Preface

Computers have revolutionised mathematics and the many scientific, engineering, and economic fields in which mathematics is applied. In the applications of mathematics the role of computation has long been obvious and prominent. Now, the development of theorem proving software is increasing the prominence of computing in pure mathematics. What this means is that the ability to write computer programs well is an indispensable part of the toolkit of a 21st century mathematician and, indeed, scientist or engineer.

Regrettably, university courses in mathematics and its sibling disciplines have often failed to reflect this revolution in the way that mathematics is practised. Far too often the sum total of programming education in an undergraduate degree is a computational methods module in which students are shown elementary programming constructs such as functions, loops, and perhaps plotting. These are introduced in the context of solving particular computational mathematics or statistics problems, and the programming constructs introduced are driven by what is needed to solve those problems.

The assumptions underpinning this approach seem to be some combination of a belief that maths degrees should only teach maths, as well a feeling that the mathematical algorithms are some how the difficult part and that programming is a mere technical skill that students will somehow pick up along the way. The consequence of this approach is that many maths students and graduates end up without the programming, software development, and debugging skills that they need to make effective use of computers in their further studies or working careers.

What is this book for?

This book is the result of a significant change in the mathematics curriculum at Imperial College London. Rather than assume that students will somehow acquire programming skills along the way, we have introduced first and second year courses with the sole objective of teaching students to programme well. This book is the text for the second of these courses. Named “Principles of Programming”, the course aims to take students with a knowledge of basic Python programming (functions, loops, plotting and so forth) and introduce them to higher level programming concepts.

The objective of this course is to graduate better programmers. Material on new programming constructs and concepts is accompanied by chapters on good programming style, so that students learn how to write code that they and others can understand, and on errors, exceptions and debugging, so that students learn how to get themselves out of trouble.

An underlying theme throughout this book is that programming has strong connections with mathematics. In particular, both mathematics and programming depend on building higher level, more abstract objects which encapsulate and hide the underlying complexity of operations. It is in taking this mathematical approach to programming that this book is “for mathematicians”.

The examples are chosen from across mathematics. For example, it appears to be traditional in object oriented programming books to use a telephone directory as the initial example of classes. We eschew this in favour of a class implementing polynomials. This provides the opportunity to introduce encapsulation and operator overloading for a familiar mathematical object. In contrast to the traditional computational methods courses, the examples are chosen to illustrate and explain the programming concepts we study rather than the converse.

Who is this book for?

This book is for anyone with mathematical, scientific, or engineering interests who would like to learn to be a more capable programmer. The mathematical examples assume that you know how to differentiate functions of one variable, but very little beyond that. Where examples or exercises employ other mathematics, such as cellular automata in Chapter 4 and groups in Chapter 7, enough of the mathematics will be introduced that the reader should be able to understand the programming concept being explained, without necessarily understanding all of the mathematical details of the example.

This is not an introduction to basic Python: it’s assumed that the reader knows the sort of basic Python usually covered in a first programming or computational methods course. In particular the reader will be assumed to be familiar with writing functions, variable assignments, loops, and list comprehensions. The reader is also assumed to have used numeric and string data values, as well as dictionaries, sets, lists, and tuples.

Many introductory Python courses exclusively use Jupyter notebooks, so nothing beyond that is assumed. Getting set up with a working Python installation is covered in Chapter 1 while the Python command line and using a text editor to create programmes in files are introduced from scratch in Chapter 2.

How to use this book

You can, of course, simply sit down and read this book from cover to cover, or dip in to see what it has to say on particular subjects. However, reading a book about programming will not teach you to program. For that, you need to get your hands dirty writing code and debugging your mistakes. The videos and exercises throughout the book are designed to help you do this.

The videos

The videos were created to accompany the course at Imperial College London. They’re not primarily lecture videos but are instead practical demonstrations of the programming concepts being introduced at the relevant point. Usually it’s better to watch the video after reading the relevant section.

The exercises

At the end of each chapter are exercises. These usually depend on a skeleton code which is available on GitHub. Sometimes you might be asked to complete a piece of code while on other occasions you’ll need to write a whole Python module from scratch. Each set of exercises will come with a matching set of tests. These are small programs which check whether your code produces the correct responses to a range of inputs. Tests like this provide immediate feedback and enable you to know how you are doing. Links to the skeleton code for each chapter are provided at:

Conventions employed

Each chapter starts by introducing new material, supported by the videos and exercises. At the end of each chapter is a glossary containing many of the key concepts introduced in that chapter. Terms to be found in a glossary are given in italics and can be looked up in the index.

Python has excellent official online documentation, and we link to that throughout the text. External links show up in purple while internal links to other parts of the notes are blue.

The text sometimes introduces counterexamples: illustrations of code errors or bad implementation ideas. These will be flagged with a big red cross:

print "Hello World"

Conversely, if it’s necessary in context to highlight which approach is the correct one, the code will come with a big green tick:

print("Hello World")

Teaching this course elsewhere

The course of which this book forms the text has been given to master’s students at the University of Oxford, as well as to undergraduate students at Imperial College London. Instructors are welcome to use this material to teach elsewhere, and are encouraged to contact the author for assistance with access to materials.

Acknowledgements

The course Principles of Programming, and the notes on which this book is based, were first delivered in spring 2020, when university teaching was completely online during the COVID pandemic. I’d like to thank teaching fellow Dr Matthew Woolway who worked tirelessly with me on the module and who put together many of the tests on the exercises, and the graduate teaching assistants Miguel Boland, Sophia Vorderwuelbecke, and Connor Ward whose professionalism in delivering the course in very complex circumstances was outstanding. Pulling out all the stops to deliver the written and video materials for online learning meant a lot of evenings and weekends. I am exceptionally grateful to my wife Gebina Ham for disproportionately picking up our childcare responsibilities in that period in order to make this possible. I’d also like to thank Dr Aaron Pereira and Reuben Nixon-Hill for their eagle-eyed corrections to the text.

This is a textbook about programming in Python, so it would be remiss of me not to also thank the developers of the Python language, its CPython reference implementation, and all the third party packages which on which this book depends. In that regard, the developers of Numpy, Flake8, Pytest, PDB++ and IPython deserve particular mention.

This book is typeset using the Sphinx documentation system. Among other things this facilitates generating the web, PDF, and print versions of the book from a single source. Thanks are due to its authors as well as those of the underlying LaTeX and TeX typesetting systems.

Changes in the second and third editions

The second edition was a minor update correcting numerous small issues that have been pointed out over the last year. A more substantive change was that the explanation of packaging in Section 2 was modernised to use pyproject.toml in place of setup.py. Section 3.5 was added in response to confusion expressed by a number of students about the distinction between instantiating new objects and assigning new variable names to existing objects.

The third edition is similarly a minor update. In documents Visual Studio Code workspaces in Section 2.2.1, and the Flake8 extension in Section 4.2.2. Together, these provide correctly integrated code linting. It also consistently uses python -m to invoke tools such as pip, pytest and flake8. This is less error-prone for students. The description of finally in Section 6.5.2 has been also improved.