PYTHON IN HIGH SCHOOL

ARNAUD BODIN

A L G O R I T H M S A N D M AT H E M AT I C S

Exo7
Python in high school
Let’s go!
Everyone uses a computer, but it’s another thing to drive it! Here you will learn the basics of programming.
The objective of this book is twofold: to deepen mathematics through computer science and to master
programming with the help of mathematics.

Python
Choosing a programming language to start with is tricky. You need a language with an easy handling,
well documented, with a large community of users. Python has all these qualities and more. It is modern,
powerful and widely used, including by professional programmers.
Despite all these qualities, starting programming (with Python or another language) is difficult. The best
thing is to already have experience with the code, using Scratch for example. There is still a big walk to
climb and this book is there to accompany you.

Objective
Mastering Python will allow you to easily learn other languages. Especially the language is not the most
important, the most important thing is the algorithms. Algorithms are like cooking recipes, you have to
follow the instructions step by step and what counts is the final result and not the language with which the
recipe was written. This book is therefore neither a complete Python manual nor a computer course, nor is
it about using Python as a super-calculator.
The aim is to discover algorithms, to learn step-by-step programming through mathematical/computer
activities. This will allow you to put mathematics into practice with the willingness to limit yourself to the
knowledge acquired during the first years.

Mathematics for computer science

Computer science for mathematics
Since computers only handle numbers, mathematics is essential to communicate with them. Another
example is the graphical on-screen display that requires a good understanding of the coordinates (x, y),
trigonometry. . . .
Computers are a perfect match for mathematics! The computer becomes essential to manipulate very large
numbers or to test conjecture on many cases. In this book you will discover fractals, L-systems, brownian
trees and the beauty of complex mathematical phenomena.

You can retrieve all the activity Python codes and all the source files on the Exo7 GitHub page:
GitHub: Python in high school
 Contents

I Getting started 1

1 Hello world! 2

2 Turtle (Scratch with Python) 9

II Basics 17

3 If ... then ... 18

4 Functions 24

5 Arithmetic – While loop – I 33

6 Strings – Analysis of a text 40

7 Lists I 50

III Advanced concepts 58

8 Statistics – Data visualization 59

9 Files 68

10 Arithmetic – While loop – II 77

11 Binary I 82

12 Lists II 89

13 Binary II 95

IV Projects 98

14 Probabilities – Parrondo’s paradox 99

15 Find and replace 102

16 Polish calculator – Stacks 107

17 Text viewer – Markdown 119

18 L-systems 126

19 Dynamic images 134

20 Game of life 140

21 Ramsey graphs and combinatorics 146

22 Bitcoin 155

23 Random blocks 164

V Guides 171

24 Python survival guide 172

25 Main functions 181

26 Notes and references 197

Index
Summary of the activities
Hello world!
Get into programming! In this very first activity, you will learn to manipulate numbers, variables and code your
first loops with Python.

Turtle (Scratch with Python)

The module turtle allows you to easily make drawings in Python. It’s about ordering a turtle with simple
instructions like “go ahead”, “turn”. . . It’s the same principle as with Scratch, but with one difference: you no
longer move blocks, but you write the instructions.

If ... then ...

The computer can react according to a situation. If a condition is met, it acts in a certain way, otherwise it does
something else.

Functions
Writing a function is the easiest way to group code for a particular task, in order to execute it once or several
times later.
Arithmetic – While loop – I
The activities in this sheet focus on arithmetic: Euclidean division, prime numbers . . . This is an opportunity to
use the loop “while” intensively.

Strings – Analysis of a text

You’re going to do some fun activities by manipulating strings and characters.

Lists I
A list is a way to group elements into a single object. After defining a list, you can retrieve each item of the list
one by one, but also add new ones. . .

Statistics – Data visualization

It’s good to know how to calculate the minimum, maximum, average and quartiles of a series. It’s even better to
visualize them all on the same graph!

Files
You will learn to read and write data with files.

Arithmetic – While loop – II

Our study of numbers is further developed with the loop “while”. For this chapter you need your function
is_prime() built in the part “Arithmetic – While loop – I”.
Binary I
The computers transform all data into numbers and manipulate only those numbers. These numbers are stored
in the form of lists of 0’s and 1’s. It’s the binary numeral system of numbers. To better understand this binary
numeral system, you will first understand the decimal numeral system better.

Lists II
The lists are so useful that you have to know how to handle them in a simple and efficient way. That’s the
purpose of this chapter!

Binary II
We continue our exploration of the world of 0 and 1.

Probabilities – Parrondo’s paradox

You will program two simple games. When you play these games, you are more likely to lose than to win.
However, when you play both games at the same time, you have a better chance of winning than losing! It’s a
paradoxical situation.

Find and replace

Finding and replacing are two very frequent tasks. Knowing how to use them and how they work will help you
to be more effective.
Polish calculator – Stacks
You’re going to program your own calculator! For that you will discover a new notation for formulas and also
discover what a “stack” is in computer science.

Text viewer – Markdown

You will program a simple word processor that displays paragraphs cleanly and highlights words in bold and
italics.
L-systems
The L-systems offer a very simple way to code complex phenomena. From an initial word and replacement
operations, we arrive at complicated words. When you “draw” these words, you get beautiful fractal figures.
The “L” comes from the botanist A. Lindenmayer who invented the L-systems to model plants.

Dynamic images
We will distort images. By repeating these distortions, the images become blurred. But by a miracle after a
certain number of repetitions the original image reappears!

Game of life
The game of life is a simple model of the evolution of a population of cells that born and die over time. The
“game” consists in finding initial configurations that give interesting evolution: some groups of cells disappear,
others stabilize, some move. . .

Ramsey graphs and combinatorics

You will see that a very simple problem, which concerns the relationships between only six people, will require a
lot of calculations to be solved.
Bitcoin
The bitcoin is a dematerialized and decentralized currency. It is based on two computer principles: public key
cryptography and proof of work. To understand this second principle, you will create a simple model of bitcoin.

Random blocks
You will program two methods to build figures that look like algae or corals. Each figure is made up of small
randomly thrown blocks that stick together.
 PA R T I

G E T T I N G S TA R T E D

1
 Chapter

1
Hello world!

Get into programming! In this very first activity, you will learn to manipulate numbers, variables
and code your first loops with Python.

Lesson 1 (Numbers with Python).

Check in the console that Python works correctly, by typing the following commands in a Python console:
>>> 2+2
>>> "Hello world!"
Here are some instructions.
• Addition. 5+7.
• Multiplication. 6*7; with brackets 3*(12+5); with decimal numbers 3*1.5.
• Power. 3**2 for 32 = 9; negative power 10**-3 for 10−3 = 0.001.
• Real division. 14/4 is equal to 3.5; 1/3 is equal to 0.3333333333333333.
• Integer division and modulo.
– 14//4 returns 3: it is the quotient of the Euclidean division of 14 by 4, note the double slash;
– 14%4 returns 2: it is the remainder of the Euclidean division of 14 by 4, we also say “14
modulo 4”.
Note. Inside the computer, decimals numbers are encoded as “floating point numbers”.

Activity 1 (First steps).

Goal: make your first calculations with Python.

1. How many seconds are there in a century? (Do not take leap years into account.)
2. How far do you have to complete the dotted formula to obtain a number greater than one billion?
(1 + 2) × (3 + 4) × (5 + 6) × (7 + 8) × · · ·
3. What are the last three digits of
123456789 × 123456789 × · · · ?
| {z }
7 instance of 123456789
4. 7 is the first integer such that its inverse has a repeating decimal representation of period 6:
1
= 0. |
142857
{z } 142857
| {z } 142857
| {z } . . .
7
Find the first integer whose inverse has a repeating decimal representation of period 7:
1
= 0.00 a bcd e f g a bcd e f g . . .
??? | {z } | {z }
HELLO WORLD! 3

Hint. The integer is bigger than 230!

5. Find the unique integer:
• which gives a quotient of 107 when you divide by 11 (with integer division),
• and which gives a quotient of 90 when you divide by 13 (with integer division),
• and which gives a remainder equal to 6 modulo 7.

Lesson 2 (Working with an editor).

From now on, it is better if you work in a text editor dedicated to Python. You must then explicitly ask
to display the result:
print(2+2)
print("Hello world!")
In the following you continue to write your code in the editor but we will no longer indicate that you
must use print() to display the results.

Lesson 3 (Variables).
Variable. A variable is a name associated with a memory location. It is like a box that is identified by a
label. The command “a = 3” means that I have a variable “a” associated with the value 3.
Here is a first example:
a = 3 # One variable
b = 5 # Another variable

print("The sum is",a+b) # Display the sum

print("The product",a*b) # Display the product

c = b**a # New variable...

print(c) # ... that is displayed

Comments. Any text following the sharp character “#” is not executed by Python but is used to explain
the program. It is a good habit to comment extensively on your code.
Names. It is very important to give a clear and precise name to the variables. For example, with the right
names you should know what the following code calculates:
base = 8
height = 3
area = base * height / 2
print(area)
# print(Area) # !! Error !!
Attention! Python is case sensitive. So myvariable, Myvariable and MYVARIABLE are different vari-
ables.
Re-assignment. Imagine you want to keep your daily accounts. You start with S0 = 1000, the next day
you earn 100, so now S1 = S0 + 100; the next day you add 200, so S2 = S1 + 200; then you lose 50, so on
the third day S3 = S2 − 50. With Python you can use only one variable S for all these operations.
S = 1000
S = S + 100
HELLO WORLD! 4

S = S + 200
S = S - 50
print(S)
You have to understand the instruction “S = S + 100” like this: “I take the contents of the box S, I add
100, I put everything back in the same box”.

Activity 2 (Variables).

Goal: use variables!

1. (a) Define variables, then calculate the area of a trapezoid. Your program should display "The
value of the area is ..." using print("The value of the area is",area).
b=4

h=3

B=7

(b) Define variables to calculate the volume of a box (a rectangular parallelepiped) whose
dimensions are 10, 8, 3.
(c) Define a variable PI equals to 3.14. Define a radius R = 10. Write the formula for the area of
a disc of radius R.
2. Put the lines back in order so that, at the end, x has the value 46.

(1) y = y - 1
(2) y = 2*x
(3) x = x + 3*y
(4) x = 7

3. You place the sum of 1000 dollars in a savings account. Each year the interest on the money
invested brings in 10% (the capital is multiplied by 1.10). Write the code to calculate the capital
for the first three years.
4. I define two variables by a = 9 and b = 11. I would like to exchange the content of a and b.
Which instructions agree so that at the end a equals 11 and b equals 9?

c = a
c = b c = a
a = b a = c
a = b a = b
b = a c = b
b = c b = c
b = c

Lesson 4 (Use functions).

• Use functions of Python.
You already know the print() function that displays a string of characters (or numbers). It can be
HELLO WORLD! 5

used by print("Hi there.") or through a value:

string = "Hi there."

print(string)

There are many other functions. For example, the function abs() calculates the absolute value:
abs(-3) returns 3, abs(5) returns 5.
• The module math.
Not all functions are directly accessible. They are often grouped in modules. For example, the
module math contains the mathematical functions. For instance, you will find the square root
function sqrt(). Here’s how to use it:

from math import *

x = sqrt(2)
print(x)
print(x**2)
p
The first line imports all the functions of the module named math, the second line calculates x = 2
(in approximate value) and then displays x and x 2 .
• Sine and cosine.
The module math contains the trigonometric functions sine and cosine and even the constant pi
which is an approximate value of π. Be careful, the angles are expressed in radians.
Here is the calculation of sin( π2 ).

angle = pi/2
print(angle)
print(sin(angle))

• Decimal to integer.
In the module math there are also functions to round a decimal number:
– round() rounds to the nearest integer: round(5.6) returns 6, round(1.5) returns 2.
– floor() returns the integer less than or equal to: floor(5.6) returns 5.
– ceil() returns the integer greater than or equal to: ceil(5.6) returns 6.

Activity 3 (Use functions).

Goal: use functions from Python and the module math.
1. The function Python for gcd is gcd(a,b) (for greatest common divisor). Calculate the gcd of
a = 10 403 and b = 10 506. Deduce the lcm from a and b. The function lcm does not exist, you
must use the formula :
a×b
lcm(a, b) = .
gcd(a, b)
2. Find by trial and error a real number x that checks all the following conditions (several solutions
are possible):
• abs(x**2 - 15) is less than 0.5
• round(2*x) returns 8
HELLO WORLD! 6

• floor(3*x) returns 11
• ceil(4*x) returns 16
Hint. abs() refers to the absolute value function.
3. You know the trigonometric formula
cos2 θ + sin2 θ = 1.
Check that for θ = π7 (or other values) this formula is numerically true (this is not a proof of the
formula, because Python only makes approximate computations of the sine and cosine).

Lesson 5 (Loop “for”).

The loop “for” is the easiest way to repeat instructions.

keywords "for" and "in"

a variable
list to browse, here 0, 1, 2, . . . , n − 1
colon
for i in range(n) :
instruction_1 indented block of instructions
instruction_2 will be repeated n times
... for i = 0, then i = 1. . . up to i = n − 1
other instructions
indentation
program continuation

Note that what delimits the block of instructions to be repeated is indentation, i.e. the spaces at the
beginning of each line that shift the lines to the right. All lines in a block must have exactly the same
indentation. In this book, we choose an indentation of 4 spaces.
Don’t forget the colon “:” at the end of the line of the for declaration!
• Example of a “for” loop.
Here is a loop that displays the squares of the first integers.

for i in range(10):
print(i*i)

The second line is shifted and constitutes the block to be repeated. The variable i takes the value 0
and the instruction displays 02 ; then i takes the value 1, and the instruction displays 12 ; then 22 ,
32 . . .
In the end this program displays:
0, 1, 4, 9, 16, 25, 36, 49, 64, 81.
Warning: the last value taken by i is 9 (and not 10).
• Browse any list.
The loop “for” allows you to browse any list. Here is a loop that displays the cube of the first prime
numbers.
HELLO WORLD! 7

for p in [2,3,5,7,11,13]:
print(p**3)

• Sum all.
Here is a program that calculates
0 + 1 + 2 + 3 + · · · + 18 + 19.

mysum = 0
for i in range(20):
mysum = mysum + i
print(mysum)

Understand this code well: a variable mysum is initialized at 0. We will add 0, then 1, then 2. . . This
loop can be better understood by filling a table:
Initialisation: mysum= 0
i mysum
0 0
1 1
2 3
3 6
4 10
... ...
18 171
19 190
Display: 190

• range().
– With range(n) we run the entire range from 0 to n − 1. For example range(10)
corresponds to the list [0, 1, 2, 3, 4, 5, 6, 7, 8, 9].
Attention! the list stops at n − 1 and not at n. What to remember is that the list contains
n items (because it starts at 0).
– If you want to display the list of items browsed, you must use the command:
list(range(10))
– With range(a,b) we go through the elements from a to b − 1. For example
range(10,20) corresponds to the list [10, 11, 12, 13, 14, 15, 16, 17,
18, 19].
– With range(a,b,step) you can browse the items a, a + step, a + 2step. . . For example
range(10,20,2) corresponds to the list [10, 12, 14, 16, 18].

• Nested loops.
It is possible to nest loops, i.e. inside the block of a loop, a new loop is used.

for x in [10,20,30,40,50]:
for y in [3,7]:
print(x+y)
HELLO WORLD! 8

In this small program x is first equal to 10, y takes the value 3 then the value 7 (so the program
displays 13, then 17). Then x = 20, and y again equals 3, then 7 again (so the program displays
23, then 27). Finally the program displays:
13, 17, 23, 27, 33, 37, 43, 47, 53, 57.

Activity 4 (“for” loop).

Goal: build simple loops.

1. (a) Display the cubes of integers from 0 to 100.

(b) Display the fourth powers of integers from 10 to 20.
(c) Display the square roots of integers 0, 5, 10, 15,. . . up to 100.
2. Display the powers of 2, from 21 to 210 , and memorize the results!
3. Experimentally search for a value close to the minimum of the function
1
f (x) = x 3 − x 2 − x + 1
4
on the interval [0, 1].
Hints.
• Build a loop in which a variable i scans integers from 0 to 100.
i
• Defined x = 100 . So x = 0.00, then x = 0.01, x = 0.02. . .
• Calculate y = x − x 2 − 41 x + 1.
3

• Display the values using print("x =",x,"y =",y).

• Search by hand for which value of x you get the smallest possible y.
• Feel free to modify your program to increase accuracy.
4. Seek an approximate value that must have the radius R of a ball so that its volume is 100.
Hints.
• Use a scanning method as in the previous question.
• The formula for the volume of a ball is V = 43 πR3 .
• Display values using print("R =",R,"V =",V).
• For π you can take the approximate value 3.14 or the approximate value pi of the module
math.

Activity 5 (Loop “for” (continued)).

Goal: build more complicated loops.

1. Define a variable n (for example n = 20). Calculate the sum

12 + 22 + 32 + · · · + i 2 + · · · + n2 .
2. Calculate the product:
1 × 3 × 5 × · · · × 19.
Hints. Begin by defining a variable myproduct initialized to the value 1. Use range(a,b,2) to
get every other integer.
3. Display multiplication tables between 1 and 10. Here is an example of a line to display:
7 x 9 = 63
Use a display command of the style: print(a,"x",b,"=",a*b).
 Chapter

Turtle (Scratch with Python)

2

The module turtle allows you to easily make drawings in Python. It’s about ordering a turtle
with simple instructions like “go ahead”, “turn”. . . It’s the same principle as with Scratch, but with
one difference: you no longer move blocks, but you write the instructions.

Lesson 1 (The turtle Python).

The turtle is the ancestor of Scratch! In a few lines you can make beautiful drawings.
from turtle import *

forward(100) # Move forward

left(90) # Turn 90 degrees left
forward(50)
width(5) # Width of the pencil
forward(100)
color('red')
right(90)
forward(200)

exitonclick()

Here is a list of the main commands, accessible after writing:

from turtle import *
• forward(length) advances a number of steps
• backward(length) goes backwards
• right(angle) turns to the right (without advancing) at a given angle in degrees
• left(angle) turns left
TURTLE (SCRATCH WITH PYTHON) 10

• setheading(direction) points in one direction (0 = right, 90 = top, −90 = bottom, 180 =

left)
• goto(x,y) moves to the point (x, y)
• setx(newx) changes the value of the abscissa
• sety(newy) changes the value of the ordinate
• down() sets the pen down
• up() sets the pen up
• width(size) changes the thickness of the line
• color(col) changes the color: "red", "green", "blue", "orange", "purple". . .
• position() returns the (x, y) position of the turtle
• heading() returns the direction angle to which the turtle is pointing
• towards(x,y) returns the angle between the horizontal and the segment starting at the turtle
and ending at the point (x, y)
• exitonclick() ends the program as soon as you click
The default screen coordinates range from −475 to +475 for x and from −400 to +400 for y; (0, 0) is in
the center of the screen.
y

300

200

100

(0, 0)

−400 −300 −200 −100 100 200 300 400 x

−100

−200

−300

Activity 1 (First steps).

Goal: picture your first drawings.

Trace the first letters of Python, for example as below.

TURTLE (SCRATCH WITH PYTHON) 11

Activity 2 (Figures).

Goal: drawing geometric shapes.

1. Pentagon. Draw a first pentagon (in blue). You have to repeat 5 times: advance 100 steps, turn 72
degrees.
Hint. To build a loop, use
for i in range(5):
(even if you do not use the variable i).
2. Pentagon (bis). Define a variable length which is equal to 200 and a variable angle which is
equal to 72 degrees. Draw a second pentagon (in red), this time advancing by length and turning
by angle.
3. Dodecagon. Draw a polygon having 12 sides (in purple).
Hint. To draw a polygon with n sides, it is necessary to turn an angle of 360/n degrees.
4. Spiral. Draw a spiral (in green).
Hint. Build a loop, in which you always turn at the same angle, but on the other hand you move
forward by a length that increases with each step.

Activity 3 (Function graph).

Goal: draw the graph of a function.

TURTLE (SCRATCH WITH PYTHON) 12

Plot the graph of the square function and the sine function.
In order to get a curve in the turtle window, repeat for x varying from −200 to +200 :
1
• set y = 100 x 2,
• go to (x, y).
For the sinusoid, you can use the formula
1
‹

y = 100 sin x .
20
By default Python does not know the sine function, to use sin() you must first import the module
math:
from math import *
To make the turtle move faster, you can use the command speed("fastest").

Activity 4 (Sierpinski triangle).

Goal: trace the beginning of Sierpinski’s fractal by nesting loops.

Here is how to picture the second drawing. Analyze the nesting of the loops and draw the next pictures.

for i in range(3):
color("blue")
forward(256)
left(120)

for i in range(3):
color("red")
forward(128)
left(120)
TURTLE (SCRATCH WITH PYTHON) 13

Activity 5 (The heart of multiplication tables).

Goal: draw the multiplication tables.

We set an integer n. We are studying the 2 table, that is to say we calculate 2 × 0, 2 × 1, 2 × 2, up to

2 × (n − 1). In addition, the calculations will be modulo n. We therefore calculate
2 × k (mod n) for k = 0, 1, . . . , n − 1
How to draw this table?
We place on a circle, n points numbered from 0 to n − 1. For each k ∈ {0, . . . , n − 1}, we connect the
point number k with the point number 2 × k (mod n) by a segment.
Here is the layout, from the table of 2, modulo n = 10.

3 2

4 1

5 0

6 9

7 8

For example:
• the 3 point is linked to the 6 point, because 2 × 3 = 6 ;
• the 4 point is linked to the 8 point, because 2 × 4 = 8 ;
• the 7 point is linked to the 4 point, because 2 × 7 = 14 = 4 (mod 10).

Draw the table of 2 modulo n, for different values of n.

Here is what it gives for n = 100.
TURTLE (SCRATCH WITH PYTHON) 14

Hints. For calculations modulo n, use the expression (2*k) % n.

Here’s how to get the coordinates of the vertices. This is done with the sine and cosine functions (available
from module math). The coordinates (x i , yi ) of the vertex number i, can be calculated by the formula :
2iπ 2iπ
 ‹  ‹
x i = r cos et yi = r sin
n n
These points will be located on the circle of radius r, centered at (0, 0). You will have to choose r rather
large (for example r = 200).

(x 1 , y1 )

(x i , yi ) 2iπ


yi = r sin n

(x 0 , y0 )
2iπ (0, 0) x


x i = r cos n

(x n−1 , yn−1 )

Lesson 2 (Several turtles).

Several turtles can be defined and move independently on their own. Here’s how to define two turtles
(one red and one blue) and move them.
turtle1 = Turtle() # with capital 'T'!
turtle2 = Turtle()
TURTLE (SCRATCH WITH PYTHON) 15

turtle1.color('red')
turtle2.color('blue')

turtle1.forward(100)
turtle2.left(90)
turtle2.forward(100)

Activity 6 (The pursuit of turtles).

Goal: draw tracking curves.

Program four turtles running one after the other:

• the turtle 1 runs after the turtle 2,

• the turtle 2 runs after the turtle 3,
• the turtle 3 runs after the turtle 4,
• the turtle 4 runs after the turtle 1.

Here are the starting positions and orientations:

turtle 4 turtle 3

turtle 1 turtle 2
TURTLE (SCRATCH WITH PYTHON) 16

Hints. Use the following piece of code:

position1 = turtle1.position()
position2 = turtle2.position()
angle1 = turtle1.towards(position2)
turtle1.setheading(angle1)

• You place turtles at the four corners of a square, for example in (−200, −200), (200, −200),
(200, 200) and (−200, 200).
• You get the position of the first turtle by position1 = turtle1.position(). Same for the
other turtles.
• You calculate the angle between turtle 1 and turtle 2 by the command angle1 =
turtle1.towards(position2).
• You orient the first turtle according to this angle: turtle1.setheading(angle1).
• You advance the first turtle by 10 steps.
Improve your program by drawing a segment between the chasing turtle and the chased turtle each time.
PA R T I I

BASICS

17
 Chapter

3
If ... then ...

The computer can react according to a situation. If a condition is met, it acts in a certain way,
otherwise it does something else.

Lesson 1 (If ... then ...).

Here’s how to use the test “if” with Python:

keyword "if"

a condition

colon

if condition :
instruction_1 indented block
instruction_2 will be executed only
... if the condition is true
other instructions
program continuation

Here is an example, which warns a driver if a variable speed is too large.

if speed > 110:

print("Warning, you are driving too fast.")

Instructions can also be executed if the condition is not met using the keyword “else”.

if condition :
instruction this block will be executed
instruction if the condition is true
...
else:
instruction this block will be executed
...
if the condition is not true
other instructions

Once again, it is the indentation that delimits the different blocks of instructions. Here is an example that
displays the sign of a number x.
IF ... THEN ... 19

if x >= 0:
print("The number is positive (or zero).")
else:
print("The number is negative.")

Lesson 2 (Keyboard entry).

To be able to interact with the user, you can ask him to enter a text on the keyboard. Here is a small
program that asks for the user’s first name and age and displays a message like “Hello Kevin” then “You
are a minor/adult!” according to age.

first_name = input ("What's your name? ")

print("Hello",first_name)

age_str = input("How old are you? ")

age = int(age_str)

if age >= 18:

print("You're an adult!")
else:
print("You're a minor!")

Explanations.
• The command input() pauses the execution of the program and waits for the user to send a text
(ended by pressing the “Enter” key).
• This command returns a string.
• If you want an integer, you have to convert the string. For example, here age_str can be equal to
"17" (it is not a number but a sequence of characters), while int(age_str) is now the integer
17.
• The reverse operation is also possible, str() converts a number into a string. For example str(17)
returns the string "17"; if you set age = 17, then str(age) also returns "17".

Lesson 3 (The module “random”).

The random module generates numbers as if they were randomly drawn.
• Here is the command to place at the beginning of the program to call this module:
from random import *
• The command randint(a,b) returns a random integer between a and b.
For example after the instruction n = randint(1,6), n is a random integer with 1 6 n 6 6. If
you repeat the instruction n = randint(1,6), n takes a new value. It’s like rolling a 6-face dice.
• The command random(), without argument, returns a floating number (i.e. a decimal number)
between 0 and 1. For example, after the instruction x = random(), then x is a floating point
number with 0 6 x < 1.
IF ... THEN ... 20

Activity 1 (Multiplication quiz).

Goal: program a small test on the multiplication tables.

• Define a variable a, to which you assign a random value between 1 and 12.
• Same thing for a variable b.
• Display the question on the screen: “How much is the product a × b?”. (Replace a and b by their
value!)
• Retrieve the user’s answer and transform it into an integer.
• If the answer is correct, display “Well done!”, otherwise display “Lost! The correct answer was . . . ”.

