COMPUTER SCIENCE

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
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
 COMPUTER SCIENCE
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

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

CS.3.A.Arrays.Basics
Basic building blocks for programming

any program you might want to write

objects

functions and modules

graphics, sound, and image I/O

arrays
Ability to store and process
conditionals and loops huge amounts of data

Math text I/O

primitive data types assignment statements

3
Your first data structure

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

An array is an indexed sequence of values of the same type.

Examples. index value

• 52 playing cards in a deck. 0 2♥

• 100 thousand students in an online class. 1 6♠

• 1 billion pixels in a digital image. 2 A♦

• 4 billion nucleotides in a DNA strand. 3 A♥

• 73 billion Google queries per year. ...
• 86 billion neurons in the brain. 49 3♣
• 50 trillion cells in the human body. 50 K♣
• 6.02 × 1023 particles in a mole.
51 4♠

Main purpose. Facilitate storage and manipulation of data.

4
Processing many values of the same type

10 values, without arrays

double a0 = 0.0;
double a1 = 0.0;
double a2 = 0.0; 10 values, with an array 1 million values, with an array
double a3 = 0.0;
double a4 = 0.0; double[] a; double[] a;
double a5 = 0.0; a = new double[10]; a = new double[1000000];
double a6 = 0.0; ... ...
double a7 = 0.0; a[4] = 3.0; a[234567] = 3.0;
double a8 = 0.0; ... ...
double a9 = 0.0; 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;
...
double x = a4 + a8; scales to handle huge amounts of data
an easy alternative

tedious and error-prone code

5
Memory representation of an array

An array is an indexed sequence of values of the same type.

A computer's memory is also an indexed sequence of memory locations. stay tuned for many details

• Each primitive type value occupies a fixed number of locations.

• Array values are stored in contiguous locations.

a for simplicity in this lecture, think of a as the memory address of the first location
the actual implementation in Java is just slightly more complicated.

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

Critical concepts
• Indices start at 0.
• 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
(stay tuned for details)

6
Java language support for arrays

Basic support operation typical code

Declare an array double[] a;

Create an array of a given length a = new double[1000];

Refer to an array entry by index a[i] = b[j] + c[k];

Refer to the length of an array a.length;

Initialization options
no need to use a loop like
for (int i = 0; i < 1000; i++)
a[i] = 0.0;
operation typical code

Default initialization to 0 for numeric types a = new double[1000];

BUT cost of creating an
array is proportional to
its length.
Declare, create and initialize in one statement double[] a = new double[1000];

Initialize to literal values double[] x = { 0.3, 0.6, 0.1 };

7
Copying an array

To copy an array, create a new array , then copy all the values.

double[] b = new double[a.length];

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

a i b i

0.3 0.6 0.99 0.01 0.5 0.3 0.6 0.99 0.01 0.5

Important note: The code b = a does not copy an array (it makes b and a refer to the same array).

double[] b = new double[a.length];

b = a;

a b

0.3 0.6 0.99 0.01 0.5

8
Programming with arrays: typical examples

Access command-line args in system array For brevity, N is a.length and b.length in all this code.
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];
for (int i = 0; i < N; i++)
b[i] = a[i];

Create an array with N random values

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++)
a[i] = Math.random(); System.out.println(a[i]);

Compute the average of array values Find the maximum of array values
double sum = 0.0; double max = a[0];
for (int i = 0; i < N; i++) for (int i = 1; i < N; i++)
sum += a[i]; if (a[i] > max) max = a[i];
double average = sum / N;

9
Pop quiz 1 on arrays

Q. What does the following code print?

public class PQarray1

{
public static void main(String[] args)
{
int[] a = new int[6];
int[] b = new int[a.length];

b = a;
for (int i = 1; i < b.length; i++)
b[i] = i;

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

System.out.print(a[i] + " ");
System.out.println();

