See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/226810761

Visualization and art in the mathematics classroom

Article in ZDM: the international journal on mathematics education · February 2003

DOI: 10.1007/BF02652763

CITATIONS READS

4 330

4 authors, including:

János Karsai Éva Veronika Rácz

University of Szeged Széchenyi István University
62 PUBLICATIONS 992 CITATIONS 13 PUBLICATIONS 31 CITATIONS

SEE PROFILE SEE PROFILE

Angela Schwenk
Berlin University of Applied Sciences
15 PUBLICATIONS 310 CITATIONS

SEE PROFILE

All content following this page was uploaded by János Karsai on 17 February 2014.

The user has requested enhancement of the downloaded file.

 IS - APMEF 2002
International Symposium
Anniversary of Pollack Mihály Engineering Faculty
May 31-June 1, 2002
Pécs, Hungary

VISUALIZATION AND ART IN THE MATHEMATICS CLASSROOM

János Karsai1

Department of Medical Informatics

University of Szeged, Szeged, Hungary
e-mail: karsai@silver.szote.u-szeged.hu

Éva Veronika Rácz

Department of Environmental Engineering

István Széchenyi University, Gyor, Hungary
e-mail: raczev@szif.hu

Angela Schwenk and Norbert Kalus

Department of Mathematics, Physics and Chemistry
University of Applied Sciences, Berlin, Germany
e-mail: schwenk@tfh-berlin.de, kalus@tfh-berlin.de

Key words: Computer-aided education, modelling, experiments, visualization, Mathematica

Mathematics subject classification: 97U50, 97U70

Abstract. In this paper we summarize our concepts and practice on computer-aided

mathematical experimentation, and illustrate them by Mathematica projects that we have
developed for our research and the courses ”Computer-aided mathematical modelling” and
“Computer Algebra I-II” held for students of life sciences at University of Szeged and
computational engineering at University of Applied Sciences Berlin.

1
Denotes the speaker.
 J. Karsai, É. Rácz, A. Schwenk, N. Kalus

1 Introduction

Computer visualization and modelling tools on high powered desktop computers with realistic
graphical capabilities have become an organic part of scientific research and education, and
their influence is still increasing (for example, see [2,3,4,6,8,10,13]). Such techniques are
widely used in almost every field of scientific research and engineering, even in the most
abstract disciplines such as mathematics. They help to understand and illustrate problems
since the phenomena in applied field s can be described by quite complicated mathematical
models. In particular, they can increase the effectiveness of the mathematics courses for life
science and engineering students.
Realizing this fact, several years ago we introduced computer tools in our traditional
classes, introduced new ones on computing methods, visualization and modelling. Although
engineering and life sciences seem to be far from each other, it turned out that they strengthen
each other. Hence, our joint research and developments are fruitful for both fields.
In this paper we summarize the cornerstones of our computer-aided teaching and some of
the results of our practice. Then, we illustrate the ideas with some of our experimental
projects prepared in Mathematica. We believe that they can be applied for other courses.

2 Teaching concepts

The fact that mathematics is "only" a tool for engineering and applied sciences can be
generally accepted. Students work with mathematical models and often do hard calculations
without a very deep study of the mathematical theory behind. Explorations, experiments, and
visualization can fasten the understanding since they are often more convincing than a
rigorous mathematical proof. Then, computer programs can substitute tiresome manual
calculations. In addition, we note that mathematical programming and computer
implementations of numerical methods are of basic importance for engineering students.
Hence, combining the traditional treatment and manual drill with experimental studies can
result in a deeper and more applicable mathematical knowledge, and it improves the creative
thinking.
According to the above statements, our teaching has three essentially different but more or
less overlapping phases:
- The theoretical mathematical courses essentially follow the traditional lines, where the
computer applications as demonstrational materials help to understand the theoretical
concepts and methods.
- In the computer-based courses (Computer-aided mathematical modelling [10,11,12],
Computer-algebra I-II [9]) the students learn special mathematical packages or general
computer-algebra systems, learn how to use them to perform mathematical algorithms.
- Finally, they solve practical problems during their whole study with the help of
computerized mathematical procedures and methods.