Test of equality. To test if two numbers x and y are equal, the instruction is:
if x == y:
The equality test is written with the double symbol equal “==”. For example “x == 3” returns “True”
if x is equal to 3 and “False” otherwise.
Attention! The instruction “x = 3” has nothing to do with it, this instruction stores 3 in the variable
x.

Activity 2 (The words of the turtle).

Goal: guide the turtle with a word, each character corresponding to an instruction.
You give the Python turtle a word, for example FlFrfflFrfF, in which each character (read from left to
right) corresponds to an instruction that the turtle must execute.
• F : go forward of 100 by tracing,
• f : go forward of 100 without tracing,
• l : turn left by 90 degrees,
• r : turn right by 90 degrees.
Example. Here is the drawing that the turtle must trace when
word = "FflFlFrFlFFlfFF"

Hints. Here’s how to scan the letters of a word and test if a letter is the character F:

for c in word:
if c == "F":
instructions...
IF ... THEN ... 21

Lesson 4 (Booleans).
• A bolean is a data that is either equal to the value “True” or the value “False”. In Python the values
are True and False (with a capital letter).
• We obtain a boolean for example as a result of the comparison of two numbers. For example 7 < 4
is equal to False (because 7 is not smaller than 4). Check that print(7 < 4) displays False.
Here are the main comparisons:
– Test of equality: a == b
– Strict lower test: a < b
– Large lower test: a <= b
– Higher test: a > b a >= b
or
– Test of non equality: a != b
For example 6*7 == 42 is equal to True.
•

ATTENTION! The classic mistake is to confuse “a = b” and “a == b”.

– Assignment. a = b puts the content of the variable b in the variable a.
– Test of equality. a == b tests if the contents of a and b are the same, and is equal to
True or False.

• We can compare something other than numbers. For example, “char == "A"” tests if the variable
char is equal to "A"; “its_raining == True” tests if the variable its_raining is true. . .
• Booleans are useful in the test “if . . . then . . . ” and in the loops “while . . . then . . . ”.
• Operations between booleans. If P and Q are two booleans, new booleans can be defined.
– Logical and. “P and Q” is true if and only if P and Q are true.
– Logical or. “P or Q” is true if and only if P or Q is true.
– Negation. “not P” is true if and only if P is false.
For example “(2+2 == 2*2) and (5 < 3)” returns False, because even if we have 2+2 = 2×2,
the other condition is not satisfied because 5 < 3 is wrong.

Activity 3 (Digits of an integer).

Goal: find numbers whose digits verify certain properties.
1. The following program displays all integers from 0 to 99. Understand this program. What do the
variables u and t represent?

for t in range(10):
for u in range(10):
n = 10*t + u
print(n)
2. Find all integers between 0 and 999 that check all the following properties:
• the integer ends with 3,
• the sum of the digits is greater than or equal to 15,
• the tens digit is even.
3. Modify your previous program to count and display the number of integers checking these proper-
ties.
IF ... THEN ... 22

Activity 4 (Triangles).

Goal: determine the properties of a triangle from the three lengths of the sides.

We give ourselves three lengths a, b and c. You will determine the properties of the triangle whose
lengths would be a, b, c.

a
b

Define three variables a, b and c with integer values and a 6 b 6 c (or ask the user for three values).
1. Order. Ask Python to test if the lengths check a 6 b 6 c. Display a sentence for the answer.
2. Existence. There is a triangle corresponding to these lengths if and only if:
a + b > c.
Ask Python to test if this is the case and display the answer.
3. Rectangle triangle. Ask Python to test if the triangle is a rectangular triangle. (Think of Pythagoras’
theorem.)
4. Equilateral triangle. Test if the triangle is equilateral.
5. Isosceles triangle. Test if the triangle is isosceles.
6. Acute triangle. Tests if all angles are acute (i.e. less than or equal to 90 degrees).
Hints.
• The cosine law allows to calculate an angle according to the lengths:

a
b
γ
β

α c

−a2 + b2 + c 2 a2 − b2 + c 2 a2 + b2 − c 2
cos α = , cos β = , cos γ = .
2bc 2ac 2a b
• To test if the angle α is acute just check cos α > 0 (in the end we never calculate α, but only
cos α).

Find examples of lengths a, b, c to illustrate the different properties.

Activity 5 (The mystery number).

Goal: code the essential game when learning to program. The computer chooses a random number.
The user must guess this number by following the indications “larger” or “smaller” given by the
computer. As this game is quickly boring, we introduce variants where the computer is allowed to
lie or cheat!
1. The classic game.
IF ... THEN ... 23

• The computer randomly chooses a mystery number between 0 and 99.

• The player offers an answer.
• The computer replies “the number to find is greater” or “the number to find is smaller” or
“bravo, it’s the right number!”.
• The player has seven attempts to find the right answer.
Program this game!
Hints. To leave a loop for before the last proposal, you can use the command break. Use this
when the player finds the right answer.
2. The computer is lying.
To complicate the game, the computer has the right to lie from time to time. For example, about
one in four times the computer gives the wrong indication “larger” or “smaller”.
Hints. To decide when the computer is lying, each turn, draw a random number between 1 and 4,
if it is 4 the computer will lie!
3. The computer is cheating.
Now the computer is cheating (but it no longer lies)! Each turn the computer changes a little bit
the mystery number to find.
Hints. Each round, draw a random number, between −3 and +3 for example, and add it to the
mystery number. (Be careful not to exceed the 0 and 99 limits.)
 Chapter

4
Functions

Writing a function is the easiest way to group code for a particular task, in order to execute it once
or several times later.

Lesson 1 (Function (start)).

A computer function is a portion of code that performs a specific task and can be used one or more times
during the rest of the program. To define a function with Python, it’s very simple. Here are two examples:

def print_squares():
def say_hello():
for i in range(20):
print("Hello world!")
print(i**2)
return
return

The instructions are grouped into an indented block. The word return (optional) indicates the end of
the function. These instructions are executed only if I call the function. For example, each time I execute
the command say_hello(), Python displays the sentence “Hello world!”. Each time I execute the
command print_squares(), Python displays 0, 1, 4, 9, 16, . . ., i.e. the numbers i 2 for i = 0, . . . , 19.

keyword "def"
well-chosen name
parentheses
colon
def my_function () :
instruction_1
instruction_2 instructions
...
return
end of the function

Lesson 2 (Function (continued)).

Functions achieve their full potential with:
• an input, which groups variables that serve as parameters,
• an output, which is a result returned by the function (and which will often depend on the input
parameters).
Here are two examples:
FUNCTIONS 25

def display_month(number):
if number == 1:
print("We are in January.")
if number == 2:
print("We are in February.")
if number == 3:
print("We are in March.")
# etc.
return

When called this function displays the name of the month based on the number provided as input. For
example display_month(3) will display "We are in March.".

def compute_cube(a):
cube = a * a * a # or a**3
return cube

This function calculates the cube of a number, for example compute_cube(2) does not display anything
but returns the value 8. This value can be used elsewhere in the program. For example, what do the
following instructions do?

x = 3
y = 4
z = compute_cube(x) + compute_cube(y)
print(z)

In mathematical terms, we put x = 3, y = 4, then we calculate the cube of x, the cube of y and add them
up:
z = x 3 + y 3 = 33 + 43 = 27 + 64 = 91
Thus the program displays 91.

# Definition of the function

def my_function (param) :
instruction_1
instruction_2 a parameter
...
return result
returns the result

# Call of the function

x = 7
val = my_function (x)
argument

call of the function

returned result

The advantages of programming using functions are as follows:

FUNCTIONS 26

• you write the code of a function only once, but you can call the function several times;
• by dividing our program into small blocks, each with its own use, the program is easier to write,
read, correct and modify;
• you can use a function written by someone else (such as the sqrt() function) without knowing
all the internal details of its programming.

Activity 1 (First functions).

Goal: write very simple functions.

1. Function without parameter or output.

(a) Program a function called print_table_of_7() that displays the multiplication table by
7: 1 × 7 = 7, 2 × 7 = 14. . .
(b) Program a function called print_hello() that asks the user for his first name and then
displays “Hello” followed by his name.
Hint. Use input().
2. Function with one parameter and no output.
(a) Program a function called print_a_table(n) that depends on a parameter n and displays
the multiplication table by this integer n. For example, the command print_a_table(5)
must display: 1 × 5 = 5, 2 × 5 = 10. . .
(b) Program a function called say_greeting(sentence) that depends on a parameter
sentence. This function asks for the user’s first name and displays a greeting followed
by the first name. For example, say_greeting("Hi") would display “Hi” followed by the
first name given by the user.
3. Function without parameter and with output.
Program a function called ask_full_name() that first asks for the user’s first name, then his last
name and returns the complete identity with the last name in upper case as a result. For example, if
the user enters “Dark” then “Vador”, the function returns the string "Dark VADOR" (the function
displays nothing).
Hints.
• If string is a string, then string.upper() is the transformed string with characters in
capital letters. Example: if string = "Vador" then string.upper() returns "VADOR".
• You can merge two strings by using the operator “+”. Example: "Dark" + "Vador" is equal
to "DarkVador". Another example: if string1 = "Dark" and string2 = "Vador" then
string1 + " " + string2 is equal to "Dark Vador".

Lesson 3 (Function (continuation and end for now)).

A function can have several parameters and can return several results. For example, here is a function
that calculates and returns the sum and product of two given input numbers.

def sum_product(a,b):
""" Computes the sum and product of two numbers. """
s = a + b
p = a * b
return s, p
FUNCTIONS 27

mysum, myprod = sum_product(6,7)

The last line calls the function with arguments 6 (for parameter a) and 7 (for parameter b). This function
returns two values, the first one is assigned to mysum (which is therefore equal to 13) and the second
one to myprod (which is equal to 42).

one or several parameters

def my_function (param1,param2) :

""" My function does this and that. """
if param1 == param2: immediate termination of the
return 0,0,0
function in this case
other instructions...
return value1, value2, value3

returns several values

function documentation

So let’s remember:
• There can be several input parameters.
• There can be several results at the output.
• Very important! Do not confuse displaying and returning a value. The display (by the command
print()) just displays something on the screen. Most functions do not display anything, but return
one (or more) value. This is much more useful because this value can be used elsewhere in the
program.
• As soon as the program encounters the instruction return, the function stops and returns the result.
There may be several times the instruction return in a function but only one will be executed. It
is also possible not to put an instruction return if the function returns nothing.
• In the instructions of a function, you can of course use other functions!
• It is important to comment well on your programs. To document a function, you can describe what
it does starting with a docstring, i.e. a description (in English) surrounded by three quotation marks:
""" My function does this and that. """
to be placed just after the header.
• When defining a function, the variables that appear between the parentheses are called the pa-
rameters; when calling the function, however, the values between the parentheses are called the
arguments. There is of course a correspondence between the two.

Activity 2 (More features).

Goal: build functions with different types of input and output.
1. Trinomials.
(a) Write a function trinomial_1(x) that depends on a parameter x and returns the value of
the trinomial 3x 2 − 7x + 4. For example, trinomial_1(7) returns 102.
FUNCTIONS 28

(b) Write a function trinomial_2(a,b,c,x) that depends on four parameters a, b, c and x

and returns the value of the trinomial a x 2 + b x + c. For example, trinomial_2(2,-1,0,6)
returns 66.
2. Currencies.
(a) Write a function conversion_dollars_to_euros(amount) which depends on a parame-
ter and which for a sum of money amount, expressed in dollars, returns its value in euros
(you will take for example 1 dollar = 0.89 euro).
(b) Write a function conversion_dollars(amount,currency) which depends on two pa-
rameters amount and currency and converts the amount given in dollars, to the desired
currency. Examples of currencies: 1 dollar = 0.89 euros; 1 dollar = 0.77 pounds; 1 dollar =
110 yen. For example, conversion_dollars(100,"pound") returns 77.
Take care to give an intelligible name to your functions and variables. Don’t forget to document
each function by adding a small explanatory text between triple quotation marks at the very
beginning of your function.
3. Volumes.
Build functions that calculate and return volumes:
• the volume of a cube according to the length of one side,
• the volume of a ball according to its radius,
• the volume of a cylinder according to the radius of its base and its height,
• the volume of a rectangular parallelepiped box according to its three dimensions.
For the value of π, you will take either the approximate value 3.14, or the approximate value
provided by the constant pi of the module math.
4. Perimeters and areas.
(a) Write a function whose use is perimeter_area_rectangle(a,b) and which returns the
perimeter and area of a rectangle with dimensions of a and b.
(b) Same question with perimeter_area_disc(r) for the perimeter and area of a disc of
radius r.
(c) Use your previous function to guess from which radius, the area of a disc is larger than the
perimeter of that disc.
Hint. If you want to scan the radius by incrementing the value of 0.1 each time, you can build
a loop as follows:
for R in range(0,30):
then make a call to the function by perimeter_area_disc(R/10).

Activity 3 (Turtle).
Goal: define some functions that draw geometric shapes. Creating a function is similar to creating
a block with Scratch.

1. Program a function triangle() that draws a triangle (in red, each side measuring 200).
FUNCTIONS 29

2. Program a function square() that draws a square (in green, each side measuring 200). Use a
loop “for” so you don’t have to rewrite the same instructions several times.
3. Program a hexagon(length) function that draws a hexagon (in blue) of a given side length (the
angle to turn is 60 degrees).
4. Program a function polygon(n,length) that draws a regular polygon of n sides and a given
side length (the angle to rotate is 360/n degrees).

Activity 4 (Functions again).

Goal: create new functions.

1. (a) Here is the discount for the price of a train ticket based on the age of the passenger:
• reduction of 50% for those under 10 years old;
• reduction of 30% for 10 to 18 years old;
• reduction of 20% for 60 years old and over.
Write a function reduction() that returns the reduction according to age and whose
properties are recalled in the box below:
reduction()
Use: reduction(age)
Input: an integer corresponding to age
Output: an integer corresponding to the reduction
Examples:
• reduction(17) returns 30.
• reduction(23) returns 0.

Deduce a amount() function that calculates the amount to be paid based on the normal fare
and the traveler’s age.
amount()
Use: amount(normal_rate,age)
Input: a number normal_rate corresponding to the price without discount
and age (an integer)
Output: a number corresponding to the amount to be paid after reduction
Note: uses the function reduction()
Example: amount(100,17) returns 70.

A family buys tickets for different trips, here is the normal fare for each trip and the ages of
the passengers:
• normal price 30 dollars, child of 9 years old;
• normal price 20 dollars, for each of the twins of 16 years old;
• normal price 35 dollars, for each parent of 40 years old.
What is the total amount paid by the family?
2. We want to program a small quiz on the multiplication tables.
(a) Program a function is_calculation_correct() that decides if the answer given to a
multiplication is right or not.
FUNCTIONS 30

is_calculation_correct()
Use: is_calculation_correct(a,b,answer)
Input: three integers, answer being the proposed answer to the calculation
of a × b.
Output: “True” or “False”, depending on whether the answer is correct or not
Examples:
• is_calculation_correct(6,7,35) returns False.
• is_calculation_correct(6,7,42) returns True.

(b) Program a function that displays a multiplication, asks for an answer and displays a short
concluding sentence. All this in English or a foreign language!
test_multiplication()
Use: test_multiplication(a,b,lang)
Input: two integers, the chosen language (for example "english" or
"french")
Output: nothing (but display a sentence)
Note: uses the function is_calculation_correct()
Example: test_multiplication(6,7,"french") asks, in French, for the
answer to the calculation 6 × 7 and answers if it is correct or not.

Bonus. Improve your program so that the computer offers random operations to the player. (Use
the randint() function of the module random.)

Activity 5 (Experimental equality).

Goal: use the computer to experiment with equality of functions.

1. (a) Build a function absolute_value(x) that returns the absolute value of a number (without
using the function abs() of Python).
p
(b) Build a function root_of_square(x) which corresponds to the calculation of x 2 .
(c) Two functions (of one variable) f and g are said to be experimentally equal if f (i) = g(i)
for i = −100, −99, . . . , 0, 1, 2, . . . , 100. Check by computer that the two functions defined by
p
|x| and x2
are experimentally equal.
2. (a) Build a two-parameter function F1(a,b) that returns (a + b)2 . Same thing with F2(a,b)
that returns a2 + 2a b + b2 .
(b) Two functions of two variables F and G are said to be experimentally equal if F (i, j) = G(i, j)
for all i = −100, −99, . . . , 100 and for all j = −100, −99, . . . , 100. Check by computer that
the functions defined by (a + b)2 and a2 + 2a b + b2 are experimentally equal.
(c) I know that one of the following two identities is true:
(a − b)3 = a3 − 3a2 b − 3a b2 + b3 or (a − b)3 = a3 − 3a2 b + 3a b2 − b3 .
Help yourself from the computer to decide which one it is!
3. (a) Build a function sincos(x) that returns (sin(x))2 + (cos(x))2 and another one(x) that
always returns 1. Are these two functions experimentally equal (in the sense of the first
question)? Find out what could be the cause of this answer.
FUNCTIONS 31

(b) Fix ε = 0.00001. It is said that two functions (of one variable) f and g are experimentally
approximately equal if | f (i) − g(i)| 6 ε for i = −100, −99, . . . , 100. Do the two functions
defined by sincos(x) and one(x) now check this criterion?
(c) Experimentally check and experimentally approximately check the identities:
π 
sin(2x) = 2 sin(x) cos(x), cos − x = sin(x).
2
(d) Bonus. A counter-example. Show that the functions defined by g1 (x) = sin(πx) and g2 (x) =
0 are experimentally equal (with our definition given above). But also show that you don’t
get g1 (x) = g2 (x) for every x ∈ R.

Lesson 4 (Local variable).

Here is a very simple function that takes a number as an input and returns the number increased by one.

def my_function(x):
x = x + 1
return x

• Of course my_function(3) returns 4.

• If I define a variable by y = 5 then my_function(y) returns 6. And the value of y has not
changed, it is still equal to 5.
• Here is the delicate situation that you must understand:

x = 7
print(my_function(x))
print(x)

– The variable x is initialized to 7.

– The call of the function my_function(x) is therefore the same as my_function(7) and
logically returns 8.
– What is the value of x at the end? The variable x is unchanged and is still equal to 7! Even if
in the meantime there has been an instruction x = x + 1. This instruction changed the x
inside the function, but not the x outside the function.

• Variables defined within a function are called local variables. They do not exist outside the function.
• If there is a variable in a function that has the same name as a variable in the program (like the x
in the example above), it is as if there were two distinct variables; the local variable only exists
inside the function.
To understand the scope of the variables, you can color the global variables of a function in red, and the
local variables with one color per function. The following small program defines a function that adds one
and another that calculates the double.
FUNCTIONS 32

global variable
x = 7
def add_one( x ):
x = x +1
local variable
return x

def double( x ):
x = 2 * x
another local variable
return x

print( x )
print(add_one( x ))
print(double( x ))
print( x )

The program first displays the value of x, so 7, then it increases it by 1, so it displays 8, then it displays
twice as much as x, so 14. The global variable x has never changed, so the last display of x is still 7.

It is still possible to force the hand to Python and modify a global variable in a function using the keyword
global. See the chapter “Polish calculator – Stacks”.
 Chapter

Arithmetic – While loop – I

5

The activities in this sheet focus on arithmetic: Euclidean division, prime numbers . . . This is an
opportunity to use the loop “while” intensively.

Lesson 1 (Arithmetic).
We recall what Euclidean division is. Here is the division of a by b, a is a positive integer, b is a strictly
positive integer (with an example of 100 divided by 7) :

quotient

q 14

b a 7 100
remainder
r 2

We have the two fundamental properties that define q and r:

a = b×q+r and 06r < b
For example, for the division of a = 100 by b = 7: we have the quotient q = 14 and the remainder r = 2
that verify a = b × q + r because 100 = 7 × 14 + 2 and also r < b because 2 < 7.

With Python :
• a // b returns the quotient,
• a % b returns the remainder.

It is easy to check that:

b is a divisor of a if and only if r = 0.

Activity 1 (Quotient, remainder, divisibility).

Goal: use the remainder to find out if an integer divides another.

1. Program a function quotient_remainder(a,b) that does the following tasks for two integers
a > 0 and b > 0 :
• It displays the quotient q of the Euclidean division of a per b,
• it displays the remainder r of this division,
ARITHMETIC – WHILE LOOP – I 34

• it displays True if the remainder r is positive or zero and strictly less than b, and False
otherwise,
• it displays True if you have equality a = bq + r, and False if not.
Here is for example what the call should display quotient_remainder(100,7):
Division of a = 100 per b = 7
The quotient is q = 14
The remainder is r = 2
Check remainder: 0 <= r < b? True
Check equality: a = bq + r? True
Note. You have to check without cheating that we have 0 6 r < b and a = bq + r, but of course it
must always be true!
2. Program a function is_even(n) that tests if the integer n is even or not. The function returns
True or False.
Hints.
• First possibility: calculate n % 2 and discuss as appropriate.
• Second possibility: calculate n % 10 (which returns the digit of units) and discuss.
• The smartest people will be able to write the function with only two lines (one for def...
and the other for return...).
3. Program a function is_divisible(a,b) that tests if b divides a. The function returns True or
False.

Lesson 2 (Loop “while”).

The loop “while” executes instructions as long as a condition is true. As soon as the condition becomes
false, it proceeds to the next instructions.

keyword "while"

a condition

colon

while condition :
instruction_1 indented block will
instruction_2 be executed as long as
... the condition is true
other instructions
program continuation
ARITHMETIC – WHILE LOOP – I 35

Example.
Here is a program that displays the countdown
n = 10
10, 9, 8, . . . 3, 2, 1, 0. As long as the condition n > 0 is
while n >= 0:
true, we reduce n by 1. The last value displayed is n = 0,
print(n)
because then n = −1 and the condition “n > 0” becomes
n = n - 1
false so the loop stops.
This is summarized in the form of a table:
Input: n = 10
n “n > 0” ? new value of n
10 yes 9
9 yes 8
... ... ...
1 yes 0
0 yes −1
−1 no
Display: 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0

Example.
n = 100
This piece of code looks for the first power of 2 greater
p = 1
than a given integer n. The loop prints the values 2, 4,
while p < n:
8, 16,. . . It stops as soon as the power of 2 is higher or
p = 2 * p
equal to n, so here this program displays 128.
print(p)
Inputs: n = 100, p = 1
p “p < n” ? new value of p
1 yes 2
2 yes 4
4 yes 8
8 yes 16
16 yes 32
32 yes 64
64 yes 128
128 no
Display: 128
ARITHMETIC – WHILE LOOP – I 36

Example.
For this last loop we have already prepared a function
is_even(n) which returns True if the integer n is even n = 56
and False otherwise. The loop does this: as long as while is_even(n) == True:
the integer n is even, n becomes n/2. This amounts to n = n // 2
removing all factors 2 from the integer n. As here n = print(n)
56 = 2 × 2 × 2 × 7, this program displays 7.
Input: n = 56
n “is n even” ? new value of n
56 yes 28
28 yes 14
14 yes 7
7 no
Display: 7

For the latter example, it is much more natural to start the loop with
while is_even(n):
Indeed is_even(n) is already a value “True” or “False”. We’re getting closer to the English sentence
“while n is even...”

Operation “+=”. To increment a number you can use these two methods:
nb = nb + 1 or nb += 1
The second writing is shorter but makes the program less readable.

Activity 2 (Prime numbers).

Goal: test if an integer is (or not) a prime number.
1. Smallest divisor.
Program a function smallest_divisor(n) that returns, the smallest divisor d > 2 of the integer
n > 2.
For example smallest_divisor(91) returns 7, because 91 = 7 × 13.
Method.
• We remind you that d divides n if and only if n % d is equal to 0.
• The bad idea is to use a loop “for d ranging from 2 to n”. Indeed, if for example we know that
7 is a divisor of 91 it is useless to test if 8, 9, 10 . . . are also divisors because we have already
found a smaller one.
• The good idea is to use a loop “while”! The principle is: “as long as I haven’t got my divisor, I
keep looking for”. (And so, as soon as I find it, I stop looking.)
• In practice here are the main lines:
– Begin with d = 2.
– As long as d does not divide n then pass to the next candidate (d becomes d + 1).
– At the end d is the smallest divisor of n (in the worst case d = n).
2. Prime numbers (1).
Slightly modify your function smallest_divisor(n) to write a first function is_prime_1(n)
which returns “True” if n is a prime number and “False” otherwise.
ARITHMETIC – WHILE LOOP – I 37

For example is_prime_1(13) returns True, is_prime_1(14) returns False.

3. Fermat numbers.
Pierre de Fermat (∼1605–1665) thought that all integers of the form Fn = 2(2 ) + 1 were prime
n

numbers. Indeed F0 = 3, F1 = 5 and F2 = 17 are prime numbers. If he had known Python he

would probably have changed his mind! Find the smallest integer Fn which is not prime.
Hint. With Python b c is written b ** c and therefore a(b ) is written a ** (b ** c).
c

We will improve our function which tests if a number is prime or not, it will allow us to test more
quickly lots of numbers or very large numbers.
4. Prime numbers (2).
Enhance your previous function to a is_prime_2(n) that does not test all divisors d from 2 to n,
p
but only up to n.
Explanations.
• For example, to test if 101 is a prime number, just see if it admits divisors among 2, 3, . . . , 10.
The gain is significant!
• This improvement is due to the following proposal: if an integer is not prime then it admits a
p
divisor d that verifies 2 6 d 6 n.
p
• Instead of testing if d 6 n, it is easier to test if d 2 6 n.
5. Prime numbers (3).
Improve your function into a function is_prime_3(n) using the following idea. We test if d = 2
divides n, but from d = 3, we just test the odd divisors (we test d, then d + 2. . . ).
• For example to test if n = 419 is a prime number, we first test if d = 2 divides n, then d = 3
and then d = 5, d = 7. . .
• This allows you to do about half less tests!
• Explanations: if an even number d divides n, then we already know that 2 divides n.
6. Calculation time.
Compare the calculation times of your different functions is_prime() by repeating for example
a million times the call is_prime(97). See the course below for more information on how to do
this.

Lesson 3 (Calculation time).

