advanced

' . APPLICATIONS AND PROBLEMS

JAMES S. COAN
CID@JWCIDITU©®@J
~&~O©
Hayden Computer Programming Series

COMPREHENSIVE STANDARD FORTRAN PROGRAMMING

James N. Haag
BASICS OF DIGITAL COMPUTER PROGRAMMING (Second Ed)
John S . Murphy
BASIC BASIC: An Introduction to Computer Programming in BASIC Language (Second Ed.)
James S. Coan
ADVANCED BASIC: Applications and Problems
James S Coan
DISCOVERING BASIC: A Problem Solving Approach
Robert E. Smith
ASSEMBLY LANGUAGE BASICS: An Annotated Program Book
Irving A. Dodes
PROGRAMMING PROVERBS
Henry F. Ledgard
PROGRAMMING PROVERBS FOR FORTRAN PROGRAMMERS
Henry F. Ledgard
FORTRAN WITH STYLE
Henry F, Ledgard and Louis J. Chmura
COBOL WITH STYLE: Programming Proverbs
Louis J, Chmura and Henry F. Ledgard
SNOBOL: An Introduction to Programming
Peter R Newsted
FORTRAN FUNDAMENTALS: A Short Course
Jack Steingraber
THE BASIC WORKBOOK: Creative Techniques for Beginning Programmers
Kenneth E. Schoman, Jr.
BASIC FROM THE GROUND UP
David E. Simon
 CID@JWCIDITU©®@J
lID&~D©
APPLICATIONS AND PROBLEMS

JAMES S. COAN
Community Computer Corporation
Germantown Friends School

[JJ
HAYDEN BOOK COMPANY, INC.
Rochelle Park, New Jersey
Library of Congress Cataloging in Publication Data

Coan, James S
Advanced BASIC.

(Hayden computer programming series)

Bibliography: p.
Includes index.
1. Basic (Computer program language).
2. Mathematics--Data processing. I. Title.
QA76. 73. B3C6 001. 6'425 76-7435
ISBN 0-8104-5856-X
ISBN 0-8104-5855-1 pbk.

Copyright© 1977 by HAYDEN BOOK COMPANY, INC. All rights reserved.

No part of this book may be reprinted, or reproduced, or utilized in any
form or by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying and recording, or in any infor-
mation storage and retrieval system, without permission in writing from
the Publisher.
Printed in the United States of America

6 7 B 9 PRINTING

78 79 BO 81 82 83 84 85 YEAR
 PREFACE

Students are finding that after they have been through the intro-
ductory texts presenting BASIC they are pretty much on their own.
They may get additional bits and pieces of assistance from other texts,
finding that each adds something to their repertoire, but there is still
all that introductory material to sort through. The present text is in-
tended for those who have been introduced to the BASIC language and
want to go further with the language. It is also intended for those who
have already learned another language (such as FORTRAN or COBOL)
and need only a brief introduction to what is for them a second or
third programming language.
A review chapter is placed at the beginning of the book for those
whose coverage of BASIC may have been somewhat limited or distant
in time, and for those who come to it with competence in one or more
other programming languages. In fact, this text can even be used by
students with no former programming experience, if tutorial assistance
is available, by a more extensive use of the review chapter.
Some extended features of BASIC are presented so that students
working on systems providing extensions will have an opportunity to
experiment with or master new techniques. Strings and files are intro-
duced in Chapters 3 and 4. Due to the variety of implementations for
use of both strings and files, two systems are presented in each area:
General Electric and Hewlett Packard.
In addition to other topics, the applications chapters cover
coordinate geometry, area, sequences and series, polynomials, graphing,
simulation, and games. These chapters may be studied independently
and in any order, although, where appropriate, the student may occa-
sionally be referred to another section in the text. Generally these
topics evolve from dr build on the ground of a second course in algebra
and beyond.
'.Vhe topic of efficiency is treated explicitly on several occasions,
though not necessarily to the point of optimizing execution time. The
topic of structured programming is treated implicitly in that all exam-
ple programs incorporate a clear programming style with minimal un-
conditiona l branching and maximal use of appropriate data and control
structures. These two topics provide good stepping-off points in a
computer science course.
Appendix A presents an abbreviation of the ASCII character
codes. Appendices B and C summarize flowchart shapes and program
statements in BASIC. Appendix Dis an index of the nearly 100 demon,
strati on programs in the text. Some solution programs for selected
problems in the text follow these appendices.
I wish to thank Community Computer Corporatio n for computer
time, General Electric Informatio n Services for assistance, and those
who commented on the first draft for invaluable suggestions.

JAMES s. COAN
Philadelphia
 CONT ENTS

1 Review of BASIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1-1 Introductio n, 1
1-2 Some Simple Programs, 1
1-3 Functions, Loops, and Lists, 7
1-4 User Defined Functions, Subroutines, and Arrays, 12
1-5 MAT Statements in BASIC, 18
2 Some Extended Features of BASIC. . . . . . . . . . . . . . . . . . . . . . . . 23
2-1 Introductio n, 23
2-2 TAB, 23
2-3 Print Using, 25
2-4 Logical Operations, MAX and MIN, 27
2-5 Multiple Line Defined Function, 29
3 Strings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3-1 Introductio n, 31
3-2 The String Simple Variable, 31
3-3 The String Subscripted Variable, 34
3-4 The Substring Scheme, 35
3-5 The String Array Scheme, 40
4 Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4-1 Introductio n, 46
4-2 Hewlett Packard Files, 46
4-3 General Electric Files, 54
5 Plotting on the Terminal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5-1 Introductio n, 64
5-2 Plotting a Function, 64
5-3 Plotting Using Data Stored in an Array, 68
6 Area Under a Curve. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
6-1 Introductio n, 72
7 Coordinate Geometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
7-1 Points in a Plane, 78
7-2 Pairs of Points, 78
7 -3 Parallel and Perpendicular Lines, 84
7-4 Point of Intersection of Two Lines, 88
7-5 Three Points in a Plane, 93
 8 Polynomials. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
8-1 Introduction, 101
8-2 Finding a Real Zero of a Polynomial, 104
8-3 Synthetic Division to Get Further Zeros, 108
8-4 Miscellaneous Aids, 112
9 Sequences and Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
9-1 Sequences, 115
9-2 Convergence and Divergence, 119
9-3 Series, 122
9-4 More on Series, 125
10 Matrices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
10-1 Introduction, 127
10-2 Solving Simultaneous Linear Equations Using MAT INV, 131
11 Some Topics of Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
11-1 Introduction, 137
11-2 Average, Variance, and Standard Deviation, 137
11-3 Median, 140
11-4 Coefficient of Linear Correlation, 143
12 Simulation and Games. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
12-1 Introduction, 147
12-2 Lines at the Bank, 147
12-3 Magic Squares, 152
12-4 Games, 154
Appendix A ASCII Character Set Printable on Model 33 161
Appendix B Summary of Flowchart Shapes . . . . . . . . . . . . . . . . . . . 162
Appendix C Summary of Statements in BASIC . . . . . . . . . . . . . . . . . 163
Appendix D Index of Programs in Text . . . . . . . . . . . . . . . . . . . . . . 169
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Answers to Selected Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
 1
REVIEW OF
BASIC

1-1 Introduction
This chapter is intended to serve several purposes. The student who is al-
ready competent in BASIC will move quickly through the chapter and perhaps
write a few programs. The student who has studied BASIC, but not recently or
not extensively, will want to move more slowly and write more programs. The
student who comes to BASIC with competence in another language will be
mainly interested in the differences between the language with which he or she is
familiar and BASIC.
There are numerous implementations of BASIC available in schools, col-
leges, and businesses today. These implementations have many common fea-
tures and some differences. This chapter will concentrate primarily on those
features which are almost universally available. We will mention some of the
more common variations (you can usually determine which features are part of
your system by writing very short programs to see what works), but beyond that
the student should obtain the specific features for his system from the BASIC
reference manuals supplied by the computer center or the vendor.
It is also true that the specifications sometimes change as the people
responsible for maintaining the computer update its language capabilities. Since
these changes tend to be additions, however, programs previously tested will
usually still run.
1-2 Some Simple Programs
We can demonstrate many features of BASIC by writing a program that
will compare two numbers to determine whether the first is greater than, less
than, or equal to the second. See program C</>MP AR.
2 Advanced BASIC

CI ST
CGMPAR

94 RF:M * THIS IS A SAMPLF PRCrHAM TC

95 REM CC MPARE. J',JC N'JM'1E ~ S FC: I< 0 HO:O: I<
100 READ A,fl
I I0 IF A = •0 l TYFN 220
l?.O JF A = '1 T4FN 140
130 IF A < g THFN !RO
1'•0 PRINT A; "IS r·ffl'TF:R T4A'J"; 9
150 GOTO lCJO
140 PRINT A; "IS 'IJ'IAL TC"; ~
170 GOTC 100
!RO PRINT A; "IS L'SS T4AN"; R
190 GOTO 100
192
194 R'"
~00 DATA 3,4, 1.7,1.1, J1,11, -J,~, o,o
?.10 DATA .01,0
220 F:ND
RIJN
r.OMPAR

3 IS LESS THAN 4
J .7 JS GRF:ATF:K THA~J J. J
31 IS EOIJAL. TC 31
-3 IS LESS THAN ?.
0 JS FO'IAL TC 0

The first item shown is LIST. This is a system command rather than a
program statement, and we type it to instruct the computer to print out the
program exactly as it stands. Next the computer automatically prints C<t'>MPAR,
which is the program name. Some computers also print the time of day and date
along with the program name. On some systems we assign program names with
the system command NAME-C<t'>MPAR. On others, to name a new program,
NEW Cq.>MP AR is typed before the program itself is. Having typed the pro-
gram name, the computer goes on to list the program itself. Let us examine the
program statements.
Read-Data
Line 100 is a READ statement. In this case we want the computer to
READ two numbers into two variables A and B. Those numbers must come
from one or more DATA statements. We provide data in lines 200 and 210.
Conditional Transfer
Lines 110, 120, and 130 are all examples of the conditional transfer in
BASIC. Line 110 is used to terminate the execution of the program itself. We
send the computer to the END statement only if the value of A is .01. This is an
example of the use of dummy data to control program execution. Lines 120
and 130 direct the computer to the appropriate PRINT statement according to
the relation between the values of A and B. BASIC also allows "greater than"
(>), "greater than or equal to" (>=) and "less than or equal to" (<=). To test
for "not equal to" use (< >). Some systems also allow"#" for "not equal to."
In place of the algebraic symbols just mentioned, some systems require
special symbols as relational operators. They are (\LT) or (LT) for "less than,"
 Review of BASIC 3

(\LE) or (LE) for "less than or equal to," (\GT) or (GT) for "greater than,"
(\GE) or (GE) for "greater than or equal to," (\EQ) or (EQ) for "equal to"
and (\NE) or (NE) for "not equal to." The paired relational operators listed
above are not generally interchangable. The first is for certain systems, and the
second is for others. So there are three sets of possible relational operators.
You can quickly tell which works on your computer by consulting the vendor-
supplied manual or by trial and error.
Another conditional transfer is available on many systems. It is called the
computed G</JT</J. It takes the form,
100 </JN K G</JT</J niln 2 ,n 3 , etc.
or

or
100 G</JT</J n 1 ,n 2 ,n 3 , etc, </JN K
At line 100 computer control passes to line n 1 if K = 1, 1 2 if K = 2, etc. If K
is not in the range from one to the number of line numbers named, some sys-
tems terminate with an error message, while others simply pass control to the
line after 100.
Print
Lines 140, 160, and 180 result in printed output to the terminal for the
operator to see. We may mix literal output with numeric results by enclosing
literal messages in quotation marks. Replacing semicolons in the PRINT state-
ment with commas would result in wider spacing of printed output. On many
systems, using commas to separate printed results causes the page to be divided
into five columns of 15 characters each. For terminals with only 72 character
spaces, the fifth column is 12 characters wide. Generally speaking, use of a semi-
colon to separate printed output results in closer spacing than with a comma.
Unconditional Transfer
Lines 150, 170, and 190 are examples of unconditional transfer. The
G</JT</J statement in BASIC serves to name the number of the next line to be
executed. When the computer gets to line 150, the next line the computer exe-
cutes is 100. The same is true for lines 170 and 190. Generally speaking, good
programs try to minimize the number of G</JT</J statements. We shall see ways
to do this later.
End
The final statement in our example is the END statement. On most sys-
tems the highest numbered statement must be an END statement. After the
END statement we see the system command RUN, which is typed by us to cause
the computer to actually carry out the instructions of the program. In response
to the RUN command, the computer has printed the program name, C</JMPAR
(some systems will also print the date and time), followed by the printed output
4 Advanced BASIC

specified by the instructions of our program. (Some systems will follow pro-
gram results with information about the computer resources used by this RUN
of the program.)
Variables
In program C</>MP AR we used A and B as variables to name stored nu-
meric values. BASIC allows us to use every letter of the alphabet and also every
letter of the alphabet followed by a single digit. (Some systems also provide ad-
ditional symbols as variables.) Every letter or letter plus digit allows us to store
a single number at any one time in a program. However, variables may be used
over and over again to store new values as long as we have no further need for a
particular old value. In C</>MPAR, the variables A and B were each used for six
different numbers.
Prettyprinting
Note the overall appearance of program Cl/.>MPAR. An attempt has been
made to provide spacing within the program statements to facilitate readability
and therefore clarity of thought. The practice of indenting and spacing to
achieve this goal is called "prettyprinting." In program C</>MPAR, the IF state·
men ts have been indented three spaces. Spaces have been inserted in the PRINT
statements to avoid a crowded appearance. The blank REM in line 194 is used
to offset the DATA portion of the program, and the data has been grouped in
the DATA statements to show just how the values will be read in the READ
statement. Some systems do not allow prettyprinting. However, if your system
does allow prettyprinting, you should work on developing a style of spacing to
enhance program readability. Prettyprinting becomes more and more worth-
while as programs become longer and more complex.
Arithmetic Operations
The computer is often used to perform arithmetic opera.!;ions on numbers.
The operations allowed are exponentiation (**) or (t), multiplication (*),
division (/), addition (+) and subtraction (- ). The priorities assigned these opera-
tors are the same as those assigned in conventional algebra, that is, first exponen-
tiation followed by multiplication and division followed by addition and sub-
traction. Program </>PRATN shows a use of each of the arithmetic operators.

CPRATN

94 RFM * T41S IS A SAMPLF P~OrRAM TO DEMONSTRATF.

95 l'F.M 'ISF: Cl' ARITHMF:TIC flpf:;,ATCRS IN RASJr;
100 PRINT "AtR", "A*R", "A/B", 0
A+R", "A-A"
110 l'EAD A,8
120 JI' A <> 0 THEN 150
130 JI' Fl <> 0 T4f:N l~O
140 STOP
150 PRINT " A ="; A; "''l ="; 8
l~O PRINT Atfl, A•'l, A/'l, A+B, A-8
170 PRINT
lf!O GOTO 110
1R2
184 REM
190 DATA 1,2, 3,~ .. 2 .. ~o .. 1.~,~s.2
200 DATA O,O
210 END
 Review of BASIC 5

RIJN
OPRATN

AtB A*B A/R A+B A-9

A = l A =2
2 o.s 3 - l

A 3 8 = 4
-1
81 12 0.75

A 2 8 = 40
l. 0995 l E. + 12 BO o.os 42 - 38

A = l. 4 B =
65.2
66.6 -63. R
3.36936E+9 91·2R 2.14724E-2

The RUN of <l>PRATN produces three examples of what is called E-format.

For A 2 and B = 40, A**B results in l.09951E+12. That means 1.09951
times 10 to the twelfth power, or 1.09951 X 10 12 • Systems vary, but many
provide from six to nine significant digits of numeric output.
In <l>PRATN, line 140 is equivalent to G<l>T<l> 210. The ST<l>P statement
in BASIC is used to terminate execution of a program at some point other than
the highest numbered line. This line of the program is called a "logical end" to
differentiate it from the physical end.
Assignment
Thus far, the way that we have gotten numeric values to be stored in vari-
ables has been to READ values from DATA. We can also assign values directly
as follows:
100 x 3 or 100 LET X 3
200 y 3*X+5 or 200 LETT 3*X+5
These are examples of the assignment statement in BASIC. (On some sys.
terns the LET is required. On others it is optional.) Used in this way, the equals
sign is called the assignment operator. The assignment capability greatly en-
hances the power of any programming language since it permits us to retain
values for later use. For example, we can sum up any number of data items as in
program ADD.

.ADD

9L.! REM * T4IS PRcc:~Atv! ADOS f\JIJ·\11f:\i:"~S r;'f~:J',1 DATA

- 100 LFT S = ()
11 n 1<0-on A
120 IF A = - .()I TY"'J 16()
140 L"T S = S+o
150 GCTC 110
160 PRINT "S'IM IS"; S
172
174 RF.M
!RO DATA 21, 39, 11.3, 24.6, 91.J
190 DATA -.01
~00 END
R!JN
ADD

SIJM IS 1R7.2
6 Advanced BASIC

In line 100 of ADD the summing variable S is initialized at zero. (Some

implementations of BASIC automatically initialize all variables to zero when the
program is run.) We now generalize program ADD to count the number of num-
bers in the previous program. See lines 100, 130, and 140 of ADDI.

ADDI

94 ~C:M *T41 S PRCG~l\M AOOS ANO f;C'l'ITS

95 R<oM N'IM'lFRS l'RQ'l DATA
-100 u:T s =
r; o=
I I0 RF:AD A
120 IF A = - .01 THF:'-1 160
-130 u:T c = c+1
-140 Ll'T S = S+A
150 GOTO 110
160 PRINT "S'l" IS"; S

I72
17 4 REM
!RO DATA 21, 39, 11.J, :'!4.~, 91·3
190 DATA -.QI
200 END
RIJC)
ADDI

S'J'l IS I 8 7 • ?.
THERE ARE 5 CJJ'1Rl'riS

Line 100 of ADDl is an example of a multiple assignment statement. It

allows us to assign the rightmost value to all of the variables separated by equals
signs. (Some systems require commas instead of equals signs for all but the
rightmost equals signs. Our statement 100 would be 100 LET S,C = 0 on such
a system. You may be able to assign different values to different variables on
one line, for example, 100 LETH = 4, Y = 9*K, B = 81, or even 100 LET
H,I,J 3, T = -32.)
Input
The final statement of this review section which results in variables con-
taining numeric values is the INPUT statement. It is this statement which allows
the operator to interact with a program during execution. When the computer
executes an input statement, it prints a question mark at the terminal and awaits
information from the keyboard. If we replace READ A with INPUT A in pro-
gram ADD and remove the DATA statement, we have a program that behaves a
little like an adding machine. By printing a marker such as # followed by a
semicolon in line 130 we can type our selected numbers on the same line as
the marker, as in program ADD2.

ADD2

94 REM * THIS PROGRAM WORKS A LI TTL~

95 REM LIKE AN ADDING MACHINE
I 00 PRINT "IN P'JT q = - • 0 I TC 0 fl TA IN TC TAL"
I IO PRINT
120 LET S = O
-130 PRINT "#";
140 INPUT A
150 IF A =
-.QI THEN 180
160 LET S = S+A
 Review of BASIC 7

170 GOTO 130

180 PRINT "SUM IS"; S
190 END
RUN
ADD2

INPIJT H -.01 TC C'lTAIN TCTAL

#?45
#?7 g
#? 34
#?98-12
#?43
#? - • 0 I
S'JM IS 29~·12

Note that input statements may be used to call for several values. 100 IN-
PUT X,Y,B9 calls for three numbers to be typed, separated by commas at the
keyboard.
Summary of Sec. 1-2
We have looked at three system commands to give a program a name, in-
struct the computer to RUN a program, and to LIST a program. We have used
the following eight statements in programs; READ, DATA, IF-THEN, PRINT,
END, ST</>P, LET, and INPUT. With just these statements we are able to write
substantial programs (nevertheless, the language does contain tremendous ad-
ditional power in other statements to come in the next sections). The concepts
of variable and program control have been discussed. We have looked at three
kinds of operators: arithmetic, relational, and assignment.
Problems for Sec. 1-2
1) Write a program to average numbers entered as DATA or on INPUT.
2) Write a program to find the largest and/or smallest number of a set of
DATA.
3) Write a program to repeatedly average groups of numbers.
4) Write a program to add all positive integers from 1 ton, where n is an
item of DATA.
5) Write a program to calculate n factorial. Be sure to make the value of
O! equal 1.
6) Write a program to solve equations of the form, ax + b = ex + d.
7) Write a program to solve quadratic equations, ax 2 + bx + c = 0.
8) Write a program to find the sum of the reciprocals of the first n positive
integers.
1-3 Functions, Loops and Lists
BASIC provides computer functions or subroutines for special purposes.
INT(X), SGN(X), ABS(X), and SQR(X) are among them. On some systems the
INT(X) function takes the greatest integer not greater than X, such as 4 for 4.3
and - 2 for -1.6. Other systems simply give the integral part of X, such as 4 for
4.3 and -1 for -1.6, by removing the decimal part. For non-negative numbers,
the two are equivalent. SGN(X) becomes +l if Xis positive, zero if Xis zero,
and -1 if Xis negative. ABS(X) becomes X if Xis non-negative and becomes - X
if X is negative. SQR(X) becomes the principle square root of X so long as Xis
8 Advanced BASIC

non·negative. Each of these functions is useful. However, it is likely that of

the four, INT(X) is most frequently used.
One use of INT(X) is to round off numeric results. For example, to round
to the nearest integer, use INT(X+.5). To round to the nearest hundredth, use
INT(lOO*X+.5)/100, etc. See line 150 of program R</HJND.

HO'JND

94 REM *
THIS PHOGRAM DF.MONSTRATF.S A USE CF THF.
95 REM INT< l FIJNCTICN JN BASH; FC·{ HC'J'.'JDING
96 REM NUMRERS
100 LET E 0 =
110 LET X =
t.R2564
120 PRINT "RO'JND"; ~
130 PRINT
140 PRINT "POWER CF TENTHS", " ROUNDS TC"
-150 LET R = INT< ~•IOtF.+.5 )/!OtE
160 PRINT E, R
162
164 REM *
LINF. 170 JNCRFMF.NTS TYE EXPONUIT OF 10
170 LET E E+l
!RO JF F. <= 5 THF.N 150
190 $ND
RIJN
RO'JND

ROU'JD J. R25~4

POWER OF TENTHS R0 1 l•'JDS TO

0 ?.
1 t.'l
2 t.R3
3 l·R'26
4 t.R251\
5 t.R2564

INT(X) is also used frequently to test numbers for divisibility by other

numbers. For example:
100 IF X/10 = INT(X/10) THEN 200
transfers the computer to line 200 if Xis divisible by 10, but the computer goes
to the line immediately following 100 if X is not divisible by 10. One problem
solution that uses this is that of finding all factors of a certain number. We
simply try all integers from 2 to the number and test for divisibility. If the divi-
sor goes evenly, we print it; if not, we go to the next divisor. This is left as an
exercise.
Random Numbers
BASIC provides a routine to generate pseudo-random numbers. These are
very useful as a source of data for simulating random events. The procedure
varies somewhat from system to system. All systems provide decimal numbers
in the range, 0 to 1. One procedure introduces random numbers by using
RND(X) in a statement such as 100 LET A = RND(X). The particular ran-
dom numbers are duplicated or different from one run to the next according
to the value of X. If X is negative, then you get a different set of random num-
bers from run to run. If Xis 0, then each run of the program produces the same
succession of random numbers. And if X is positive, then the set of random
 Review of BASIC 9

numbers is based on the value of X. A second procedure does not require an

argument for RND. The statement 100 LET A = RND selects a random num-
ber and assigns it to A. However, used alone it generates the same set of random
numbers from one run to the next. To obtain a different set of random num-
bers from run to run, simply introduce the companion statement 10 RAND<l>-
MIZE into the program. The ability to reproduce the same set of random
numbers is useful for finding errors in the program debugging process. Once the
program has been perfected, it can then be modified to produce different results
for each run.
Most of the uses for random numbers are for numbers in some range other
than 0 to 1. So we have to do the appropriate calculations to change the range.
For example, to "roll a die" we would use 100 LET R = INT(6*RND(-l )+l)
or 100 LET R = INT(6*RND+l) . This would give integers in the range 1 to 6
inclusive. In the statement, 200 LET N = INT(A*RND+B), A specifies the
possible number of random integers and B specifies the smallest possible random
integer.
Loops
Suppose we place ten slips of paper numbered 1 to 10 into a hat and have
five people draw one slip of paper, note the number on it, and return the slip to
the hat. This we can easily do with a program, as in program DRAW.
Program DRAW and several other programs we have looked at contain
examples of repetitive steps. This is a computer loop, and BASIC provides the
F<l>R-NEXT statement pair to set up loops. Program DRAWOl uses F<l>R-NEXT
to accomplish the same purpose as program DRAW. Note that both programs
simulate returning the slips to the hat.

DRAW

94 REM * THIS PROGRAM SIMULATES RANDCM DRAWING

95 REM FIVE NUMBERS FROM AMONG 10 WITH REPLACEMENT
100 RANDOMIZE
110 LET x = 1
120 LET R = INT< RND*lO+l )
130 PRINT Rl
140 LET x X+l
150 IF x <= 5 THEN 120
160 END
RUN
DRAW
6 9 6 10 B

DRAWOI

94 REM * THIS PROGRAM DIFFERS FHOM DRAW

95 REM IN THAT FOR-NEXT IS NOW INTRCDUCED
100 RANDIJMIZE
-110 FOR X = 1 TO 5
120 LET R = INT< RND•lO+l l
130 PRINT RI
140 NEXT X
142
150 END
RIJN
DRAWOl

2 9 5 2 10
10 Advanced BASIC

Lists
Now suppose that the five people who drew slips of paper from the hat
did not replace them. We must make sure that no number is drawn twice. In
order to achieve this goal we need to be able to keep track of what numbers have
been drawn and what numbers are left. This can be done rather nicely using a
subscripted variable called a "list." A list, designated by a single letter, reserves
space for more than one number. At the time that we first designate a list
variable, the computer automatically reserves 10 locations (some computers re·
serve 11 by allowing zero as a subscript). We can simulate the numbered slips
of paper in the hat by establishing a 10-item list with the integers 1 through 10
stored in locations 1through10. This is done by the following routine:

100 FOR I = I TO 10
110 LETL<Il=l
120 NEXT I

Line 110 sets up the subscripted variable L( ) so that L(l) = 1, L(2) = 2, etc.,
through L(lO) = 10. If we need more than 10, we can get them by using the
DIMension statement. This is usually placed at the very beginning of the pro-
gram. For example, if we want 25 slips of paper in our hat, we would begin with
10 DIM L(25), and 25 locations would be available. Note that we could dimen-
sion for 25 and then use only 10 of them, but we cannot use more than the
DIM statement specifies. Any number of lists may be dimensioned on the same
line as: 10 DIM A(36), B(43). Systems allow a maximum number of storage
locations from a few thousand to many thousands.
Now we can develop a procedure to select five numbers at random without
replacement. For the first draw there will be 10 numbers from which to draw;
for the second draw one has been removed, leaving nine. The process is re-
peated until there remain only six from which to draw for the last draw. This
can be done with another FG)R-NEXT pair. We can step backwards with FG)R
J = 10 TG) 6 STEP -1. We may specify all three numbers on the right of the
equals sign by variable or formula.
If we design a program that merely checks to see if the latest number has
already been drawn, then we get into a trial and error situation. And if we try
to draw 10 numbers from among 10, or worse yet, 100 from 100, the drawing
gets slower and slower as we get nearer and nearer to the last draw (adding sus-
pense and expense). For efficiency's sake we should avoid this pure trial and
error scheme. Thus consider the following procedure for eliminating trial and
error entirely.
For the first draw we may select a number R at random from one to 10.
We may use the value of L(R) as the number on our randomly selected slip of
paper. If we draw the same value of R later on, we need a method that does not
require testing to see if L(R) has been used. We can acheive this by simply re-
placing the value of L(R) with the value of L(J) where J is the number of slips
from which the drawing is being made after each drawn number has been
printed. The important consideration here is that this scheme allows us to use
every number that is drawn. We have taken care here to develop an efficient
algorithm to solve the problem submitted to us. See in particular line 160 in
program DRA W02.
 Review of BASIC 11

94 REM * THIS PROGRAM SIMULATES HANDOM DHAWING WITHOUT

9S REM REPLACEMENT AND WITHCIJT TRIAL AND E.RRCR
96
97 REM* l!NFS 100-120 SIM'ILATE 10 SLIPS
98 REM OF PAPER IN A HAT
99 RANDOMIZE
100 HJ R I = 1 TO 1 O
110 LET Lill I =
120 NEXT I
122
124 REM * NOW DRAW FIVE NUMAERS AT RANDOM
130 FOR J = 10 TO 6 STEP -1
140 LET R =
INTI RND•J+l I
!SO PRINT LIRll
!S2
154 REM * NOW REPLACE THE NIJMBER JUST PRINTED
!SS REM WITH THE LAST NUMBER IN THE LIST
~160 LET LIRI = LIJI
170 NEXT J
172
180 END
RUN
DRAW02

9 7 3 4 s

We can see from a run of DRA W02 that no number has been drawn more
than once, but one run is not a certainty. A further check can be made by draw-
ing all ~O. We leave this as an exercise.
More Functions
In addition to the functions already described, the following are univer-
sally available: SIN, C</JS, ATN, L</JG, and EXP. SIN(X), C</JS(X) and TAN(X)
give the sine, cosine, and tangent of X, where X is taken as an abstract dimen-
sionless number or the measure of an angle in radians. ATN(X) gives the princi-
ple arctangent in radians, where X is the tangent of the required number.
L</JG(X) gives the natural logarithm of a non-negative X and EXP(X) gives ex.
Besides these there are other functions which may or may not be available on a
particular system. Among these will be log base 10, secant, cosecant, etc. On
some systems CLK(X) gives the present time using a 24-hour clock, and TIM(X}
gives the elapsed time of program run in seconds. It is best to consult the
manual for your system to determine the full extent of function capabilities.
Summary of Sec. 1-3
BASIC generally provides INT, SQR, SGN, ABS, SIN, C</JS, TAN, ATN,
EXP, and L</JG as standard functions, where trigonometric functions are based
on radian measure and exponentials use e as the base. The routine RND is avail-
able to provide random numbers. The language created loop is F</JR A = B
T</J C STEP D, where the loop variable A first takes on the value of B and
increments by D until A passes C and control passes to the next line following
NEXT A which closes the BASIC loop. The default STEP value is always 1.
The list is available as a subscripted variable to allow block storage of several
numbers attached to a single letter. The DIM statement is required for sub-
scripts exceeding 10.
12 Advanced BASIC

Problems for Sec. 1-3

1) Write a program to find the absolute value without using any of the
functions introduced in this section.
2) Write a program to find the greatest integer in X without using the
INT function. Be certain it works for negative numbers.
3) Write a program to roll two dice.
4) Write an efficient program to deal four 13 card hands from a standard
52 card bridge deck.
5) Write a program to store 25 random numbers in a list, print them all,
and print the largest and the smallest along with their respective
positions in the list.
6) Write a program to print all factors of an integer entered on INPUT.
7) Write a program to print prime integers in the range 2001 to 2501.
8) Find the greatest common factor for pairs of integers.
9) Find the least common multiple for pairs of integers.
10) Generate 100 integers at random from one to 10. Use a list to tabu-
late their frequency of occurrence.
11) Write a program to print a table of trigonometric values in a 10 degree
range by intervals of 20 minutes. Pick any one function.
12) Modify DRAW02 to draw all 10 slips of paper.
13) Write a program to produce the results of C</JMPAR using
SGN(A B) and the computed G</JT</J statement.

1-4 User Defined Functions, Subroutines, and Arrays

User Defined Functions
Another kind of computer function is available in BASIC. It is called the
"user defined function" and has the following form:
100 DEF FNA(X) = [formula]
If the A is positioned in FNA(X), you may use instead any letter of the alphabet
as the identifier, thus designating up to 26 functions in any one program. For
example, we might want to round off results to the nearest hundredth in several
places in a program. We would then use
100 DEF FNH(X) = INT( X*l00+.5)/100
and then use FNH( ), placing whatever variable we want rounded off in the
parentheses. Generally speaking, programmers place DEF statements near the
beginning of the program, and some systems require that this be done. Some
systems allow only a single argument; some allow two or more; some allow none.
We may use any variable or legal BASIC expression, including a defined func-
tion, as an argument.
Suppose we define a polynomial function, select a few values of X
between -10 and 10 at random and round the results to the nearest tenth. See
program PRT.
Defined functions are useful whenever we wish the computer to return a
single value.
 Review of BASIC 13

PRT

94 REM * T41S PROGRAM D~MONSTRATES T4E SINGLE

95 REM LINE DEFINED FIJNCTION IN BASIC
99 RANDOMIZE
-·100 DEF FNPCl<l = 3.41*K•3 + 4.32•K•2 - 11·2•K + 16.73
-110 DEF FNRCKl = INT< K•RND+l l
-120 DEF FNTCKl = INT< '<*10+.5 l/10
130 PRINT "X", "FNP<Xl", "f"NTCFNP<Xl )"
140 f"OR C = I TO 5
150 LET X = f"NRC21l-11
160 PRINT x, FNPCX>. f"NT<FNPCXll
170 NEXT C
172
lflO END
RUN
PRT

x f"NPCXl f"NT<f"NPCXl l
B 1949.53 1949.5
6 841 .61 841. 6
-5 -245.52 -245.S
10 3746.73 37116.7
-R - 136 3. 11 -1363.t

Subroutines
If we want the computer to return with two or more values, we cannot
use a defined function. Also, in the unlikely case that we want more than 26
functions, we need a new capability. In such cases we use a subroutine. A sub-
routine amounts to a detour in the program that returns to the statement im-
mediately following the one that caused the detour in the first place. 'I'his
capability permits a set of program statements to be accessed from more than
one point in a program. The statement pair G</>SUB-RETURN accomplishes
this.
One significant use of the subroutine is to improve the flow of control
through a program. A subroutine may be used to eliminate the proliferation of
G</>T<l> statements by replacing them with RETURN statements where appropri-
ate. This technique often simplifies the writing of programs. Programs written
in this way are also easier to read and thus easier to correct or modify later (see,
for example, programs GRAPH2 and GRAPH3 of Chap. 5).
Suppose that we want the product and the sum of two numbers modulo
M. We can write a subroutine that calculates and prints both values. Then we
can "call" the subroutine from anywhere in the program with G</>SUB n where
n is the first line of the subroutine. See lines 140 and 210 in program M</>D.

M0D

94 REM * THIS PROGRAM DEMONSTRATES GOSUB

95 REM WITH MODULAR ARITHMETIC
99 RANDOMIZE
100 D~F FNRCXl = INT< RND•X+l l
110 PRINT "f"IND A•B AND A+B MOD M"
120 PRINT "A, s,, M" J
130 INPUT A, 8, M
-140 GOSIJB 500
150 PRINT
160 PRINT "NOW DQ f"0UR .v1NDOM CALCULATIONS"
170 f"O R I = 1 TO 4
180 LET M = FNRC9l+l
14 Advanced BASIC

190 LET A= FNR<M-IJ

200 LET B = FNR<M-ll
-210 GOSLJA 500
220 NEXT I
222
224 REM * THE STOP STATEMENT PREVENTS
225 REM ILLEGAL SUBROUTINE l':NTRY
230 STOP
492
494 REM *
MODULAR ARITHMETIC SUBROUTINE
500 LET P = A•B
510 IF P < M THEN 540
520 LET P = P-M
530 GOTO 510
540 LET S = A+B
550 IF S < M THEN 580
560 LET S = S-M
570 GOTO 550
580 PRINT A; "*"; 8; "="; P; A; "+"; B; "="J S; "MOD"; M
590 RETURN
592
999 ENO
RUN
MOD

FIND A•B AND A+B MOD M

A,9,M? 2,5,6
2 *
5 = 4 2 + 5 = I MC D 6

NOW 00 FOUR RANDOM r.ALCIJLAT ICN S

3 2 2 3 + 2 I MOD 4
2 * I 2 2 + I 3 MCD 5
I * 3 3 I + 3 4 MOD Fl
2 * 4 2 + 4 6 MCD 9
* Fl

Subroutines may in turn call other subroutines as in both lines 510 and
550 of MQ>Dl. One thing to avoid in subroutines is inadvertent use of variables
that have been used elsewhere in the program. Subroutines called from other
subroutines are called "nested subroutines." (What else?) In our example
Ml,l>Dl we have nested them two deep. Systems vary, but many have a limit on
how deep subroutines may be nested.

M0DI

94 REM *THIS PROGRAM INTR0DUCES A NESTED

95 REM GOSUB INT0 PR0GRAM MOD
99 RAND0M I ZE
100 DEF' FNR<XJ = INT< RND•X+I J
110 PRINT "FINO A•B AND A+B MOD M"
120 PRINT "A,B,M"J
130 INPUT A, 8, M
140 GOSUB 500
150 PRINT
160 PRINT "NQl,1 DO FOUR RANDOM CALCULATIClNS"
170 F'ClR I = I TO 4
180 LET M = FNRC9J+I
190 LET A= FNRCM-ll
200 LET 8 = FNR<M-IJ
210 GOSIJB 500
220 NEXT I
222
224 REM *
THE STOP STATEMENT PREVENTS
225 REM ILLEGAL SUB ROUT !NE ENTRY
230 STOP
492
494 REM *
MODULAR ARITHMETIC SIJARCUT!NE
500 LET N = A•B
 Review of BASIC 15

-510 GO SUB 700

520 LET P = N
540 LET N = A•B
- 550 GOSUfl 700
560 LET S =
N
580 PRINT A; "*"J BJ "="J Pl A; "+"; BJ "="J SJ "MOD"J ~
590 RETURN
592
694 REM * REM REDUCES N TO A NIJMBER MOD M
700 IF N < M THEN 730
710 LET N = N-M
720 GOTO 700
730 RETIJRN
732
999 END
RUN
MOD I

FIND A*fl ANO A+B MOD M

A,8,M? 3,4,5
3 * 4 = 2 3 + 4 = 2 MOD 5

NOW DO FOUR RANDOM CALCl/LATICNS

2 * 3 6 2 + 3 5 MCD 8
4 I 4 4 + I 5 MCD 9
2 * 2 * 4 ?. + 2 4 MCD 6
6 * 8 8 6 + R 4 MOD I 0

Some systems have a computed G<,1>SUB statement similar in format to

the computed G<,1>T<,1>. The format is:
100 qlN K GqlSUB n 1 ,n2,n 3 , etc.
or

or

Arrays
An array is simply a two dimensional list. For this we use two subscripts
separated by a comma and enclosed in parentheses. The first subscript desig-
nates the row and the second subscript designates the column. As with lists,
we can use a subscript as high as 10 without the need for a DIM statement.
Suppose you recorded the temperature at 6AM, 12NOON and 6PM for
one week. This data could easily be stored in an array to enable a program to
carry out various calculations, for example, see program WETHR.

•,JET HR

94 RF.M * THIS PROGRAM FINDS AVF.HAGE TEMPERATURE

95 RE.M TO DEMONSTf~ATE A IJSE OF ARRAYS
100 DIM AC6,4l
102
104 REM * READ DATA
I I0 FDR D = I TO 5
120 FOR I = I TO 3
130 READ ACD,Jl
140 Nt.XT I
150 NF:XT D
152
16 Advanced BASIC

154 REM *
C:ALC 11LATF: AV<:RAGF: F:ACH DAY
160 FOR D = 1 TO 5
170 LET T = 0
180 FOR I = 1 TO 3
190 LET T = T+A<D, I)
200 NEXT I
210 LET A<D, 4) T/3
220 NEXT D
222
224 REM *
CALCULATE AVERAGF EACH READING
230 FOR I = 1 TO 3
240 LET T = O
250 FOR D = I TO 5
260 LETT= T+ACD,I>
270 NEXT D
?.80 LET AC6, I> = T/5
290 NEXT I
292
300 PRINT "DAY\TIME", "6AM", "12NOCN", "6PM", "DAILY AVG"
310 PRINT
320 FOR D = 1 TC 5
322
324 REM *
~EEP TFRMINAL ON THIS LINE
,125 REM FCR THE NEXT PRINTED RESULT
326 REM WITH TRAILING COMMA IN LINE 330
330 PRINT "DAY"; Q,
340 FOR I = 1 TO 3
350 PRINT A<D, I),
360 NEXT I
370 PRINT A<D,4l
380 NEXT D
382
390 l'Q R I = I TO 5 0
400 PRINT "-";
410 NEXT I
420 PRINT
422
430 PRINT "AVE'<AGF:S",
440 FOR I = 1 TO 3
450 PRINT AC6, I),
460 NEXT I
470 PRINT
472
474 RF:M
4RO DATA ?.7,J6,J4 .. 40,50,55, so,s?.,4~

490 DATA 43,L!l,37, JO,J3,~R

500 END
RIJN
1 •.!F:THR

DAY\ TIME 6AM 12NOON 6PM DAILY AVG

DAY 1 27 36 34 32. 3333

DAY 2 40 50 55 48.3333
DAY 3 so S2 48 so
DAY 4 43 41 37 ,,o. 3333
DAY 5 30 33 28 30.3333

AVERAGES 38

Even though the computer permits a subscript as high as 10 for rows and
for columns, it does not require that we use them all. In WETHR we used only
six rows and four columns.
Note that we used loops within loops, or nested loops, several times in
WETHR. The requirement here is that the loops must be nested entirely within
 Review of BASIC 17

other loops. If a program contains Ffl>R X = A T</J B followed by F</JR Y

= C T'l> D, the NEXT Y statement must appear before the NEXT X statement.
Once the data is stored, it is a simple matter to obtain other information.
For example, we can find the highest temperature reading for the week as in
WETHRl.

WETHRl
94 REM * THIS PR0GRAM FINDS HIGHEST TEMPERATURE
95 REM IN A FIVE DAY PERI0D USING ARRAYS
100 DIM AC6,4l
102
104 REM * READ DATA
110 FOR D = 1 TO 5
120 FOR I = 1 TO 3
130 READ ACO, I>
140 NEXT I
150 NEXT D
152
154 REM * THE FIRST ENTRY IS THE HIG4EST SO FAR
160 LETH= AC1,l>
170 LET R = C = 1
180 FOR D = 1 TO 5
190 r0R I = 1 T0 3
200 IF ACD•I> <= H THEN 240
202
204 REM * IF THE CU~RENT ENTRY IS HIGHER THEN
205 REM SAVE DATA IN H, R, AND C
210 LET H ACD,J)
220 LET R D
230 LET C = I
240 NEXT I
250 NEXT D
252
260 PRINT "HIGHEST TEMPERATURE ="I H
262
264 REM * USE Or 24-HOIJR CLOCK SIMPLIFIES
265 REM PRINTING THE TIME
270 PRINT "OCCURS AT"J C*600; "HOURS ON DAY"; R
474 REM
480 DATA 27.36.34. 40,50,55, so,s2,4g
490 DATA 43,41,37, 30,33,28
500 END
RUN
WETHRl

HIGHEST TEMPERATURE = 55
OCCURS AT 1800 HOURS ON DAY 2

Summary of Sec. 1-4

BASIC allows tremendous flexibility in the user defined functions. We
may have up to 26 functions defined with DEF FNA(X) followed by an equals
sign and any legal BASIC formula.
For calculations that don't lend themselves to function definition, such
as calculations requiring more than one value in the result, the subroutine
entered with G</JSUB is available. The end of a subroutine is indicated by the
RETURN statement.
Two dimensional variables are available in BASIC. A(I,J) designates the
Ith row and the Jth column of A. For subscripts greater than 10, a DIM state-
ment is required.
18 Advanced BASIC

Problems for Sec. 1-4

1) Write a program to read six test scores for each of five students into an
array, one student to a row. Find the test average by student and by
test, and print the scores and the results in easily readable form.
2) Write a program to locate the maximum and the minimum point of
any function in a domain specified as input. Also specify the incre-
ments as input.
3) Write a program to set up a tic-tac-toe board and keep track of the
play of two external players.
4) Write a program to fill a square array with zeros except along the
diagonal from the upper left to the lower right corners, which should
contain ones. Print this array (called the "identity array").
5) Write a program to produce the results of program C<l>MPAR in Sec.
1-1 using a simple Gq>SUB and only one G<l>T<l>.
6) Write a program to produce the results of program C<l>MP AR in Sec.
1-1 using SGN(A - B) and the computed Gq>SUB.
7) If your system allows two arguments in user defined functions, write
a single function to allow rounding off numbers to any desired pre-
cision. Try the same thing with a single argument function. Write
a program to verify your functions.

1~5 MAT Statements in BASIC

Arrays and lists are used so routinely in programming and in mathematics
applications that most implementations of BASIC include a group of special
statements to handle them. Note that since BASIC treats lists as special arrays,
we may not use the same letter for a one-dimensional list that we use for a two-
dimensional array. Instead of reading values into the elements of an array entry
by entry with loops, we may simply use MAT READ A or MAT READ A,
B,C as long as the array or the arrays are dimensioned to be compatible with the
data. BASIC also allows us to alter the dimensions of arrays in the MAT READ
statement. MAT READ A,B(3,7),C(N,M) reads mat A according to previously
set dimensions; dimensions B to three rows and seven columns; reads the array
and dimensions C to N rows and M columns, N and M having been previously
defined; and reads that array.
MAT PRINT A prints the complete array with comma format. To get
semicolon format, MAT PRINT A; must be used. A single statement may
specify printing of more than one array, as in MAT PRINT A,B;C, which will
result in A being printed with comma spacing, followed by B printed with
semicolon spacing, followed by C printed with comma spacing (see program
MATOl).
MAT INPUT allows us to type entries from the keyboard of our ter-
minal with all of the dimensioning options of MAT READ. MAT INPUT
A has one additional option in some implementations of BASIC. If A is a
list, you may not know the number of entries required when the program is
written. In some systems, the function NUM takes on the number of elements
entered in the most recent MAT INPUT statement in the program.
 Review of BASIC 19

MATO!

94 REM *
THIS PROGRAM DEMONSTRATES
95 REM MAT READ AND MAT PRINT
100 DIM A(3,4i, 8(6, IOJ, C<8,9l
110 READ R,c
-120 MAT READ A. 8(2,5), cm.Cl
-130 MAT PRINT A, BJ C
132 REM
140 DATA 4,3
150 DATA 1.,2,3,4,5,6,7,g,9,10,11,12
160 DATA 13, 14, IS, 16, 17, IR, 19,20,21,22
170 DATA 23,24,25,26,27,28,29,30,31,32,33,34
180 END
RUN
MATO!

2 3 4

5 6

9 10 11 12

13 14 15 16 17

18 19 20 21 ~~

23 24 25

26 27 2!l

29 30 31

32 33 34

Other statements that may be used to assign values to the elements of

arrays are as follows:

MATA ZER Fills array A with zeros according to

previously specified dimensions.
MATA ZER(2,3) Redimensions A and fills it with zeros.
MATA ZER (B,C) Redimensions A and fills it with zeros.
MATA ZER(5) Redimensions A and fills it with zeros.
MATA ZER(X) Redimensions A and fills it with zeros.
MAT A = C</>N fills the array A with ones. As shown, the dimensions must
have been previously determined. C</>N has all of the redimensioning options
shown for ZER.
MAT A = IDN is used only for a square array (where the number of rows
is equal to the number of columns). A is filled with ones where the row num-
ber and the column number are equal and all other locations are set equal to
zero. IDN may be redimensioned with IDN(X,X) or IDN(5,5). This is called
the "identity matrix" in matrix algebra.
In matrix algebra, addition and subtraction are defined for like-dimen-
sioned arrays as the sum or difference, respectively, of elements having the same
location. That is, if array X is the sum or difference of Y and Z, then for all
I,J we have X(I,J) = Y(I,J) ± Z(I,J). This could be done with nested loops;
in BASIC, however, we merely type:
20 Advanced BASIC

100 MAT X y + z for addition

100 MAT X y- z for subtraction
Matrix multiplication has a more complicated definition. For arrays Y
and Z, the number of columns in Y must equal the number of rows in Z. The
(R,C)th entry of X is the sum of the products Y(R,T)*Z(T,C), where T goes
from one to the number of columns in Y, which is also the number of rows in Z.
The product matrix will have the same number of columns as X and the same
number of rows as Y. A BASIC program for matrix multiplication requires
triple nested loops.
The following statement accomplishes the same result:
100 MAT X = Y*Z
Note that if multiplication works for Y*Z, it will work for Z*Y only if the
arrays are square. In any case, the product has the number of rows of the left
multiplier and the number of columns of the right multiplier (see program
MAT02).
In matrix algebra, instead of dividing Y by Z, we first find the inverse of
Z. The inverse of Z, written z- 1 , is the matrix such that Z*z- 1 equals the

MAT02

100 DIM xc10.1oi. YClO,!Q), zc10.10>. P<10,10>

102
104 REM *
READ MATRIX DIMENSIONS
110 READ I,J, K,L
120 IF J =
K THEN 150
130 PRINT "PRODUCT UNDEFINED"
140 SHJP
142
144 REM * READ MATRIX ELEMENTS
150 MAT READ YCioJl, Z<K,Ll
152
154 REM *
INITIALIZE PRODUCT MATRIX
160 MAT X =
ZER<I,L>
162
164 REM *
MULTIPLY USING NESTED LOOPS
170 FOR T =
1 TO J
180 FOR R =
l TO I
190 FOR C = l TO L
200 LET X<R,C> =
X<R,Cl+YCR,Tl*Z<T,C>
210 NEXT C
220 NEXT R
230 NEXT T
232
240 PRINT "USING TRIPLE NESTED LOOPS"
250 MAT PRINT Xl
260 PRINT
262
264 REM *
DEMONSTRATE MAT PRODUCT IN BASIC
270 MAT P =
ZERCl,L>
280 MAT P Y*Z =
290 PRINT "USING MAT MULTIPLY STATEMENT"
300 MAT PRINT Pl
302
304 REM
310 DATA 2,3, 3,4
320 DATA 1,2,3,405,6
330 DATA 1,2,3,4o5o6o7o8o9o 10, l lo 12
340 END
 Review of BASIC 21

RUN
MAT02

USING TRIPLE NESTED LOO PS

38 44 50 56

83 98 113 128

USING MAT MIJLTJPLY STATEMENT

38 44 50 56

83 98 113 128

identity matrix. Having found the inverse of Z, we multiply Y by it. The in-
verse (if it exists) is obtained with the following statement:
100 MAT X = INV(Z)
BASIC allows us to find the transpose of a matrix with the following
statement:
100 MAT X = TRN(Z)
The transpose consists of an array X such that the columns of X are the rows of
Z and the rows of X are the columns of Z. Note that if the dimensions of Z are
(R,C), then the dimensions of X must be (C,R).
We can duplicate an array with
.100 MAT X = Z
and can multiply every entry of an array by the same number with
100 MAT X = (C)*Z
where C is any legal BASIC formula.
In the last seven cases just presented, some systems use the statement
itself to dimension or redimension X, whereas others require that the program
make the dimensions of X compatible in advance.