2
 Visualization and Art in the Mathematics Classroom

We need to take into account in our curriculum development that the main interest of our
students is directed toward their specific fields in engineering, physics, biology, chemistry,
medicine, pharmacy, etc.
What concerns the theoretical courses, we emphasize that computer is not used instead of
blackboard and chalk. We do use computer only if the traditional tools are not useful enough.
For the computer-based courses, the main points are as follows:
- the students would like to gain maximal benefit with "minimal" mathematical and
computer-technical effort;
- we emphasize, and prefer, computer-aided experimentation and visual arguments to the
formal mathematical methods;
- consequently, a large set of user- friendly modelling "skeletons" (recipes) and case studies
are prepared and used;
- in particular, the programming tools which are fundamental for the engineering students
are learned via applications;
- a teacher and an instructor hold the class together: the teacher gives the main line and the
instructor helps to solve the arising individual problems;
- the teacher and the students work on computers simultaneously;
- students use and improve the schemes to problems according to their particular interest.
We need to mention that the synchronization of courses has a significant influence on the
success since the technical tools should be used parallel to their learning. This problem has no
general resolution. Intensive courses cannot be usual held in any regular university teaching
system, but they are real solutions in some special cases. For "computational engineering"
postgraduate students at University of Applied Sciences Berlin and "mathematics" Ph.D.
students at University of Szeged we started to organize intensive Computer Algebra courses.
Such a course consists of two one-week long parts (beginning and advanced) at the beginning
of the semesters. These students own the necessary mathematical knowledge, and they can
immediately apply the learned methods and tools in their applied studies.

3 On the project development

There are a great number of books and computer developments published on computer
algebra applications (see the references). Most of them either deal with the details of the
technical tools, or use computer-algebra systems to help in the understanding of the
mathematical theory. Some deal with modelling problems, giving quite deep knowledge of
the considered problems, but do not give easily applicable "recipe-collections". Hence, using
the experiences with the published works, we have developed Mathematica projects that fit
completely the stated concepts of our courses. The projects can roughly be classified into the
following groups:
- illustrations to mathematical theories;
- introduction to the technical tools including the power of graphics;
- computer implementation of mathematical concepts and procedures, the modelling steps;
- projects to investigate elementary continuous and discrete models;
- projects to investigate advanced models;
- mathematical programming;

3
 J. Karsai, É. Rácz, A. Schwenk, N. Kalus

- tools and projects for interactive modelling.

The structure and the style are uniform in each project. During the development we follow
the following concepts:
- hyper- media structure;
- short mathematical and technical summaries;
- user-friendly computer implementation of theoretical methods;
- graphical explanations and arguments emphasizing the role of visualization methods such
as animation, colouring, and interactive methods;
- experimentation: constructive methods are preferred;
- numerical solution of mathematical systems is preferred; attention is paid to the
theoretical and numerical limitations;
- dangers of trusting only experiments are treated;
- well-prepared statements, programming examples, avoiding typing and syntactical
difficulties;
- simple and complex modelling schemes (skeletons), case-studies are presented.
These projects can help to understand the qualitative as well as the quantitative properties
of the considered phenomena. We believe that our projects improve the students' creativity
and scientific thinking.

4 Examples

Let us consider now some examples from the simplest up to the very advanced ones. It is not
possible to present the illustrations and projects in details, and the colouring and animations
cannot be shown either. Hence, we only summarize the main objects and the features.

4.1 Experiments with computer plots

Doing the simplest but essential experiments, the students learn the basic plotting features, as
well as the use and the dangers of the computer plots. For example, changing the number of
sample points can cause a dramatic change on the quality of the plots (Fig. 1). Pay attention to
possible singularities, watch the extremely large growth rate, error messages (zero divide, for
example). Note that some singularities cannot appear on the computer plots.
6 6

4 4

2 2

-6 -4 -2 2 4 6 -6 -4 -2 2 4 6

-2 -2

-4 -4

- 6 -6

Figure 1. High and low quality plots of tan(x)

4
 Visualization and Art in the Mathematics Classroom

4.2 Zooming in and out

It is a well-known method in video technique. “Zooming out” gives us a global plot of
functions, while “zooming in” visualizes the invisible (Fig. 2) and can help to understand
local properties such as continuity and linear approximation at a given point (differentiation).
3
Steigung
-1 - 2 - 3- 4 54 3 2
2.5
0.75

2 0.5

0.25
1.5

S t eigung
0.