There are two ways to run programs faster: a good and a bad one. The bad way is to buy a more powerful
computer. The good method is to find a more efficient algorithm!
With Python, it is easy to measure the execution time of a function in order to compare it with the
execution time of another. Just use the module timeit.
Here is an example: we measure the computation time of two functions that have the same purpose, test
if an integer n is divisible by 7.
# First function (not very clever)
def my_function_1(n):
divis = False
for k in range(n):
if k*7 == n:
divis = True
return divis
ARITHMETIC – WHILE LOOP – I 38

# Second function (faster)

def my_function_2(n):
if n % 7 == 0:
return True
else:
return False

# Measurement of execution times

import timeit

print(timeit.timeit("my_function_1(1000)",
setup="from __main__ import my_function_1",
number=100000))
print(timeit.timeit("my_function_2(1000)",
setup="from __main__ import my_function_2",
number=100000))
Results.
The result depends on the computer, but allows the comparison of the execution times of the two functions.
• The measurement for the first function (called 100 000 times) returns 5 seconds. The algorithm is
not very clever. We’re testing if 7 × 1 = n, then test 7 × 2 = n, 7 × 3 = n. . .
• The measurement for the second function returns 0.01 second! We test if the remainder of n divided
by 7 is 0. The second method is therefore 500 times faster than the first.
Explanations.
• The module is named timeit.
• The function timeit.timeit() returns the execution time in seconds. It takes as parameters:
– a string for the call of the function to be tested (here we ask if 1000 is divisible by 7),
– an argument setup="..." which indicates where to find this function,
– the number of times you have to repeat the call to the function (here number=100000).
• The number of repetitions must be large enough to avoid uncertainties.

Activity 3 (More prime numbers).

Goal: program more loops “while” and study different kinds of prime numbers using your function
is_prime().
1. Write a function prime_after(n) that returns the first prime number p greater than or equal to
n.
For example, the first prime number after n = 60 is p = 61. What is the first prime number after
n = 100 000?
2. Two prime numbers p and p + 2 are called twin prime numbers. Write a function
twin_prime_after(n) that returns the first pair p, p + 2 of twin prime numbers, with p > n.
For example, the first pair of twin primes after n = 60 is p = 71 and p + 2 = 73. What is the first
pair of twin primes after n = 100 000?
3. An integer p is a Germain prime number if p and 2p + 1 are prime numbers. Write a function
germain_after(n) that returns the pair p, 2p + 1 where p is the first Germain prime number
p > n.
ARITHMETIC – WHILE LOOP – I 39

For example, the first Germain prime number after n = 60 is p = 83, with 2p + 1 = 167. What is
the first Germain prime number after n = 100 000?
 Chapter

Strings – Analysis of a text

6

You’re going to do some fun activities by manipulating strings and characters.

Lesson 1 (Characters and strings).

1. A character is a unique symbol, for example a lowercase letter "a", a capital letter "B", a special
symbol "&", a symbol representing a number "7", a space " " that we will also note " ".
To designate a character, it must be put in single quotation marks 'z' or double quotation marks
"z".
2. A string is a sequence of characters, such as a word "Hello", a sentence 'It is sunny.', a
password "N[w5ms}e!".
3. The type of a character or string is str.

Lesson 2 (Operations on strings).

1. The concatenation, i.e. the end-to-end placing of two strings, is done using the operator +. For
example "umbr" + "ella" gives the string "umbrella".
2. The empty string "" is useful when you want to initialize a string before adding other characters.
3. The length of a string is the number of characters it contains. It is obtained by the function len().
For example len("Hello World") returns 11 (a space counts as a character).
4. If word is a string then you can retrieve each character by the command word[i]. For example, if
word = "plane" then:
• word[0] is the character "p",
• word[1] is the character "l",
• word[2] is the character "a",
• word[3] is the character "n",
• word[4] is the character "e".

Letter p l a n e
Rank 0 1 2 3 4

Note that there are 5 letters in the word "plane" and that you access it through the ranks starting with
0. The indices are therefore here 0, 1, 2, 3, and 4 for the last letter. More generally, if word is a string,
characters are obtained by word[i] for i varying from 0 to len(word)-1.

Lesson 3 (Substrings).
You can extract several characters from a string using the syntax word[i:j] which returns a string
STRINGS – ANALYSIS OF A TEXT 41

formed by characters ranked i to j − 1 (beware the character number j is not included).

For example if word = "wednesday" then:
• word[0:4] returns the "wedn" substring formed by the characters of ranks 0, 1, 2 and 3 (but not
4),
• word[3:6] returns "nes" corresponding to indices 3, 4 and 5.

Letter w e d n e s d a y
Rank 0 1 2 3 4 5 6 7 8

Another example: word[1:len(word)-1] returns the word but with its first and last letter cut off.

Activity 1 (Plurals of words).

Goal: write step by step a program that returns the plural of a given word.

1. For a string word, for example "cat", displays the plural of this word by adding a "s".
2. For a word, for example "bus", display the last letter of this string (here "s"). Improve your
program of the first question, by testing if the last letter is already a "s":
• if this is the case, then add "es" to form the plural ("bus" becomes "buses"),
• otherwise you have to add "s".
3. Test if a word ends with "y". If so, display the plural in "ies" (the plural of "city" is "cities").
(Exceptions are not taken into account.)
4. Collect all your work from the first three questions in a function plural(). The function displays
nothing, but returns the word in the plural.

plural()
Use: plural(word)
Input: a word (a string)
Output: the plural of the word

Examples:
• plural("cat") returns "cats"
• plural("bus") returns "buses"
• plural("city") returns "cities"

5. Write a function conjugation() that conjugates a verb to the present continuous tense.

conjugation()
Use: conjugation(verb)
Input: a verb (a string, exceptions are not taken into account)
Output: no result but displays conjugation in the present continuous tense

Example: conjugation("sing"), prints "I am singing, you are

singing,..."
STRINGS – ANALYSIS OF A TEXT 42

Lesson 4 (A little more on strings).

1. A loop for ... in ... allows you to browse a string, character by character:
for charac in word:
print(charac)
2. You can test if a character belongs to a certain list of characters. For example:
if charac in ["a", "A", "b", "B", "c", "C"]:
allows you to execute instructions if the character charac is one of the letters a, A, b, B, c, C.
To avoid some letters, we would use:
if charac not in ["X", "Y", "Z"]:

Activity 2 (Word games).

Goal: manipulating words in a fun way.

1. Distance between two words.

The Hamming distance between two words of the same length is the number of places where the
letters are different.
For example:
SNAKE STACK
The second letter of SNAKE is different from the second letter of STACK, the fourth and fifth ones
are also different. The Hamming distance between SNAKE and STACK is therefore equal to 3.
Write a function hamming_distance() that calculates the Hamming distance between two words
of the same length.

hamming_distance()
Use: hamming_distance(word1,word2)
Input: two words (strings)
Output: the Hamming distance (an integer)
Example: hamming_distance("SHORT","SKIRT") returns 2

2. Upside down.
Write a function upsidedown() that returns a word backwards: HELLO becomes OLLEH.

upsidedown()
Use: upsidedown(word)
Input: a word (a string)
Output: the word backwards
Example: upsidedown("PYTHON") returns "NOHTYP"

3. Palindrome.
Deduce a function that tests whether a word is a palindrome or not. A palindrome is a word that
can be written from left to right or right to left; for example RADAR is a palindrome.
STRINGS – ANALYSIS OF A TEXT 43

is_palindrome()
Use: is_palindrome(word)
Input: a word (a string)
Output: “True” if the word is a palindrome, “False” otherwise.
Example: is_palindrome("KAYAK") returns True

4. Pig latin. Pig latin is word play, here are the rules according to Wikipedia:
• For words that begin with vowel sounds, one just adds "way" to the end. Examples are:
– EAT becomes EATWAY
– OMELET becomes OMELETWAY
– EGG becomes EGGWAY
• For words that begin with consonant sounds, all letters before the initial vowel are placed at
the end of the word sequence. Then, "ay" is added, as in the following examples:
– PIG becomes IGPAY
– LATIN becomes ATINLAY
– BANANA becomes ANANABAY
– STUPID becomes UPIDSTAY
– GLOVE becomes OVEGLAY
Write a function pig_latin() that transforms a word according to this procedure.

pig_latin()
Use: pig_latin(word)
Input: a word (a string)
Output: the word transformed into pig latin.
Examples:
• pig_latin("DUCK") returns "UCKDWAY"
• pig_latin("ALWAYS") returns "ALWAYSWAY"

Activity 3 (DNA).
A DNA molecule is made up of about six billion nucleotides. The computer is therefore an essential
tool for DNA analysis. In a DNA strand there are only four types of nucleotides that are noted A,
C, T or G. A DNA sequence is therefore a long word in the form: TAATTACAGACACCTGAA...
1. Write a function presence_of_A() that tests if a sequence contains the nucleotide A.
presence_of_A()
Use: presence_of_A(sequence)
Input: a DNA sequence (a string whose characters are among A, C, T, G)
Output: “True” if the sequence contains “A”, “False” otherwise.
Example: presence_of_A("CTTGCT") returns False

2. Write a function position_of_AT() that tests if a sequence contains the nucleotide A followed
by the nucleotide T and returns the position of the first occurrence found.
STRINGS – ANALYSIS OF A TEXT 44

position_of_AT()
Use: position_of_AT(sequence)
Input: a DNA sequence (a string whose characters are among A, C, T, G)
Output: the position of the first sequence “AT” found (starting at 0); None if
not found.
Example:
• position_of_AT("CTTATGCT") returns 3
• position_of_AT("GATATAT") returns 1
• position_of_AT("GACCGTA") returns None

Hint. None is assigned to a variable to indicate the absence of value.

3. Write a function position() that tests if a sequence contains a given code and returns the position
of the first occurrence.
position()
Use: position(code,sequence)
Input: a code and a DNA sequence
Output: the position of the beginning of the code found; None if not found
Example: position("CCG","CTCCGTT") returns 2

4. A crime has been committed in the castle of Adeno. You recovered two strands of DNA, from two
distant positions from the culprit’s DNA. There are four suspects, whose DNA you sequenced. Can
you find out who did it?
First code of the culprit: CATA
Second code of the culprit: ATGC
DNA of Colonel Mustard:
CCTGGAGGGTGGCCCCACCGGCCGAGACAGCGAGCATATGCAGGAAGCGGCAGGAATAAGGAAAAGCAGC
Miss Scarlet’s DNA:
CTCCTGATGCTCCTCGCTTGGTGGTTTGAGTGGACCTCCCAGGCCAGTGCCGGGCCCCTCATAGGAGAGG
Mrs. Peacock’s DNA:
AAGCTCGGGAGGTGGCCAGGCGGCAGGAAGGCGCACCCCCCCAGTACTCCGCGCGCCGGGACAGAATGCC
Pr. Plum’s DNA:
CTGCAGGAACTTCTTCTGGAAGTACTTCTCCTCCTGCAAATAAAACCTCACCCATGAATGCTCACGCAAG

Lesson 5 (Character encoding).

A character is stored by the computer as an integer. For ASCII/unicode encoding, the capital letter “A” is
encoded by 65, the lowercase letter “h” is encoded by 104, the symbol “#” by 35.
Here is the table of the first characters. Numbers 0 to 32 are not printable characters. However, the
number 32 is the space character " ".
STRINGS – ANALYSIS OF A TEXT 45

33 ! 43 + 53 5 63 ? 73 I 83 S 93 ] 103 g 113 q 123 {

34 " 44 , 54 6 64 @ 74 J 84 T 94 ^ 104 h 114 r 124 |

35 # 45 - 55 7 65 A 75 K 85 U 95 _ 105 i 115 s 125 }

36 $ 46 . 56 8 66 B 76 L 86 V 96 ‘ 106 j 116 t 126 ~

37 % 47 / 57 9 67 C 77 M 87 W 97 a 107 k 117 u 127 -

38 & 48 0 58 : 68 D 78 N 88 X 98 b 108 l 118 v

39 ’ 49 1 59 ; 69 E 79 O 89 Y 99 c 109 m 119 w

40 ( 50 2 60 < 70 F 80 P 90 Z 100 d 110 n 120 x

41 ) 51 3 61 = 71 G 81 Q 91 [ 101 e 111 o 121 y

42 * 52 4 62 > 72 H 82 R 92 \ 102 f 112 p 122 z

1. The function chr() is a Python function that returns the character associated to a code.

python : chr()
Use: chr(code)
Input: a code (an integer)
Output: a character
Example:
• chr(65) returns "A"
• chr(123) returns "{"

2. The function ord() is a Python function corresponding to the reverse operation: it returns the
code associated with a character.
python : ord()
Use: ord(charac)
Input: a character (a string of length 1)
Output: an integer
Example:
• ord("A") returns 65
• ord("*") returns 42

Activity 4 (Upper case/lower case).

Goal: convert a word to upper or lower case.

1. Decodes by hand the encrypted message hidden under the following codes:
80-121-116-104-111-110 105-115 99-64-64-108
2. Write a loop that displays characters encoded by integers from 33 to 127.
3. What does the command chr(ord("a")-32) return? And chr(ord("B")+32)?
STRINGS – ANALYSIS OF A TEXT 46

4. Write a function upper_letter() that transforms a lowercase letter into its uppercase letter.

upper_letter()
Use: upper_letter(charac)
Input: a lowercase character among "a",...,"z"
Output: the same letter in upper case
Example: upper_letter("t") returns "T"

5. Write a function uppercase() which from a sentence written in lower case returns the same
sentence written in upper case. Characters that are not lowercase letters remain unchanged.

uppercase()
Use: uppercase(sentence)
Input: a sentence
Output: the same sentence in upper case
Example: uppercase("Hello world!") returns "HELLO WORLD!"

Do the same work for a function lowercase().

6. Write a function format_full_name() that returns the first name and last name formatted
according to the style First_name LAST_NAME.

format_full_name()
Use: format_full_name(somebody)
Input: a person’s first name and surname (without accent, separated by a
space)
Output: the full name following the format "First name LAST NAME"
Example:
• format_full_name("harry Potter") returns "Harry POTTER"
• format_full_name("LORD Voldemort") returns "Lord
VOLDEMORT"

Activity 5.

Goal: determine the language of a text from the analysis of letter frequencies.

1. Write a function occurrences_letter() that counts the number of times the given letter appears
in a sentence (in upper case).
STRINGS – ANALYSIS OF A TEXT 47

occurrences_letter()
Use: occurrences_letter(letter,sentence)
Input: a letter (a character) and a sentence in capital letters (a string)
Output: the number of occurrences of the letter (an integer)
Example: occurrences_letter("E","IS THERE ANYBODY OUT
THERE") returns 4

2. Write a function number_letters() that counts the total number of letters that appear in a
sentence (in upper case). Do not count spaces or punctuation.

number_letters()
Use: number_letters(sentence)
Input: a sentence in capital letters (a string)
Output: the total number of letters from “A” to “Z” (an integer)
Example: number_letters("IS THERE ANYBODY OUT THERE") re-
turns 22.

3. The frequency of appearance of a letter in a text or sentence is the percentage given according to
the formula :
number of occurrences of the letter
frequency of appearance of a letter = × 100.
total number of letters
For example, the sentence IS THERE ANYBODY OUT THERE contains 22 letters; the letter E
appears there 4 times. The frequency of appearance of E in this sentence is therefore:
number of occurrences of E 4
fE = × 100 = × 100 ' 16.66
total number of letters 22
The frequency is therefore about 17%.
Write a function percentage_letter() that calculates this frequency of appearance.

percentage_letter()
Use: percentage_letter(letter,sentence)
Input: a letter (a character) and a sentence in capital letters (a string)
Output: the frequency of appearance of the letter (a number lower than 100)
Example: percentage_letter("E","IS THERE ANYBODY OUT
THERE") returns 16.66. . .

Use this function to properly display the frequency of appearance of all letters in a sentence.
4. Here is the frequency of appearance of letters according to the language used (source:
en.wikipedia.org/wiki/Letter_frequency). For example, the most common letter in English is
“E” with a frequency of more than 12%. The “W” represents about 2% of letters in English and
German, but almost does not appear in French and Spanish. These frequencies also vary according
to the text analyzed.
STRINGS – ANALYSIS OF A TEXT 48

Letter English French German Spanish

a 8.167% 8.173% 7.094% 12.027%
b 1.492% 0.901% 1.886% 2.215%
c 2.782% 3.345% 2.732% 4.019%
d 4.253% 3.669% 5.076% 5.010%
e 12.702% 16.734% 16.396% 12.614%
f 2.228% 1.066% 1.656% 0.692%
g 2.015% 0.866% 3.009% 1.768%
h 6.094% 0.737% 4.577% 0.703%
i 6.966% 7.579% 6.550% 6.972%
j 0.153% 0.613% 0.268% 0.493%
k 0.772% 0.049% 1.417% 0.011%
l 4.025% 5.456% 3.437% 4.967%
m 2.406% 2.968% 2.534% 3.157%
n 6.749% 7.095% 9.776% 7.023%
o 7.507% 5.819% 3.037% 9.510%
p 1.929% 2.521% 0.670% 2.510%
q 0.095% 1.362% 0.018% 0.877%
r 5.987% 6.693% 7.003% 6.871%
s 6.327% 7.948% 7.577% 7.977%
t 9.056% 7.244% 6.154% 4.632%
u 2.758% 6.429% 5.161% 3.107%
v 0.978% 1.838% 0.846% 1.138%
w 2.360% 0.074% 1.921% 0.017%
x 0.150% 0.427% 0.034% 0.215%
y 1.974% 0.128% 0.039% 1.008%
z 0.074% 0.326% 1.134% 0.467%

In your opinion, in which languages were written the following four texts (the letters of each word
were mixed).
TMAIER BERACUO RSU NU REBRA PRCEEH EIANTT NE ONS EBC NU GAOFREM EIMATR RERNAD APR L
RDUOE LAHECLE UIL TTNI A EUP SREP EC LGNGAEA TE RBONUJO ERMNOUSI DU UBRACEO QUE
OVSU EEST LIJO UQE OUVS EM MSZELBE BAEU ASNS MIERNT IS RVETO AGRAME ES PRARPTOE A
OEVTR AMGUPLE VUOS SEET EL PNIHXE DSE OSHET ED CSE BIOS A ESC MSOT LE OUBRCEA NE ES
ESTN ASP DE IEJO TE OUPR ERRNOTM AS BELEL XOVI IL OREVU NU RGLEA ECB ILESSA EBOMTR AS
PIOER EL NRDAER S EN ISIAST TE ITD MNO NOB EUSRMNOI NRPEEAZP QEU UTOT EUTLRFTA IVT XUA
SPNEDE DE UECIL UQI L TECEOU TECET NEOCL VATU BNEI UN GMAEORF SNAS TUOED LE EOABURC
OHENTXU TE NSCOFU UJRA SMIA UN EPU TRDA UQ NO EN L Y ARRPEIDNT ULSP
WRE TREITE SO TSPA CUDHR AHNCT UND WIND SE STI RED AEVRT MTI ESEIMN IDNK RE ATH END
NEABNK WLOH IN EMD AMR ER AFTSS HIN IHSERC RE AHTL HIN MRWA EINM SHNO SAW SRTIBG UD
SO NGBA DNEI EIHSGTC ESISTH RAETV UD DEN LERNIOKG NITHC NDE LOENINKGRE TIM OKRN UDN
CHWFSEI NEIM NSOH ES STI IEN BIFTRLSEEEN DU BILESE IKDN OMKM EHG MIT MIR RAG ECHNOS
EPELSI EIPSL IHC ITM RDI HNCMA BEUTN MBLUNE DINS NA DEM TNDRAS NMIEE UTETMR AHT
CAMHN UDNGEL GDAWEN MIEN EATRV MENI VEART DUN OSTHER DU CINTH SAW KNNOEIREGL RIM
ILEES PRSTVRCIEH ISE IHGRU BEEILB RIGUH MNEI KNDI NI RDNEUR NATBRLET STAESUL EDR WNID
DSNOACAIF ORP ANU DAEDALRI DNAAEIMTI EQU NNCOSETE EL RSTEOUL SMA AACTFAITNS UQE EL
TSVAO OINSRVUE DE US ANIGIICANOM EIORDP TOOD RTEIENS RPO LE ITOABOLRROA ED QIUAMALI
USOP A NSSRCAEAD LA TMREAAI NXTADAUEE ROP GOARLS EMESS DE NNAMICLUIAPO Y LOVOIV A
STRINGS – ANALYSIS OF A TEXT 49

RES LE RHMEOB EOMDNEERPRD DE LOS RSOPMRIE OMTSIPE UEQ CIIDADE LE RTDAAOZ ED LSA
CELSAL Y LA NICOIOPS ED LAS UESVNA SSACA Y ES ITRMNEEOD QEU AERFU EL UEQIN IIIRDEGAR LA
NAIORTREICP DE AL RRTEIA
IMTRUESMME DNA TEH LNGIIV SI EYAS SIFH REA GJPNUIM DNA HET TTNOCO IS GHIH OH OUYR
DDADY SI IRHC DAN ROUY MA SI DOGO GKOILON OS USHH LTLIET BBYA NDOT OUY CYR NEO OF
HESET GNSRONIM YUO RE NANGO SIER PU SNIGING NAD OULLY EPADRS YUOR GINSW DAN LYOLU
KATE OT HET KSY TUB ITLL TATH MGNIRNO EREHT NATI INTGOHN ACN AHMR OYU TWIH DADYD NDA
MYMMA NSTIDGAN YB
 Chapter

7
Lists I

A list is a way to group elements into a single object. After defining a list, you can retrieve each item
of the list one by one, but also add new ones. . .

Lesson 1 (List (1)).

A list is a series of elements. This can be a list of integers, for example [5,-7,12,99], or
a list of strings, for example ["March","April","May"] or objects can be of different types
[3.14,"pi",10e-3,"x",True].
• Construction of a list. A list is defined by elements between square brackets:
– mylist1 = [5,4,3,2,1] a list of 5 integers,
– mylist2 = ["Friday","Saturday","Sunday"] a list of 3 strings,
– mylist3 = [] the empty list (very useful to complete it later).
• Get an item. To get an item from the list, simply write mylist[i] where i is the rank of the
desired item.
Beware! The trap is that you start counting from the rank 0.
For example after the instruction mylist = ["A","B","C","D","E","F"] then
– mylist[0] returns "A"
– mylist[1] returns "B"
– mylist[2] returns "C"
– mylist[3] returns "D"
– mylist[4] returns "E"
– mylist[5] returns "F"

"A" "B" "C" "D" "E" "F"

rank : 0 1 2 3 4 5

• Add an element. To add an item at the end of a list, just use the command
mylist.append(element). For example if primes = [2,3,5,7] then primes.append(11)
adds 11 to the list, if you then execute the instruction primes.append(13) then the list primes
is now [2,3,5,7,11,13].
• Example of construction. Here is how to build the list that contains the first squares:

list_squares = [] # Start from the empty list

for i in range(10):
list_squares.append(i**2) # Add squares one by one
LISTS I 51

At the end list_squares is:

[0, 1, 4, 9, 16, 25, 36, 49, 64, 81]

Lesson 2 (List (2)).

• Length of a list. The length of a list is the number of elements it contains. The
command len(mylist) returns the length. The list [5,4,3,2,1] is 5 long, the list
["Friday","Saturday","Sunday"] has length 3, the empty list [] has length 0.
• Browse a list. Here is the easiest way to scan a list (and here to display each item):
for item in mylist:
print(item)
• Browse a list (bis). Sometimes you need to know the rank of the elements. Here is another way
to do it (which here displays the rank and the element).
n = len(mylist)
for i in range(n):
print(i,mylist[i])
• To get a list from range() you have to write:
list(range(n))
• It’s a bad idea to name your list “list” because this word is already used by Python.

Activity 1 (Simple or compound interests).

Goal: create two lists to compare two types of interests.

1. Simple interest. We have a sum of S0 . Each year this investment earns interest based on the initial
amount.
For example, with an initial amount of S0 = 1000 and simple interest of p = 10%. The interest is
100. So after one year, I have a sum of S1 = 1100, after two years S2 = 1200. . .
Program a simple_interest(S0,p,n) function that returns the list of amounts for the n first
years. For example simple_interest(1000,10,3) returns [1000, 1100, 1200, 1300].
2. Compound interest. A sum of S0 brings in compound interest. This time the interest is calculated
each year on the basis of the sum of the previous year, i.e. according to the formula:
p
I n+1 = Sn ×
100
Program a function compound_interest(S0,p,n) which returns the list of amounts of the n first
years. For example compound_interest(1000,10,3) returns [1000, 1100, 1210, 1331].
3. I have the choice between a simple interest investment of 10% or a compound interest investment
of 7%. What is the most advantageous solution depending on the duration of the placement?

Lesson 3 (List (3)).

• Concatenate two lists. If you have two lists, you can merge them by the operator “+”. For example
with mylist1 = [4,5,6] and mylist2 = [7,8,9]
mylist1 + mylist2 is [4,5,6,7,8,9].
LISTS I 52

• Add an item at the end. The operator “+” provides another method to add an item to a list:
mylist = mylist + [element]
For example [1,2,3,4] + [5] is [1,2,3,4,5]. Attention! The element to be added must be
surrounded by square brackets. It is an alternative method to mylist.append(element).
• Add an element at the beginning. With :
mylist = [element] + mylist
the item is added at the beginning of the list. For example [5] + [1,2,3,4] is [5,1,2,3,4].
• Slicing lists. You can extract a whole part of the list at once: mylist[a:b] returns the sublist of
items with ranks a to b − 1.

"A" "B" "C" "D" "E" "F" "G"

rank : 0 1 2 3 4 5 6

For example if mylist = ["A","B","C","D","E","F","G"] then

– mylist[1:4] returns ["B","C","D"]
– mylist[0:2] returns ["A","B"]
– mylist[4:7] returns ["E","F","G"]
Once again, it is important to remember that the rank of a list starts at 0 and that the slicing
mylist[a:b] stops at the rank b − 1.

Activity 2 (Manipulate lists).

Goal: program small routines that manipulate lists.

1. Program a function rotate(mylist) that shifts all the elements of a list by one rank (the last
element becoming the first). The function returns a new list.
For example, rotate([1,2,3,4]) returns the list [4,1,2,3].
2. Program a function inverse(mylist) that inverts the order of the elements in a list.
For example, inverse([1,2,3,4]) returns the list [4,3,2,1].
3. Program a function delete_rank(mylist,rank) that returns a list of all elements, except the
one at the given rank.
For example, delete_rank([8,7,6,5,4],2) returns the list [8,7,5,4] (the item 6 that was
at rank 2 is deleted).
4. Program a function delete_element(mylist,element) returning a list that contains all items
except those equal to the specified element.
For example, delete_element([8,7,4,6,5,4],4) returns the list [8,7,6,5] (all items equal
to 4 have been deleted).

Lesson 4 (Manipulate lists).

You can now use the Python functions which do some of these operations.
• Invert a list. Here are three methods:
– mylist.reverse() modifies the list in place (i.e. mylist is now reversed, the command
returns nothing);
– list(reversed(mylist)) returns a new list;
– mylist[::-1] returns a new list.
LISTS I 53

• Delete an item. The command mylist.remove(element) deletes the first occurrence found
(the list is modified). For example with mylist = [2,5,3,8,5] the call mylist.remove(5)
modifies the list which is now [2,3,8,5] (the first 5 has disappeared).
• Delete an element (bis). The command del mylist[i] deletes the element of rank i (the list is
modified).

