027 Go To Index

TITLE
Ptolemy's Inequality

SOURCE
Dr John Pulham, Department of Mathematical Sciences University of Aberdeen
e-mail pulham@maths.abdn.ac.uk

AREA
history, metric spaces

KEYS
Ptolemy, inequalities, metric spaces

LEVEL
final year

LENGTH
1/8 final year (but short). 30? pages

PREREQ
a basic knowledge of metric spaces would be an advantage.

HISTORY
used once

DESCRIPTION

This short project really starts off from Apostol's paper [1]. This should be studied and reported on.
A general description of Ptolemy's work and of Alexandrian mathematics in general should be included.

References:

[1] T.M. Apostol "Ptolemy's inequality and the chordal metric" Math Mag. 40 (1967)
[2] C.B. Boyer "A History of Mathematics" Wiley 1948
[3] D.J. Schattschneider "A multiplicative metric" Math Mag 49 (1976)
[4] G.J. Toomer "Ptolemy's Almagest" Duckworth (1984)


028 Go To Index

TITLE
Square functions and curves

SOURCE
T.P.McDonough, Department of Mathematics, University of Wales, Aberystwyth.
e-mail:tpd@aber.ac.uk

AREA
Analysis

KEYS
periodic functions, approximating functions, Fourier series.

LEVEL
elementary undergraduate.

LENGTH
One of three investigations undertaken by students in a group of about four. The course is of one semester duration and is 1/12 of a student's degree commitment for the year. Each student presents a report on each of the topics investigated by the group. This report is the basis of the student's assessment.

PREREQ
A-level Mathematics.

HISTORY
The project has actually been used in this form.

DESCRIPTION

The graphs of two simple periodic functions are sketched -- one is a step function, the other is formed with line segments at $pi/4$ to the axes.
The periodic functions themselves may be described using functions such as sgn($x$), abs($x$) ( = $|x|$) and int($x$) ( = $\lfloor x \rfloor$) and the elementary functions. Find such descriptions for these and similar functions. (Perhaps, you might choose some simpler ones to start with).
These functions may also be approximated by continuous functions such as $$a_1 \sin kx + a_2 \sin 2kx + a_3 \sin 3kx$$ by choosing the constants $a_1$, $a_2$, $a_3$ and $k$ appropriately. Make a number of choices and quantify how well your approximations fit the functions.
The locus of the point ($\cos t$, $\sin t$), with $t$ real, is a circle centred at the origin. This may be considered as an approximation to a square curve obtained by inscribing a square in the circle.
Can you find a better approximation of the square curve using the trigonometric functions sine and cosine?
Can you describe the square curve parametrically in the form ($f(t),\ g(t)$) where $f(t)$ and $g(t)$ are expressions involving the simple functions sgn, abs and int?
Generalise to the case of a regular hexagon and a regular octagon.

References:

"Square functions"; J.Thomas; Math.Gaz.vol.72; no.462
"Square waves"; S.Morley; Math.Gaz.vol.71; no.456
"The Mathematical Experience"; P.J.Davis \& R.Hersh (Fourier Analysis)




029 Go To Index

TITLE
Constructibility of the Regular Polygons

SOURCE
Dr John Pulham, Department of Mathematical Sciences University of Aberdeen
e-mail pulham@maths.abdn.ac.uk

AREA
Geometry, Algebra, number theory

KEYS
geometric constructions, field extensions

LEVEL
final year

LENGTH
1/8 final year. 30-40 pages

PREREQ
a first course in rings and fields. Galois theory is not needed.