1 -0.25

-0.5
0.5
-0.75

-1.
0.5 1 1.5 2 2.5 3

Figure 2. “Zooming in” illustrating the differentiability

4.3 Computer-aided investigation of function

After doing some graphical experiments on the function considered, the students use analytic
and/or numeric methods to describe the properties and find specific points (zeros, maxima,
minima, etc.). The importance of the initial guess for these points is emphasized.
1 1

0.5 0.5

0.5 1 1.5 2 2.5 3 0.5 1 1.5 2 2.5 3

- 0.5 - 0.5

-1 -1

Figure 3. Different initial points for the Newton iteration

4.4 Experiments with series expansions

Taylor approximation around a point, or Fourier series expansion of periodic technical
functions (square and triangle signals, etc.) are essential. We developed projects to estimate
iteratively the coefficients by graphical information. Other projects visualize the series
expansion and the spectrum by animations (Fig. 4). They can give an intuitive and visual
understand in these difficult mathematical constructions.

Cosine - Sine
1 0.2
0.15
0.75
0.1
0.5 0.05
0.25
-4 -2 2 4
-0.05
2 4 6 8 10 12 14 -0.1
-0.25
-0.15
-0.5
-0.2

5
 J. Karsai, É. Rácz, A. Schwenk, N. Kalus

Cosine - Sine
1 0.2
0.15
0.75
0.1
0.5 0.05
0.25
-4 -2 2 4
-0.05
2 4 6 8 10 12 14 -0.1
-0.25
-0.15
-0.5
-0.2

Figure 4. Two frames of a spectrum animation of a triangular wave

4.5 Experiments with wavelets

Basic ideas of wavelet theory can be taught at an undergraduate level by use of computer
experiments. This is an algorithmic approach. The Daubechies wavelet can be easily
m−1
constructed by the recursion formulae for the scaling function ϕ: ϕ (t ) = 2 ∑α ϕ (2t − k)
k =0
k

1+ 3 3+ 3 3− 3 1− 3
with the coefficients α 0 = ,α1 = ,α 2 = ,α 3 = .
4 2 4 2 4 2 4 2
This formulae can be evaluated iteratively at dyadic points k/2j, k,j ε ∧ starting with
1+ 3 1− 3
ϕ (0) = 0,ϕ (1) = , ϕ(2) = , ϕ (3) = 0, ϕ ( t ) = 0for t ∉ [0,3[ . Then from the scaling
2 2
m −1
function the wavelet ψ can be constructed by the formulae: ψ (t ) = 2 ∑ βk ϕ (2t − k ) . Both
k=0
functions are continuous everywhere and almost everywhere differentiable. However at the
dyadic points they are only differentiable from the left. This fact can be visualized by
zooming techniques [14,15].

1.5 1.5

1 1

0.5 0.5

-0.5 0.5 1 1.5 2 2.5 3 3.5 -0.5 0.5 1 1.5 2 2.5 3 3.5
-0.5 -0.5

-1 -1

Figure 5. Scaling function and wavelet.

4.6 Some mathematical graphics: functions in different coordinate systems

Plotting different surfaces can sometimes be an active relaxation among serious modelling
problems. Meanwhile our students can understand the different coordinate systems and
transformations. For example, plotting a function in Cartesian, cylindrical, polar, and toroid

6
 Visualization and Art in the Mathematics Classroom

systems can result in interesting surfaces (Fig. 6). We can enjoy the results. In addition, our
students can improve their visual thinking.

Figure 6. Waves on different surfaces

4.7 Visualizing trajectories

This simple but useful project is completed by the students on the third class. Essentially, it
contains the plot of a parametric curve and the tangent vectors. The students learn how to
visualize a fourth dimension (the time) in a 3D plot by using colouring, animations and
stroboscopic plots (Fig. 7). Later on, replacing the given curve by a solution of a differential
equation, this project is used for experiments for real modelling problems.
1
0.5
0
- 0.5 4
-1
1
3
0.5
0 2
- 0.5 1
-1
-1 1 2 3 4
- 0.5
0
0.5
1

Figure 7. Stroboscopic plots

4.8 Curve fitting projects

The students graphically and analytically investigate datasets (Fig. 8), use transformations and
filters if needed (they also learn these methods here), solve curve fitting problems, and try to
improve the fitting by finding better fitting families.