Activity 3 (Bubble sort).

Goal: order a list from the smallest to the largest element.

The bubble sort is a simple way to order a list, here it will be from the smallest to the largest element.
The principle is as follows:
• We go through the list from the beginning. As soon as you encounter two consecutive elements in
the wrong order, you exchange them.
• At the end of the first pass, the largest element is at the end and it will not move anymore.
• We restart from the beginning (until the penultimate element), this time the last two elements are
well placed.
• We continue this way. There is a total of n − 1 passages if the list is of length n.

9 7 before

7 9 after

Here is the bubble sort algorithm:

Algorithm.
• – Input: a list ` of n numbers
– Output: the ordered list from the smallest to the largest
• For i ranging from n − 1 to 0:
For j ranging from 0 to i − 1:
If `[ j + 1] < `[ j] then exchange `[ j] and `[ j + 1].
• Return the list `.

Program the bubble sort algorithm into a bubble_sort(mylist) function that returns the or-
dered list of elements. For example bubble_sort([13,11,7,4,6,8,12,6]) returns the list
[4,6,6,7,8,11,12,13].
Hints.
• Begin by defining new_mylist = list(mylist) and works only with this new list.
• For the index i to run backwards from n − 1 to 0, you can use the command :
for i in range(n-1,-1,-1):
Indeed range(a,b,-1) corresponds to the decreasing list of integers i satisfying a > i > b (as
usual the right bound is not included).
LISTS I 54

Lesson 5 (Sorting).
You can now use the sorted() function from Python which orders lists.

python : sorted()
Use: sorted(mylist)
Input: a list
Output: the ordered list of elements
Example: sorted([13,11,7,4,6,8,12,6]) returns the list
[4,6,6,7,8,11,12,13].

Attention! There is also a mylist.sort() method that works a little differently. This command returns
nothing, but on the other hand the list mylist is now ordered. We are talking about a modification in
place.

Activity 4 (Arithmetic).

Goal: improve some functions of the chapter “Arithmetic – While loop – I”.

1. Prime factors. Program a function prime_factors(n) that returns a list of all the prime factors
of an integer n > 2. For example, for n = 12 936, whose decomposition into prime factors is
n = 23 × 3 × 72 × 11, the function returns [2, 2, 2, 3, 7, 7, 11].
Hints. Consult the chapter “Arithmetic – While loop – I”. The core of the algorithm is as follows:
As long as d 6 n:
If d is a divisor of n, then:
add d to the list,
n becomes n/d.
Otherwise increment d by 1.
2. List of prime numbers. Write a function list_primes(n) that returns the list of all prime
numbers less than n. For example list_primes(100) returns the list:
[2,3,5,7,11,13,17,19,23,29,31,37,41,43,47,53,59,61,67,71,73,79,83,89,97]
To do this, you will program an algorithm that is a simple version of the sieve of Eratosthenes:

Algorithm.
• – Input: an integer n > 2.
– Ouput: the list of prime numbers < n.
• Initialize mylist by a list that contains all integers from 2 to n − 1.
• For d ranging from 2 to n − 1:
For k in mylist:
If d divides k and d 6= k, then remove the element k from mylist.
• Return mylist.

Hints.
• Start from mylist = list(range(2,n)).
• Use mylist.remove(k).
LISTS I 55

Explanations. Let’s see how the algorithm works with n = 30.

• At the beginning the list is
[2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]
• We start with d = 2, we eliminate all the numbers divisible by 2, unless it is the number 2:
so we eliminate 4, 6, 8,. . . , the list is now [2, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29].
• We continue with d = 3, we eliminate multiples of 3 (except 3), after these operations the
list is: [2, 3, 5, 7, 11, 13, 17, 19, 23, 25, 29].
• With d = 4, we eliminate multiples of 4 (but there are no more).
• With d = 5 we eliminate multiples of 5 (here we just eliminate 25), the list becomes
[2, 3, 5, 7, 11, 13, 17, 19, 23, 29].
• We continue (here nothing happens anymore).
• At the end, the list is [2, 3, 5, 7, 11, 13, 17, 19, 23, 29].

Lesson 6 (Plot a list).

With the matplotlib module it is very easy to visualize the elements of a list of numbers.

import matplotlib.pyplot as plt

mylist1 = [3,5,9,8,0,3]
mylist2 = [4,7,7,2,8,9]

plt.plot(mylist1,color="red")
plt.plot(mylist2,color="blue")
plt.grid()
plt.show()
Explanations.
• The module is matplotlib.pyplot and is given the new simpler name of plt.
• Attention! The matplotlib module is not always installed by default with Python.
• plt.plot(mylist) traces the points of a list (in the form of (i, `i )) that are linked by segments.
• plt.grid() draws a grid.
LISTS I 56

• plt.show() displays everything.

To display points (x i , yi ) you must provide the list of abscissa then the list of ordinates:
plt.plot(mylist_x,mylist_y,color="red")
Here is an example of a graph obtained by displaying coordinate points of the type (x, y) with y = x 2 .

import matplotlib.pyplot as plt

mylist_x = [2, 3, 5, 7, 9]
mylist_y = [4, 9, 25, 49, 81]
plt.plot(mylist_x,mylist_y,color="red")
plt.grid()
plt.show()

Activity 5 (Ballistics).

Goal: visualize the firing of a cannonball.

LISTS I 57

A cannonball has been fired from the point (0, 0). The trajectory equation is given by the formula:
1 1
y(x) = − g 2 x 2 + tan(α)x
2 v cos2 (α)
where
• α is the angle of the shot,
• v is the initial speed,
• g is the gravitational constant: we will take g = 9.81.

y(x)

v
α

(0, 0) x x

1. Program a function parabolic_shot(x,v,alpha) which returns the value y(x) given by the
formula.
Hint. Be careful with the units for the angle α. If for example you choose that the unit for the angle
is degrees, then to apply the formula with Python you must first convert the angles to radians :
2π
αradian = αdegree
360
2. Program a function list_trajectory(xmax,n,v,alpha) that calculates the list of ordinates
y of the n + 1 points of the trajectory whose abscissa are regularly spaced between 0 and x max .
Method. For i ranging from 0 to n:
x
• calculate x i = i · max
n ,
• calculate yi = y(x i ) using the trajectory formula,
• add yi to the list.
3. For v = 50, x max = 270 and n = 100, displays different trajectories according to the values of α.
What angle α allows to reach the point (x, 0) at ground level as far away from the shooting point
as possible?
 PA R T I I I

A D VA N C E D C O N C E P T S

58
 Chapter

8
Statistics – Data visualization

It’s good to know how to calculate the minimum, maximum, average and quartiles of a series. It’s
even better to visualize them all on the same graph!

Activity 1 (Basic statistics).

Goal: calculate the main characteristics of a series of data: minimum, maximum, mean and
standard deviation.
In this activity mylist refers to a list of numbers (integer or floating point numbers).
1. Write your own function mysum(mylist) which calculates the sum of the elements of a given list.
Compare your result with the function sum() described below which already exists in Python.
Especially for an empty list, check that your result is 0.

python : sum()
Use: sum(mylist)
Input: a list of numbers
Output: a number
Example: sum([4,8,3]) returns 15

You can now use the function sum() in your programs!

2. Write a function mean(mylist) that calculates the average of the items in a given list (and returns
0 if the list is empty).
3. Write your own function minimum(mylist) that returns the smallest value of the items in a given
list. Compare your result with the Python function min() described below (who can also calculate
the minimum of two numbers).

python : min()
Use: min(mylist) or min(a,b)
Input: a list of numbers or two numbers
Output: a number
Example :
• min(12,7) returns 7
• min([10,5,9,12]) returns 5

You can now use the function min(), and of course also the function max() in your programs!
STATISTICS – DATA VISUALIZATION 60

4. The variance of a data series (x 1 , x 2 , . . . , x n ) is defined as the average of the squares of deviations
from the mean. That is to say:
1
v = (x 1 − m)2 + (x 2 − m)2 + · · · + (x n − m)2


n
where m is the average of (x 1 , x 2 , . . . , x n ).
Write a function variance(mylist) that calculates the variance of the elements in a list.
For example, for the series (6, 8, 2, 10), the average is m = 6.5, the variance is
1
v = (6 − 6.5)2 + (8 − 6.5)2 + (2 − 6.5)2 + (10 − 6.5)2 = 8.75.


4
5. The standard deviation of a series (x 1 , x 2 , . . . , x n ) is the square root of the variance:
p
e= v
where v is the variance. Program a function standard_deviation(mylist). With the example
p p
above we find e = v = 8.75 = 2.95 . . .
6. Here are the average monthly temperatures (in Celsius degrees) in London and Chicago.
temp_london = [4.9,5,7.2,9.7,13.1,16.6,18.7,18.2,15.5,11.6,7.7,5.6]
temp_chicago = [-5,-2.7,2.8,9.2,15.2,20.7,23.5,22.6,18.4,12.1,4.8,-1.9]

Calculate the average temperature over the year in London and then in Chicago. Calculate the
standard deviation of the temperatures in London and then in Chicago. What conclusions do you
draw from this?

Lesson 1 (Graphics with tkinter).

To display this:

The code is:

# tkinter window
root = Tk()

canvas = Canvas(root, width=800, height=600, background="white")

canvas.pack(fill="both", expand=True)

# A rectangle
canvas.create_rectangle(50,50,150,100,width=2)

# A rectangle with thick blue edges

canvas.create_rectangle(200,50,300,150,width=5,outline="blue")

# A rectangle filled with purple

STATISTICS – DATA VISUALIZATION 61

canvas.create_rectangle(350,100,500,150,fill="purple")

# An ellipse
canvas.create_oval(50,110,180,160,width=4)

# Some text
canvas.create_text(400,75,text="Bla bla bla bla",fill="blue")

# Launch of the window

root.mainloop()
Some explanations:
• The tkinter module allows us to define variables root and canvas that determine a graphic
window (here width 800 and height 600 pixels). Then describe everything you want to add to the
window. And finally the window is displayed by the command root.mainloop() (at the very
end).
• Attention! The window’s graphic marker has its y-axis pointing downwards. The origin (0, 0) is the
top left corner (see figure below).
• Command to draw a rectangle: create_rectangle(x1,y1,x2,y2); just specify the coordinates
(x 1 , y1 ), (x 2 , y2 ) of two opposite vertices. The option width adjusts the thickness of the line,
outline defines the color of this line, fill defines the filling color.
• An ellipse is traced by the command create_oval(x1,y1,x2,y2), where (x 1 , y1 ), (x 2 , y2 ) are
the coordinates of two opposite vertices of a rectangle framing the desired ellipse (see figure). A
circle is obtained when the corresponding rectangle is a square!
• Text is displayed by the command canvas.create_text(x,y,text="My text") specifying
the coordinates (x, y) of the point from which you want to display the text.

(0, 0) x

(x 1 , y1 )

(x 2 , y2 )
(x 1 , y1 )

(x 2 , y2 )
y

Activity 2 (Graphics).

Goal: visualize data by different types of graphs.

STATISTICS – DATA VISUALIZATION 62

Bar graphics Cumulative graph

Percentage graphics Sector graph

1. Bar graphics. Write a function bar_graphics(mylist) that displays the values of a list as
vertical bars.
Hints.
• First of all, don’t worry about drawing the vertical axis of the coordinates with the figures.
• You can define a variable scale that allows you to enlarge your rectangles, so that they have
a size adapted to the screen.
• If you want to test your graph with a random list, here is how to build a random list of 10
integers between 1 and 20 :
from random import *
mylist = [randint(1,20) for i in range(10)]
2. Cumulative graph. Write a function cumulative_graphics(mylist) that displays the values
of a list in the form of rectangles one above the other.
3. Graphics with percentage. Write a function percentage_graphics(mylist) that displays the
values of a list in a horizontal rectangle of fixed size (for example 500 pixels) and is divided into
sub-rectangles representing the values.
4. Sector graphics. Write a function sector_graphics(mylist) that displays the values of a list
as a fixed size disk divided into sectors representing the values.
The create_arc() function, which allows you to draw arcs of circles, is not very intuitive. It is
necessary to think that we draw a circle, by specifying the coordinates of the corners of a square
which surrounds it, then by specifying the starting angle and the angle of the sector (in degrees).
canvas.create_arc(x1,y1,x2,y2,start=start_angle,extent=my_angle)
STATISTICS – DATA VISUALIZATION 63

extent=θ

(x 1 , y1 )

start=θ0

(x 2 , y2 )

The option style=PIESLICE displays a sector instead of an arc.

5. Bonus. Gather your work into a program that allows the user to choose the diagram he wants
by clicking on buttons, and also the possibility to get a new random series of data. To create and
manage buttons with tkinter, see the lesson below.

Lesson 2 (Buttons with tkinter).

It is more ergonomic to display windows where actions are performed by clicking on buttons. Here is the
window of a small program with two buttons. The first button changes the color of the rectangle, the
second button ends the program.
STATISTICS – DATA VISUALIZATION 64

The code is:

from tkinter import *
from random import *

root = Tk()
canvas = Canvas(root, width=400, height=200, background="white")
canvas.pack(fill="both", expand=True)

def action_button():
canvas.delete("all") # Clear all
colors = ["red","orange","yellow","green","cyan","blue","violet","purple"]
col = choice(colors) # Random color
canvas.create_rectangle(100,50,300,150,width=5,fill=col)
return

button_color = Button(root,text="View", width=20, command=action_button)

button_color.pack(pady=10)

button_quit = Button(root,text="Quit", width=20, command=root.quit)

button_quit.pack(side=BOTTOM, pady=10)

root.mainloop()
Some explanations:
• A button is created by the command Button. The text option customizes the text displayed on
the button. The button created is added to the window by the method pack.
• The most important thing is the action associated with the button! It is the option command that
receives the name of the function to be executed when the button is clicked. For our example
command=action_button, associate the click on the button with a change of color.
• Attention! You have to give the name of the function without brackets: commande=my_function
and not command=my_function().
• To associate the button with “Quit” and close the window, the argument is command=root.quit.
• The instruction canvas.delete("all") deletes all drawings from our graphic window.

Activity 3 (Median and quartiles).

Goal: calculate the median and quartiles of some data.

1. Program a function median(mylist) which calculates the median value of the items of a given
list. By definition, half of the values are less than or equal to the median, the other half are greater
than or equal to the median.
Background. We note n the length of the list and we assume that the list is ordered (from the
smallest to the largest element).
• Case n odd. The median is the value of the list at the rank n−1 2 . Example with mylist =
[12,12,14,15,19]:
– the length of the list is n = 5 (indices range from 0 to 4),
– the middle index is the rank 2,
STATISTICS – DATA VISUALIZATION 65

– the median is the value mylist[2], so it is 14.

n
• Case n even. The median is the average between the value of the list at rank 2 − 1 and rank
2 . Example with mylist = [10,14,19,20]:
n

– the length of the list is n = 4 (indices range from 0 to 3),

– the middle indices are 1 and 2,
– the median is the average between mylist[1] and mylist[2], so it is 14+19 2 = 16.5.
2. The results of a class are collected in the following form of a number of students per grade:
grade_count = [0,0,1,2,5,2,3,5,4,1,2]
The rank i range is from 0 to 10. And the value at rank i indicates the number of students who
received the grade i. For example here, 1 student got the grade 2, 2 students got the grade 3, 5
students got 4, . . . Write a grades_to_list(grade_count) function that takes the numbers of
students for each grade as input and returns the list of all grades. For our example the function
must return [2,3,3,4,4,4,4,4,5,5,6,6,6,7,...].
Deduce a function that calculates the median of a class’s scores from the numbers of students for
each grade.
3. Write a function quartiles(mylist) that calculates the quartiles Q 1 , Q 2 , Q 3 of the items in a
given list. The quartiles divide the values into: one quarter below Q 1 , one quarter between Q 1 and
Q 2 , one quarter between Q 2 and Q 3 , one quarter above Q 3 . For the calculation, we will use that:
• Q 2 is simply the median of the entire list (assumed ordered),
• Q 1 is the median of the sublist formed by the first half of the values,
• Q 3 is the median of the sublist formed by the second half of the values.
For the implementation, it is necessary to discuss again whether the length n of the list is even or not.
Deduce a function that calculates the quartiles of a class’s grades from a list of a number of students per
grade.

Activity 4 (Box plot).

Goal: drawing box plots.

A box plot is a diagram that represents the main characteristics of a statistical series: minimum, maximum,
median and quartiles. The schematic diagram is as follows:

Q1 median Q3
minimum maximum

Write a box_plot(grade_count) function that draws the box plot of a class’s grades from a list of a
number of students per grade (see previous activity).
STATISTICS – DATA VISUALIZATION 66

Activity 5 (Moving average).

Goal: calculate moving averages in order to draw “smooth” curves.

1. Simulates the stock market price of the index NISDUQ over 365 days. At the beginning, the day
j = 0, the index is equal to 1000. Then the index for a day is determined by adding a random
value (positive or negative) to the value of the previous day’s index:
index of the day j = index of the day ( j − 1) + random value.
For this random value, you can try a formula like:
value = randint(-10,12)/3
Write a function index_stock_exchange(), without parameter, which returns a list of 365 index
values using this method.
2. Trace point by point the index curve over a year. (To draw a point, you can display a square with a
size of 1 pixel.)

3. Since the daily index curve is very chaotic, we want to smooth it out in order to make it more readable.
For this we calculate moving averages.
The moving average at 7 days for the day j, is the average of the last 7 indices. For example: the
moving average (at 7 days) for the day j = 7 is the average of the day’s indices j = 1, 2, 3, 4, 5, 6, 7.
You can change the duration: for example the moving average at 30 days is the average of the last
30 indices.
Write a function moving_average(mylist,duration) that returns a list of all moving averages
in a data set with respect to a fixed time.
4. Trace point by point on the same graph: the index curve over a year (in red below), the moving
average curve at 7 days (in blue below) and the moving average curve at 30 days (in brown below).
Note that the longer the duration, the more the curve is “smooth”. (Of course the moving average
curve at 30 days only starts from the thirtieth day.)
STATISTICS – DATA VISUALIZATION 67
 Chapter

9
Files

You will learn to read and write data with files.

Lesson 1 (Write to a file).

Writing to a file is almost as easy as displaying a sentence on the screen. On the left is a program that
writes two lines in a file called my_file.txt; on the right is the resulting file that is displayed in a text
editor.

fi = open("my_file.txt","w")

fi.write("Hello world!\n")

line = "Hi there.\n"

fi.write(line)

fi.close()

Explanations.
• The command open allows you to open a file. The first argument is the name of the file. The second
argument here is "w" to say that you want to write to the file.
• We do not work with the file name, but with the value returned by the function open. Here we
have named this object file fi. It is with this variable fi that we work now.
• We now write in the file almost as we would display a sentence on the screen. The instruction is
fi.write() where the argument is a string of characters.
• To switch to the next line, you must add the line terminator character "\n".
• It is important to close your file when you are finished writing. The instruction is fi.close().
• The data to be written are strings, so to write a number, you must first transform it with
str(number).

Lesson 2 (Read a file).

It is just as easy to read a file. Here is how to do it (left) and the display by Python on the screen (right).
FILES 69

fi = open("my_file.txt","r")
Hello world!
for line in fi:
print(line)
Hi there.
fi.close()

Explanations.
• The command open is this time called with the argument "r" (for read), it opens the file in reading.
• We work again with an object file named here fi.
• A loop goes through the entire file line by line. Here we just ask for the display of each line.
• We close the file with fi.close().
• The data read is a string, so if you want a number, you must first transform it with int(string)
(for an integer) or float(string) (for a decimal number).

Activity 1 (Read and write a file).

Goal: write a file of grades, then read it to calculate the averages.

1. Generates at random a grades file, named grades.txt, which is composed of lines with the
structure:
Firstname Name grade1 grade2 grade3
For example:

Tintin Vador 15.0 5.0 19.0

Bill Croft 15.0 14.5 10.5
Hermione Skywalker 10.5 7.0 19.5
Lara Parker 12.5 13.0 14.5
Hermione Croft 11.5 18.5 9.5
Robin Vador 8.0 8.0 11.0

Hints.
• Build a list of first names list_firstnames = ["Alice","Tintin","James",...].
Then choose a first name at random by the instruction firstname =
choice(list_firstnames) (you have to import the module random).
• Same thing for names!
• For a grade, you can choose a random number with the command randint(a,b).
• Please note! Don’t forget to convert the numbers into a string to write it to the file:
str(grade).
2. Read the file grades.txt that you produced. Calculate the average of each person’s grades and
write the result to a file averages.txt where each line is of the form:
Firstname Name average
For example:
FILES 70

Tintin Vador 13.00

Bill Croft 13.33
Hermione Skywalker 12.33
Lara Parker 13.33
Hermione Croft 13.17
Robin Vador 9.00

Hints.
• For each line read from the file grades.txt, you get the data as a list by the command
line.split().
• Please note! The data read is a string. You can convert a string "12.5" to the number 12.5
by the instruction float(string).
• To convert a number to a string with only two decimal places after the dot, you can use the
command '{0:.2f}'.format(number).
• Don’t forget to close all your files.

Lesson 3 (Files with format csv).

The format csv (for comma-separated values) is a very simple text file format containing data. Each line of
the file contains a series of data (numbers or text). On the same line the data are separated by a comma
(hence the name of the format, even if other separators are possible).
Example. Here is a file that contains the names, first names, years of birth, height and number of Nobel
Prize medals:

CURIE,Marie,1867,1.55,2
EINSTEIN,Albert,1879,1.75,1
NOBEL,Alfred,1833,1.70,0

Activity 2 (csv format).

Goal: write a data file, with csv format, then read it for a graphic display.

1. Generates a sales.csv file of sales figures (randomly drawn) from a sports brand.
Here is an example:

Best sales of the brand 'Pentathlon'

,2015,2016,2017,2018,2019,2020

Mountain bike,560,890,500,550,650,970
Surfboard,550,690,750,640,710,790
Running shoes,850,740,790,1000,680,540
Badminton racket,710,640,620,550,790,880
Volley ball,900,550,790,510,930,800

• The data starts from the fifth line.

FILES 71

• The produced file respects the format csv and must be readable by LibreOffice Calc for example.

2. Reads the file sales.csv to display the sales curves.

Hints.
• The matplotlib package allows you to easily display graphics, it is often called with the
instruction:
import matplotlib.pyplot as plt
• Here is how to view two data lists mylist1 and mylist2:

plt.plot(mylist1)
plt.plot(mylist2)
plt.grid()
plt.show()

Lesson 4 (Images bitmap).

There is a simple file format, called the bitmap format, which describes an image pixel by pixel. This
format is available in three variants depending on whether the image is in black and white, grayscale or
color.
Black and white picture, format “pbm”.
The image is described by 0 and 1.
Here is an example: the file image_bw.pbm on the left (read as a text file) and on the right its visualization
FILES 72

(using an image reader, here very enlarged).

P1
4 5
1 1 1 1
1 0 0 0
1 1 1 0
1 0 0 0
1 1 1 1

Here is the description of the format:

• First line: the id P1.
• Second line: the number of columns, then the number of rows (here 4 columns and 5 rows).
• Then the color of each pixel line by line: 1 for a black pixel, 0 for a white pixel. Warning: this is
contrary to the usual convention!
Grayscale image, format “pgm”.
The image is described by different values for different grayscales. Here is an example: the file
image_gray.pbm on the left and on the right its visualization.

P2
4 5
255
0 0 0 0
192 192 192 192
192 255 128 128
192 255 64 64
192 0 0 0

Here is the description of the format:

• First line: the identifier is this time P2.
• Second line: the number of columns, then the number of rows.
• Third line: the maximum value of the grayscale (here 255).
• Then the grayscale of each pixel line by line: this time 0 for a black pixel, the maximum value for a
white pixel and intermediate values give intermediate grays.
Image in colors, format “ppm”.
The image is described by three values per pixel: one for red, one for green, one for blue.
Here is an example: the file image_col.ppm on the left and on the right its visualization.

P3
3 2
255
255 0 0 0 255 0 0 0 255
0 128 255 255 128 0 128 255 0

Here is the description of the format:

• First line: the id is now P3.
FILES 73

• Second line: the number of columns, then the number of rows.

• Third line: the maximum value of the color levels (here 255).
• Then each pixel is described by 3 numbers: the level of red, green and blue (RGB system). For
example, the first pixel is encoded by (255, 0, 0) so it is a red pixel.

Activity 3 (Images bitmap).

Goal: define your own images pixel by pixel.

1. Generate a file image_bw.pbm that represents a black and white image (e.g. of size 300 × 200)
according to the following pattern:

Hints. If i designates the line number and j the column number (from the top left), then the pixel
in position (i, j) is white if i + j is between 0 and 9, or between 20 and 29, or between 40 and
49,. . . This is obtained by the formula :
col = (i+j)//10 % 2
which returns 0 or 1 as desired.
2. Generate a file image_gray.pgm that represents a grayscale image (for example of size 200 × 200
with 256 grayscale) according to the following pattern:

Hints. This time the formula is:

col = (i**2 + j**2) % 256
which returns an integer between 0 and 255.
3. Generate a file image_col.ppm that represents a color image (e.g. of size 200 × 200, with 256
red, green and blue levels) according to the following pattern:
FILES 74

Hints. This time the formula is:

R = (i*j) % 256
G = i % 256
B = (i + j)//3 % 256
which gives the red, green and blue levels of the pixel (i, j).
4. Write a function inverse_black_white(filename) that reads a black and white image file
.pbm and creates a new file in which the white pixels have turned black and vice versa.
Example: on the left the start image, on the right the finish image.

5. Write a function colors_to_gray(filename) that reads a color image file in .ppm format and
creates a new file in .pgm format in which the color pixels are transformed into grayscale.
You can use the formula:
g = 0, 21 × R + 0, 72 × G + 0, 07 × B
where
• R, G, B are the red, green and blue levels of the colored pixel,
• g is the grayscale of the transformed pixel.
Example: on the left the starting image in color, on the right the arrival image in grayscale.

Activity 4 (Distance between two cities).

Goal: read the coordinates of the cities and write the distances between them.

1. Distance in the plan.

Write a program that reads a file containing the coordinates (x, y) of cities, then calculates and
writes in another file the distances (on the map) between two cities.
FILES 75

The formula for the distance between two points (x 1 , y1 ) and (x 2 , y2 ) of the plan is:
Æ
d = (x 2 − x 1 )2 + ( y2 − y1 )2 .
Example. Here is an example of an input file:

And here is the output file produced by the program:

We read on this file that the distance between Lille and Marseille is 1077 kilometers.
Below is the map of France that provided (very approximate) data for the input file. The origin is
at the bottom left, each side of a square represents 100 km. For example, in this reference, Paris
has the coordinates (500, 800).