Summary
BASIC provides a number of MAT statements that greatly simplify pro-
grams which deal with arrays. They are as follows:
MAT READ Reads data into a variable list.
MAT INPUT Enters data from the keyboard into a variable list.
MATX = ZER Fills X with zeros.
MATX = Cq'.>N Fills X with ones.
MATX = IDN Creates the identity matrix.
MAT PRINT Prints the contents of a variable list to the terminal.
MATX = Y Copies the contents of Y into X.
MATX = Y + Z Enters the sum of Y and Z into X.
MAT X = Y - Z Enters the difference of Y and Z into X.
MAT X = Y*Z Enters the product of Y and Z into X.
MAT X = (C)*Y Multiplies each entry of Y by C and enters result in X.
22 Advanced BASIC

MAT X = INV(Y) Enters the inverse of Y into X.

MAT X TRN(Y) Enters the transpose of Y into X.
The first five statements above have optional redimensioning capabilities.
Problems For 1-5
1) Write a program to add or subtract two arrays without using matrix
statements from this section.
2) Write a program to enter 25 random integers into a five by five array
and print it. Now find the largest number in each row and its column
number, and find the largest number in each column and its row
number.
3) Write a program to print a times table modulo six.
4) Write a program to print a times table modulo five.
5) MAT READ the integers one through nine into a three by three array.
Copy this into the upper left corner of a four by four array. Use the
fourth row to enter the column sums, and use the fourth column to
enter the row sums. Print the resulting array.
6) Write a program to create the transpose of a given matrix without using
the TRN function.
 2
SOME EXTENDED
FEATURES OF BASIC

2-1 Introduction
None of the features to be discussed in this chapter is absolutely necessary
to writing programs in BASIC. Other language statements may be assembled to
achieve the same effect of each one. Some of them may not even be available on
the computer you are using. However, if your system does provide any of the
features presented here, you are encouraged to experiment since they save
programming effort and storage space and make programs more readable.
(Note about problems: The author has not attempted to contrive problems
which would require the reader to use the various features presented in this
chapter. Rather it is recommended that you be alert to possibilities for using
the material of this chapter in programs you will be writing throughout the rest
of the book.)
2-2 TAB
The T AB(X) function is available on many BASIC systems. T AB(X)
placed in a PRINT statement causes the printing mechanism of the terminal to
be located in the xth space of the current line, provided it has not already passed
that point. Note that on most systems the leftmost space is numbered zero and
the counting is modular, with the mod being typically 72 or 75, but sometimes
more. Some systems are not modular but start on a new line when the TAB
argument exceeds the system line length. An explicit number or BASIC formula
may be entered as the argument of the TAB function. If the value of the argu-
ment is not an integer, most systems use only the integer part. See program
TABOl.
23
24 Advanced BASIC

TABOl

94 REM *
PRINT SCALE TC AID COUNTING SPACES
100 FOR I = 1 TO 5
110 FOR J = 1 TO 9
120 PRINT "-";
130 NEXT J
140 PRINT "!";
150 NEXT I
160 PRINT
162
164 HEM *
PRINT IN LOCATIONS READ FROM DATA
170 FCR K = 1 TO 5
180 READ T
190 PRINT TAflCTlJ "X";
200 NEXT K
210 PRINT
212
214 REM *
SHOW TAfl MORE THAN ONCE IN
215 REM THE SAME PRINT STATEMENT
220 READ A, 9, C
-·230 PRINT TABCAJ; "A"; TAB<Bll "fl"J TA8CCll "C"
234 REM
240 DATA 3, s, 12, 28, 48
250 DATA 2, 7, 15
260 F.:ND
RUN
TABOl

---------!---------!---------!---------!---------!
x x x x x
A B c

Examining the output of program TABOl and counting the leftmost space
as zero, we see that the X's are printed in the spaces numbered 3, 8, 12, 28, and
48, as designated in the DATA ofline 240.
The uses of the TAB function are not limited to literal output. We may
also format numeric output or a mixture of numeric and literal output. See
program TAB02.

TAB02

94 REM *
TAfl DEMONST~ATION PROGRAM
100 DIM AC4l
102
104 REM * HEAD FOUR ITEM A LIST
1 10 F 0 R I = 1 TC 4
120 READ AC!l
130 NEXT I
132
140 PRINT TABC15lJ "CO'IMA SPACING"
150 FOR I = 1 Tr. 4
160 PRINT A<IJ,
170 NEXT I
l'lO PRINT
182
190 PRINT TAAClOll "IJS!NG TAB FOR 10 CHARACTER PRINT ZCNE
200 FDR I = 1 TO 4
210 PRINT TAR< 10•(!-l l l l AC!);
220 NF.XT I
23D PR INT
232
24D PRINT TA'lC6lJ "SF:'11COLCN SPACI'JG"
25 0 FDR I = 1 Tn ,,
260 PR!'JT A<Ill
?.70 ~l".XT I
280 PRINT
 Extended Features of BASIC 25

2fl2
~90 Rf'M
300 DATA -3, 11,0,406
310 END
RllN

COM"IA SPACING
-3 I I CJ 40~
llSING TAR FCR 10 r.HARAC:FR PRICIT ZCNES
-3 11 0 406
SEMICOLON SPACING
- 3 1I 0 406

Note that systems differ in the number of spaces generated by the comma
and semicolon delimiters.
2-3 Print Using
While it is possible with the string facilities available in BASIC (see Chap-
ter 3) to print output in any desired form, formatting can require tedious pro-
gramming. It is for this reason that many systems provide PRINT USING and
image statements.
PRINT USING permits the format for printing output for the current line
to be specified in yet another line. The specifications for printing are called the
image. The image can be defined in an image statement, which begins with a
colon. Very simply, a pound sign is used to specify where we want digits printed.
For pound signs placed to the right of the decimal point, zeros to the right of
the last nonzero digit are printed. For pound signs placed to the left of the
decimal point, zeros to the left of the leading nonzero digit are not printed.
One of the common uses of PRINT USING is to obtain zeros to the right
of the decimal point in printed output, since as a numeric the trailing zero is
dropped in BASIC. If we want $3.10 printed, we may use an image to force the
printing of the zero, as shown in our first demonstration program, IMAGEOl.

!MAGED I

94 REM * DF"IONSTRATFS ELEMENTARY

95 REM EXAMPLF FC~ PRINT USING
100 LE'T D = 3.1
II 0 PR INT "TH IS IS '1l ITHO'JT I MAl-E"
120 PRINT "THE. AMO' INT IS S"; D
130 PRINT
140 PR!NT "THIS IS '.;ITH IMAFF:"
150 :THE AMCllNT IS S##.##
160 PRINT 'lo!NG ISO, D
170 END
R•JN
IMAG~O I

THIS IS WITHO'IT !MAG~

THF AMOUNT IS S 3.1

THIS IS WITH IMAGE

THE AMOUNT IS S 3.IQ

Next we present program IMAGE02 to show a variety of numbers printed in a

variety of images.
26 Advanced BASIC

IMAGF:02

9LJ REM * DEMONSTRATES A VARIETY CF IMAGF:S

JOO : 5$$##.### ###.## ####. #### ##
102
110 FOR I = 1 TO 5
120 READ X
130 PRINT "THE NUMRF:R IS"; X
lLJO PRINT USING 100, X, X, x, x, x, X
150 PRINT
160 NEXT I
162
16LJ REM
170 DATA J, 12, 12.3, 123LJ.56, .01
180 END
RUN
!MAGE02

THE NUMBER rs I
$ I. 000 1.00 1.

THE NUMBER rs 12
$12.000 12.00 12. 12 12 •12
THE NUMBER rs 12.J
$12.300 12.30 12. 12 12 •12

THE NUMBER rs I 23LJ. 56

$123LJ.560 •123LJ.56 1235. I 23LJ •123LJ •1234
THE NUMBER rs 0.01
$ .010 .01 o. 0 0 0

Note that where several dollar signs appear only one of them gets printed.
The one that is printed is as far right as possible in a space that has a dollar sign
in the image. When there is not room to the right of the decimal point in the
image to print all of the digits, the excess is dropped. No rounding off is done.
Where there is not room to the left, all digits get printed anyway, and an asterisk
is printed to call attention to the fact that the number did not fit the image.
The image can be a string. In this case, the line number which calls the
image in the PRINT USING statement is replaced by the string variable that con-
tains the image, as in program IMAGE03.

!MAGE03

9LJ REM DEMONSTRATES IMAGf: STORED

95 REM * IN A STRING VARIA8LE
100 LET IS = ""###. ## S$S#. ##"
110 LET N = 23.4
120 PRINT NJ .. .
IS THF: NIJMRE:R"
130 PRINT I SJ IS THE IMAGE"
140 PRINT USING rs, N, N
150 END
RUN
IMAGE03

23.4 rs THE NUMBER

#UH.## $5$#.## JS THF: IMAGE
23.40 $23.40

We can force the computer to print numeric output using E-format. Four
up arrows are used for this, as shown in program IMAGE04.
 Extended Features of BASIC 27

IMAGE04

94 REM *
PRINTS E-FORMAT
100 : ##,##tttt
110 LET N =
24.3
120 PRINT USING 100, N
130 PRINT NJ "IS THE NUMBER"
140 END
RUN
IMAGE04

2.43E+OI
24·3 IS THE NUMBER

Strings may be formatted with images also. An apostrophe must be used

to begin the printing for a string image. Then control letters E, L, R, and C may
be used to indicate fields of specific characteristics. Control character E calls for
left justification and allows the string to overflow, whereas L calls for left justifi·
cation but truncates on the right. Control character R calls for right justification
and truncates any excess, and C calls for centering the string output and truncates
on the right if there is an excess. The apostrophe is needed to delimit the con-
trol characters because any other alphabetic characters will be printed just as
they appear in the image. See program IMAGE05 for a demonstration of for.
matting string output in the image.

IMAGE05

94 REM * DEMONSTRATES PRINTING STRINGS

95 REM USING AN IMAGE STRING
100 LET 1$ = " 'LLLL 'RRRRR 'CCCCC 'EE"
110 PRINT IS
120 FOR I = I TO 5
130 READ SS
140 PRINT USING 1$, S$, S$, S$, SS
150 NEXT I
152
154 REM
160 DATA THIS, IS, A, SAMPLE, PROGRAM
170 END
RUN
IMAGE05

'LLLL 'RRRRR •ccccc 'EE

THIS THIS THIS THIS
IS IS IS IS
A A A A
SAM PL SAMPLE SAMPLE SAMPLE
PRO GR PROGRA PROGRA PROGRAM

2-4 Logical Operations, MAX and MIN

Truth Values
Some BASIC implementations include a set of logical operations. Let's
look at the already familiar IF-THEN statement. The statement IF X=l THEN
200 transfers control to line 200 only if X=l is true. If X=l is true, BASIC
assigns the statement X=l a value of 1 to designate 'true.' If X=l is false, then
BASIC assigns it a value of 0 to designate 'false.' So we could replace our pro-
28 Advanced BASIC

gram statement with IF X THEN 200, and when X=l the computer goes to 200;
otherwise control passes to the next line in sequence, provided X can be only 0 or 1.
We may even print the logical value of statements. If we print the value of
N/X=INT(N/X), we should get a one when Xis a factor of Nanda zero when X
is not a factor of N (see line 150 of program L<!>GICl).

UJGICI

94 REM * THIS PROGRAM DEMONSTRATES THE TRUTH VALUE

95 REM 0F THE STATEMENT N/X = INTCN/Xl
100 PRINT "INPUT N"l
110 INPUT N
114 REM* PRINT HEADINGS
120 PRINT "X"."'N/X","INTCN/Xl","TRUTH VALUE ClF"
130 PRINT TABC45ll"N/X = INT<N/Xl"
140 FDR X=I TO N
--150 PRINT X1N/X1INTCN/X),N/X=INTCN/Xl
160 NEXT X
170 END
RUN
LOGIC!

INPUT N?6
x N/X INT<N/Xl TRUTH VALUE OF
N/X INT<N/Xl
I 6 6 I
2 3 3 I
3 2 2 I
4 1. 5 0
5 1.2 0
6

The fourth column prints the logical value of N/X=INT(N/X). As we ex-

pected, that value is one whenever X is a factor of N and zero whenever X is not a
factor of N. This gives us a novel approach for counting factors of N. See line
140 of program L<!>GIC2. Where available, all of the relational operators may be
used in a similar manner.

LOG I C2

94 REM * THIS PROGRAM COUNTS FACTORS CF INTEGERS

95 REM USING THE TRUTH VALUE OF N/X = INTCN/Xl
100 PRINT "INPUT AN INTEGER"!
110 INPUT N
120 LET T=O
130 FCRX=ITON
-140 LET T=T+CN/X=INTCN/Xll
150 NEXT X
160 PRINT Tl"FACTORS"
170 END
RIJN
LOGIC2

INPIJT AN INTEGER? 240

20 FACTO RS

RUN
LOGIC2

INPIJT AN INTEGER? 1949

2 FACTORS
 Extended Features of BASIC 29

Logical Operators
The logical operators AND, fl>R, and Nfl>T are available on some BASIC
systems and may be used in a wide variety of applications. One rather straight-
forward application is an extension of IF-THEN statements. We may often com-
bine several IF-THEN's into one statement. For example, the pair,
100 IF X= 3 THEN 200
110 IF Y<2 THEN 200

becomes
100 IF X=3 OR Y<2 THEN 200

Or the four statements,

100 IF X <> 3 THEN 130
110 !FY>=2TH ENl30
120 G0T0 510
130 PRINT "MESSAGE"

become
100 IF X=3 AND Y<2 THEN 510
110 PRINT "MESSAGE"

We can negate a statement with Nfl>T. That is, X<>3 and NIPT(X=3) are
equivalent. (When you are unsure of the order of operations, it may be best to
use parentheses if only to make the statement easier to read.) What about X and
Nfl>T X'? In this case, Xis either equal to zero or it isn't. So in the case IF Nfl>T
X THEN 200, control passes to line 200 for X=O (because Nfl>T 0 equals 1) and
passes to the next statement in sequence for all other values.

MAX and MIN

For a system which provides MAX and MIN functions, the value of A
MAX B becomes the larger of the two numbers. The value of A MIN B becomes
the smaller of the two numbers. If you don't have these functions, then you
may use
.5*(A+B-ABS(A-B)) for A MIN Band .5*(A+B+ABS(A-B)) for A MAX B
2-5 Multiple Line Defined Function
Some versions of BASIC provide multiple line user-defined functions that
permit defining of functions that require two or more program statements to
define. The first line must be DEF FN followed by the function identifying
letter and the function argument or arguments in parentheses. The last state-
ment must be FNEND, and in between there must be at least one assignment
statement with FN and the same letter specified in the DEF statement on the
left of the equals sign.
For example, we may write a multiple line function that will do modular
multiplication. The function of lines 100 through 160 in program FUNC'I'Ifl>N
does just that.
As with other user defined functions, the multiple line function may be
30 Advanced BASIC

save program
accessed at any point of the program. In this way we may often
to avoid repeatin g a set of program state-
space, just as G(j>SUB is often used
defined function is accesse d directly whenev er its name is
ments. However, the
control to it.
used and thus it does not require a special stateme nt to transfer
value is re-
The defined functio n may be used in all situatio ns where a single
of parame ters, whereas a subrout ine is more
quired for each parame ter or set
appropr iate if two or more variable s must be returne d.

FIJNCT ION
OF THE
94 REM * THIS PROG'lAM IJFMONST RATES A USE
95 REM MIJLTIPL E LINE 'JSER DEFINED FIJNCTIQ N
96 160
IN L.INES 100 THiW'JGH
97 REM * THE F'JNCTIO N IS DEF INF.D
DEF F NM CX, Y, 7l
=
-{m
LET P X*Y
IF P < Z TH~N 150
LET P = P-Z
140 GOTC 120
150 LET FNM = P
160 FNEND
192
200 PRINT "MULTIP LY A TINIES 8 MOD M TC C:ET"
210 READ A, 8, M
220 IF M =
0 THEN 260
TA8C27l ; FNM(A,
230 PRINT TABC8l; Al TASC16 ll 8; TA8C22) ; M;
240 GOTO 210
242
244 REM
250 DATA 1,2,3, 3,2,4, 3,3,4, Q,O,Q
260 END
R•JN
FIJNCT ION

MULTIPL Y A TIMES B MCD M TO GET

1 2 3 2
3 '-? It ?.
3 3 4
 3
S INGS

3-1 Introduction
A string is simply any set of characters that is not to be treated as a number
but is to be otherwise manipulated by the computer. The characters are referred
to as alphameric or alphanumeric, since both alphabetic and numeric symbols
may be used, as are most other symbols permitted by your terminal, even includ-
ing the nonprinting characters. It is the purpose of this chapter to outline a few
of the possible uses of strings and string variables in BASIC even though the
specifications are not uniform from system to system. The user should deter-
mine the capabilities of his or her own system before planning extensive pro-
gramming activity.
3-2 The String Simple Variable
To distinguish the string variable from the numeric variable, virtually all
systems use a trailing dollar sign ($) in the variable name for a string. Legal sim-
ple string variable names are A$, K$, B3$, and C8$. Some systems are restricted
to the 26 letters of the alphabet, not permitting such variables as B3$. Others
allow additional variables such as &$. The number of characters you are allowed
to store in one simple string varies typically from 18 to thousands. Most of the
manipulations allowed for numeric variables are allowed for string variables ex-
cept for arithmetic operations. String variables can be READ, INPUT, assigned,
printed, and compared for order. Order comparison is accomplished according
to ASCII (American Standard Code for Information Interchange) specifications,
which place the digits in order 0 through 9 ahead of the letters of the alphabet
in alphabetical order. (See Appendix A for relevant parts of the code.)
One of the uses of strings is to permit person-computer "conversation."
We do not need to numerically code our answers to questions put to the pro-
gram user. If we wish to give the program operator options, the answers can be
words which the computer processes directly. If for example, we write a pro-
gram that will require a lot of yes-no answers from the keyboard, we may write
31
32 Advanced BASIC

a subroutine that prints the question, checks the answer to be sure that it is
either a "YES" or a "N't)," and rerequests the answer for all other responses.
Program DECIDE uses such a subroutine repeatedly as the computer as-
sembles the information fed into it by the user. The subroutine outlined above
begins in line 800.

DECIDE

94 REM * DEMONSTRATICM PROGRAM INTRODIJCING srnINGS

95 REM SHO',JS ASSIGNMENT, INPUT, COMPARISIO"J AND
96 REM PRINTING CF STRINGS
I 00 GO SIJFl 200
110 LET X$ =OS
120 GO SUB 400
130 LET YS = OS
140 PRINT "PROGRAM NAME"l
150 INPUT N$
160
-170
PRINT
PRINT XS; YSJ.. - tt; NS
180 STOP
192
194 REM * LANGUAGE DECISION SUBRCUTINE
200 LET OS = "FORTRAN"
210 GOSIJB BOO
22 0 IF AS = "YES" THEN 3 I 0
230 LET 0$ = "COBOL"
240 GO SUB 800
250 IF AS = ••yES" THEN 310
260 LET OS = "BASIC"
270 GO SUB 800
280 IF AS = "YES" THEN 310
?.90 PRINT "FORTRAN, COBOL OR BASIC ONLY"
300 GOTO 200
310 RETURN
392
394 REM *MAKE 'OLD - NEW' DECISION HERE
400 LET 0$ = " OLD"
410 GOSIJB 800
420 =
IF' AS "YES" THEN 480
430 LET OS = " NEW"
440 GO SUR 800
450 IF As = "YES" THEN 480
460 PRINT "OLD OR NEW ONLY"
470 GO T!l 400
480 RETURN
7B2
784 REM * YES-NO DECISION SUBROUTINE
785 REM ENTER WITH GOSUB 800
790 PRINT "YES OR NO"
-soo PRINT OS;
BIO INPUT A$
820 IF AS "YES" THEN 850
830 IF A$ "NO" THEN 850
840 GO TCl 790
850 RETURN
999 END
RUN
DEC !DE

FORTRAN?NO
COBOL? NO
BASIC? YES
OLD? YES
PROGRAM NAME?DEC!DE

BASIC OLD - DECIDE

 Strings 33

Strings are read from data in exactly the same manner that numerics are
read from data. String and numeric data may be intermixed. Some systems re·
quire that string data be enclosed in quotes. On systems that do not require
quotes, any expression that contains a comma will have to be enclosed in quotes
anyway, since the comma will otherwise be interpreted as the end of the string.
Look at line 170 of program DECIDE. Note that the string variables are
separated by semicolon delimiters in that PRINT statement. You should see
that the printed output has no spaces other than those actually placed in string
Y$ and the literal expression" - "by the program itself. When printing numeric
values, the computer always provides some space automatically when using semi-
colons as delimiters. When printing strings, the computer does not provide any
space automatically. We can print string characters right next to each other by
using semicolons. We can print strings in 15-character columns using commas, or
we can use the TAB function for other spacing. Some systems also provide
PRINT USING for yet another formatting capability.
We can easily construct a program to do for two strings exactly what pro-
gram C</.>MPAR in Sec. 1-2 did for two numerics, as shown in program C</.>MPR$.
All we want is to determine if A$ is less than, greater than, or equal to B$. Look
carefully to see what happens to the leading and trailing blanks in the strings. In
this case, the leading blank makes" ALPHA" less than "ALPHA", but the trail-
ing blank in "BETA " has no effect. Some systems will not ignore trailing
blanks. Also note that the string 0123456789 is placed in quotes. When calling
for a string read, some computers will ignore such a string unless it is in quotes
because it 'looks like' a numeric.

COM PR$

94 REM THIS PROGRAM COMPARES mo

95 REM
*STRINGS F'OR ORDER
100 READ A$, B$
110 IF' A$ = "STOP" THEN 260
120 IF A$ = B$ THF:.N 160
130
140 PRINT ASJ ..
IF' A$ < B$ THEN 180
IS GREATER THAN "J BS
ISO GOTO 100
160 PRINT A$J ..
IS l"QIJAL TO B$
170 GOTO JOO
!RO PRINT AS; ..
IS LESS THAN "; 8$
190 t;crc 100
192
194 REM
200 DATA NUMERIC, ALPHABE'TIC
210 DATA
-220 DATA ..
"0123456789", ABCDFF' GH I JK
..
ALPHA", "ALPHA"
-230 DATA "Bl!TA , "'BETA"
240 DATA ENO, END
250 DATA STOP, STOP
260 END
RUN
CCMPRS

NUMl!RIC IS GREATER THAN ALPHABETIC

0123456789 IS LESS THAN ABCDEFGHIJK
- ALPHA IS Ll!SS THAN ALPHA
-BETA IS EQUAL TO 8ETA
END IS EOIJAL TO END
 34 Advanced BASIC

Summary of Sec. 3-2

We have seen that the simple string variable can be used in BASIC programs
in many of the same ways that simple numeric variables can. We may use simple
string variables in READ, DATA, INPUT, PRINT, and IF-THEN statements. In
DATA and INPUT statements, some systems require quotes to delimit strings
under certain conditions.
Problems for Sec. 3-2
1) Write a program to find the highest and lowest ranking strings in a set
of data.
2) Write a program that reads two strings and then prints them in ascend-
ing order.
3) Write a program that will read three strings from data and print them in
ascending order (more on ordering strings later).
4) How many print statements would be required to print n strings in
alphabetical order using only the methods of this section?
5) Write a program to request two peoples' names in two strings for each
person, first name first. Then have the names printed in alphabetical
order. Be sure to handle William Smith and George Smith properly.
3-3 The String Subscripted Variable
It is at this point that we must distinguish the ways in which various systems
treat subscripted string variables. There are two fundamentally distinct ways to
deal with A$(1,J) and B$(I). (Some systems don't even permit the double sub-
script.)
One concept, as shown in demonstration program ARRAY$, considers
A$(1,J) as an element of a string array just as A(I,J) is an element of a numeric

ARRAYS

94 REM * THIS PRCGNAM IS A DE~SNST~ATICN CF

95 REM A STRING ARRAY
96
97 REM *
HERE THF: D!MF:NSION SPF.CIFif.S THF: N'lt>IRF.R
98 REM OF STRINGS THAT MAY A~ ST~RED IN A LIST
100 DIM A$C3l
102
104 REM * READ THE STRING LIST FRCM D~TA
1 10 FDR I = 1 TO 3
120 READ ASCll
130 NEXT I
132
134 REM *NO~ PRINT THE STRING LIST
140 FC R I = 1 TC 3
150 PRINT "AS<"; I; ") = "; AH!l
160 NEXT I
172
17 4 REM
180 DATA F lilST, SECOND, THlilD
190 END
RUN
ARRAY$

AS C 1 l = F IR ST
A$( 2 l = SECOND
AS< 3 l THIRD
 Strings 35

array. This makes B$(I) the Ith element of a string list called B$ which may con-
tain up to 11 strings before the DIMension statement is required.
The other concept, as shown in program SUBSTR, considers A$(I,J) as a
substring having J- I+l characters going from the Ith to the Jth character inclu-
sive. So A$(I,I) is the single Ith character. Using this concept, B$(I) is the sub-
string beginning with the Ith character and continuing to the end of the string.
The next two sections will treat these two concepts separately.

SiJ8STR

94 RE~ * TYIS P~SGRAV IS A D~~O~ST~ATI~N

95 REM CF AS(J,Jl AS .4 S•l.'JSTRING

97 REM * IN T4IS CASE TH~ OI~ENSION SPECI~IES TYE

98 REM NUMBER Q" CHARACTERS IN THE STRING VARIA~lf A!
100 DIM AS( l4J
110 LET AS="THIS IS A TEST"
120 PRINT "AS = ";AS
130 FOR !=I TO 14 ST1'P 4
140 LET J=I+I~TCRNDC-1l*4+ll
150 PRINT "AS<"; I;",";J;") '";ASl J,JJ;"'"
160 NEXT I
16 4 REM
170 E'"D
RUN
S'JflSTR

ASC I
ASC 5
4
6
.
AS = T!1IS IS A TS:ST
) 'THIS'
I•
'
AS( 9
ASC 13
12
17. 'A T".,
'ST

3-4 The Substring Scheme*

One advantage in the substring concept is that we can deal with any part or
parts of the string directly. One disadvantage is that we cannot deal with
hundreds or thousands of strings without the use of files. Strings may be read
from data. When placed in data, strings must be enclosed in quotes. The same is
true if we type more than one string to the keyboard in response to an INPUT
statement.
Generally speaking, the DIMension statement is required for string variables
to permit the computer to allocate space. For example, 100 DIM A$(10),
B$(58),A(l5} provides for 10 characters in A$, 58 characters in B$, and 15
numerics in the A list, the latter showing that string and numeric dimensioning
may be intermixed.
A program to arrange the letters of a string in alphabetical order is shown
in program ALPHA. Line 100 provides for up to 72 characters in the string vari-
able A$. Line 120 uses the LEN() function. This function measures the actual
number of characters in the string. Note that lines 160, 170, and 180 exchange
two characters which are not in the proper order. Since B$ is used for only a
single character, it need not appear in the DIMension statement. Note that
spaces have a lower ASCII code than the letter A and so appear first in the result
of program ALPHA.

*The programs of Sec. 3-4 were run on a Hewlett Packard Computer.

36 Advanced BASIC

ALPHA

94 REM *
THIS PROGRAM ALPHABETIZES THE
9S REM CHARACTERS OF A STRING
-· 100 DIM ASl72J
110 INPUT AS
-120 LET L=LENCASl
122 REM
124 REM *LINE lSO TESTS ALL PCSSIFlLE PAIRS CF
12S REM CHARACTERS TO SFE IF THEY ARF IN ORDER
130 FOR I=! TO L-1
140 FOR J= I+ 1 TO L
150 IF AS(!,!J <= AS[J,JJ THEN 190
!S2 REM
!S4 REM* IF ASC!,Il IS NOT LESS THAN ASCJ,Jl THEN
!SS REM WE EXCHANGE THOSE CHARACTERS
-160 LET BS=AS[ I, I l
·-170 LET ASC I, ll=AS(J,JJ
-180 LET AS[J,Jl=FlS
190 NEXT J
200 NEXT l
204 REM
210 PRINT AS
220 END
RUN
ALPHA

?THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG

ABCDEEEFGHHIJKLMNOOOOPORRSTTUUVWXYZ

Program REVRS simply prints the characters of a string in reverse order

but keeps the string intact. Program END shows the printing of A$(I) for a
sample string.

REV RS

94 REM *
THIS PROGRAM DEMONSTRATES PRINTING THE
9S REM CHARACTERS OF A STRING JN REVERSE ORDER
96 REM WITHOUT MODIFYING THE CONTENTS OF
97 REM THE STRING
100 DIM ASC72J
110 INPUT AS
120 PRINT
130 PRINT " THE STRING CONTENTS JN REVERSE ORDER:"
134 REM
140 FOR I=LENCASl TO 1 STEP -1
ISO PRINT AS(I,Ill
160 NEXT I
164 REM
170 PRINT
180 PRINT
190 PRINT " THE CONTENTS PRINTED IN TACT:"
200 PRINT AS
210 END
RUN
REV RS

?THIS PROGRAM PRINTS IN REVERSE

THE STRING CONTENTS IN REVERSE ORDER:

ESREVER NI STNIRP MARGORP SIHT

THE CONTENTS PRINTED IN TACT:

THIS PROGRAM PRINTS JN REVERSE
 Strings 37

END

94 REM *
THIS PROGRAM DEMONSTRATES THE USE OF
95 REM PRINT A$CI> IN THE SUBSTRING SCHEME
100 DIM AS(72l
110 INPUT A$
120 FOR I=LENCA$> TO I STEP · I
130 PRINT "ASC"Jll"> "JA$(!J
140 NEXT I
142 REM
150 END
RUN
END

?THIS IS IT
A$( 10 ) T
A$( 9 ) IT
A$( 8 IT
AS< 7 s IT
A$< 6 IS IT
A$( 5 ) = IS IT
AS< 4 ) = s JS IT
AS< 3 JS IS IT
AS< 2 HIS IS IT
AS< I THIS IS IT

The ability to compare strings and substrings allows us to pack several

items of information into a single string. For example, we might want to work
with the days of the week. As shown in program WEEKA, we could use
110 LET A$ = "SUNM$NTUEWED'l'HUFRISA T"

\!EEKA

94 REM *THIS PROGRAM DEMONSTRATES CCMPARING A

95 REM STRING AGAINST A SUBSTRING
-100 DIM D S ( I 5 J , AS ( 2 I J
110 LE.T AS="S!JNMONTUE\JEDTHUFRI SAT"
120 INPUT 0$
130 LET D=O
-140 FOR I= I TO 19 STEP 3
150 LET D=D+I
-·160 IF DS <> A$(J,J+2J THEN 190
170 PRINT "DAY #"D
180 STOP
190 NEXT I
194 REM
200 PRINT "DAY NOT FCIJND"
210 END
RIJN
''IEEKA

?TUE
DAY # 3

Then we could step through the string with

140 F$R I = 1 T$ 19 S'I'EP 3
comparing some test string with A$(1,1+2) to see if the test string matches those
three characters of A$.
38 Advanced BASIC

We can even test parts of two strings. We could test the first three charac-
ters of D$ successively against groups of three characters in A$ looking for a
match by changing line 160 to
160 IF D$(1,3) <> A$(I,I+2) THEN 190
This is left for the reader to try.
It should be clear that by the use of strings we can control what is printed
in every space of the paper on our terminal. If we convert numeric values to
string variables, then we gain some added flexibility for the printing of numeric
results since these can be printed imbedded in strings to avoid extra spaces. If
you have PRINT USING, this step is not necessary.
We shall develop here the beginnings of a routine to convert a numeric to
a string. The fundamental idea is simply to pick off the digits one at a time as
numbers and use the numbers to store string equivalents in the correct positions
of a string variable. The way to get the string equivalent is to use what is called
a dummy string. In this case the dummy string to use contains all the digits.
As shown in program C(/>NVRT, it is D$="0123456789". The string A$ in pro-
gram WEEKA is another example of a dummy string. If we want a "3" stored
in the sixth position of string 8$, we use the following statement:
100 LET 8$(6,6) = D$(4,4)
since the "3" is really in the fourth position of D$. To get the digit "I" in the
Jth position of 8$, we use:
200 LET S$(J,J) = D$(1+1,I+l)
as shown in line 200 of program C(/>NVRT.

CCNIJRT

94 REM *
THIS PRr.GRAM CGNVO:RTS A NIJMFR!r:
95 REM TO A STRING
100 DIM SSC6J,DS( !OJ
110 LET DS="Ol?.34567q9"
120 INP•.JT N
130 IF N=INTCNl THEN 160
140 PRl'IT "INTSGF:RS CNLY"
150 GOTG 120
160 P~INT "*"lNl"*"
170 FOR E=5 TO 0 STEP -I
180 LFT ,J=6-E
190 L~T l=INTCN/!OtEl
-200 U:T S$[.J,Jl=D$[l+ld+ll
210 LET N=N-l*IOtF.
220 NICXT E
:::!30 PRINT "$ 11 ; SS; 0
S"
240 ENO
R•JN
CONVRT

?975310
* 975310. *
$975310$

To append a string onto thP. end of another string, use the technique of
program C(/>NCAT.
 Strings 39

C0NCAT

94 REM * THIS PROGRAM DEMONSTRATES CONCATFNATION

100 DIM A$(5QJ,B$!50l
110 INPUT AS, B$
120 PRINT "BEFORE CONCATENATION''.
130 PRINT A$,B$
140 LET ASCLEN<ASl+ll=BS
150 PRINT
160 PRINT "AFTER CONCATENATION",
170 PRINT AS.BS
180 END
RUN
CON CAT

? "CON CAT", "ENAT ION"

BEFORE CONCATENATION CONCAT EN AT ION

AFTER CONCATENATION CONCATENATION ENATION

Summary of Sec. 3-4

The substring scheme permits us to look at string variables in one of the
following four ways:
A$ The whole string.
A$(1,J) The substring from the Ith to the Jth characters inclusive.
A$(1,I) The single Ith character of the string A$.
A$(1) The substring beginning at the Ith character and continu-
ing to the end. To obtain the beginning of the string, use
A$(1,I).
We can build up new strings from parts of old ones. We can compare
strings and substrings for equality and for order.
Problems for Sec. 3-4
1) Modify ALPHA to eliminate duplicates.
2) Write a program to accept abbreviations for the days of the week and
respond with the full correct spelling.
3) As written, C<j')NVRT will print $000001$ if we input 1 for N. Elimi-
nate these leading zeros and end up with a string only as long as needed.
4) Modify C<j')NVRT to accept negative numbers and insert the minus sign
in the string.
5) Modify C<j')NVRT to accept decimal numbers representing dollars and
cents.
6) Write a program to convert a numeric string to a true numeric stored in
a numeric variable.
7) Write a program to multiply two six-digit integers and print the answer
exactly.
8) Write a program like that for problem 7, but not limited to six-digit
integers. (Allow 10- or 20-digit integers).
9) Write a program to encode and decode messages using a keyword.
10) Modify program WEEKA to test the first three characters of the input
string from line 120 against successive sets of three characters from A$.
11) In comparing strings for order, BASIC compares the two strings one
40 Advanced BASIC

character at a time. So for nonequal strings, the first nonequal charac-

ters determine order. Therefore when comparing such strings as '60'
and '100', BASIC will evaluate '60' as greater than '100'. Write a pro-
gram to overcome this flaw. Be sure to also provide for proper order-
ing of strings like 'A6' and 'AlO.'
3-5 The String Array Scheme*
Even where systems use the same general scheme, we will find differences.
Therefore we present programs here which use features found on many time-
sharing systems. Strings and string arrays may be manipulated in many of the
same ways that numerics and numeric arrays may be. That is, we may use
INPUT, READ, PRINT, assignment, and comparison for order. We generally
may not use arithmetic operations with the possible exception of addition since
some computers allow LET A$ "ABC" + "DEF" to assign "ABCDEF" to
string variable A$. Generally speaking, we find the same dimensioning require-
ments for string arrays as for numeric arrays. So 10 DIM A$(15),B$(3,4 7),
C(3,8) provides for up to 16 strings in A$ (generally zero subscripts are allowed),
up to 192 strings in B$, and up to 36 numerics in C. Note that we may intermix
strings and numerics in the dimension statement. Some systems allow only one-
dimensional string arrays. The number of characters allowed in each string will
vary from system to system. The LEN() function, if available; evaluates the
number of characters actually stored in a string. An alternative approach uses
the CHANGE statement. CHANGE A$ T</J A places the ASCII numeric code for
the characters in the string in the list positions of A. So if A$ = "XYZ," then
list A carries in position 1 the ASCII code for X, in position 2 the code for Y,
and in position 3 the code for Z. Moreover, A(O) will contain the number of
characters in the string A$. CHANGE also works in the other direction. Change
A T</J A$ takes the codes stored in list A and enters the equivalent string charac-
ter in A$. With this information, we can write some sample programs.
To arrange the letters of a string in alphabetical order we may first store
the coded values in a list. We then arrange the coded values in numeric order
and convert the numeric list back to a string for printing, as shown in program
ALPHAl.
Similarly, we can reverse the order of the letters in a string with the proce-
dure of program REVRSl. Look particularly at line 150 to see that the B list is
filled in the reverse order from the contents of the A list.
We can compare strings as a means of coding information. For instance,
we can number the days of the week by storing their names as the elements of a
list-the first day stored in the first element, the second day in the second
element, and so on. We then know which number belongs to a particular string
by which subscript we use to obtain a match, as shown in program WEEKAl.
We ran romp2re p~rts 0f sttl!!gs by p!ope! l!se ~f the CH~t\~!GE statement
and a list for each string being compared. Program WEEKBl looks at only the
first three letters of two strings in lines 190, 200, and 210.

*The programs of Sec. 3-5 were run on the General Electric Information
Services time sharing system.
 Strings 41

ALPHA I

94 REM * THIS PROGRAM ORDERS TH~ CHARACTERS

95 HEM OF A STHING
100 DIM A<75l
110 INP'JT A$
120 CHANG~ AS TO A
122
130 LET L = A<Ol
140 FOR I = I TO L-1
15 0 FOR J = I+ I TO L
160 IF A< I l <= .A<Jl THEN 200
162
164 RFM *
IF O'JT CF C'IDFR THF'N FXCHA\JGE
170 LFT X = A<Il
!RO LET A<Il ACJl
190 LET ALJl = X
200 NEXT J
210 NEXT I
212
220 CHANCiE A TC AS
230 PRINT AS
240 END
RIJN
ALPHA I

? THE ~JICK AHOWN FOX JUMPED OVFH THF LAZY DOGS

ABCDDEEEEFGHH J .JKLMNCOCC PORR STTI llJV\·!XYZ

HEVRSI

94 REM * THIS PROGRA"1 1<F.IJFl{S!"S THC:

95 REM CHAHACTERS CF A STRING
100 DIM A(75l,AC75l
110 INP 1JT AS
120 CHANGE AS TO A
122
124 REM *
STClff THF: N'IMAEi{ '1' C4A"<ACTF.ilS IN L AND 'HO>
130 LET L = A<Ol = A<Ol
140 FOR I = I TO L
-150 LET BCL-l+ll = A<Il
160 NEXT I
162
170 CHANGE B TC AS
!RO PRINT AS
190 E:ND
RIJN
REV HS I

? GOCD THINGS COMF IN SMALL PACKAGFS

SEGAKCAP LLAMS NI EMOC SGN!4T DCCG

\•!EEKA I

94 REM * THIS PHOG~AM FINDS THE DAY NUMOER

95 REM FHCM THE DAY NAME
100 DIM ASC7l
I I0 FQ R D = TC 7
l~O HEAD ASCOi
130 NF:XT D
ltiO PRINT "DAY";
150 INPIJT BS
16 0 FO .~ D = I TC 7
170 IF RS <• AS<Dl T4EN ?00
!RO PRINT "DAY # "; D
190 STCP
200 NEXT D
210 PRINT BS; "NOT FOIJND"
42 Advanced BASIC

212
214 REM
220 DATA SUNDAY, ·'10NflAY, T 1 1ESDAY, ''!EDNESOAY
230 DATA TH•JRSDAY, Fl<JDAY, SAT•JRDAY
240 END
RIJN
WEEKAI

DAY? T•JESDAY
DAY # 3

94 i"~M * T4IS P1'0GRA"l fJIFF>!~S F"CI·~ '·~FKAl RECl'.\IJ~;o:

95 HEM IT CHECKS ONLY THF FIRST T4~FF CHAHACTO:HS
9~ REM OF TWO STH!NGS. SFF LINFS 190 TO 210
100 DIM ASC7J,AC9),R(9)
102
104 HEM *
READ DAYS CF T4F WFFK INTC AS LIST
I 10 FDR D = I Tr 7
120 "EAD AS<Dl
130 NEXT D
132
-140 PRI'JT "DAY"J
150 INPIJT RS
160 CHANGF RS TC g
17 0 FDR l = 1 TO 7
180 CHANGE ASl!l TO A
lfl2
184 REM *
TEST FIRST THRFF CHARACTERS
190 FDR K = I TO 3
-200 IF ACKl <> 91KI THFN 240
-·210 NEXT K
220 PRINT ASCilJ " IS LJAY #";
230 STCP
240 N~XT I
250 PRINT 3SJ " NGT FCIJN!l"

?54 ilFM
260 DATA SIJNDAY, MONDAY, T 1 JFSOAY, 1•TDNESO.l\Y
270 D.l\TA THl!RSDAY, FRIDAY, SAT•J!WAY
2fl0 END
RllN

DAY? SIJNDAF
SUNDAY IS DAY # 1

We can easily construct a program to alphabetize strings by using exactly

the same approach that we use for sorting numbers except that we use sub·
scripted string arrays instead of subscripted numeric arrays, as shown in program
'l>RDER$. (For more about sorting see Sec. 11-3). Note that this program stops
comparing as soon as the list is in order, an optimizing feature which obviously
leads to shorter execution time.

94 REM THIS PRC GRA'! ALPHAflF:TIZF:S

95 '<EM
*
A LI ST OF STRINGS
96 RFM THE Cl IRRFMT Ll'!IT IS 200 STil!NGS
100 DIM LS<2001
102
104 llEM Rl':AD DATA
*
 Strings 43

110 FOR I = 1 TO 200

120 READ LSl!I
130 IF LSl!I = "STOP" THEN 150
140 NEXT I
142
150 LFT N = I = I-1
152
154 REM *
T'IRN S'.·!!TC4 OFF, INCRFMF.MT N ANO SORT
160 LET S = 0
170 LET N N -1=
1RO FOR J = 1 TO N
190 IF l$1JI LSIJ+ll THEN 240
192
19/i ~F'1 *
EXCHANGF: Fl'':t~FNTS AND TllRN r.N S'-'ITC4
200 LFT SS = LSI JI
210 LET LSIJI = LSIJ+l I
220 !_FT LT.IJ+ll =SS
2~0 LFT S = 1
240 NFXT .J

244 REM * CH1'CI{ s·,11 rr:y ll="FF. l=ON

250 IF S = 1 THEN 160
252
254 RF:M *
THF SWITCH IS OFF - LIST JS CRDFRFD
26 0 FC R X = 1 TC l
270 PRINT LllXI
280 NEXT X
~R~
8R4 REM
290 DATA WILLIAMS, JCN~S. SMIT'-!
300 DATA YOUNGMAN, STOP
:310 END
RUN
ORDER$

JONES
SMITH
WILLIAMS
Y0UNGMAN

Since strings in the scheme we are now considering may be treated as array
variables, we should realize that we may MAT READ and MAT PRINT string
arrays with all of the same options that these two statements allow for numerics.
That is, we may MAT READ A$(M,N) to redimension the array in the MAT
READ statement. And we may MAT PRINT with semicolon or comma spacing.
Use semicolon spacing with care, however, as this will result in strings being
printed with no spaces. For most systems, quotes are not required when placing
strings in DATA statements unless the string contains a comma or is a numeral.

MAT$

94 REM *
THIS PROGRAM D•MONSTHATFS NAT RFAD
95 REM AND MAT PRINT FCR STHING A~RAYS
100 DIM ASl5, IOI
110 READ R,C
120 MAT READ ASCH.Cl
130 MAT PRINT AS;
140 PRINT
150 MAT PRINT A$,
152
15 4 REM
160 DATA 2,5
170 DATA THIS. rs, A. SAMPLE. PROGRAV
180 DATA WITH, STRING, MAT RFAD, AND, MAT PRINT
190 END
44 Advanced BASIC

RUN
MAT$

THISISASAMPLFPHOGRAM
WITHSTHINGMAT READANDMAT P~INT

THIS IS A SAl"PLE: PliCGRAM

WITH STRING MAT HEAD .l\:>JD

Generally speaking, you need not be intimately familiar with the ASCII
code because the ASC( ) function gives the numeric code for the character in
parentheses. Alternatively, we may write a program to print the ASCII code for
alphanumeric characters stored in a string variable, as shown in program ASC.
Note that lines 130, 150, and 160 are used to store single characters from A$ in
the string variable B$. Some time sharing systems provide the EXT$ function.
The statement B$ = EXT$(A$,I,I) would result in the Ith character of A$ being
stored in B$. Some systems use SEG$ for this. You may find numerous other
functions for working with string data available in your system.

ASC

94 REM * THIS PROGRAM PRINTS SAMPLE

95 REM ASCII CODE VALUES
100 DIM A(9)
110 LET A$= "6 :/C[ABZ"
120 CHANGE AS TO A
-130 LET BCO> = l
140 FOR X = l TO ACO>
-150 LET B<l> =ACX>
-160 CHANGE 8 TO 8$
170 PRINT ""'; 8$; ""'; 8(1)
180 NEXT X
190 END
RUN
ASC

'6. 54
32
••• 58
• /. 47
• (. 40
'(. 91
'A' 65
'A 1 66
'Z' 90

Summary of Sec. 3-5

row I, column J of a string array. We may consider individual characters of the

string by using CHANGE, which places equivalent numeric codes for the charac-
ters of the string in the positions of a numeric list corresponding to the position
of the character in the string. In addition, the zero position of the list contains
the number of characters in the string.
 Strings 45

Problems for Sec. 3-5

1) Modify program ALPHAl to eliminate duplications.
2) Suppose you have typed a large quantity of data consisting of peoples
names, first name first, and you suddenly find that you should have
typed them last name first and in alphabetical order. Write a program
to make the change. Your program should work whether or not you
used middle initials.
3) Write a program to convert a string to a numeric (even though your
computer may provide a function for this).
4) Write a program to accept two six-digit integers, multiply them digit
by digit, and print the exact product.
5) Write a program to accept two numbers as strings not limited to six
digits, find the exact product, and print it.
6) Write a program to generate license plate number and letter combina-
tions. Establish a pattern, such as six characters with the first three
alphabetic and the last three numeric. Select a few at random to print.
7) Write a program to generate 'words' where vowels and consonants
alternate. Print a few at random.
8) In comparing strings for order, BASIC compares the two strings one
character at a time. So for nonequal strings, the first nonequal charac-
ters determine order. Therefore, when comparing such strings as '60'
and '100,' BASIC will evaluate '60' as greater then '100.' Write a pro-
gram to overcome this flaw. Be sure to also provide for proper order-
ing of strings like 'A6' and 'AlO.'
 4
FILES

4-1 Introduction
There are tremendous differences in file handling from system to system.
Files give a tremendous extension of power to the BASIC language in two sig-
nificant ways. First, the use of files allows us to handle far more data than we
could ever store in DATA statements of a single program because of computer
space limitations. Second, the use of files allows us to process the same data
using several different programs. We might use one program only to enter data
into a file. Another program might be used to make corrections. Another might
be used to extract a particular piece of information, and yet another might be
used to modify the data in the files. And so on and so on. There is no limit to
the number of programs which may work on the data in a single file or group of
files.
We have chosen to present here file programs on two different systems.
You will want to determine which one resembles your system before attempting
to write programs. The systems chosen here are Hewlett Packard and General
Electric since both are in common use and differ in many ways.
4-2 Hewlett Packard Files
Serial Data Files
Serial files have the advantage that they are quite efficient in the use of
computer storage space. We simply print entry after entry until all desired
entries are in the file or files. Serial files have the disadvantage that we cannot
easily go into the file to change an existing entry or insert a new entry. To do
this to a serial file, we must copy the entire contents into another file making
the changes on the way and then copy the corrected data back into the original
file. Naturally, this is a serious problem only if we are working with a large
amount of data.
Let us construct an inventory file and work on it. Suppose that for each
item of inventory we have a part name, a part number, a price, and a quantity.
46
 Files 47

Program ENTERl will enter an initial set of inventory data into a file called
INVOl.