HISTORY
used

DESCRIPTION
A regular polygon is one which has all its sides of the same length and all its vertex angles the same size (and is convex?). A problem that was considered by the Greeks and which remained of interest for a long time afterwards was to determine which of the regular polygons can be constructed using a ruler and compass. It was finally solved by Gauss. The required result is that an $n$-gon can be so constructed iff $n=2^kp_1p_2\ldotsp_r$ where the $p$s are distinct primes of the form $2^{2^a}+1$.
Your job is to write an essay centred on this result. Space and personal choice will determine exactly what you put in and what you leave out, but some side issues that you might like to consider are (1) does it make a difference if the compasses hold their shape (modern ones) or collapse (Greek ones) when lifted off the paper? (2) Constructions with compasses alone. (3) Constructions using a ruler and a fixed circle. (4) Constructions where the rules are relaxed slightly in various ways.
The main elements of the proof of the main theorem are fields and groups. So the essay will be a mixture of geometry and algebra with a bit of history thrown in to set the scene.

References:

[1] Jones, Morris and Pearson : "Abstract Algebra and Famous impossibilities"
[2] C.R. Hadlock : "Field Theory and its Classical Problems"
[3] H. Dorrie : "100 Great problems of Elementary Mathematics"
[4] Dan Pedoe : "Geometry and the Liberal Arts"
[5] F. Richman : "Number Theory: An Introduction to Algebra"






030 Go To Index

TITLE
Chord envelopes

SOURCE
T.P.McDonough, Department of Mathematics, University of Wales, Aberystwyth.
e-mail:tpd@aber.ac.uk

AREA
Geometry

KEYS
Geometry, straight lines. envelopes, cardioid, epicycloid.

LEVEL
elementary undergraduate.

LENGTH
One of three investigations undertaken by students in a group of about four. The course is of one semester duration and is 1/12 of a student's degree commitment for the year. Each student presents a report on each of the topics investigated by the group. This report is the basis of the student's assessment.

PREREQ
A-level Mathematics.

HISTORY
The project has actually been used in this form.

DESCRIPTION

The circle can be described parametrically by $x(t) = \cos t$,$y(t) = \sin t$ and the corresponding point may be loosely referred to as the point $'t'$.
If $f(t)$ and $g(t)$ are two functions of $t$, the chord $C(t)$ joining the points $'f(t)'$ and $'g(t)'$ is also a function of $t$. As $t$ varies, the chords $C(t)$ are seen to be tangents to another curve which is known as their envelope.
Plot a selection of such chords when the functions $f(t)$ and $g(t)$ are of the following simple forms
\halign{\tabskip=20pt#&\hfill#&&${#}$\hfill\cr &(i)&f(t)=at+b&g(t)=ct+d\cr &(ii)&f(t)=at+b&g(t)=t^2 \cr }
Where possible, find parametric descriptions of these envelopes and plot them directly.
Make some other simple choices for the functions for $f(t)$ and $g(t)$ involving, for example, trigonometric functions, exponential functions, etc.
In the case of envelopes which are closed curves, find the area enclosed by them -- approximately, if necessary. By taking sufficiently many values of the parameters and finding the areas approximately, can you suggest how the area depends on the parameters?

References

[1] "Circles,chords and epicycloids"; A.F.\& L.A.Beardon; Math.Gaz.vol.73; no.465
[2] "The area of a cardioid"; W.W.Wilson; Math.Gaz.vol.67; no.441



031 Go To Index

TITLE
Factorization of Polynomials

SOURCE
Dr John Pulham, Department of Mathematical Sciences University of Aberdeen
e-mail pulham@maths.abdn.ac.uk

AREA
algebra, computer algebra

KEYS
computer algebra, polynomials, factorization

LEVEL
final year

LENGTH
1/8 final year. 30-40 pages

PREREQ
Basic knowledge of polynomials, modular arithmetic and Chinese remainder theorem

HISTORY
used once

DESCRIPTION
Computer algebra systems are now very efficient at factorizing large polynomials over the integers (and more general rings). Surprisingly, this process is essentially algorithmic whereas factorizing over the reals is certainly not. Factorizing $x^{15}+8x^{14}+18x^{13}+28x^{12}+23x^{11}+9x^{10}-25x^9 -49x^8-61x^7-23x^6+15x^5+41x^4+22x^3+3x^2-7x-3$$ as $$ (x-1)(x^2+5x-3)(x^6+x^5+x^4+x^3+x^2-2x-1)(x^2+x+1)^3$$ is almost trivial.
You are to study at least one of the currently available methods for performing such a factorization (e.g. Berlekamp's method). Present the mathematical theory of the method and show, by useful examples, how it can be applied.
You could also investigate the actual methods used by such packages as REDUCE, Maple and Mathematica.