Lille

Strasbourg
Paris
Nantes

Lyon

Bordeaux

Toulouse

(0, 0) Marseille
100 km

2. Distance on the sphere.

On Earth, the distance between two cities is a distance on a great circle along the surface of the
sphere and not along a straight line. This is the distance a plane travels to connect two cities.
Write a program that reads the latitudes and longitudes of cities, then calculates and writes to
another file the distances (on the Earth’s surface) between two cities.
Example. Here is an example of an input file:
FILES 76

And here is the output file produced by the program:

Implementation and explanations.

• The input file contains the latitude (noted ϕ) and longitude (noted λ) in degrees for each
city. For example, Paris has a latitude of ϕ = 48.853 degrees and a longitude of λ = 2.350
degrees.
• For the formulas, it will be necessary to use the angles in radians. The formula for converting
degrees to radians is:
2π
angle in radians = × angle in degrees
360
• Formula for an approximate distance.
There is a simple formula that gives a good estimate for the shortest distance between two
points of a sphere with a radius of R. First, set:
ϕ + ϕ 
1 2
x = (λ2 − λ1 ) · cos and y = ϕ2 − ϕ1
2
The approximate distance is then
Æ
d̃ = R x 2 + y 2
where (ϕ1 , λ1 ) and (ϕ2 , λ2 ) are the latitudes/longitudes of two cities expressed in radians.
• Exact distance formula.
The bravest can use the exact formula to calculate the distance. First, set:
  ϕ + ϕ 2 λ2 − λ1 2
  ‹‹
1 2
a = sin + cos(ϕ1 ) · cos(ϕ2 ) · sin
2 2
The exact distance is then:
p p

d = 2 · R · atan2 a, 1 − a
where atan2( y, x) is the function “arctangent” which is obtained by the command
atan2(y,x) from the module math.
• For the Earth radius we will take R = 6371 km.
 Chapter

Arithmetic – While loop – II

10

Our study of numbers is further developed with the loop “while”. For this chapter you need your
function is_prime() built in the part “Arithmetic – While loop – I”.

Activity 1 (Goldbach’s conjecture(s)).

Goal: study two Goldbach conjectures. A conjecture is a statement that you think is true but you
can’t prove it.
1. Goldbach’s good guess: Every even integer greater than 4 is the sum of two prime numbers.
For example 4 = 2 + 2, 6 = 3 + 3, 8 = 3 + 5, 10 = 3 + 7 (but also 10 = 5 + 5), 12 = 5 + 7,. . . For
n = 100 there are 6 solutions: 100 = 3 + 97 = 11 + 89 = 17 + 83 = 29 + 71 = 41 + 59 = 47 + 53.
No one can prove this conjecture, but you will see that there are good reasons to believe it is true.
(a) Program a function number_solutions_goldbach(n) which for a given even integer n,
finds how many decompositions n = p + q there are with p and q two prime numbers and
p 6 q.
For example, for n = 8, there is only one solution 8 = 3 + 5, but for n = 10 there are two
solutions 10 = 3 + 7 and 10 = 5 + 5.
Hints.
• It is therefore necessary to test all p including 2 and n/2;
• set q = n − p;
• we have a solution when p 6 q and p and q are both prime numbers.
(b) Proves with the machine that the Goldbach conjecture is verified for all even integers n
between 4 and 10 000.
2. Goldbach’s bad guess: Every odd integer n can be written as
n = p + 2k2
where p is a prime number and k is an integer (possibly zero).
(a) Program a function is_decomposition_goldbach(n) that returns “True” when there is a
decomposition of the form n = p + 2k2 .
(b) Show that Goldbach’s second guess is wrong! There are two integers smaller than 10 000
that do not have such decomposition. Find them!

Activity 2 (Numbers with 4 or 8 divisors).

Goal: disprove a conjecture by doing a lot of calculations.

Conjecture: Between 1 and N , there are more integers that have exactly 4 divisors than integers that have
exactly 8 divisors.
ARITHMETIC – WHILE LOOP – II 78

You will see that this conjecture looks true for N rather small, but you will show that this conjecture is
false by finding a large N that contradicts this statement.
1. Number of divisors.
Program a function number_of_divisors(n) that returns the number of integers dividing n.
For example: number_of_divisors(100) returns 9 because there are 9 divisors of n = 100 :
1, 2, 4, 5, 10, 20, 25, 50, 100
Hints.
• Don’t forget 1 and n as divisors.
• Try to optimize your function because you will use it intensively: for example, there are no
divisors strictly larger than 2n (except n).
2. 4 or 8 divisors.
Program a function four_and_eight_divisors(Nmin,Nmax) that returns two numbers: (1)
the number of integers n with Nmin 6 n < Nmax that admit exactly 4 divisors and (2) the number
of integers n with Nmin 6 n < Nmax that admit exactly 8 divisors.
For example four_and_eight_divisors(1,100) returns (32,10) because there are 32 inte-
gers between 1 and 99 that admit 4 divisors, but only 10 integers that admit 8.
3. Proof that the conjecture is false.
Experiment that for “small” values of N (up to N = 10 000 for example) there are more integers
with 4 divisors than 8. But calculate that for N = 300 000 this is no longer the case.
Hints. As there are many calculations, you can split them into slices (the slice of integers 1 6
n < 50 000, then 50 000 6 n < 100 000,...) and then add them up. This allows you to share your
calculations between several computers.

Activity 3 (121111. . . is never prime?).

Goal: study a new false conjecture!

We call Uk the following integer:

Uk = 1 2 1
| 1 1 .{z
.. 111
}
k occurrences of 1
formed by the digit 1, then the digit 2, then k times the digit 1.
For example U0 = 12, U1 = 121, U2 = 1211, . . .
1. Write a function one_two_one(k) that returns the integer Uk .
Hint. You can notice that starting with U0 = 12, we have the relationship Uk+1 = 10 · Uk + 1. So
you can start with u = 12 and repeat a number of times u = 10*u + 1.
2. Check with the machine that U0 , . . . , U20 are not prime numbers.
You might think it’s still the case, but it’s not true. The integer U136 is a prime number! Unfortunately
it is too big to be verified with our algorithms. In the following we will define what is a almost prime
number to be able to push the calculations further.
3. Program a function is_almost_prime(n,r) that returns “True” if the integer n does not admit
any divisor d such as 1 < d 6 r (we assume r < n).
For example: n = 143 = 11 × 13 and r = 10, then is_almost_prime(n,r) is “True” because n
does not allow any divisor less than or equal to 10. (But of course, n is not a prime number.)
Hint. Adapt your function is_prime(n)!
4. Find all the integers Uk with 0 6 k 6 150 which are almost prime for r = 1 000 000 (i.e. they are
not divisible by any integer d with 1 < d 6 1 000 000).
ARITHMETIC – WHILE LOOP – II 79

Hint. In the list you must find U136 (which is a prime number) but also U34 which is not prime but
whose smallest divisor is 10 149 217 781.

Activity 4 (Integer square root).

Goal: calculate the integer square root of an integer.

Let n > 0 be an integer. The integer square root of n is the largest integer r > 0 such as r 2 6 n. Another
p
definition is to say that the integer square root of n is the integer part of n.
Examples:
• n = 21, then the integer square root of n is 4 (because 42 6 21, but 52 > 21). In other words,
p
21 = 4.58 . . ., and we only keep the integer part (the integer to the left of the dot), so it is 4.
• n = 36, then the integer square root of n is 6 (because 62 6 36, but 72 > 36). In other words,
p
36 = 6 and the integer square root is of course also 6.
p
1. Write a first function that calculates the integer square root of an integer n, first calculating n,
then taking the integer part.
Hints.
• For this question only, you can use the module math of Python.
• In this module sqrt() returns the real square root.
• The function floor() of the same module returns the integer part of a number.
2. Write a second function that calculates the integer square root of an integer n, but this time
according to the following method:
• Start with p = 0.
• As long as p2 6 n, increment the value of p by 1.
Test carefully what the returned value should be (beware of the offset!).
3. Write a third function that still calculates the integer square root of an integer n with the algorithm
described below. This algorithm is called the Babylonian method (or Heron’s method or Newton’s
method).

Algorithm.
Input: a positive integer n
Output: its integer square root
• Start with a = 1 and b = n.
• as long as |a − b| > 1:
– a ← (a + b)//2 ;
– b ← n//a
• Return the minimum between a and b: this is the integer square root of n.

We do not explain how this algorithm works, but it is one of the most effective methods to calculate
square roots. The numbers a and b provide, during execution, an increasingly precise interval
p
containing of n.
Here is a table that details an example calculation for the integer square root of n = 1664.
ARITHMETIC – WHILE LOOP – II 80

Step a b
i=0 a=1 b = 1664
i=1 a = 832 b=2
i=2 a = 417 b=3
i=3 a = 210 b=7
i=4 a = 108 b = 15
i=5 a = 61 b = 27
i=6 a = 44 b = 37
i=7 a = 40 b = 41
In the last step, the difference between a and b is less than or equal to 1, so the integer square
root is 40. We can verify that this is correct because: 402 = 1600 6 1664 < 412 = 1681.
Bonus. Compare the execution speeds of the three methods using timeit(). See the chapter “Functions”.

Lesson 1 (Exit a loop).

It is not always easy to find the right condition for a loop “while”. Python has a command to immediately
exit a loop “while” or a loop “for”: this is the instruction break.
Here are some examples that use this command break. As it is rarely an elegant way to write your
program, alternatives are also presented.

Example.
Here are different codes for a countdown from 10 to 0.
# Better (with a flag)
# Countdown
n = 10 # Even better
n = 10
finished = False # (reformulation)
while True: # Infinite loop
while not finished: n = 10
print(n)
print(n) while n >= 0:
n = n - 1
n = n - 1 print(n)
if n < 0:
if n < 0: n = n - 1
break # Immediate stop
finished = True

Example.
Here are programs that search for the integer square root of 777, i.e. the largest integer i that satisfies
i 2 6 777. In the script on the left, the search is limited to integers i between 0 and 99.
# Integer square root # Better
n = 777 n = 777
for i in range(100): i = 0
if i**2 > n: while i**2 <= n:
break i = i + 1
print(i-1) print(i-1)
ARITHMETIC – WHILE LOOP – II 81

Example.
Here are programs that calculate the real square roots of the elements in a list, unless of course the
number is negative. The code on the left stops before the end of the list, while the code on the right
handles the problem properly.
# Better with try/except
# Square root of the elements
mylist = [3,7,0,10,-1,12]
# of a list
for element in mylist:
mylist = [3,7,0,10,-1,12]
try:
for element in mylist:
print(sqrt(element))
if element < 0:
except:
break
print("Warning, I don't know how to
print(sqrt(element))
compute the square root of",element)
 Chapter

Binary I
11

The computers transform all data into numbers and manipulate only those numbers. These numbers
are stored in the form of lists of 0’s and 1’s. It’s the binary numeral system of numbers. To better
understand this binary numeral system, you will first understand the decimal numeral system better.

Lesson 1 (Base 10 system).

We usually denote integers with the decimal numeral system (based on 10). For example, 70 685 is
7 × 10 000 + 0 × 1 000 + 6 × 100 + 8 × 10 + 5 × 1 :
7 0 6 8 5
10000 1000 100 10 1
104 103 102 101 100
(we can see that 5 is the number of units, 8 the number of tens, 6 the number of hundreds. . . ).
It is necessary to understand the powers of 10. We note 10k for 10 × 10 × · · · × 10 (with k factors).

d p−1 d p−2 ··· di ··· d2 d1 d0

10 p−1
10 · · · 10 · · · 10 10 100
p−2 i 2 1

We calculate the integer corresponding to the digits [d p−1 , d p−2 , . . . , d2 , d1 , d0 ] by the formula :
n = d p−1 × 10 p−1 + d p−2 × 10 p−2 + · · · + di × 10i + · · · + d2 × 102 + d1 × 101 + d0 × 100

Activity 1 (From decimal system to integer).

Goal: from the decimal notation, find the integer.

Write a function decimal_to_integer(list_decimal) which from a list representing the decimal

notation calculates the corresponding integer.

decimal_to_integer()
Use: decimal_to_integer(list_decimal)
Input: a list of numbers between 0 and 9
Output: the integer whose decimal notation is the list
Example: if the input is [1,2,3,4], the output is 1234.

Hints. It is necessary to sum up elements of type:

di × 10i for 0 6 i < p
BINARY I 83

where p is the length of the list and di is the digit at rank i counting from the end (i.e. from right to left).
To manage the fact that the index i used for the power of 10 does not correspond to the rank in the list,
there are two solutions:
• understand that di = mylist[p − 1 − i] where mylist is the list of p digits,
• or start by reversing mylist.

Lesson 2 (Binary).
• Power of 2. We note 2k for 2 × 2 × · · · × 2 (with k factors). For example, 23 = 2 × 2 × 2 = 8.
27 26 25 24 23 22 21 20
128 64 32 16 8 4 2 1

• Binary system: an example.

Integers admit a binary numeral system, i.e. a notation where the only numbers are 0 or 1. In
binary, the digits are called bits. For example, 1.0.1.1.0.0.1 (pronounce the numbers one by one) is
the binary numeral system of the integer 89. How to do this calculation? It’s like the base 10, but
using the powers of 2.
1 0 1 1 0 0 1
64 32 16 8 4 2 1
26 25 24 23 22 21 20
So the system 1.0.1.1.0.0.1 represents the integer:
1 × 64 + 0 × 32 + 1 × 16 + 1 × 8 + 0 × 4 + 0 × 2 + 1 × 1 = 64 + 16 + 8 + 1 = 89.
• Binary system: formula.

b p−1 b p−2 ··· bi ··· b2 b1 b0

2 p−1
2 · · · 2 · · · 2 2 20
p−2 i 2 1

We calculate the integer corresponding to the bits [b p−1 , b p−2 , . . . , b2 , b1 , b0 ] as a sum of terms
bi × 2i , by the formula :
n = b p−1 × 2 p−1 + b p−2 × 2 p−2 + · · · + bi × 2i + · · · + b2 × 22 + b1 × 21 + b0 × 20
• Python and the binary. Python accepts that we directly write the integers in binary numeral
system, just use the prefix “0b”. Examples:
– with x = 0b11010, then print(x) displays 26,
– with y = 0b11111, then print(y) displays 31,
– and print(x+y) displays 57.

Activity 2 (From binary numeral system to integer).

Goal: find the integer from a binary notation.

1. Calculate integers whose binary numeral system is given below. You can do it by hand or help
yourself from Python. For example 1.0.0.1.1 is equal to 24 + 21 + 20 = 19 as confirmed by the
command 0b10011 which returns 19.
• 1.1, 1.0.1, 1.0.0.1, 1.1.1.1
• 1.0.0.0.0, 1.0.1.0.1, 1.1.1.1.1
• 1.0.1.1.0.0, 1.0.0.0.1.1
BINARY I 84

• 1.1.1.0.0.1.0.1
2. Write a function binary_to_integer(list_binary) which from a list representing binary
system calculates the corresponding integer.

binary_to_integer()
Use: binary_to_integer(list_binary)
Input: a list of bits, 0 and 1
Output: the integer whose binary numeral system is the list
Examples:
• input [1,1,0], output 6
• input [1,1,0,1,1,1], output 55
• input [1,1,0,1,0,0,1,1,0,1,1,1], output 3383

Hints. This time it is necessary to sum up elements of type:

bi × 2i for 0 6 i < p
where p is the length of the list and bi is the bit (0 or 1) at rank i of the list counting from the end.
3. Here is an algorithm that does the same work: it allows the calculation of the integer from the
binary notation, but it has the advantage of never directly calculating powers of 2. Program this
algorithm into a binary_to_integer_bis() function that has the same characteristics as the
previous function.

Algorithm.
Input: mylist : a list of 0 and 1
Output: the associated binary number
• Initialize a variable n to 0.
• For each element b from mylist :
– if b is 0, then make n ← 2n,
– if b is 1, then make n ← 2n + 1.
• The result is the value of n.

Lesson 3 (Base 10 system (bis)).

Of course, when you see the number 1234 you can immediately find the list of its digits [1, 2, 3, 4]. But
how to do it in general from an integer n?
1. We will need the Euclidean division by 10:
• calculate by n % 10 the remainder, we also call it n modulo 10
• calculate n // 10 the quotient of this division.
BINARY I 85

quotient

123

10 1234
remainder
4

2. Commands with Python are simply n % 10 and n // 10.

Example: 1234 % 10 is equal to 4; 1234 // 10 is equal to 123.
3. • The digit of the units of n is obtained as the remainder modulo 10: it is n%10. Example
1234%10 = 4.
• The tens figure is obtained from the quotient of n per 10, then by taking the figure of the
units of this number: it is therefore (n//10)%10. Example: 1234//10 = 123, then we have
123%10 = 3; the figure of tens of 1234 is indeed 3.
• For the figure of hundreds, we calculate the quotient of the division of n by 100, then we take
the figure of the units. Example 1234//100 = 12 ; 2 is the figure of the units of 12 and it is
the figure of the hundreds of 1234. Note: dividing by 100 means dividing by 10, then again
by 10.
• For the figure of thousands we calculate the quotient of the division by 1000 then we take the
figure of the units. . .

Activity 3 (Find the digits of an integer).

Goal: decompose an integer into a list of its figures (based on 10).

Program the following algorithm into a function integer_to_decimal().

integer_to_decimal()
Use: integer_to_decimal(n)
Input: a positive integer
Output: the list of its digits
Example: if the input is 1234, the output is [1,2,3,4].

Algorithm.
Input: an integer n > 0
Output: the list of its digits
• Start from an empty list.
• As long as n is not zero:
– add n%10 at the beginning of the list,
– make n ← n//10.
• The result is the list.
BINARY I 86

Lesson 4 (Binary system with Python).

Python calculates very well the binary numeral system of an integer using the function bin().

python : bin()
Use: bin(n)
Input: an integer
Output: the binary numeral system of n in the form of a string starting with
'0b'
Example:
• bin(37) returns '0b100101'
• bin(139) returns '0b10001011'

Lesson 5 (Binary system calculation).

To calculate the binary numeral system of an integer n, it is the same method as to calculate the decimal
numeral system but by replacing the divisions by 10 by divisions by 2.
So we need:
• n%2 : the remainder of the Euclidean division from n to 2 (also called n modulo 2); the remainder
is either 0 or 1.
• and n//2: the quotient of this division.
Note that the remainder n%2 is either 0 (when n is even) or 1 (when n is odd).
Here is a general method to calculate the binary numeral system of an integer:
• We start from the integer whose binary notation we want.
• A series of Euclidean divisions by 2 are performed:
– for each division, you get an remainder 0 or 1;
– we get a quotient that we divide again by 2. . . We stop when this quotient is zero.
• The binary system is read as the sequence of the remainders, but starting from the last one.
BINARY I 87

Example.
Calculation of the binary notation of 14.
• We divide 14 by 2, the quotient is 7, the remainder is 0.
• We divide 7 (the previous quotient) by 2: the new quotient is 3, the new remainder is 1.
• We divide 3 by 2: quotient 1, remainder 1.
• We divide 1 by 2: quotient 0, remainder 1.
• It’s over (the last quotient is zero).
• The successive remainders are 0, 1, 1, 1. we read the binary numeral system backwards, it is
1.1.1.0.
The divisions are done from left to right, but the remainders are read from right to left.

7 3 1 0

2 14 2 7 2 3 2 1

0 1 1 1

Example.
Binary system of 50.

25 12 6 3 1 0

2 50 2 25 2 12 2 6 2 3 2 1

0 1 0 0 1 1

The successive remainders are 0, 1, 0, 0, 1, 1, so the binary notation of 50 is 1.1.0.0.1.0.

Activity 4.

Goal: find the binary numeral system of an integer.

1. Calculate by hand the binary numeral system of the following integers. Check your results using
the bin() Python function.
• 13, 18, 29, 31,
• 44, 48, 63, 64,
• 100, 135, 239, 1023.
2. Program the following algorithm into a function integer_to_binary().
BINARY I 88

Algorithm.
Input: an integer n > 0
Output: its binary numeral system in the form of a list
• Start from an empty list.
• As long as n is not zero:
– add n%2 at the beginning of the list,
– make n ← n//2.
• The result is the list.

integer_to_binary()
Use: integer_to_binary(n)
Input: a positive integer
Output: its binary notation in the form of a list
Example: if the input is 204, the output is [1,1,0,0,1,1,0,0].

Make sure your functions are working properly:

• start with any integer n,
• calculate its binary numeral system,
• calculate the integer associated with this binary notation,
• you have to find the starting integer!
 Chapter

12
Lists II

The lists are so useful that you have to know how to handle them in a simple and efficient way. That’s
the purpose of this chapter!

Lesson 1 (Manipulate lists efficiently).

• Slicing lists.
– You already know mylist[a:b] that returns the sublist of elements from the rank a to the
rank b − 1.
– mylist[a:] returns the list of elements from rank a until the end.
– mylist[:b] returns the list of elements from the beginning to the rank b − 1.
– mylist[-1] returns the last element, mylist[-2] returns the penultimate element, . . .
– Exercise.
7 2 4 5 3 10 9 8 3
rank : 0 1 2 3 4 5 6 7 8

With mylist = [7,2,4,5,3,10,9,8,3], what do the following instructions return?

– mylist[3:5]
– mylist[4:]
– mylist[:6]
– mylist[-1]
• Find the rank of an element.
– mylist.index(element) returns the first position at which the item was found. Example:
with mylist = [12, 30, 5, 9, 5, 21], mylist.index(5) returns 2.
– If you just want to know if an item belongs to a list, then the statement :
element in mylist
returns True or False. Example: with mylist = [12, 30, 5, 9, 5, 21], “9 in
mylist” is true, while “8 in mylist” is false.
• List comprehension.
A set can be defined by listing all its elements, for example E = {0, 2, 4, 6, 8, 10}. Another way is to
say that the elements of the set must verify a certain property. For example, the same set E can be
defined by :
E = {x ∈ N | x 6 10 and x is even}.
With Python there is such a way to define lists. It is an extremely powerful and efficient syntax.
Let’s look at some examples:
– Let’s start from a list, for example mylist = [1,2,3,4,5,6,7,6,5,4,3,2,1].
LISTS II 90

– The command mylist_doubles = [ 2*x for x in mylist ] returns a list that con-
tains the double of each item of the mylist list. So this is the list [2,4,6,8,...].
– The command mylist_squares = [ x**2 for x in mylist ] returns the list of
squares of the items in the initial list. So this is the list [1,4,9,16,...].
– The command mylist_partial = [x for x in mylist if x > 2] extracts the list
composed only of elements strictly greater than 2. So this is the list [3,4,5,6,7,6,5,4,3].
• List of lists.
A list can contain other lists, for example:
mylist = [ ["Harry", "Hermione", "Ron"], [101,103] ]
contains two lists. We will be interested in lists that contain lists of integers, which we will call
arrays. For example:
array = [ [2,14,5], [3,5,7], [15,19,4], [8,6,5] ]
Then array[i] returns the sublist of rank i, while array[i][j] returns the integer located at
the rank j in the sublist of rank i. For example:
– array[0] returns the list [2,14,5],
– array[1] returns the list [3,5,7],
– array[0][0] returns the integer 2,
– array[0][1] returns the integer 14,
– array[2][1] returns the integer 19.

Activity 1 (Lists comprehension).

Goal: practice list comprehension. In this activity the lists are lists of integers.

1. Program a function multiplication(mylist,k) that multiplies each item in the list by k. For
example, multiplication([1,2,3,4,5],2) returns [2,4,6,8,10].
2. Program a function power(mylist,k) that raises each element of the list to the power k. For
example, power([1,2,3,4,5],3) returns [1,8,27,64,125].
3. Program a function addition(mylist1,mylist2) that adds together the elements of two lists
of the same length. For example, addition([1,2,3],[4,5,6]) returns [5,7,9].
Hint. This one is an example of a task where lists comprehension is not used!
4. Program a function non_zero(mylist) that returns a list of all non-zero elements. For example,
non_zero([1,0,2,3,0,4,5,0]) returns [1,2,3,4,5].
5. Program a function even(mylist) that returns a list of all even elements. For example,
even([1,0,2,3,0,4,5,0]) returns [0,2,0,4,0].

Activity 2 (Reach a fixed amount).

Goal: try to reach the total of 100 in a list of numbers.
We consider a list of n integers between 1 and 99 (included). For example the list of 25 integers:
[16,2,85,27,9,45,98,73,12,26,46,25,26,49,18,99,10,86,7,42]
which was obtained at random by the command:
mylist_20 = [randint(1,99) for i in range(20)]
We are looking for different ways to find numbers from the list whose sum is exactly 100.
1. Program a function sum_twoinarow_100(mylist) that tests if there are two consecutive ele-
ments in the list whose sum is exactly 100. The function returns “True” or “False” (but it can also
LISTS II 91

display numbers and their position for verification). For the example given the function returns
False.
2. Program a function sum_two_100(mylist) that tests if there are two items in the list, located at
different positions, whose sum is equal to 100. For the example given the function returns True
and can display the integers 2 and 98 (at ranks 1 and 6 of the list).
3. Program a function sum_seq_100(mylist) that tests if there are consecutive elements in the
list whose sum is equal to 100. For the example given the function returns True and can display
the sequence of integers 25, 26, 49 (at ranks 11, 12 and 13).
4. (Optional.) The larger the size of the list, the more likely it is to get the total of 100. For each of
the three previous situations, determines from which size n of the list, the probability of obtaining
a sum of 100 is greater than 1/2.
Hints. For each case, you get an estimate of this integer n, by writing a function proba(n,N) that
performs a large number N of random draws of lists having n items (with for example N = 10 000).
The probability is approximated by the number of favorable cases (where the function returns
true) divided by the total number of cases (here N ).

Activity 3 (Arrays).

Goal: working with lists of lists.

In this activity we work with arrays of size n × n containing integers. The object array is therefore a list
of n lists, each having n elements.
For example (with n = 3):
array = [ [1,2,3], [4,5,6], [7,8,9] ]
represents the array:
1 2 3
4 5 6
7 8 9
1. Write a function sum_diagonal(array) that calculates the sum of the elements located on the
main diagonal of an array. The main diagonal of the example given is 1, 5, 9, so the sum is 15.
2. Write a function sum_antidiagonal(array) that calculates the sum of the elements located on
the other diagonal. The anti-diagonal of the example given is composed of 3, 5, 7, the sum is still
15.
3. Write a function sum_all(array) that calculates the total sum of all elements. For the example
the total sum is 45.
4. Write a function print_array(array) that displays an array properly on the screen. You can
use the command:
print('{:>3d}'.format(array[i][j]), end="")
Explanations.
• The command print(string,end="") allows you to display a string of characters without
going to the next line.
• The command '{:>3d}'.format(k) displays the integer k on three characters (even if
there is only one digit to display).