!':.NTF:"ll

94 REM *
THIS P~OGRAM "NT~RS DATA IN AN EMPTY FILE
JOO FIL>'.S I"JVOI
110 DIM PSPSl
120 READ PS,"J,P,Q
130 IF N=O TH~N 240
134 R<:M *
LIN~ 140 Ptl!NTS IT""1 AFTE'l IT~"1 CLOSE PACKED
135 llFM IN THE: FILE MAKIN(; IT si;:RJAL IN STR'JCT'JR'C
140 P~INT #t;PS1N1P10
ISO GOTO 120
194 Rr:M
~00 D!\TC\ "F~AMIS'',10r)t,~.J,l80()," 1 ·llDGi::T'',t00?.1L!.1i,tlOO
210 DATA "RFG'ILATO;<", IOOJ,J.4R,9QQ,"5LIC"JEP", I0r)4,.r)4,9QQO
220 DATA "F-lASMALATO-l". I QOS. I q. 49, qoo. ""'HATS IT". I -Jf)~. I 0. 9R. 1000
230 DATA "STOP",Q,t),O
240 END
'l'IN
"NTF..Rl

DCNE

Line 100 is necessary to make file INVOl available to the present program.
The file must be in existence at the time the program is run. (Some HP systems
provide for ASSIGNing files during program execution.) In this case, we typed
the executive command <tJPEN-INVOl,50 to create the file. That command
specified 50 discrete storage blocks referred to as "sectors" or "records." Each
of our 50 records will hold as many as 32 numerics, where one numeric occu-
pies the same space as 4 string characters. We must add one string character each
for delimiters at the beginning and end of a string and add one if the number of
characters in the string is odd. The record size varies from computer to com-
puter, and some computers allow the user to specify its length. The maximum
number of records allowed varies also.
Line 120 reads from program data statements in the usual way.
Line 130 checks for dummy data.
Line 140 prints the data as read in line 120 but to the external file instead
of to the paper on the terminal in front of us. The #1 is used because the file we
use is the first named in the files statement. If we want to print to the second
file named in a files statement, then we use PRINT #2. We can also use PRINT
#F where F is the file number. File names are separated by commas in the
FILES statement. For example,
100 FILES FILE1,FILE2,FILE3
makes three files available to the program in which this statement appears. Sys-
tems vary as to the number of files that may be named in a files statement. Most
allow at least 8.
This is the very first time that we have run a program which failed to print
anything and yet did something useful. Generally speaking, we should print
something to the terminal. To prove that the data really is in that file, we pre-
sent program READ to read the data out of file INVOl.
48 Advanced BASIC

94
I 00
110 rJ IM PS("" l
l~O P<I\JT "P.'\n \J'\M:O:","P!\H #","Pnc1c","JJANTITY"
l".:!Ll ~C:"'i *
RO:AD DA fl\ '"<8'1 TY" F !IJ
-130 RF.A[) H};P·~ .. ~,;.i,J
134 ·<F.'1 *
Pn\JT T'l'" l)l\TA TC Fl~ TF. <MP.J!\L
140 P ·~ Tr,J T PS, ~, P, ')
1~0 f;'.JTC 111
-140 C::N!J
'J\J

PD.f~T \Jt:\"11'..: PA ~l ~Ir;::: 11<\IJT!TY

F' <'I'; IS I 011 " t)

~. '1 I "JCJ
'.• rnr-r. T I IJO~ I~ • ll 111 11
·~>-: r: 11_ C\ TG i~ lrJ<)J J •I~~ 91)0
SL! CNEP 1004 .04 9000
l"~ASMAL.A TOR I 005 18.49 ~00
WHATS IT 1006 10.98 3000

END-OF-FILE/END OF ~ECORD IN LINE 1:30

We get all of the information from the file and an error message to boot.
We may avoid the error message by inserting an IF END statement. We will use
105 IF END #1 THEN 160. This statement, executed just once, sets a flag so
that if line 130 tries to read nonexistent data or tries to read past the physical
end of the file, the computer will next execute line 160. Alternatively, of
course, we could have placed our own flag in the file by printing dummy data
into the file at the end of the real data.

-105 IF END #I THEN 160

RIJN
READ I

PA~T NAM<: PART H PRICE 0 1 JANTITY

FRAM IS 1001 2.3 11300
'#!DGET 1002 4.4 1100
"EG'll.ATOR 1003 3.4~ 900
SL! CNEP 1004 ·04 9000
"RASMAL.ATOR 1005 I fl. 49 800
',IHAT SIT 100~ Io. 9fl 3000

We can now sit down and dream up little programs to process the data in
file INVOl. We might want to know the number of parts in inventory. You can
easily verify that program READ2 does that.

READ2

94 REM* THIS PROGRAM FINDS THE TOTAL.

?S REM ~~M3ER CF ~A~TS IN FILE !NVO!
100 FILES !NVOI
110 DIM P$[25J
120 LET T=O
130 IF END HI THEN 170
140 READ #IJP$,N,P,Q
144 REM * RUNNING TOTAL CALCULATED IN LINE 150
150 LET T=T+O
160 GOTO 140
 Files 49
170 PRINT T1"PARTS"
180 END
RIJN
READ2

16600 PARTS

It is left as an exercise for the reader to find the total value of inventory
with a program.
Now suppose we want to add inventory to INVOl. Using ENTERl again
with different data will simply replace what is in INVOl with new data. (Actu-
ally, the new data might replace only part of what is already in there, and this
result has other complications.) What we must do is read out to the end of the
data already in the file and then begin printing new data to the file at that point,
as shown in program ENTER2.
Note that program ENTER2 may also be used to enter data into an empty
file. We do not need a different program for this purpose. It is a good practice
to provide for printed output to the terminal in a program of this kind. We run
READl again to make sure that the new data was properly added to the end of
the file.

ENTER2

94 REM * THIS PROGRAM ENTERS DATA INTO AN EMPTY OR

95 REM PARTIALLY FILLED SERIAL FILE
100 FILES INVOI
110 DIM PS[25J,Q$[25J
114 REM* READ TO THE END OF EXISTING DATA
--120 IF END #1 THEN 150
130 READ #1JP$,N,p,Q
1-.0 GOTO 130
144 REM *ENTER NEW DATA INTO THE FILE
-· 150 READ P$,N,P,Q
160 IF N=O THEN 200
170 LET OS=P$
180 PRINT #lJPS,N,P,Q
190 GOTO 150
200 PR INT "LAST ITEM IS "J Q$
204 REM
210 DATA "LIFTER".1007 •• 29, 10000."DRCJPPER"· l008. ,g9, 1500
220 DATA "WHDSIT", 1009.16, 12,"HOLOER", 1010• ,47, 1141
230 DATA "STOP".D.o.o
240 END
R'JN
ENTER2

LAST ITEM IS HOLDER

RIJN
READ!

PART NAME PART # PRICE D'JANTITY

FRAM IS IDOi 2.3 1800
WIDGET 1002 4,4 1100
REGULATOR 1003 3,4g 900
SLICNEP 1004 .04 9000
FRASMALATOR 1005 IR· 49 800
WHATS IT 1006 10098 3000
LIFTER 1007 .29 10000
DROPPER 1008 • 89 1500
WHOSI T 1009 16 12
HOLDER 1010 ,47 1141
50 Advanced BASIC

The next area we might get into is that of making changes in file INVOl
according to business activity. We can do this with the procedures mentioned
earlier, that is, by copying the entire contents of the file into another file,
making changes on the way. However, if the amount of data becomes very
great, the time lost by this method tends to be more important than the space
saved by using serial files. Furthermore, there are many devices for utilizing ran-
dom access files more efficiently. So let's talk about random access files.
Random Access Files
Files are serial or random access according to the structure the program-
mer creates for them. In fact, the same file may be treated as serial by one
program and random access by another. File INVOl was serial because we
printed the data in serially. We may instead use the discrete records of the file
for specifically arranged data.
In our inventory example we might consider placing one inventory item on
one record. Since most of our sample data are equivalent to about 5 numerics,
this would be wasteful of computer space. However, in practice, an item of in-
ventory will contain more data, and thus it is often wise to allow more file space
than is needed at first. Then we will not have to restructure the entire storage
plan as we might if we had placed several inventory items on a single record.
In fact, if programmers anticipate that additional data may be required, they
go one step further. They incorporate the extra variables into all programs,
using zeros for numbers and empty or null strings for anticipated alphameric
data. Then the existing programs will carry all of the variables needed and will
not have to be rewritten to accommodate the new data structure. Of course,
changes will have to be made to actually utilize the newly activated variables
later. Moreover, in some situations we can use space more efficiently by storing
numerics in one file and strings in another with a scheme to link the data from
the two files to each other. That way numeric data may be MAT READ out
of the file for very easy processing.

ENTER3

94 REM * THIS PROGRAM ENTERS DATA INTO AN EMPTY

95 REM rILE IN RANDOM ACCESS rORMAT
100 PILES INV02
110 DIM P$C25l
-120 LET R=O
130 READ P$,N,P,O
140 IF N=O THEN 190
-·150 LET R=R+I
-160 READ #1,R
-170 PRINT #IJP$,N,p,Q
!BO GOTO 130
190 PRINT R1"RECIJRDS USED"
194 REM
2CC !).l'l.,'f!'. "F'P.!'.M!S": !00!.-?.-:':l· '800·"WTf'1f;FT".100?"4·4~ 1100
210 DATA "REGULATOR", 1003,3.46,900,"S LICNEP''. 1004, .04,9000
220 DATA "rRASMALATIJR"· 1005. 1B.49,soo,"WHAT SIT"• 1006.10.96, 3000
230 DATA "STOP",o,o.o
240 ENO
RUN
ENTER3

6 RECORDS USED
 Files 51

To keep things simple, suppose that we look at a program to place one

inventory item per record, such as program EN'fER3, which PRINTs inventory
data into file INV02.
Note that we have the computer print the number of records used to
assure us that the program has performed as expected. Lines 120 and 150 keep
track of the record to which we are going to print. Statement 160 sets a pointer
in the file to the beginning of record R. Statement 170 prints data to the file
beginning at the pointer. The pair of statements 160 and 170 may be replaced
with the single statement, PRINT #1,R;P$,N,P,Q, the difference between the
two being that if 170 attempts to print more than will fit on the record, the
excess goes on the next record, whereas PRINT #1,R;P$,N,P,Q will result in an
error message if the data to be printed will not fit on record R.
Now program READl may still be used to read the contents of INV02
by changing line 100 to FILES INV02.
In order to add data to the end of file INV02, we have several techniques
available to us. One is exactly analogous to that for serial files. Another uses
the TYP() function. The TYP() function 'looks' at the next item of informa-
tion in a file and identifies it as follows:
TYP(N) 1 Next item is a numeric
TYP(N) 2 Next item is a string
TYP(N) 3 Next item is end of file
TYP(N) 4 Next item is end of record
where N is the position that the file name occupies in the FILES statement. If
N is positive, the value 4 is never returned; to detect end of record, N must be
negative.
The TYP( ) function is used in line 150 of program ENTER4 to determine

ENTER4

94 REM * THIS PROGRAM ENTERS DATA IN A PARTIALLY FILLED

9S REM FILE IN RANDOM ACCESS FORM
100 FILES INV02
110 DIM P$(2SJ
120 LET R=O
130 LET R=R+l
140 READ N!,R
144 REM *
LINE !SO DIRECTS THE COMPUTER TO LINE
l4S REM 130 IF WE ARE NOT AT THE END OF DATA
-ISO IF TYP< I> <> 3 THEN 130
160 LET R=R-1
170 READ P$,N,p,Q
180 IF N=O THEN 230
190 LET R=R+I
200 READ #!,R
210 PRINT NlJP$,N,P,Q
220 GOTO 170
230 PRINT Rl "RECORDS IJSED"
234 REM
240 DATA "LIFTER''. 1007, .29, !0000,"DROPPER", 1008, .89, 1500
2SO DATA "WHOSIT", !009, 16, !2,"HOLDER", !OI0,.47, 1141
260 DATA "STOP",o,o.o
270 END
RUN
ENTER4

10 RECORDS USED
52 Advanced BASIC

the first record that has an end of file (sometimes referred to as EOF) marker
and begin printing the new data on that record. Here again, as a check on the
program, we have the computer tell us how many records have been used.
Now we are in a position to explore some possibilities for editing the ran-
dom access file. Note that we selected part numbers, so that if we subtract
1000, we get the number of the record on which that part will be found. Often
in data processing it is helpful to organize data so that something about the data
tells us where to find it. This method is sometimes called "content addressing."
It could consist of an extra item of data for file management purposes only.
Let's write a program that allows us to go into the inventory file and
change anything but the part number. We do this with program UPDATE. The
part number is requested in line 150, and the part is found in line 200. Line
210 prints the part name and offers to make a change possible. The same is
done for the price in line 260 and for the quantity in line 310. When all infor-
mation is correct, it is re,entered into the file in line 360. The IF END state-
ment in line 130 provides for attempting to access a nonexistent record or for
attempting to access a record that has no data on it.

IJPDATE

94 REM * THIS PROGRAM EDITS AN INVENTORY FILE

100 FILES INV02
110 DIM P$l25J,A$[ 31
120 PRINT "INPUT PART NIJMBER ZERO TO 0'JIT"
-130 IF END #I THEN 380
140 PRINT
-150 PRINT "PART #"J
160 INP'JT NI
170 IF NI <> INTCNI l THEN 150
180 IF Nl=O THEN 860
184 REM *CALCIJLATE RECORD FROM PART NIJMBER
190 LET R=Nl·IOOO
-200 READ #l,RJP$,N,p,Q
210 PRINT P$1
220 GO SIJB 800
230 IF A$="YES" THEN 260
240 PRINT "CHANGE TCJ"l
250 INP'JT P$
-·260 PRINT "PRICE = $"!Pl
270 GO SUB 800
2fl0 IF A$="YES" THEN 310
290 PRINT "CHANGE TCJ $"J
300 INPUT P
-310 PRINT "QTY ="I OJ
320 GOSUB 800
330 IF A$= "YES" THEN 36 0
340 PRINT "CHANGE TO"J
350 !NPIJT Q
354 REM *PRINT CORRECTED INFORMATION BACK TO THE FILE
-360 PRINT #1,RJPS,N,p,Q
370 Gorn 140
380 PRINT "NO SUCH PART #"
390 GOTO 140

790 PRINT "YES OR NO"

800 PRINT " OK"l
810 INPUT A$
820 IF A$="YES" THEN 850
830 IF A$="N(J" THEN 850
840 GOTO 790
850 RETURN
860 END
 Files 53
RUN
IJPDATE

INPIJT PART NIJMBER ZERO TO OIJIT

PART #?1001
F'RAMIS OK?YES
PRICE = $ 2.3 OK?NO
CHANGE T0 $? 3, 24
QTY = 1800 OK?YES

PART #?I 003

REGIJLATOR OK? YES
PRICE = $ 3.48 OK?YES
QTY = 900 OK?NO
CHANGE HJ? 87 B
PART #?0

A run of READl confirms that the proper changes were indeed made in
the file.

100 F'IL1'S INV02

RIJN
READ!

PART NAME PART H PRICE OIJANTI TY

F'RAMIS I rJO I 3.24 1800
WIDGET 1002 4.4 1100
REG'JLATOR 1003 3,4g 878
SLI CNE:P 1004 .04 9000
F'RASMALATOR I 005 tB.49 800
WHAT SIT I 006 10.98 3000
LIF'TER 1007 .29 10000
DROPPER I 003 .g9 1500
WHl'JSIT 1009 16 12
H3LDER 1010 .47 1141

Summary of Sec. 4-2

We have seen that files are serial or random access depending only on the
approach that a program takes in printing data to or reading data from the file.
In order to treat a file as random access, the data must be placed so that its lo-
cation within a record is known. Files are made available to a program with the
FILES statement. We can detect the end of data or physical end of file with the
IF END statement. Data is entered into a file with the PRINT# statement and
read from a file with the READ# statement. In addition, we may determine the
nature of the next information in the file through use of the TYP( ) function.
A file pointer can be set to the beginning of record R of file F with READ #F,R
without reading any data.
Problems for Sec. 4-2
1) Arrange 10 or more strings in alphabetical order by placing them one
to a record in a file.
2) Write a program to print the contents of a file without knowing the
structure of the file.
3) Write a program to copy the contents of one file into another. See if
you can provide for copying 'holes' too.
54 Advanced BASIC

4) Write a program to find the first empty record and determine the num-
ber of records in the file.
4-3 General Electric Files
We actually are going to talk about external data files in this section, for
in fact, programs are files too, and the data contained in DATA statements of a
program are referred to as "internal files." The creation and use of data files can
be a very complex business indeed. It is the purpose of this section to present
some of the data file concepts and provide sufficient examples so that the reader
will be able to use the power of files in future programming. (Our examples will
of necessity handle only small amounts of data, as it would not be practical to
print the entire contents of very large files.)
Files of two types are available to BASIC programs. They are called
"ASCII files" and "BINARY files." ASCII files are also referred to as "Teletype
files." This term reflects the fact that they may be created by typing data
directly at the terminal using line numbers exactly as a program is typed at the
terminal and that they may be listed directly to the terminal just as a program
may be listed to the terminal. ASCII files may also be used by programs. Binary
files may be used only under program control and may not be written to or read
from the terminal.
ASCII Files
Data may be typed directly to the terminal. All we have to do is create a
file with the command NEW and begin typing our data as if it were a program.
Each line must have a line number followed by a space followed by our data
separated by commas. String and numeric data may be intermixed. Lines may
be corrected by retyping them. Lines may be deleted by typing the line number
followed by return. For an ASCII file to be usable at some later time, it must be
saved by typing the command SA VE.
To demonstrate some of the uses of ASCII files, we have selected the
names of the ten largest cities in the UB. according to the 1960 census, their
rank, and the percentage change in population from 1960 to 1970. File CITY
has been created, the data typed to the terminal, and the file saved as described
above. Since this is a listable file, we do so below.

LIST CITY
CITY

100 BALTIM0RE MD>6•-4•7

110 CHICAG0 ILL•2•-6•3
120 CLEVELAND 0HJ0,B,-IS.7
130 DETR01T MICH.S,-10·6
140 H0UST0N TEXAs,7,29.3
150 L0S ANGELES CALJF,3,12.2
160 NEW Y0RK N,y,,J,-,1
170 PHILADELPHIA PA.,4,-3.B
160 ST L0UIS M0>10·-19
190 WASHINGTON o.c •• 9.-4.B

We may now write programs to access the data in file CITY. Probably the
simplest useful task we could perform would be to print the contents of the file
under program control.
 Files 55

In order to make the file available to the program we use the FILES state·
ment in line 100 of program READCITY. We may read data from the external
file in a manner similar to that with which we read data from internal files. To
read from the file named in the files statement, we use READ #1 followed by
the list of variables we want. This is done in line 120 of the program. Every
time such a statement is executed, a pointer is moved in the file so that the next
read statement begins to read at the pointer. Initially, all file pointers are at the
very beginning of the file. Note that we could use READ #0 to read from the
data statement of the program. Since it is possible to attempt to read past the
end of the data in the file, we use the IF MIJ)RE statement in line 140. As long
as there is more data in the file, line 140 in our program will direct the computer
to 120. When the data has all been read, line 140 will test false and control
passes to line 150 in this case.

RE ADC I TY

94 REM *
THIS PROGRAM READS THE CONTENTS Or rlLE CITY
-100 rlLES CITY
110 PRINT "CITY"l TABC20)J "RANK"J TABC25)J "% GROWTH"
-120 READ Nt, C$,R,G
130 PRINT C$J TABC20)J Rl TABC25ll G
-140 Ir MORE NI THEN 120
t 50 END
RUN
READC!TY

CITY RANK % GROWTH

BAL T !M0RE MD 6 -4.7
CHICAG0 ILL 2 -6. 3
CLEVELAND 0Hl0 8 -15.7
DETRO IT MI CH 5 -10.6
HOUSTON TEXAS 7 29.3
L0S ANGELES CAL!r 3 12.2
NEW YORK N.y. I -o. I
PHILADELPHIA PA. 4 -3.8
ST LOIJI S MO 10 -19
WASHINGTON D. c. 9 -4.8

It turns out that ASCII files are always sequential (serial) in format. This
means that data must be accessed by reading from the first set of data step by
step until the desired information is obtained. There is no way to begin at some
intermediate point of the file. We can, however, add data to the end of the file
with an append statement. APPEND #1 sets the file pointer to the end of data in
the file and prepares the file for writing. For any data processing that requires
tabulating information from each line of data, the sequential nature of ASCII
files is ideal. Furthermore, ASCII files are very easy to edit from the keyboard,
as described earlier. Having seen how to read an ASCII file, we will next see how
to write to such a file with a program.
Let's write a program to transfer the data from file CITY to file CITYl,
rearranging the data so that the order in which the cities appear in the new file
will be according to decreasing percentage growth. One way to create the new
file is with the NEW command. Once you have named the file, type 100, press
the space bar, then the return key, and save the file. Now we have to provide
access to two files in one program. This is done with the FILES statement. Up
to eight files may be named in a FILES statement as long as they are separated
56 Advanced BASIC

with semicolons. These files may then be designated by number according to

the order in which they are named in the FILES statement, as shown by line 100
of program GR'{>WTH.
Since it is easy to sort numbers in a one-dimensional list, the strategy we
will use here is to pass through the file once, entering the percentage growth for
each city in the array A as we go. Thus A(l) becomes the percentage growth of

GROWTH
94 REM * THIS PROGRAM ARRANGES TEN CITIES ACCORDING TO
95 REM GROWTH RATE USING TWO SEQUENTIAL FILES
IDO FILES CITYJ CITY!
110 DIM ACIQ), BCIO>
112
114 REM* ENTER GROWTH DATA IN ARRAY A AND ORIGINAL
115 REM POSITION IN FILE CITY IN ARRAY B
120 FOR I = I TO 10
130 READ Nt, N$,R,G
140 LET AC!> G
150 LET BCI> = I
160 NEXT I
162
164 REM * BEGINNING OF SORT
170 LET N = I 0
180 LET s = 0
190 FOR I = I TO N-1
200 IF AC!> >= AC!+ I> THEN 280
202
204 REM * EXCHANGE OUT OF ORDER DATA
205 REM AND TURN SWITCH ON
210 LET SI = AC!>
220 LET AC!>= ACI+I>
230 LET AC!+I) = SI
240 LET SI = BC!>
250 LET BC!>= BCI+I>
260 LET BC!+I> = SI
270 LET S = I
280 NEXT I
282
290 LET N = N-1
300 IF S = I THEN 180
302
304 REM * GROWTH RATE IS IN ORDER NOW PRINT
305 REM DATA IN FILE CITY!
31 0 SCRATCH H2
320 FOR I = I TO I 0
330 RESTORE #I
340 FOR J = I TO BC!>
350 READ Nt, N$,R,G
360 NEXT J
362
370 WRITE #2, N$,R,G
380 PRINT N$
390 NEXT I
392
400 END
RIJN
GROWTH
HOUSHlN tEXAS
LOS ANGELES CALIF
NEW YORK N.Y.
PHILADELPHIA PA.
BALTIMORE MD
WASHINGTON D· c.
CHICAGO ILL
DETRO IT MICH
CLEVELAND OHIO
ST LOUIS MO
 Files 57

the first city in the file and A(lO) becomes the percentage growth of the tenth
city in the file. Entering is done in lines 120 through 160. We store the original
position in a B array.
Now that the list is in order we know that whatever city has the percentage
growth stored in A(l) goes first in our second file and whatever city has the
growth rate in A(I) goes in the Ith position of file CITYl. However, in order to
prepare a file for printing, we must use the SCRATCH statement. That state-
ment erases whatever is in a file, prepares it for write mode, and sets the file
pointer at the beginning of the file. Since we want to write to the second file
in our FILES statement, we use SCRATCH #2 in line 310. The next complica-
tion we face is that the pointer in file CITY is now at the end of the data. We
move the pointer to the beginning of the file and keep the file in read mode with
the REST</)RE statement in line 330. REST</)RE #N acts on the Nth file in the
FILES statement. The REST</)RE statement must be executed every time we
want to reread the file. The actual entry of the data to file CITYl takes place
in line 370 of program GR</)WTH when we have found the city with the growth
rate in the Ith location of the arranged list A by reading to that city's position
in the original file as determined by list B.
We now run READCITY on file CITYl to confirm the contents of that
file.

100 rILES CITY!

RUN
REAOCITY

CITY RANK % GROWTH

HOUSTON TEXAS 7 29.3
LOS ANGELES CALIF" 3 12.2
NEW YORK N•Y• I -o. I
PHILADELPHIA PA. 4 -3.8
BAL TI MORE MD 6 -4.7
WASHINGTON D.c. 9 -4. 8
CHICAGO ILL 2 -6.3
DETRO IT MI CH 5 -10.6
CLEVELAND OH'IO 8 -15.7
ST LOIJ IS MO 10 -19

Binary Files
Binary files are available only under program control. Whereas ASCII files
may be only sequential, binary files may be either sequential or random access.
Random access means that any data item may be accessed without reading all
data from the beginning of the file up to that data item and that data may be
written to any point in the file directly in the same manner.
An ASCII file or a binary file may be used as a sequential file as deter-
mined by the first WRITE statement which applies to that file after the file is
made available for write mode by the SCRATCH statement. To work with a
file as a sequential binary file simply use a colon (:) where the pound sign (#)
occurs in the READ, REST</>RE, SCRATCH, IF M</>RE and WRITE statements.
The distinction between ASCII and binary files, regardless of sequentiality or
randomness, has to do with the code that is used to store the data on some
device peripheral to the computer, and does not generally effect the programmer
at the level of programming in BASIC.
58 Advanced BASI<;;

Random Access Files

Thus far we have not been concerned with how much space our data has
occupied because our files simply expanded to fit whatever we had written to
the file. We are not likely to exceed the maximum space available in a single file
for some time. However, since random access files must have a structure allow-
ing data to be placed in physical locations that will be known to us for future
access, space requirements for data storage become important. The structure of
a random access file is a little like that of a two-dimensional array except that in
a file we may intermix numeric and string data. In some situations, however, we
may choose to place numeric data in one file or group of files and related string
data in another file or group of files, with a scheme for relating the two
structures.
Random access files may be segmented into components called "records."
We may assign the amount of storage space in each record at the time we create
the file. Storage space is measured in computer words, and storage is required
for data as follows:
1 word per numeric
1 word per 4 string characters or fraction thereof
1 word for control per string for internal computer purposes.
For example we could store 120 numerics in any file in which the product of the
number of records and the number of words per record is at least 120. Since we
are going to have control over the space in each record, we generally select some
record size that is appropriate to our data structure, just as we usually dimension
an array to fit the structure of the data we intend to store in it.
Let us begin by writing 10 random numbers to a random access file, proving
that the numbers are really there and then arranging them in increasing order by
using the random accessibility of the file to do the sorting.
First, the file must exist. The system command CREATE is required here.
Let's provide two records which will handle 10 words each. Then our ten num-
bers will all fit on the first record. The executive command,
CRE RAND,(RAN(l0,2))
accomplishes this. The numbers in parentheses specify the number of words per
record first and the number of records second. Program RNDENTER is identi-
cal to a program that would write ten random numbers to an ASCII file except
that a colon appears where a pound sign would appear in a program writing to a
sequential file and no SCRATCH statement is required.

RNDENTER

94 RF:M * THIS PROGRA"l WRITES 10 RANDOM N 1JMBERS

95 REM TO A RANDO"l ACCESS FILE
100 FILES RAND
110 RANDOMIZE
120 FOR I =
I TO 10
130 WRITE : I, RND
140 NEXT I
142
150 END
RUN
RNDENTER
 Files 59

Note that the program generates no printed output. (It is not good general
procedure to write programs with no visible output except to prove, as with pro-
gram RNDENTER, that the program did in fact do something invisible.) We
now need a program to read file RAND. Program RNDREAD does that.

RN DREAD

94 REM *THIS PROGRAM READS NUMERICS FROM A

95 REM RANDOM ACCESS FILE SEQ•JO:•'JTIALLY
100 F"IU::S RAND
110 READ : l, A
-120 IF" A=O THEN 150
130 PRINT A
140 GOTO 110
ISO END
RUN

o.~~0239
0.257207
0.705748
0.717468
0.143835
o. 349935
0.704994
0.649726
0.974231
0.852828

RNDREAD is like a program to read a sequential file, with one important

difference. Look at line 120. Since we didn't put any zero values into the file,
why test for zero? We would have used the IF M<tlRE statement in an ASCII
file. IF M<tlRE does not determine an end of data condition in a random access
file effectively because the CREATE command filled the file with binary zeros.
The SCRATCH statement will have the same effect. Thus we are able to termi-
nate our little program by reading values of A until we come up with zero.
IF M<tlRE does determine whether or not we have reached the physical end of
the file, however. Thus in our file RAND, if we had read 20 numerics, then IF
MC/>RE :1 would test false. A companion statement IF END :1 would test true
in this condition. We should be beginning to see, and will soon see even more
clearly, that the total structure of the data storage is the complete responsibility
of the programmer. We must know when we have reached the end of our real
data. We may want to count the number of data items or we may want to place
dummy data at the end of the file just as we often do for data statements of a
program.
Now let us arrange the numbers of file RAND in ascending order by
operating on the data within the file as we go. The basic sorting scheme is ex-
actly like that we just used for sorting cities and that we will use again in chapter
eleven. The only difference is that we are now comparing numbers that have
addresses in a file instead of in a list. In order to set the file pointer to the
proper location of our file, we use the SETW statement. SETW N T<t> X moves
the pointer in file N to the beginning of the Xth word of storage counting from
the beginning of the file. This operation is totally independent of the number of
words per record. Thus if a file has nine words per record, then SE'l'W N T</J 12
moves the pointer to the beginning of the third word on the second record of
60 Advanced BASIC

file number N. Once the pointer is set, the next read or write statement begins
reading or writing from that point. In our problem we will be reading a single
numeric value.
In program RNDS'l>RT, line 130 sets the pointer to the Ith word of the
file so that the Ith number may be read by line 140. Line 150 sets the pointer
to the I+ 1st word so that the I+ 1st number may be read out. If they are in the
desired order, line 170 directs the computer to line 230, which causes the next
pair of adjacent numbers to be tested. Should the test in line 170 fail lines 180

RNDSORT

94 REM * THIS PROGRAM ARRANGES 10 NUMBERS IN

95 REM ORDER IN A RANDOM ACCESS FILE
100 FIL.ES RAND
110 LET S = 0
120 F0R I = I TCl 9
-130 SETW I TO I
-140 READ : 1, A
-150 SETW I TO l+I
160 READ : 1, B
-110 IF A <= B THEN 230
172
174 REM *
EXCHANGE OUT OF ORDER DATA
160 SETW I TO I
190 WRITE : 1, B
-- { 200 SETW I TCl !+I
210 WRITE : 1, A
220 LET S I=
-230 NEXT I
232
-240 IFS= I THEN 110
250 PRINT "DONE"
260 END
RUN
RNDSORT

DCJNE

through 210, exchange the positions that the two values occupied in the file.
Then a switch is turned on by setting S equal to one. When the computer has
passed through the list, we test in line 240 to see if any exchanges have been
made. If there have been no exchanges, then the numbers are in order and we
have the computer print "D'l>NE". In the present situation, we rerun
RNDREAD to verify that the program did in fact sort the file.

RUN
RNDREAD

a. 143835
a. 220289
0.251201
Q.349935
Q.649726
0.704994
Q.705748
Q.717468
0.852828
Q.974231

For our final example, let us take the data in the ASCII file CITY and
enter it into a random access file. To do this we must plan very carefully, as we
 Files 61

must be able to specify the position of all data in the file exactly. It is usually
convenient to group the data that belongs together on a single record. In our
problem we have three variables: the city name, its rank, and its percentage
growth. The rank and percentage growth are both numerics and so occupy one
word of computer storage each. The city names are strings of different lengths
and so require different amounts of space. A reasonable approach is to allow
space for the largest string and dimension our file accordingly. Clearly L(,t)S
ANGELES CALIF is the longest string we use. The string has four groups of
four characters and one character left over. That requires five words. We must
add one word for internal control plus the two for the numerics. That puts the
required space at eight words of storage per city. So we must CREATE a file
with at least 10 records at eight words per record. We do this with
CRE CITY2,(RAN(8,10))
In order to set the file pointer to the beginning of the Ith record where the
records contain eight words each, all we have to do is point 8(I-1)+1 words from
the beginning of the file, as shown in line 140 of program XFER. This program
transfers the data from ASCII file CITY to random access file CITY2. Notice
that there is no difficulty whatever in accessing two different kinds of file with
the same program.

XFER

94 REM * THIS PROGRAM COPIES DATA FROM ASCII FILE CITY

95 REM TO RANDOM ACCESS FILE CITY2
100 FILES CITYJ CITY2
110 LET I = 0
120 READ #I, cs.R.G
130 LET I = I+I
132
134 REM * THE FORMULA IN LINE 140 SETS THE
135 REM POINTER TO THE ITH RECORD IN THE FILE
140 SETW 2 TO 8*CI-l>+I
150 WRITE 12, C$,R,G
160 IF MORE #I THEN 120
170 PRINT IJ "ENTRIES"
180 END
RUN
XFER

10 ENTRIES

Our final task is to arrange the cities according to rank without using a
second file and without using the technique of saving the numbers to be sorted
in an array. This program, program RANK, is very similar to RNDS(,t)RT except
that the pointer is a formula and the read and write statements work with three
variables instead of one.
RANK

94 REM * THIS PROGRAM ORDERS CITIES ACCORDING TO RANK

95 REM IN A RANDOM ACCESS FILE
100 FILES CITY2
110 LET NI = 0
120 LET N = 10
130 LET S = 0
140 FOR I = I TO N-1
62 Advanced BASIC

ISO SEHi I TO 8*<1-1)+ 1

160 READ : 1, cs,R,G
170 SETW I TO 8*I+I
180 READ I I, c1:1;,R1,G1
190 IF R <= RI THEN 250
192
194 REM * EXCHANGE OUT OF ORDER DATA
200 SETW I Hl 8*< I- I>+ I
210 WRITE 11, Cl$,Rl,G I
220 SETW I TO B*l+I
230 WRITE : t, C$, R, G
240 LET s = I
250 NEXT I
252
260 LET N = N-1
270 LET NI = NI+ I
280 IF s = I THEN 130
282
290 PRINT "CITIES ORDERED ON RANK"
300 PRINT "IN"J Nil "PASSES "
310 END
RUN
RANK

CITIES ORDERED ON RANK

IN 7 PASSES

Program REDCITY2 tabulates the results directly from the file.

REDCITY2
94 REM * THIS PROGRAM READS THE CONTENTS OF FILE CITY2
95 REM AND PRINTS TO THE TERMINAL
100 Fil.ES CITY2
110 PRINT "CITY"; TABC20)J "RANK"J TABC25)J "% GROWTH"
120 FOR I = I TO I 0
130 SETW I TO 8*CI-l) +I
-140 READ 11, C$,R,G
--150 PRINT CSJ TABC20JJ RJ TABC26J; G
160 NEXT I
162
170 END
RUN
REDC!TY2
CITY RANK % GROWTH
NEW YORK N.y, I -0.J
CHICAGO ILL 2 -6.3
LOS ANGELES CALIF 3 12.2
PHILADELPHIA PA. 4 -3.8
DETR01T MICH 5 -10.6
BAL T IM0RE MD 6 -4,7
HOUST0N TEXAS 7 29.3
CLEVELAND 0Hl0 8 - 15· 7
WASHINGT0N D. C. 9 -4.8
ST L0UIS M0 10 -19

Summ arv of Sec. 4-3

and
ASCII and binary files are used to store and to arrange both string
either by the ter-
numeric data. We have seen that ASCII files can be accessed
files are available only under pro·
minal directly or by a program, whereas binary
but ASCII
gram control. Binary files may be either sequential or random access,
ial. In order to make any file available to a program , the
files are only sequent
 Files 63

FILES statement must name the files we want, separated by semicolons. The
files are numbered according to the order of appearance in the files statement.
As many as eight files may be accessed by a single program simultaneously.
(Provision may be made for substituting new files for previously named old ones
within a program.)
The following special statements apply to ASCII files: REST</>RE #N
moves the file pointer to the beginning of file N and places the file in read mode.
READ #N,A,B,C$ reads values from file N for variables A,B,C$. WRITE
#N,X,B$ enters the contents of X and B$ into file N. SCRATCH #N erases
the contents of file N and places the file in write mode at its beginning. IF
M</>RE #N tests for more data in file N. IF END #N tests for the end of data
condition in file N. All the above statements can be used for binary files by
changing the pound sign (#)to a colon(:). There are some differences when it
comes to random access files, however. SCRATCH :N, when applied to a
random file, fills the file with zeros, sets the pointer to the beginning of the file,
and places it in write mode. IF END and IF M</>RE work only at the physical
end of the file. In addition, there is a special statement for random files to place
the file pointer under program control. SETW N TC/> I may be used to place the
file pointer of file N at the beginning of the Ith word of the file.
Problems for Sec. 4-3
1) Enter student names followed by a fixed number of test scores in a
sequential file. Arrange the students by test average and write the re-
sults to another file.
2) Suppose you have entered a large number of names in a file, last name
first, one name to a string for alphabetizing. Write a program to print
a list of names in alphabetical order but first name first.
3) Write a program to read numerics from two ordered files and print a
single merged and ordered list to the terminal.
4) Write a program that will insert an item of data into an already ordered
random access file so that the new item is in order.
5) You have a random access file with unknown contents. Write a pro-
gram to determine the number of words the file contains.
6) Write a program to enter inventory data into a random access file. The
data should include, for every part, a number, name, price, quantity,
low order point, and reorder quantity. The low order point is the
quantity that should trigger reordering for that part, and the reorder
quantity is the quantity that should be ordered. Write a program or
programs to do any or all of the following: modify quantity and or
price according to business activity, edit the low order point and re-
order quantities, process the file to find the total dollar value of the
inventory, and process the file to determine what parts must be re-
ordered and the quantities to be ordered.
 5

E INAL

5-1 Introductio n
There are many sophisticated mechanical plotters on the market offered
by a number of manufacturers. These generally use a pen which draws very
short (0.01 inch is typical) straight line segments as determined by a program.
The smaller the segments, the smoother the curve. However, this chapter is
limited to using the terminal itself as a plotter. There are a number of disad-
vantages to using the terminal for this purpose. Plotting is slow and may there-
fore be expensive. The graphs we get are imprecise, and we can't easily use
graph paper. Be all that as it may, if we have no other plotter, we can get very
helpful plots from the terminal. The terminal has the advantage that it is con-
veniently available.
5-2 Plotting a Function
We can greatly simplify plotting by choosing a function, since functions
have the feature that for any value of x there is exactly one value for y. So, if
we think of y as increasing across the page from left to right and x increasing
down the page from top to bottom, when the plot is completed, we just turn
the paper ninety degrees counterclockwise to obtain the conventional orienta-
tion.
To plot a single function, we first locate the origin and then concern our-
selves with having the terminal mark the axes. Let's plot y = t x + 3. The
b2s!c p!ctt!~g !::; VC"tj" ::;:rnp!c. All that is i1Ccessary is to use Lile T AB(I~) printing
function to get the printing mechanism out to the proper location on the paper
and then print some symbol. Since the TAB function begins counting at the
left margin and we would like to have negative values available for y on the
graph, it will be necessary to move the x axis to the right. How far we move it
64
 Plotting on the Terminal 65

will depend on the particular graph. We simply add some constant value to
all TAB arguments. This value is set in line 110 of program GRAPHl. If your
version of BASIC does not have a TAB(N) function, then you may put
PRINT" ";in a loop to get the printing head to the proper spot on the paper.
We print a plus sign to indicate the origin. We will therefore have to check to
see if we are at x = 0 and then determine whether the point of the graph is to
the left or the right of zero.

GRAPH I

94 REM * THIS IS A RUDIMENTARY Pl.0TTING PR0GRAM

100 DEr rNrCX> = ·5•X+3
102
104 REM * K M0VES THE 0RIGIN K SPACES T0 THE RIGHT
-110 LET K =10
120 r0R X = •7 T0 7
130 LET y = rNr c x)
-140 Ir X <> 0 THEN 210
150 Ir Y < .,5 THEN 180
152
154 REM * X = 0 AND Y >= ··5 50 PRINT 0RIGIN
160 PRINT TABCK)J "+"J
170 Ir Y < ,5 THEN 220
180 PRINT TABC INTCY+o5>+K )J "*"J
190 Ir Y < -.s THEN 160
200 G0T0 220
210 PRINT TABC INTCY+.S)+K )J "*"J
220 PRINT
230 NEXT X
232
240 END
RUN
GRAPH I

*
*
*
*
*
*
+ *
*
**
*
*
*
*
*
GRAPHl seems to do the job we set out to do. (Be sure to tum the re-
sults ninety degrees counterclockwise to view the graph.) However, there are
many improvements that we can make. Let's put in a set of axes. Since the
y-axis is the line where x = 0, the y-axis provision can go after line 140 of
GRAPHl. In order to put the x-axis in, we have to check for each value of x
whether or not the plotted point is below the x-axis. This is done in line 200 of
GRAPH2. As long as we are working on the program, let's add a little flexibility
by putting the position of the x-axis and the domain in as data. See lines 110,
130, and 600 of GRAPH2. Note that a distinguishing character is used to mark
every tenth location in each of the axes. This greatly improves the readability of
the plot. Lines 210 and 420 determine where those marks are printed.
66 Advanced BASIC

GRAPH2

94 REM * THIS PR0GRAM PL0TS GRAPHS 0F FUNCTI0NS

95 REM WITH AXES PRINTED
100 DEF FNFCX> •5*X+3 =
102
104 REM * THE VALUE 0F N DETERMINES LENGTH 0F Y-AXIS
105 REM K M0VES THE GRAPH K SPACES T0 THE RIGHT
106 REM F IS THE FIRST VALUE 0F X F0R THE PL0T
107 REM L IS THE LAST VALUE 0F X F0R THE PL0T
-110 READ N, K• F, L
120 PRINT TABCK-l>J "X•"J F
-130 F0R X • F T0 L
140 LET Y =
FNFCX>
150 IF X <> 0 THEN 200
152
154 REM * X • 0 S0 THIS IS THE Y-AXIS
160 F0R YI • -K T0 N-K
170 G0SlJB 400
180 NEXT YI
190 G0T0 280
192
194 REM *
P0INTS 0FF Y-AXIS ARE PRINTED
195 REM IN LINES 200 T0 260
-200 IF Y < ,5 THEN 260
-210 IF X/10 <> INTCX/IO> THEN 240
220 PRHIT TABCK>J "-"J
230 GlilT0 250
240 PRINT TABCK>J "!"J
250 IF Y < ,5 THEN 260
260 PRINT TABC INTCY+.Sl+K ll "*"I
270 IF Y < -·5 THEN 210
280 PRINT
290 NEXT X
292
300 PRINT TABCK-l)J "X •"I L
310 ST0P
392
394 REM *
SUBR0UTINE 10 PRINT Y-AXIS
400 IF YI • 0 THEN 490
410 IF ABSCY-Yll < ,5 THEN 470
-420 IF Yl/10 • INTCYl/IO> THEN 450
430 PRINT "-"J
440 RETURN
450 PRINT "l"J
460 RETURN
470 PRINT "*"J
460 RETURN
490 PRINT "+"J
500 RETURN
502
594 REM
-600 DATA 25, 10, -11, 12
610 END

Three runs of G RAPH2 show a variety of results. For the first RUN of
GRAPH2, we present the line y = 1-
x + 3 from program GRAPHl. For the
second RUN, we re-define the function in line 100 to plot the parabola, y =
-.4x 2 - x + 8. For the third RUN, we again re-define the function in line
100 to get the graph of y = .2x 3 - 2x 2 + x + 5.
We have in GRAPH2 a sat!sfacto:ry bas!c graph!?!g p!og!~!!!, =.!!d. the !e-
sults may be enhanced by sketching a curve through the plotted points by hand.
We can still improve on the program itself, though. Consider what the sine
graph would look like using the program as it stands. The graph would show us
little. Clearly we could do better by putting a scale capability in for both axes.
This is left as an exercise.
 Plotting on the Terminal 67

RIJN 100 DEf FNFCX> = -·4*X•2-X+8

GRAPH2 600 DATA 55, 40, -12, 9
RUN
X=-11 GRAPH2
* I
*- X=-12
• I

*'*'*
* •
I

* •
'*'* *
*
*
*
I
I
* *
* I •
1---------+--*-
*
-----1----- *
! *
* *
--------1---------1---------1---------·-------•-1---
* I *
* *
* !
*
*
'*
* I
* * I
* *
* *
* *
x = 12 * x = 9

100 DEF FNFCXl = .2*X•3-2*Xt2+X+ 5

600 DATA 69, 30, -3, 11
RUN
GRAPH2
X=-3
*
I *
*
1---------1---------1---------+! ----·----1------
*
---1---------1---------
* '*
*
*
*
*
* !
*
*
x = 11
*

Summary of Sec. 5-2

We have developed a program to use the terminal as a plotter. This ap-
pears to give us rough but satisfactory graphs for a variety of functions.
Problems for Sec. 5-2
1) Modify GRAPH2 to allow a change of scale. Allow for two different
scales for the two axes. Use the new program to plot any of the
following:
a) y sin x
b) y = cos x
68 Advanced BASIC

c) y = x - INT(x)
d) y = x + INT(x)
e) Y = Vx
2) You may have noticed that the mechanics of your terminal are such
that the vertical and horizontal scales are different. On many termi-
nals there are 10 characters per inch on the horizontal line and six
lines per inch on the vertical page. Use the scaling ability of problem
1 to provide equal scale for both axes relative to the graph.
3) Modify GRAPH2 to handle more than one function.

5-3 Plotting Using Data Stored in an Array

There are some other things we can do with plotting. For instance, we
can plot a graph that does not have to be rotated to achieve standard orienta-
tion. And we might want to plot some nonfunction relation.
One way to achieve such a capability is to set up a computer array in such
a way that each storage location of the array corresponds to a coordinate point
of the graph. We will have some adjustment problems, however, because for an
array the "starting point" is the upper left corner and for a graph the "starting
point" is generally nearer the center of things. This difficulty can be handled
by shifting the "starting point" of the graph to the left and up. Also for an
array, row numbers increase from top to bottom, and the reverse is true for a
graph. This problem can be handled by putting the graph into the array "up-
side down".
We can scan through the array inserting numbers which will later be inter-
preted to print specific characters. Let us make the graph cover an odd number
of coordinate points in each direction and make the middle element of the array
correspond to the origin of the graph. We c,an begin by setting every entry of
the array to zero, to be ignored when the printing takes place. Then we use a
'1' to signify the origin, a '2' to signify the horizontal axis, a '3' to signify the
vertical axis, and a '4' to signify the plotted point. This scheme will allow more
relations to be handled by using numbers greater than 4 for the additional
graphs.
All these features are incorporated in program GRAPH3. The axes are
shifted, and the y-axis is turned upside down by line 230. The y-axis is entered
in line 160. The x-axis is entered in line 170. The origin is entered at line 190.
The use of the variable Ll is to save program storage. It takes less computer
storage to set Ll equal to L + 1 and then use Ll (rather than L + 1) as long
as we need this value several times, and moreover it saves typing. On some sys-
tems, storage will not be a limitation for a program like this, but if it is, then
such a procedure is worth using. Line 220 tests to see if the coordinate position
of the array is on the graph. This depends on our choice for the value oft. If

However, by increasing the value oft, we can have more points plotted. This
tends to "fill in" the graph. Lines 110 and all other references to D are
employed to limit the actual use to only that part of the array that seems neces·
sary for the graph requested. This will save printing time. As written, the pro-
gram causes the printing mechanism to scan the entire line. Considerable print-
 Plotting on the Terminal 69

GRAPH3

94 REM *
GRAPH F'RDM DATA ENTERED IN AN ARRAY
100 DIM AC4B,4Bl
102
104 REM * 0 SPECIF'IES WIDTH AND HEIGHT OF' GRAPH
105 REM L IS THE NUMBER OF' SPACES EACH SIDE OF' ZERO
106 REM T DETERMINES GRAPH RESOLUTION
107 REM LI IS USED TO SAVE TYPING L+I REPEATEDLY
110 LET D =
21
120 LET L = <D-1 l/2
130 LET T = •4
140 LET LI = L+I
142
144 REM * ENTER AXES AND ORIGIN IN THE ARRAY
150 F'0R I = I TD D
- 160 LET A<Ll• I> 3
-170 LET A(!, Lt> 2
!BO NEXT I
-190 LET ACLl•LI> =I
192
194 REM * NOW INSERT 4' S TD DESIGNATE PLOTTED POINTS
200 F'OR Y = -L TD L
210 F'0R X = -L TO L
-220 IF' ABS< •5*X+3-Y > > T THEN 300
-230 LET AC X+LI , D-CY+Ll > = 4
300 NEXT X
310 NEXT Y
312
314 REM * THE VALUES ARE STORED NOW PRINT
320 PRINT TABCLllJ "Y"
330 F'DR Y = I TO D
340 F'OR X = I T0 D
350 PRINT TAB<X>J
360 IF' ACX,Y> 0 THEN 380
370 G0SUB 500
380 NEXT X
390 PRINT
400 NEXT Y
410 PRINT TAB<L>J "-Y"
420 ST0P
492
494 REM * PRINTING SUBROUTINE
500 IF' ACX,Y> 2 THEN 550
510 IF' A<X•Y> 3 THEN 580
520 IF' ACX,Y> =
4 THEN 610
530 PRINT "+"J
540 RETURN
550 IF' <X-Lll/10 INT< CX-Lll/10 >THEN 590
560 PRINT "-''J
570 RETURN
580 IF' <Y-Lll/10 INT< <Y-Lll/10 > THEN 560
590 PRINT "!";
600 RETURN
610 PRINT "*"J
620 RETURN
692
700 END

ing time could be saved by sending the printing head back to the beginning of
the new line as soon as the last point is printed. This is left as an exercise.
Three runs of GRAPH3 are presented. The first RUN is the straight line of
the first RUN of program GRAPH2; the second RUN is the parabola of the
second RUN of GRAPH2; and the third RUN is a circle with radius 10 and
center at the point (2,- 3). Note that as with the graphs of Sec. 5-2, these graphs
are distorted by the fact that the space occupied by a single character on the
printed page is higher than it is wide.
70 Advanced BASIC

RIJN 130 LET T = • 6

GRAPH3 ~>o IF ABSC -.4*Xt2-X+8-Y l > T THEN 300
RIJN
y GRAPH3

y
*
*
*
! ** * *
* *!
!
! *
*
!---*-----+---------! *
* !-~~*-----+---~-----!
*
*
*
*
*
*
-Y
*
* -Y

110 LET D = 29
130 LET T 5 =
220 IF ABSC CX-2lt2+CY+3lt2-IOO l > T THEN 300
RUN
GRAPH3

* *****
! *
* *
* *
* *
* *
-~~-!---------+---~-----!----
* ! *
* *
* *
* *
* *

* *
* *
**
*
* !
*****
-Y
 Plotting on the Terminal 71

Summary of Sec. 5-3

GRAPH3 gives us the ability to deal with many kinds of graphs fairly
easily. The graph can be oriented in the conventional manner, and we can see
that the ability to graph more than one relation on one set of axes is a direct
extension of the current program.
Problems for Sec. 5-3
1) Modify GRAPH3 to stop printing when the last character of the cur-
rent line has been printed.
2) Modify GRAPH3 to permit two relations to be plotted. Use your pro-
gram to find the approximate points of intersection of y = 2x 2
+ x - 1 and y = 3x + 4.
3) Use the ideas of GRAPH3 to write your name. That is, store points
to be printed in an array. If you have a long name, maybe you'd
rather use the word BASIC. This exercise may be a little tedious, but
fun.
4) Rewrite GRAPH3 so that the origin does not have to be the center
point of the array.
5) We get an interesting effect if instead of plotting points, we plot spaces.
That is, where there is no point on the graph, print an asterisk, and
where there is a point, leave the location on the paper blank. Try this.
6) Modify GRAPH3 to allow different scales for the two axes. Then plot
a large circle to see how well you can d.o.
7) On some systems, the array size allowed is limited enough to make
some plots not practical if we use the methods of this section. One
way to program around this is to notice that there is a tremendous
amount of wasted storage in each element of the array itself. Note
that for up to six relations GRAPH3 requires only a one digit number
to store the information required for graphing. Since most computers
provide at least six digits, by using each of those digits we can increase
the storage by a factor of six. Write a graphing program to use this
additional storage space.
8) One method for obtaining larger graphs is described in problem 7.
Another procedure would be simply to analyze the graph one line at a
time. Try this.
 A EA UN
A c

6-1 Introduction
Let us consider a moving object. At constant speed, the distance traveled
is simply its speed multiplied by time. If an object travels 15 feet per second for
5 seconds, it will travel 75 feet. However, it is often true that the speed of an
object is not constant. Suppose the graph of Fig. 6-1 represents an object in
nonconstant motion. The distance the object has traveled at time t is the area
under the graph from the origin to t. For Fig. 6-1 that area is a convenient
geometric shape. We get Area = A = t*60*3 = 90 ft.

