisciplinary Approach

ROBERT SEDGEWICK
1.4 K E V I N WAY N E

http://introcs.cs.princeton.edu

CS.3.C.Arrays.2D