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

System.out.print(b[i] + " ");
System.out.println();
}
}

10
Pop quiz 1 on arrays

Q. What does the following code print?

public class PQarray1

{
public static void main(String[] args)
{
int[] a = new int[6];
int[] b = new int[a.length];

b = a; After this, b and a refer to the same array

for (int i = 1; i < b.length; i++)
b[i] = i;

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

System.out.print(a[i] + " ");
System.out.println();

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

System.out.print(b[i] + " "); A. % java PQ4_1
System.out.println();
0 1 2 3 4 5
}
} 0 1 2 3 4 5

11
Programming with arrays: typical bugs

Array index out of bounds double[] a = new double[10];

for (int i = 1; i <= 10; i++)
a[i] = Math.random();

No a[10] (and a[0] unused)

Uninitialized array double[] a;

for (int i = 0; i < 9; i++)
a[i] = Math.random();

Never created the array

Undeclared variable a = new double[10];

for (int i = 0; i < 10; i++)
a[i] = Math.random();

What type of data does a refer to?

12
 COMPUTER SCIENCE
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

Image sources

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

CS.3.A.Arrays.Basics
 COMPUTER SCIENCE
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

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

CS.3.B.Arrays.Examples
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" };

• Ranks. String[] suit = { "♣", "♦", "♥", "♠" };

• Suits.
String[] deck = new String[52];
• Full deck. j

Use nested for loops to put all the cards in the deck.

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
0 1 2 3
for (int i = 0; i < 13; i++)
suit ♣ ♦ ♥ ♠
deck[i + 13*j] = rank[i] + suit[j];
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 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
Example of array use: create a deck of cards

public class Deck

{
public static void main(String[] args)
{
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
for (int i = 0; i < 13; i++)
deck[i + 13*j] = rank[i] + suit[j];

for (int i = 0; i < 52; i++)

System.out.print(deck[i] + " ");
System.out.println();
} % java Deck
}
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♠
%

16
Pop quiz 2 on arrays

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 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];

17
Pop quiz 2 on arrays

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 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];

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♠

18
Pop quiz 3 on arrays

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

% 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♠
10♣ 10♦ 10♥ 10♠ J♣ J♦ J♥ J♠ Q♣ Q♦ Q♥ Q♠ K♣ K♦ K♥ K♠ A♣ A♦ A♥ A♠
%

19
Pop quiz 3 on arrays

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

% 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♠
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♣ ...

20
21
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.
22
Array application: random sequence of cards

public class DrawCards

{
public static void main(String[] args)
{
int N = Integer.parseInt(args[0]);

String[] rank = {"2", "3", "4", "5", "6", "7", "8", "9","10", "J", "Q", "K", "A" };
String[] suit = { "♣", "♦", "♥", "♠" };

String[] deck = new String[52];

for (int i = 0; i < 13; i++) % java DrawCards 10
for (int j = 0; j < 4; j++)
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♦
int r = (int) (Math.random() * 52);
System.out.print(deck[r] + " ");
} % java DrawCards 10
System.out.println(); 6♠ 10♦ 4♥ A♣ K♥ Q♠ K♠ 7♣ 5♦ Q♠
}
}
% java DrawCards 10
A♣ J♣ 5♥ K♥ Q♣ 5♥ 9♦ 9♣ 6♠ K♥

Note: Sample is with replacement (same card can appear multiple times). appears twice
23
Array application: shuffle and deal from a deck of cards

Problem: Print N random cards from a deck.

Algorithm: Shuffle the deck, then deal.

• Consider each card index i from 0 to 51.
• Calculate a random index r between i and 51.
• Exchange deck[i] with deck[r]
• Print the first N cards in the deck.