70
60
.,,~ 50
::: 40
30
20
10

2 3 4 5

Figure 6-1

Now suppose a graph does not provide such a convenient geometric shape,
as in Fig. 6-2.
72
 Area Under a Curve 73

ft/sec

seconds
Figure 6-2

To find the total area of such a shape, we can subdivide the area into many
smaller segments, find the area of each segment, and sum up the individual areas.
All we have to do is decide what kind of smaller segments to use and how large
they should be. Consider Fig. 6-3.

Figure 6-3

Figure 6-3 subdivides the area into rectangles so constructed that each one
falls completely within the desired area. Thus we know that the sum of these
areas will be less than the actual area. We could alternatively place the rectangles
as shown in Fig. 6-4. But that would have the opposite effect on the accuracy.
Another possibility is to construct the rectangles so that the midpoint of the top

Figure 6-4
74 Advanced BASIC

side is a point of the curve, as shown in Fig. 6-5. Now we have some area in-
cluded by the rectangles that is not included by the actual curve and some area
included by the curve but excluded by the rectangles. Thus we expect some
cancelling-out effect.

Figure 6-5

Let us begin by writing a program to sum up the area shown in Fig. 6-5.
In order to test the program, we can make our first function contain a known
2 2
area. We choose a semicircle derived from the circle (x - 5) + y = 25.
This circle has a radius of 5 and its center at the point (5,0), as shown in
f 2
Fig. 6-6A, and so the area we expect is n r 2 or (rr*5 )/2.

(0,0) (10,0)

Figure 6·6A

The curve we are considering is in fact a function from 0 to 10 for x.

Solving (x - 5) 2 + y 2 = 25 for y 2 , we get
y2 = 25 - (x -5) 2

and
y -./25 - (x - 5) 2
 Area Under a Curve 75
For the shaded area of Fig. 6-6A, the function is
y = f(x) = y25 - (x - 5) 2
Thus we can easily define a computer function as follows:
100 DEF FNS(X) = SQR(25 - (X-5)t2))
For this problem we can begin by taking 10 intervals one unit wide. Then the
midpoint of an interval will be x - .5, and the height of the rectangle will be
FNS(X - .5). See Fig. 6-6B and program AREAl.

Figure 6-68

AREAi

94 REM * THIS PROGRAM COMPARES THE AREA OF A

95 REM SEMICIRCLE FOUND BY FORMULA AND BY
96 REM SUMMING AREAS 0F RECTANGLES.
-100 DEF FNSCXl = SQRC 25-CX-5lt2 l
110 LET A= 3.14159*25*•5
120 PRINT "AREA OF SEMICIRCLE BY FORMULA ="J A
130 LET A = 0
132
134 REM * THE LENGTH OF THE BASE IS B
140 LET B = I
!'so F0R X = I T0 I 0
152
154 REM * THE HEIGHT IS H
160 LET H = FNSCX-.S>
162
164 REM * THE AREA OF CURRENT RECTANGLE IS B*H
170 LET A = A + B*H
180 NEXT X
182
190 PRINT TABC3lJ "AREA BY SUMMING RECTANGLES ="I A
200 END
RUN
AREAi

AREA OF SEMICIRCLE BY FORMULA = 39.2699

AREA BY SUMMING RECTANGLES = 39.6499

The relative error is about .38 in 39 or about 1%. We can improve on this
by taking smaller intervals within the accuracy of the computer.
Let's rewrite program AREAl to allow varying widths of intervals. We can
take the b = 1 out of the x loop and allow the value of b to be read from data.
76 Advanced BASIC

Now since the width of interval will change according to the value of b, the mid-
point of the interval will be x - b/2, and the value of x must be successively
incremented by b, the interval width. While we are at it, let us allow the value of
x to have a variable range read as data. This is done in lines 140 and 160 of
AREA2.

AREA2

94 REM * THIS PR0GRAM SUMS RECTANGLES T0

95 REM APPR0XIMATE AN AREA ALL0WING THE
96 REM WIDTH 0r THE RECTANGLES T0 BE READ
97 REM AS DATA.
100 DEr rNS<X> = SQR< 25-<X-5>•2 )
110 PRINT "F'R0M"• "IN'TERVAL"• "T0 .. ,, "AREA"
120 READ 8
130 Ir 8 =
0 THEN 250
-140 READ P-.T
ISO LET A = 0
-160 r0R x = r+B T0 T STEP 8
170 LET H • rNS< X-812 >
180 LET A = A + B*H
190 NEXT X
192
200 PRINT r, 9, T. A
210 G0T0 120
212
214 REM
220 DATA 1,0,10. o5o0ol0
230 DATA .1.o, 10, .01.0, 10
240 DATA 0
250 END
RUN
AREA2
rR0M INTERVAL T0 AREA
0 I 10 39.6499
0 ,5 10 39.4051
0 .1 10 39.2115
0 .01 10 39. 26 74

The results of AREA2 do indeed give successively more accurate approxi-

mations of the area.
Now we run AREA2 for the function,
f~) = 2x 3 - 2x 2 + x + 5

from -3 to 11 for x. For a graph of this function, see Sec. 5·2.

100 DEr rNS<X> = 2*X•3 - 2*X•2 + x + 5

220 DATA 1,-3, 11, .5.-3.11
230 DATA .05.-3.11
RUN

rR0M INTERVAL Tl1J AREA

-3 I II 6475
-3 .5 II 6494.25
-3 .05 II 6500.62
 Area Under a Curve 77

Summary of Chapter 6
We have developed a routine that approximates the area under a curve for
functions. The method used is to sum up areas of rectangles whose heights are
determined by the midpoint of the intervals. (There are several other methods,
which are left as exercises.)

Problems for Chapter 6

1) Find the area enclosed by y = x + 3 and y = x 2 - 8x + 17.
2) Find the area enclosed by y = - 3x 2 + 4x + 2 and the line y = - 3.
3) Another procedure for finding area under a curve is called the trapezoid
method. This is done by inscribing trapezoids instead of rectangles
under the curve. Write a program to use the trapezoid method.
4) Write a program that begins with a specified width and automatically
makes the subdivisions smaller until the new approximation does not
differ from the old approximation by more than some percentage error,
say .01%.
5) Write a program to approximate area by making the first interval the
whole domain and each subsequent interval one half the previous
interval. Have the computer stop when the percentage change is less
than, say, .01%.
6) A method always as accurate and usually more accurate than the trape-
zoid method is Simpson's rule. Simpson's rule requires an even number
of intervals and is given by the following (see Fig. 6-7):

A = gw (f1 + 4f2 + 2{3 + 4{4 + · · · + 2fn-2 + 4fn-1 + fn)

Write a program to use Simpson's rule.

Figure 6·7

7) Use the ideas of this section to find the length of the graphed line of a
continuous function instead of the area.
 7
c
OME

7-1 Points in a Plane

In the Cartesian coordinate system, points in a plane are named by ordered
pairs of real numbers. A point is labeled p(x,y), where the first number is called
the x-coordinate and the second is called they-coordin ate. As we consider the
ordered pairs of numbers associated with a point or set of points, many geo-
metric relationships unfold. For two points in a plane, we can look at the dis-
tance between them or consider the straight line they determine. For three
points in a plane, we may be interested in whether or not they fall on a straight
line, that is, are collinear. If not collinear, then three points determine a tri-
angle which has many properties of interest. We may examine many other com-
mon geometric figures such as quadrilaterals, circles, parabolas, etc. It is the
purpose of this chapter to study some of these topics, using the computer to
assist us.
7-2 Pairs of Points
Consider the two points A(3,5) and B(3, -1), as shown in Fig. 7-1. Clearly
the distance AB from A to B is six units, usually stated simply as 6. We define
the distance between two points with the same x-coordinate as the absolute
value of the difference in Y-coordinates. (For two points with the same y-
coordinate, the distance is defined as the absolute value of the difference in
x-r.oordinates.) Thus in Fig. 7-1.
AB = 15 - (-l)J = 15 + 11 = 6.
Left only with this definition we would have a very limited ability to determine
distance. We should, however, be interested in determining distance for any
two points whatever their coordinates.
78
 Coordinate Geometry 79

A(3,5)
e

•8(3, -1)
Figure 7-1

For the points A(3,5) and B(-1,2), we may find the distance by plotting
the points and constructing a right triangle, as shown in Fig. 7-2.

A(3, 5)

I
1 Ac= s- 21=3
J

I
8(-1,2) ___ ..de
BC= I3 - (-1) I= 4

Figure 7-2

Using the Pythagorean Theorem,

AB 2 = BC 2 + AC 2

or
AB = VBC 2 + AC 2
so that
AB = V4 2 + 32
AB = y'25
AB= 5
It is evident that for any two points A(x 1 ,yi) and B(x 2 ,y 2 ) we have the
relationships shown in Fig. 7-3.
80 Advanced BASIC

I
ii
I AC= IY2 - y
I
______ .de
BC=lxz-xd

Figure 7~3

Using AB VBC 2 + AC 2 , we get

2
AB = Vlx2 - X11 + IY2 - Y11 2
but since the square of a number equals the square of its absolute value, we get

This is known as the distance formula.

We shall now find distances for pairs of points. The distance formula
translates readily into the BASIC statement:
200 LET D = SQR ( (X2 - Xl) t2) + (Y2 - Yl) t2)
All that remains is to get coordinate pairs read into the computer and results
printed out. This is done in program DISTl.

DIST!

94 REM *
THIS PR0GRAM FINDS THE DISTANCE AB
95 REM r0R THE C00RDINATES CXJ,Yl> AND CX2,Y2)
100 PRINT "T0 rIND THE DISTANCE BETWEEN TW0 POINTS"
150 PRINT
160 PRINT "P01NT A"J
170 INPUT XI.YI
160 PRINT "POINT B"J
190 INPUT X2.Y2
192
194 REM *
CALCULATE DISTANCE AND PRINT IT
-200 LET D = SQRC CX2-Xll•2 + CY2-Yll•2 >
210 PRINT "DISTANCE AB ="J D
222
230 END
RUN
DIST!
T0 r!ND THE DISTANCE BETWEEN TW0 P01NTS

P01NT A?J,5
P01NT B?-1,2
DISTANCE AB = 5
 Coordinate Geometry 81

Program DISTl works fine for a single pair of points. But suppose we
had several pairs of points. We may simply request the number of calculations
desired and use FIJ}R-NEX'l', as shown in DIST2.

DIST2
94 REM * FINDS DISTANCE r0R PAIRS 0F P0INTS
95 REM PERMITS M0RE THAN 0NE SET 0r DATA
100 PRINT "T0 FIND THE DI STANCE BETWEEN TW0 P0 INTS"
110 PRINT
120 PRINT "N0W MANY PAIRS"J
130 INPUT N
132
140 F0R I = I T0 N
ISO PRINT
160 PRINT "P0 INT A"J
170 INPUT XI.YI
IBO PRINT "P0 INT B"J
190 INPUT x2,y2
192
194 REM * CALCULATE DISTANCE AND PRINT IT
-200 LET D = SQRC CX2-Xl)t2 + CY2-YJ>t2 >
210 PRINT "DISTANCE AB ="J D
220 NEXT I
222
230 END
RUN
DIST2
T0 rIND THE DISTANCE BETWEEN TW0 P0INTS
N0W MANY PAIRS?2
P0 INT A? 3, 4
P01NT B?o.o
DI STANCE AB = 5
P01NT A?-3.4, 5.75
P0 INT B? 3. 1251 2
DISTANCE AB = 7.52583

For any two points A and B, it can be shown that the coordinates of the
midpoint of segment AB are found by taking the average of the corresponding
coordinates of A and B. 'l'hus the midpoint of AB for A(x 1 ,yi) and B(x 2 ,y 2 ) is

It is left as an exercise for the reader to write a program to give coordinates of

midpoints.
Where coordinates of two points are known, another property of interest
is the slope of the line they determine. 'l'he slope is the ratio of the change in
y-coordinates to the change in x-coordinates, or

Y2
Slope = m = - ---Y1
-

Thus for A(l,3) and B(5,6),

82 Advanced BASIC

6 - 3 3
m
5 - 1 4

and for A(3, -4) and B(l,6),

6 - (-4) 10
m = -5
1 - 3 -2

Note that slopes up and to the right are positive and slopes up and to the
left are negative. A short program can easily be written to make the above cal-
culation, as shown by program SLQ}PE.

SLOPE

100 PRINT "THIS PROGRAM FINDS THE SLOPE OF AB"

110 PRINT
120 PRINT "H0W MANY PR0BLF.:MS"J
130 INPUT N
140 F0R I = I T0 N
142
150 PRINT
160 PRINT "P0 INT A"J
170 INPUT XI, YI
180 PRINT "PO INT B"J
190 INPUT X2,Y2
192
194 REM * CALCULATE SL.OPE AND PRINT IT
-- 230 LET M = CY2·Yll/CX2·XI>
240 PRINT "SLOPE ="J M
250 NEXT I
252
260 END
RIJN
SUJPE

THIS PROGRAM FINDS THE SL0PE OF AB

H0W MANY PROBLEMS?2

POINT A?O,O
POINT 8?4,5
SUJPE = 1.25

POINT A?3,6
POINT 8?5,-3
SL.OPE =-4.5

Look at line 230 in program SLl;1>PE. Note that we instruct the computer
to perform division. When two points have the same x-coordinate, division by
zero is required, which is a mathematically, and thus computationally, unde-
fined condition. The program should be modified to test the value of x 2 - x 1
before allowing division to take place. This is left as an exercise for the reader.
The idea that two points determine a line has been referred to several
times. It can be shown that any line in a plane can be described by an equation
of the form ax + by + c = 0, where a, b, and c are constants. This is called
the general form of the equation. It can also be shown that all nonvertical lines
can be described by an equation of the form y = mx + k, where m is the slope
as defined earlier and k is the value of y when x = 0, that is, the value of y
 Coordinate Geometry 83

Figure 7-4

where the line crosses the y-axis, called the y-intercept, as shown in Fig. 7-4.
'I'he form y = mx + k is called the slope-intercept form.
If we are given two points, we can find m in y = mx + k. Then we can
get a value for k by solving y = mx + k for k to get k = y - mx. With a
value for m and a value for k, we can write the equation of the line in slope-
intercept form by using program LINEl.

LINEI

94 REM * THIS PROGRAM PR0DUCES THE EQUATI0N 0F A LINE

95 REM IN SL0PE-INTERCEPT FORM GIVEN COORDINATES FOR
96 REM TWO P0INTS 0N THE LINE· THE PR0GRAM USES X
97 REM A CONSTANT F0R VERTICAL LINES
100 PRINT "EOUATHJN OF A STRAIGHT LINE IN"
110 PRINT "SL0PE-INTERCEPT FORM GIVEN TWO POINTS"
120 READ Xl1Yl1 X2,Y2
130 IF XI = .QOI THEN 250
140 PRINT
150 PRINT "C"J XIJ ","; YlJ ">, C"J X2J ","J Y2J ">"
152
154 REM * TEST FOR DEFINED SLOPE
-160 IF X2-XI <> 0 THEN 190
170 PRINT "EQUATHJN IS: X ="J XI
ISO G0T0 120
182
184 REM * CALCULATE SLOPE AND INTERCEPT
185 REM THEN PRINT SOLUTI0N EGUATI0N
190 LET M = CY2-Yll/CX2-XI>
200 LET K = YI - M*XI
210 PRINT "EQUATION IS: Y ="J MJ "*X+C"J KJ ">"
220 G0T0 120
222
224 REM
230 DATA 3,4, S,6, -1,6,
240 DATA .001.0, o.. o
250 END
RUN
LINEI

EQUATI0N 0F A STRAIGHT LINE IN

SL0PE-INTERCEPT FORM GIVEN TWO P0 INTS
C3 ,4 ),(5 ' 6
EQUATI0N IS: Y = I *X+C I
C-1 • 6 >.C-1 • 3
EQUATI0N IS: X =-I
84 Advanced BASIC

Summary of Sec. 7-2

For two points we have found the coordinates of the midpoint of the seg-
ment they determine. We have found the distance, the slope, and the equation
of the line in slope-intercept form.
Problems for Sec. 7-2
1) Modify program SLC/:>PE to accommodate data for which the slope is
undefined.
2) Write a program to calculate the coordinates of the midpoint for pairs
of points.
3) Modify program LINEl to treat a horizontal line as a special case.
4) Modify program DISTl or DIST2 to give the distance in simplified
radical form.
5) Modify LINEl to give the slope as a decimal, integer, or fraction
reduced to lowest terms as appropriate.
6) For an equation in slope-intercept form, have the computer give a
table of (x,y) values suitable for graphing.
7) For an equation in general form, have the computer give a table of
(x,y) values suitable for graphing.
8) For a set of n points, write a program to give equations for all pos·
sible pairs of points. Enter the n points as DATA and store them in
a 2 by n array or in a pair of lists.
9) For problem 8, have the computer eliminate duplicate points and
therefore duplicate lines in the output.
10) Write a program to write the equation of a line, given two points, in
ax + by + c = 0 form. Treat vertical and horizontal lines as special
cases.
11) Modify program LINEl to find the slope and y-intercept as fractions
reduced to lowest terms, if appropriate.
7-3 Parallel and Perpendicular Lines
What happens when we begin to think about more than one line in a plane?
Lines either intersect or they don't. If they don't, then the lines are parallel,
in which case their slopes are equal unless the lines are vertical, making both
the slopes undefined. If the lines intersect, then they might be perpendicular.
What about the situation for perpendicular lines? Consider Fig. 7 -5.
Note that the slope of 11 is positive and the slope of 12 is negative. Thus,
d1
(7-1)
t
and
dz
m2 = -- (7-2)
t
1
Since 11 and 1z are perpendicular, right triangles ABC and AB C are similar and

(7-3)
 Coordinate Geometry 85

Figure 7-5

because ratios of lengths of corresponding sides of similar triangles are equal.

Since

= - !!:2 (7-4)
t
we get
t 1
(7-5)

Substituting from (7-1) and (7-5) above into (7-3), we get

1
m1 =-- (7-6)
m2
which gives

(7-7)
That is, for perpendicular lines having slopes m 1 and m2 , the product of
the slopes is -1, or the slopes are negative reciprocals of each other.
We can use the parallel and perpendicular properties for a variety of prob·
!ems. We can test lines to see if they are either parallel or perpendicular. We can
find the equation of a line through a fixed point parallel or perpendicular to
another line. We can write the equation of a line which is the perpendicular
bisector of a segment.
We shall now present a program to give an equation for the line through a
given point perpendicular to the line determined by a pair of given points. There
are many conditions that we must account for. Suppose the given points deter-
mine a vertical or a horizontal line? We might even be given the same point
86 Advanced BASIC

Start
PERP

(_EN_D)

READ X1, Y1, X2, Y2

The points are ( X1, Y1)
and(X2,Y2)

Begin G¢sus
equation equation
subroutine subroutine

Yes Yes

No

Yes Calculate slope

and y-intercept
of solution line

PRINT PRINT Print the

"x="; XO ''y=";YO equation in
Solution line Solution line slope- intercept
is vertical is horizontal form

Figure 7-6 Flowchart for finding the equation of a line given one point on
the line and two points determining a line perpendicular to it.
 Coordinate Geometry 87

twice. A flowchart should be helpful in organizing these conditions. See Fig.

7-6 and program PERP.
PERP

94 REM * THIS PROGRAM ATTEMPTS T0 WRITE AN EQUATION

95 REM r0R A STRAIGHT LINE CONTAINING THE GIVEN POINT
96 REM cxo.YO) AND PERPENDICULAR TO THE LINE
97 REM DETERMINED BY THE GIVEN POINTS ext.YI)
98 REM AND CX2,Y2).
100 READ xo,yo
110 Ir XO = .001 THEN 800
120 READ x1,y1, x2,y2
130 PRINT
140 PRINT "LINE THROUGH C"1 XD1 "•"1 YD1 ">"
150 PRINT "PERPENDICULAR HJ THE LINE THR0UGH POINTS"
160 PRINT "<"J XlJ ","J YlJ ")11 C"J X2J ","J Y2J ")"
170 PRINT EQUATION IS: "1
180 GD SUB 500
190 GDT0 I DO
492
494 REM * SUBROUTINE T0 DETERMINE AND PRINT EQUATION
500 LET V Y2-YI
510 LET H X2-XI
520 IF" V <> 0 THEN 560
530 ff H <> 0 THEN 590
532
534 REM * THE TW0 GIVEN POINTS COINCIDE
540 PRINT "N0T UN I QUE"
550 RETURN
560 Ir H <> 0 THEN 610
562
564 REM * SOLUTION LINE IS H0RIZ0NTAL
570 PRINT "Y ="1 YO
580 RETURN
582
584 REM * SOLUTION LINE IS VERTICAL
590 PRINT "X ="1 XO
600 RETURN
602
604 REM *DEFINED NON-ZERO SLOPE
610 LET M = V/H
620 LET MO= -1/M
630 LET K = YO - MO*XD
640 PRINT "Y ="J MOJ "*X+C"J KJ ")"
650 RETURN
652
704 REM
710 DATA 1,2, 41111141?11 4, 3, -1.9,4,5
720 DATA 5, ~2, 6,4,6,-8 .. 3, g, 18,5,-11,5
730 DATA .001, 0
800 END

RUN
PERP

LINE THROUGH C I , 2 >

PERPENDICULAR TO THE LINE THROUGH POINTS
(4 ,7 ),(4 ,7)
EQUATION IS: NOT IJNIOIJE

LINE THROUGH C 4 , 3
PERPEND I CUL AR Tel THE LINE THROUGH POINTS
C-1 ,9 ),( 4 , 5
EQUATION IS: Y 1.25 *X+C-2

LINE THROUGH C 5 ,-2 )

PERPENDICULAR TO THE LINE THROUGH POINTS
6 ,4 ),(6 ,-8)
EQUATION IS: Y =-2
88 Advanced BASIC

LINE THROUGH < 3 , B >

PERPENDICULAR TD THE LINE THROUGH POINTS
(18 ,5 ),(-11 ,5)
EQUATION IS: X = 3

Summary of Sec. 7-3

We have examined pairs of lines in a plane. If they are parallel and non-
vertical, their slopes are equal. If two lines are perpendicular and neither is
ve1ticai, the piu<luct or their slopes is -1.

Problems for Sec. 7-3

1) Change program PERP to write the equation of the new line parallel
to the line determined by the given pair of points.
2) Write a program to give the equation of the perpendicular bisector of a
segment, given its endpoints.
3) For two pairs of points entered as DATA, have the computer deter-
mine if the specified lines are parallel or perpendicular. Be sure to
permit vertical lines.
4) For two lines given in ax + by + c = 0 form, have the computer
determine whether the lines are parallel, perpendicular, or neither.
7-4 Point of Intersection of Two Lines
We can define two lines to work with in several ways. One is to give a pair
of points to define each line. Another is to give an equation in the form y =
mx + k by specifying m and k for each line. Yet another is to give equations in
the form ax + by + c = 0 by specifying (a,b,c) for each line. We may become
more familiar with the general form by working with data for equations in the
form (a,b,c).
Given two lines,
0 (7-8)
and
(7-9)
and the need to find the point of intersection, we must find a coordinate pair
(x,y) that fits both equations. That is, we have to solve the equations simultan-
eously. Since the computer does not "do algebra," we will have to.
Rearranging the above equations, we get
(7-10)

(7-11)
Multiplying Eq. (7-10) through by-a?. and Eq. (7-11) by a, gives
(7-12)
(7 -13)
Adding Eqs. (7-12) and (7-13) produces
 Coordinate Geometry 89

(7-14)
Factoring, we get
y(a 1 b 2 - a2b 1) = a 2c 1 - a 1c 2

Dividing both sides by a 1 b 2 - a2 b 1 leaves

ll2C1 - ll1C2
y = (7-16)
a1 b2 - a2 b 1
Following a similar procedure to solve for x, we get
b1C2 - b2CJ
x = (7-17)
b2 a1 - b 1 a2

Let's write a program to use the above results to find the point of inter-
section, if it exists. Note that in Eqs. (7-16) and (7-17) the denominators are
equal. So we may use the BASIC statement,
180 LET D = Al*B2 - A2*Bl
This will allow us to divide the numerators from Eqs. (7-16) and (7-17) by D to
obtain the values for y and x, respectively, unless D happens to equal zero. This
situation would require division by zero, which means that the value is either
indeterminant or undefined depending on whether the numerator is zero or not.
Thus, if the value of D is not zero, we have a unique intersection, which we may
calculate using Eqs. (7 ·16) and (7-17). If D equals zero, we may obtain more
information about the two lines as outlined in the following discussion.
D = 0 means the following:
(7 -18)
and therefore
(7-19)

which may be written

~ = ll2
(7-20)
b1 b2

To see the significance of this, let's look at the original equations, (7-8)
and (7-9). They are a 1 x + b 1y + c 1 = 0 and a 2 x + b 2 y + c 2 = 0. Solving
each for y we get:

y -ll1X
- - C1
(7-21)
b1 b1

and
ll2X C2
y
bz b2
90 Advanced BASIC

Now the equations are in y mx + k form. Notice that the slope of line 1 in
Eq. (7-21) is -a 1 /b 1 and of line 2 in Eq (7-22) is -a 2 /b 2 unless bi or b2 is zero.
Clearly by multiplying both sides of Eq. (7--20) by -1 we get

- !:.!._. = - ~ (7-23)
b1 b2
Since the two sides of Eq. (7-23) are the slopes of the lines in Eqs. (7-21) and
(7 -22), that makes line 1 parallel to line 2. To sum up, if D = 0, then the lines
are paraiiei and there is no soiution uniess bi or b2 is zero.
T'Jov~' :;uppose that either b 1 or b 2 is zero when D ~ 0. Looking at Eq.
(7-19), if bi = 0 then either a i or b2 is also zero. Now, if Eq. (7-8) is not
meaningless, that means that b 2 must also be zero. If bi and b 2 are both zero,
we get
C1
aix + Ci = 0 or x = (7 -24)
ai

C2
ll2X + C2 = 0 or x = (7-25)
a2
in which case both lines are vertical because x is constant. If what we are look-
ing for is the point of intersection, there won't be any unless the two equations
define the same line, in which case all points on one line will also lie on the other.
We really have three levels of information to test for. First we want to
know if the lines intersect in a single point. If they do intersect in a single point,
then we want the coordinates of that point. If they don't intersect in a single
point, then it is useful to know whether or not they are vertical, because if one
is, they both are. And finally, we can determine whether or not they coincide.
If they coincide, the solution is indeterminan t. If they do not coincide and
are parallel, we say the equations are inconsistent. A flowchart should help to
sort out all of the above considerations. See Fig. 7-7 and program Pq'.>INT.

Start
P0JNT

END~

Figure 7-7 Flowchart for program POINT (cont'd on next page)

D= Al*B2-A2*81
Calculate the value of
the denominator to be
used later

N1 =A2*C1 -A1*C2
N2= Bl H2 - B2*C1
X =N2/D
Y =Ni/D
calculate solution

PRINT "S0LUTION IS:

xi II' II; Y; 11)11
( 111
output coordinates
of solution point
92 Advanced BASIC

POINT

94 REM * THIS PR0GRAM rINDS THE P0INT 0r INTERSECTION

95 REM r0R TW0 LINES GIVEN IN AX+BY+C=O r0RM· HANDLES
96 REM INDETERMINANT AND INC0NSI STENT CASES.
100 READ T
110 Ir T = 0 THEN 500
I 12
114 REM * READ C0Err!CIENTS r0R B0Tll LINES
120 READ AJ,Bl,CI• A2,B2.C2
130 PRINT
132

140 PRINT TABCIO>J "LINES"

i50 ?R!rrr
160 PRINT TABCIO>J "AND"
170 PRINT A2J "*X+C"J B2J ">*Y+C"J C2J 0
) O"
172
174 REM * IF Al*B2-A2*BI <> 0 THEN THERE IS
175 REM A UNIQUE S0LUTI0N
180 LET D Al*B2 - A2*BI
190 Ir D <> 0 THEN 300
192
194 REM* WE C0ULD TEST F0R B2 0 WITH THE SAME RESULT
200 Ir Bl = 0 THEN 240
210 PRINT "N0N-VERTICAL "J
220 Ir Cl/Bl = C21B2 THEN 280
230 G0T0 260
240 PRINT "VERTICAL "J
250 Ir Cl/Al = C21A2 THEN 280
260 PRINT "PARALLEL"
270 G0T0 100
280 PRINT "C0INCIDENT"
290 G0T0 JOO
292
294 REM * S0LUTION EXISTS - CALCULATE AND PRINT IT
300 LET NI = A2*CI - Al*C2
310 LET N2 = Bl*C2 - B2*CI
320 LET X = N21D
330 LET Y = NllD
340 PRINT "S0LUTI0N ISr C"J XJ "•"J YJ ")"
350 G0T0 I 00
352
394 REM
400 DATA J, 1 .. l 1 1.. 2,2,2
410 DATA J, 1 .. 2 .. 3, 4,5,6
420 DATA 1. 3,, o.. 4,, 7,, o.. -8
430 DATA 1, S.11 3, to, 10,6,-3
440 DATA 0
500 END

RUN
PO INT

LINES
*X+( l l*Y+( 0
AND
2 *X+( 2 l*Y+( 2 0
NUN-VERT!~AL ~O!N~!UENT

LINES
*X+ ( 2 l*Y+ < 3 0
AND
4 *X+( 5 l*Y+< 6 0
S0LUTI0N IS: C I ,-2
 Coordinate Geometry 93

LINES
3 *X+C 0 l*Y+< 4 0
AND
7 *X+C 0 l*Y+C-8 0
VERTICAL PARALLEL

LINES
5 *X+C 3 l*Y+C 10 0
AND
10 *X+C 6 >*Y+C-3 0
N0N-VERT!CAL PARALLEL
;

Summary of Sec. 7-4

We have found the intersection of two lines for which the equations are
given in standard form. Consideration has been given to the special cases of
parallel and coincident lines.
Problems for Sec. 7-4
1) Write a program to find the intersection of two lines with equations
given in slope-intercept form.
2) Write a program to find the intersection of two lines given two points
on each line.
3) Write a program to find the distance from a point to a line, given the
equation of the line in ax + by + c = 0 form. You may use
lax + by + cl
d =
va2 + b2

4) Write a program to find the distance from a point to a line, given two
points on the line.
7-5 Three Points in a Plane
If we are given two points, we know that they lie on a straight line, but if
we are given three points, they may or may not lie on a straight line, that is, they
may or may not be collinear. If they are not collinear, there are many questions
we may ask.
Let us look at three points, A(x 1 , y 1 ), B(x 2 , y 2 ) and C(x 3 , y 3 ), to deter-
mine if they are collinear. Suppose for the moment that if they are, the line
containing them is not vertical. One way to see if they are collinear is to look
at the slope of AB and the slope of BC. If the two slopes are equal, then the
points are on a single line.
Now suppose that we select points (1, 2), (1, 4) and (2, 3) as DATA.
Clearly the slope ciµculation will produce an undefined condition due to an
attempt to divide by zero. We need to check for equal x-coordinates. We need
onlycheckx 1 = x 2 andx 2 = x 3 • lfx 1 = x 2 ,whataboutx 1 - x 2 ? Notice
that x 1 - x 2 would be zero. Now we see that if (x 1 - x 2 )(x 2 - x 3 ) = 0,
then either AB or BC, or both, are vertical segments. Or, to put it another way,
if (x 1 - x 2 )(x 2 - x 3 ) is not zero, then neither AB nor BC is vertical, and it is
safe to instruct the computer to calculate the slopes of both segments. See
especially line 190 in program C(,t)LIN.
94 Advanced BASIC

C0LIN

94 REM *
THIS PR0GRAM DETERMINES WHETHER 0R N0T
95 REM THREE P0INTS IN A PLANE ARE C0LLINEAR.
96 REM IT ALL0WS F0R P0INTS IN A VERTICAL LINE.
100 READ T
110 Ir T = 0 THEN 900
120 PRINT
130 READ x1.v1. x2.v2. X3.Y3
140 PRINT "C"J XIJ "• 1 'J YIJ ">•"J
150 PRINT "C"J X2J "• .. , Y2J "> AND 11 J
160 PRINT ''C''J X3J ''•''J YJJ ''>''
t llJ i.F.i IJ'I

IBO LET D2 = X2•X3

184 REM *
Ir Dl*D2 N0N·ZER0 THEN NEITHER
185 REM SEGMENT IS VERTICAL
-·190 Ir Dl*D2 <> 0 THEN 300
200 Ir DI <> 0 THEN 400
210 Ir D2 <> 0 THEN 400
212
214 REM *
AT THIS P0INT B0TH SEGMENTS ARE VERTICAL
220 PRINT "VERTICALLY C0LLINEAR"
230 G0T0 I 00
292
294 REM * CALCULATE 80TH SL0PES
300 LET Ml =
<Yl-Y2)/DI
310 LET M2 =
<Y2-Y3)/D2
312
314 REM * Ir SL0PES N0T EQUAL THEN P0INTS N0N-C0LLINEAR
320 Ir Ml <> M2 THEN 400
330 PRINT "C0LLINEAR"
340 G0T0 JOO
400 PRINT "N0N-C01.LINEAR"
410 G0T0 100
492
494 REM
500 DATA 1. 6 .. s .. 0,1, -9.10
510 DATA 1. 1,,2 .. 3,4,, 5,7
520 DATA 1. 1 .. 2. 1 .. 4 .. 2,3
530 DATA 1. 5, 2. 5, -3, 5,50
540 DATA 0
900 END
RUN
CeJLIN

), ( 0 AND (-9
( 6
C0LL !NEAR
' 5 ' 7 J ' 10

( I , 2 ), ( 3 , 4 > AND ( 5 , 7
N0N-C0LLINEAR
( I , 2
N0N-C0LLINEAR
) .
( . 4 AND ( 2 , 3

( 5 , 2 ), ( 5 .-3 ) AND ( 5 , 50
VERTICALLY C0LLINEAR

Triangles (The Noncol!inear Case)

It should be clear that if three points in a plane are noncollinear, they are

properties to consider. To name only a few possibilities, we can calculate the

area using Hero's formula; we can test to see if the triangle is acute, right, or
obtuse; we can find the centroid where the medians intersect; we can find the
orthocenter where the altitudes intersect; and we can find the lengths of the
altitudes.
 Coordinate Geometry 95

Suppose we first look for a right triangle. There are at least two proce-
dures we might use. One is to test the lengths of the three sides using the
Pythagorean Theorem, c 2 = a 2 + b 2 • Another is to test the slope of each pair
of sides to see if their product is -1 or if one side is vertical and another is hori-
zontal. Either procedure will be instructive. Let's first investigate Pythagoras
with the aid of Fig. 7-8.

Figure 7-8

When we consider c 2 = a2 + b 2 we assume that the hypotenuse has

length c. Suppose we find that a is the length of the hypotenuse or that b is.
The only safe bet is to test them all. This could be done by putting the follow-
ing in an IF-THEN statement to test if AB is the hypotenuse for points A(x1,
y 1), B(x2 , Y2) and C(x 3 , y 3 ):
(Xl - X2)t2 + (Yl - Y2)t2 = ((X2 X3)t2 + (Y2 - Y3)t2)
+ ((X3 - Xl)t2 + (Y3 - Yl)t2)
and by then putting in two similar statements to see if BC or CA might be the
hypotenuse. Such a procedure ought to work, but this statement is excessively
long and cumbersome and we aren't really using the power of the computer to
do repeated operations. It seems worthwhile to devise a way of storing the co-
ordinates so that we can use computer loops to move around the triangle instead
of writing new statements for each side being considered. That seems to suggest
subscripts and lists. So instead of Xl, X2, and X3 we are going to use X(l),
X(2), and X(3). And for Yl, Y2, and Y3 we use Y(l), Y(2), and Y(3), as shown
in Fig. 7-9.
Often when trying a new procedure on a new program, it is a good idea to
do the work piecemeal. If we put a lot of new things into a program simulta-
neously, errors are harder to find than if we go in steps. Consider program
PYTHl for simply reading coordinates in and printing them out.
96 Advanced BASIC

A(X(I), Y(1))

'""'~\ C(X(3), Y(3))

Figure 7·9

PYTHI

94 REM *THIS IS A TRIAL PR0GRAM T0 SIMPLY READ

95 REM C00RDINATES r0R THE VERTICES 0r A TRIANGLE
96 REM AND PRINT THEM 0UT AGAIN USING AN X l.I ST
97 REM r0R X C00RDINATES AND A Y LIST r0R Y
98 REM C00RDINATES
100 DIM XC3), YC3)
110 READ T
120 Ir T = 0 THEN 900
130 PRINT
132
134 REM *READ C00RDINATES IN LINES 140 T0 160
140 r0R P I T0 3
- ( 150 READ XCP>. YCP>
160 NEXT P
162
164 REM PRINT C00RDINATE:S IN LINES 170
* T0 190

-{180
170
190
PRINT .. ("J
PRINT ("J
PRINT Jt (II J
fl
XC 1 >1 ", "J YC 1 >J ">, "J
X<2>J ". "J Y<2>J ") AND
X< 3)J "• .. J y ( 3) J ")"
.,
200 G0T0 110
792
794 REM
800 DATA 1. 1 .. 2, 5.--111 6. 15
810 DATA 1. o.. o.. o.. 3.o 4,0
820 DATA 0
900 END
RUN
PYTHI

< 1 , 2 ), ( 5 ' - I > AND C 6 , 15

( 0 , 0 ), ( 0 , 3 > AND < 4 , 0

To consider each side in turn as a possible hypotenuse, we want to work

our way around the triangle. First we will use the distance from point 1 to point
2, then the distance from point 2 to point 3, and finally the distance from point
3 to point 1. See the columns labeled "Hypotenuse" in Table 7-1.
 Coordinate Geometry 97

TABLE 7-1

Hypotenuse 1st Leg 2nd Leg

Segment Points Segment Points Segment Points

AB 1,2 BC 2,3 CA 3,1

BC 2,3 CA 3,1 AB 1,2
CA 3,1 AB 1,2 BC 2,3

Looking at Table 7-1, we see a progression across, from points 1 and 2 to

points 2 and 3 and then to points 3 and 1 for hypotenuse AB. For hypotenuse
BC, we go from points 2 and 3 to points 3 and 1 and then to points 1 and 2.
This pattern is completed by considering CA as the hypotenuse. If we can sim-
ply go 1-2-3-1-2-3, we can get all of the pairs we need using loops and subscripts.
Consider the following four lines of BASIC code:

200 I T0 3
F"0 R P "
210 LET X(P+3l X<P>
220 LET Y<P+3l = Y<P>
230 NEXT P

These four lines have the effect of copying the three points 1, 2, and 3
into locations 4, 5, and 6, respectively, of the X and Y lists. This will enable us
to organize the testing of each side in turn for being the hypotenuse of triangle
ABC, as shown in TABLE 7-1. See lines 200 to 230 in PYTH2.

PYTH2
94 REM *THIS PR0GRAM STORES THE C0RDINATES 0F THE
95 REM VERTICES 0F A TRIANGLE IN THE FIRST THREE
96 REM ELEMENTS 0F X AND Y LISTS AND DUPLICATED
97 REM IN THE NEXT THREE ELEMENTS 0F THOSE LISTS
98 REM T0 FACILITATE USE OF SUBSCRIPTS AND L00PS
99 REM T0 TEST PR0PERTIES OF THE TRIANGLES.
100 DIM X<6» YC6l
110 READ T
120 IF T = 0 THEN 900
130 PRINT
132
134 REM * READ COORDINATES IN LINES 140 TO 160
140 FOR P 1 T0 3
150 READ XCP» Y<P>
160 NEXT P
162
164 REM * PRINT C00RDINATES IN LINES 170 TO 190
170 PRINT "( 0 J X(J)J ",,"; YCl)J "),,"J
IBO PRINT "C"J XC2)J "•"J YC2)J "> AND "J
0
190 PRINT "C"J XC3)J ",,"; YC3)J )"

192
194 REM *DUPLICATI0N TAKES PLACE IN LINES 200 TO 230
200 F0R P = 1 TO 3
210 LET XCP+3> XCP>
- { 220 LET Y<P+3> = YCP>
230 NEXT P
232
234 REM * WE PRINT THE C0NTENTS 0F THE X AND Y LISTS
235 REM F0R DEMONSTRATION PURP0SES ONLY
98 Advanced BASIC

240 PRINT "VALUES OF X L00K LIKE THIS IN THE X LI ST"

250 F0R P = I TO 6
260 PRINT X<P>J
270 NEXT P
2BO PRINT
2B2
290 PRINT "VALUES 0F Y L00K LIKE THIS IN THE Y LIST"
300 F0R P = I T0 6
310 PRINT YCP>J
320 NEXT P
330 PRINT
332
,...,..'T" ....
J~V UU IV • •v
792
l';JLJ, Kt.M
BOO DATA 1. 1, 2 .. SJI' -1, 6. 15
BIO DATA 1. o, o, o, 3, 4, 0
B20 DATA 0
900 END
RUN
PYTH2

( I 2 .
»( 5 >- I ) AND ( 6
VALUES OF x L00K LIKE THIS IN THE x LIST
. 15

I 5 6 I 5 6
VALUES 0F Y L00K LIKE THIS IN THE Y LIST
2 -1 15 2 -1 15
( 0
VALUES
. 0 »( 0 3 ) AND ( 4.
0F X L00K LIKE THIS IN THE x LIST
. 0

0 0 4 0 0 4
VALUES 0F Y L00K LIKE THIS IN THE y LI ST
0 3 0 0 3 0

Now all that remains is to calculate the lengths of the sides of the triangle
and to store them in such a way that we may test in sequence around the tri-
angle, trying each side as possible hypotenuse. This is accomplished by lines 300
to 350 in PYTH3. Note that line 340 does for the lengths of sides what 210 and
220 do for the coordinates of the vertices.

PYTH3

94 REM *
THIS PROGRAM DETERMINES FROM THE COORDINATES
95 REM 0F THE VERTICES OF A TRIANGLE WHETHER OR
96 REM NOT IT IS A RIGHT TRIANGLE
100 DIM XC6), YC6), DC6)
110 READ T
120 IF T = 0 THEN 900
130 PRINT
132
134 REM *
READ COORDINATES IN LINES 140 TO 160
140 FOR P I TO 3
150 READ X<P>, YCPl
160 NEXT P
162
164 REM PRINT COORDINATES IN LINES 170 TO 190
170 *
PRINT "("J X< 1 )J .. , .. ,; YC 1 ); "),"J
18'0 PR!~IT "("_l 'x:('?)! ": .. ! Y'(2~! ") /\MD "!
190 PRINT ·•c"1 XC3>J "•"J YC3lJ ")"
192
194 REM *
DUPLICAT!0N TAKES PLACE IN LINES 200 T0 230
200 F0R P = I TO 3
210 LET XCP+3) XCP>
220 LET YCP+3> = YCP)
230 NEXT P
232
 Coordinate Geometry 99

294 REM *
ENTER LENGTHS IN THE D LIST IN THE
295 REM SAME F'CiJRMAT AS USED F0R COORDINATES IN
296 REM THE X AND Y LISTSJ THAT IS - DUPLICATED
300 F0R J = I TO 3
310 LET A =
X<Jl·XCJ+ll
320 LET B =
YCJl-YCJ+ll
- 330 LET DCJl = SQRC At2+Bt2 l
{ 340 LET DCJ+3l =DCJl
350 NEXT J
352
394 REM *
N0W TEST EACH SIDE FOR BEING A HYPOTENUSE
400 FOR P =
I TO 3
410 IF DCPlt2 DCP+llt2 + DCP+2l•2 THEN 500
420 NEXT P
422
424 REM *IF THE TEST FAILS FOR ALL THREE SIDES THEN
425 REM WE HAVE A NON-RIGHT TRIANGLE
430 PRINT "N0T A RIGHT TRIANGLE"
440 GOTO 110
492
500 PRINT "RIGHT TRIANGLE WITH HYPOTENUSE DETERMINED BY"
510 PRINT "P01NTS: ("J XCP)J ","; YCPlJ ") AND "J
520 PRINT "C"J XCP+l)J ","J Y<P+llJ ")"
530 GOTO 110
'792
794 REM
BOO DATA 1, 1, 2, 5,-1, 6, 15
BIO DATA 1, o, Q, o, 3, 4,0
B20 DATA 0
900 END
RUN
PYTH3

( I .2 ), ( 5
N0T A RIGHT TRIANGLE
,- I ) AND ( 6 . 15

( 0 • 0 ), < 0 , 3 ) AND < 4 , 0

RIGHT TRIANGLE WITH HYPOTENUSE DETERMINED BY
POINTS: < 0 , 3 ) AND < 4 , 0

Summary of Sec. 7-5

We have examined sets of three points for collinearity. We have looked at
the triangle formed by noncollinear points to determine whether or not it is a
right triangle. In order to do this, we have devised a technique of storing co-
ordinates in lists to take advantage of repeatable program statements using
loops.
Problems for Sec. 7-5
1) In program C(,l>LIN, provide for treating points on a horizontal line as
a special case.
2) Write a program to find the intersection of two altitudes of a triangle.
3) Write a program to find the intersection of two medians of a triangle.
4) Modify program PYTH3 to classify triangles as acute, equiangular,
right, or obtuse.
5) Have the computer print all Pythagorean Triples for a hypotenuse less
than 101. Pay particular attention to efficiency here.
6) Write a program to determine if four points are collinear.
7) Write a program to determine if n points are collinear.
100 Advanced BASIC

8) Write a program to compute the area of a triangle given the co-

ordinates of the three vertices and using Hero's formula:
A = Vs(s - a)(s - b)(s - c) where s = (a + b + c)/2
9) Write a program to calculate the area of a triangle by finding the
length of an altitude using the methods of Sec. 7-4.
10) Write a program to test triangles for right triangles by determining the
slope of all three sides to see if tht: pre.duct is -1 for any pail". You
will have to treat vertical lines as a special case.
11) Given three noncollinear points, find the center and radius of the
circle they determine.
 8
POLYNOMIA LS

8-1 Introduction
A real polynomial in x can be defined as any expression that can be
written in the following form:
2 + a1X + ao
anxn + an_ 1xn-l + · · · + a2x

For a given value of x, n is a non-negative integer and the a's are constant
values. Examples of polynomials are
2x; 3x 2 + 2; 4; 5x 11 + 3x 3 - 2
The highest exponent of x, called n, is the degree of the polynomial.
Polynomials are of interest to the mathematician and scientist alike.
The distance an object travels in a gravitational field is described by a second
degree polynomial, for example. We will now explore some ways in which
the computer may be programmed to evaluate polynomials and make some
comparisons.
Regardless of the final procedure to be used to evaluate a polynomial,
we will probably want to store the coefficients in a list. If you have zero sub-
scripts available, this is a perfect fit because the subscript of the coefficient is
equal to the exponent on x in each term. If you don't use zero subscripts,
4
simply adjust the variable used for exponents. We arbitrarily select 3x +
3 2
2x - x + 5x + 3 for our first example. The degree of the example is
four, and hence there are five coefficients. The coefficients and the degree of
the polynomial may be entered as data.
One procedure for evaluating a polynomial is to use a BASIC language
function such as

100 DEF FNPCXl = PCSl*Xt4+PC4l*Xt3+PC3l*Xt2+PC2 l*X+PCI>

This defined function will work fine for any fourth degree polynomial. For a
101
102 Advanced BASIC

degree less than four we could fill in with zeros, but for degrees more than
four, we would have to retype the line. We are l!lso limited by line length
when DEFining a function on some computer systems. Others allow multiple
line DEFinitions. So while the DEF capability is very useful in some situa-
tions, we might give some thought to developing another procedure for evalu-
ating a polynomial.
Note that the exponent of x is one less than the subscript of the list co-
efficient in each term of the DEF statement. (They are the same if you use
zero subscripts.) Even p(2)*x can be vvTitte11 p(2)~.xtl and p(l) ca11 he writ-
hm .n(l)*x to (unless x = 0). This suggests the fd!c~·!l!!g !ccp:

300 LET P2 = PC!>

310 FOR I = N+I TO 2 STEP -I
320 LET P2 = P2 + PCil•XtCI·l>
330 NEXT I

This four line routine has the desirable feature that it requires no change for
handling polynomials of differing degree.
Suppose we give some thought to the number of operations we are in-
structing the computer to perform. Look at line 320. There we can see an
addition and a multiplication and an exponentiation. For positive integers as
exponents, exponentiation amounts to successive multiplication. That means
for an exponent of I - 1 there will be I - 2 multiplications plus the multi-
plication of the coefficient. That makes I - 1 multiplications for that line.
If we take another look at 3x 4 + 2x 3 - x 2 + 4x + 3 and count the number
of operations, we get four additions and 10 multiplications for our fourth de-
gree polynomial (not accounting for how the computer might handle the ex-
ponent one as a special case). For a 10th degree polynomial there would be
10 additions and 55 multiplications.
Programmers often try to improve the efficiency of their programs because
of the limitations of the computer and the size of the job to be done or purely
for the challenge involved. Notice that in any polynomial, beginning with the
second term, each term has x as a factor one less time than the previous term.
This suggests some sort of successive factoring, as shown by the following:
3x4 + 2x3 (3x + 2)x 3
(3x + 2)x 3 - x2 ((3x + 2)x - l)x 2
((3x + 2)x - l)x 2 + 4x (((3x + 2)x - l)x + 4)x
(((3x + 2)x - l)x + 4)x + 3 (((3x + 2)x - l)x + 4)x + 3
Thus our original polynomial 3x 4 3
+ 2x - x 2 + 4x + 3 can be written in the
following form:
(((3x + 2)x - l)x + 4)x + 3
This is called nested form.
Now there are only four multiplications and four additions called for in a
fourth degree polynomial. Of course, a saving of six multiplications is not much
to a computer, but the saving accumulates as the degree of the polynomial in-
 Polynomials 103

creases and as the number of times we evaluate the polynomial increases. Con-
sider the following routine:

400 LET P3 = PCN+ l l

410 rOR I =N TO l STEP -1
420 LET P3 = P3*X + PC!l
430 NEXT I

We have another four-line routine that is more efficient than the first one.
We now have three procedures for evaluating polynomials. The first is a
defined function, the second is a loop which evaluates the polynomial using ex-
ponents term by term, and the third is a loop which uses nested form. Let's
assemble these three procedures into a program so that we can compare some
values. As we expected, we see that all values do check out by looking at the
output of program PCJ>LYOl.