7
 J. Karsai, É. Rácz, A. Schwenk, N. Kalus

10

2.5 5 7.5 10 12.5 15 17.5

Figure 8. A non-homogeneous dataset

4.9 1D differential equations

The students learn the basic steps together with the corresponding computing tools of the
investigation of 1D ODE's concerning direction field, equilibriums, and their stability
properties without solving the equation; then the properties of the solutions and their
dependence on initia l conditions.
2.5

1.5

0.5

1 2 3 4

Figure 9. The direction field and solutions of the equation x’= x(2-x)-0.2

4.10 Critical reflection of hand made and computer made results

The students know how to solve an ordinary differential equation by separation of variables,
and they can verify the result with the help of a computer algebra system. Now, they plot the
slope field and the solution (Fig. 10), and then they can recognize that the obtained solution is
not a solution on the whole domain of the function and that computer solutions should be
checked carefully. The students check whether the validity of their calculation is limited by
indirectly given conditions. More critical reflections of the use of computers in teaching areas
can be found in [1].
6

-4 -2 2 4
-1

-2

Figure 10. Slope field of y' = x y + 1 with one solution

8
 Visualization and Art in the Mathematics Classroom

4.11 Models described by planar differential equations

The students investigate 2D vector fields, the trajectories in 2D and the integral curves in 3D.
They can improve now the project prepared for 2D and 3D parametric curve s (see Example
4.5). To investigate the properties of the equilibriums, they learn the method of Liapunov
functions. Finally, the students get acquainted with the linearization methods.
2

- 10 -5 5 10
-1

-2

Figure 11. The vector field of x’’ +0.3 x’ + sin(x)=0 and the contour lines of the energy
2

- 7.5 - 5 - 2.5 2.5 5 7.5

-1

-2

Figure 12. The trajectories and the contour lines of the energy

4.12 Experiments with impulsive effects

Since their formal description and behaviour is sometimes quite complicated, impulsive
systems are rarely considered in details in graduate courses [13]. Our projects deal with some
simple but important models, and help to understand the nature of such systems. One of them
considers the problem of the repeated drug dosage. Here the students investigate the relation
between the times elapsed between the doses, the importance of the first dose (Fig. 13). They
have to find an optimal setting to avoid either the under- and overdosing.

2.5

1.5

0.5

2 4 6 8 10

Figure 13. The dependence on the first dose

9
 J. Karsai, É. Rácz, A. Schwenk, N. Kalus

4.13 Iterative problems, difference equations, cellular automata

It is well known that iteration and recursion are basic mathematical algorithms. Hence, our
students have to get acquainted with the computer tools to investigate and solve iterative
problems from the simplest to the very advanced ones (see Examples 4.3 and 4.5). With a
little effort they are able to solve and visualize difference equations. Figure 14 shows some
solutions (“integral curves”) and the phase maps of an initial set of a planar difference system
that shows some kind of pulsation effect.
x1

1
0.5
0
-0.5
-1

0.5

0
x2

-0.5

5 -1

10

15
t
20

Figure 14. Solutions and phase maps of a pulsative difference system

Engineering students and PhD students in any applied science often need to do spatial
modelling whereas they are not skilled to tract partial differential equations. By discretising
partial differential equations we obtain partial difference equations or cellular automata. In
addition, they themselves are very important models of several real phenomena (such as
reaction-diffusion problems in ecology) or engineering problems. By defining the discrete
local rules, the behaviour of such a system can be easily simulated. Note that S. Wolfram
needed a software package for his cellular automata research, and hence he developed
Mathematica [16].
Among the introductory projects in this field, Figure 15 shows some states of a discrete wave.

2 2

1 1

5 10 15 20 5 10 15 20
-1 -1

-2 -2

2 2

1 1

5 10 15 20 5 10 15 20
-1 -1

-2 -2

Figure 15. Discrete 1D wave propagation and reflection

The study of space-time systems is even more complicated in higher dimensions both
theoretically and using computer simulations. To illustrate such cases, Figures 13 and 14
shows the evolution of a cellular automaton that describes the behavior of an ecological
system of two competing species.