References

[1] D. E. Knuth The Art of Computer Programming : Vol 2 Addison-Wesley
[2] R.B.J.T. Allenby `Rings, Fields and Groups'
[3] J.F. Humphreys & M.Y. Prest `Number, Groups and codes'




032 Go To Index

TITLE
Floor polynomials

SOURCE
T.P.McDonough, Department of Mathematics, University of Wales, Aberystwyth.
e-mail:tpd@aber.ac.uk

AREA
Analysis

KEYS
Step functions, polynomials, floor function, graphs, recursion.

LEVEL
elementary undergraduate.

LENGTH
One of three investigations undertaken by students in a group of about four. The course is of one semester duration and is 1/12 of a student's degree commitment for the year. Each student presents a report on each of the topics investigated by the group. This report is the basis of the student's assessment.

PREREQ
A-level Mathematics.

HISTORY
The project has actually been used in this form.

DESCRIPTION

Polynomial functions in one real variable and with real coefficients $$ a_nx^n+a_{n-1}x^{n-1}+\ldots+a_2x^2+a_1x+a_0 $$ are well-known.
Describe the principal properties of such functions. In particular, how do the the roots (their number and location) depend on the coefficients? And how do the regions in which the function is increasing or and decreasing,respectively, depend on the coefficients?
The {\em floor polynomial functions} in one real variable and with real coefficients are defined as follows $$ a_n\lfloor x^n\rfloor +a_{n-1}\lfloor x^{n-1}\rfloor $$
Study a selection of such functions of degrees 1, 2, 3 and 4, comparing them with the corresponding polynomial functions.
Determine the roots of these floor polynomial functions. How many roots do such functions have? Or does the concept of root need to be modified for such a question?
These functions will have a zero slope almost everywhere. Can you provide a new and appropriate concept.
Can you develop formulas for integrating these functions?
Modify other familiar concepts concerning functions so that you can make meaningful statements about the floor polynomials regarding these concepts.
\vspace{7pt} Families of polynomials may be described by the following iterative process: $$ f_n(x) = A(n,x)f_{n-1}(x) + B(n,x)f_{n-2}(x), \hspace{.25cm}n\ge 2 $$ where $f_0(x)$ is a constant, $f_1(x)$ is a linear polynomial in $x$ and $A(n,x)$ and $B(n,x)$ are simple polynomials in $n$ and $x$.
Examine some of these families when the functions $A(n,x)$ and $B(n,x)$ are particularly simple polynomials in $n$ and $x$, e.g. constant and/or linear in $n$ and/or $x$.
What statements can you make about their graphs, roots, integrals, etc. and about the corresponding floor polynomials?

References:

[1] "Legendre polynomials"; B.Symons; Math.Gaz.vol.66; no.436



033 Go To Index

TITLE
The Lorentz Group

SOURCE
Dr John Pulham, Department of Mathematical Sciences University of Aberdeen
e-mail pulham@maths.abdn.ac.uk

AREA
Group theory, Relativity

KEYS
Lorentz group, relativity, matrix groups

LEVEL
final year

LENGTH
1/8 final year. 30-40 pages

PREREQ
basic group theory and some special relativity

HISTORY
used once

DESCRIPTION
The aim of the project is to study fairly thoroughly the structure of the Lorentz group and its action on Minkowski space. The Lorentz group is the group of all 4 by 4 real matrices $L$ satisfying $ = $ where $$ is the `Minkowski inner product' $x_0y_0 - x_1y_1-x_2y_2-x_3y_3$.
Particular attention should be paid to those features in which the Lorentz group differs in its behaviour from that of the Orthogonal group. For example, null rotations.
The physical significance of the results should also be explored.