Implementation for (int i = 0; i < 52; i++)

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

24
Array application: shuffle a deck of 10 cards (trace)
deck
i r
for (int i = 0; i < 10; i++) 0 1 2 3 4 5 6 7 8 9
{
int r = i + (int) (Math.random() * (10-i)); 2♣ 3♣ 4♣ 5♣ 6♣ 7♣ 8♣ 9♣ 10♣ J♣
String t = deck[r];
0 7 9♣ 3♣ 4♣ 5♣ 6♣ 7♣ 8♣ 2♣ 10♣ J♣
deck[r] = deck[i];
deck[i] = t; 1 3 9♣ 5♣ 4♣ 3♣ 6♣ 7♣ 8♣ 2♣ 10♣ J♣
}
2 9 9♣ 5♣ J♣ 3♣ 6♣ 7♣ 8♣ 2♣ 10♣ 4♣

Q. Why does this method work? 3 9 9♣ 5♣ J♣ 4♣ 6♣ 7♣ 8♣ 2♣ 10♣ 3♣

• Uses only exchanges, so the deck after
 4 6 9♣ 5♣ J♣ 4♣ 8♣ 7♣ 6♣ 2♣ 10♣ 3♣

the shuffle has the same cards as before.
5 9 9♣ 5♣ J♣ 4♣ 8♣ 3♣ 6♣ 2♣ 10♣ 7♣
• N −i equally likely values for deck[i].
6 8 9♣ 5♣ J♣ 4♣ 8♣ 3♣ 10♣ 2♣ 6♣ 7♣
• Therefore N ×(N −1)×(N −1)... ×2×1 = N !

equally likely values for deck[]. 7 9 9♣ 5♣ J♣ 4♣ 8♣ 3♣ 10♣ 7♣ 6♣ 2♣

Initial order is immaterial. 8 8 9♣ 5♣ J♣ 4♣ 8♣ 3♣ 10♣ 7♣ 6♣ 2♣

9 9 9♣ 5♣ J♣ 4♣ 8♣ 3♣ 10♣ 7♣ 6♣ 2♣

Note: Same method is effective for randomly rearranging any type of data.
25
Array application: shuffle and deal from a deck of cards

public class DealCards

{
public static void main(String[] args)
{
int N = Integer.parseInt(args[0]);

String[] rank = {"2", "3", "4", "5", "6", "7", "8", "9", "10", "J", "Q", "K", "A" };
String[] suit = { "♣", "♦", "♥", "♠" };

String[] deck = new String[52];

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

for (int i = 0; i < 52; i++)

{
int r = i + (int) (Math.random() * (52-i));
String t = deck[r];
deck[r] = deck[i];
deck[i] = t; random poker hand
}

for (int i = 0; i < N; i++) % java DealCards 5

System.out.print(deck[i]); 9♣ Q♥ 6♥ 4♦ 2♠ random bridge hand
System.out.println();
}
} % java DealCards 13
3♠ 4♥ 10♦ 6♥ 6♦ 2♠ 9♣ 8♠ A♠ 3♥ 9♠ 5♠ Q♥
26
Coupon collector

Coupon collector problem

• M different types of coupons.
• Collector acquires random coupons, one at a time, each type equally likely.
Q. What is the expected number of coupons needed to acquire a full collection?

Example: Collect all ranks in a random sequence of cards (M =

Sequence

9♣ 5♠ 8♥ 10♦ 2♠ A♠ 10♥ Q♦ 3♠ 9♥ 5♦ 9♣ 7♦ 2♦ 8♣ 6♣ Q♥ K♣ 10♥ A♦ 4♦ J♥

Collection
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♠ 22 cards needed

to complete
2♦ 5♦ 8♣ 9♥ 10♥ Q♥ A♦ collection
9♣ 10♥
27
Array application: coupon collector

Coupon collector simulation public class Coupon

{
• Generate random int values public static void main(String[] args)
{
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.
boolean[] found = new boolean[M];
while (distinct < M)
{
Key to the implementation int r = (int) (Math.random() * M);
cards++;
• Create a boolean array of length M. if (!found[r])
{
(Initially all false by default.) distinct++; % java Coupon 13
found[r] = true; 46
• When r generated, check the r th } % java Coupon 13
value in the array. } 22