10
 Visualization and Art in the Mathematics Classroom

Figure 16. A CA simulation of an ecological model of two competing species

N
1

0.8

0.6

0.4

0.2

t
20 40 60 80 100

Figure 17. Evolution of the frequencies of the two species in time

5 Math-Art in the classroom

G. H. Hardy [7] said, the beauty is the main criterion in mathematics. Ugly mathematics is
useless. The beauty of mathematics appears in the theories, but also in the form of artworks.
For common people mathematical art (different from computer art!) is kept almost identical to
iterative forms and fractals. In addition to their scientific importance, their beauty is obvious.
But math-art is much wider. We can find beauty in almost every, even in the simplest
mathematical construction. For illustration, see our examples, especially Example 4.6.
Artistic issues are very important in the mathematics classroom at all levels. Students can be
motivated to learn deeper theory, to get practice in mathematical experimentation by letting
them create artworks. The necessity of this motivation can be underlined by the fact that a big
number of students have some inhibitions about mathematics. We hope that after such math-
art experience our students will certainly be able to recognize mathematical structures in the
real world and hence the break between the abstract sciences and the real life can become
much narrower than it is now. Beyond lessons, successful exhibitions also demonstrated that

11
 J. Karsai, É. Rácz, A. Schwenk, N. Kalus

math-art approaches people to mathematics (see the web page http://silver.szote.u-szeged.hu).

Instead of more explanations we close our report with the surface described by the formula

 0. 5
  0 .5
 
 7u   7u 
 sin   + 0.4  cos (u ),  sin   + 0.4  sin (u ), 4 ( 0.02 sin ( 29u ) +1) sin (v )
0. 4

  2     2   
  

Figure 18. A flower as a parametric surface

References

[1] Berger, M., Schwenk, A., Mathematical Abilities of University Entrants and the
adapted Use of Computers in Engineering Education A., in Schriftenreihe Didaktik der
Mathematik, ed. BOROVCNIK – KAUTSCHITSCH, Volume 25, Technology in
Mathematics Teaching, Proceedings of ICTMT5 in Klagenfurt 2001, 477-480.
[2] Giordano, F. R., Weir, M. D., Fox W. P., A First Course in Mathematical Modeling,
Brooks/Cole Publishing Company, 1997.
[3] Gaylord, R. J., Wellin, P. R.., Computer Simulations with Mathematica, Telos-
Springer, 1995.
[4] Dreyer, T. P., Modelling with Ordinary Differential Equations, CRC Press, 1993.
[5] Forczek E., Karsai J., Computer visualization in the Mathematics Classroom,
Proceedings of the M/SET 2000-International Conference on Mathematics / Science
Education and Technology, San Diego CA, February 5-12, 2000.
[6] Gray, A., Mezzino, M., Pinsky, M. A., Introduction to Ordinary Differential Equations
with Mathematica, Springer-TELOS, 1997.
[7] Hardy, G. H., A Mathematician's Apology, Cambridge University Press, 1992.
[8] Hege, H. C., Polthier, K. (Eds.), Visualization and Mathematics, Experiments,
Simulations and Environments, Springer, 1997.

12
 Visualization and Art in the Mathematics Classroom

[9] Kalus, N., Computational Science and Engineering (CE) - Eine Neue Eigenständige
Disziplin für Postgraduale Masterstudiengänge, Global J. of Engng. Educ., Vol. 5, No.
3, 2001, 259-263.
[10] Karsai, J., Mathematics for Pharmacy Students (in Hungarian), A. Szent-Görgyi Med.
Univ., 1996.
[11] Karsai, J., Hantos Z., Computer curricula at Albert Szent-Györgyi Medical University:
infrastructure, programmes and development, Medical Informatics Europe '96,
Technology and Informatics 34, IOS Press, 1996, 838-842.
[12] Karsai, J. , Rácz É., Képíró I., Computer labs to improve the visual thinking and
intuition, Proceedings of the 10th SEFI MWG European Seminar on Mathematics in
Engineering Education, June 14-16, 2000, Miskolc, 73-79.
[13] Karsai, J., Mathematical models of impulsive phenomena, Mathematica experiments
and simulations, Typotex Budapest, to appear
[14] Nivergelt, Y., Wavelets Made Easy. Birkhäuser Boston, 1999.
[15] Walker, J. S., A Primer on Wavelets and their Scientific Applications, Chapman &
Hall/CRC, 1999.
[16] Wolfram, S., Mathematica, A System for Doing Mathematics by Computer, Addison-
Wesley Publishing Company, 1994.

13

View publication stats