Activity 4 (Magic Squares).

Goal: build magic squares as big as you want! You must first have done the previous activity.
LISTS II 92

A magic square is a square array of size n × n that contains all integers from 1 to n2 and satisfies that: the
sum of each row, the sum of each column, the sum of the main diagonal and the sum of the anti-diagonal
all have the same value.
Here is an example of a magic square with a size of 3 × 3 and one of size 4 × 4.

1 14 15 4 34
4 9 2 15
7 9 6 12 34
3 5 7 15
10 8 11 5 34
8 1 6 15
16 3 2 13 34

15 15
15 15 15
34 34
34 34 34 34

For a magic square of size n × n, the value of the sum is:

n(n2 + 1)
Sn = .
2
1. Examples. Define an array for each of the above examples 3 × 3 and 4 × 4 and display them on
the screen (use the previous activity).
2. To be or not to be. Define a is_magic_square(square) function that tests whether a given
array is (or not) a magic square (use the previous activity for diagonals).
3. Random squares. (Optional.) Randomly generate squares containing integers from 1 to n2 using
a function random_square(n). Experimentally verify that it is rare to obtain a magic square in
this way!
Hints. For a list mylist, the command shuffle(mylist) (from the random module) randomly
mixes the list (the list is modified in place).
The purpose of the remaining questions is to create large magic squares.
4. Addition. Define a function addition_square(square,k) which adds an integer k to
all the elements of the array. With the example of the square 3 × 3, the command
addition_square(square,-1) subtracts 1 from all the elements and returns an array that
would look like this:
3 8 1
2 4 6
7 0 5
Hints. To define a new square, start by filling it with 0:
new_square = [[0 for j in range(n)] for i in range(n)]
then fill it with the correct values by commands of the type:
new_square[i][j] = ...
5. Multiplication. Define a function multiplication_square(square,k) which multiplies
all the elements of the array by k. With the example of the square 3 × 3, the command
multiplication_square(square,2) multiplies all the elements by 2 and thus returns an
array that would be displayed as follows:
8 18 4
6 10 14
16 2 12
LISTS II 93

6. Homothety. Define a function homothety_square(square,k) which enlarges the array by a

factor of k as shown in the examples below. Here is an example of the 3×3 square with a homothety
ratio of k = 3.
4 4 4 9 9 9 2 2 2
4 4 4 9 9 9 2 2 2
4 4 4 9 9 9 2 2 2
4 9 2 3 3 3 5 5 5 7 7 7
3 5 7 −→ 3 3 3 5 5 5 7 7 7
8 1 6 3 3 3 5 5 5 7 7 7
8 8 8 1 1 1 6 6 6
8 8 8 1 1 1 6 6 6
8 8 8 1 1 1 6 6 6
Here is an example of a square 4 × 4 with a homothety ratio of k = 2.
1 1 14 14 15 15 4 4
1 1 14 14 15 15 4 4
1 14 15 4 7 7 9 9 6 6 12 12
7 9 6 12 7 7 9 9 6 6 12 12
−→
10 8 11 5 10 10 8 8 11 11 5 5
16 3 2 13 10 10 8 8 11 11 5 5
16 16 3 3 2 2 13 13
16 16 3 3 2 2 13 13
7. Block addition. Define a block_addition_square(big_square,small_square) function
that adds a small array of size n × n to the large array of size nm × nm per block as shown in the
example below with n = 2 and m = 3 (hence nm = 6). The small square 2 × 2 on the left is added
to the large square in the center to give the result on the right. For this addition the large square is
divided into 9 blocks, there is a total of 36 additions.
4 4 9 9 2 2 5 6 10 11 3 4
4 4 9 9 2 2 7 8 12 13 5 6
1 2 3 3 5 5 7 7 4 5 6 7 8 9
−→
3 4 3 3 5 5 7 7 6 7 8 9 10 11
8 8 1 1 6 6 9 10 2 3 7 8
8 8 1 1 6 6 11 12 4 5 9 10
8. Products of magic squares. Define a function product_squares(square1,square2) which
from two magic squares, calculates a large magic square called the product of the two squares.
The algorithm is as follows:
LISTS II 94

Algorithm.
• – Inputs: a magic square C1 of size n × n and a magic square C2 of size m × m.
– Output: a magic square C of size (nm) × (nm).
Set the square C3a by subtracting 1 from all elements of C2 . (Use the command
addition_square(square2,-1).)
• Define the square C3b as the homothety of the square C3a of ratio n. (Use the command
homothety(square3a,n).)
• Define the square C3c by multiplying all the terms of the square C3b by n2 . (Use the
command multiplication_square(square3b,n**2).)
• Define the square C3d by adding the square C1 to the square C3c per block. (Use the
command block_addition_square(square3c,square1).)
• Return the square C3d .

• Implement this algorithm.

• Test it on examples, checking that the square obtained is indeed a magic square.
• Build a magic square of size 36 × 36.
• Also checks that the order of the product is important: C1 × C2 is not the same square as
C2 × C1 .
 Chapter

Binary II
13

We continue our exploration of the world of 0 and 1.

Activity 1 (Palindromes).

Goal: find palindromes in binary and decimal numeral system.

In English a palindrome is a word (or a sentence) that can be read in both directions, for example
“RADAR” or “A MAN, A PLAN, A CANAL: PANAMA”. In this activity, a palindrome will be a list, which
has the same elements when browsing from left to right or right to left.
Examples:
• [1,0,1,0,1] is a palindrome (with binary numeral system),
• [2,9,4,4,9,2] is a palindrome (with decimal numeral system).
1. Program a function is_palindrome(mylist) that tests if a list is a palindrome or not.
Hints. You can compare the items at ranks i and p − 1 − i or use list(reversed(liste)).
2. We are looking for integers n such that their binary numeral system is a palindrome. For example,
the binary notation of n = 27 is the palindrome [1,1,0,1,1]. This is the tenth integer n having
this property.
What is the thousandth integer n > 0 whose binary notation is a palindrome?
3. What is the thousandth integer n > 0 whose decimal notation is a palindrome?
For example, the digits of n = 909 in decimal notation, form the palindrome [9,0,9]. This is the
hundredth integer n having this property.
4. An integer n is a bi-palindrome if its binary notation and its decimal notation are palindromes.
For example n = 585 has a decimal notation which is a palindrome and also is its binary notation
[1,0,0,1,0,0,1,0,0,1]. This is the tenth integer n having this property.
What is the twentieth integer n > 0 to be a bi-palindrome?

Lesson 1 (Logical operations).

We consider that 0 represents “False” and 1 is “True”.
• With the logical operation “OR”, the result is true as soon as at least one of the two terms is true.
This is written:
– 0 OR 0 = 0
– 0 OR 1 = 1
– 1 OR 0 = 1
– 1 OR 1 = 1
• With the logical operation “AND”, the result is true only when both terms are true. This is written:
BINARY II 96

– 0 AND 0 = 0
– 0 AND 1 = 0
– 1 AND 0 = 0
– 1 AND 1 = 1
• The logical operation “NOT”, exchange true and false values:
– NOT 0 = 1
– NOT 1 = 0
• For numbers in binary notation, these operations range from bits to bits, i.e. digit by digit (starting
with the digits on the right) as one would add (without carry).
For example:

1.0.1.1.0 1.0.0.1.0
AND 1.1.0.1.0 OR 0.0.1.1.0

1.0.0.1.0 1.0.1.1.0

If the two systems do not have the same number of bits, we add non-significant 0 on the left
(example of 1.0.0.1.0 OR 1.1.0 on the figure at the right).

Activity 2 (Logical operations).

Goal: program the main logical operations.

1. (a) Program a function NOT() which corresponds to the negation for a given list. For example,
NOT([1,1,0,1]) returns [0,0,1,0].
(b) Program a function OReq() which corresponds to “OR” with two lists of equal length. For
example, with mylist1 = [1,0,1,0,1,0,1] and mylist2 = [1,0,0,1,0,0,1], the
function returns [1,0,1,1,1,0,1].
(c) Do the same work with ANDeq() for two lists having the same length.
2. Write a function zero_padding(mylist,p) that adds zeros at the beginning of the list to
get a list of length p. Example: if mylist = [1,0,1,1] and p = 8, then the function returns
[0,0,0,0,1,0,1,1].
3. Write two functions OR() and AND() which correspond to the logical operations, but with two
lists that do not necessarily have the same length.
Example:
• mylist1 = [1,1,1,0,1] and mylist2 = [1,1,0],
• it should be considered that mylist2 is equivalent to the list mylist2bis = [0,0,1,1,0]
of the same length as mylist1,
• so OR(mylist1,mylist2) returns [1,1,1,1,1],
• then AND(mylist1,mylist2) returns [0,0,1,0,0] (or [1,0,0] depending on your
choice).
Hints. You can take over the content of your functions OReq and ANDeq, or you can first add zeros
to the shortest list.
BINARY II 97

Activity 3 (De Morgan’s laws).

Goal: generate all possible lists of 0 and 1 to check a proposition.

1. First method: use binary notation.

We want to generate all possible lists of 0 and 1 of a given size p. Here’s how to do it:
• An integer n runs all integers from 0 to 2 p − 1.
• For each of these integers n, we calculate its binary notation (in the form of a list).
• We add (if necessary) 0 at the beginning of the list, in order to get a list of length p.
Program this method.
Example: for n = 36, its binary notation is [1,0,0,1,0,0]. If you want a list of p = 8 bits, you
add two 0: [0,0,1,0,0,1,0,0].
2. Second method (optional): a recursive algorithm.
We want to generate again all the possible lists of 0 and 1 of a given size. We adopt the following
procedure: if we know how to find all the lists of size p − 1, then to obtain all the lists of size p, we
just have to add one 0 at the beginning of each list of size p − 1, then to start again by adding one
1 at the beginning of each list of size p − 1.
For example, there are 4 lists of length 2: [0, 0], [0, 1], [1, 0], [1, 1]. I deduct the 8 lists
of length 3:
• 4 lists by adding 0 at the front: [0, 0, 0], [0, 0, 1], [0, 1, 0], [0, 1, 1],
• 4 lists by adding 1 at the front: [1, 0, 0], [1, 0, 1], [1, 1, 0], [1, 1, 1].
This gives the following algorithm, which is a recursive algorithm (because the function calls itself).

Algorithm.
Use: every_binary_number(p)
Input: an integer p > 0
Output: the list of all possible lists of 0 and 1 of length p
• If p = 1 return the list [ [0], [1] ].
• If p > 2, then:
– get all lists of size p-1 by the call every_binary_number(p-1)
– for each item in this list, build two new items:
– on the one hand add 0 at the beginning of this element;
– on the other hand add 1 at the beginning of this element;
– then add these two items to the list of lists of size p.
• Return the list of all the lists with a size p.

3. De Morgan’s laws.
De Morgan’s laws state that for booleans (true/false) or bits (1/0), we always have these properties:
NOT(b1 OR b2 ) = NOT(b1 ) AND NOT(b2 ) NOT(b1 AND b2 ) = NOT(b1 ) OR NOT(b2 ).
Experimentally checks that these equations are still true for any list `1 and `2 of exactly 8 bits.
PA R T I V

PROJECTS

98
 Chapter
Probabilities – Par-
rondo’s paradox
14

You will program two simple games. When you play these games, you are more likely to lose than to
win. However, when you play both games at the same time, you have a better chance of winning
than losing! It’s a paradoxical situation.

Lesson 1 (Random – Winning – Expected value).

A player plays a game of chance: he throws a coin; depending on the result he wins or loses money. We
cannot predict how much the player will win or lose for sure, but we will introduce a value that estimates
how much the player can expect to win on average if he plays many times.
• At the start the player’s total winnings are zero: g = 0.
• Each time he draws, he throws a coin. If he wins, he gets one dollar, if he loses he owes one dollar.
• In the games we study, the coin is not balanced (it is a little rigged). The player does not have as
many chances to win as to lose.
• Draws are repeated a number of times N . After N draws, the player’s winnings (which can be
positive or negative) are totaled.
• The expected value, is the amount that the player can expect to win with each throw. The expected
value is estimated by averaging the earnings of a large number of draws. In other words, we have
the formula:
winnings afters N throws
expected value ' with large N .
N
For our games, the expected value will be a real number between −1 and +1.
• Examples:
– If the coin is well balanced (50 chances out of 100 to win), then for a large number of N
draws, the player will win about as many times as he loses; his winnings will be close to 0
and so the expected value will be close to N0 = 0. On average he earns 0 dollar per draw.
– If the coin is faked and the player wins all the time, then after N draws, he has pocketed N
dollars. The expected value is therefore NN = 1.
– If the coin is faked so that the player loses all the time, then after N draws, his win is −N
dollars. The expected value is therefore −NN = −1.
– An expected value of −0.5 means that on average the player loses 0.5 dollar per draw. This
is possible with an unbalanced coin that wins in only one out of four cases. (Check the
calculation!) If the player plays 1000 times, we can estimate that he will lose 500 dollars
(−0.5 × 1000 = −500).
PROBABILITIES – PARRONDO’S PARADOX 100

Activity 1 (Game A: first losing game).

Goal: model a first simple game, which on average makes the player loses.

Game A. In this first game, we throw a slightly unbalanced coin: the player wins one dollar in 49 cases
on 100; he loses one dollar in 51 cases on 100.
1. One draw. Write a function throw_game_A() that does not depend on any argument and that
models a draw of the game A. For this purpose:
• Pick a random number 0 6 x < 1 using the random() function of the random module.
• Returns +1 if x is smaller than 0.49; and −1 otherwise.
2. Gain. Write a function gain_game_A(N) that models N draws from game A and returns the total
winnings of these draws. Of course, the result depends on the actual runs, it can vary from one
time to another.
3. Expected value. Write a function expected_value_game_A(N) that returns an estimate of the
expected value of the game A according to the formula:
winnings afters N throws
expected value ' with large N .
N
4. Conclusion.
(a) Estimate the expected value by making at least one million draws.
(b) What does it mean that the expected value is negative?
(c) Deduct from the value of the expected value, the win (or loss) that I can expect by playing
1000 times in game A.

Activity 2 (Game B: second losing game).

Goal: model a second game that is a little more complicated which on average still makes the
player loses.
Game B. The second game is a little more complicated. At the beginning, the player starts with a zero
win: g = 0. Then, depending on this gain, he plays one of the following two subgames:
• Subgame B1. If the gain g is a multiple of 3, then he throws a very disadvantageous coin: the
player wins a dollar in only 9 cases on 100 (so he loses a dollar in 91 cases on 100).
• Subgame B2. If the gain g is not a multiple of 3, then he throws an advantageous coin: the player
wins a dollar in 74 cases on 100 (so he loses a dollar in 26 cases on 100).
1. One draw. Write a throw_game_B(g) function that depends on the win already acquired and
models a draw from game B. You can use the g%3 == 0 test to find out if g is a multiple of 3.
2. Gain. Write a function gain_game_B(N) that models N draws playing game B (starting from a
zero initial gain) and returns the total gain of these draws.
3. Expected value. Write a function expected_value_game_B(N) that returns an estimate of the
expected value of the game B.
4. Conclusion.
(a) Estimates the expected value by making at least one million draws.
(b) How much can I expect to win or lose by playing 1000 times in game B?

Activity 3 (Games A and B: a winning game!).

Goal: invent a new game by playing at each round game A or game B; strangely enough this game
makes you win on average! That’s Parrondo’s paradox.
Games AB. In this third game, we play each round either game A or game B (the choice is made at
PROBABILITIES – PARRONDO’S PARADOX 101

random). At the beginning the player starts with a zero win: g = 0. At each step, he chooses at random
(50% of luck each):
• to play game A once,
• or to play game B once; more precisely with subgame B1 or subgame B2 depending on the win
already acquired g.
1. One draw. Write a throw_game_AB(g) function that depends on the win already acquired and
models a draw of the AB game.
2. Gain. Write a function gain_game_AB(N) that models N draws from the AB game (starting from
a zero initial gain) and returns the total gain of these draws.
3. Expected value. Write a function expected_value_game_AB(N) that returns an estimate of the
expected value of the AB game.
4. Conclusion.
(a) Estimate the expected value by performing at least one million game turns.
(b) What can we say this time about the sign of expected value?
(c) How much can I expect to win or lose by playing AB game 1000 times? Surprising, isn’t it?

Reference: “Parrondo’s paradox”, Hélène Davaux, La gazette des mathématiciens, July 2017.
 Chapter

Find and replace

15

Finding and replacing are two very frequent tasks. Knowing how to use them and how they work
will help you to be more effective.

Activity 1 (Find).

Goal: learn different ways to search with Python.

1. The operator “in”.

The easiest way to know if a substring is present in a string is to use the operator “in”. For example,
the expression:
"NOT" in "TO BE OR NOT TO BE"
is equal to “True” because the substring NOT is present in the sentence.
Deduce a function find_in(string,substring) that returns “True” or “False”, depending on
whether the substring is (or is not) present in the chain.
2. The method find().
The find() method of Python is used as string.find(substring) and returns the position
to which the substring was found.
Test this on the previous example. What does the function return if the substring is not found?
3. The method index().
The index() method has the same utility, it is used in the form string.index(substring)
and returns the position to which the substring was found.
Test this on the previous example. What does the function return if the substring is not found?
4. Your function find().
Write your own function myfind(string,substring) which returns the starting position of
the substring if it is found (and returns None if it is not).
You are not allowed to use some functions of Python, you only have the right to test if two
characters are equal.

Activity 2 (Replace).

Goal: replace portions of text with others.

1. The replace() method is used in the form:

string.replace(substring,new_substring)
Each time the sequence substring is found in string, it is replaced by new_substring.
Transform the sentence TO BE OR NOT TO BE into TO BE AND NOT TO BE, then into TO HAVE
AND NOT TO HAVE.
FIND AND REPLACE 103

2. Write your own function myreplace() which you will call in the following form:
myreplace(string,substring,new_substring)
and which only replaces the first occurrence of substring found. For example,
myreplace("ABBA","B","XY") returns "AXYBA".
Hint. You can use your myfind() function from the previous activity to find the starting position
of the sequence to replace.
3. Improve your function to build a function replace_all() which now replaces all occurrences
encountered.

Lesson 1 (Regular expressions regex).

The regular expressions allow you to search for substrings with greater freedom: for example, you can
allow a wildcard character or several possible choices for a character. There are many other possibilities,
but we are only studying these two.
1. We allow ourselves a joker letter symbolized by a point “.”. For example, if we look for the expression
“P . R” then:
• PORK, EMPIRE, PURE, REPORT contain this group (for example for PORK the point plays
the role of O),
• but not the words CAR, POOR, RAP, PRICE.
2. We are still looking for groups of letters, we now allow ourselves several options. For example
“[CT]” means “C or T”. Thus the letter group “[CT]O” corresponds to the letter group “CO” or “TO”.
This group is therefore contained in TOTEM, COST, ACTOR but not in BLOCK nor VOTE. Similarly
“[ABC]” would mean “A or B or C”.

We will use regular expressions through a command:

python_regex_find(string,exp)
whose function is defined below.
from re import *

def python_regex_find(string,exp):
pattern = search(exp,string)
if pattern:
return pattern.group(), pattern.start(), pattern.end()
else:
return None
Program it and test it. It returns: (1) the found substring, (2) the start position and (3) the end position.
FIND AND REPLACE 104

python : re.search() - python_regex_find()

Use: search(exp,string)
or python_regex_find(string,exp)
Input: a string string and a regular expression exp
Output: the result of the search (the substring found, its start position, its end
position)
Example with string = "TO BE OR NOT TO BE"
• with exp = "N.T", then python_regex_find(string, exp) re-
turns ('NOT', 9, 12).
• with exp = "B..O", the function returns ('BE O', 3, 7) (the
space counts as a character).
• with exp = "[NM]O", the function returns ('NO', 9, 11).
• with exp = "[BC]..O[RS]", the function returns ('BE OR', 3,
8).

Activity 3 (Regular expressions regex).

Goal: program the search for simple regular expressions.

1. Program your function regex_find_wildcard(string,exp) who is looking for a substring

that can contain one or more wildcards “.”. The function must return: (1) the found substring, (2)
the start position and (3) the end position (as for the function python_regex_find() above).
2. Program your function regex_find_choice(string,exp) who is looking for a substring that
can contain one or more choices contained in tags “[]”. The function must return again: (1) the
found substring, (2) the start position and (3) the end position.
Hint. You can start by writing a function all_choices(exp) that generates all possibilities from
exp. For example, if exp = "[AB]X[CD]Y" then all_choices(exp) returns the list formed of:
"AXCY", "BXCY", "AXDY" and "BXDY".

Lesson 2 (Replace 0 and 1 and start again!).

We consider a “sentence” composed of only two possible letters 0 and 1. In this sentence we will search
for a pattern (a substring) and replace it with another one.
FIND AND REPLACE 105

Example.
Apply the transformation 01 → 10 to the sentence 10110.
We read the sentence from left to right, we find the first pattern 01 from the second letter, we replace
it with 10:
1(01)10 7−→ 1(10)10
We can start again from the beginning of the sentence obtained, with always the same transformation
01 → 10:
11(01)0 7−→ 11(10)0
The pattern 01 no longer appears in the sentence 11100 so the transformation 01 →10 now leaves
this sentence unchanged.
Let’s summarize: here is the effect of the iterated transformation 01 → 10 in the sentence 10110:
10110 7−→ 11010 7−→ 11100

Example.
Apply the transformation 001 →1100 to the sentence 0011.
A first time:
(001)1 7−→ (1100)1
A second time:
11(001) 7−→ 11(1100)
And then the transformation no longer modifies the sentence.

Example.
Let’s see a last example with the transformation 01 →1100 for the starting sentence 0001:
0001 7−→ 001100 7−→ 01100100 7−→ 1100100100 7−→ · · ·
We can iterate the transformation, to obtain longer and longer sentences.

Activity 4 (Replacement iterations).

Goal: study some transformations and their iterations.

We consider here only transformations of the type 0a 1 b →1c 0d , i.e. a pattern with first 0’s then 1’s is
replaced by a pattern with first 1’s then 0’s.
1. One iteration.
Using your myreplace() function from the first activity, check the above examples. Make sure
you replace only one pattern at each step (the leftmost one).
Example: the transformation 01 → 10 applied to the sentence 101, is calculated by
myreplace("101","01","10") and returns "110".
2. Multiple iterations.
Program a function iterations(sentence,pattern,new_pattern) that, from a sentence,
iterates the transformation. Once the sentence is stabilized, the function returns the number of
iterations performed and the resulting sentence. If the number of iterations does not seem to stop
(for example when it exceeds 1000) then returns None.
Example. For the transformation 0011 → 1100 and the sentence 00001101, the sentences obtained
FIND AND REPLACE 106

are:
000011011 7−→ 001100011 7−→ 110000011 7−→ 110001100 7−→ 110110000 7−→ ···
1 2 3 4

For this example, the call to the iterations() function returns 4 (the number of transformations
before stabilization) and "110110000" (the stabilized sentence).
3. The most iterations possible.
Program a function max_iterations(p,pattern,new_pattern) which, among all the sen-
tences of length p, is looking for one of those that takes the longest to stabilize. This function
returns:
• the maximum number of iterations,
• a sentence that achieves this maximum,
• and the corresponding stabilized sentence.
Example: for the transformation 01 → 100, among all the sentences of length p = 4, the maximum
number of possible iterations is 7. Such an example of a sentence is 0111, which will stabilize (after
7 iterations) in 11100000000. So the command max_iteration(4,"01","100") returns:
7, '0111', '11100000000'
Hint. To generate all sentences with a length of p formed of 0 and 1, you can consult the chapter
“Binary II” (activity 3).
4. Categories of transformations.
• Linear transformation. Experimentally check that the transformation 0011 →110 is linear,
i.e. for all sentences with a length of p, there will be at most about p iterations before
stabilization. For example, for p = 10, what is the maximum number of iterations?
• Quadratic transformation. Experimentally check that the transformation 01 →10 is
quadratic, i.e. for all sentences with a length of p, there will be at most about p2 iterations
before stabilization. For example, for p = 10, what is the maximum number of iterations?
• Exponential transformation. Experimentally check that the transformation 01 →110 is
exponential, i.e. for all sentences with a length of p, there will be a finite number of iterations,
but that this number can be very large (much larger than p2 ) before stabilization. For example,
for p = 10, what is the maximum number of iterations?
• Transformation without end. Experimentally verify that for the transformation 01 →1100,
there are sentences that will never stabilize.
 Chapter

16
Polish calculator – Stacks

You’re going to program your own calculator! For that you will discover a new notation for formulas
and also discover what a “stack” is in computer science.

Lesson 1 (Stack).
A stack is a sequence of data with three basic operations:
• push: you add an element to the top of the stack,
• pop: the value of the element at the top of the stack is read and this element is removed from the
stack,
• and finally, we can test if the stack is empty.

"Because"
Top of the stack
"Unless"
"If"
6 "While"
11 "Still"
3 "As"
7 "When"

A stack of integers A stack of strings An empty stack

b
pop()
J push(J) return the value (here "b")
J b
A A k k
Z Z s s
U Push U x Pop x

Remarks.
• Analogy. You can make the connection with a stack of plates. You can put plates on a stack one by
one. You can remove the plates one by one, starting of course with the top one. In addition, it must
be considered that on each plate is drawn a data (a number, a character, a string . . . ).
• Last in, first out. In a queue, the first one to wait is the first one to be served and comes out. Here
it’s the opposite! A stack works on the principle of “last in, first out”.
• In a list, you can directly access any element; in a stack, you can only directly access the element at
the top of the stack. To access the other elements, you have to pop several times.
POLISH CALCULATOR – STACKS 108

• The advantage of a stack is that it is a very simple data structure that corresponds well to what
happens in a computer’s memory.

Lesson 2 (Global variable).

A global variable is a variable that is defined for the entire program. It is generally not recommended to
use such variables but it may be useful in some cases. Let us look at an example.
The global variable, here the gravitational constant, is declared at the beginning of the program as a
classic variable:
gravitation = 9.81
The content of the variable gravitation is now available everywhere. On the other hand, if you want
to change the value of this variable in a function, you must specify to Python that you are aware of
modifying a global variable.
For example, for calculations on the Moon, it is necessary to change the gravitational constant, which is
much lower there.
def on_the_moon():
global gravitation # Yes, I really want to modify this variable!
gravitation = 1.625 # New value for the entire program
...

Activity 1 (Stack basic operations).

Goal: define the three (very simple) commands to use the stacks.

In this chapter, a stack will be modeled by a list. The item at the end of the list is the top of the stack.

