MATHCAD'S PROGRAM FUNCTION

and
APPLICATION IN TEACHING OF MATH
DE TING WU
Depart of Math
Morehouse College
Atlanta, GA.30314, USA
dtwu@morehouse.edu
1. Introduction
1.1 About Mathcad
Mathcad is one of popular computer algebra system (math software) in the world. Like
other CAS's, it has the capabilities to perform algebraic operations, calculus operations
and draw graph of 2 or 3 dimensions. We can use it to get numerical, symbolic and
graphic solution of math problem. Besides above common capabilities, it has some
feature: its words processor function is better than other CAS. This feature makes it look
like a scientifical words processor and it easy to create a woksheet.
Following current trend to add program function to math software, it added program
function to itself , started in Mathcad V7. This improvement makes its function more
complete and more powerful.
1.2 Role of program function in teaching of Math.
Without doubt, math softwares and technology are helpful aide in teaching of Math. They
are effecting the teaching of Math profoundly and it can be expected the emerging newer
technologies will penetrate and reconfigure the teaching of Math.
The program function is a new function of math software. It is natural we want to know
whether or not it is useful in teaching of Math? I think: although for most problems we
meet in teaching of Math it is enough to use computation ability, in certain cases the
program function becomes necessary and very helpful. For example, for Numerical
Analysis the program function is very useful but for Abstract Algebra it is less necessary.
Another example is Differential Equations, when teaching approximate method for a
solution of a differential equation the program function is necassary and teaching
existence and uniquess of the solution of a differential equation it is not necessary.
Moreover, as the power of program function is increasing the more applications will be
found.
1.3 About this presentation
This presentation is to introduce the program function of Mathcad and illustrate some of
application in teaching of Math which is based on my teaching experience in past years.

2. Program function of Mathcad

In this part the program operators of Mathcad and the structure of Mathcad program
are introduced.
2.1 Program operators of Mathcad
Mathcad program is created by using its program operators. Mathcad has 10
operators which are accassable from Program toolbar. Now, let's take a look of
these operators.
1."add line": operator to initiate a program or add a line to a program. When click it on
the programming toolbar a vertical bar and 2 placeholds will be produced.
2. " ": local assign operator which define left holder by right holder of arrow.
3. "

4."

if

": conditional operator which can be used whenever you want a program
statement to be executed only upon the occurrence of some condition.

otherwise ":

conditional operator which can be used together with "if" operator when
you want a program to execute a statement when the condition of "if"
operator is false.
" : looping operator which can be used when you know how many times you

5." for

want a program statement to be executed repeatedly.

6." while ": looping operator which can be used when you want to stop execution of a
statement upon the occurrence of a condition but don't know exactly when
the condition will occur.
7." break ": Control operator which can be used when you want to break out of a loop.
upon the occurrence of some condition and which moves execution to next
starement outside the loop.
8. " continue ": Control operator which can be used when you want to halt current iteration
of a loop upon the occurrence of a condition and force program to
continue next iteraion of a loop.
9." return

10."

": Control operator which can be used anywhere in a program when you
want to interrupt a program and to return a value different from the value
of last expression of a program.

on error

": Operator for error control which can be used to return an alternative
value when an error happened in an expression.

2.2 Structure of Mathcad program

A program is a precise sequence of instructions designed to accomplish a specific task.
Mathcad program is a sequence of statements created by using program operators, too.
But, it has special form and structure. Now, we use an example to illustrate.
Example of Mathcad program

f( x, y ) :=

x +y

v sin ( u ) + e
w 1+ u
v
w

Left is a Mathcad program which defines a function

in 2 variable. This program consists of 2 parts. Left
part is name or variable name which represent the
output of last statement in the program; Right part
is included in a vertical bar which is the body of the
program comprised of program operators. Both
parts are connected by the definition sign " := ". This
means Left part is defined by right part or the program
assigns the result to the name on the left. In fact, the
program just is a compound expression which is
represented by name on the left.

The form and structure of Mathcad program make it easy to recognize and separate
from other part of worksheet. The program operators are simple so that writing a
program is simple. This is very suitable for teaching purpose. This form and structure
is good for revising of program too. Also, its form and structure make it convenient to
interact with other parts of a worksheet.
Although Mathcad program is comprised of its program operators, a lot of its built-in
functions can be used in the program too. For example, augment, sign, random variable
and so on. This makes its program more concise and powerful. The output of Mathcad
program maybe is a number, a vector, a matrix and a function, this makes its application
wide and flexible.
However, the program function of Mathcad is as limited as other technologies. For
example, Not all its built-in functions can be used in Mathcad program and some of
them can not be used completely and symbolic operations don't work within a
program.
Anyway, Program function of Mathcad is a useful tool in teaching of Math.