POL YO l

94 REM * THIS PROGRAM EVALUATES POLYNOMIALS BY THREE

95 REM METHODS! DEr• TERM BY TERM AND NESTING
96
97 REM * SET UP DEr!NED rUNCT!ON
100 DEr rNPCX) =
PC5l*Xt4+P C4l*Xt3+PC 3l*X•2+PC2 l•X•PCll
102
104 REM * READ DATA
110 READ N
120 rOR I =
N+I TO I STEP -I
130 READ PC!>
140 NEXT I
142
144 REM * PRINT HEADINGS 11
150 PRINT "X",, "F'NPCX)", "EXP L00P 1 "NESTING"
160 rOR X =
I TO 10
192
194 REM * CALCULATE BY DEr
200 LET Pl =
rNPCX>
292
294 REM *
CALCULATE USING EXPONENTS IN A LOOP
295 REM TERM BY TERM.
300 LET P2 = PC l l
310 rOR I =
N+I TO 2 STEP -I
320 LET P2 =
P2 + PCil*XtCl- ll
330 NEXT I
392
394 REM *
CALCULATE USING NESTING
400 LET P3 =
PCN+ll
410 rOR I =
N TO l STEP -I
420 LET P3 =
P3•X + PC!>
430 NEXT I
492
494 REM *
PRINT RESULTS
500 PRINT X, Pl, P2, P3
510 NEXT X
512
514 REM
520 DATA 3,, 2 .. -1, 5,, 3
530 END
RUN
POL YO l

x fNPCXl EXP LOOP NESTING

l 12 12 12
2 73 73 73
3 306 306 306
4 903 903 903
 104 Advanced BASIC

5 2128 2128 2128

6 4317 4317 4317
7 7878 7878 7878
8 13291 13291 13291
9 21108 21108 21108
10 31953 31953 31953

Problems for Sec. 8-1

1) Write a program to tabulate the number of additions and multiplica-

with a one line DEF or with a loop which uses exponents.

2) Write a program to multiply two polynomials.
3) Write a program to add polynomials. Be sure to avoid leading zero co-
efficients.
4) Do problem 3 for subtraction.
5) Write a program to give ordered pairs (x,p(x)) suitable for graphing the
polynomial p(x ).
6) Write a program to find zeros of a second-degree polynomial using the
quadratic formula.
8-2 Finding a Real Zero of a Polynomial
Now that we are set up to evaluate a polynomial, we can explore some ap-
plications. One important consideration where polynomials are concerned is to
find values of x for which the value of the polynomial is equal to zero. If we
designate a polynomial as p(x) (read p of x) and set y equal to p(x), then we
may think in terms of polynomial equations and their graphs. The zeros of a
polynomial are the values of x where p(x) = 0. For real zeros they are the val-
ues of x where the graph of the polynomial equation crosses the x-axis. How-
ever, in the case of nonreal zeros, while the value of p(x) will be zero, the graph
of the polynomial equation in the x-y plane will not cross the x-axis. Of course,
we are familiar with special mathematical procedures for solving certain poly·
nomials. For instance, all second.degree polynomials may be solved by the
quadratic formula. We will look here at more general solutions.
It can be shown that every nth degree polynomial has exactly n complex
zeros. So one approach to finding zeros is to simply try values of x until n are
found which give p(x) = 0. Such a procedure is feasible only with the aid of a
computer because of the tedious calculations required. Even so, the approach
can be very complicated indeed. We will simplify our job for now if we begin by
looking only for real zeros. A most valuable aid to finding zeros of any contin-
uous function y = f(x) is the principle that if f(xi) and f(x 2 ) are of different
sign, then there are an odd number of zeros in the interval from x 1 to x 2 • For
real values of x it should be clear that if f(xi) > 0, the point (x 1 ,f(xi)) on the
graph is above the x-axis, whereas f(x 1 ) < 0 makes (x 2 , f(x 2 )) below the x-axis,
2.nd the g~r:.ph of u. couthiU.VU.S f;;.uctiVii iiiU.St CLOSS th& x~axis SUH!tfWht::it: i.u
between.
There is a variety of ways to isolate increments in which real zeros may be
found. Let us set up a procedure whereby we have control over where the com-
puter searches for intervals within which zeros may be found. We can store co-
efficients in a list, select limits within which to search, and have the computer
 Polynomials 105

step through the specified region, trying pairs of values for x to see if f (x 1 )*f (x 2 )
is positive. If it is not, then we have a zero between x 1 and x 2 or either x 1 or x 2
is a zero. If we always find f (x i)*f (x 2 ) positive, there are numerous possible
reasons. We may not have included the zeros in our limits, or we may have had
the computer search such large increments that the graph crosses the x-axis an
even number of times, or the graph may just touch the x~axis at a minimum or
maximum point and never cross the x-axis (the point at which it touches, how-
ever, is still a zero of the polynomial), or there are no real zeros. We must bear
all of these problems in mind as we construct and operate a program.
In program ZER<,l>Ol, line 120 reads the degree of the polynomial and line
140 reads the coefficients. Line 280 tests for sign change. The subroutine
which starts at line 800 evaluates the polynomial by nesting. Line 840 checks
for exact zeros. The general procedure here is just as valid for any continuous
function as it is for polynomial functions. Note that we could use the graphing
of Chapter 5 to assist us in finding regions where we might expect to find zeros.
ZEROOI

94 REM * THIS PROGRAM SEARCHES FOR A SIGN CHANGE

95 REM IN THE VALUE OF A POLYNOMIAL FUNCTION
100 DIM P< IS>
11p PRINT "POLYNOMIAL IS"J
--120 READ N
130 F0R I = N+I T0 I STEP -I
-140 READ P<I>
150 PRINT P<Ill
160 NEXT I
162
170 PRINT
180 PRINT
190 PRINT "FIRST, LAST, STEP"J
200 INPUT F, L• S
210 PRINT
220 LET XO = F
230 GOSUB 800
240 FOR X = F+S TO L STEP S
242
244 REM * SAVE THE VALUE OF THE FUNCTI0N AT THE
245 REM BEGINNING OF THE INTERVAL
250 LET YI = P3
260 LET XO = X
270 G0SUB 800
-280 IF Yl*P3 > 0 THEN 320
290 PRINT "SIGN CHANGE!"
300 PRINT "PC"J X-SJ tr)="J YIJ "AND PC"J XJ ")="J P3
310 STOP
320 NEXT X
322
330 PRINT "NO SIGN CHANGE F'CUND"
340 STOP
h2
'794 REM * LINES 800 TO 830 EVALUATE BY NESTING
-800 LET P3 = P<N+ I l
810 FOR I = N TO I STEP -I
R::!O LET P3 = P3*XO + P<Il
830 NFXT I
832
834 REM * CHECK FOR EXACT ZERO
-840 IF P3 <> 0 THEN 880
850 PRINT
860 PRINT XOJ "IS A ZERC"
870 STOP
880 RETURN
882
106 Advanced BASIC

884 REM
890 DATA 3, 1,2,3,4
900 END
RUN
ZE.ROOI

POLYNOMIAL IS I 2 3 4

FIRST, LAST, STEP? -IQ, IQ, I

SIGN CHANGE:
P(-2 l=-2 AMO P(-1 l= 2

Interval Halving
Now we have an interval in which we expect to find a zero. We would like
to make that interval smaller and smaller to get successively better approxima-
tions of a zero. We can simply evaluate the polynomial at the midpoint of the
interval found in program ZER</)01 and check for a change of sign against one
endpoint. If a change in sign is found, we bisect and repeat. If a change in sign
is not found, then the change must occur between the midpoint and the end-
point not tested above; therefore we bisect that interval and repeat. After each
bisection the interval is smaller and the midpoint is a better approximation of
the zero. This process is accomplished in lines 400 to 550 of program ZER</>02.
The process needs a stopping place. We need a test of the accuracy of the
current approximation that works for roots very close to zero and far away from
zero, a test, moreover, that is compatible with the precision of the computer.
We would like to get six significant digits in our results. We label the endpoints
of the interval x 1 and x 2 • If we test lx 1 - x 2 I against 10- 6 , we will not get six
significant digits for x 1 and x 2 when they are very close to zero. For x 1 and x 2
large in magnitude, say 100,000, we would require a machine with 12-digit pre-
cision. These problems are avoided by using relative error and testing lx 1 - x 2 I/
(lx 1 I + lx 2 I) against 10- 6 • See line 450 in program ZER</>02.

ZER002

9~ REM * THIS PROGRAM SEARCHES FCR APPROXIMATE ZEROS

95 RE.M IN THE INTERVALS FOUND IN PROGRAM ZEROOI
100 DIM PC15)
110 PRINT "POLYNOMIAL IS"J
120 READ N
130 FOR I = N+I TO I STEP -I
140 READ PC!)
150 PRINT PC!)J
160 NE.XT I
162
170 PRINT
180 PRINT
190 PRINT "FIRST, LAST, STEP"J
200 INPUT F, L• S
210 PRINT
220 LET XO = F
230 GO SUB 800
240 FOR X = F+S TO L STEP S
250 LET YI = P3
260 LE.T XO = X
270 GOSUB 800
280 IF Yl*P3 > O THEN 320
 Polynomials 107

290 PRINT "SIGN CHANGE:"

300 PRINT "P("J X-SJ ">="J YIJ "AND P<"J XJ ">="J P3
310 GOT0 400
320 NEXT X
322
330 PRINT "NO SIGN CHANGE FOUND"
340 ST0P
392
394 REM * LINES 400 THROUGH 550 CARRY OUT INTERVAL
395 REM HALVING. INTERVAL BOUNDARIES ARE XI AND X2
396 REM EXACT ZEROS ARE FOUND IN LINE 840
400 LET XI X-S
410 LET X2 = X
420 LET Y2 = P3
422
424 REM * HALVING TAKES PLACE IN LINE 430
430 LET XO = <Xl+X2l/2
440 GOSUB 800
450 IF ABS<Xl-X2l/<ABS<Xll+ABS<X2>> > IE-6 THEN 500
460 PRINT
470 PRINT "APPROXIMATE ZERO:"
480 PRINT "PC"J XOJ ">="J P3
490 ST0P
500 IF Yl*P3 > 0 THEN 530
510 LET X2 = XO
520 G0T0 420
5 30 LET XI = XO
540 LET YI = P3
550 G0T0 430
792
794 REM * EVALUATE BY NESTING
800 LET P3 = P<N+ I>
810 F0R I = N TO I STEP -I
820 LET P3 = P3*XO + P<I>
830 NEXT I
832
840 IF P3 <> 0 THEN 880
850 PRINT
860 PRINT XOJ "IS A ZERO"
870 ST0P
880 RETURN
882
884 REM
890 DATA
900 END
RUN
ZER002

POLYNOMIAL IS I 2 3 4

FIRST, LAST, STEP? -2,-1,.1

SIGN CHANGE:
PC-1.7 >=-0.233 AND PC-lo6 >= Q.224

APPR0XIMATE ZERO:
PC-1.65063 >=-4o7632E-6

Summary of Sec. 8-2

We have used the principle that if f(xi) and f(x 2 ) are of opposite signs,
there must be a value of x between x 1 and x 2 such that f(x) = 0 if we are to
find intervals within which zeros exist for polynomial functions. We have fur-
ther reduced interval size by successive halving to obtain approximate zeros. All
of the methods used here can easily be applied to other finite continuous func-
tions as well.
108 Advanced BASIC

Problems for Sec. 8-2

1) Find a zero for any or all of the following:
a) x 3 + 6x 2 - 49x + 66
b) x 4 + 2x 3 - 13x 2 - l4x + 24
c) x 3 + llx 2 - 68x - 672
d) 2x 3 + 3x 2 + 4x - 5
e) xs + 13x 4 - 37x 3 - 457x 2 + 492.x + 2340
f) x 3 + 2x 2 - 11x 12
g) XS + 4.8x 4 - 4.2x 3 - 29.6x 2 + 7.2.x + 28.8
2) Modify ZERQ)02 to search for aii reai zeros automatically after the val-
ues of F, L, and Shave been specified. Be sure to stop if all zeros have
been found without further testing values of x.
3) Modify ZER<,t>02 to use linear interpolation instead of interval halving
to obtain approximate zeros.
4) Modify ZER<,tl02 to find zeros for any continuous function rather than
just a polynomial function.

8-3 Synthetic Division to Get Further Zeros

We present the Remainder Theorem, Factor Theorem, and synthetic divi-
sion as aids to finding zeros after a first zero is known.
The Remainder Theorem
If p(x) is divided by (x - z), the remainder is the value of the polynomial
when z is substituted for x.
p(x) r
--'---'--- = q(x) + - - -
(x - z) (x - z)
That is, p (x) divided by (x - z) yields a quotient polynomial plus a remainder.
Multiplying through by (x - z) we get:
p(x) = (x - z)q(x) + r
and when z is substituted for x that produces:
p(z) = (z - z)q(z) + r
or simply
p(z) r

Factor Theorem
We note that when the value of r is zero, we have p (z) = 0, and that makes
z a solution of the equation p (x) = 0 or a zero of the polynomial Rin~f! t.he rf!-
mainder is zero after dividing by (x - z), it follows that (x - z) must be a
factor ofp(x), or
p(x) = (x - z)q(x)

Having found the first zero using the procedure of Sec. 8-2, we need only find
 Polynomials 109

q(x) and then use the procedure of 8-2 on it, repeating until all real zeros are
found.
Synthetic Division
We now develop the synthetic division algorithm using x 4 + 5x 3 +
2
9x + 8x + 4 divided by x + 2 as an example. It is this division that will
enable us to find q (x) above. We begin by performing the division "long hand":
x 3 + 3x 2 + 3x + 2
+ 5x 3 + 9x 2 + 8x + 4
+ 2x3

3x 3 + 9x 2
3x 3 + 6x 2
3x 2 + 8x
3x 2 + 6x
2x + 4
2x + 4

Notice that a great many things will always be written twice. We will always
"bring down" 9x 2 and 8x and 4. For division by x + 2, we will always get x 4
3
and 3x and 3x 2 and 2x repeated. So let us eliminate these repetitions and
compress the problem vertically:
x3 + 3x 2 + 3x + 2
x + 2)x4 + 5x 3 + 9x 2 + 8x + 4
2x3 + 6x 2 + 6x + 4
3x3 + 3x 2 + 2x

Now if we line things up properly, there is no need to write the x's and their
exponents. And there will always be an x in the divisor, so we don't need that.
Let's condense again:
1 + 3 + 3 + 2
+ 2)1 + 5 + 9 + 8 + 4
2 + 6 + 6 + 4
3 + 3 + 2
Since the coefficient of the first term of the quotient is always the same as the
coefficient of the first term of the original polynomial, we can make the bottom
line of figures exactly agree with the top line by simply bringing down the first
figure. Now we eliminate the top line to get:

+ 2)1 + 5 + 9 + 8 + 4
2 + 6 + 6 + 4
1 + 3 + 3 + 2
110 Advanced BASIC

Recognizing that subtraction is the same as "changing the sign and adding," we
can change the 2 to a -2 and add instead of subtracting. That leaves us with:

- 2)1 + 5 + 9 + 8 + 4
-2-6-6-4
---··--· ----
1 + 3 + 3 + 2
The sequence of operations is as folknvs: Bring d{nvn the 1, !!!U1t!p1y the 1 by
- 2, and '.'!rlte it under the 5. Adrl fi anrl - 2 to get 3. Multiply the 3 by - 2 and
write it down under the 9. Add 9 and -6 to get 3. Multiply 3 by -2 and write it
down under the 8. Add 8 and - 6 to get 2. Multiply 2 by - 2 and write it down
under the 4. Add 4 and -4 to get a remainder of 0. Division done by this
algorithm is called "synthetic division."
Since synthetic division is an iterative process, it is especially suited to the
computer. Notice that most of the process consists of multiplying and adding.
Not only is that an iterative process, but it is the same iterative process used for
evaluating polynomials by the nesting method. All that is required to adapt the
subroutine 800 in program ZER'l>02 to synthetic division is to store the sub·
totals along the way in an appropriate list element. This is done in line 820 of
program DIVIDE. Look at program DIVIDE and compare subroutine 800 in
it with the subroutine 800 in ZER'l>02.

DIV IDE
94 REM * THIS PROGRAM USES SYNTHETIC DIVISION
100 DIM PCIS), QCl5>
I I0 READ N
120 PRINT "PCX)•"J
130 FDR I • N+I TO I STEP -I
140 READ PC!>
150 PRINT PCl)J
160 NEXT I
162
170 PRINT
l!lO PRINT "DIVIDE BY X-"I
190 INPIJT XO
200 PRINT "QUOTIENT "J
210 GOSUB 800
220 ST0P
792
794 REM *THE DIVISION TAKES PLACE IN LINES 800 TO 830
795 REM * NOTE THE SIMILARITY TD NESTING
800 LET P3 • QCN+I) • P<N+ll
810 FOR I • N TO I STEP - I
-·820 LET OC!l • P3 • P3•XO + PC!)
830 NEXT I
832
834 REM * PRINT RESULTS
840 FOR I • N+I TO 2 STEP -I
850 PRINT OC!ll
860 NEXT I
862
870 PRINT "REMAINDER •"J QC I>
880 RETURN
882
884 REM
890 DATA 4, 1,5,9,3,4
900 END
 Polynomials 111

RIJN
DIVIDE

PCXl= I 5 9 8 4
DIVIDE BY X-? -2
QUOTIENT I 3 3 2 REMAINDER =0

Program DIVIDE works well, but why use a program to perform division
in the first place? Why not simply use program ZERQ)02 to look for zeros until
we find all of them? It is true that program ZER</)02 would easily find - 2 as a
zero of f(x) = x 4 + 5x 3 + 9x 2 + 8x + 4, but then what? Program ZERQ)02
will find no further zeros. We may use the results of a run of program DIVIDE
to see why.
The output of program DIVIDE above tells us that
4
x + 5x 3 + 9x 2 + 8x + 4 = (x + 2)(x 3 + 3x 2 + 3x + 2)

So -2 is a zero of our function f(x) = x 4 + 5x 3 + 9x 2 + 8x + 4. Next we

want a zero of q (x) = x 3 + 3x 2 + 3x + 2. Let's use program ZER</)02 to find
that zero.

890 DATA 3, 1,3,3,2

RUN
ZER002

POLYNOMIAL JS I 3 3 2

F'JRST. LAST. STEP? -s.s.1

-2 IS A ZERO

A run of ZER</)02 with the new data reveals why we could not have found one
of the other zeros. We now see that - 2 is a zero of the function f twice and our
program had no way of determining that fact. Values that occur more than once
as zeros are called multiple zeros. Why wouldn't ZER</)02 find the other two
zeros? Let's divide q(x) by (x + 2) and find out.

890 DATA 3, 1,3,3,2

RUN
DIVIDE

PCXl= I 3 3 2
DIV IDE BY X-? -2
QUOTIENT I I I REMAINDER 0

Thus one factored form of our original polynomial is

(x + 2)(x + 2)(x 2 + x + 1)

The first two factors tell us that - 2 is a zero twice and analyzing the third factor
using the quadratic formula tells us that the final two zeros are nonreal numbers.
They are -t
+ trJ3 and -t -
trJ3.
112 Advanced BASIC

Thus we see that the ability to divide polynomials by (x - z) where z is a

zero of the polynomial function makes more information available to us than we
would have if we limited ourselves to the procedures of program ZER'1>02. The
ability to divide enables us to find multiple zeros and, where we are able to
divide so that the quotient polynomial is a quadratic, we are able to determine
nonreal zeros. We note that caution must be exercised when dividing by ap·
proximate zeros. Each succeeding division will be susceptible to additional error.

Tht: Fa(;tf;::: Tht:t":re111 teHs us that \;;;hen ·\Ive fh:.d a zero z, ;ve may divide the
polynomial by (x - z) to obtain a new polynomial with one less zero. Synthetic
division has been shown to duplicate the steps of evaluating a polynomial by the
nesting method and so enables us to easily perform the division indicated by the
results of the factor theorem. This enables us to find multiple zeros and, in
some cases, nonreal zeros.
Problems for Sec. 8-3
1) Write a program to find the zeros of third-degree polynomials by find·
ing the first zero and then finding the remaining zeros by using the
quadratic formula.
2) Incorporate synthetic division into program ZER'1>02 so that when·
ever a zero is found, the division is performed and a search is begun for
the next zero.
3) Use the methods of this section to find as many zeros as possible for
the problems of problem 1 in Sec. 8-2.
8-4 Miscellaneous Aids
Integral Zeros
It can be shown that for an nth degree polynomial with zeros Zn,

Zn-I, ••• ,z2, Z1 that the following is true:

(x - Zn )(x - Zn-d ... (x - Z2 )(x - zi)

= anxn + an-1Xn-l + ... + a1X + ao

Considering the product of n binomials on the left we can see that (-zn) ·
(-zn-d ... (-z 2 )(-zi) is the constant term in the product which must equal the
constant term on the right, or a0 • If there is at least one integral zero and a0 is
an integer, that means that all integral zeros of a polynomial must be factors of
a0 • So we could write a program somewhat simpler than ZER<l>Ol that would
search only for integral zeros by first determining all integral factors of a0 • In
program ZER<!>Ol we used p (1) for a0 .
Descartes' Ruie of Signs
Consider the polynomial x 2 + 8x + 1. It should be clear that no matter
2
what positive value we might try for x, we can never make x + 8x + 1 equal
2
zero because x and 8x are both positive for x positive. Thus in searching for
real zeros we need not consider any positive numbers. The same is true for
-2x 2 - 5x - 1. Now, what about x 2 + 3x - 1? Can there be more than one
 Polynomials 113

positive value of x for which x 2 + 3x - 1 = O? No! Consider as another

example x 2 + 1. There is no positive replacement for x to make x 2 + 1 equal
to zero, and likewise there is no negative replacement for x that makes x 2 + 1
equal to zero since x 2 would have to equal -1. That tells us that since x 2 + 1
is a second-degree polynomial and has two complex zeros, they must both be
nonreal. Descartes observed all this and more and left us with Descartes' Rule
of Signs.
We may define the variation v in a sequence of numbers as the number of
changes in sign found by comparing successive pairs of adjacent numbers. For
example, for the sequence 1, 3, 4, -8, 2, the value of vis 2. There is no change
for 1 to 3 or 3 to 4. There is one change for 4 to - 8 and for -8 to 2. If zeros
appear in the sequence, we drop them. The sequence - 2, 8, 0, 5, - 3, 6 becomes
- 2, 8, 5, - 3, 6 in order to determine the number of variations, which is 3.
Descartes' Rule of Signs says that for

the number of positive zeros depends on the number of variations in the se-
quence, an, an-I, ... , a 1 , a0 , in the following manner. If v is the number of
variations, then the number of positive zeros is either v or v - 2 or v - 4, etc.,
but not less than zero. This may be written v - 2i where i is a positive integer.
It turns out that we may find a corresponding number for negative zeros
by finding positive zeros for p (- x). Substituting - x for x will change the sign of
all terms which have an odd exponent for x. Thus if p(x) = -4x 5 - 3x4 +
5x 3 - 2x 2 + x 3, the value of v is 4 and there must be 4 or 2 or 0 positive
zeros. Now we find that p(-x) = +4x 5 - 3x4 - 5x 3 - 2x 2 - x - 3 and
that v is 1. Thus there must be exactly one negative zero. For example, in
3x 6 - 2x 5 + x 4 - 2x 3 + 5x 2 - x + 1, we might expect to find as many as
six positive zeros, but under no conditions would we look for negative zeros
since p(-x) = 3x 6 + 2x 5 + x 4 + 2x 3 + 5x 2 + 1, which gives zero varia-
tions. All of this gives us a great deal of information. Sometimes the informa-
tion is exact, as when we get 0 or 1 as the number of variations. At other times
we get only a guide, as with x 3 + 2x 2 + 2x + 1, which has no positive real
zeros and three variations for p(-x), which gives 3 or 1 negative real zeros. A
solution of the problem will yield one negative zero and two nonreal zeros in
this particular example.
Problems for Sec. 8-4
1) Write a program to produce polynomials of random degree when zeros
are all random integral values.
2) Write a program to produce an nth degree polynomial given n integral
zeros.
3) Modify program ZER<JlOl to find all integral zeros by having x go from
-a 0 to a0 STEP SGN (ao ).
4) In problem 3, how many additions and multiplications would be re-
quired in a0 = 100 for nesting compared to the use of exponents, not
counting the loop operations?
5) Modify program ZER<JlOl to find all integral zeros by having x take on
only values which are factors of a0 •
114 Advanced BASIC

6) Analyze the number of operations called for in problem 5 as compared

to problem 3 for selected values of a 0 •
7) Write a routine to use Descartes' Rule of Signs which merely prints the
number of possible zeros in each of the following categories: complex,
positive, negative, and zero. Be careful about zero coefficients.
8) Incorporate the routines of problem 7 in ZERCj>02 so that the com-
puter stops searching if it has found enough positive zeros and if it has
found enough negative zeros. Be sure to have the program check for
both positive and negative possibilities.
9) An upper bound may be piaced on zeros by finding a non,negative
value of z such that after dividing by (x - z), all nonzero terms in the
third line of the synthetic division are of the same sign, provided the
leading coefficient ofp(x) is positive. A lower bound may be found by
taking the negative of the upper bound for p (-x). Write a program to
find upper and lower bounds.
 9
SEQUENCES AN
SERIES

9-1 Sequences
A sequence is simply a list of numbers. It is, of course, a natural for use in
computers, although a computer list may not be essential for a particular
application.
Sequences come in two kinds, finite and infinite. Obviously we will only
be able to evaluate a finite number of terms for any sequence used in a computer.
We routinely work with sequences. The set of counting numbers is a
sequence, as is the set of odd integers and the set of even integers. If we were to
consider 10 random numbers, they could constitute a sequence. It is more usual
that the numbers in a sequence follow some relatively simple pattern. One such
sequence you've probably seen goes 1, 1, 2, 3, 5, 8, where every number from
the third on is the sum of the previous two. This is called the Fibonacci sequence.
The numbers in this sequence have widespread significance in mathematics, art,
and nature. We can easily write a program to compute elements of this sequence,
store them in a list, and print them, as shown in Program FIBOl. One can
generate a Fibonacci type sequence by selecting any two integers for F(l)
and F(2).
F'l BO 1

94 REM * THIS PR0GRAM PRINTS THE ,IRST

95 REM 30 ,IB0NACCI NUMBERS
100 DIM ,<JO>
102
104 REM * ST0RE THE NUMBERS IN A LIST
110 LET ,<ll = ,<2l= 1
120 '0R X = 3 T0 30
130 LET ,<Xl = ,CX-ll + ,CX·2l
140 NEXT X
142
115
116 Advanced BASIC

144 REM *
N0W PRINT THE LIST
150 F'0R X = I T0 30
160 PRINT F"CX»
170 NEXT X
172
IBO END
RUN
F'lBO I

I I 2 3 5
B IJ 21 34 55
B9 144 233 377 610
00, t c;a"T L "1'£ C:
2584 ~!8!
10946 17711 26657 46J6tl 75025
:.:: : .,,;.-.,, : 96 .!~ i 6 .:; i 75 ii 5i~22~ 032(:40

Probably one of the simplest sequences is the set of counting numbers. It

begins 1, 2, 3, .... This is an example of an "arithmetic sequence." An arith-
metic sequence is one in which we get from one term to the next by adding a
constant, called the common difference. The general form for the nth term of
such a sequence is tn = t 1 + (n - l)d, and we will discuss this later. For the
counting sequence, d = 1.
Some sequences are motivated by physical problems. Suppose you are
200 feet from a fixed point. How long will it take to reach that point if you
cover half the remaining distance every minute? It should be clear that you can-
not ever reach that point, because at every minute along the way there is still a
distance separating you from the point and half of that is still a nonzero distance.
So there is always a distance remaining, and you can only cover half of it in the
next minute. However, it is also clear that at some point you can reach out and
touch the point. So we might ask how long it will take to be within six inches.
Consider program HALF.

HALF"
94 REM * THIS PROGRAM HALVES DISTANCE
95
96 REM * INITIAL CONDITIONS - 200 FEET AT TIME ZERO
100 LET D = ?.flO
110 LET T = O
112
114 REM* N0W COVER HALF THE REMAINING DISTANCE
115 REM EVERY MINUTE
120 LET T = T+I
130 LET D = D*Cl/2J
140 IF" D > 6112 THEN 120
150 PRINT "'DISTANCE ="'J DJ "'FEET"
160 PRINT "'IN"'J TJ "'MINUTES"'
17D END
RUN
HALF"

DISTANCE = 0.390625 F"EET

IN 9 MINIJTES

If we change HALF to look at the intermediate values of D, we get an

example of another common type of sequence. Consider program HALFl.

135 PRINT Tl D
RUN
HALF"!
 Sequences and Series 117

100
2 50
3 25
4 12.5
5 6.25
6 3. 125
7 1.5625
8 o.78125
9 o.390625
DISTANCE = 0.390625 rEET
IN 9 MINUTES

Each term in the sequence is calculated by multiplying the previous term

by a constant. In our problem the constant is t.
Such sequences are called
"geometric sequences." The constant term is called the "common ratio." The
nth term of a geometric sequence can be found by the formula tn = t 1 r<n - 1 ),
where r is the common ratio.
A common situation described by a geometric sequence is compound
interest. If you put money in a savings account, the bank adds interest to the
account at regular intervals, and the amount becomes the principle for the
next interest period. If the interest rate per interest period is r, we get the
amount at the end of that period asp + pr, which equals p(l + r), and at the
end of n periods we get p(l + rt. If you put $100 in a bank for one year at
5% interest compounded monthly, the rate for each interest period would be
.05/12, and the number of interest periods would be 12. Consider Program INT.

INT

94 REM * THIS PR0GRAM CALCULATES C0MP0UND

95 REM INTEREST BY r0RMULA
100 READ N
110 r0R I = I T0 N
120 READ p, RI> NI> Y
130 LET R = RI/NI
140 LET N =
Nl*Y
150 LET A= P•Ct+R/IOO>tN
160 PRINT "$"J PJ "AT"J RU "% F'0R"J YJ "YEARS"
170 PRINT "C0MP0UNDED"J Nil "TIMES ANNUALLY BEC0MES $"J A
180 PRINT
190 NEXT I
192
194 REM
200 DATA 3
210 DATA 100. 5, 4, 2
220 DATA 100. 5, 12. 2
230 DATA IQO, 5, 365. 2
240 END
RUN
INT

$ 100 AT 5 % r0R 2 YEARS

C0MP0UNDED 4 TIMES ANNUALLY BECOMES$ 110.449

$ 100 AT 5 % r0R 2 YEARS

C0MP0UNDED 12 TIMES ANNUALLY BEC0MES $ lt0.494

$ 100 AT 5 % r0R 2 YEARS

C0MP0UNDED 365 TIMES ANNUALLY BECOMES$ 110·516

For some sequences we are merely given a formula. For example, consider
the sequence for which the nth term is (2n - 1)/(2n + 1). We see the first 20
terms in program SEQOl.
118 Advanced BASIC

SEQOI

94 REM * THIS PROGRAM PRINTS 20 TERMS 0F THE

95 REM SEQUENCE C2•N-l>/C2*N+I)
100 DEF FNSCN> =
C2•N-t> / C2•N+I>
110 F0R I = I TO 20
120 PRINT FNSCI » 2•I- IJ "/"J 2*I+ I
130 NEXT I
132
140 END
RUN
SEQO!

0.333333 I / 3
U.(; 3 .i ~

0.714286 5 / 7
o. 777778 7 / 9
0.818182 9 / II
0.846154 II / 13
0.866667 13 / 15
0.882353 15 / 17
0.894737 17 / 19
0.904762 19 / 21
0.913043 21 / 23
0.92 23 / 25
0.925926 25 / 27
0.931034 27 / 29
0.935484 29 / 31
o. 939394 31 / 33
o.942857 33 / 35
0.945946 35 / 37
0.948718 37 / 39
0.95122 39 / 41

Summary of Sec. 9-1

Sequences are defined as lists of numbers. Sequences may be motivated
by a study of some natural phenomenon or simply an interest in the relation-
ships of numbers.
Problems for Sec. 9-1
1) Beginning with the second Fibonacci number, and continuing to the
eighteenth term of the sequence find (a) the square of the term,
(b) the product of the immediately preceding and the immediately
following terms, and (c) the difference of the numbers obtained in (a)
and (b).
2) For at least 25 Fibonacci numbers, find, for all possible pairs, the
greatest common factor.
3) For at least 15 Fibonacci numbers, print the ratios of adjacent terms.
4) Have the computer print at least 20 elements of a general Fibonacci
sequence making f(l) = 1 and f(2) = 3, or any f(l) and f(2) you
prefer.
5) For a general Fibonacci sequence like that in problem 4, print the
:r:atio uf at.lj<it;talt tt:rHiS.
6) For a general Fibonacci sequence like that in problem 4, print the
square of each term from the second to the next to last, the product
of the one just before and the one just after, and the difference of
these results.
7) Find the accrual to $1000 after seven years at 6% interest compounded
monthly and also compounded daily.
 Sequences and Series 119

8) Find how many minutes it will take for you to get to within six
inches of a point if you start 200 feet away and every minute you
cover one-third of the remaining distance.
9) Suppose that a rubber ball is known to bounce to three-fourths of the
height from which it is dropped. If the ball is dropped from 10 feet,
how many bounces will it take to bounce back less than one inch?
10) If you were to place one grain of rice on the first square of a chess
board, two grains on the second, four grains on the third, doubling
the number from each square to the next, how many grains would
you have to place on the 64th square? (Could you?)
11) Print a few terms of any of the following:
(a) 2n + 3 (d) nn
{b} 2n 2 - n + 1 (e) n(l fn)
(c) 2n /n 2 (f) (1 + 1/nt
9-2 Convergence and Divergence
Looking at the sequences of the last section and the results of the prob-
lems, we can see some differences. Note that for the Fibonacci sequence, the
numbers get larger and larger, and for the sequence in which the distance is halved
each minute, the terms get closer and closer to zero. For the sequence generated
by (2n - l}/(2n + 1), the terms seem to get closer and closer to one. The later
two sequences are examples of converging sequences, and the Fibonacci sequence
is an example of a diverging sequence.
If a sequence converges on some value, then that value is called the limit of
the sequence. In the formal study of limits, methods are developed for deter-
mining whether or not a sequence has a limit and for finding the limit if it does
exist. However, we will take a somewhat informal approach here.
The limit of a sequence of values s 1 , s2 , •• • Sn is denoted by
lim Sn
n-+"'
Thus for Sn (2n - l)/{2n + 1), we write:
2n - 1
lim - - - = 1
+ 1
n-+oo 2n

Note that there is no integer for which (2n - l}/(2n + 1) actually equals one,
but the larger n gets, the closer to one the value of (2n l)/(2n + 1) becomes.
We can get some insight into the behavior of sequences by looking at a few terms,
even if that insight is that we need to look at still more terms. In order to be
certain about the properties of some sequences, you should pursue a formal
study of limits. However, one clear benefit of using the computer is that we
may look at hundreds or even thousands of terms without tedious hand calcu-
lations. Of course, one way to save computer time is to print only those values
that we are interested in seeing. Sequences converge and diverge at vastly
different rates. Therefore don't be too quick to infer too much from just a few
terms.
We can look at .lim (-
1-+oo
t/ by writing a very short program to print some
terms. See program LIMO!.
120 Advanced BASIC

LIMO!

94 REM * THIS PROGRAM PRINTS SAMPLE VALUES

95 REM FOR THE SEQUENCE S = C-3/5ltH
100 DEF F"NLCHl = C-3/5ltH
110 FOR I = I TO I 0
120 PRINT IJ F"NLIIl
130 NEXT I
132
140 PRINT 1001 FNLCIOO>
150 PRINT IOIJ F"NLC IOI l
!60 EMO
RUN
LIMO!

I -0.6
2 o. 36
3 -0.216
4 0.1296
5 -0.07776
6 o. 046656
7 -2.79936E-2
B 1.67962E-2
9 - I. 00777E-2
10 6.Q4662E-3
100 6.53319E-23
101 -3.91991E-23

Looking at just the first 10 terms we can see that each term is closer to
zero than the one before and that the values alternate between positive and
negative. For the one-hundredth term, we get 6.5 X 10- 23 and for the one-
hundred and first term we get - 3.9 X 10-23 , making us more and more confi-
dent that the limit is zero.
Not all converging sequences converge on zero. Consider lim (2 + t").
n-+oo

We can see that (i-)" converges on zero, and therefore 2 + (i-)" converges on
2 + 0, or 2. Look at program LIM02, and see how much faster it converges
than program LIMOl.

LIM02

94 REM * THIS PROGRAM PRINTS SAMPLE VALUES

95 REM FOR THE SEQUENCE S = C2+ C1/5 l t H>
100 DEF FNLCHl = C2+C!/5ltHl
110 FOR I = I HJ I 0
120 PRINT II F"NLC!l
130 NEXT I
132
140 END
RUN
LIM02

I 2.2
2 2.04
3 2.oos
4 2.0016
5 2.00032
6 2.00006
7 2.00001
B 2.
9 2.
10 2.

It should be clear that if any number n such that In I < 1 is raised to

higher and higher powers, the closer to zero n is, the faster the result approaches
zero.
 Sequences and Series 121

We will look at an example of divergence before we leave this section.

Suppose that we have a magic ball that bounces to 110 percent of the height
from which it is dropped. If we drop this ball from 10 feet, after one bounce it
reaches 11 feet. After the second bounce, it reaches 12.1 feet, etc. How high
will it go after 10 bounces? See program B<!>UNCE.
B0UNCE

94 REM * THIS PR0GRAM PRINTS HEIGHTS FOR A BALL

95 REM RECOVERING 11 OX 0F ITS HEIGHT EACH BOUNCE
100 LET H = I 0
110 F0R I = I TO I 0
120 LETH= H*l.10
130 PRINT !J H
140 NEXT I
142
150 END
RUN
BOUNCE

I II
2 12.1
3 13.31
4 14.641
5 16. I 051
6 17.7156
7 19.48'72
8 21.4359
9 23.5795
10 25.9374

Summary of Sec. 9-2

We have been judging divergence and convergence of sequences by looking
at successive terms.
Problems for Sec. 9-2
1) Compare the convergence of (- ~g~)" with <lot.
2) Print a few terms of 1 + (f
)"and (1 + (f ))". 2
3) Write a program to give enough terms of (2 + (n - 2)/n ) to decide on
convergence.
4) Write a program to examine any or all of the following:
n2 + 3n - 1
(a)
n
n - 1
(b) (n + l)(n - 2)

( c) 1/n!
1
(d) 1 - 1/3! + 1/5! + ... + (-1)"+ /(2n - 1)!
(e) (1 + (1/n)")
{f) sin(n)
n
5) It can be shown for the Fibonacci sequence that the nth term is:
122 Advanced BASIC

Verify this for a reasonable number of terms.

9-3 Series
A series is what you get if you write the terms of a sequence with plus
signs between them. In other words, a series is the sum of a sequence. Thus the
series for the counting sequence is
1 + 2 + 3 + 4 + 5 + ...
The sum of the first five terms is 15. Note that if we look at successive sums, we
can form yet another sequence called the sequence of partial sums. For the
counting sequence, we get
1, 1 + 2, 1 + 2 + 3, ( + 2 + 3 + ~ 1 + 2 + 3 + 4 + 5, ....
--.,......... '---y----1 ~
1 3 6 10 15 ...
We could look at the series associated with that sequence and so on without end.
For some series, we can evaluate the sum by formula, and for others, to do
so is difficult. It is even possible to evaluate the sum by formula for certain
infinite sequences.
Looking at an arithmetic series of n terms, we see that
Sn = t1 + (t1 + ld) + (t1 + 2d) + ... + (t1 + (n - l)d) (8·1)
and looking at the same series in reverse order we see that
l)d) + ... + (t1 + 2d) + (t1 + ld) + t1 (8-2)
Adding (8-2) to (8-1) we get
2sn = (2t 1 + (n - l)d) + ... + (2t 1 + (n - l)d) + (2t 1 + (n - l)d)

+ (2t 1 + (n l)d)
which is therefore twice the sum sn. Note that 2t 1 + (n - l)d occurs n times.
Thus we get 2sn = n(2t 1 + (n - l)d) or 2sn = n(t1 + (t1 + (n - l)d).
Sincethenthte rmist 1 + (n - l)d,weget2sn = n(t 1 + tn)or
Sn (n/2)(t1 + tn)
This gives us a choice for finding the sum of an arithmetic series. We may
use either the formula, or add terms as we generate them in a computer loop.
For a geometric series we have
.L •• 2 • Jn-1\
'JI T lt/ (8-3)
Multiplying both sides by r we get
rsn = t 1r + t 1 r 2 + t 1r 3 + . . . + t 1rn (8-4)
Subtracting (8-4) from (8-3) we get
 Sequences and Series 123

which simplifies to

Again we have the choice of finding the sum by formula or by having the
computer add terms as it generates them.
We can get some idea about how the arithmetic and geometric series
behave by picking two terms and treating them as the first and second terms for
both kinds of sequences. This is done for two pairs of terms in two runs of
program GEq.>ARI.
Note that in both cases the sequence of sums for the arithmetic sequence
seems to diverge. In fact, all arithmetic sequences ford <> 0 diverge and so
do their associated sequences of sums. We can see that one of the geometric
sequences diverges and the other seems to converge. All geometric sequences
and their associated series for which lrl < 1 converge.

GEO ARI

94 REM * THIS PROGRAM PRINTS VALUES FOR GEOMETRIC AND

95 REM ARITHMETIC SEQUENCES AND SERIES WITH THE SAME
96 REM FIRST TWO TERMS
100 PRINT "FIRST TW0 TERMS";
110 INPUT AC I>. AC2>
120 LET SCI> =
TCI> =
GCll = ACJ>
130 LET GC2l =
AC2>
132
134 REM * FIND COMM0N DIFFERENCE f0R ARITHMETIC SEQUENCE
140 LET D =
AC2l - ACll
142
144 REM * FIND COMM0N RATIO F0R GEOMETRIC SEQUENCE
150 LET R =
GC2l/GCI>
160 PRINT "TERM"• "GEO SEQ''. "GEO SERIES''.
170 PRINT "ARITH SEQ''. "ARITH SERIES"
172
174 REM * STORE SUCCESSIVE VALUES IN LISTS
I BO F0 R I = 2 T0 I 0
190 LET GCI> GCI-l>*R
200 LET SCI> SCI-I> + GCI>
210 LET AC!> ACI-1> + D
220 LET TC!> TCI-1> +AC!>
230 NEXT I
232
234 REM *PRINT RESULTS
240 FOR I = I TO I 0
250 PRINT J. GCI>. SCI>. AC!>• Tell
260 NEXT I
262
270 END

RUN
GE0ARI

FIRST TWO TERMS? 112

TERM GEO SEQ GEO SERIES ARITH SEQ ARITH SERIES
I I I I I
2 2 3 2 3
3 4 7 3 6
4 B 15 4 10
5 16 31 5 15
6 32 63 6 21
7 64 127 7 28
8 128 255 8 36
9 256 511 9 45
JO 512 1023 10 55
124 Advanced BASIC

RUN
GEil AR I

FIRST TW0 TERMS? 4, 3

TERM Gm s<:o GEO SERIES AR!TH SEQ AR!TH SERIES
I 4 4 4 4
2 3 7 3 7
3 2.25 9.25 2 9
4 l o6B75 10.9375 I I0
5 1.26562 12.2031 0 I0
6 ·949219 13.! 523 -1 9
7 • 711914 t 3.13643 -2 7
8 ~533936 14.3932 -3 4
9 .400452 14.7986 -4 0
•v ,, 30033-? I 5., U'J') -~ -5

Looking at the formula for the sum of a geometric sequence,

Sn =ti(~)
1 - r
we can see for lrl < 1 that rn gets closer and closer to zero as n gets larger
and larger. Thus 1 - rn gets closer and closer to one, that is, Jim 1 - r n = 1,
n-+~

and

Jim
n-+ 00
Sn = n-+
Jim 00
ti(~)
1 - ,. = ti (
1
1
_ r)

Thus for an infinite sequence with lrl < 1,

1
s =ti ( -)
1 - r
Finally, we will look at some other series in the next section.
Summary of Sec. 9-3
We have defined a series as what we get by replacing commas with plus
signs in sequences. We can find the successive sums of a series, which is the
sequence of partial sums. For an arithmetic series the nth sum is sn = (n/2) ·
(ti + tn), and for a geometric series the nth sum is Sn = ti (1 - rn)/(1 - r).
Series either converge on some real number value or they diverge. All
arithmetic series diverge, and geometric series diverge for a common ratio greater
than or equal to 1 or less than or equal to -1.
Problems for Sec. 9-3
1) For the magic ball bounding to 110 percent of its height in Sec. 9~2,
find the total distance traveled before reaching the height of the tenth
b('t!n~e.
2) Find the sum of the first n positive integers for the following values of
n: 10, 100, 1000, and 10000.
3) In the song "The Twelve Days of Christmas," gifts are bestowed upon
the singer in the following pattern: the first day she receives a partridge
in a pear tree; the second day two turtle doves and a partridge in a pear
 Sequences and Series 125

tree; the third day three French hens, two turtle doves, and a partridge
in a pear tree. This continues for 12 days. On the twelfth day she
receives 12 + 11 + · · · + 2 + 1 gifts. How many gifts were there
altogether? Note that the figure asked for here is the twelfth term of
the sequence of partial sums of the sequence of partial sums of the
sequence of positive integers.
9-4 More on Series
There are many series that have importance in the field of mathematics
which are neither arithmetic nor geometric.
It can be shown that the cosine is the sum of an infinite sequence:
x2 x4 x2n-2
cos(x) = 1 - - + - - · · · + (-l)n+l + ···
2! 4! (2n - 2)!
While this is a very tedious calculation by hand, it is relatively simple with the
aid of a computer program. As with finding zeros for polynomials, we use a
relative comparison to decide when to stop (see line 180 of program Cft>SINE).
However, in the case of polynomial evaluation, we were using the value at the
midpoint of an interval which we knew contained the true value. In that case we
had a measure of relative error. In the present situation we have somewhat less
information. We have only the sum of n terms to compare with the sum of
n - 1 terms. Thus we are saying that the magnitude of the most recent term is
small enough to stop summing. When the magnitude of the most recent term is
very small compared to the sum so far, we may expect the error to be small also,
but not necessarily as small. We have not attempted to measure the cumulative
effect of the remaining terms, although methods exist for evaluating it.

COSINE

94 REM * THIS PROGRAM APPROXIMATES CO SINE X

95 REM USING TERMS Of A SERIES
100 PRINT ''FIND COSINE Of"J
110 INPUT X
120 LET SI = S = N = f = I
130 PRINT NJ S
140 LET N = N+l
150 LET SI = S
152
154 REM * f IS THE VALUE Of FACTORIAL 2*N-2
160 LET f = f*C 2*N-2 >*< 2*N·3 >
170 LETS= S + CC·!ltCN+ll * XtC 2*N·2 l/f l
-180 If ABS<S·Sl l/CABSCSl+ABSCS I ll > lE-6 THEN 130
190 PRINT "BY COMPUTER COSINE fUNCTION"J CGSCXl
200 END
RUN
COSINE

FIND CGSINE Of? t.57

I 1
2 -0.23245
3 2.07055E·2
4 ·9.45914E·5
5 8.20947E·4
6 7.95873E·4
7 7.96341E·4
8 7.96335E-4
BY CGMPUTER COSINE FUNCTION 7.96334E·4
126 Advanced BASIC

Similarly it can be shown that

x3 xs x211-1
sin(x) = x - - + - - ... + (-1) 11 + 1 - - - -
3! 5! (2n - 1)!
Evaluation of this is left as a problem.
Problems for Sec. 9-4
1) Vlrite a program to calculate sin(x) using the series in this section.
2) Each of the following series approaches pi as n increases.

(a)4(1-t+t- ... +(-l)"-1(2nl 1) + .. '.)

+ __!__)
nz

+
(2n ~ 1) 2)

Write a program to compare convergence for each of the above.

3) The constant e, whose value is approximately 2.718, is of importance
in calculus. It can be shown that
1 1 1 1
-+-+-+ +-
0! 1! 2! n!
and

both approach e as n increases. Compare convergence for the two

different methods of calculating e.
4) One method of approximating pi is to select random points in a square
and find how many of them fall within a quarter circle whose radius is
a side of the square. The number of points that fall within the quarter
circle should be proportional to the area. The area of the quarter circle
is t7Tr 2 , whereas the area of the square is r2 • Thus four times the ratio
of circle to square should approximate pi. This is a variation of the
Monte Carlo method. Write a program to select from 1,000 to 10,000
such points to obtain an approximate value for pi.
 10
MATRICES

10-1 Introduction
A matrix is simply a rectangular array of numbers. As such, an array is a
set of numbers arranged in rows and columns. This is, of course, exactly the ar-
ray we have been using whenever we have used subscripted numeric variables in
BASIC. A matrix may also consist of a single row or a single column. Such
arrays are sometimes called "row vectors" and "column vectors."
It is the purpose of this chapter to study some of the properties of matri-
ces and see some applications. We will not concern ourselves with theorems and
their proofs. A brief review of Sec. 1-5 at this time might be helpful.
Matrices may be studied in a totally abstract setting, but it is useful to
have a concrete example. Suppose that we are operating the Framis Corpora-
tion, which employs three salesmen (it could be 300): Brown, Jones, and Smith.
Brown, Jones, and Smith are selling clevises, hammer handles, shoehorns, and
whipsockets. In a given week the three salesmen turn in orders for merchandise
as shown in Table 10-la.
One fundamental piece of information is the amount of money brought in
by each salesman. To calculate that we need the prices as shown in Table 10-lb.
We can find subtotals by multiplying the price of an item by the number of
items. The subtotals are shown in Table 10-2a. Now to obtain each salesman's
total, we simply add across the rows to get the figures shown in Table 10-2b.