7
11
2
stack = [4,2,11,7]
4

A stack Model as a list

The stack will be stored in a global variable stack. It is necessary to start each function that modifies
the stack with the command:
global stack
1. Write a function push_to_stack() that adds an element to the top of the stack.

push_to_stack()
Use: push_to_stack(item)
Input: an integer, a string. . .
Output: nothing
Action: the stack contains an additional element
Example: if at the beginning stack = [5,1,3] then, after the instruction
push_to_stack(8), the stack is [5,1,3,8] and if you continue with the
instruction push_to_stack(6), the stack is now [5,1,3,8,6].
POLISH CALCULATOR – STACKS 109

2. Write a function pop_from_stack(), without parameters, that removes the element at the top
of the stack and returns its value.

pop_from_stack()
Use: pop_from_stack()
Input: nothing
Output: the element at the top of the stack
Action: the stack contains one less element
Example: if initially stack = [13,4,9] then the instruction
pop_from_stack() returns the value 9 and the stack is now [13,4]; if
you execute a new instruction pop_from_stack(), it returns this time the
value 4 and the stack is now [13].

3. Write a function is_stack_empty(), without parameter, that tests if the stack is empty or not.

is_stack_empty()
Use: is_stack_empty()
Input: nothing
Output: true or false
Action: does nothing on the stack
Example:
• if stack = [13,4,9] then the instruction is_stack_empty() re-
turns False,
• if stack = [] then the instruction is_stack_empty() returns
True.

Activity 2 (Operations on a stack).

Goal: handle the stacks using only the three functions push_to_stack(), pop_from_stack()
and is_stack_empty().
In this exercise, we work with a stack of integers. The questions are independent.
1. (a) Starting from an empty stack, arrive at a stack [5,7,2,4].
(b) Then execute the instructions pop_from_stack(), push_to_stack(8),
push_to_stack(1), push_to_stack(1), push_to_stack(3). What is the stack
now? What does the instruction pop_from_stack() now return?
2. Start from a stack. Write a function is_in_stack(item) that tests if the stack contains a given
element.
3. Start from a stack. Write a function that calculates the sum of the elements of the stack.
4. Start from a stack. Write a function that returns the second last element of the stack (the last
element is the one at the very bottom; if this second last element does not exist, the function
returns None).

Lesson 3 (String manipulation).

1. The function split() is a Python method that separates a string into pieces. If no separator is
POLISH CALCULATOR – STACKS 110

specified, the separator is the space character.

python: split()
Use: string.split(separator)
Input: a string string and possibly a separator separator
Output: a list of strings
Example :
• "To be or not to be.".split() returns ['To', 'be',
'or', 'not', 'to', 'be.']
• "12.5;17.5;18".split(";") returns ['12.5', '17.5', '18']

2. The function join() is a Python method that gathers a list of strings into a single string. This is
the opposite of split().
python: join()
Use: separator.join(mylist)
Input: a list of strings mylist and a separator separator
Output: a string
Example :
• "".join(["To", "be", "or", "not", "to", "be."]) returns
'Tobeornottobe.' Spaces are missing.
• " ".join(["To", "be", "or", "not", "to", "be."]) re-
turns 'To be or not to be.' It’s better when the separator is a
space.
• "--".join(["To", "be", "or", "not", "to", "be."])
returns 'To--be--or--not--to--be.'

3. The function isdigit() is a Python method that tests if a string contains only numbers. This allows
to test if a string corresponds to a positive integer. Here are some examples: "1776".isdigit()
returns True; "Hello".isdigit() returns False.
Remember that you can convert a string into an integer by the command int(string). The
following small program tests if a string can be converted into a positive integer:
mystring = "1776" # A string
if mystring.isdigit():
myinteger = int(mystring) # myinteger is an integer
else: # Problem
print("I don't know how to convert this string to an integer!")

Activity 3 (Sorting station).

Goal: solve a sorting problem by modeling a storage area by a stack.
A train has blue wagons with a number and red wagons with a letter.

7 W 13 G S 5

Unsorted train.
POLISH CALCULATOR – STACKS 111

The stationmaster wants to separate the wagons: first all the blues and then all the reds (the order of the
blue wagons does not matter, the order of the red wagons either).

7 5 13 W S G

Sorted train.

For this purpose, there is an output station and a waiting area: a wagon can either be sent directly to the
output station or temporarily stored in the waiting area.
7 W 13 G S 5

Output station Input station

Waiting area

Here are the instructions from the stationmaster.

• Phase 1. For each wagon in the train:
– if it is a blue wagon, send it directly to the output station;
– if it is a red wagon, send it to the waiting area.
• Phase 2. Then, move the (red) wagons one by one from the waiting area to the output station by
attaching them to the blue ones.
7 13 G S 5

Output station Input station

W

Phase 1

Waiting area

7 13 5

Phase 2
S
G
W
POLISH CALCULATOR – STACKS 112

Here is how we will model the train and its waiting area.
• The train is a string of characters made up of a series of numbers (blue wagons) and letters (red
wagons) separated by spaces. For example train = "G 6 Z J 14".
• The list of wagons is obtained by asking train.split().
• We test if a wagon is blue and check if it is marked with a number, by the function
wagon.isdigit().
• The train reconstructed by the sorted wagons is also a string of characters. At first, it is the empty
string.
• The waiting area will be the stack. At the beginning the stack is empty. We’re only going to add the
red wagons. At the end, the stack is drained towards the tail end of the reconstituted train.
Following the station manager’s instructions and using stack operations, write a sort_wagons() function
that separates the blue and red wagons from a train.

sort_wagons()
Use: sort_wagons(train)
Input: a string with blue wagons (numbers) and red wagons (letters)
Output: blue wagons first and red wagons second.
Action: use a stack
Example :
• sort_wagons("A 4 C 12") returns "4 12 C A"
• sort_wagons("K 8 P 17 L B R 3 10 2 N") returns "8 17 3
10 2 N R B L P K"

Lesson 4 (Polish notation).

Writing in Polish notation (of its real name, reverse Polish notation) is another way to write an
algebraic expression. Its advantage is that this notation does not use brackets and is easier to
handle for a computer. Its disadvantage is that we’re not used to it.
Here is the classic way to write an algebraic expression (left) and its Polish notation (right). In any case,
the result will be 13!
Classic: 7+6 Polish: 7 6 +
Other examples:
• Classic: (10 + 5) × 3 ; Polish: 10 5 + 3 ×
• Classic: 10 + 2 × 3 ; Polish: 10 2 3 × +
• Classic: (2 + 8) × (6 + 11) ; Polish: 2 8 + 6 11 + ×

Let’s see how to calculate the value of an expression in Polish notation.

• We read the expression from left to right:
2 8 + 6 11 + ×
−−−−−−−−−−−−−−−→

• When you meet a first operator (+, ×, . . . ) you calculate the operation with the two members just
before this operator:
2 8 +
| {z } 6 11 + ×
2+8
POLISH CALCULATOR – STACKS 113

• This operation is replaced by the result:

10
|{z} 6 11 + ×
result of
2+8
• We continue reading the expression (we are looking for the first operator and the two terms just
before):
10 6 {z +
| 11 } × becomes 10 17 × that is equal to 170
6+11=17
• At the end there is only one value left, it’s the result! (Here 170.)

Other examples:
• 8 2 ÷ 3 × 7 +
8 2 ÷
| {z } 3 × 7 + becomes 4 3 ×
| {z } 7 + becomes 12 7 + that is equal to 19
8÷2=4 4×3=12
• 11 9 4 3 + − ×
11 9 4 3 +
| {z } − × becomes 11 | 7 −
9 {z } × becomes 11 2 × that is equal to 22
4+3=7 9−7=2
Exercise. Calculates the value of expressions:
• 13 5 + 3 ×
• 3 5 7 × +
• 3 5 7 + ×
• 15 5 ÷ 4 12 + ×

Activity 4 (Polish calculator).

Goal: program a mini-calculator that evaluates expressions in Polish notations.

1. Write a function operation() that calculates the sum or product of two numbers.

operation()
Use: operation(a,b,op)
Input: two numbers a and b, one operation character "+" or "*"
Output: the result of the operation a + b or a * b
Example:
• operation(2,4,"+") returns 6
• operation(2,4,"*") returns 8

2. Program a Polish calculator, according to the following algorithm:

POLISH CALCULATOR – STACKS 114

Algorithm.
• – Input: an expression in Polish notation (a string).
– Output: the value of this expression.
– Example: "2 3 + 4 *" (the calculation (2 + 3) × 4) returns 20.
• Start with an empty stack.
• For each element of the expression (read from left to right):
– if the element is a number, then add this number to the stack,
– if the element is an operation character, then:
– pop the stack once to get a number b,
– pop a second time to get a number a,
– calculate a + b or a × b depending on the operation,
– push this result to the stack.
• At the end, the stack contains only one element, it is the result of the calculation.

polish_calculator()
Use: polish_calculator(expression)
Input: an expression in Polish notation (a string)
Output: the result of the calculation
Action: uses a stack
Example:
• polish_calculator("2 3 4 + +") returns 9
• polish_calculator("2 3 + 5 *") returns 25

Bonus. Change your code to support subtraction and division!

Activity 5 (Expression with balanced brackets).

Goal: determine if the parentheses of an expression are placed appropriately.

Here are some examples of well and bad balanced bracketed expressions:
• 2 + (3 + b) × (5 + (a − 4)) is well balanced parenthesized;
• (a + 8) × 3) + 4 is incorrectly bracketed: there is a closing bracket “)” alone;
• (b + 8/5)) + (4 is incorrectly bracketed: there are as many opening parentheses “(” than closing
parentheses “)” but they are badly positioned.
1. Here is the algorithm that decides if the parentheses of an expression are well placed. The stack
acts as an intermediate storage area for opening parenthesis "(". Each time we find a closing
parenthesis ")" in the expression we delete an opening parenthesis from the stack.
POLISH CALCULATOR – STACKS 115

Algorithm.
Input: any expression (a string).
Output: “True” if the parentheses are well balanced, “False” otherwise.
• Start with an empty stack.
• For each character of the expression read from left to right:
– if the character is neither "(", nor ")" then do nothing!
– if the character is an opening parenthesis "(" then add this character to the stack;
– if the character is a closing parenthesis ")":
– test if the stack is empty, if it is empty then return “False” (the program ends
there, the expression is incorrectly parenthesized), if the stack is not empty
continue,
– pop the stack once, it gives a "(".
• If at the end, the stack is empty then return the value “True”, otherwise return “False”.

are_parentheses_balanced()
Use: are_parentheses_balanced(expression)
Input: an expression (string)
Output: true or false depending on whether the parentheses are correctly
placed or not
Action: uses a stack
Example :
• are_parentheses_balanced("(2+3)*(4+(8/2))") returns
True
• are_parentheses_balanced("(x+y)*((7+z)") returns False

2. Enhance this function to test an expression with parentheses and square brackets. Here is a correct
expression: [(a + b) ∗ (a − b)], here are some incorrect expressions: [a + b), (a + b] ∗ [a − b).
Here is the algorithm to program a function are_brackets_balanced().
POLISH CALCULATOR – STACKS 116

Algorithm.
Input: an expression (a string).
Output: “True” if the parentheses and square brackets are well balanced, “False” otherwise.
• Start with an empty stack.
• For each character of the expression read from left to right:
– if the character is neither "(", nor ")", nor "[", nor "]" then do nothing;
– if the character is an opening parenthesis "(" or an opening square bracket "[",
then add this character to the stack;
– if the character is a closing parenthesis ")" or a closing square bracket "]", then:
– test if the stack is empty, if it is empty then return “False” (the program ends
there, the expression is not correct), if the stack is not empty continue,
– pop the stack once, you get a "(" or a "[",
– if the popped (opening) character does not match the character read in the
expression, then return “False”. The program ends there, the expression is not
consistent. To match means that the character "(" corresponds to ")" and "["
corresponds to "]".
• If at the end, the stack is empty then return the value “True”, otherwise return “False”.

This time the stack can contain opening parentheses "(" or opening square brackets "[". Each
time you find a closing parenthesis ")" in the expression, the top of the stack must be an opening
parenthesis "(". Each time you find a closing square bracket "]" in the expression, the top of the
stack must be a opening square bracket "[".

Activity 6 (Conversion to Polish notation).

Goal: transform a classic algebraic expression with parentheses into a Polish notation expression.
The algorithm is a much improved version of the previous activity. We will not give any justification.
You are used to writing “(13 + 5) × 7”; you have seen that the computer can easily calculate “13 5 + 7 ×”.
All that remains is to switch from classical algebraic expression (with parentheses) to Polish notation
(without parentheses)!
Here is the algorithm for expressions with only additions and multiplications.
POLISH CALCULATOR – STACKS 117

Algorithm.
Input: a classic expression
Output: an expression in Polish notation
• Start with an empty stack.
• Start with an empty string polish which at the end will contain the result.
• For each character of the expression (read from left to right):
– if the character is a number, then add this number to the output string polish;
– if the character is an opening parenthesis "(", then add this character to the stack;
– if the character is the multiplication operator "*", then add this character to the stack;
– if the character is the addition operator "+", then:
while the the stack is not empty:
pop an element from the stack,
if this element is the multiplication operator "*", then:
add this item to the output string polish
otherwise:
push this element to the stack (we put it back on the stack after removing it)
end the “while” loop immediately (with break)
finally, add the "+" addition operator to the stack.
– if the character is a closing parenthesis ")", then:
while the stack is not empty:
pop an element from the stack,
if this element is an opening parenthesis "(", then:
end the “while” loop immediately (with break)
otherwise:
add this item to the output string polish

• If at the end, the stack is not empty, then add each element of the stack to the output string
polish.

polish_notation()
Use: polish_notation(expression)
Input: a classic expression (with elements separated by spaces)
Output: the expression in Polish notation
Action: use the stack
Example:
• polish_notation("2 + 3") returns "2 3 +"
• polish_notation("4 * ( 2 + 3 )") returns "4 2 3 + *"
• polish_notation("( 2 + 3 ) * ( 4 + 8 )") returns "2 3
+ 4 8 + *"

In this algorithm, each element between two spaces of an expression is miscalled “character”. Example:
the characters of "( 17 + 10 ) * 3" are (, 17, +, 10, ), * and 3.
POLISH CALCULATOR – STACKS 118

You see that the addition has a more complicated treatment than the multiplication. This is due to the
fact that multiplication takes priority over addition. For example, 2 + 3 × 5 means 2 + (3 × 5) and not
(2 + 3) × 5. If you want to take into account subtraction and division, you have to be careful about
non-commutativity (a − b is not equal to b − a; a ÷ b is not equal to b ÷ a).

End this chapter by checking that everything works correctly with different expressions. For example:
• Define an expression exp = "( 17 * ( 2 + 3 ) ) + ( 4 + ( 8 * 5 ) )"
• Ask Python to calculate this expression: eval(exp). Python returns 129.
• Convert the expression to Polish notation: polish_notation(exp) returns
"17 2 3 + * 4 8 5 * + +"
• With your calculator calculates the result: polish_calculator("17 2 3 + * 4 8 5 * + +")
returns 129. We get the same result!
 Chapter

17
Text viewer – Markdown

You will program a simple word processor that displays paragraphs cleanly and highlights words in
bold and italics.

Lesson 1 (Text with tkinter).

Here’s how to display text with Python and the graphics window module tkinter.

The code is:

from tkinter import *
from tkinter.font import Font
# tkinter window
root = Tk()
canvas = Canvas(root, width=800, height=600, background="white")
canvas.pack(fill="both", expand=True)
# Font
myfont = Font(family="Times", size=30)
# Some text
canvas.create_text(100,100, text="Text with Python!",
anchor=SW, font=myfont, fill="blue")
# Launch the window
root.mainloop()
Some explanations:
• root and canvas are the variables that define a graphic window (here of width 800 and height
600 pixels). This window is launched by the last command: root.mainloop().
• We remind you that for the graphic coordinates, the y-axis is directed downwards. To define a
rectangle, simply specify the coordinates (x 1 , y1 ) and (x 2 , y2 ) from two opposite vertices (see figure
below).
• The text is displayed by the command canvas.create_text(). It is necessary to specify the
coordinates (x, y) of the point from which you want to display the text.
• The text option allows you to pass the string to display.
• The anchor option allows you to specify the text anchor point, anchor=SW means that the text
box is anchored to the Southwest point (SW) (see figure below).
TEXT VIEWER – MARKDOWN 120

• The fill option allows you to specify the text color.

• The option font allows you to define the font (i.e. the style and size of the characters). Here are
some examples of fonts, it’s up to you to test them:
– Font(family="Times", size=20)
– Font(family="Courier", size=16, weight="bold") in bold
– Font(family="Helvetica", size=16, slant="italic") in italics

(0, 0) x

(x 1 , y1 )

South West anchor

(x 2 , y2 )
y

Activity 1 (Display a text with tkinter).

Goal: display text with the graphics module tkinter.

1. (a) Define a window tkinter of size 800 × 600 for example.

(b) Draw a gray rectangle (which will be our background box) with a size of back_width ×
back_height (for example 700 × 500).
(c) Define several types of fonts: title_font, subtitle_font, bold_font, italic_font,
text_font.
(d) Display texts with different fonts.
2. Write a function text_box(word,font) that draws a rectangle around a text. To do this, use
the canvas.bbox(myobject) method which returns the x 1 , y1 , x 2 , y2 coordinates of the desired
rectangle. (Here myobject = canvas.create_text(...)).
3. Write a function length_word(word,font) that calculates the length of a word in pixels (this
is the width of the rectangle of the previous question).
4. Write a function font_choice(mode,in_bold,in_italics) that returns the name of an
adapted font (among those defined in the first question) according to a mode (among "title",
"subtitle", "text") and according to booleans in_bold, in_italics.
For example, font_choice("text",True,False) returns the font bold_font.

Lesson 2 (Markdown).
The Markdown is a simple markup language that allows you to write your own easy to read text file and
possibly convert it to another format (html, pdf. . . ).
TEXT VIEWER – MARKDOWN 121

Here is an example of a text file with the Markdown syntax with just below its graphic rendering.
# On the Origins of Species

## by Charles Darwin

When on board H.M.S. * Beagle * as naturalist, I was much struck with

certain facts in the distribution of the inhabitants of South America,
and in the geological relations of the present to the past inhabitants
of that continent. These facts seemed to me to throw some light on the
** origin of species ** that mystery of mysteries, as it has been called
by one of our greatest philosophers.

## Chapters

+ Variation under domestication.

+ Variation under nature.
+ Struggle for existence.
+ Natural selection.
+ ...

The syntax is simple, with a clear and clean text file. Here are some elements of this syntax:
• a bold text is obtained by surrounding the text with two asterisks **;
• a text in italics is obtained by surrounding the text with one asterisk *;
• the line of a title begins with one sharp #;
• the line of a subtitle begins with two sharps ##;
• for the elements of a list, each line starts with a special symbol, for us it will be the symbol “plus” +.
• There is also a syntax to display links, tables, code. . .
In the following we will use the simplified syntax as described above.

Activity 2 (View Markdown).

Goal: display text with our simplified Markdown syntax.

1. Write a function print_line_v1(par,posy) that displays one by one the words of a paragraph
par (at the line of ordinate posy).
TEXT VIEWER – MARKDOWN 122

Hints.
• These words are obtained one by one with the command par.split().
• The displayed line starts at the very left, it overflows to the right if it is too long.
• After each word we place a space and then the next word.
• On the picture above the words are framed.
2. Improve your function in print_line_v2(par,posy) to take into account titles, subtitles and
lists.

Hints.
• To know in which mode to display the line, simply test the first characters of the line. The
line of a title begins with #, that of a subtitle with ##, that of a list with +.
• For lists, you can get the bullet character “•” by the unicode character u'\u2022'. You can
also indent each item of the list for more readability.
• Use the font_choice() function of the first activity.
• In the image above, each line is produced by a call to the function. For example
print_line_v2("## And there a subtitle",100).
3. Further improve your function in print_line_v3(par,posy) to take into account the words in
bold and italics in the text.

Hints.
• Words in bold are surrounded by the tag **, words in italics by the tag *.
In our simplified syntax, tags are separated from words by spaces, for example:
"Word ** in bold ** and in * italics *"
• Define a boolean variable in_bold that is false at the beginning; each time you encounter the
tag ** then reverse the value of in_bold (“True” becomes “False”, “False” becomes “True”,
you can use the operator not).
• Still use the font_choice() function of the first activity.
• In the image above, each line is produced by a call to the function. For example
print_line_v3("+ Apples and ** pears ** and * ananas *",100)
TEXT VIEWER – MARKDOWN 123

4. Further improve your function in print_paragraph(par,posy) which manages the display of

a paragraph (i.e. a string that can be very long) on several lines.

Hints.
• As soon as you place a word that exceeds the length of the line (see those that come out of
the frame in the image above), then the next word is placed on the next line.
• The function will therefore modify the variable posy at each line break. At the end, the function
returns the new value of posy, which will be useful for displaying the next paragraph.
5. Ends with a function print_file(filename) that displays the paragraphs of a text file with our
simplified Markdown syntax.

Activity 3 (Justification).
Goal: understand how it is possible to “justify” a text, i.e. to ensure that the words are well aligned
on the left and right sides of the page. To model the problem we work with a series of integers that
represent the lengths of our words.
In this activity:
• list_lengths is a list of integers (for example a list of 50 integers between 5 and 15) which
represent the lengths of the words;
• we set a constant line_length which represents the length of a line. For our examples, this
length is equal to 100.
In previous activities, we moved to the next line after a word had passed the end of the line. We represent
this by the following figure:

You’re going to try to place the words more nicely!

The drawings are based on the example:

list_lengths = [8, 11, 9, 14, 8, 8, 15, 10, 14, 11, 15, 15, 5, 12, 9, 9, 15, 10, 14, 5, 12, 8, 8, 13, 10,
11, 8, 13, 7, 5, 6, 11, 7, 7, 13, 6, 6, 9, 8, 12, 5, 8, 7, 6, 6, 15, 13, 11, 7, 12]
TEXT VIEWER – MARKDOWN 124

which was obtained by random integers:

from random import randint
list_lengths = [randint(5,15) for i in range(50)]
1. Write a function hyphenation_simple() that calculates the indices to make the hyphenations
corresponding to the figure below, i.e. an alignment on the left (without spaces) and without
exceeding the total length of the line (here of length 100).

hypenation_simple()
Use: hypenation_simple(list_lengths)
Input: a sequence of lengths (a list of integers)
Output: the list of indices where to make a cut
Example: hypenation_simple(list_lengths) where list_lengths
is the example given above, returns the list [0, 9, 17, 27, 39, 49,
50]. That is to say that:
• the first line corresponds to indices 0 to 8 (given by range(0,9)),
• the second line corresponds to indices 9 to 16 (given by
range(9,17)),
• the third line corresponds to indices 17 to 26 (given by
range(17,27)),
• ...
• the last line corresponds to the index 49 (given by range(49,50)).

2. Modify your work into a function hyphenation_spaces() that adds a space (with
space_length = 1) between two words of the same line (but not at the beginning of the line,
nor at the end of the line). This corresponds to the following drawing:

For our example, the hyphenations returned are [0, 8, 16, 24, 34, 45, 50].
TEXT VIEWER – MARKDOWN 125

3. In order to be able to justify the text, you allow spaces to be larger than the initial length of 1. On
each line, the spaces between the words are all the same length (greater than or equal to 1) so that
the last word is aligned on the right. From one line to another, the length of the spaces can change.

Write a function compute_space_lengths() that returns the length that the spaces of each line
must have in order for the text to be justified. For our example, we obtain the list [2.43, 1.43,
2.14, 1.67, 1.40, 1.00], i.e. for the first line the spaces must be 2.43 long, for the second
line 1.43,. . .
To find the right formula, simply take the results of the hyphenation_spaces() function and
then, for each line, count the total length of words it contains, as well as the value it lacks to arrive
at a total of 100.

You now have everything you need to visualize text written with the syntax Markdown and justify
it (see the displayed text of Darwin in the lesson). It’s still a lot of work! You can also improve the
support of Markdown syntax: code, numbered lists, sublists, bold and italic words at the same time. . .
 Chapter

L-systems
18

The L-systems offer a very simple way to code complex phenomena. From an initial word and
replacement operations, we arrive at complicated words. When you “draw” these words, you get
beautiful fractal figures. The “L” comes from the botanist A. Lindenmayer who invented the L-systems
to model plants.

Lesson 1 (L-system).
A L-system is the data of an initial word and replacement rules. Here is an example with the starting
word and only one rule:
BlArB A → ABA
The k-iteration of the L-system is obtained by applying k times the substitution to the starting word.
With our example:
• First iteration. The starting word is BlArB, the rule is A → ABA: we replace the A by ABA. We get
the word BlABArB.
• Second iteration. We start from the word obtained BlABArB, we replace the two A by ABA: we get
the word BlABABABArB.
• The third iteration is BlABABABABABABABArB, etc.
When there are two (or more) rules, they must be applied at the same time. Here is an example of a
two-rules L-system:
A A → BlA B → BB
With our example:
• First iteration. The starting word is A, we apply the first rule A → BlA (the second rule does not
L-SYSTEMS 127

apply, because there is no B yet): we get the word BlA.

• Second iteration. We start from the word obtained BlA, we replace the A by BlA and at the same
time the B by BB: we get the word BBlBlA.
• The third iteration is BBBBlBBlBlA, etc.

Lesson 2 (Optional argument for a function).

I want to program a function that draws a line of a given length, with the possibility to change the
thickness of the line and the color.
One method would be to define a function by:
def draw(length, width, color):
I would then call it for example by:
draw(100, 5, "blue"):
But since my features will most of the time have a thickness of 5 and a blue color, I lose time and legibility
by giving this information back each time.
With Python it is possible to give optional arguments. Here is a better way to do this by giving default
values:
def draw(length, width=5, color="blue"):
• The command draw(100) draws my line, and as I only specified the length, the arguments width
and color take the default values (5 and blue).
• The command draw(100, width=10) draws my line with a new thickness (the color is the default
one).
• The command draw(100, color="red") draws my line with a new color (the thickness is the
default).
• The command draw(100, width=10, color="red") draws my line with a new thickness and
a new color.
• Here is also what you can use:
– draw(100, 10, "red"): do not specify the names of the options if you pay attention to
the order.
– draw(color="red", width=10, length=100): you can name any variable, the named
variables can pass as arguments in any order.

Activity 1 (Draw a word).

Goal: make a drawing from a “word”. Each character corresponds to a turtle instruction.