• If true, ignore it (not new). System.out.println(cards); % java Coupon 13

} 54
• If false, count it as new distinct
 } % java Coupon 13
value (and set r th entry to true) 27

28
Array application: coupon collector (trace for M = 6)

found
r distinct cards
0 1 2 3 4 5

F F F F F F 0 0

boolean[] found = new boolean[M]; 2 F F T F F F 1 1

while (distinct < M)
{ 0 T F T F F F 2 2
int r = (int) (Math.random() * M);
cards++; 4 T F T F T F 3 3
if (!found[r])
0 T F T F T F 3 4
{
distinct++;
1 T T T F T F 4 5
found[r] = true;
} 2 T T T F T F 4 6
}
5 T T T F T T 5 7

0 T T T F T T 5 8

1 T T T F T T 5 9

3 T T T T T T 6 10

29
Simulation, randomness, and analysis (revisited)

Coupon collector problem

• M different types of coupons.
• Collector acquires random coupons, one at a time, each type equally likely.
Q. What is the expected number of coupons needed to acquire a full collection?
Pierre-Simon Laplace
1749-1827
A. (known via mathematical analysis for centuries) About M ln M + .57721M .

type M expected wait

% java Coupon 4
11
playing card suits 4 8
% java Coupon 13
playing card ranks 13 41 38

baseball cards 1200 9201 % java Coupon 1200

8789
Magic™ cards 12534 125508
% java Coupon 12534
125671
Remarks
• Computer simulation can help validate mathematical analysis.
Example: Is Math.random()
• Computer simulation can also validate software behavior. simulating randomness?
30
Simulation, randomness, and analysis (revisited)

Once simulation is debugged, experimental evidence is easy to obtain.

Analogous code for coupon collector, this lecture
Gambler's ruin simulation, previous lecture
public class CouponCollector
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]);
int goal = Integer.parseInt(args[1]); int cards = 0;
int trials = Integer.parseInt(args[2]); boolean[] found;

int wins = 0; for (int i = 0; i < trials; i++)

for (int i = 0; i < trials; i++) {
{ int distinct = 0;
int t = stake; found = new boolean[M];
while (t > 0 && t < goal) while (distinct < M)
{ {
if (Math.random() < 0.5) t++; int r = (int) (Math.random() * M);
else t--; cards++;
} if (!found[r])
if (t == goal) wins++; {
} distinct++;
System.out.println(wins + " wins of " + trials); found[r] = true;
} }
} }
}
System.out.println(cards/trials);
}
}
31
Simulation, randomness, and analysis (revisited)

Coupon collector problem

• M different types of coupons.
• Collector acquires random coupons, one at a time, each type equally likely.
Q. What is the expected number of coupons needed to acquire a full collection?

Predicted by mathematical analysis Observed by computer simulation

type M M ln M + .57721M
% java CouponCollector 4 1000000
playing card suits 4 8 8

% java CouponCollector 13 1000000

playing card ranks 13 41 41

% java CouponCollector 52 100000

playing cards 52 236
236

baseball cards 1200 9201 % java CouponCollector 1200 10000

9176

magic cards 12534 125508 % java CouponCollector 12534 1000

125920

Hypothesis. Centuries-old analysis is correct and Math.random() simulates randomness.

32
 COMPUTER SCIENCE
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

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

CS.3.B.Arrays.Examples
 COMPUTER SCIENCE
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

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

CS.3.C.Arrays.2D
Two-dimensional arrays

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

grade

Examples 0 1 2 3 4 5 ...
0 A A C B A C
• Matrices in math calculations.
1 B B B B A A

student ID
• Grades for students in an online class. 2 C D D B C A
• Outcomes of scientific experiments. 3 A A A A A A
4 C C B C B B
• Transactions for bank customers.
5 A A A B A A
• Pixels in a digital image. ...
• Geographic data
• ...

y-coordinate
Main purpose. Facilitate storage and manipulation of data.
x-coordinate 35
Java language support for two-dimensional arrays (basic support)

operation typical code

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

Create a two-dimensional array of a given length a = new double[1000][1000];