TABLE 10-1a ARRAYS

Hammer
Salesman Clevis handle Shoehorn Whipsocket

Brown 30 800 50 20
Jones 50 31 40 10
Smith 0 500 50 90

127
128 Advanced BASIC

IADLC i0·2a

Hammer
Salesman Clevis handle Shoehorn Whipsocket

Brown $30 $312.00 $24.50 $ 75.80

Jones 50 12.08 19.60 37.90
Smith 0 195.00 24.50 341.10

TABLE 10-1b ARRAY P TABLE 10-2b ARRAY M

Item Price Salesman Dollar sales

Clevis $1.00 Brown $442.30

Hammer Jones 119.59
handle 0.39 Smith 560.60
Shoehorn 0.49
Whipsocket 3.79

To make the discussion easier, we label the array of Table 10-la as S, the
array of Table 10-lb as P, and the array of Table 10-2b as M. Thus to get
M(l,1), we add the products S(l,I)*P(I,l), for I going from one to four, and to
get M(2,l), we sum the products 8(2,I)*P(I,l), for I going from one to four.
That makes M(J,1) the sum of the products S(J,I)*P(I,1) for J going from one to
three and for I going from one to four. If the P array had had a second column,
such as the salesmen's commission per item, we could carry out the above pro-
cess for the second column, getting M(J,K) by summing up the products
S(J,I)*P(I,K), for K going from one to the number of columns in P, J going from
one to the number of rows in S, and I going from one to the number of columns
in S. Note that the number of columns in S must equal the number of rows in
P and that the product array has the number of rows in S and the number of
columns in P. This is exactly the definition for matrix multiplication. There-
fore, instead of constructing triply-nested loops to perform the algorithm de-
scribed above, we take advantage of the BASIC matrix multiplication statement
(see line 290 of program SALESl). To make the run of the program more
readable, we have used string lists to provide labeling.

SALES I

94 REM * THIS PROGRAM CALCULATES SALES T0TALS

95 REM USING MATRICES
100 DIM sc3,4), p(4, I ) , MC3, I ) , !$(4), S$(3)
110 MAT READ s, p, I$, S$
112
114 REM* PRINT RAW DATA IN LINES 120 THROUGH 260

130 F0R I =
I T0 4
140 PRINT !$CI>,
150 NEXT I
152
16 0 F0 R I = I T0 3
170 PRINT S$C!l,
180 F0R J = I T0 4
 Matrices 129

190 PRINT SCI,J>,

200 NEXT J
202
210 NEXT I
220 PRINT
222
230 PRINT "ITEM"• "PRICE"
240 FOR I = I TO 4
2SO PRINT !$CI>. PCI.J>
260 NEXT I
270 PRINT
272
274 REM *CALCULATE T0TALS IN LINE 290
2BO PRINT "SALESMAN"• "DOLLAR SALES"
2B2
-- 290 MAT M = S•P
300 F0R I = I T0 3
310 PRINT S$CI>• MCl1ll
320 NEXT I
322
324 REM
330 DATA 30, BOO, SO~ 20
340 DATA so. 31. 40. 10
3SO DATA o. soo. so. 90
3S2
360 DATA 1 . . . 39, ,49, 3,79
362
370 DATA CLEVIS• HAMMER HANDLE
3BO DATA SHOE HORN, WHIPSOCKET
382
390 DATA BROWN, JONES, SMITH
400 END
RUN
SALES I

SALESMAN\ ITEM CLEVIS HAMMER HANDLE SHOE HORN WHIPSOCKET

BROWN 30 BOO so 20
JONES so 31 40 10
SMITH 0 soo so 90

ITEM PRICE
CLEVIS I
HAMMER HANDLE o. 39
SHOE HORN 0.49
WHIPSOCKET 3.'79

SALESMAN DOLLAR SALES

BROWN 442·3
JQNES 119.S9
SMITH 560.6

It may strike you that SALESl is rather long, and indeed it is. However,
most of the program is devoted to producing nicely formatted output. If all that
is required is the three dollar amounts at the end of the printout, that can be
done with an extremely short program having just one calculation statement and
one print statement, as shown in program SALES2.

SALES2

100 DIM SC3,4), PC41 I>. MC 3, I)

110 MAT READ s, p
-290 MAT M = S•P
310 MAT PRINT M
330 DATA 30. BOO• so. 20
340 DATA so. 31. 40. 10
3SO DATA o. soo. so. 90
130 Advanced BASIC

360 DATA 1, .39, .49, 3.79

400 END
RUN
SALES2

442.3
119.59
560.6

There are rnany properties of rnatrices \i\fhich corn.e to light lhrough the
process of just experimenting with different relationships. So it is recommended
that you do as many probiems as possible and that you work on problems of
your own throughout this chapter.
Summary of Sec. 10-1
Matrices are not only convenient as storage areas for data, they possess
mathematical properties that are both interesting and of practical value. The
mathematics of matrix algebra is fairly complex, containing numerous abstract,
involved, and intricate theorems. The thrust of our work will be to use familiar
properties to develop new facts.

Problems for Sec. 10-1

1) As manager of the FramisCorporation of this section, you are interested
only in total sales. Create a row vector T to contain the sales totals by
item for the week. Then find T*P and print it.
2) MAT READ the integers 1through12 into both a column vector C and
a row vector R. Find R *C and C*R and print them both.
3) Write a program to print integer powers of an array. How must the
number of rows and columns be related for this?
4) A light fixture manufacturer makes three different fixtures requiring
parts as shown in the table and wishes to make 800 of fixture A, 200
of fixture B, and 1,500 of fixture C. Write a program to find how
many bulbs, switches, meters of wire, and screws will be needed.

A B c
Bulbs 3 1 4
Switches 1 1 2
Meters of wire 9 2 3
Screws 15 8 12

5) Find any or all of the following products:

<·r m; :J 2 (blO 2
l 0
1 0
5
8
1
0 1
0 1 m; 5
8 :J
 Matrices 131

(<l~ ~ n~ :J
6) Find the result:
:
~1 [~:
800 50
.39
20]] . [1.00]
31 40 10
1 1]
500 50 90 .49
3.79
and compare with the results of problem 1.

10-2 Solving Simultaneous Linear Equations Using MAT INV

The matrix equation,

(10-1)

can be multiplied out on the left side to obtain

a1x + b1Y + C1Z] [d1] d2. (10-2)

ll2X + b2Y + C2Z =
[ +
ll3X b3y + C3Z d3

We say that two matrices are equal if each entry of one equals the corresponding
entry of the other. (Each entry of a matrix is often referred to as an element of
the matrix.) For MAT A = MAT B that means A(I,J) = B(I,J) for all values of
I and J. Therefore, we may say that
ll1X + b1Y + C1Z = di
a2x + b2Y + C2Z = d2 (10-3)
ll3X + b3y + C3Z = d3

Equation (10-3) constitutes a system of three linear equations in three un-

knowns. Actually Eqs. (10-1), (10-2), and (10-3) are simply three different ways
of writing the same equality.
We want to solve the set of equations (10-3). It will be easier to discuss
the solution if we assign variables to the matrices of Eq. (10-1) as follows:
132 Advanced BASIC

iviairix C may be referred to as the coefficient matrix, : : ; the solution matrix, and
K the matrix of constants. Now we may rewrite Eq. (10-1) in the form
C*S = K
and proceed to solve for S.
It would be very convenient if we could just divide both sides by C. But
it turns out that the division of one matrix by another is not an easily describ-
able process. However, division by C is equivalent to multiplication by the in-
verse of C and the inver~e of C, if it exists, is easily obtained in BASIC.
Before we use an inverse to solve simult.anP.ons linP.ar P.rprntiom:: !i:>t's !ook
more closely at just what the inverse of a matrix is. The inverse of a matrix C
is the matrix c- 1 such that the product of C and c- 1 is the identity matrix. The
identity matrix has the same number of rows as columns and is filled with zeros
except for the upper left to lower right diagonal, which is filled with ones. Note
that in order for a matrix to have an inverse, it must be square.
Let's find, for example, the inverse of

We are looking for a matrix with entries a, b, c, and d such that

[: :J .[~ :J = [~ ~]
Finding the product on the left we get

[ :: : ~: :: : :: ] = [ ~ ~]
If two matrices are equal, then their corresponding entries are equal. So we get
the following four equations with four unknowns:
5a + 7b 1 6a + Sb 0
5c + 7d 0 6c + 8d 1
We can easily solve these equations to get a -4, b 3, c 3.5, and d
- 2.5. Therefore,

[~ ~]
or

r-4 3 1
L3.5 -2.5J
We can easily verify this with the demonstration program MATINV. (Note
that the computer in some cases, unlike our example here, is susceptible to slight
roundoff errors when the MAT INV statement is used.)
Now to get back to solving simultaneous linear equations. We had the
 Matrices 133

MAT INV

94 REM *
THIS IS A PR0GRAM TO DEMONSTRATE MAT INV
100 DIM xc2.2>. AC2.2), PC2.2>
110 MAT READ A
120 MAT X = INVCAl
122
130 PRINT "ORIGINAL MATRIX"
140 MAT PRINT A
150 PRINT
152
160 PRINT "INVERSE MATRIX"
170 MAT PRINT X
!BO PRINT
182
190 PRINT "THE PRODUCT IS"
200 MAT P =
X*A
210 MAT PRINT P
212
214 REM
220 DATA
230 END
RUN
MA TI NV

0RIGINAL MATRIX

5 6
7 B

INVERSE MATRIX

-4 3
3,5 -2.5

THE PRODUCT IS

I 0
0 I

matrix equation, C *S K. Now

c- 1 *C*S = c- 1 *K (10-5)
and a matrix times its inverse gives the identity matrix, sometimes designated I.
Thus c- 1 *C = I. The identity matrix has the property that for any matrix M
with dimensions compatible with the dimensions of I,
I*M = M*I = M
Thus Eq. (10·5) becomes
I*S c- 1 *K
or
S = c- 1 *K
and we have solved Eq. (10-4) for S, which was our purpose. It is now a relatively
simple matter to write a program (see program S</)LVE) to solve the system,
x+2y- z=2
3x 3y + 2z 3 (10-6)
4x y + 2z 8
134 Advanced BASIC

S0LVE

94 REM * THIS PROGRAM SOLVES SIMULTANEOUS LINEAR

95 REM EQUATIONS USING THE MAT INV STATEMENT
100 DIM CC3,3>. SC3. I>. KC3, I>. NC3.3>
110 MAT READ C• K
120 MAT N = INVCC>
130 MAT S = N*K
140 PRINT "SOLUTIONSt"
150 MAT PRINT S
152
154 REM

170 DATA 2,3,g

RUN
S0LVE

S0LUTI0NS1

In program S<;!>LVE, the column vector,

s-m
translates back to x = 1, y = 2, and z = 3. We may now substitute these
values in Eq. (10-6) to verify that they do in fact solve the system of equations.
Summary of Sec. 10-2
We have seen that sets of simultaneous linear equations may be solved by
considering an equivalent matrix equation C*S = K, where C is the coefficient
matrix, S is a column vector which contains the values of the variables in the
original set of linear equations, and K is a column vector containing the constant
terms in the original set of linear equations. We may solve for S by finding the
inverse of matrix C, so that S c- 1 *K. The inverse may be found with the
BASIC statement MAT I = INV(C). For systems of simultaneous linear equa-
tions having a unique solution, MAT C will always be square, which is one of the
requirements for having an inverse.
Problems for Sec. 10-2
1) Let

A [~ -~ J
Find the print A- 1 , A*A- 1 , and A- 1 *A.
2) Let

B [-80 -1-3]
=
 Matrices 135

Find B- 1 and print it. Verify by hand-computing the inverse of B.

Find and print B*B- 1 and B- 1 *B.
3) Solve for x and y:
-2x - 5y = -16
- x + 4y = 31
4) Solve for x, y, and z:
2x - 9y 5z 2
7x 6y + 5z -35
9x 6y + 5z -39
5) Solve for x, y, and z:
3x + 4y + z = 7
5x - 6y + 3z = 8
3x + 4y + z = -3
6) Solve for w, x, y, and z:
6w + 3x + 6y + 5z = -12
-7w + 5x - 7y - z = 77
-3w + x + 3y + 6z = 31
-2w - 4x + 4y - 7z = -76
7) Solve for w, x, y, and z:
-3w + 6x - 5y - z = -32
w + 9x - 5y - 2z = 9
w + 6y + 5z = 2
-7w + 4x - y + 5z -86
8) Solve for x, y, and z:
2x + 4y - 3z = -11.9
-9x - 3y 58.5
-9x + 8y + 5z = 66.6
9) Solve for v, w, x, y, and z:
7v + 6w - 3x - y + 9z = 26.3
-9v + 2w + 9x + 5y + z = 91.1
-3v + 4w + 5x + 5z = 62.9
6v - 8x - 2y - 6z = - 55.6
-3v - 9w + 5x + 7y + 3z = -25.9
10) Let
136 Advanced BASIC

A = [~ :~ -!] •nd B

Find and print (A*Bf 1 and B- 1 *A- 1

11) Write a program that can solve sets of simultaneous linear equations
having different numbers of equations. Provide an item of data that is
the nurnber of equation~ and redimension aH matrices accordingly.
 11
SOME TOPICS
OF STATISTICS

11-1 Introduction
The possibilities for using the computer to analyze and summarize large
amounts of data are virtually unlimited. This chapter will introduce just a few
fundamental statistical calculations.
11-2 Average, Variance, and Standard Deviation
One of the most common measures of statistical information is the average
or arithmetic mean. The average is the sum of the measures divided by the num-
ber of measures. In some cases the mere task of counting the number of mea-
surements may be a job in itself. So we can even use the computer to do the
counting for us. All that is necessary is to append an item of artificial data as a
flag to stop counting and calculate the average, as shown in program AVG.

AVG

94 REM *THIS PROGRAM COUNTS DATA AND

95 REM CALCULATES AVERAGE
100 LET N = T = 0
110 READ D
-120 IF" D = .01 THEN 160
130 LET N = N+ I
140 LET T = T+D
150 G0TO 110
160 PRINT " MEASUREMENTS"; N
170 LET A = TIN
190 PRINT "AVERAGE MEASURE"J A
492
494 REM
500 DATA 93, so, 73, 92, 77, 34, 33, 79, 37, 73
510 DATA 99, 63, 63, 92, s 1, 93, 47, 53, 89, 100
520 DATA 93, 71, 73, .Qt
530 END t
137
138 Advanced BASIC

RUN
AVG
MEASUREMENTS 23
AVERAGE MEASURE B0.3478

The average for a set of data gives no idea of the spread or dispersion of
the data. The average of zero and 100 is 50, and the average of 49 and 51 is also
50. We could get some idea by having the computer find for us the largest and
the smallest rneasures. Even that inf0:rrnatiun eou!d be misleading" since the
largest measure could be much larger than the next largest, or the smallest could
be much smaiier than the next smallest. One way to gain some insight into the
distribution of the measures is to find the average of the amount by which each
measurement differs or deviates from the average of the measures. There is a
flaw here, however, as some will deviate by a positive amount and some will
deviate by a negative amount, thus cancelling each other out. Using mean ab-
solute deviation would avoid this difficulty. However, expressions involving
absolute value are difficult to work with algebraically, and statisticians thus find
the average of the squares of the deviations. This figure is called the "sample
variance." In order to write a formula for variance, we use the Greek letter
sigma, L , which indicates summation. Defining average using summation nota-
tion looks like this:
n
.2: xi
i=l
A =
n
The average A is the sum of all values of xi for i going from 1 to the number of
measurements, which is n, divided by the number of measurements. We define
variance in terms of the average as follows:
n
L (xi - A)2
i=l
V=
n
Even the variance, which gives an indication of how measurements are distrib-
uted, doesn't indicate actual dispersion. It indicates the square of dispersion.
Thus we take the square root of V, getting a number called "standard
deviation":
n
L (xi - A)2
i=l
s2
n
and

s
 Some Topics of Statistics 139

Now if we try to apply a computer program directly to the formula for 8 2 , we

soon find that we will have to READ the DATA twice, once to find the average,
and again to get each value of xi - A. This is not a problem for small amounts
of data; but since it can be avoided, let's do so.
It can be shown that
11

I: (xi - A)2 11
i=l 1 x~I - Az
n
I:
n i=l

This means that we can, alternatively, have the computer sum up the squares of
the measures rather than the squares of the deviations. This can easily be incor-
porated into program A VG. See lines 150 and 210 of program VAR.

VAR

94 REM *THIS PROGRAM COUNTS DATA AND CALCULATES

95 REM THE AVERAGE, VARIANCE AND STANDARD DEVIATION
100 LET N = T = Tl = 0
110 READ D
120 IF" D = • 0 I THEN I 7 0
130 LETN=N+l
140 LET T = T+D
-150 LET Tl = Tl + Dt2
160 GOTO 110
170 PRINT " MEASUREMENTS"J N
172
IBO LET A = TIN
190 PRINT " AVERAGE MEASURE" J A
192
200 LET Al = Tl/N
-210 LET V = Al - At2
220 PRINT VARIANCE"J V
222
230 LET S = SQRCV l
240 PRINT "STANDARD DEVIAT!ON"J S
492
494 REM
500 DATA 98, so, 73, 92, 77, 84, 83. 79, 87, 73
510 DATA 99, 63. 63. 92, B 1, 93, 47, 53, 89, 100
520 DATA 98. 7), 7 3, .01
530 END
RUN
VAR

MEASUREMENTS 23
AVERAGE MEASURE 80.3478
VARIANCE 202.314
STANDARD DEVIATION 14·2237

Summary of Sec. 11-2

In this section, the average, or arithmetic mean, variance, and standard
deviation have all been defined. We have written programs to calculate the
average and standard deviation.
Problems for Sec. 11-2
1) Modify program AVG so that the computer tells us the highest and
the lowest measures.
140 Advanced BASIC

2) Modify program VAR so that we get the deviations of the largest and
smallest measures from average in terms of the number of standard
deviations. (If the largest measure is 91 for an average of 70 and the
standard deviation is 7, then the largest measure would be three stan -
dard deviations from average.)
3) Often in practice we use what is called the "weighted average." Sup-
pose that tests count three quizzes and that the final exam counts two
tests. Find the weighted average for quiz marks 70, 80, and 73, test
marks 63, 82, and 91, and a final exam of 83.
4) Generate 100 random numbers from 1 to 201. Caicuiate the average
and standard deviation.
5) Generate 100 random numbers from -100 to 100 and calculate the
average and standard deviation.
6) Devise a scheme without reading data twice for finding accurately
the average and standard deviation for the following data: 9999.12,
9999.36, 9999.64, 10000.03, 10000.41, 9999.83, 9999.51, 9999.13
and 10000.08. Due to roundoff error, many programs will give a large
standard deviation for the above data. (Hint: we can simply measure
differences from 10000 instead of from zero.)
11-3 Median
The median is the middle value. Sometimes the median is presented as an
item of statistical information, such as median income or median weight. If
there are an even number of data items, then the median is the average of the
middle two values. One reason for using the median is that it tends to be less
affected by a few widely dispersed items of data than the average. There are
no particularly difficult calculations required to find the median. What does
have to be done, though, is to first arrange the data in numerical order. Thus
let us develop an ordering routine.
There are many, many ways of ordering. Some ordering procedures are
very elaborate and some are very simple. As the number of items to be sorted
increases, the need for efficiency increases. The study of sorting is a fascinating
and intriguing one. However, we hesitate to become too involved at this time.
We will instead develop an ordering routine that works with only a little atten-
tion to efficiency and defer a more sophisticated study of ordering for another
time and place.
If we test every adjacent pair of numbers in a list and find that they are
in order, then we know that the entire list is in order. This is called a "bubble
sort." If we find any adjacent pair that is not in order, then we can direct the
computer to exchange those two elements so that they are in order. If every
time that we make such an exchange, we turn a switch on by letting S = 1,
J:hen we Cl\n oeterinine ::it the enfl of Checking thro11gh the list th::it lln P}(Ch::inge
has been made and that the list might not be in order yet. If after scanning the
entire list we find that switch Sis still zero, then we know that no exchange has
been made and the list must be in order. After the first scan through the list,
we know that the number at the end of the list is the highest or lowest depend-
ing on which order we specify. That is, it is the number that will be there when
the list is finally ordered. Thus we do not need to check the entire list the next
time; we can check one less item. See program <!>RDER.
 Some Topics of Statistics 141

ORDER

94 REM *
THIS PROGRAM ORDERS UP TO 200 NUMBERS
100 DIM LC200)
102
104 REM *READ ANO COUNT DATA
110 LET I = 0
120 LETI=I+l
130 Ir I <= 200 THEN 160
140 PRINT "CURRENT LIMIT IS 200 NUMBERS"
150 ST0P
160 READ L< I>
170 I f L<I> <> .at THEN 120
IBO LET N = I = 1-1
IB4 REM *TURN SWITCH Orr ANO BEGIN SORT
190 LET S = 0
200 LET N = N-1
210 f0R J = I T0 N
-220 Ir L<J> >= LCJ+I) THEN 270
222
224 REM * EXCHANGE ELEMENTS AND TURN SWITCH ON
230 LET St = L<J>
240 LET L<J> = LCJ+t>
250 LET L <J+ I) = SI
-260 LET S = I
270 NEXT J
272
274 REM * CHECK SWITCH
275 REM S = 0 SORT C0MPLETE
276 REM S = I S0RT N0T C0MPLETE
-2BO If S = I THEN 190
2B2
2B4 REM *
THE LIST IS IN 0ROER - PRINT IT
290 r0R X = I T0 I
300 PRINT LCX>J
310 NEXT X
492
494 REM
500 DATA 9B. BO. 73, 92. 11. B4. 83. 79, B7. 73
510 DATA 99, 63. 63. 92. st. 93, 47, 53. B9• 100
520 DATA 9B. 11. 73, .01
530 ENO
RUN
0ROER

100 99 9B 9B 93 92 92 89 B7 B4 B3 Bl BO 79 77 73 73 73
71 63 63 53 47

Note that in line 220 we check for greater than or equal to. What would
happen if we only checked for greater than? If there are two equal numbers in
the list, the switch will always get turned on and cause the routine to be re-
peated endlessly. In program </)RDER we have done two things in the interest
of efficiency. We do not scan that part of the list that we know to be in order,
and we quit when we know the entire list is in order.
There are other things that may be done to improve the efficiency of
program </)RDER. One is to sort "up" the list as well as "down." However,
in spite of these precautions, the general procedure here is satisfactory only for
relatively small amounts of data. If we are to order thousands or hundreds of
thousands of data items, then there are far more efficient algorithms which we
would have to use. The fundamental weakness in the procedure we have used
is that on each pass the computer checks only one less pair of data items than
on the previous pass. Much can be gained by partitioning the data to be ordered
in such a way that only a small fraction of the data need be scanned each time
an item is placed in its final spot in the list. However, such procedures generally
require much more programming effort.
142 Advanced BASIC

If all we want is the median, then there is no need to actually print the
data in order. The middle number for I odd is L(INT(I/2)+1) or L(INT((I+ 1)/2)).
See line 320 of program MEDIAN.

MEDIAN

94 REM *
THIS PROGRAM FINDS THE MEDIAN FOR
95 REM AN ODD NUMBER OF DATA ITEMS
!00 D!M L<200l
i iG LEi ;;; 0
120 LET I = I+ 1
l3U 11• 1 <= ;;!UU IHt.N 160
140 PRINT "CURRENT LIMIT IS 200 NUMBERS"
150 ST0P
160 READ LC I>
170 IF L<I> <> .01 THEN 120
180 LET N = I = 1-1
190 IF N/2 <> INTCN/2) THEN 220
200 PRINT "N EVEN"
210 STOP
220 LET S = 0
230 LET N N-1 =
240 FOR J I TO N =
250 IF L<J> >= LCJ+I> THEN 300
260 LET SI = L<J>
270 LET LCJ) l.CJ+I> =
280 LET L<J+I> = SI
290 LET S = I
300 NEXT J
310 IF S = I THEN 220
312
-320 PRINT "MEDIAN"J LC INTCCI+l>/2> )
492
494 REM
500 DATA 98> so, 73, 92, 77, 84> 83, 79, 87, 73
510 DATA 99, 63, 63, 92, 81, 93, 47, 53, 89, 100
520 DATA 98. 71, 7 3, .01
530 END
RUN
MEDIAN

MEDIAN 81

As written, MEDIAN does not properly account for an even number of

data items.

Problems for Sec. 11-3

1) Modify program MEDIAN to allow for both even and odd numbers of
data items.
2) Modify q'>RDER so that it can be used to arrange in either ascending or
descending order determined by an INPUT request.
3) Modify MEDIAN as in problem 1 and print the largest, smallest, and
average value.
A\ £'"1 .... ..,..,"'! .... ....,"-,, 1.L.V'J
fl(\ vu1~hr\v<- f\·n.~
.... ,.. ..... ,....l ..... ._,. .l.L\A,.l.l.l.IU'.._,.._.., 1 tr-. 1 ()() "'.l1"'rl ftn.rl tho marll!ln
-XJ \...A'VJ.J.\..>.LUIJ'V LU.l.lUVlJ..L .l."-'-"-'".l _,_ .,..,.,,. ...._ •..., ... , "''"'""" ... _ ... ..,,..,,.. ____ _
.....,,.,........_ .._,.. ... ,. .....

5) As written, program q'>RDER is efficient for a set of numbers in which

only the first number in the list is out of order. But suppose only the
last number is out of order. Then the program is slow. We can im-
prove it by inserting an upward sort, taking the number that belongs
at the top of the list up to the top in one pass and then not scanning
that element again. Incorporate this step into the program.
 Some Topics of Statistics 143

11-4 Coefficient of Linear Correlation

Very often people have a set of data consisting of two or more figures for
each object of the study and would like to know if there is a relation between
them. For example, suppose we have test scores for five people for whom we
also have IQ scores, as shown in Table 11-1. We would like to know if one score
is a good predictor of the other score.

TABLE 11-1 IQ's and Test Scores

for Five People

IQ TEST

110 80
105 84
134 92
128 71
92 83

There is a variety of formula and techniques for finding correlations. We

present here a computational formula for finding the degree of linear correlation
between two sets of data, X and Y.
11 11 11
n_L X1Y1 - .L X1 .L Y1
i=I i=I i=I
rxy (11-1)
VxVy

where rxy is the linear correlation coefficient, Vx is the variance of the X data,
and Vy is the variance of the Y data.
Let's look at the correlation coefficient for the data of Table 11-1. See
program Cft>RREL. The correlation is about .03. That may safely be taken to
indicate that there is no correlation between these two sets of data. That is, IQ
is unrelated to the test score. We can see by inspection that no obvious pat-
tern is present.
We can get some idea of how the value of r is affected by various patterns
in the data by simply using Cfl>RREL with a variety of data. We present two
additional runs for your observation.
CEJRREL

94 REM * THIS PROGRAM CALCULATES CEJRRELATIEJN CEJEFFICIENT

100 LET N = =S St =
S2 Tl =
T2 0 = =
110 PRINT "IQ TEST SCEJRE"
120 READ x. Y
130 IF X =0 THEN 220
140 PRINT XJ TABC6lJ Y
150 LET N = N+t
152
154 REM * SUMMATIEJNS DONE IN LINES 160 THREJUGH 200
160 LET S = S + X*Y
170 LET St = St+X
180 LET S2 = S2+Y
190 LET Tl = Tl + Xt2
200 LET T2 = T2 + Yt2
210 GEJTEJ 120
212
144 Advanced BASIC

214 REM * VARIANCES CALCULATED IN LINES 220 THROUGH 270

220 LET Al Sl/N
230 LET A2 S2/N
240 LET Bl Tl/N
250 LET B2 T2/N
260 LET VI SQR( Bl - Al•2 l
270 LET V2 SQR( B2 - A2t2 >
272
260 LET R = < N*S - Sl*S2 l/( CNt2l*Vl*V2 l
290 PRINT
300: CORRELATION n.nuu
310 PRINT USING 300• R
3!2
314 REM

330 DATA 126.71• o. 0

340 END
RUN
CORREL

IQ TEST SCORE
110 60
105 64
134 92
126 71
92 63

C0 RRELAT ION .021

RUN
C0RREL
IQ TEST SCCJRE
134 92
126 64
110 63
105 60
92 71

CORRELATION ,930

RUN
CORREL

IQ TEST SC0RE
I -2
2 -3
4 -5
5 -6

CORRELATION *-1.000

We can see that correlation coefficients range from -1 to .930 in the

sample of runs. It turns out that -1 to 1 is the true maximum range. A coef-
ficient of one indicates perfect correlation. A coefficient of zero indicates no
correlation, and a coefficient of negative one indicates perfect negative correla·
tion. Generally, values between - .40 and .40 are considered to indicate that the
variables are unrelated, whereas for -1 to - .9 and for .9 to 1 the variables are
coi1sidei"ed tv bG extremely clcsoly rc!G.tcd in :i linc:!r f!lsh!cn. The !arge! the
number of data items we use to calculate the value of r, the more reliable its
value. For a small amount of data, the coefficient is more likely to be affected
by a single stray or inaccurate item of data.
Often it may happen that we have more than two sets of data with which
to work. With a slight modification of Cq'>RREL, we can easily find the linear
 Some Topics of Statistics 145

correlation coefficient for columns 1 and 2, 1 and 3, and 2 and 3 for the data
shown in Table 11-2. See program runs labelled WEATHER. The program is
actually CQ'.>RREL with the READ statement modified as shown in each run.
An alternative approach would be to RESTQ'.>RE the data and use the entire
program as a subroutine three times.
TABLE 11-2 Run Table Showing Normal, Record High, and Record Low
Temperatures for 24 Selected Dates of the Year

Normal High Low Normal High Low

32 61 4 32 63 5
46 81 14 53 91 27
33 61 2 34 63 -2
58 90 35 63 90 37
36 68 11 41 86 10
68 97 44 72 98 44
74 102 52 76 98 56
62 89 34 55 89 34
76 97 56 74 94 56
50 84 30 44 72 21
71 97 49 67 93 44
39 68 11 33 60 7

120 REAO x. y, z 120 READ x .. z, y 120 READ z.. x, y

RUN RUN RUN
WEATHER WEATHER WEATHER

x y x y x y
32 61 32 4 61 4
32 63 32 s 63 s
46 Bl 46 14 B1 14
S3 91 S3 27 91 27
33 61 33 2 61 2
34 63 34 -2 63 -2
SB 90 SB 3S 90 3S
63 90 63 37 90 37
36 6B 36 11 6B II
41 B6 41 10 B6 10
6B 97 68 44 97 44
72 98 72 44 9B 44
74 102 74 S2 102 S2
76 9B 76 S6 98 S6
62 B9 62 34 89 34
SS 89 S5 34 89 34
76 9'7 76 56 97 S6
74 94 74 56 94 56
50 B4 so 30 84 30
44 72 44 21 72 21
71 97 71 49 97 49
67 93 67 44 93 44
39 68 39 II 6B II
33 60 33 7 60 7

CORRELATION .944 CO RRELATl0N ·984 C0RRELATHlN .920

For correlation normal to high, we get .944; for normal to low, we get
.984; and for high to low, we get .920. These all indicate a high degree of linear
correlation.
146 Advanced BASIC

If we were to require the correlation coefficients for all pairs of columns

for more than three columns, it is a bit awkward to use the method we used in
program WEATHER. It would be a lot cleaner to read all the data into an array
and manipulate the column subscripts to get all possible pairs. This is left as an
exercise.
Summary of Sec. 11-4
We have presented a computational formula for obtaining the linear cor-
reiation coefficient. This is also referred to as the Pearson r. Values close to
zero ii1di(;ate a lov; dt:gree of linear ;.;ui'i'eiatiuu, whcn:!a~ values vvith absolute
value close to one indicate a high degree of linear correlation.
Problems for Sec. 11-4
1) Write a program to generate 25 pairs of random numbers and compute
the correlation coefficient. What value do you expect? Run the pro-
gram several times.
2) Write a program to calculate the correlation coefficient for the integers
2 through 100 and the number of factors.
3) Do problem 2 for prime factors not including 1.
4) Write a program to find correlation coefficients for all pairs of columns
by first reading the data into a single array as described in the section.
Use the following data:
a b c d
39 12 2 1978
43 8 5 1749
25 4 1 1462
22 4 1 1288
21 11 11 1241
21 7 3 1176
32 10 2 1086
37 7 12 1026
18 2 1 1003
30 10 3 971
5) For the data given, find any or all of the following linear correlations:
(a) aandb a b
(b) a 2 and b
1 1.04631
(c) a 3 and b
(d) a4 andb 2 16.5958
(e) aandlog(b) 3 84.0632
4 266.206
5 651.343
6 1353.51
7 2512.56
8 4294.3
9 6890.5
10 10519
11 15424
 12
SIMULATION A
GAMES

12-1 Introduction
'I'he ability of the computer to store information, generate random num-
bers, and make decisions makes it well suited for simulations of all kinds. Com-
puters can be programmed to play games. Programs can be written to simulate
business activity, social phenomena, and numerous activities in the physical
sciences. Computers can be used to conduct gambling enterprises, schedule
classes, and manage production schedules. Some situations are dealt with by
having the computer investigate all possible alternatives. Other situations are so
complex that a procedure must be found that enables the computer to make a
best reasonable decision which may not be the best possible decision. For
example, it is possible to write an unbeatable tic.tac-toe program. However, the
game of chess allows so many possible sequences of moves that it is impossible
to write a program for existing computers to investigate them all.
'l'he purpose of this chapter is to present a few examples of simulation and
to suggest areas for further investigation.
12-2 Lines at the Bank
As the manager of a new bank branch, you are interested in knowing what
to expect in the way of teller requirements. You are presented a bank that has
five windows. As a preliminary trial, you make the following estimates and
assumptions:
1. Assume that there is always a customer waiting with a four-minute
transaction when the bank opens at 9 AM.
2. Always open two windows at 9 AM.
3. Customers will tolerate only as many as 10 persons per line; thus when
all lines are full, a new window must be opened.
147
148 Advanced BASIC

4. During every minute of the day one or two or no customers will enter
with equal probability.
5. Every customer after the first has transactions that will last one, two,
or three munutes with equal probability.
6. A new customer upon entering the bank goes to the line with the
fewest persons. In case of a tie, the customer takes the line closest to the door.
'l. The bank closes at 3 PM.
Our joh now is to wr·ite a c•)mputer program that reacts to each of the
restrictions above and keeps track of the day's business. One possible simulation
is to use two arrays-one from the customer's point of view and one from the
window's point of view. Both arrays have five columns, one per window. The
customer array has 10 rows to allow as many as 10 people in line. The window
array has four rows. The first row contains a '1' to signify that the window is
open and a 'O' to signify that the window is closed. The second row contains the
number of people in line for that window. The third row contains the number
of people served since the run began. The fourth row contains the number of
minutes that the window has been open. The customer array contains the
number of minutes each customer in line will take.
In program LINES, all loops F</>R Wl = 1 T</> 5 scan all five windows.
Lines 200 and 210 open two loops to keep track of time. H stands for hours
and Ml stands for minutes of that hour. Line 240 looks to see if a window is
open, and line 250 adds one minute to open time for the window. Line 260
looks to see if anyone is in line at the open window. Lines 390 through 500
search for the line having the fewest people. Numerous other relevant comments
appear in the REM statements of program LINES.
LINES
94 REM * THIS PR0GRAM SIMULATES LINES AT
95 REM TELLER WINDOWS 0, A BANK
96
97 REM *
ARRAY L IS THE CUSTOMER ARRAY
98 REM ARRAY R IS THE WINDOW ARRAY
100 DIM LCI0,5>. RC4,5J. A$C4l
I 05 RAND0M I ZE
110 MAT READ A$
120 MAT R = ZER
130 MAT L = ZER
132
134 REM *
0PEN WIND0WS I AND 2 BY PLACING A
135 REM I IN ROW I OF COLUMNS I AND 2 IN ARRAY R
140 LET RCl,ll = RCJ.2> = I
142
144 REM *
THE FIRST CUSTOMER ENTERS WITH A
145 REM FOUR MINUTE TRANSACTION
150 LET LCI, I> = 4
160 LET RC2, I> = I
162
164 REM * SET UP L00PS Tel KEEP TRACK lff TIME
200 roR H 0 TO 5
210 rnR Ml I TO 60
220 LET T = 60*H + Ml
222
224 REM * THIS L0BP ADJUSTS TIME FOR
225 REM CUST0MERS AND WIND0WS
230 ,0 R WI = I T0 5
240 ff RC 1,w1 > = 0 THEN 350
250 LET RC4,Wll = RC4,Wll+I
 Simulation and Garnes 149

260 I' RC2.Wll = 0 THEN 350

262
264 REM * THERE IS A LINE
265 REM REDUCE 'IRST PERSON'S TIME
270 LET LCl.Wl> LCl.Wll-1
280 I, L<l.Wll <> 0 THEN 350
282
284 REM *
ENO 0' TRANSACTION - MOVE PEOPLE UP IN LINE
290 ,OR P= I TO RC2.Wll-l
300 LET LCP.Wll = LCP+l.Wll
310 NEXT P
320 LET LC RC2.WI l.WI l = 0
322
324 REM * ONE MORE TRANSACTHlN - ONE LESS PERS0N
330 LET RC3.Wll RC3.Wll+I
340 LET RC2.Wll =
R<2.Wll-l
350 NEXT WI
352
354 REM *ENTER o. I OR 2 CUSTOMERS
360 LET C = INT< RNDC-1>*3 l
370 rQRCI ITOC
372
374 REM * SELECT TRANSACTION TIME
380 LET TI = INT< RND<-1>*3+ I l
382
384 REM * NOW 'IND THE SHORTEST LINE WITH
385 REM LESS THAN TEN PECJ PLE
390 LET N = R<2.ll
400 LET NI = I
410 ,OR WI = 2 TO 5
420 ff R <I• WI l = 0 THEN 46 0
430 I, R<2.Wll >= N THEN 460
432
434 REM *
CURRENT LINE IS SHORTER
435 REM SAVE WINDOW # ANO # 0, PEOPLE
440 LET N = RC2.Wl l
450 LET NI = WI
460 NEXT WI
462
500 I' N <= 9 THEN 590
510 ,OR WI = I TO 5
520 If RCl.Wll = 0 THEN 570
530 NEXT WI
532
540 PRINT "ALL WINDOWS ,ULL AT"I TJ "MIN 1JTES"
550 MAT PRINT RJ
560 ST0P
562
564 REM * OPEN ANOTHER WINDOW
570 LET RCl.Wll I
580 LET NI = WI
590 LET RC2.Nll RC2.Nll+I
600 LET LC R<2.Nll.Nl l = Tl
610 NEXT Cl
620 NEXT Ml
622
6 30 GCl SIJB 800
640 NEXT H
650 ST0P
652
800 PRINT TAB< 11 l l "AT THE ENO O'"J T/60J "HOURS"
810 PRINT WINDOW NO. ClNE TW0 THREE rnuR '1VE"
820 ,0R I = I T0 4
830 PRINT A$CilJ
840 ,OR J = I TO 5
850 PRINT TA9(15+CJ-ll*6lJ R<I.JlJ
860 NEXT J
870 PRINT
880 NEXT I
890 PRINT
900 RETURN
972
150 Advanced BASIC

974
980
REM
OATA
..
"!=OPEN O= CLOSED", .. PEO PL.E IN LINE"
990 DATA PEOPLE SSRVED 0 ~ " MINUTES OPENED"
999 END

RUN
LINES

AT THE END Or I H0URS

WINDOW NO. 0NE TW0 THREE FOUR FIVE
1=0PEN O= CLIJ SED I I 0 0 0
PE~PLE J,\'I t..J.1'11:. 4 4 0 0 0
PE0PLE SERVED 31 25 0 0 0
M!~!!JTE:S 0?Er-!E:!} 60 60 0 0 0

AT THE END Or 2 HOURS

WIND0W N0. 0NE TW0 THREE FOUR F"IVE
1=0PEN O=CL0SED I I 0 0 0
PE0PLE IN LINE 8 8 0 0 0
PE0PLE SERVED 59 52 0 0 0
MINUTES 0PENED 120 120 0 0 0

AT THF. END 0r 3 H0URS

W!ND0W NO. 0NE TWO THREE FOUR FIVE
1=0PEN O=CL0SED I I 0 0 0
PE0PLE IN LINE 9 9 0 0 0
PE0Pl.E SERVED 92 79 0 0 0
MINUTES OPENED ISO 180 0 0 0

AT THE END OF 4 HOURS

WIND0W N0. 0NE TVJO THREE F"CJUR FIVE
!=OPEN O=CUlSED I I I 0 0
PEOPLE IN LINE I 0 0 0 a
PEOPLE SERVED 122 108 23 0 0
MINUTES OPENED 240 240 50 0 0

AT THE END 0F 5 HOURS

1 0NE TWO THREE FOUR F"IVE
11IND0W Nlil•
!=OPEN O=CLO SED I I I 0 0
PEOPLE IN LINE I I I 0 0
PEOPLE SERVED 148 I 31 36 0 0
MINUTES OPENED 300 300 110 0 0

AT THE END 0F 6 HOURS

WINDOW NO. 0NE TW0 THREE r01JR FIVE
l=OPEN O=CLOSED I I I 0 0
PEOPLE IN LINE I I 0 0 0
PEOPLE SERVED 178 153 51 0 0
MINUTES 0PENED 360 36D 170 0 0

The run shows that ten minutes into the fourth hour the third window
was opened. We can see that at the end of the day there was one person waiting
at window 1, which had served 178 people during the day. Window 2 also had
one person in line, but had served only 153 customers. Similarly, the third win-
dow had served 51 customers and left none in line at the end of the day's
business.
While the results of LINES provide some interesting information, there are
virtually unlimited possibiiities for extracting more information and for testing
changes in the original list of assumptions and estimates.
There have been no provisions for closing a window. We might want to
close a window due to lack of activity or to allow employees time to lunch. In
practice, bank tellers close windows but service those customers already in line.
 Simulation and Games 151

Program LINES does not provide for having a window closed with people
standing in line. We could use a "-1" in the window array to signify this
condition.
It is fairly obvious that the assumption of random arrival of customers
is an oversimplification of the true pattern. Clearly, large numbers of people
conduct bank business during their lunch hour. Not only does the arrival of
customers vary during the day, but it varies with the days of the week and of
the month. Fridays tend to be heavier, and the first of the month is heavy.
The limit of 10 persons per line was thus arbitrary and perhaps unrea-
sonable as an absolute limit. The program could be modified to open a new
window when all the lines contain 10 customers, but when all windows are
open and all lines contain 10, then we should allow the lines to grow.
In practice, a new customer generally steps into the shortest line, but
the customer is not obligated to stay there. Thus we could make provision in
our simulator for customers to move to a faster moving line. (We know from
experience that fast-moving lines immediately become slow-moving lines when
we step into them and slow-moving lines immediately become fast-moving lines
when we step out of them.) This points up the fact that although a customer
enters a line based on the number of customers in it, what he really cares about
is how long he has to wait. We could add a row to the window array giving
maximum waiting time so far. We could cause a certain waiting time to trigger
opening a new window.
As the program is written, when a new window opens only new customers
may enter that line. Generally when a new window opens, a whole bunch of
people swarm into the new line. Sometimes the new line quickly exceeds the
old lines in length. We could modify the simulation of LINES to allow an
orderly shift of customers from all lines to a newly opened window.
It is easy to see that we could go on and on at great length, making our
simulation program more and more like what we believe to be the real life
activity. What about drive-in windows, automatic tellers, etc.?
Based on many runs of simulations like this, a business person is in a better
position to make decisions about hiring, opening hours, business procedures, and
other aspects of management than he would be without the computer. Once we
are convinced that a simulation is realistic, then we can experiment with innova-
tive procedures using computer results to warn us of poor changes without
actually having to use customers as guinea pigs.
Similar simulations could be set up for toll booths, grocery store check-
outs, post offices, gas stations, and stores and businesses of all kinds.
Summary of Sec. 12-2
We have looked at a much simplified set of rules for lines at the tellers'
windows of a bank and written a program to imitate the activities of bank
customers for a sample business day. Random numbers are used to simulate the
random nature of the arrival of people at the bank and the random nature of
transactions. Arrays have proved very useful for keeping track of many of the
activities of our banking model. We recognize that simulations usually must be
simplifications of the real activity under study.
152 Advanced BASIC

Problems for Sec. 12-2

The possibilities for making changes in program LINES and developing
other models are so varied and so numerous that no attempt will be made to
enumerate a specific set of problems. Instead you should select one or more of
the improvements outlined in this section and implement them, along with any
changes not mentioned that you would like to make. You should obtain several
runs of your final program to get a range of results. Experiment with differing
transaction times, differing rates of arrival of customers, and so forth.

12-3 Magic Squares

Magic squares have provided entertainment and been the source of wonder
for more than a thousand years. They have been considered to have magic
powers and therefore have been used to keep away evil spirits. Magic squares are
square arrays of numbers (usually integers) so arranged that all row sums equal
all column sums and these equal each of the sums of the elements of the two
main diagonals.
Of course, this feat can be accomplished by simply entering the same
number in each position of the array, but this is trivial and of little interest. The
simplest magic square of real interest is the following three-by-three magic square:
8 1 6
3 5 7
4 9 2
Note that all integers from 1 through 9 have been used and that the magic sum is
15. It turns out that while we can rotate this to get a total of eight different
positions, there is no other arrangement of these integers that will produce a
magic square, even though there are 45,360 possible different arrangements.
The magic sum can be found for integers 1 through n 2 by the formula,
n3 + n
s = ---
2
An odd order magic square from 3 up can be generated by a procedure
called the De la Loubere method. For a 5 X 5 magic square, this method pro-
duces only one magic square, but millions are possible. The De la Loubere
method uses the integers 1 to n 2 for an nth-order magic square and may be
described with the following set of rules:
1. Begin by entering a "1" in the center column of the first row.
2. Always move diagonally up one and to the right one and enter the next
larger integer there unless the move (a) is the move immediately following an
entry that is a multiple of the order of the magic square, in which case the new
number goes directly beneath the previous one, or, (b) takes us out the side of
the square, in which case the new number goes to the extreme left of the new
row, or, (c) takes us out the top of the square, in which case the new number
goes to the bottom of the new column.
3. Proceeding in this way, n 2 should always be placed in the middle
column of the bottom row, and we know it is time to stop.
 Simulation and Games 153

In program MAGIC for n = 5, the middle column is determined in line

160, and the row is set to 1 in line 170 so that the first entry in line 190 follows
rule 1 above. Line 210 checks to see if the condition in rule 2a has occurred.
Line 250 checks for the condition of rule 2b, and line 300 checks on rule 2c.
Line 200 determines the stopping point as per rule 3.

MAGIC

94 REM *THIS PR0GRAM GENERATES 0DD 0RDER

95 REM MAGIC SQUARES BY THE DE LA L0UBERE
96 REM METH0D
100 DIM M(47.47)
110 PRINT "WHAT 00D SIZE"J
120 INPUT N
130 IF N/2 = INTCN/2) THEN 110
140 MAT M = ZER<N•N>
150 LET Cl = 0
-160 LET C = INT(N/2)+1
-110 LET R = I
180 LET Ct " Ct+!
--190 LET M<R. C > = CI
-200 IF Cl =
Nt2 THEN 330
--210 IF Cl/N <> INT<Cl/N) THEN 240
220 LET R " R+t
230 G0T0 180
240 LET C C+l
-250 IF C <= N THEN 290
260 LET C = I
270 LET R = R-1
280 G0T0 lf!O
290 LET R = R-1
--300 IF R > 0 THEN 180
310 LET R = N
320 G0T0 180
330 PRINT
340 LET T =
0
342
344 REM *
ADD 0NE COLUMN T0 FIND MAGIC NUMBER
350 F0R I = I T0 N
360 LET T = T+M Cl• l>
370 NEXT I
372
380 PRINT "MAGIC NUMBER IS"J T
390 PRINT
400 MAT PRINT MJ
410 END
RUN
MAGIC

WHAT 0DD SIZE?5

MAGIC NUMBER IS 65

17 24 8 15

23 5 7 14 16

4 6 13 20 22

10 12 19 21 3

II 18 25 2 9

As written, this program will arrange and print magic squares up to 4 7 by

47. Of course, we can't print 47 numbers across the page. So we would have to
do some rearranging to make the results easy to read.
154 Advanced BASIC

Problems tor Sec. 12-3

1) The magic squares of this section used 1 as the first number. It is easy
to show that beginning with any integer will also produce a magic
square. Modify program MAGIC to allow beginning with any integer.
Be careful about rule 2a.
2) Another way to generate odd order magic squares may be described as
follows. Place the first number in the array position directly beneath
the central element, and proceed down one row and to the right one
column unless this move: (a) takes you both out the bottom and out
Lire righL side of the square, in which case the new entry goes in posi-
tion (2,n), or (b) takes you out the bottom of the square, in whicl:v
case the new entry goes to the top of the new column, or (c) takes you
out the right side of the square, in which case the new entry goes to the
extreme left in the new row, or (d) the new location is already occupied,
in which case the new entry goes in the second row directly below the
previous entry. (Note that this may take you out the bottom.) Write
a program to generate this type of odd-order magic square.
3) Squares of the type described in problem 2 can also be generated by
beginning with any integer. Modify your program for problem 2 to
do this.
4) There are 880 different 4-by-4 magic squares using the integers 1
through 16. One of them can be generated by the following simple
procedures: MAT READ the integers 1 through 16 into a 4-by-4 array
and then make these exchanges:

A(l,1) ~ A(4,4)
A(2,2) ~ A(3,3)
A(3,2) A(2,3)
A(4,1) A(l,4)

Write a program to do this.