3. Application in teaching of Differential Equations

Mathcad's program function can be applied to the teaching of Differential Equations. Although Mathcad has several built-in functions(or routines) for solving of differential
equations, such as Odesolve, rkfixed, rkadapt, etc, there're cases we need its program
function. For example, when teaching approximate methods solving differential equations
and they are different from those which are basis of above built-in functions, Mathcad
program function becomes necessary and helpful. Esspecially, in comparison of different
approximate methods. In this part we'll illustrate its application by exampltes.
3.1 Appication to one differential equation
In this section we 'll use Mathcad's program function to write programs to perform
Euler method and Improved Euler method for a first order differential equation and
compare their results with the exact solution and the results of built-in function through
an example.
Let's consider following example.
2

y
Given: y'=1+ y +

and y(1)=0, x in [1,2]; Find: its approximate solution over given

interval by Euler method and Improved Euler method with step size h=0.1 and errors.
To make programs of Euler and Improved Euler method
Program of Euler method

Eu( f , a , b , c , h ) :=

Program of Improved Euler Method

IEu( f , a , b , c , h ) :=

x a
0

i+ 1

x +h

y + h f x , y

y1

( i i)

P augment ( x, y)
P

for i 0.. n 1

for i 0 .. n 1

i+ 1

ca

ca

i+ 1

y b

y b
n

x a

x +h
i

i+ 1

i+ 1

( i i)

y + h f x , y
i

y +
i

h
2

( ( i i) ( i+ 1 , y1i+ 1))

f x ,y + f x

P augment ( x, y )
P

Note: in above programs Left side is the output of the program which is a matrix of xvalue and approximate solution values. Right side is the program which uses the
input: 'f' function of ODE, 'a' initial x value; 'b' initial y value; 'c' end point of the interval;
'h' step size and perform approximate method to produce approximate solution.
Now, let's us these programs to get approximate solutions of above example.

In above example,
Let

f( x, y) := 1 +

A := Eu( f , a , b , c , h )

y
x

a := 1

b := 0

c := 2

h := 0.1

B := IEu( f , a , b , c , h )

then, Approximate solution produced by Euler mathod:

0

A = 0

1.1

0.1 0.2099 0.3305 0.4624 0.6063

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

0.763 0.9335 1.1185 1.3193

Approximate solution produced by Improved Euler method:

0

B = 0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

0.105 0.2208 0.3483 0.4885 0.6421 0.8104 0.9944 1.1955 1.4152

It's not hard to find its exact solution:

g( x) := x tan ( ln( x) )

i := 0 .. 10

x := 1 + i 0.1
i

Also, find its solution by built-in function Odesolve'

Given

y' ( x)

1+

y ( x)
x

y ( x)
x

y( 1)

y := Odesolve ( x, 2)

( i)

y x =

0.1052

0.1052

0.2212

0.2212

0.3491

0.3491

0.4897

0.4897

0.6439

0.6439

0.8128

0.8128

0.9975

0.9975

1.1994

1.1994

1.4201

1.4201

10 1.6613

Now, we graph these 4 solutions in a picture.

( i)

g x =

10 1.6613

(A 1 )i

1.5

(B 1 )i

g( xi)
y ( xi)

0.5

1.2

1.4

1.6

1.8

xi

Next, let's analyze their error. we define error of Euler method and Improved Euler Method as
ErEu and ErIEu.

( i) (

ErEu := g x A
i

ErEu =

ErIEu =

1
0 5.159810 -3

0 2.011410

ErEui

-4

)i

0.0113

0.0187

0.0273

0.0376

4.71410

-4

8.117610

-4

1.228610

0.1

ErIEui
0.05

1.2

( i) ( 1 )i

ErIEu := g x B

1.4

1.6
xi

1.8

-3

1.732210 -3

3.2 Application to system of differential equations

In this section we'll write a program to implement Improved Euler method to find the
approximate solutions of a system of 3 differential equations and analyze the error.
First, write the program
IEu( f1, f2, f3, a , b , c , d , l, h ) :=

t a
0

u1 b
0

u2 c
0

u3 d
0

la
h

for i 0.. n 1
t

i+ 1

t +h
i

( i i i i)
u21
u2 + h f2( t , u1 , u2 , u3 )
i+ 1
i
i
i
i
i
u31
u3 + h f3( t , u1 , u2 , u3 )
i+ 1
i
i
i
i
i
u11

i+ 1

u1 + h f1 t , u1 , u2 , u3
i

u1

u1 +

u2

u2 +

u3

u3 +

i+ 1
i+ 1
i+ 1

i
i

h
2
h
2
h
2

( (i

( (i

( (i

( i+ 1 , u11i+ 1 , u21i+ 1 , u31i+ 1))

