Principles of Programming

Jon Reades

According to the BBC & Google, programming is about:

  • Decomposition: breaking down a complex problem or system into smaller, more manageable parts.
  • Pattern recognition: looking for similarities among and within problems.
  • Abstraction: focusing on the important information only, ignoring irrelevant detail.
  • Algorithms: developing a step-by-step solution to the problem, or the rules to follow to solve the problem.

What About Computational Thinking?

Computational thinking is not thinking like a Computer Scientist. It is about recognising how to code can help us to understand, and manipulate, the world.

Key Features

Aspects of computational thinking include:

  • Recognising how one problem connects to other problems.
  • Recognising when and how to make things simpler and faster
  • Recognising how different ways of tackling a problem gives you power to tackle new problems.

See this keynote by Lorena Barba (2014); esp. from 52:00 onwards.

Why Are These Virtues?

According to Larry Wall the three virtues of the programmer are:

  1. Laziness
  2. Impatience
  3. Hubris

These are not to be taken literally (see Larry Wall’s “Three Virtues of a Programmer” are Utter Bull💩).

Four Quotes to Remember

  1. Computers make very fast, very accurate mistakes.
  2. A computer program does what you tell it to do, not what you want it to do.
  3. Only half of programming is coding. The other 90% is debugging.
  4. Weeks of coding can save you hours of planning.

And One More…

> Experience is the name everyone gives to their mistakes.

Oscar Wilde

Algorithm as Recipe…

Following a Recipe is Easy, Right?

Source

Calculating the Mean

Given these numbers, what’s the average?

1, 4, 7, 6, 4, 2

As a Recipe

  1. Take a list of numbers
  2. Start a count of numbers in the list at 0
  3. Start a sum of numbers in the list at 0
  4. Take a number from the list:
    • Add 1 to the count
    • Add the value of the number to the sum
  5. Repeat step #4 until no numbers are left in the list.
  6. Divide the sum by the count
  7. Report this number back to the user

As Python

numbers = [1, 4, 7, 6, 4, 2]
total   = 0
count   = 0
for num in numbers:
  total = total + num 
  count = count + 1
mean = total / count
print(mean)

Why We Still Haven’t ‘Solved It’

Source

Computers are good at:

  • Doing exactly what they are told.
  • Doing it over and over indefinitely.

Computers are not good at:

  • Solving problems that you can’t define.
  • Creating new knowledge.

Languages

Computer languages come with all of the ‘baggage’ of human languages; they have:

  • A vocabulary (reserved words)
  • A grammar (syntax)
  • Rules about the kinds of things you can say (grammar)
  • Styles and idiosyncrasies all their own (history)

In this module we will use the Python programming language. We could also teach this same content in R.

Python

numbers = [1, 4, 7, 6, 4, 2]
total   = 0
count   = 0
for num in numbers:
  total = total + num 
  count += 1 # An alternative to count = count + 1
mean = total / count
print(mean)

R

numbers = c(1, 4, 7, 6, 4, 2)
total   = 0
count   = 0
for (num in numbers) {
  total = total + num
  count = count + 1
}
mean = total / count
print(mean)

Finally: Style

Although all programmers develop their own style (see: writing in any language), Python encourages coders to use a consistent style so that others can pick up your code and make sense of what’s going on (see: comments!).

Two useful resources:

> It’s okay to figure out murder mysteries, but you shouldn’t need to figure out code. You should be able to read it.

Steve McConnell

Source: xkcd

> Science is what we understand well enough to explain to a computer. Art is everything else we do.

Donald Knuth, Things a Computer Scientist Rarely Talks About

Principles of Programming • Jon Reades