12-4 Games
There are hundreds of games which may be played with computers. There
are games played with cards, dice, dominoes, and numbers. There are board
games and two- and three-dimensional tic-tac-toe. Programs have been written
to play casino gambling games and to simulate slot machines. There are programs
which simulate horse races. Programs can be written to play word games such as
Hangman or Geography using strings and files. Using computer files, game-
p!aying prog:rf!.ms ~2n be dev!setl i,x,rhich mcdify st!'.'atcgy depending en ccnsc··
quences of previous decisions. It is not the purpose of this section to present
any comprehensive or systematic study of games or game strategy. Rather, it is
the purpose of this section to arouse the sleeping giant of gamesmanship that
may lie within the reader by exploring two examples.
 Simulation and Garnes 155

Battle of Numbers
The game, Battle of Numbers, begins with two integers, such as 63 and 11,
where one should always be somewhat larger than the other. Two players take
turns subtracting an integer in the range 1 to 11 from 63 and subsequent new
remainders. The last player to subtract loses. The feature that makes this game
intriguing is that usually the first player may assure a win by applying proper
strategy on the very first move. Working out the strategy is fairly straightforward
if we look at the last few moves. Suppose it is your turn, and you may subtract
up to 11 from 15. If you subtract 2 leaving 13, you win because your opponent
must leave you a number in the range of 2 to 12, subtracting in the range 11 to 1.
Now it is your turn again and you can be assured of leaving your opponent a "1,"
which he must subtract, thereby losing. Now one pair of plays earlier you can
assure yourself of leaving your opponent with 13 by leaving 25 and before that
37, and so forth. That is, you want to leave (11 + 1) i + 1, or one more than
an integral multiple of one more than the largest number you are allowed to
subtract. For subtracting in the range 1 to a, then leave (a + 1) i + 1. This
means that if the human player goes first in competition with the computer, the
human can always win except when the original larger number is one more than
an integral multiple of the largest subtractable number. However, one slip-up,
and the computer can always win.
Thus in our game with 63 as the starting total and 11 as the maximum
subtractable integer, divide 63 by 12 to get 5 as the integral quotient. Since 5
times 12 is 60, if we are faced with 61 we can't win and should subtract some
random integer. But since we are not faced with 61, we want to leave our
opponent with 61 by subtracting 2. No matter what our opponent does, we will
leave 49, then 37, then 13, then 1. The calculations and testing here are done in
lines 370 through 410 of program BATTLE. Note that lines 280 through 320
assure that the human player inputs a number in the range 1 to a. If you would
like to play the game, but don't want to type the program, many timesharing
systems include the game under the name BATNUM.
BATTLE

94 REM * THIS PROGRAM PLAYS BATTLE 0r NUMBERS

100 PRINT TAB<ISJJ "BATTLE 0r NUMBERS"
I 05 RAND0M I ZE
110 PRINT "00 Y0U KN0W THE RULES"J
120 INPUT AS
130 ff A$ = "YES" THEN I BO
140 PRINT "WE TAKE TURNS SUBTRACTING AN INTEGER IN THE"
150 PRINT "INTERVAL I T0 S0ME NUMBER A rR0M AN0THER NUMBER"
160 PRINT "B WITH THE DffrERENCE LEAVING B r0R THE NEXT"
1'70 PRINT "TURN. THE LAST PERS0N T0 TAKE LlilSES."
!BO PRINT "HERE WE G0. ***"
IB2
IB4 REM * SELECT RANGE AND STARTING NUMBER
!BS REM THE RESTRICTING CONSTANTS HERE ARE ARBITRARY
190 LET A= INT< RNDC-1>*14+7 J
200 LET B = INTC RND<-1>*77+41 >
210 PRINT
220 PRINT "INTERVAL = I T0"J A
230 PRINT
240 PRINT "STARTING TOTAL IS"J B
250 PRINT
156 Advanced BASIC

260 PRINT "YOU G0"J

270 INPUT P
272
274 REM * CHECK FOR AN INTEGER IN THE LEGAL RANGE
275 REM LESS THAN 0R EQUAL T0 THE REMAINING T0TAL
280 IF P <> !NTCP> THEN 310
290 IF INT«P-1 )/Al <> 0 THEN 310
- 300 IF P <= B THEN 330
{ 310 PRINT "ILLEGAL MOVE"
320 GOTO 250
330 LET B =
B-P
340 IF B > 0 THEN 370

360 STOP
,.,. , .........
,n .. t,,, ..
,,.... t ,"-

-{~~g
TM'T'' ~,I\•
'

LET C = B - C CA+ll*l+I l
Ir C > 0 THEN 420
400 LET C = INT C RNDC-1 HA+ I l
410 Ir B-C < 0 THEN 400
420 LET B =
B-C
430 PRINT "I TAKE"J C
440 Ir B =0 THEN 470
450 PRINT "LEAVING A TaTAL Or"J B
460 GCJTa 250
470 PRINT "*** YClU 'A IN ***"
480 END

RUN
BATTLE

BATTLE Clr NUMBERS

DIJ YOIJ KNOW THE RULES? YES
HERE WE GO. ***
INTERVAL = I TO 20

STARTING TOTAL IS 68

YalJ GO? 4
I TAKE 6
LEAVING A TOTAL aF 58

Y0!J G0? 15
I TAKE 7
LEAVING A TOTAL OF 36

YOU G0? 14
I TAKE 6
LE.AV ING A TOTAL OF 16

Y0U GCJ? 15
I TAKE I
*** YCJU 'NIN ***

The Knight's Tour

The game of chess is played on a square board having 64 smaller squares,
eight on a side. The various pi(;!ces belonging to the two sides are assigned specific
moves~ The knight moves !!! an !..-sh2.ped path, rr!cving en~ ~qu!!r~ in ~ny
direction and two squares in a direction perpendicular to the first move. Thus
from a position near the center of the chess board, a knight may move to any of
eight possible positions. If the knight occupies position (3, 4), then he may
move to any one of the following: (4, 6), (4, 2), (2, 6), (2, 2), (5, 5), (5, 3), (1, 5)
or (1, 3). In general, if the knight occupies position (r, c), then he may move to
 Simulation and Garnes 157

any of the following: (r + 1, c + 2), (r + 1, c - 2), (r - 1, c + 2), (r - 1,

c - 2), (r + 2, c + 1), (r + 2, c - 1), (r - 2, c + 1), or (r - 2, c - 1),
unless the new position is off the board. An ancient and intriguing challenge is
to move the knight about the board in such a way that it visits all 64 squares of
the chess board exactly once.
This is a difficult feat, known as the Knight's Tour, but it can be done. We
will here contrive only to select moves randomly until the knight reaches a
dead end due to the fact that all reachable squares have already been visited. We
use an 8-by-8 array B to simulate the board. Initially all entries are zero to
indicate open positions. We will place the move numbers in the squares as the
knight moves about the board. The 8-by-2 array U stores all eight possible
moves from the present position as described in the previous paragraph. These
eight moves are scanned for legal use in lines 220, 230, and 240 of program
T<!>UR. Note that IN'f ((Rl - 1)/8) = 0 in line 220 is equivalent to (O<Rl

T0UR

94 REM * THIS PR0GRAM CARRIES 0UT A RAND0M

95 REM KNIGHT'S T0UR T0 DEAD END
100 DIM Bca.0>.Tc2.0>.u cs.2>
110 MAT B = ZER
120 MAT READ U
130 LET M =I
140 PRINT "BEGIN WHERE"J
150 INPUT R, C
160 LET BCR.C> = M
170 MAT T = ZER
172
174 REM * Kl C0UNTS THE NUMBER 0r LEGAL M0VES
160 LET Kl = 0
162
164 REM * ENTER ALL LEGAL MOVES IN T ARRAY
190 r0R T = I T0 B
200 LET RI= R + UCT,J>
210 LET Cl = c + UCT.2>
220 Ir INTC CRl-l>IB > <> 0 THEN 260
230 Ir INTC CCl-ll/6 > <> 0 THEN 260
240 Ir B(RJ,CI> <> 0 THEN 260
250 LET Kl = Kl+!
260 LET TCJ,KJ> RI
270 LET TC2•Kl> = Cl
260 NEXT T
262
290 Ir Kl = 0 THEN 350
292
294 REM * SELECT A LEGAL MOVE AT RANDOM
300 LET T = INTC RNDC-l>*Kl+I >
310 LET R = TCJ,T>
320 LET C = TC2.T>
330 LET M = M+I
340 G0Tlil 160
342
350 PRINT "G0T T0"J M
360 PRINT "PRINT IT"J
370 INPUT A$
380 Ir A$ <> "YES" THEN 420
390 MAT PRINT BJ
392
394 REM
400 DATA 1,2, J,-2, -1•2• -J,-2
410 DATA 2, 1, 2,-J, -2, 1, -2,-1
420 END
RUN
T0UR
158 Advanced BASIC

BEGIN WHERE?4,4
G0T T0 41
PRINT IT? YES
0 6 15 0 0 24 39 0
14 II 0 7 36 0 0 25
0 16 9 12 0 26 23 40
10 13 16 6 37 D D
17 0 5 36 27 22 41 0
D 0 2 19 32 35 26 0
0 0 0 4 0 30 21 34
0 3 0 31 20 33 0 29

AND R1<9). If a move is found to be legal, then we enter it into the T array.
When all legal moves are in the Array T, Kl is the number of legal moves. If
Kl is zero, then the knight has reached a dead end and we may print the tour
or not. We present a flowchart in two parts. Figure 12-1 details the sorting out
of legal next moves. Figure 12-2 shows where a legal move chosen at random
is incorporated into the tour.

Summary of Sec. 12-4

We have seen programs to play Battle of Numbers and simulate the
Knight's Tour. In the first case there is a guaranteed strategy which we exploit
in our program. In the second case we have not employed strategy of any kind
but merely progress from step to legal step at random with no procedure for
maximizing results.
Projects for Sec. 12-4
Some of the projects listed will require considerable study and planning
before the actual coding of the program takes place. Be sure to allow a reason-
able amount of time should you attempt any of the longer projects. Some of
the solution programs can be very long indeed. The reader needn't feel limited
to projects proposed here.
1) Modify the game of Battle of Numbers so that the last person to take
away wins. Be sure to change the computer's strategy.
2) Write a program to play the game of Nim.
3) Write a program to play Tic-Tac-Toe. First decide whether or not you
want an unbeatable program.
4) Write a program to play three-dimensional Tic-Tac-Toe.
5) Write a program to lengthen the knight's tour by backing up every time
the knight hits a dead end.
6) Write a program to pl:ty th~ gam<:> of Geogr!!p!!y using st!ings :!nd fil::!~.
In this game two players take turns naming places where the first
letter of the new place must be the last letter of the last place named.
7) Write a program to make the computer the dealer in a blackjack game.
8) Write a program to play craps.
9) Write a program to simulate a Roulette-wheel.
 Start
T,0'UR

MAT B=Z:l
Clear the b:~J

M=1
Start move counter

B(R,C) = M
Place move
on the board

MATT =ZER
Clear possible
move array

Find all possible

moves and select
one at random
(see Fig.12-18)

M=M+1
Bump move
counter

Figure 12-1 Flowchart for keeping track of Knight's Tour.

 K1=0
Set up legal
move counter

R1=R+U(T,1) F0R T = iT,08

Cl=C+U(T,2) Check 8
Find next possible potential moves
move

v ••
1c.:i

Yes

Kl=Klg+
Bump legal
move counter

T(1,K1l=R1
T(2, K1)= C1 NEXT T
Enter point as
possible move

Print move number

Yes and offer to
print ihe whole
touri then quit

T =INT (RNDH) *Kl+ 1)

Select a move at random END
from legal moves

Return to main
flowchart at
this point

Figure 12-2 Flowchart for selecting moves at random for Knight's Tour.
 APPENDIX A
ASCII Character Set Printable on Model 33

C0DE CHARACTER C0DE CHARACTER

32 64 i
33 65 A
34 66 B
35 67 c
36 $ 68 D
37 % 69 E
38 & 70 F"
39 71 G
40 72 H
41 73 I
42 74 J
43 *
+ 75 K
44 76 L
45 77 M
46 78 N
47 I 79 0
48 0 80 p
49 I 81 Q

50 2 82 R
51 3 83 s
52 4 84 T
53 5 85 u
54 6 86 v
55 7 87 w
56 8 68 x
57 9 89 y
58 90 l.
59 91 c
92
60
61 93 '
l
62 94
63 95

N0TE SPECIAL CHARACTERS I

LI NEF"EED 0R Cl'RL J 10
F"0RMF"EED 0R CTRL L. 12
RETURN 0R CTRL M 13
SPACE 32

161
 Summary of flowchart Shapes

Used for beginning and ending of program.

( Terminal )

Indicates data entered into the computer or results

returned by the computer.
Input
Output READ MAT READ READ#
PRINT MAT PRINT READ:
INPUT MAT INPUT WRITE#
WRITE:

Indicates that a decision is being made.

IF xxxxxx THEN yyy

Indicates a sequence of program statements not in-

Predefined cluded in the flowchart. May be used for Gl;l>SUB
process statement.

Connector. Indicates transfer from one statement to

another other than the next higher numbered
statement in the program. N matches another N
elsewhere in the same flowchart.

Used for anything not already specified.

Operation
Opens loop NEXTX
Closes loop LET
RETURN
STIJ>P

162
 APPEND IX C
Summary of Statements in BASIC

NOTE: Not all statements which appear in this appendix will run on all systems
and the list here does not cover every statement for some systems.

END It is the highest numbered statement of every BASIC

program. It is optional on a few systems and re-
quired on most.
PRINT Prints values of variables, calculated values, and
literal expressions inside quotes. Spacing is con-
trolled by commas, semicolons, and TAB. More
spacing functions are available on some systems.
PRINT USING n Prints according to format specified in line n.
Specifies printing for PRINT USING statements.
READ Enters values stored in DAT A statements into vari-
ables named in the READ statement. All legal
BASIC variables (string and numeric) may be read
in a single READ statement by separating them
with commas.
DATA Stores values for READ statements. Items of data
must be separated by commas. Some systems re-
quire that strings be in quotes.
G</JT(/J n Names n as the next line number to be executed by
the computer.
(/JN X G</>T(/J ni,n 2 ,n 3 , Computed G</JT</J goes to the line number in the xth
etc., or position in the list of line numbers ni, n 2 , n 3 , etc.
G</JT</J X </JF n 1 ,n 2 ,n 3 , If available, one of these should work. They are
etc., or not interchangeable.
G</JT</J n 1 ,n 2 ,n 3 , etc.,
</>N X
LET Assignment statement. The word LET is optional on
many systems. Stores the value on the right of an
equals sign in the variable named on the left. May
be used to assign string variables. Multiple assign-
ment is available on most systems.
REM Permits the programmer to remark upon the program
in the program itself without affecting the pro-

163
164 Appendix C

gram operation. Some systems allow' to serve the

same purpose.
IF THEN n Tests the truth of an algebraic sentence placed be-
tween the IF and the THEN. Sends the computer
to line n if the sentence is true. Control passes to
the next line if the sentence is false.
F</>R X =A Tc/> B Opens a machine loop with first value for X at A, last
STEPC number B, and increment C. If C is omitted, the
~tep defat!lts to an increment cf 1.
NEXT X Closes machine loop, Sends the ~ompnter to the
corresponding F<,l>R statement to increment and
test X.
G</>SUB n Sends the computer to a subroutine beginning at line
n. Upon executing a RETURN statement, the
computer returns to the line immediately follow-
ing G</>SUB n.
</>N X G<,l>SUB ni,n 2 ,n 3 , Computed G</>SUB goes to the subroutine beginning
etc., or at the xth line number in the list. Upon executing
G<,l>SUB X <,l>F ni,n 2 ,n 3 , a RETURN statement, control goes to the line
etc., or immediately following this statement. If avail-
G<,l>SUB X <,l>N ni,n 2 ,n 3 , able, one of these should work. They are not
etc. interchangeable.
RETURN Closes all subroutines.
DEF FNA(X) Program-defined function. The letter pair FN desig-
nates that a function is called for. The function
name is A, and the argument is X. Any letter of
the alphabet may be used. Some systems permit
multiple arguments separated by commas.
DEF FNA(X) Opening line of a multiple-line program-defined
function. Several arguments may be permitted,
separated by commas. A value must be assigned
to FNA in the lines to follow.
FNEND Closing statement of a multiple-line, program-defined
function.
ST</>P Execution of the ST</>P statement causes termination
of the RUN at that point.
DIM A( ),B$( ).... Declares dimensions for one- or two-dimensional
numeric arrays or string arrays or both. One
number is required in the parentheses for a list
and two numbers separated by a comma are
required for a two,dimensional array.
INPUT Same as READ except that data is to be typed on the
keyhoarci oi the remote terminal.
REST</> RE Restores all data in the program. The next item of
data to be read will be the very first data item in
the program.
REST</> RE* Restores numeric data only.
REST<,l>RE$ Restores string data only.
 Appendix C 165

REST<J>RE n Restores all data from line n on.

CHANGE A$ T<f> A Stores the ASCII code of the characters of the string
A$ in the array A with the length of the string in
characters stored in A(O).
CHANGE A T<J> A$ Stores a string in A$ with length specified in A(O)
and characters determined by the ASCII code
stored in the array elements of the A list.
RAND<!> MIZE Causes the random numbers generated in successive
runs of the same program to vary.

MATRIX INSTRUCTIONS

MAT READ Enters data into numeric and string arrays. Several
arrays can be read in the same MAT READ state·
ment by separating the array names with commas.
MAT PRINT Prints the array(s) listed, separated by commas or
semicolons. The delimiter used specifies spacing
for the preceding array. Numeric and string arrays
are allowed.
MAT INPUT Enters data into an array (string or numeric) from the
keyboard. Some systems allow more than one
array listed here; others do not.
MAT C =A* B Enters the produc t of A and B into array C.
MAT A B + C Enters the sum of B and C into array A.
MAT A B C Enters the difference of Band C into array A.
MA1' A (K)*B Multiplies each entry of B by the scalar K and enters
the result into A.
MAT A= ZER Creates the zero matrix (fills each entry of A with
zero). ZER may be followed by redimensioning
specifications in parentheses.
MAT A C<f>N Fills each element of A with 1. C<!>N may be fol-
lowed by redimensioning specifications in
parentheses.
MATE IDN Forms the identity matrix E. E must be square. All
elements with equal row and column numbers are
1 and all other elements are 0. IDN may be fol-
lowed by redimensioning specifications in
parentheses.
MAT X = INV(A) Finds the inverse of A and enters it in X (if it exists).
MAT A = TRN(B) Fills A with the transpose of B.

FUNCTIONS

SQR(X) Computes the non-negative square root of X. X must

be non-negative.
ABS(X) Computes the absolute value of X.
SGN(X) Returns the value 1 for X positive, 0 for X equals
zero, and -1 for X negative.
i66 Appendix C

!N'l'(X) Returns integer part of X. For some systems this is

the mathematically greatest integer function. For
others, the computer simply chops off the digits
to the right of the decimal point. (The results
are the same for non-negative numbers.)
RND(X) Generates a random number. In some systems the
set of random numbers accessed is determined by
the value of X. Some systems generate the same
set cf numbers each time the progra1n is run,
wherP.a~ others provide a. different set 3nd still
others provide an option. See RND below.
SIN(X),C<{>S(X),TAN(X) Computes the sin, cos, or tan of X, where X must be
in radians.
ATN(X) Computes the arctan of X. ATN(X) is in radians.
The program must be written to determine the
correct quadrant for the result.
1'/>G(X) Computes the logarithm of X using base e.
EXP(X) Computes the number whose L!/>G base e is X.
TAB(X) Moves the printing mechanism to the (X + l)st posi-
tion of the carriage unless the printing mechanism
is already past that point, in which case there is
no effect.
ASC( ) Returns the ASCII code for the character placed in
parentheses.
LEN(A$) Returns the number of characters in the string A$.
EXT$(A$,I,J) String extract function. Isolates a substring in A$
from the Ith to the Jth character inclusive.
NUM Returns the number of elements typed in response to
the most recent MAT INPUT statement executed
in the program.
DET Returns the determinan t of the most recent matrix
for which the inverse has been found with the
MAT INV ( ) statement. No argument required.
RND Returns a random number. The numbers will be the
same on successive runs of the program if the
RAND!/>MIZE statement is not present in the pro-
gram and different on successive runs if the
RAND!/>MIZE statement is present.

FILES

Hewlett Packard Files

FILES Names files to be used by the present program and

makes them available for access. File names are
separated by commas.
READ#N, R Sets the file pointer to the beginning of the Rth
record of the Nth file named in the files statement.
In addition, when followed by a semicolon and
 Appendix C 167

variable list, this statement reads values from the

file to the variables.
READ#N; When followed by a variable list, this statement reads
from the file at a point previously established.
MAT READ# Reads values from a file with the same options al-
lowed for READ #.
PRINT #N,R Sets the file pointer in the Nth file named in the files
statement to the beginning of the Rth record and
erases the contents of that record. In addition,
when followed by a semicolon and a variable list,
this statement causes the contents of the variables
to be printed into the file.
PRINT #N; When followed by a variable list this statement
causes the contents of the variables to be printed
wherever the file pointer has been previously set.
MAT PRINT# Prints values from a matrix to a file with the same
options as for PRINT #.
IF END #N THEN n When executed, this statement sets a flag. If at any
later time an attempt is made to read past the end
of data or past the physical end of the file or to
print past the physical end of the file, control
passes to linen.
TYP(N) The TYP (N) function takes on values from 1 to 4,
depending on the nature of the next information
in the file. TYP(N) becomes 1 for number, 2 for
string, and 3 for end of file. If the argument is
negative, the value 4 will be returned for end of
record.

General Electric Files

FILES Names files to be used by the current program and

makes them available for access. File names are
separated by semicolons.

ASCII Files

READ #N, Reads data from the Nth file named in the program
into the variables of the variable list following the
comma.
WRITE #N, Writes data from the variable list following the
comma to the file. The variables in the list may be
separated by semicolons or commas to achieve
corresponding spacing in the file.
IF Mct>RE #N Determines whether or not there is more data in the
file.
IF END #N Determines whether or not the end of the file has
been reached.
APPEND #N Allows additional data to be written to an existing
168 Appendix C

file by setting the file pointer to the end of the

Nth file and placing the file in write mode.
SCRATCH #N Sets the pointer of the Nth file to the beginning of
the file, erases the file, and places it in write mode.
RESTORE #N Sets the pointer of the Nth file to the beginning of
the file and places it in the read mode.

Binary Sequential Files

Rinary SP.l)UP.ntiftl files may be processed by a!! of the 2bcve &t:!tcmcnt:; by

substituting a colon (:) for the pound sign (#). Binary files should be less ex-
pensive to work with; however, ASCII files are very convenient due to the fact
that they may be listed at the terminal.

Random Access Files

READ :N, Same as ASCII.

WRITE :N, Same as ASCII.
IF MqlRE :N Tests true, except when the file pointer is at the
physical end of file.
IF END :N Tests false, except when the file pointer is at the
physical end of file.
SCRATCH:N Places the file pointer at the beginning of the file and
fills the file with binary zeros.
RESTqlRE :N Places the file pointer at the beginning of the file
without altering the contents of the file.
SETW N Tql X Places the file pointer to the Xth word of file N. To
access a random file by record, the formula
W*(R - 1) + 1 places the pointer at the begin-
ning of the Rth record if there are W words per
record.
 APP END IXD
Index of Programs in Text

Descriptio n Section
Program
Adds Numbers from Data . . . . . . . . . . . . . 1-2
ADD 1-2
ADDl Adds and Counts Numbers from Data . . . . .
Adds Numbers from Input . . . . . . . . . . . . 1-2
ADD2 3-3A
ALPHA Arranges Character s of String . . . . . . . . . .
First Use of Change Statemen t . . . . . . . . . 3-3B
ALPHAl
Area by Summing Rectangle s . . . . . . . . . . 6
AREAl
Areal with Variable Interval Width . . . . . . 6
AREA2 3-3
ARRAY$ Demonstr ates a String Array . . . . . . . . . . .
Prints Sample ASCII Codes . . . . . . . . . . . . 3-3B
ASC
AVG Averages Numbers from Data . . . . . . . . . . 11-2
Plays Battle of Numbers . . . . . . . . . . . . . . 12-4
BATTLE
B(,l)UNCE Prints Successiv e Heights for a Magic Ball .. 9-2
CITY First GE Data File . . . . . . . . . . . . . . . . . 4-3
C(,l)LIN Determin es Colineari ty of Three Points
in a Plane . . . . . . . . . . . . . . . . . . . . . . . 7-5
C(,l)MPAR Compares Two Numerics . . . . . . . . . . . .. 1-2
C(,7)MPR$ Program COMPAR with Strings . . . . . . . . . 3-2
C(,l)NCAT Appends One String on Another . . . . . . . . 3-3A
C(,l)NVRT Converts Numeric to String . . . . . . . . . . .· 3·3A
C(,l)RREL Calculate s Linear Correlatio n Coefficie nt . . 11-4
C(,l)SINE Compares Computer cos(x) Function and
Series Evaluatio n . . . . . . . . . . . . . . . . . . 9-4
DECIDE First Demonstr ation of Strings . . . . . . . . . 3-2
DISTl Uses Distance Formula for Two Points ... . 7-2
DIST2 Finds Distance for Several Pairs of Points .. 7-2
DIVIDE Demonstr ates Synthetic Division . . . . . . . . 8-3
DRAW Draws 5 Numbers at Random from 10 with
Replacem ent . . . . . . . . . . . . . . . . . . . . . 1-3
DRAWOl Draws with FOR-NEX T . . . . . . . . . . . . . 1-3
DRAW02 Draws Numbers from io Without
Replacem ent . . . . . . . . . . . . . . . . . . . . . 1-3
END Prints A$(I) in Substring Scheme . . . . . . . . 3-3A
ENTERl Prints Data Serially in File INVOl (HP) ... . 4-2
ENTER2 Adds Data to File INVOl (HP) . . . . . . . . . 4·2
ENTER3 Prints Data to Random Access File
INV02 (HP) . . . . . . . . . . . . . . . . . . . . . . 4-2
ENTER4 Adds Data to File INV02 (HP) . . . . . . . . . 4-2
FIB(,7)1 Prints the First 30 Fibonacci Numbers ... . 9-1
FUNCTI(,l)N Demonstr ates Multiple Line Defined
Function . . . . . . . . . . . . . . . . . . . . . . . . 2-5
GE(,l)ARI Compares Geometri c and Arithmeti c
Sequence s and Series . . . . . . . . . . . . . . . . 9-3
GRAPHl Bare Plot, No Axes, Origin Only . . . . . . . . . 5-2
GRAPH2 Graphl with Axes . . . . . . . . . . . . . . . . . . 5-2
GRAPH3 Plotting from an Array with Axes . . . . . . . 5-3
GR(,l)WTH Orders Contents of City into File
CITYl (GE) . . . . . . . . . . . . . . . . . . . . . 4-3
HALF Halves Remainin g Distance of Separatio n .. 9-1
IMAGE01 -
IMAGE05 Demonstr ates Print Using . . . . . . . . . . . . . 2·3
INT Compute s Compoun d Interest by Formula . 9-1
LIMOl Prints Powers of (-3/5) . . . . . . . . . . . . . . 9-2
LIM02 Prints Terms of (2+(1/5)1 H) . . . . . . . . . . . 9-2
LINEl Gives Equation of Line Given Two Points 7-2
169
Program Description Section
LINES Simulates Bank Teller Window Activity 12-2
L(iJGICl Prints Truth Vaiues . . . . . . . . . . . . . . ::l-4
L(,l)GIC2 Gets Factors Using Truth of
N/X=INT(N/X) . . . . . . . . . . . . . . . . 2-4
MAGIC Prints Odd Order De La Loubere Magic
Squares . . . . . . . . . . . . . . . . . . . . . . . . . 12-3
MAT$ Demonstrates MAT READ and MAT PRINT
for String Arrays . . . . . . . . . . . . . . . 3-3B
MATOl Demonstration MAT READ and MAT
PRINT . . . . . . . . . . . . . . . . . . . . . . 1-5
MAT02 Demonstrates
iviATI:N~l
......... . . ...Matrix Product . . . . . . .
.......................... ...
uemonstrates 1v11y1·
1-5
11'~ v 10-~
MEDIAN Finds Median for Odd Number of Data . 11-3
M(}.>D Modular Arithmetic with One Subroutine 1-4
M(,l)Dl Modular Arithmetic with Nested
Subroutines . . . . . . . . . . . . . . . . . . . 1-4
(,l)PRATN Uses t ,*,/,+,- . . . . . . . . . . . . . . . . . . . 1-2
(,l)RDER Orders Numbers from Data . . . . . . . . . . . 11-3
(,l)RDER$ Alphabetizes Names in a String Array ... . 3-3B
PERP Equation of Line Given Point and Points
on Perpendicular Line . . . . . . . . . . . . . . 7-3
P(,l)INT Finds Intersection of Two Lines in
AX+BY+C=O Form . . . . . . . . . . . . . . . . 7-4
P(,l)LYOl Compares Values of F(X) by Three Methods 8-1
PRT Demonstration of User Defined Function .. 1-4
PYTHl Reads and Prints Coordinates of Three
Points . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5
PYTH2 Sets Up Three Points in Six List Positions .. 7-5
PYTH3 Checks Three Points for Vertices of
Right Triangle . . . . . . . . . . . . . . . . . . . . 7-5
RANK Arranges the Contents of File CITY2 (GE) 4-3
READ Reads File INVOl (HP) . . . . . . . . . . . 4-2
READl Reads with IF END Trap (HP) . . . . . . . . . 4-2
READ2 Finds Number of Parts in Inventory File
INVOl (HP) . . . . . . . . . . . . . . . . . . . 4-2
READCITY Reads File CITY (GE) . . . . . . . . . . . . 4-3
REDCITY2 Reads Contents of File CITY2 (GE) .. . 4-3
REVRS Prints String in Reverse Order . . . . . . . 3-3A
REVRSl Reverses Characters of a String Using Change 3-3B
RNDENTER Writes 10 Random Numbers to Binary File
RAND (GE) . . . . . . . . . . . . . . . . . . . . 4-3
RNDREAD Reads File RAND (GE) . . . . . . . . . . . . . 4-3
RNDS(,l)RT Arranges Contents of File RAND (GE) .. . 4-3
R(,l)UND Rounds to Various Precisions . . . 1-3
SALESl Matrix Demonstration . . . . . . . . . . . . 10-1
SALES2 Salesl Stripped to Bare Essentials ... . 10-1
SEQOl Prints Ratios of Successive Odd Integers 9-1
SL(,l)PE Finds Slope of Lines Given Two Points . 7-2
S(,l)LVE Solves Simultaneous Linear Equations Using
MAT INV . . . . . . . . . . . . . . . . . . 10-2
SUBSTR Demonstrates A$(I,J) As a Substring . . . . . 3-3
TABOl Demonstrates Tab Printing of String
Ch?..r?~-cters . ., . . . . . . . ., . . . . . . . . 2-2
TAB02 Prints Numerics Using the Tab Function .. . 2-2
T(,l)UR Knight's Tour-Prints at Dead End . . . . . . . 12-4
UPDATE Edits Inventory File INV02 (HP) . . . . . . . . 4-2
VAR Calculates Variance and Standard Deviation 11-2
WEATHER Calculates Linear Correlation for
Temperature Data . . . . . . . . . . . . . . . . . . 11-4
170
Program Description Section
WEE KA Matches String and Substring . . . . . . . . . . 3-3A
WEEKAl Finds Day Number from String Day . . . . . . 3-3B
WEEKBl WEEKAl Checking Only First Three
Letters of Input String . . . . . . . . . . . . . . . 3·3B
WET HR Array Demonstra tion with Weather Data .. . 1-4
WETHRl Array Demonstra tion with Weather Data .. . 1-4
XFER Transfers Contents of City to Random
File CITY2 (GE) . . . . . . . . . . . . . . . . . . 4-3
ZER<l>Ol Search for Change of Sign in Polynomial
Function . . . . . . . . . . . . . . . . . . . . . . . . 8-2
ZER(,))02 Uses Interval Halving to Find an Approxi-
mate Zero . . . . . . . . . . . . . . . . . . . . . . . 8·2

BIBLIOGRAPHY

Albrecht, Robert L., LeRoy Finkel and Jerold R. Brown, BASIC, Wiley,
1973. Programmed instruction is used to introduce BASIC.
Coan, James S., Basic BASIC, Hayden, 1970. An introductio n to BASIC
in a mathematics setting.
Gateley, Wilson Y., and Gary G. Bitter, BASIC for Beginners, McGraw-Hill
Book Company, 1970. Brief introductio n to the rudiments of BASIC.
Gruenberger, Fred, and George Jaffray, Problems for Computer Solution,
Wiley, 1965. A book of problems from a wide variety of topics.
Kemeny, John G., and Thomas E. Kurtz, BASIC Programming, 2nd Ed.,
Wiley, 1971. BASIC programming presented by the originators of the
language.
Knuth, Donald E., The Art of Computer Programming, Vol. 3, Sorting and
Searching, Addison-Wesley, 1973. A comprehensive treatment on
arranging data in order and locating items in an ordered data structure.
Ledgard, Henry F., Programming Proverbs, Hayden, 1975. Writing pro-
grams that work the first time.
Nolan, Richard L., Introduction to Computing Through the BASIC
Language, Holt, Rinehart and Winston, 1969. Introduction to com-
puters and computing as well as BASIC.
Pavlovich, Joseph P. and Thomas E. Tahan, Computer Programming in
BASIC, Holden-Day, 1971. An introduction to BASIC assuming a
traditional background in high-school mathematics.
Pegels, C. Carl, BASIC A computer Programming Language, Holden-Day,
1973. With business and management applications.
Sage, Edwin R., Problem Solving with the Computer, Entlek, 1969. An
introductio n to BASIC using mathematics problems.
Smith, Robert E., Discovering BASIC, Hayden, 1970. Introduces BASIC
using a wide variety of interesting problems.
Spencer, Donald D., A Guide to BASIC Programming; A Time-Sharing
Language, Addison-Wesley, 1970. Introduces the language, solves
problems from general interest topics, and presents problems for
reader solution.
Spencer, Donald D., Game Playing with Computers, Rev. 2nd Ed., Hayden,
1975. Mostly Fortran with some BASIC. Detailed anaylsis of a wide
variety of games.
171
..__,

"' ANSWERS TO SELECTED PROBLEMS

Each two-page spread should be read from top to bottom as an individual page.

C!i.1plcr1 ;>;umber.! 230 DATA 1001• !JQQ, 010

232
'lt tam !~.! q4 REM • FINO GREATEST INTEGER WITHOUT INT !"UNCTION 234 REM • THE FUCLJ!ll' f·N ALC-ORITHM
100 REAO X 235 REM WOULD BE MC fif E.i'F' IC I ENT
110 IF' X"' -.001 THEN 290 240 END
;-.,;, mbt•r.! 120 PRINT "INTC"J XI ") IS "1 RUN
I JO IF" X o 0 THEN 160 NC.8
• f"!ND LAfWEST i'l'JU S'll'LLEST rJUMhEK f~;CM OATA 140 PRINT X
~ECONt G.C.F'.
"
91' HEM
READ
-
~I
RF.AO FIRST ~l'JMBEH
ISO
152
160
GOTO 100

LET C = 0
F"IRST I
50
28
JS 5
I

JO.-; R£.,., • l'"JRST VAL'JE IS LARC:E~T AND SMALLEST SO FAR

170 IF' X > 0 THEN 230
LET C = C-1
1001 "
1300 IJ

110 LET s " L "' N 190 LET X :: X+I

120 READ N 200 IF' X < 0 THEN 180
120 t• N : -.Ol T'if.N 200 210 PRINT C
S <= N THEN liO 220 GOTO 100 Number 10
222
l'·•l REM • CURRENT NU~BER S,_,ALLER SO SAVE 230 LET C = C+I 94 REM • COUNT FRECUE~JCY OF' OCCURANCE F'OP. RANDOM INTEC ERS
l'.:0 LET S " N 2·10 LET X = X-1 IDD OtM SCIO>
160 GOTO 120 250 IF" X >: I THEN 230 102
170 IF L ~= N THEN 120 260 PRINT C C;JUNT lN(' LI ST
10•
1'72 270 GOTO 110
!"ir. REM • NU~rn£R LARC-EF SO SAVE 272 120 LET
l''G LET L " 274 lJO NEXT I
190 GOTO 280 DATA 21 l.234561 -2.8712• J, 01 -4, -.OOI 132
290 END 134 REM• GENE::<ATE ICO RANDOM INTEGERS COUNT IN LINE HO
200 PRINT " LARf.EST ="= L RUN 140 FOR I = I TJ lOC
210 PRINT "SMALLEST :"1 S N0.2 LET N::: JNTC fi:>0<-1>•10+1
150
212 LET SCNl :: SCNJ + l
160
REM INTC 2 ~ IS 2 170 NEXT 1
220 DATA SJ. 54, 711 q2, -J, -.01 HHC 1.23456 IS I 172
2Jfl INTC-2.8712 IS -J 174 REM • PRINT RESULTS
H\J'-1 lNTC i IS •SO FOR i = I TJ 10
NC.2 WTC 0 1 IS PRINT II S<I l
'90
INTC-<l • IS -4 NEXT I
200
U·RGEST " 92 202
S~•·LLEST ::-3 210 ENO
Number-I RUN
NO. JO
94 REM • TO DEAL FOUR 1 3 CARD HANDS
100 DIM OC52l
1'1mlwr.J
110 PRINT " HANDS" "
J
• SU'1 INTEGERS FROi-<
''SUM I TO"I
N
I 10 N 120
'30
1·10
PRINT "ONE"• "TWO"• "THREE", "FOUR"
FOR I = I TO 52
LET CCI> = I
'
10
12
I: 0 150 NEXT l 10
!JO LET ( " 1 +I 152
160 LET N : 52 a •
I•
150 II' 1 < N THEN 130 170 FOR R = l TO 13
'10 a
160 PRINT T lfiO
 170 ENO 182 Chaplcrl
RUN I 64 REM • SELECT A CARD AT RANDOM
NC.4 190 LET C = INTC RNDC-1 hN+ I } Sectmnl-·1
200 LET Cl " DCC}
SUM l T0?5 202 Number:.'.
JS 204 REM • F"INO SUiT
210 LETS:: INTC CCl-D/13 +I l 94 REM *
F'INOS MAX ANO MIN FOR OEFINEO F'UNCTION
212 100 DEF F'NR<Xl = t3•X•3 - 2•SINCX•2l - LOGCABSCIS•Xll
Numbers 214 REM • F'JNO CARO IN SUIT 110 PRINT "F'•L•l"!
220 LET C2 "' Cl - CS-lh13 120 INPUT F'•L• I
94 REM • F'INO F'ACTORIAL N 230 IF' C2 > 10 THEN 270 130 LET M"' Ml = F'NRCF'>
100 PRINT "F'ACTORIAL"l 240 IF" C2 = i THEN 560 140 LET P Pl " F'
110 INPUT N 250 PRINT C2J 150 F'OR X : F'+I TO L STEP I
120 LET F' = 1 260 GOTO 280 160 IF M <= FNRCXl THEN 200
130 LET t : 0 270 GOTO c2-10 OF' 500.520.S40 170 LET P " X
140 LET I : J+I 280 GOTO S OF' 600.620.640,660 180 LET M " F'NROO
150 LET F' = F'•l 290 LET OCC) : DCNl 190 GOTO 230
150 IF' I < N THEN 140 300 LET N = N- l 200 IF' Ml :.: FNRCXl THEN 230
170 PRINT F' 310 NEXT H 210 LETPl=X
180 ENO 320 PRINT 220 LET Ml :: F'NRCXl
RUN 330 NEXT R 230 NEXT X
NQ,S 332 232
340 STOP 240 PRINT MJ "MINIMUM AT"I P
F'ACTORJAL?5 342 250 PRINT MIJ "MAXIMUM AT"J Pl
120 SOD PRINT "JACK 260 ENO
SID GOTO 280 RUN
S20 PRINT "QUEEN "I N0.2
Numberh S30 GOTO 280
S40 PRINT "KING F',L~I?t.22;.S
94 REM • TO SOLVE EQUATIONS OF' THE F'ORM 5SO GOTO 280 S.60901 MINIMUM AT 1
9S REM AX+B = CX+O S60 PRINT "ACE MAXIMUM AT 22
100 READ A, B. C. 0 S70 GOTO 280
110 IF' A <:> 0 THEN 140 572
120 IF' B o 0 THEN 140 600 PRINT "CLUBS", NumlwrJ
130 STOP 610 GOTO 290
140 PRINT Al "X+C"I Bl "l :: t"t Cl ")X+C": DI ")
620 PRINT "DIAMONDS''. 94 REM • KEEPS SCORE IN A GAME OF TIC TAC TOE
150 IF' A-C o 0 THEN 180 630 GOTO 290 100 DIM AC3.3); P<2>; WC4)
160 PRINT " NO SOLUTION" 640 PRINT "HEARTS". I ID PRINT "THE BOARD"
170 GOTO 100 6SO GOTO 290 120 F'OR l "' I TO 3
172 660 PRINT "SPADES", 130 F'OR J " I TO 3
PRINT " SOLUTION X " "1 <O-Bl/CA-Cl 670 GOTO 290 140 READ /\CI,Jl
190 672 ISO NEXT J
192 700 160 NEXT 1
194 REM RUN 1'2
200 DATA 1,3,2,-1, 3,2, 1,-J 170 LET Pl : 0
210 DATA 2.-1,2,-3, O.O,Q,O 180 GOSUB 400
220 ENO HANDS 190 F'OR Z "' 1 TO 2
RUN ONE Tl.ID THREE F'OUR 200 LETPl=Pl+1
N0.6 3 CLUBS 9 DIAMONDS 3 SPADES QUEEN CLUBS 210 IF Pl < 10 THEN 250
JACK CLUBS SPADES 6 HEARTS CLUBS 220 PRINT "IT'S A DRAW"
I X+l 3 l ,. < 2 lX+t-1 l SOLUTION X = 4 7 SPADES 5 CLUBS 230 GOSUB 400
3 X+t 2 > : I )X+r-3 ; SOLUTION X " -2.S ACE OIAMONOS 5 CLUBS QUEEN SPADES 7 DIAMONDS 240 STOP
2 X+i-1 l " t 2 lX+t-3 • NO SOLUTION 2 DIAMONDS 5 HEARTS ACE CLUBS 7 CLUBS 250 GOSUB 500
KING SPADES 10 CLUBS KING CLUBS 9 SPADES 260 GOSU8 600
10 SPADES 4 SPADES JACK DIAMONOS 7 HEARTS 270 NEXT Z
Chaptcrl 2 SPADES 272
JACK HEARTS
CLUBS 2 ' CLUBS
HEARTS
ACE
QUEEN
HEARTS
DIAMONDS 3 HEARTS 280 GOTO 180
Section l·J DIAMONDS KING HEARTS QUEEN HEARTS ACE SPADES 392
HEARTS 5 DIAMONDS OIAMONDS 2 CLUBS 394 REM • PRINT THE SOARD
Number I
9
KING DIAMONDS 10 HEARTS '4 HEARTS JACK SPADES 400 F'OR I " I TO 3
DIAMONDS 10 DIAMONDS 4 DIAMONDS HEARTS 410 FORJ:ITOJ
N REM • F'JND ABSCXl WITHOUT COMPUTER F'UNCTIONS
J
' '120 PRINT A(J,J)J
100 READ X 430 NEXT J
110 IF' X: -.001 THEN 190 Number!\ 44D PRINT
450 NEXT I
'"
IW
PRINT "ABS<"J XI " l " "J
IF' X < 0 THEN 160 94 REM • F"INO GREATEST COMMON F'ACTOR
"FIRST #", "SECOND •"1 "G.c.F"."
452
'160 RETURN
10 PRINT X 100 PRINT
IH GOTO 100 110 READ A. B 490 PRINT "ILLEGAL MOVE"
IH 120 IF' A = 0 THEN 240 492
IU PRINT -X 130 PRINT A. B. 494 REM • INPUT MOVES ANO CHECK F'OR I TO 9
IH GOTO 100 140 F'OR 1 : A TO 2 STEP - I 500 PRINT " I " ! ZJ
IH 150 IF A/1 o INTCA/I) THEN 170 SIO INPUT PCZ>

-
174 REM 160 IF B/ I :: ltlTCB/I l THEN 200 S20 IF' INTC CPCZ)-1)/9 ) <> 0 THEN 490
~ DATA -5, o, 3, -2• 2.4, -.OOI 170 NEXT I 530 LET P :: -z
S40 RETURN
IN

NOo I
END
"'
180
190
PRINT 1 592
S94 REM • ENTER NEl.I MOVE ANO CHECK F'OR WIN
192 600 LET C "' 0
ABSC-S 200 PRINT ) 610 F'OR I "' I TO 3
ABS< 0 210 GOTO 110 620 F'OR J : I TO 3
ABS< 3 3 212 630 LET C " C + I
__,
~
ABSC-2 2 214 REM 6'10 IF' C < PCZ> THEN 710
(,J ABSC 2.4 ) :: 2.4 220 DATA 50• JS, 650 IF' Act.Jl " PCZ> THEN 690
 R•JN
.._, .)<:,tion 1-·l lum!'dl Chaptcr3
N:l.11
-"" .'ii mber J icont'ch '.-<-'ltlOll l-~ AS? 10
CS?6
660 PRINT "SPACE TAKEN'' 6 IO
670 GOSUB 500 i'.wnbcr 1
680 GOTO 600 AS?6
94 REM • "/IGHEST A'<D L0';1£5T STRl'lGS 1~; OATA
662 CS? 10
100 RE:AD OS
68•1 REM • ENTER NEW MOVE
LET Act~J) = l IO LET HS=DS ' to
6~0 P
GOTO 730 120 LET LS:>OS
7CO AS?AIOO
122 REM
710 NEXT J CS?A60
720 NEXT ~ 1:10 PRINT OS
ASO AJO'.J
140 READ OS
122
12·1 REM • F'OR WIN 150 IF' OSe"LAST" THE'l 220
AS?ASO
160 If." LS <= OS T>iEN 190
7~0
7.·,Q
FOR l ::::
LET '
'0 170 LET LS=OS CS?AlOO
ASO AIOO
NEXT 1 ISO GOTO 130
·~2
162 REM
190 IF' HS>" OS THE:-J 130 AS? STOP
F"OR I = 1 TO 3
7l2 200 LET HS,,OS
1(4 REM * CHECK DIAGONALS 210 GOTO 130
770 IF AO.I> <> P THEN 212 RE.,
Chapter]
7CO LET IHll = W<ll + 1 220 PiHNT
7'.'0 ff A(J,4-ll <> P THEr, 810 230 PRINT " LOWEST : "ILS
Section J-5
600 LET W(2) = WC2) + I 240 PRINT ''HIG'-IEST = "P·IS
8112 242 RE"I
Nurnber
60.t REM * CHECK ROWS AND COLUM!JS 250
250
DATA "O'l£", "TWO"• "THREE"• "FO'JR", "FIVE"• "LAST"
ENO
810 LET WC3> = IH4> = 0 100 DIM A<75J
8'.~0 FOR J = 1 TO 3 R•JN
110 INPUT AS
8'.10 If" A(l,J) <> P THJ:N 850 NO.J
120 CHANGE AS fO A
8·10 LET WC3l = \HJ) + I 130 LET L = At1})
8~10 IF ACJ, I> <> P THCN 870 ONS
132
St10 LET WC4) = W(4) + I 140 FOR t = i iO L-1
870 NEXT J 150 FORJ=l+!TJL
872 160 11" ACJJ <; A<J> THEN 200
SHO FOR K TO 4 = i 170 LET { = A<J l
6'10 IF WCKJ < J THEN )30 urn LET HI) = A<J>
9·l0 PRINT "YOU WIN l"J LO\..IEST = FIVE 190 LET HJ> : X
910 GOSUB 400 HIGHEST = TWO 200 NEXT J
9:!0 STOP 2>0
9JO NEXT K
212
9-lO NEXT I t>.umberS 220 LET X = 1
9-12 230 IF ACX)<> AC1<•1> THEN 290
9)0 RETURN 100 PRINT 232
9-;_2 110 HEAO As.as.cs.os 23' REM* I~ ACX>= ACX+ll THEN MOVE EACH
9;,4 REM 120 IF" AS="STOP" HiEl-1 250 REM 1EM UP HE POSITION
235
9~0 DATA 1,2,3,4,5,6,;,5,9 !JO tF BS<OS THE~ 190 FOR (::: X .. I TO --1
2<0
970 END 140 tF BS>OS T'-IE~ 160 LETACil=A<IHl
250
R'JN ISO IF AS<CS TKEN 190 260 NEXT !
NJ.3 152 REM 270 LET i.. ::: L- !
160 PRINT OSI"• "1C.S 280 GOTO 230
T-tE BOARD 170 PRINT SSI"• "IAS
I 2 282
160 GOTO 100 290 LET X = X.. !
5 182 REH 300 IF X < '.-1 T-t::·~ 230
8 190 PRINT BSI"• "IAS
I ?3 310 LET ACOl = L
200 PRINT OSI"• "JCS 312
• 2 ?2 210 GOTO 100
-2 -I 320 CHANGE A T J AS
212 REM 330 PRINT AS
5 6
'7 220 DATA "\.llLLIA"l","SMITH","GEORGE"•"SMITH " 340 ENO
'I
8
?5 ' 230
240
DATA "ALtCE''."JONES''.''ROBERTA",".J:JNES "
OATA "STOP"•'"'•""•'"'
RUN
NO• I
• 2 71 250 ENO
-2 -2 -I
R'JN .. THE OUlCK BROm F'OX JUMPED OVER THE LAZY DOGS
-I 6 NO,S ABCOEFGHl JKLM1l lPORSTUVWXYZ
8
'
'I
YOU WIN I
-2
" _, I S'ilTHo GEORGE
Number:!
SMITH, WILLIAM
-I 6
100 OJM NS{IOO>. NC:JO)
8
' J:l'IES, ALICE
JONES• ROBERTA 110 REAON
,'.11rnber6 120 MAT READ llS<N>
130 F'OR 1 ::: l TO N
94 REM *
COMPAR WITH COMPUTED GOSUB
140 PRINT ~S<l>J TA8(20lJ
JOO READ A18
Ch.ip1er3 ISO CHANGE 'IS< I) T>J N
110 IF" A= ,QI THEN 320
152
120 GOSUB SGNCA-81+2 OF 240,2201200
!JO GOTO 100 Scc!mnJ--1 15' REM *SEARCH F>JR FIRST SPACE
160 FOR J"' N<O> HJ I STEP -I
200 PRINT Al "IS GREATER THAN"! B
:-.:umbcrU 170 !=- NC.J> = 32 THEN 190
2 !O RETURN
Al "IS EQUAL TO": 6 180 NEXT J
220 PRINT
"4 REM • WJLTlPLIES TWO l"HEGEHS 'J? TO 20 OIGlTS EACH 190 LET NSCIJ = EXTS<NSC1>.J+1,NC0l) • "• " • EXTS<NS0>.11J)
2JO RETURN ~SCil
AJ "IS LESS THAN"J S 100 otM Asr201.ssr201.osc 101.At201.sr201.cr;io1 200 PRINT
2·~0 PRINT
 250 RETURN 110 LET 05="01234567'39" 210
292 120 READ As.as 212
294 REM 130 IF AS="STDP" TYEN 530 220 FOR I = I TO N-1
300 DATA 3,4, l•7• I• I• :,31,31, -3,2, o.o 140 PRINT AS1"•"16S1" = "J 230 F'OR J = l + l TO N
310 DATA .01.0 142 REM 240 IF' NSCil <" NSCJ) TfiEN 280
320 END 144 REM *
CO:-.!VEiH AS TO Lt ST A 250 LET AS = NS CI}
RUN 150 LET A=LEN<ASl 260 LET NSCll = NSCJ>
NO.& 160 MAT A=ZERtAl 270 LET NSCJl : AS
170 FOR '=I TO A 280 NEXT J
l S LESS THAN 4 180 FOR J= I TO 10 290 NEXT I
i.7 IS GREATER THAN l•I 190 IF AS(I,Il=DS{J,J) THE'll 210 292
3t IS EQUAL TO 31 200 NEXT J 300 PRINT
-3 IS LESS THAN 2 210 LET ACA+l-ll=J-t 310 PRINT "OROEREDt"
0 IS EQUAL TO 0 220 NEXT I 320 PRINT
222 REM 330 F'OR I = I TO N
224 REM • CONVERT BS TO LI ST B 340 PRINT NS CI)
Chi1plcr1 230 LET B=LE'H8Sl 3SO NEXT I
240 MAT 8=ZERt91 352
~l'l[J()[J !-5 360 DATA 3
250 FOR I:I TO B
260 F'OR J= I TO 10 370 DATA GEORGE WASHINGTON.- JOHNNY APPLESEED• JOHN O. AOAMS
Number.!
270 1F' BS[J,JJ:OStJ,JJ HIEN 290 380 ENO
94 REM • FIND LARGEST RA!'JDOM INTEGER IN ROWS AND COLUMNS 280 NEXT J RUN
100 DIM A<S.5) 290 LET aCB+l-ll=J-t NOo2
102 JOO NEXT .1
104 REM *
F'ILL ARRAY WITH RA.'llDOM INTEGERS 302 REM GEORGE WASHINGTON WASHINGTON. GEORGE
110 F'OR I "' I TO 5 304 RE~ *M'.JLTIPLY DIGIT BY DIGIT JOHNNY APPLESEED APPLESEED, JOHNNY
120 F'OR J "' I TO S 310 LET N=A+B JOHN Q, ADAMS ADAMS• JOHN Oo
130 LET ACl.J> = INT< RN0<-1U51-2S) 320 MAT C;ZERtNl
140 NEXT J 330 F'OR ): I TO A ORDERED:
ISO NEXT l 340 FOR J=I TO 8
152 350 LET So: 1 +J-1 ADAMS.- JOHN O.
160 MAT PRINT AJ 360 LET CESl=CtSl+ACll•BCJl APPLESEED• JOHNNY
162 370 IF CCS1<10 TfiEN 410 WASHINGTON> GEORGE
164 REM *
F'INO LARGEST INTEGER JN EACH ROW 380 LET C(Sl=CtSl-10
170 F'OR R I TO 5 390 L.ET C( S+l l,,.CtS+l l+ I
180 LET Cl "' I <lOO GOTO 370 Number 5
190 LET L = ACR.o I) 410 NEXT J
200 F'OR C = 2 TD 5 NEXT I 100 DIM O<IO>. AC!S>.. 8Cl5>. PC30>. OC30l
210 IF' A<R.C> <= L THEN 240 REM 110 LET OS = "0123456789"
220 LETCl=C REM • PRINT RESULTS 120 CHANGE OS TO 0
230 LET L = ACR~Cl <13{} IF C[NJ <> 0 Tl-li':N 4SO 130 Rf:AO AS.- SS
240 NEXT C 440 LET N;N-1 140 PRINT A.SJ . . . "J es
250 PRINT "ROW": RJ "LARGEST IS"J Lj "IN COLUMN"! Cl FOR l:"I TO I STEP - I ISO CHANGE AS TO A
260 NEXT R PRINT OSCCEIJ+l,Ctll+lll 152
270 PRINT NEXT J 160 FOR I = I TO ACOl
212 480 PRINT 110 FOR J = I TO 10