Refer to an array entry by index a[i][j] = b[i][j] * c[j][k];

Refer to the number of rows a.length;

can be different
Refer to the number of columns a[i].length;
for each row

Refer to row i a[i] no way to refer

to column j

a[][]

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

36
Java language support for two-dimensional arrays (initialization)

no need to use nested loops like

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];
for numeric types BUT cost of creating an
array is proportional to
its size.
Declare, create and initialize

double[][] a = new double[1000][1000];
in a single statement

double[][] p =
{
{ .92, .02, .02, .02, .02 },
{ .02, .92, .32, .32, .32 },
Initialize to literal values
{ .02, .02, .02, .92, .02 },
{ .92, .02, .02, .02, .02 },
{ .47, .02, .47, .02, .02 },
};

37
Application of arrays: vector and matrix calculations

Mathematical abstraction: vector Mathematical abstraction: matrix

Java implementation: 1D array Java implementation: 2D array

Vector addition Matrix addition

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

for (int i = 0; i < N; i++) for (int i = 0; i < N; i++)
c[i] = a[i] + b[i]; for (int j = 0; j < N; j++)
c[i][j] = a[i][j] + b[i][j];

.70 .20 .10 .80 .30 .50 1.5 .50 .60

.30 .60 .10 + .50 .10 .40 = .80 .70 .50 .30 .60 .10 + .10 .40 .10 = .40 1.0 .20

.50 .10 .40 .10 .30 .40 .60 .40 .80

38
Application of arrays: vector and matrix calculations

Mathematical abstraction: vector Mathematical abstraction: matrix

Java implementation: 1D array Java implementation: 2D array

Vector dot product Matrix multiplication

double sum = 0.0; double[][] c = new double[N][N];

for (int i = 0; i < N; i++) for (int i = 0; i < N; i++)
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];
.30 .60 .10 · .50 .10 .40 = .25

i x[i] y[i] x[i]*y[i] sum

0 0.3 0.5 0.15 0.15

.70 .20 .10 .80 .30 .50 .59 .32 .41
1 0.6 0.1 0.06 0.21
.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
Pop quiz 4 on arrays

Q. How many multiplications to multiply two N-by-N matrices?

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

for (int i = 0; i < N; 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

2. N 2

3. N 3

4. N 4

40
Pop quiz 4 on arrays

Q. How many multiplications to multiply two N-by-N matrices?

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

for (int i = 0; i < N; 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

2. N 2

3. N 3 Nested for loops: N × N × N

4. N 4

41
Self-avoiding random walks

A dog walks around at

random in a city, never
revisiting any intersection.

Q. Does the dog escape?

escape
Model: a random process in an N-by-N lattice
• Start in the middle.
• Move to a random neighboring intersection

but do not revisit any intersection.
• Outcome 1 (escape): reach edge of lattice.
• Outcome 2 (dead end): no unvisited neighbors.

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.
42
Self-avoiding random walks

43
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%
}
}
50%
System.out.println(100*deadEnds/trials + "% dead ends");
}
25%
}
0%
10 20 30 40 50 60 70 80 90 100

44
Simulation, randomness, and analysis (revisited again)

Self-avoiding walk in an N-by-N lattice

• Start in the middle.
• Move to a random neighboring intersection (do not revisit any intersection).

Applications
• Model the behavior of solvents and polymers.
• Model the physics of magnetic materials.
• (many other physical phenomena)
Paul Flory
1910-1985
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
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
Your first data structure

Arrays: A basic building block in programming

• They enable storage of large amounts of data (values all of the same type).
• With an index, a program can instantly access a given value.
• Efficiency derives from low-level computer hardware organization (stay tuned).

Some applications in this course:

^ digital images N-body simulation

LFSR

digital audio

46
 COMPUTER SCIENCE
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

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

CS.3.C.Arrays.2D
 COMPUTER SCIENCE
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
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