( i+ 1 , u11i+ 1 , u21i+ 1 , u31i+ 1))

( i+ 1 , u11i+ 1 , u21i+ 1 , u31i+ 1))

f1 t , u1 , u2 , u3 + f1 t
i

f2 t , u1 , u2 , u3 + f2 t
i

f3 t , u1 , u2 , u3 + f3 t
i

P augment ( t , u1)
P augment ( P , u2)
P augment ( P , u3)
P

Now, we use this program to find approximate solutions of following example.

Given: u1'= u2
u1(0)=1
t
u2'= -u1-2 e +1
u2(0)=0
t in [0,1]
u3'=-u1- et+1
u3(0)=1
Find: its approximate solutions by Improved Euler method with step size
h=0.1
In this example
t

f1( t , u1 , u2 , u3) := u2
a := 0

b := 1

f2( t , u1 , u2 , u3) := u1 2 e + 1

c := 0

d := 1

l := 1

h := 0.1

f3( t , u1 , u2 , u3) := u1 e + 1

Let

A := IEu( f1, f2, f3, a , b , c , d , l , h )

then its approximate solution is

shown in the right matrix.
The ist column is t-value ;
The 2nd column is u1-value;
The 3rd column is u2-value;
The 4th column is u3-value.

A=

2
1

0.1

0.99 -0.2105

0.8947

0.2

0.9579 -0.4411

0.7805

0.3

0.9018 -0.6918

0.6583

0.4

0.8196 -0.9627

0.5295

0.5

0.7093 -1.2539

0.3953

0.6

0.5689 -1.5657

0.2571

0.7

0.3963 -1.8983

0.1163

0.8

0.1893 -2.2524 -0.0258

0.9 -0.0541 -2.6286 -0.1678

10

1 -0.3363 -3.0278 -0.3081

Its exact solutions are :

t

g1( t ) := cos ( t ) + sin ( t ) e + 1

i := 0 .. 10

g2( t ) := sin ( t ) + cos ( t ) e

( i)

g3( t ) := sin ( t ) + cos ( t )

t := 0.1 i
i

g1 t =

0.9897

-0.21

0.8952

0.9573

-0.44

0.7814

0.901

-0.69

0.6598

0.8187

-0.9602

0.5316

0.7083

-1.2506

0.3982

0.5679

-1.5614

0.2607

0.3953

-1.8931

0.1206

0.1885

-2.2462

-0.0206

-0.0547

-2.6213

-0.1617

10 -0.3365

( i)

g2 t =

10 -3.0195

( i)

g3 t =

10 -0.3012

Next, we make 4 graphes. Ist one is graph of 3 approximate solutions; 2nd one is the
graph of u1 and g1; 3rd one is the graph of u2 and g2; 4th is graph of u3 and g3.

2
1

(A 1 )i
0

(A 2 )i
(A 3 )i

0.5

(A 1 )i

0.5

g1( t i)

0.5

0.5

(A 0 )i

(A 2 )i
g2( t i)

ti

0.5

(A 3 )i

0.5

g3( t i)
0

0.5

0.5
ti

ti

Now, let's analyze the error. We define error in u1, u2, u3 as Eru1, Eru2, Eru3.
Eru1 :=
i

(A 1 )i g1(ti)
0

Eru1 =

Eru2 =

Eru2 :=
i

(A 2 )i g2(ti)

Eru3 :=
i

1
2
3
4
0 3.333410 -4 6.134310 -4 8.327510 -4 9.845510 -4

1
0 5.169210

Eru3 =

2
-4

1
0 4.292910

1.116410

3
-3

2
-4

9.318910

1.794610

4
-3

3
-4

1.503110

2.547310

2.137510

5
1.06310 -3

5
-3

4
-3

(A 3 )i g3(ti)

3.369910 -3

5
-3

2.829410 -3

The graph of these errors as follows:

0.008

Eru1i 0.006
Eru2i
Eru3i

0.004

0.002

0.2

0.4

0.6
ti

0.8

4. Application in teaching of Numerical Analysis

Numerical Analysis is another area in which Mathcad's program function can display its
full potential. Since Numerical Analysis often needs massive, repeated and tedious
computation the program function can best meet this requirement. In this part we just
illustrate its application in finding approximate solution of non-linear equations and
Newton's interpolation polynomial.
4.1 Application of finding approximate solution of non-linear equations
In this section we'll use program function of Mathcad to write a program to perform
Secant method to find approximate solution of non-linear equations. And, we'll
compare the result with built-in function's result.