--
* F'INO LARGEST INTEGER lN EACH COLUMN <190 PRINT !BO IF AC!l <> DCJ) THEN 210
~·
REM
•o
no
F'OR C "' I TO S
LET RI :: I
500
502
GOTO 120
R.EM
190
200
LET AC!) = J~l
GOTO 220
LET L"' ACl.-Cl 5 I0 DATA "I 000"•" I 000" • "9999999", "9999999" 210 NEXT J
3 10 F'OR R : 2 TO 5 520 DATA "STOP",'"' 220 NEXT I

= IF' ACR.-Cl <= L THEN 350

LET RI "' R
530
RUfl
ENO 222
230 CHANGE as TO a
LET L "' A<R. C} NOo8 2•0 F'CR I : I TO BCOl
~ NEXT R 250 FOR J = I TO 10
"COLUMN"J CJ "LARGEST lS"J LJ "IN ROW"J RI 1000•1000 " 260 IF BCll <> OCJ) THEN 290

=
~ PRINT
9999999•9999999 = 99999980000001 270 LET sen : J-1

-
= NEXT C
280 GOTO 300
290 NEXT J
~- 300 NEXT l
N 0.-2 Xumber 11 302
310 MAT P :: ZERCACOl+BCO))
-IS -19 -14 94 REM *
ORDERS N'.J'1ERIC CYARACTE:RS I"l STRING VARIABLES 320 F'OR 1 : BCOl TO 1 STEP -I
100 DI"! AS£25J.CSt25l 330 F'OR J : ACOl TO I STEP -I
23 17 ! JO PHHlT 340 LET K = l+J
120 PRINT "AS"I 350 LET PCK> : POO + ACJHaC!)
-10 130 INPUT AS 360 IF' PCK> < 10 THEN 390
140 IF AS="STOP" THEN 290 310 LET PCK-1} :: PCK-1> + INTcPCK)/10)
13 -9 23 15 -10 150 PRINT "CS"! 360 LET PCK) = PCK) - lNTCf\CK)/10>*10

_, 25
160
170
IN?tJT CS
IF LEN<AS);LEN<CS> HIEN 260
390
400
NEXT J
NEXT I
3
' 23
180
190
F'OR G:J TO L~N<A$)
IF' AS(G,Gl o CSCG.Gl THEN 210
002
ROW i LARGEST IS 3 IN COLUMN 4 010 LET PCO> " ACO> + BCO>
ROW 2 LARGEST IS 23 IN COLUMN 4 200 NEXT G •20 IF' PC!) <> 0 THEN 470
ROW 3 LARGEST l S 20 IN COLUMN 2 210 IF' AStG•Gl>"d" T!-IE"I 260 430 FOR I :: I TO P<Ol -I
ROW o LARGEST IS 23 IN COLUMN 3 220 IF' csrG.Gl>"!J" TH€'."I 260 440 LET PCI > = PCl +I}
ROIJ 5 LARGEST IS 25 IN COLUMN 4 230 IF LE/\ICASl<Lf::.•HCSl Tf.IE'J 270 ..ol50 NEXT I
240 PRINT CSJ" "JAS '52
COLUMN I LARGEST IS 13 IN ROW 4 250 GOTO 110 460 LET PCOl :: PCOl - 1
COLUMN 2 LARGEST l S 20 IN ROW 3 260 IF' AS>CS THEN 240 470 MAT 0 ZERCPCO))
03LUMN 3 LARGEST IS 23 IN ROW 4 270 PRINT A.SJ" "JCS 480 F'DR I = I TO PCO>
COLUMN 4 LARGEST IS 2S IN ROW 5 280 GOTO 110 490 LET OCI>: OCPCJ>+ll

"
CJ1 COLUMN 5 LARGEST 1S 23 IN ROW 5 290 ENO 500 NEXT l
 S"cl!m1 J-5 lc:ont'dl 12 REM Y-AXIS.
-..j 96 READ s1.s2 N0.6
(J) 100 DEF' F'NF'CXl = C!/S2>•SINCX>
~.t1mher 5( cont'd)
115 l.ET K "' IVS2 II OF' INCREMENT:;??.
5J2 117 LET N = NIS2 AREA ISi 6500•17
510 LET OCO> " PCO> IJO F'OR X = F TO L STEP SI
520 CHANGE 0 TO OS 160 FOR YI = -K•52 Ta CN-K>•S2 STEP 52
530 PRINT "THE PRODUCT IS "1 410 IF' ABSCY-Yll < .5•52 THEN 470 Chaplcr7
5J2 600 DATA .S;.2
540 DATA "9999999"• "9999999" 610 DATA a, 3, -4, 12 Sechon 7-2
550 ENO 620 END
Numbcr4
"'"
NJ.S
100 PRINT "TO FIND .. HE DISTANCE BETWEEN nio POINTS"
9"99999 • 9999999 '50
T-lE PRODUCT IS 99999980000001 ••o
170
RUN 100
1'.\imberU
NO. I 190
JOO DIM AC IQ), BC )0) 200 LET DI ' <X2-:(1)•2 + CY2-Yllt2

." .
X=-4 210
110 PRINT
120 PRINT "AS"I '' 220
230
130 INPUT AS
\•lO IF AS = "STOP" THEN 32( 232
l•\2
240 F'OR X = If!T<OJ.':!l TO 2 STEP - t
250 IF' .iORCXl "> INTCSORCX)) THEN 270

.--·,
150 PRINT "BS"I I 260 IF lll/X -" INTCDllX> THEN JOO
! 60 INPUT BS I
CHANGE AS TO A 270 NEXT X
HANGE BS TO 8 ' 272
1'10 IF A(Q) = BCOl THEN 27C 280 PRINT "SOilC"'I
200 FOR G = I TO A<O> 290 GOTO 330
JOO PRINT SOF:'.X)J "~50RC"J 01/Xl "l"
210 IF A<G) = S(G) THEN 260
310 GCTO 330
<·12
:!14 REM• ASCCO) = 64 CSEE ASCll 320 PRltlT D
220
~JO
240
IF A<G> > 64 THEN 210
IF' BCG> ,. 64 THEN 210
ff A(Q} < BCD) THEN JOO . '" '
330
RUN
N0.4
END

2•12

260
270
GOTO 280
NEXT G
IF' AS < as THP:N JOO
.. 1
1
I
TO FIND THE 01 ;TANC;: 3ETWEEN TWO POHITS

POHIT A? 15,0
POINT B? o, 15
280 PRINT BSJ "
GOTO 110
"; AS " I DISTANCE AB : lS •S:l'H 2 >
PRINT ASJ " 1
GOTO I 10 t-<umbcrB
I
~ :rn ENO I
1 UN
'. 100
I 10
DIM X(20) • YC20l
LET N = 0
120 READ A1 !l
rs? 100 130 IF A : -.001 THEN 180
[ $? 60 140 LET N = N•I
to 100 150 LET XCNl ' A
rs? A60 160 LET Y<N> ' 8
FS? AJOO 12 170 GOTO 120
(>(,Q AlOO 172
!BO FOR P = l
rs? STOP 190 FOR Pl = P+l TO N
200 PRl'IT "C"1 XCP>J .. , .. , YC?)J "b C"l
210 ?R!TlT XC?llJ "•"J YC?l)J "l"
Chapter-I 220 !F XCP\ <> XC?l > THEN 280
230 !F YC?) <> YC?I l THEN 260
'c•(l!on·l-2 2.iO ?RfllT "PJl'lTS COINCIDE"
:.umber·l 250 GOTIJ 340
260 ?Rl!IT "EJJ.;TJotJ IS X :"J XCP>
!00 "1LES TRY Chilptcr.'i 270 GOT<l 340
110 IF' ENO II THE•~ 200 280 LET M = Cl'C?l>-YC?))ICXC?Jl-XCPll
1?.0 LET 1=0 ~ectam5-J 290 LET 8 = y( ?l - M•XCPl
130 LET l=l+l 300 [F M C> 0 THEN 330
1•10 READ 11, l Number6 310 ?Rt IT "EQJAT!ON IS Y ="1 8
150 IF' TYPC-ll <> 3 THEN 130 320 GOTJ 340
160 PRUIT "F'lRST EMPTY RECORD IS"J I 10 REM • THESE CHANGES IN GRAPHJ PROVlOE F'OR SCALE 330 PR! 1T "ECJATION IS Y ='"; MJ "*X+C"; Bl "l"
170 LET I=I+l 11 REM FACTORS OF" SI ON THE X-AXIS ANO S2 ON THE 340 PRI!IT
1'>0 R!::AO 11, 1 12 REM Y-AXIS· 350 NEXT Pl
190 GOTO liO 110 LET D ::: 47 360 NEXT P
:'.OO PRINT "F'ILE SIZE !S''1t-11"RECOROS" 130 LET T " 5 362
C!O END 195 READ 511 S2 495 REM
215 LET XI = X•SI 500 OATA 3,.1, 5,5, -1.6
CPEN-TRY.11
J«JN 217 LET YI = Y•S2 510 DATA -J, 3, 6,3. 516
220 IF ABSC Xlt2+CYl+llt2 - 169 1 :. T THEN 300 520 OATA -.o)l,O
550 IF ABSHX-LIHSl/10-INTCCX-Ll>•Sl/10)) < .OS THEN 590 530 ENO
F JRST EMPTY RECORD IS I 580 IF" A8SCCY-Ll>•S2/I0-1NTCCY-Lih'S2/10)) < .OS THEN 560 RUN
FILE SIZE IS 11 RECOROS 630 DATA .6, I NO.B
 RUN ! 3 • <l >. ~ s • 6 )
Ch;ipter4
N0•6 ECUATION IS Y : I •X+r I )

Sectmn·l-J I 3 1 4 ),(-! • 6)
EQUATION IS y =-o.s •X+C s.s l
NumbL•rJ
r 3 , 4 hC-j , 3 i
100 F"ILES F'ILEIJ F'ILE2
l!O PRINT "F'IRST F'ILE1"
EQUATION l S Y : o. 25 .:<;:+r 3• 25
120 READ IJ, A
r 3 , <l >. ~ 6 • 3 )
130 PRINT A
IF' MORE 11 THEN 120 EQUATION IS Y "-0.2 •X•C
140
150 PRINT
! 3 • 4 ), ( 5 • 6
160 PRlNT "SECOND F'lLEt"
170
160
190
200
READ 12• A
PRINT A
IF' MORE 12 THEN 170
PRINT
.......
'
EQUATION IS Y

t 5 • b ,, (-j •
EQUATION IS Y :
= I
6
6
•XH

l
I )

210 RESTORE II

..
( 5. 6 ),l-1. 3}
220 RESTORE 12
230
240
250
260
PRINT "MERGED LISTS:"
READ 11, A
READ 12; B
IF' A <: 8 THEN 340
.. EQUATION IS Y = o.S •X-+t

' 5 • b 8 • 3 l
), (
EQUATION IS Y =-I •X+t 1!
3.5 l

270 PRINT a
r 5 , 6 ), l 5 • 6
2BO IF' HORE 12 THEN 250
POINTS COINCIDE
290 PRINT A
300
310 STOP
IF' HORE I I THEN 320 ,_,. 6 ),l-1. :l,
320 READ 11, A
EQUATION rs x =-I
330 GOTO 290
I- I • b ), ~ 8 1 3 )
340 PRINT A )

...
EQUATION IS Y =-0.:)33333 •X+t 5.66667
350 IF' HORE II THEN 410
360 PRINT B

..
1-1. b l . l 5 . b )
370 IF' MORE 12 THEN 390
360
390
-400
410
-420
STOP
READ 12, 8
GOTO 360
READ 11, A
GOTO 260
. EQUATION IS Y = 6

t-1 •
EQUATION l S Y
3 ;,t B,
=
3)
3

1-1 • 3 ),{ 5 • I> l

END }
EQUATION IS Y : Q.S •X+( 3.5

r 6 , 3 >. 1 5 , o ;
EQUATION IS Y =-I •X+i 11
F'lRST F'ILE:

Cli.1ptcri
~ecl!on 7-J

0:wnber2
SECOND F'ILE1
lo! 100 PRINT
2.2 110 READ Xl1Yl1 X2.oY2
120 IF XI = .OQI THEN 530
4.4 130 PRINT "PERPENDICULAR BISECTOR CF'"
5.5 Ch;iplcr6 140 PRINT ''LINE SEGMENT - GIVEN HID EtlO POHHS"
!SO PRINT "t"J XIJ ","! YIJ "} ANO (''J X2J "•"I Y21
MERGED LtSTSt NumhL•rb 160 PRINT "EQUATION IS "J
! 17Q LET XO : CX1+X2>12
1.1 94 REM • SIMPSON'S RULE 180 LET YO = <Yl-+Y2ll2
2 100 DEF' F'NSCX) : 2•XT3 - 2•Xt2 + X + S 190 IF' Xl-X2 <> 0 THEN 250
2.2 110 READ F';T 200 IF" Yl-Y2 <> 0 THEN 23Q
3 120 PRINT " I OF' INCREMENTS"J 202
3.3 130 INPUT l 204 REM *
THE TWO GIVEN POINTS COINCIDE
140 LET W = CT-F')/J 210 PRINT "NOT UN I OUE"
4.4 150 LET Al = F'NSCF') 220 GOTO 100
5 160 F'OR C = i TO l-1 222
5.5 170 LET X = F + C•W 224 REM • HORIZONTAL LINE
6 180 IF' C/2 = INT<C/2) THEN 210 230 PRINT "Y =" J YO
7 190 LET Al : Al • 4•FNS(Xl 240 GOTO 100
200 GOTO 220 250 1F' Y!-Y2 <> 0 THEN 280
210 LET Al = Al • 2•FtJSCXl 252
220 NEXT C 254 REM • VERTICAL LINE
Ch.ipler5 222 260 PRINT "X =": XO
230 LET Al = Al • F'NS<Tl 270 GOTO 100
Sechon 5-2 2-40 LET A : Al•CW/3) 2BO LET M = CY2-YlllCX2-Xll
250 PRINT "AREA I St"! A 290 LET MO= -JIM
Numherl 252 300 LET 8 = YO - MO•XO
254 REM 310 PRINT "y ="I MDI "•X+C"l BJ ")"
-......j 10 REH • THESE CHANGES IN GRAPH2 PROVIDE FOR SCALE 260 DATA -3, I I 320 GOTO 100
-.....J I J REM FACTORS OF' SI ON THE X-AXIS ANO S2 ON THE 270 ENO 322
 Chaptcr7 Chapter{)
~'-'' 1ion 7-J (cont'd!

"
())
Nuinber 2 fcon!'dl
:iec!1on ;:-.:;

Number·\
Sect1onil-I
;-..'umb1.>r2
4'/•\ REM
500 DATA 1.2, 4,7, <1,7, 4,3 '22 100 DIM PC20), FCIO:, SCIOl
510 DATA -3,9, 4,5, 11,-7, 11,-1 REM • THE FOLLOWING NEW LINES IN PROGRAM 110 l'"OR W = I 10 10
520 DATA .001.0.0.0 425 REM PYTHJ WILL PRODUCE THE DESI RED RESULTS 120 LET F'CWl " SCl.O ~ P('.l) " PHl+IOl
SJO END 430 FOR P = l TO 3 130 NEXT :./
RWl 435 IF OCPlT2 > D<P+l)T2+0CP+2lt2 THEN 475 132
NC.2 NEXT P 134 REM *
READ A'JO f'f:JNT COEF"FIC!ENTS
"2 140 READ A
•MS LET 0 = OCJ) 150 ff A = C THEil •120
PERPENDICULAR BISECTOR OF 450 FOR P = 2 TO 3 160 F'OR X =A 10 1 !iTEP -l
LHIE SEGMENT - GlVEN TWO END POINTS 455 IF ASSCO-OCPl>,. .QOOOOI THEN 465 170 R!:AD F'O l
C ! • 2 ! ANO t 4 , 1 1 460 NEXT P 180 PRINT F'CX)J
ECUATION IS Y "'-0.6 •X+C 6 l '62 190 NEXT X
465 PRINT "EQUILATERAL ANO EQUIANGULAR" 200 PRINT
PERPENDICULAR BISECTOR OF 470 GOTO 110 210 READ B
LlrlE SEGMENT - GIVEN nm END PO!NTS 475 PRINT "OBTUSE TRIANGLE WITH LONG SIOE DETERMINED BY" 220 F'OR y " B 10 i '.;'."EP -I
< -~ , 7 • AND t 4 , J 1 460 GOTO 510 230 READ SC) l
E(UATHIN IS Y : 5 4BS PRINT "ACUTE TRI ANGLE" 240 PRINT 51 Y>J
490 GOTO 110 250 NEXT Y
PERPENDICULAR BISECTOR OF' 260 PRINT
"2
Ll!lE SEGMENT - GIVEN TWO ENO POlNTS REM 262
(- 3 • 9 ·, ANO ~ 4 , 5 1
EfUATION IS '( = 1.75 'l'X+( 6.12:5 l
'"
BOO
BIO
DATA
DATA
1.
1.
o,o.
1.2.
2:.o.
5,-1.
1. j,73205
6.15
264
270
REM *
F"IND A'JO i>JUNT PRODUCT
PRINT "PRONJCT '"';
,,o
PERPENDICULAR BISECTOR OF'
LHIE SEGMENT - GIVEN TWO END POJNTS
820
830
DATA
DATA ,,
1. o.o.
-2.6.
o.3,
1.2.
280
290
300
F'OR I " I ··o A
l'"OR J = I TO !l
LET !'Cl+J-ll" PCI+J-1> + F"CIHS(J)
6'0 DATA 0
( -1 ,-7 l ANO ;: 4 ,-7 l RUN 310 NEXT J
EfUATION IS NOT UNIQUE N(h4 320 NEXT I
33D F"OR Z : A+B-1 TO 1 STEP -1
340 PRINT F• Zl I
1':umber.t ( 0 • 0 '• { 2 • 0 1-73205 350 NEXT Z
EQUILATERAL ANO EQUIANGULAR 360 PRINT
PRINT 370 PRINT
GOSUB 5000 ,-1 l ANO t 6 , 15 380 GOTO 110
'1
LET 1<1 : I< OBTUSE LONG SIDE DETERHHlEO BY 382
IF' I< <> 3 THEN 150 ( t ANO ( 6 , 15 384 REM
POINTS:
LET H2 " M 390 DATA 5, J. 2,Q, !, l•
PRINT TAB< 15lJ "AND" • 3 I AND i 4 • 0 400 DATA 2, J, 2, 2, 2• 3
GOSUB 5000 HYPOTENUSE DETERMWED BY 410 DATA 0
ON I< GOTO 600· 700. aoo. 1:100 t ANO ( 4 , 0 l 420 ENO
RUN
ON i<l GOTO 1100• 1000• 1200, 1300 <-2 , 6 i,t 1 , 2 i ANO ( q • 8 N0.2
ON I< I GOTO 1000, 11 oo, 12:0•1• 1300
RIGHT TRIANGLE WITH HYPOTEtlUSE DETERMHlEO BY
ON I<\ GOTO 1200• 1200• 140•l• 1300 POINTS1fQ ,a >A.'JOC-2 .6 l 3 2 0 2 I
2 5 2
PRINT "PERPENDICULAR" PRODUCT = 5 3 12 " 2
GOTO 100 ~>:umlH'r 5
PRINT "PARALLEL" 3 2
GOTO 100 100 FOR A " 3 TO 98
2 3
PRINT "NEITHER" 110 F'OR B" A+! TO 99
PRODUCT " 6 13 6
GOTO 100 12:0 LET Kl : SORCA12 + Bt2l
IJO IF' Kl <> INT<l<ll THEN 160
PRINT "MEANINGLESS"
GOTO 100 1'0 IF I< I > 100 THEN 170
IF' M•M2 "' - I THEN I 000 150 PRINT A; BJ Kl
160 NEXT B Ch.1pkr9
IF' M = M2 THEN 1100
GOTO 1200 170 NEXT A
160 END Sectmn 'l-l
ilUN
5'J00 READ A. B, C ,-...:llmber I
5010 IF' A "' -.001 THEN 9999 N0.5
5J20 PRINT AJ "•X+C"l BJ "l•Y+C"1 CJ "l:Q "l
100 DIM F'(50)
5J30 IF' A= 0 THEN 5100 4 5
102
5J40 IF' B " 0 THEN 5200 12 13
104 REM • STOfE F"IE:i·rncc1 NUMBERS IN A LIST
5J50 LET H " -AIB B 10
110 LET F'C l l " F'!2) " !
5J60 LET K " J u e 120 l'"OR X = 3 ro II!
15
5J70
5J80
PRINT "SLOPE :••1 M
RE1URN 12
0
"
15 130
140
LET F'CO = F'C<:-1 > + F"CX-21
NEXT X
5100
5110
IF' B = O THEN 5300
LET I< = 2
'10
11
u
60
"
H
61
1'2
144 REM *
NO'« "'RlNT R::SUL TS
5120 PRINT "HORIZONTAL" "DIF'F'ERENCE"
12 16 u 150 PRINT "FC:OT2"• ";;"CX-ll•F'CX+ll''.
5130 RETURN 160 FOR x " 2: ro 17
5200 LET I< : I 12 B H
13 ~ e 110 LET A " F'CXh2
5210 PRINT "VERTICAL" 180 LET 8 " F'CX-\HF(X+ll
a so
5220
5300
RETURN
LET I< "' 4 "
15
15
u
~
e
H
190
200
PRINT /\, B, A-3
NEXT X
5310 PRINT "MEANINGLESS" 202
5320 RETURN
16 H M
210 ENO
5322
5995 REM
16
1B
1B
63
N
u
"aH RUN

6000 DATA 1.1,t. 2:;2•4• Q,J,2• t.0.2

 6010
6020
6030
9999
RUN
OATA
DATA
DATA
END
i,2,3, 2.-1.-3,
o, O,
I• 2• 3, 4
- • 00 I•
Q, 0
3,4,5, g,9,2
m
21
21

"
-
H
n
•
H
H
2
H
"H
F'00•2
I

'9
25
F'O(-ll•F'CX .. 1>
2
3
10

" I

.
0 u 65 -I

"n"
N0.4 SI
w 169
,.,
166

I •XH I >•Y+t I >=O SLOPE ,,_ J n

H
60

0
65
0
53
"'
1156
3025
1155
3026
I
-I
ANO
H 7921 7920
2 •X•C 2 )*'(•C 4 ):0
PARALLEL
SLOPE=-!
m
m
"n" H
H
IM
20736
54289·
20737
54288.
-I
I
0 •XH l
>•Y•C 2 ):::0
ANO
HORIZONTAL
•
~
60
M
u
H
142129·
372100·
l42JJO •
372099.
-I

I •XH 0 l•Y•t 2 >=O VERTICAL ~ H 65 974169.

PERPENDICULAR 2 8' 91 2o55041E•06
60
! •X•C 2 >•Y•C 3 ):0
ANO
2 •X•t-1 )*Y+C-3 >=O
SLOPE =-0.5

SLOPE = 2 H -
"
"
42
H "n
65

H
NunibN2

100 DIM F'C2S» LC20)

""
PERPENDICULAR
H H 110 LET F'CI > = F'C2l "' I

--
3 •X•C 4 >•Y•f 5 >=O SLOPE =-0.?S 42 H w 120 F'OR X = 3 TO 25
0 60 H IJO LET FCX> = F'CX-1)
ANO
8 •X•C 9 >*YH 2 ):0 SLOPE =-0.888889 H 13 140 NEXT X
u 142
NEITHER

O •X•t O >*YH I >=O MEANINGLESS

51
5'
u
n
"
H
90
ISO
160
FOR X "' 2 TO 24
FOR y = X•I TO 2S
162
AND
2 •X•C 3 >•YH 4 >"'O SLOPE =-0.66666?
~
60
60
H
u "
87
IH
\64
165
REM • lJSE THE £UCLIOEAN ALGORIT4M TO
RE'1 FINO THE GCF"
MEAN INGLE SS
65 "n 97 170
180
LET N = F"CXl
LET 0 = F'CY>
190 LET l = l~HCN/Ol
Number/ 200 LETR=N-l•O
210 IF"R=OTHEN250
94 REM •DOES NOT HANDLE PAIRS OF' POINTS 220 LET N = 0
Chaplcri 230 LET D = R
95 REM ON A VERTICAL LINE
100 DIM XC20), YC20) 240 GOTO 190
St!ct1on 7--l 110 PRINT ~.ii::'

120 READ T 24• HEM * THE LATEST GCF' TO THE L LIST

130 IF"T:DTHEN530 245 RE'1 IS NOT ALREADY T4ERE.
~umber
140 LET N "' 0 25D Z = l TO ~l I
150 READ XI• Yl 260 IF" LCZl = D THE'l 300
IDD PRINT 160 IF' XI : -·001 THEN 220 no NEXT t
llD READ Mh Bl no LET N "' N• I 2BO LET NI : NI + I
120 IF' Ml = -.001 THEN 260 180 LET XCN> : XI 29D LET LCN I l = 0
IJD PRINT "LINE i Y "' C"t Mil ">•X•C"J 81J 190 LET YCN) = YI NEXT Y
i.D READ M2> 82 200 PRINT "f"J XIJ ".o"! YIJ ")" 310 NEXT X
ISO PRINT "LINE 2 Y : f"J M2J "H•X•C"J 821 21D GOTO ISO 312
If Ml <> M2 THEN 190 220 LET Ml: CYC2l-YCJ}) .1 CXC2)-X(l)) ALL DIF"F"ERE'JT GCF''S.
16D
17D
IBD
PRINT "THE LINES ARE PARALLEL"
GOTO 100
230
240
F'OR I = 2 TO N-1
LET M2: (YCI•l>-YCJ)) ,- CXCJ+ll-XC!))
'"
32D
330
19D LET Xi::: CB2-Bll/CMl-M2> 2SO JF' Ml <:> M2 THEN 300 3'D
2DD LET Y "' Ml•X•BI 260 LET Ml : M2 3'2
210 PRINT "INTERSECT AT C"J XJ ''>"! YI "l" 2?0 NEXT l 350 ENO
220 GOTO 100 272 RIJN
222 280 PRINT NJ "POINTS COLLINEAR" N0.2
22' REM 290 GOTO 110
23D DATA 3•-7• -4• 14• 1.2. 3, 4 300 PRINT "POINTS NON-COLLINEAR" 1 2 3 5 8 13 21 34 SS 89 144
2'D DATA 5.3. 5,3, 2.-1 i. 5.-11 310 GOTO 110
25D OATA -.001.0 32D
26D END 494 REM NumberJ
RUN 500 DATA 999, 1.2. 3,4, s.r.. 7,3, 9.10. -.001.0
NO. I 510 OATA 999• i•8• 3,9, 2,7, 3,2, -.001.0 100 DIM FC50)
520 DATA 0 110 LET F'C!l: FC2) = I
530 END 120 F'OR X = 3 TO 30
LINE I Y : 1 3 >•XH-i RUN 130 LET F'CX> = F'CX-1) • l"CX-2>
LINE 2 Y = 1-4 »"'X•C14 NOo7 140 NEXT X
INTERSECT AT 1 3 , 2 142
150 FOR X = 2 TO 20
LINE i Y = r 1 >•X•C 2 160 PRINT F'CXl/FCX-1 >.
LINE 2 Y = f 3 t 5 • 6 } l?D NEXT X
INTERSECT AT 1 - I ! 1 • 8 ) 172
' 9 • 10 } 180 ENO
LINE 1 Y = I 5 H•X+l 3 5 PO JNTS COLLINEAR RUN
LINE 2 Y : r 5 >•XH 8 NQ,J
THE LINES ARE PARALLEL 1 I, 8;
t 3. 9) I 2 1.5 1.66667 1.0
LINE I Y ::: t 2 >•X•{-11 r 2 ., 7 ~ 1.625 1.6\538 !.61905 1.61?65 1.61818
...., LINE 2 Y : c 5 l•XH-11 r 3 ~ 2 l 1.61798
1.61803
1.61806 io6J80J t.61804 J.6!603
INTERSECT AT 1 0 .-11 Pt!INTS NON-COLLINEAR \061803 !.61!303
"'
~ •:ct1on9-I (cun!'d) 110 FOR I = I TO 12 INVCA>
O> 120 LET S = S+I
0 -,umber·! 130 LET SI "' Sl+S 0038587 -2. 1739 IE-2 0.173913
NEXT I 7o06522E-2 Oo 108696 Oo 130435
JOO DIM F"CSOl 142 -6.52174£-2 Q, JJ0435 -4.34763£-2
l 10 LET FC I l = I 150 PRINT Sil "GIF'TS ALL TOGET~ER"
!20 LET F'C2> = J 160 ENO A•lNV!Al
!30 F'OR I = 3 TO 20 RUN
!40 LET FCI>: F<I-ll + FC:-2) j 0 -J,72529E-<1
!SO NEXT I 7o4505SE-9 1. 9, 3132JE-10
'52 364 GIFTS ALL TOGETHER 7.45058£-Q -J.72529£-9 I
l 60 F'OR l = t TO 20
l70 PRINT F'CJ),
160 NEXT ! Chaptcr9
<B2 I -9.JIJ~JE-<1 S.58794£-9
190 ENO J.72529E-9 I• -9,Jl323E-<1
Sec11on'J-4
·~'JN 9.JIJ2JE-IO -2.79JJ7E-9 I•
Numbcr4
3 II f>,;umbcr-l
100 PRINT "riOW M"-'<Y POINTS"J
29 47 76 123
"
199
2207
322
3571
521
5778
843 1364
15127
110
120
INPUT NI
LET C = 0
110
DIM
MAT
CC3,J),
READ c,
K<::J.!),
I{
SC3.!>. 1(3,J)
130 RANDOMIZE
140 F'OR N = ! TO NI !20 MAT I = !WCCl
•umber 9 ISO LET X = R'ID 130 MAT S = !•K
160 LET Y ::: RND 140 MAT PR!Nf S
100 LET 0 = 10 170 IF' x~2 + Y•2 >:: 1 THSN 190 \02
I 10 LET B = 0 ISO LET C = C+I 145 REM
!20 LET 0 = 3•0/4 190 NEXT N 150 DATA 2,-:;i,-5, 1,-6,5,
130 LET 8 B+I 192 160 DATA 2,-35,-39
140 ff 0 > 1112 THEN 120 200 PRINT CJ "IN THE CIRCLE:"I 4•C/N\
150 LET D = INTC D•IOOO l/100 J 210 ENO
H:iO PRINT BJ "BOUNCES"; OJ "H:IGHT" R'J:-.1
170 ENO N0.4
<UN
'11).9 HO\.I MA°"Y POINTS? 2500
1968 IN THE CIRCLE J.1486 1.
17 BOUNCES Q.075 HEIGHT -3·

'umber 10 RlJ:-.1
N0-4
Number 10
!00 LET A = 2J63
110 PRINT Al "GRAINS" H:J\.I MA'lY POINTS? 2500
!00 DIM A(3,)), 8C:!;J), CC:J.J), DC3,Jl• £{3,3)
120 ff A > 1000 THEN 150 1958 IN THE CIRCLE 3.1328
110 MAT READ A, 8
130 PRINT "YES THEY WILL FJT"
120 "lAT C = ,,•8
140 STOP
150 PRlNT "NO THEY WILL NEVER FIT IN THAT LITTLE SPACE" 130 MAT D : !'iVCCl
Chapter 10 140 PRINT
!60 END
150 PRINT "I:l<HA•E>"
·!UN
Section 10·1 160 MAT PRl~T D
'l[l.10
170 MAT C = !'IVCAl
Numbl'r 160 MAT D = !'IVCS>
9.22337£+18 GRAINS
190 MAT E = IHC
'Hl THEY WILL NEVER F'IT IN THAT LITTLE SPACE 100 DIM SCJ14l.P<4• ll.MC3. I l.TC J,4},AC 11 I l 200 PRINT
110 MAT READ 5, ? 210 PRINT "l'.hfCBl•l~NCAl"
i:tiapter<J 120 F'OR R = l TO .J 220 MAT PRINT E
130 F'ORC=IT04 222
'(.'Cl!on'J-l 140 LET TCl.tC) = TU.Cl + so~.Cl 224 REM
150 NEXT C 230 DATA 1,-2,3, s.-1,-2. 0,3,4
-..umber:! 160 NEXT R 240 DATA 2.-4.0. -J· 1.2. s.2.-s
162 250 ENO
100 PRINT "N''. "1+(2/JltN''. ''Cl+C2/Jl)tN" 170 MAT A : T•P RUN
110 FOR N = J TO 10 180 MAT PRINT A N0.10
!20 GOSUB 180 182
!JO NEXT N 164 REM
132 190 DATA JO, 6001 SO. 20, SO, 31•40t Io, o, 500, SO, 90 INV<A•Bl
140 F'OR N = 20 TO 70 STEP 10 200 OATA 1,. J9,, 49, J. 79
!SO GOSU8 180 210 ENO !050575 -1.2c11s -2.72989
!60 NEXT N RUN 0.747126 -o.6~9425 -1.38506
162 NOo I 1.11012 - I .~5.-.02

!70 STOP
172 1122.49 lNVCB>*lNVCA)
!80 PRINT N.o l+C2/3)TN, Cl+C2'J))TN
!90 RETUR.'I !.50575 -J.2C 115 -2.72989
?00 ENO Number4 00747127 -O.El9425 -I .J8506
~'JN 1.11012 -1 • .os.-.02 -3- 3046
'JJ.2 100 DIM A(J, !l, 8(4,Jl. CC41 ll. NSC4l
a,
., l+C2/JltN : 1+(2/J))TN
110
120
MAT READ A,
MAT C :: 6•A
NS

I J.66667 !066667 122 Number 11

2 1.44444 2.17?78 130 F'OR I = I TO 4
3 1.2963 4062963 !40 PRINT GCI. l>J TABUOlJ NSCI > 100 DIM C<JO~IO>• -<CIO.t>• ICJ0,10» S<t0.1>
lo 19753 1. 71605 ISO NEXT I 110 READ N
 5 1.13169 12.0601 152 120 ff N = 0 THEN 310
6 )008779 2104335 15' REH 130 Ml\T READ CCN.N>. K<N. t}
7 J.05853 35.7225 160 DATA aoo. 200. 1soo 140 PRINT "COEF"F'ICIENT MATRIX"
8 lo03902 59.5374 170 DATA 3,1 ••h 1.1.2. 9,2.J. 1s.a.12 150 MAT PRINT C

'
IO
20
1.02601
1.01734
1.0003
99.2291
165.382
27351.1
180
190
RUN
DATA
ENO
BULBS• SWITCHES• METERS OF' ~HRE• SCREWS 160
170
180
PRWT
PRINT "CONSTA"JT TERMS"
MAT PRINT K
JO 1.00001 •h52338E+06 190 MAT 1 "' HIVCC}
<O I 7o<l8085E+08 200 MAT S : !ti(
50 I J.23720E:+l I 8600 BULBS 210 PRINT
60 I 2004610£+ 13 4000 Sl.JITCHES 220 PRINT "SOLUTIONS"
J,J8388E .. 15 12100 METERS OF WIRE 230 MAT PRINT S
31600 SCREWS 240 PRINT
Numbers 250 GOTCl 110
252
100 DIM F'C20l 254 REM
110 LET F'CI) "'F'C2): I
Numberb
260 DATA 2, 3,1, S,-J
120 F'OR N "' 3 TO 20 270 DATA ;,21
100 DIM ACl.3»BC3,4»CC4,J>.DC!.4l>ECl,\l
130 LET F'CNl "' F'CN-)) 280 DATA 3, 2,J,-1. J,Q,1, 1,-2,-S
110 MAT READ A. B. C 290 DATA 20.- Q, 6
140 NEXT N
1'2 120 MAT 0 :: A•B
300 DATA 0
ISO PRINT "BY A001Tl0N"• "BY FORMULA" 130 MAT C : D•C 310 ENO
160 LET SO : SORCSl 140 MAT PRINT E
RUN
1'2
170 F'OR N = I TO 20 NQ, JI
LET SI "' CCl•SOhN-Cl-SO)tN)/((21N>•SOl 145 REM
ISO
PRINT F'CN), SI
!SD DATA I, I• I
190
16D DATA JD. aoo. so. 20. so. 31. 40. Io.
200 NEXT N
170 OATA 1,.39,.49,3.79 COEFFICIENT MATRIX
202
210 END 180 END
RUN
RUN
NO.S

BY ADDITION BY FORMULA 1122.49

CONSTA.'H TERMS
I
I 7
2 21
J
5 Chaplcr10
a SOLUTIONS
Secl1or1!0-2
J.
"SS34 34.
ss.
Number -2.

89 59, DIM ACJ,J), BCJ,J>. CCJ,J)

144 )44. 110 MAT R£AO A COEF"F'ICIENT MATRIX
233 233. 120 MAT B = INVCAl
377 377. \JO PRINT "INVCAl"
610 610. \40 MAT PRINT B 0 I
937 987. ISO MAT C : A•B -2 -s
1597 1597· 160 PRINT
2584 2SB4. 170 PRINT "A•INVCA>" CONSTANT TERMS
4181 4181. 180 MAT PRINT C
676S 676S. 190 MAT C :: B•A 20
200 PRINT 0
Chapter!> 210 PRINT "INVCAl•A"
220 MAT PRINT C
Sectmn'kl 222
224 REM SOLUTIONS
Number J 230 DATA 4,-4,4, I• 1.7,
240 ENO 1.
94 REM *THE TWELVE DAYS OF CHRISTMAS RtJ.'J s.
100 LET S = SI : 0 NO.I -J.

~
 INDEX

ABS(X) function, 7 Computed G(/JT(/J state- De la Loubere, 152

Algorithm, 10 ment, 3 Delimiter; comma, 3, 25,
Alphabetization, 42 Computer functions: 33
AND, 29 ABS(X), 7 in a file, 47
Append statement, 55 ASC(I$), 44 semicolon, 3, 25, 33
Area, 72 ATN(X), 11 Descartes' Rule of Signs,
Arithmetic mean, 137 CLK(X), 11 112f
Arithmetic operators, 4 C(/JS(X), 11 DIM statement, 10, 35,
Arithmetic sequence, DEF, 12 40
116 EXP(X), 11 Dispersion, 138
Arithmetic series, 122 EXT$( ), 44 Distance, 72, 78f
Array; numeric, 15, 127 INT(X), 7, 8 Distance formula, 80
string, 34 LEN(A$), 35, 40 Divergence, 119
ASCII code, 31, 35, 40, L(/JG(X), 11 Dummy data, 2, 47, 48,
44 MAX, 29 59, 137
ASCII files, 54 MIN, 29 Dummy string, 38
Assignment operator, 5 NUM, 18
Assignment statement, 5, RND,9 Efficiency, 10, 42, 46,
31,40,47 RND(X), 8 102,140
ATN(X) function, 11 SEG$( ), 44 E-format, 5, 26
Average, 137 SGN(X), 7 END statement, 2
Axes, 64, 68 SIN(X), 11 EOF, 52
SQR(X), 7 Error, 75, 112, 125
Battle of numbers, 155f TAB(X), 23, 64 Error message, 48, 51
Binary files, 54, 57 TAN(X), 11 EXP(X) function, 11
Bubble sort, 140 TIM(X), 11 EXT$( ) function, 44
TYP(N), 51
Cartesian coordinates, 18 C(/)N, 19 Factor Theorem, 108
Census, 54 Concatenation, 39 Fibonacci sequence, 115
CHANGE statement, 40 Constant matrix, 132 Files, 46, 54
Chess, 156 Content addressing, 52 ASCII, 54
Circle, 74 Continuous function, binary, 54, 57
CLK(X) function, 11 104 random access, 50, 57f
Coefficient matrix, 132 Convergence, 119 sequential, 57
Coefficients, 101 Correlation coefficient, serial, 46, 55, 57
Coinciding lines, 90 linear, 143, 146 teletype, 54
Collinear points, 78 C(/JS(X) function, 11 FILES statement, 4 7, 55
Column vector. 127 Cosine, 125 FNEND statement, 29
Comma delimiter, 3, 33 CREATE command, 58, Formatting, 25
Common difference, 116 61 F<,l>R-NEXT statement, 9
Common ratio, 117 Functions, computer; see
Compound interest, 117 DATA statement, 2, 43 Computer functions
Computed G(/JSUB state- Degree of a polynomial, Function, mathematical,
ment, 15 101 64
182
 Index 183

Games, 147,154-159 Matrix algebra, 19 Random numbers, 147

Geometric sequence, 117 MAX function, 29 RAND(,Z)MIZE statement,
Geometric series, 122 Median, 140 9
G(,Z)SUB statement, 13 Midpoint, 75, 81 READ statement, 2, 31,
G(,Z)T(,Z) statement, 3 MIN function, 29 40,55
Graph, 64 Monte Carlo, 126 Record of storage, 47, 58
Multiple assignment Relational operators, 2f
Hero's formula, 100 statement, 6 REM statement, 4
Hypotenuse, 95 Multiple zeros, 111 Remainder Theorem,
108
Identity matrix, 19, 132 NAME command, 2 REST(,l)RE statement, 57
IDN, 19 Nested form of a poly- RETURN statement, 13
IF END statement, 48, nomial, 102, 110 Right triangle, 79, 84, 95
59 Nested loops, 16 RND function, 9
IF M(,Z)RE statement, 55 NEW command, 2, 54 RND(X) function, 8
IF-THEN statement, 27 Noncollinear points, 94 Row vector, 127
Inconsistent equations, Nonreal zeros, 104 RUN command, 3
90 N(,Z)T, 29
Indeterminant solution NUM function, 18 SA VE command, 54
89f Scale, 66
Initialization, 6 Operators SCRATCH statement, 57
INPUT statement, 6, 31, arithmetic, 4 Sectors of storage, 4 7
40 assignment, 5 SEG$( ) function, 44
Integral zeros, 112 logical, 27 Semicolon delimiter, 3,
Interest, compound, 117 relational, 2, 3 33
Intersecting lines, 84 (,l)R, 29 Sequence; arithmetic,
INT(X) function, 7, 8 Ordering data, 140 116
INV, 21, 131 Origin of a graph, 64 defined, 115
Inverse of a matrix, 132 Fibonacci, 115
Parabola, 66 geometric, 11 7
Knight's tour, 156ff Parallel lines, 84 of partial sums, 122
Pearson r, 146 Sequential files, 57
LEN(A$) function, 35, Perpendicular bisector, Serial files, 5 5
40 85 Series; arithmetic, 122
LET statement, 5, 40 Perpendicular lines, 84 defined, 122
Limit of a sequence, 119 Plot, 64 geometric, 122
Linear correlation coef- Plotters, 64 SETW statement, 59
ficient, 143, 146 Pointer, 51, 55, 59 SGN(X) function, 7
List, 10 Polynomial, 12, 101 Significant digits, 5
LIST command, 2 Polynomial equation, Similar triangles, 85
Logical operators: 104 Simpsons' rule, 77
AND, 29 Prettyprinting, 4 Simulation, 147
(,Z)R, 29 PRINT statement, 2, 40, Simultaneous linear
N(,l)T, 29 47,81 equations, 88, 131
Logical value, 28 PRINT USING state- SIN(X) function, 11
L(,l)G(X) function, 11 ment, 25 Slope, 81
Loops, 9 Pseudo-random numbers, positive, 82
nested, 16 8 negative, 82
Pythagorean Theorem, Slope-intercept, 83
Magic squares, 152ff 95 Solution matrix, 132
MAT equals, 21 Spacing; comma, 43
MAT INPUT, 18 Quadratic formula, 104, semicolon, 43
MAT PRINT, 18, 43 111 Standard deviation, 137,
MAT READ, 18, 43 Quotes, 33, 35 138
Matrix, 127 Statements:
coefficient, 132 Radians, 11 APPEND, 55
constant, 132 Random access files, 57, ASSIGN, 47
identity, 19, 132 58 assignment, 5, 31, 40
184 Index

Statemen ts (continue d) Storage, 59, 68 Undefine d slope, 84, 97

CHANGE , 40 String, 25, 31 Undefine d solution, 89
computed GQ.>SUB, 15 dummy, 38 User-defi ned function,
DATA, 2, 43 null,50 12
DIM, 10, 35, 40 SQR(X) function, 7 multiple line, 29
END,2 Subroutin e, 13
FILES, 47, 55 Subscript ed variable, 10 Variables , 4
FNEND, 29 Synthetic division, 109 numeric, 31
FQ.>R-NEXT, 9 System command , 2 string, 31
flff\CiTTn
\_T\fH~"JU,Lt;
i
..LU
•") CREATE , 58, n1 :;ubscripted ntl!T!E-d£::
1

GQ.>T(/J, 3 LIST, 2 10
.,.TA 1\11'1.' 0
IF Ef~D, 48, 59 J. "4 .t"'.. J..~'.! !.:.1' .::. subscriiJLt!d stihig, 34
IF M(/JRE, 55 NEW, 2, 54 Variance, 138, 142
IF-THEN , 27 Q.>PEN, 47 Variation , 113
image, 25 RUN,3 Vector; column, 127
INPUT, 6, 31, 40 SAVE, 54 row, 127
LET, 5, 40
MAT, 18, 19 TAB(X) function, 23, 64 WRITE statemen t, 57
multiple assignmen t, 6 TAN(X) function, 11
PRINT, 2, 31, 40 Teletype files, 54 X-axis, 65, 68
PRINT USING, 25 TIM(X) function, 11 X-coordin ate, 78
RAND(/JM IZE, 9 Transfer; condition al, 2
READ, 2, 31, 40, 55 unconditi onal, 3
Trapezoid method, 77 Y-axis, 65, 68
REM,4 Y-coordin ate, 78
RESTQ.>RE, 57 Trial and error, 10
RETURN , 13 Triangles, similar, 85
SCRATC H, 57 TRN, 21 ZER, 19
SETW, 59 Truth values, 27 Zeros; integral, 112
STQ.>P, 5 TYP(N) function, 51 multiple, 111
WRITE, 57 nonreal, 104
STQ.>P statement , 5 Undefine d condition , 82 real, 104
ADVANCED BASIC:
Applications and Problems
James S. Coan
Now you can extend your expertise in the BASIC language with
this book of advanced techniques and applications of the BASIC
language. It allows you to gradually increase both your understand-
ing of concepts and your ability to write programs. The develop-
ment of each topic progresses from simple to more sophisticated
problems and is accompanied by many sample programs to clarify
the discussions.
ADVANCED BASIC opens with a review chapter on the elemen-
tary points of BASIC. Then, the book covers extended features,
strings and files, and the applications of the BASIC language in
such areas as: coordinate geometry, area, sequences and aeries,
polynomials, graphing, simulations and games. Summaries and
problem exercises end each section.

Other books ot Interest •••

BASIC BASIC:
An Introduction to Computer Programming In
BASIC Language
James S. Coan
This popular text for high school and college students integrates pro-
gramming in BASIC language and the teaching of mathematics. Each topic
begins with a short. complete program and moves to more sophisticated
problems. The use of flowcharts is encouraged as an aid in writing programs.
Summaries and questions review important concepts. #5872-1. pa,,.r,
# 5873-X, cloth, 256 pages, 6 x 9, illus.

THE BASIC WORKBOOK:

Creative Technlqu• for Beginning Programmers
Kenneth E. Schoman. Jr.
A hands-on approach to learning BASIC and the fundamentals of
problem-solving using a computer. Through many exercises. you develop
a workable BASIC vocabulary, a feeling for the logic and intrigue of pro-
gramming algorithms. and the self-confidence needed to use a computer
in a variety of applications. # 5104-2, paper, 128 pages, 8~ x 11, Illus.

[{]
HAYDEN BOOK COMPANY, INC.
Rochelle Park, New Jersey

ISBN 0·8104-5855·1