034 Go To Index

TITLE
Evaluation of Pi

SOURCE
Dr John Pulham, Department of Mathematical Sciences University of Aberdeen
e-mail pulham@maths.abdn.ac.uk

AREA
Analysis

KEYS
Elliptic Functions, Analysis

LEVEL
final year

LENGTH
1/8 final year, 30-40 pages

PREREQ
elementary analysis, perhaps some complex analysis

HISTORY
used once

DESCRIPTION
How to calculate the value of $\pi$ to a million decimal places.
Most of the traditional methods for calculating $\pi$ converge rather slowly. This makes it time-consuming to obtain even a few hundred decimal places of the value. One important exception to this rule is the Gauss AM/GM sequence method which provides good quadratic convergence and hence the possibility of calculating a huge number of decimal places.
The method relies on a link between the AM/GM process and the properties of elliptic integrals. This will provide the student with an excuse for studying the properties of complete elliptic integrals as a prelude to the $\pi$ calculations. A more extensive project could start from the theory of elliptic functions, though this would require a significant background in complex analysis.
Multiprecision real arithmetic is now available in many places (e.g. Maple and Mathematica) so there is no reason why the student should not implement the method.

References:

[1] Abramowitz and Stegun (eds) "Handbook of Mathematical Functions" (17) Dover


035 Go To Index

TITLE
Cubes and hypercubes

SOURCE
T.P.McDonough, Department of Mathematics, University of Wales, Aberystwyth.
e-mail:tpd@aber.ac.uk

AREA
Geometry

KEYS
Geometry, cubes, hypercubes, cross-sections.

LEVEL
elementary undergraduate.

LENGTH
One of three investigations undertaken by students in a group of about four. The course is of one semester duration and is 1/12 of a student's degree commitment for the year. Each student presents a report on each of the topics investigated by the group. This report is the basis of the student's assessment.

PREREQ
A-level Mathematics.

HISTORY
The project has actually been used in this form.

DESCRIPTION

Determine the different types of plane cross-section of a cube. Are there other 3-dimensional solids with the same types of cross-section exactly?
What are the different types of 1-dimensional cross-section of the cube?
Describe the 4-dimensional analogue of the cube -- the 4-cube. Determine the different types of 1-, 2- and 3-dimensional cross-section of the 4-cube.
Continue this investigation for the $n$-cube where $n = 5, 6, \dots$

References:

[1] "A Steiner-type net for a cube"; J.Costello; Math.Gaz.vol.75; no.474
[2] "How to iron a hypercube"; A.W.F.Edwards; Math.Gaz.vol.75; no.474
[3] "The Mathematical Experience"; P.J.Davis \& R.Hersh (Four-dimensional Intuition)




036 Go To Index

TITLE
Integration in REDUCE - the Risch Algorithm

SOURCE
Dr John Pulham, Department of Mathematical Sciences University of Aberdeen
e-mail pulham@maths.abdn.ac.uk

AREA
computer algebra

KEYS
computer algebra, REDUCE, integration

LEVEL
final year.

LENGTH
1/8 final year. 30-40 pages

PREREQ
Some experience of field extensions is useful. Access to REDUCE is important.

HISTORY
used

DESCRIPTION
A computer algebra system like REDUCE is very effective at the formal evaluation of integrals. The full internal process being used, which should be studied, is complex but has at its heart a methodology based on the Risch Algorithm.
Study this algorithm and its application.

References

[1] R.H. Risch "The problem of integration in finite terms" Trans AMS 139 167-189
[2] Davenport, Siret and Tournier "Computer Algebra: systems and algorithms for algebraic computation" Academic Press
[3] Buchberger, Collins, Loos and Albrecht "Computer Algebra: symbolic and algebraic computation" Springer
[4] MacCallum and Wright "Algebraic computing with REDUCE" OUP
[5] Geddes, Czapor & Labahn "Algorithms for Computer Algebra" Kluwer