Program for Secant method.

ApS ( f , a , b , r) :=

p0 a

Note:
Input: f(x)=0- given equation
a,b-endpoints of [a,b]
that bracket the solution
r-desired accuracy
Output: Approximate solution
satifying accuracy

p1 b
while p1 p0 r
p

p0 f( p1) p1 f( p0)
f( p1) f( p0)

p0 p1
p1 p
p

Now, let's see an example: Find approximate solution of cos(x)-x=0 with accuracy 10 5.
f( x) := cos ( x) x

f( 0) = 1

f( 1) = 0.45969769

Use graph to locate solution in [0,1]

f( x)

b := 1

x := ApS ( f , a , b , r)

x = 0.73908513

Let's compare withresult from built-in function.

r := root ( f( x) , x, 0 , 1)

r := 10

The desired approximate solution:

a := 0

r = 0.73908513

Let's check the error

4
x

er := x r

13

er = 1.59317004 10

4.2 Application in finding Newton interpolation polynomial

In this section we'll use the program function to make a program to produce the coeffients of Newton
interpolation polynomial then use symbolic operation to form Newton interpolation polynomial. Also,
the graph of Newton polynomial is provided. We illustrate by below example.
Example.Given: f(1)=0.7651977, f(1.3)=0.620086, f(1.6)=0.4554022, f(1.9)=0.2818186,
f(2.2)=0.1103623. Construct interpolating polynomial of degree 4 by Newton's
interpolatory divided difference formula and use it to approximate f(2).
Sol'n: First, enter the data:
a := 1

a := 1.3

a := 1.6

b := 0.7651977

a := 1.9

b := 0.620086

b := 0.4554022

a := 2.2
b := 0.2818186

b := 0.1103623

Second, use following program to figure out divided-differences. In following program,

use a for-loop to enter above set of data, and then use double for-loop to figure out the
divided-differences. The output of the program is a matrix and the elments on the diagonal
are the divided-differences needed in Newton's formula.

f :=

for i 0 .. 4
x a
i

i, 0

0.7651977

0.620086 0.48370567

for j 1 .. i
i, j

0
f = 0.4554022 0.548946 0.10873389
0.2818186 0.578612 0.04944333 0.0658784

for i 1 .. 4

0.1103623 0.571521

i , j 1

0
0

i1 , j 1

x x

i j

Then, use Newton's interpolatory divided-difference formula to construct polynomial.

i 1

p4( t) := f

0, 0

p4( t)

f
i =1

i, i

0.01181833 0.06806852 1.82510288 10

(t ak)

k =0

simplify
-2
2
-2 3
-3 4
.977735 + 7.3391310

t .343047t
+ 5.5292810

t + 1.8251010

t
float , 6

Then, find the approximate value of f(2) is

p4( 2) = 0.22387536

Finally, Plot the data points and Newton's interpolation polynomial in a graph below:
i := 0.. 4

t := 1 , 1.01.. 2.2
Graph of data points and Newton's poly.

bi
p4 ( t ) 0.5

0.5

1.5

2.5

ai , t

5. Reference
[1] R.L.Burden & J.D.Faires:"Numerical Analysis" 7th edition, Brooks/Cole ,2001.
[2] L.Fausett: "Numerical Methods Using Mathcad", Prentice Hall, 2002.
[3] R.K.Nagle, E.B.Staff & A.D.Snide: "Fundamentals of Differential Equations" 6th
edition , Addition Wesley, 2004
[4} D.T.Wu: "Using Mathcad in Teaching Differential Equations" , Electronic Proceedings
of ICTMT4, 1999.
Web site: www.tech.plym.ac.uk/maths/CTMHOME/ictmt4.html
[5] D.T.Wu: "Teaching Numerical Analysis with Mathcad", "International Journal of
Computer Algebra in Mathematics Education" Vol 10 No.1, 2003.