You are given a word (for example AlArAArArA) in which each letter (read from left to right) corresponds
to an instruction for the Python turtle.
• A or B: advance of a fixed quantity (by tracing),
• l: turn left, without moving forward, by a fixed angle (most often 90 degrees),
• r: turns right by a fixed angle.
The other characters do not do anything. (More commands will be added later on.)
Program a function draw_lsystem(word,angle=90,scale=1) which displays the drawing corre-
sponding to the letters of word. By default the angle is 90 degrees, and each time you advance, it is of
100×scale.
For example: draw_lsystem("AlArAArArA") displays this:
L-SYSTEMS 128

Activity 2 (Only one rule – Koch’s snowflake).

Goal: draw the Koch snowflake from a word obtained by iterations.

1. Program a function replace_1(word,letter,pattern) that replaces a letter with a pattern

in a word.
For example with word = "ArAAl", replace_1(word,"A","Al") returns the word AlrAlAll:
each letter A has been replaced by the pattern Al.
2. Program a function iterate_lsystem_1(start,rule,k) which calculates the k-iteration of
the L-system associated with the initial word start according to the rule rule which contains
the couple formed by the letter and its replacement pattern. For example, with:
• start = "A"
• rule = ("A","AlArArAlA") i.e. A → AlArArAlA
• for k = 0, the function returns the starting word A,
• for k = 1, the function returns AlArArAlA,
• for k = 2, the function returns :
AlArArAlAlAlArArAlArAlArArAlArAlArArAlAlAlArArAlA
• for k = 3, the function returns: AlArArAlAlAl... a word of 249 letters.
3. Trace the first images of the Koch’s snowflake given as above by:
start: A rule: A → AlArArAlA
Here the images for k = 1 up to k = 5. For k = 1, the word is AlArArAlA and you can check the
trace on the first image.
L-SYSTEMS 129

4. Trace other fractal figures from the following L-systems. For all these examples the starting word is
"ArArArA" (a square) and the rule is to be chosen among:
• ("A","ArAlAlAArArAlA")
• ("A","AlAArAArArAlAlAArArAlAlAAlAArA")
• ("A","AArArArArAA")
• ("A","AArArrArA")
• ("A","AArArArArArAlA")
• ("A","AArAlArArAA")
• ("A","ArAArrArA")
• ("A","ArAlArArA")

Invent and trace your own L-systems!

Activity 3 (Two rules – Sierpinski triangle).

Goal: calculate more complicated L-systems by allowing two replacement rules instead of one.

1. Program a function replace_2(word,letter1,pattern1,letter2,pattern2) that re-

places a first letter with a pattern and a second letter with another.
L-SYSTEMS 130

For example with word = "ArBlA", replace_2(word,"A","ABl","B","Br") returns the

word ABlrBrlABl: each letter A has been replaced by the pattern ABl and at the same time each
letter B has been replaced by the Br.
Warning! Do not get ABrlrBrlABrl. If this is the case, it is because you used the replace_1()
function to first replace the A, then a second time for the B (but after the first replacement new B
appeared). A new function must be reprogrammed to avoid this.
2. Program a function iterate_lsystems_2(start,rule1,rule2,k) which calculates the k-
iteration of the L-system associated with the initial word start according to the rules rule1 and
rule2. For example, with:
• start = "ArBrB"
• rule1 = ("A","ArBlAlBrA") i.e. A → ArBlAlBrA
• rule2 = ("B","BB") i.e. B → BB
• for k = 0, the function returns the starting word ArBrB,
• for k = 1, the function returns ArBlAlBrArBBrBB,
• for k = 2, the function returns:
ArBlAlBrArBBlArBlAlBrAlBBrArBlAlBrArBBBBrBBBB
3. Trace the first pictures of the Sierpinski triangle given as above by:
start: ArBrB rules: A → ArBlAlBrA B → BB
The angle is −120 degrees. Here are the images for k = 0 up to k = 4.

4. Trace other fractal figures from the following L-systems.

• The dragon curve:
start="AX" rule1=("X","XlYAl") rule2=("Y","rAXrY")
The letters X and Y do not correspond to any action.
• A variant of the Sierpinski triangle, with angle = 60 :
start="A" rule1=("A","BrArB") rule2=("B","AlBlA")
• The Gosper curve, with angle = 60:
start="A"
rule1=("A","AlBllBrArrAArBl")
rule2=("B","rAlBBllBlArrArB")
L-SYSTEMS 131

Invent and trace your own L-systems with two rules!

Lesson 3 (Stacks).
A stack is a temporary storage area. Details are in the chapter “Polish calculator – Stacks”. Here are just a
few reminders.

(−100, 0), 60

Top of the stack
(0, 0), 0


(100, 100), 0


(0, 0), 90


(0, 0), 30

A stack

• A stack is like a stack of plates; elements are placed one by one above the stack; the elements are
removed one by one, always from the top. It’s the principle: “last in, first out”.
• We model a stack by a list.
• At the beginning the stack is empty: stack = [].
• Push. We add the items at the end of the list: stack.append(element) or stack = stack +
[element].
• Pop. An item is removed by the command pop():
element = stack.pop()
which returns the last item in the stack and removes it from the list.


• On the drawing and in the next activity, the elements of the stack are of the type (x, y), θ that
will store a state of the turtle: (x, y) is the position and θ its direction.

Activity 4 (L-system, stack and turtle).

Goal: improve our tracings by allowing us to move forward without tracing and also by using a
kind of flashback, to trace plants.
L-SYSTEMS 132

1. Forward without tracing.

Increase the possibilities by allowing the turtle to move forward without drawing a line, when the in-
struction is the letter a (in lowercase). (It is sufficient to modify the function trace_lsystems().)
Then trace the following L-system:
• start = "ArArArA"
• rule1 = ("A","AlarAAlAlAAlAalAAralAArArAArAarAAA")
• rule2 = ("a","aaaaaa")

2. Return back.
We now allow square brackets in our words. For example AlA[lAAA]A[rAA]A. When you encounter
a opening bracket “[”, you memorize the position of the turtle, then the commands in brackets are
executed as usual, when you find the closing bracket “]” you go back to the position stored before.
Let us understand the example of the route of the AlA [lAAA] A [rAA] A

P AlA [lAAA] A [rAA] A

• AlA: we start from the point O, we move forward, we turn, we move forward.
L-SYSTEMS 133

• [lAAA]: we retain the current position (the point P) and also the direction; we turn, we
advance three times (we trace the red segment); at the end we return the turtle to the position
P (without tracing and with the same direction as before).
• A: from P we advance (green segment).
• [rAA]: we retain the position Q and the direction, we turn and we trace the purple segment.
We come back in Q with the old state.
• A: from Q we trace the last segment.
Here is how to draw a word containing brackets using a stack:
• At the beginning the stack is empty.
• We read one by one the characters of the word. The actions are the same as before.
• If the character is the opening bracket “[” then add to the stack the current position and
direction of the turtle (x, y), θ that you get by (position(), heading()).

• If the character is the closing bracket “]” then pop (i.e. read the top element of the stack and
remove it). Put the position of the turtle and the angle with the read values, use goto() and
setheading().
3. Trace the following L-systems, where the starting word and rule (or rules) are given. The angle is
to be chosen between 20 and 30 degrees.
• "A" ("A","A[lA]A[rA][A]")
• "A" ("A","A[lA]A[rA]A")
• "A" ("A","AAr[rAlAlA]l[lArArA]")
• "X" ("X","A[lX][X]A[lX]rAX") ("A","AA")
• "X" ("X","A[lX]A[rX]AX") ("A","AA")
• "X" ("X","Ar[[X]lX]lA[lAX]rX") ("A","AA")

Invent your own plant!

 Chapter

Dynamic images
19

We will distort images. By repeating these distortions, the images become blurred. But by a miracle
after a certain number of repetitions the original image reappears!

Lesson 1 (Transformation of the photo booth).

We start from an array n × n, with n even, each element of the table represents a pixel. The rows are
indexed from i = 0 to i = n − 1, the columns from j = 0 to j = n − 1. From this image we calculate a new
image by moving each pixel according to a transformation, called the photo booth transformation.
We cut the original image into small squares of size 2 × 2. Each small square is therefore composed of
four pixels. Each of these pixels is sent to four different locations in the new image: the pixel at the top
left remains in an area at the top left, the pixel at the top right of the small square, is sent to an area at
the top right of the new image,...
j =0 j =1 j 0 =0 j 0 =1

i =0 A B a b i 0 =0 A a B b
i =1 C D c d i 0 =1

C c D d

For example, the pixel in position (1, 1) (symbolized by the letter D) is sent in position (4, 4).
Let’s explain this principle through formulas. For each couple (i, j), we calculate its image (i 0 , j 0 ) by the
photo booth transformation according to the following formulas:
• If i and j are even: (i 0 , j 0 ) = (i//2, j//2).
• If i is even and j is odd: (i 0 , j 0 ) = (i//2, (n + j)//2).
• If i is odd and j is even: (i 0 , j 0 ) = ((n + i)//2, j//2).
• If i and j are odd: (i 0 , j 0 ) = ((n + i)//2, (n + j)//2).
Here is an example of an array 4 × 4 before (left) and after (right) the transformation of the photo booth.
1 2 3 4 1 3 2 4
5 6 7 8 9 11 10 12
9 10 11 12 5 7 6 8
13 14 15 16 13 15 14 16
DYNAMIC IMAGES 135

Here is an image 256 × 256 and its first transformation:

Here is what happens if you repeat the photo booth transformation several times:

The image becomes more and more blurred, but after some number of repetitions of the transformation,
we fall back on the original image!

Activity 1 (Transformation of the photo booth).

Goal: program the transformation of the photo booth that decomposes an image into sub-pictures.
When this transformation is iterated, the image gradually disintegrates, then suddenly re-formed.
1. Program a function transformation(i,j,n) that realize the formula of the photo booth trans-
formation and returns the coordinates (i 0 , j 0 ) of the pixel image (i, j).
For example, transformation(1,1,8) returns (4,4).
2. Program a function photo_booth(array) that returns the table calculated after transformation.
For example, the array on the left is transformed into the array on the right.
1 2 3 4 1 3 2 4
5 6 7 8 9 11 10 12
9 10 11 12 5 7 6 8
13 14 15 16 13 15 14 16
Hints. You can initialize a new table with the command:
new_array = [[0 for j in range(n)] for i in range(n)]
Then fill it with commands of the type:
new_array[ii][jj] = array[i][j]
3. Program a function photo_booth_iterate(array,k) that returns the table calculated after k
iterations of the photo booth transformation.
4. To be finished after completing activity 2.
DYNAMIC IMAGES 136

Program a function photo_booth_images(image_name,kmax) that calculates the images cor-

responding to the transformation, for all iterations from k = 1 to k = kmax .
5. Experiment for different values of the size n, to see after how many iterations we find the original
image.
Here is the starting image size 256 × 256 and the images obtained by iterations of the photo booth
transformation for k = 1 up to k = 8. After 8 iterations we find the initial image again.

Activity 2 (Conversion array/image).

Goal: switch from an array to an image file and vice versa. The format to display the images is
the format “pgm” which has been manipulated in the chapter “Files”.
1. Array to image.
Program a function array_to_image(array,image_name) that writes an image file in format
“pgm” from a grayscale table.

P2
5 5
255
128 192 128 192 128
224 0 228 0 224
228 228 228 228 228
224 64 64 64 224
192 192 192 192 192

For example with array = [ [128,192,128,192,128], [224,...] ], the command

array_to_image(array,"test") writes a file test.pgm (on the left) that would be displayed
as the image on the right.
2. Image to array.
DYNAMIC IMAGES 137

Program a function image_to_array(image_name) which from an image file in format “pgm”,

returns an array of gray levels.

Lesson 2 (Baker’s transformation).

We start from an array n× n, with n even, each element representing a pixel. We will apply two elementary
transformations each time:
• Stretching. The principle is as follows: the first two lines (each with a length of n) produce a single
line with a length of 2n by mixing the values of each line by alternating an upper element and a
lower element.

stretch

Here is how two lines mix into one:

A B
A a B b
a b

Formulas. An element at position (i, j) of the target array, corresponds to an element (2i, j//2) (if j
is even) or (2i + 1, j//2) (if j is odd) of the source array, with here 0 6 i < 2n and 0 6 j < 2n.
Example. Here is an array 4 × 4 on the left, and the stretched array 2 × 8 on the right. The rows 0
and 1 on the left give the row 0 on the right. The rows 2 and 3 on the left give the row 1 on the
right.
1 2 3 4
5 6 7 8 1 5 2 6 3 7 4 8
9 10 11 12 9 13 10 14 11 15 12 16
13 14 15 16
• Fold. The principle is as follows: the right part of a stretched array is turned upside down, then
added under the left part. Starting from an array 2n × 2n you get an array n × n.

A B
fold A
B
Formulas. For 0 6 i < 2n and 0 6 j < n the elements in position (i, j) of the array are kept. For
n n


2 6 i < n and 0 6 j < n an element of the array (i, j), corresponds to an element 2 −i−1, 2n−1− j
of the source array.
DYNAMIC IMAGES 138

Example. From the stretched array 2 × 8 on the left, we obtain a folded array 4 × 4 on the right.
1 5 2 6
1 5 2 6 3 7 4 8 9 13 10 14
9 13 10 14 11 15 12 16 16 12 15 11
8 4 7 3
The transformation of the baker is the succession of a stretching and a folding. Starting from an array
n × n we still get an array n × n.
Let’s see an example of the action of several transformations of the baker. On the left the initial image
size 128 × 128, then the result of k = 1, 2, 3 iterations.

Here are the images for k = 12, 13, 14, 15 iterations:

Activity 3 (Baker’s transformation).

Goal: program a new transformation that stretches and folds an image. Once again, the image
becomes more and more distorted but, after a certain number of iterations, we find the original
image again.
1. Program a function baker_stretch(array) that returns a new array obtained by “stretching”
the table given as input.
2. Program a function baker_fold(array) that returns a table obtained by “folding” the table
given as input.
3. Program a function baker_iterate(array,k) that returns the table calculated after k iterations
of the baker’s transformation.
For example, here is a table 4 × 4 on the left, its image by the transformation (k = 1) and its image
after a second transformation (k = 2).

1 2 3 4 1 5 2 6 1 9 5 13
5 6 7 8 9 13 10 14 16 8 12 4
9 10 11 12 16 12 15 11 3 11 7 15
13 14 15 16 8 4 7 3 14 6 10 2
4. Program a function baker_images(image_name,kmax) that calculates the images correspond-
ing to the transformation of the baker, with iterations ranging from k = 1 to k = kmax .
5. Experiment for different values of the size n, to see after how many iterations we find the original
image.
DYNAMIC IMAGES 139

Caution! It sometimes takes a lot of iterations to get back to the original image. For example with
n = 4, we find the starting image after k = 5 iterations; with n = 256 it is k = 17. Conjecture
a return value in the case where n is a power of 2. However, for n = 10, you need k = 56 920
iterations!
Here is an example with an image of size 256 × 256, first the initial image, then one transformation
(k = 1) and a second iteration (k = 2).

k = 3, 4, 5 :

k = 15, 16, 17 :

For k = 17 you find the original image!

This chapter is based on the article “Blurred images, recovered images” by Jean-Paul Delahaye and Philippe
Mathieu (Pour la Science, 1997).
 Chapter

20
Game of life

The game of life is a simple model of the evolution of a population of cells that born and die over time.
The “game” consists in finding initial configurations that give interesting evolution: some groups of
cells disappear, others stabilize, some move. . .

Lesson 1 (Rules of the game of life).

The game of life takes place on a grid. Each square can contain one cell. Starting from an initial configura-
tion, each day cells will be born and others will die depending on the number of its neighbors.
Here are the rules:
• For an empty square on the day j and having exactly 3 neighboring cells: a cell is born on the day
j + 1.

Has 3 neighbors Birth

Initial position (day j) Number of neighbors Evolution (day j + 1)

Birth

• For a square containing a cell at day j, having either 2 or 3 neighboring cells: then the cell continues
to live. In other cases the cell dies (with 0 or 1, it dies of isolation, with more than 4 neighbors, it
dies of overpopulation).
GAME OF LIFE 141

Has 4 neighbors Has 0 neighbor

Has 4 neighbors Deaths Births

Initial position (day j) Numbers of neighbors Evolution (day j + 1)

Three deaths (and two births)

Here is a simple example, the “blinker”.

0 0 0 0 0
1 2 3 2 1
1 1 2 1 1
1 2 3 2 1
0 0 0 0 0
Initial position (day j) Numbers of neighbors Evolution (day j + 1)

Here is another example (with directly the number of neighbors), the first evolution and the evolution of
the evolution.

0 1 2 2 1 1 2 3 2 1
1 3 3 2 1 2 2 3 1 1
1 3 4 4 1 3 4 6 3 1
1 3 2 2 0 2 2 2 1 0
0 1 1 1 0 1 2 2 1 0
Day 0 Day 1 Day 2

Activity 1 (Display).
Goal: define and display arrays.
We model the living space of the cells by a double entry table, containing integers, 1 to indicate the
presence of a cell, 0 otherwise. Here is an example of the configuration “toad” and its array:
index j

j=0 j=1 ··· j = p−1

i=0 0 0 0 0 0 0
i=1 0 0 1 1 1 0
index i n
..
. 0 1 1 1 0 0
i = n−1 0 0 0 0 0 0

p
Cells Array
GAME OF LIFE 142

1. • Initialize two variables n (the height of the table) and p (the width) (for example at 5 and 8).
• Define a two-dimensional table filled with zeros by the command:
array = [[0 for j in range(p)] for i in range(n)]
• By instructions of the type array[i][j] = 1, fill in the table to define the configuration of
the blinker, the toad . . .
2. Program the display of a given array on the screen. For example, the blinker is displayed as follows:

00000000
00000000
00111000
00000000
00000000
Hint. By default the command print() passes to the next line on each call (it adds the character
"\n" which is the end of line character). It can be specified not to do so by the option print("My
text",end="").

Activity 2 (Graphic display).

Goal: create a graphical display of a cell configuration.

1. Program the creation of a tkinter window (see chapter “Statistics – Data visualization”). Write a
function draw_grid(), without parameters, that displays a blank grid of coordinates. You will
need a constant scale (for example scale = 50) to transform the indices i and j into graphic
coordinates x and y, depending on the relationship: x = s × j and y = s × i (s being the scale).
Bonus. You can also mark the values of the indices i and j at the top and the left to make it easier
to read the grid.
2. Build a function draw_array(array) that graphically displays the cells from a table.
Bonus. Add a button “View” and a button “Quit” (see chapter “Statistics – Data visualization”).

Activity 3 (Evolution).
Goal: calculate the evolution of a configuration to the next day.
1. Program a function number_neighbors(i,j,array) that calculates the number of living neigh-
GAME OF LIFE 143

boring cells to the cell (i, j).

Hints.
• There is a maximum of 8 for possible neighbors. The easiest way is to test them one by one!
• For the counting, it is necessary to be very careful with the cells that are placed near one edge
(and have less than 8 neighbors).
2. Program a function evolution(array) that receives a table at the input and returns a new array
corresponding to the situation on the next day, according to the rules of the game of life described
at the beginning. For example, if the table on the left corresponds to the input, then the output
corresponds to the table on the right:
00000000 00000000
00000000 00010000
00111000 evolves in 00010000
00000000 00010000
00000000 00000000
Hints. To define a new table, use the command:
new_array = [[0 for j in range(p)] for i in range(n)]
and then modify the table as desired.

Activity 4 (Iterations).
Goal: end the graphic program so that the user can define configurations and make them evolve
with a simple click.
1. Improve the graphic window to make the user’s life easier:
• A button “evolve” which displays a new evolution with each click.
• Buttons to display predefined configurations (on the screenshot below is the configuration
“pentadecathlon”).

2. Perfect your program so that the user can draw the configuration he wants with mouse clicks.
Clicking on an extinguished cell turns it on, clicking on an living cell turns it off. You can break
this work down into three functions:
• on_off(i,j), which switches the cell (i, j) ;
• xy_to_ij(x,y) that converts graphic coordinates (x, y) into integers coordinates (i, j) (use
variable scale and the integer division).
GAME OF LIFE 144

• action_click_mouse(event) to retrieve the coordinates (x, y) with a mouse click (see

course below) and switch the clicked box.

Lesson 2 (Mouse click).

Here is a small program that displays a graphic window. Each time the user clicks (with the left mouse
button) the program displays a small square (on the window) and displays “Click at x = . . ., y = . . .” (on
the console).
from tkinter import *

# Window
root = Tk()
canvas = Canvas(root, width=800, height=600, background="white")
canvas.pack(side=LEFT, padx=5, pady=5)

# Catch mouse click

def action_mouse_click(event):
canvas.focus_set()
x = event.x
y = event.y
canvas.create_rectangle(x,y,x+10,y+10,fill="red")
print("Click at x =",x,", y =",y)
return

# Association click/action
canvas.bind("<Button-1>", action_mouse_click)

# Launch
root.mainloop()
Here are some explanations:
• The creation of the window is usual. The program ends with the launch by the command
mainloop().
• The first key point is to associate a mouse click to an action, that’s what the line does:
canvas.bind("<Button-1>", action_mouse_click)
GAME OF LIFE 145

Each time the left mouse button is clicked, Python executes the function action_mouse_click.
(Note that there are no brackets for the call to the function.)
• Second key point: the action_mouse_click function retrieves the click coordinates and then
does two things here: it displays a small rectangle at the click location and displays the (x, y)
coordinates in the terminal window.
• The coordinates x and y are expressed in pixels; (0, 0) refers to the upper left corner of the window
(the area delimited by canvas).

Here are some ideas to go further:

• Make sure that the grid automatically adapts to the screen width, i.e. calculate scale based on n and
p.
• Add an intermediate step before evolving: color in green a cell that will be born and in black a cell
that will die. For this purpose the elements of array can take values other than 0 and 1.
Here are some interesting configurations.

Square Spaceship Pulsar

You will find many more on the Internet but above all have fun discovering new ones!
In particular, find configurations:
• that remain fixed over time;
• which evolve, then become fixed;
• which are periodic (the same configurations come back in loops) with a period of 2, 3 or more;
• who travel;
• that propel cells;
• whose population is increasing indefinitely!
 Chapter
Ramsey graphs
21
and combinatorics

You will see that a very simple problem, which concerns the relationships between only six people,
will require a lot of calculations to be solved.

Lesson 1 (Ramsey’s problem).

Proposition. In a group of 6 people, there is always 3 friends (the three know each other two by
two) or 3 foreigners (all three are strangers to each other).
The purpose of this chapter is to have the computer demonstrate this statement. For that we will model
the problem by graphs, then we will check the statement for each of the 32 768 possible graphs.

We consider n people. For two of them, either they know each other (they are friends) or they do not
know each other (they are strangers to each other). We schematize this by a graph:
• a person is represented by a vertex (numbered from 0 to n − 1);
• if two people are friends, the corresponding vertices are connected by a green solid edge;
• otherwise (they are strangers), the corresponding vertices are connected by a red dotted edge.
The graph below means that 0 is friend with 2; 1 is friend with 3. The other pairs don’t know each other.

3 0

2 1

A graph checks the Ramsey problem, if there are among its vertices, either 3 friends, or either 3 foreigners.
RAMSEY GRAPHS AND COMBINATORICS 147

j0 i0

i
Three stangers

k Three friends k0

Here is an example of 5 peoples that verify the statement: there is group of 3 strangers (the 0, 2 and 4
vertices), even if there is no group of three friends.

1
A graph with n = 5 that satifies Ramsey’s statement

Lesson 2 (Model).

We model a graph by an array, containing 0 and 1.

Let be a graph with n vertices, numbered from 0 to n − 1. The array of the graph is a table of size n × n
in which we place a 1 in position (i, j) if the vertices i and j are connected by an edge, otherwise we
place a 0.

First example below: the vertices 0 and 2 are friends (because they are connected by a green edge) so the
table contains a 1 in position (0, 2) and also in (2, 0). Similarly 1 and 3 are friends, so the table contains
a 1 in position (1, 3) and (3, 1). The rest of the table contains 0.

index j

3 0 j=0 j=1 j=2 j=3

i=0 0 0 1 0

i=1 0 0 0 1
index i n
i=2 1 0 0 0

i=3 0 1 0 0
2 1

Here is a more complicated graph and its array:

RAMSEY GRAPHS AND COMBINATORICS 148

j=0 j=1 j=2 j=3 j=4

3
i=0 0 1 0 1 0

i=1 1 0 1 0 1

i=2 0 1 0 1 0
0
i=3 1 0 1 0 1

i=4 0 1 0 1 0
2

Activity 1 (Build graphs).

Goal: define graphs and test if three given vertices are friends.

1. Define the graph array for the four examples above. You can start by initializing the array with
array = [[0 for j in range(n)] for i in range(n)]
Then add commands:
array[i][j] = 1 and array[j][i] = 1
Don’t forget that if the i vertex is connected to the j vertex by an edge, then you have to put a 1 in
position (i, j) but also in position ( j, i).
2. Define a function print_graph(array) that allows you to display on the screen the table of a
graph. Thus the third example above (with n = 5) should be displayed as follows:
RAMSEY GRAPHS AND COMBINATORICS 149

0 0 1 1 0
0 0 1 0 1
1 1 0 0 0
1 0 0 0 1
0 1 0 1 0
3. We set three vertices i, j, k of a graph. Write a function have_3_fix_friends(array,i,j,k)
that tests if the vertices i, j, k are three friends (the function returns “True” or “False”). Do the
same work with a have_3_fix_strangers(array,i,j,k) function to find out if these vertices
are strangers.
Find by hand on the fourth example, three friend or foreign vertices and check your answer using
the functions you have just defined on these vertices.

Activity 2 (Draw nice graphs).

Goal: draw a graph! Optional activity.

Program the graphical display of a graph by a function display_graph(array).

Hints. This activity is not necessary for the next steps, it just helps to visualize the graphs. It is neces-
sary to use the module tkinter and the functions create_line(), create_oval() and possibly
create_text().
The most delicate point is to obtain the coordinates of the vertices. You will need the sine and cosine
functions (available in the module math). The coordinates (x i , yi ) of the vertex number i of a graph
with n elements can be calculated by the formulas :
2iπ 2iπ
 ‹  ‹
x i = r cos and yi = r sin .
n n
These vertices are located on the circle of radius r, centered at (0, 0). You will have to choose r large
enough (for example r = 200) and shift the circle to display it on the screen.
RAMSEY GRAPHS AND COMBINATORICS 150

(x 1 , y1 )

(x i , yi )
yi = r sin( 2iπ
n )

(x 0 , y0 )
x i = r cos( 2iπ (0, 0) x
n )

(x n−1 , yn−1 )

Activity 3 (Binary notation with zero-padding).

Goal: convert an integer to binary notation with possible leading zeros.

Program an integer_to_binary(p,n) function that displays the binary notation of an integer p on n

bits. The result is a list of 0 and 1.

Example.
• The binary notation of p = 37 is 1.0.0.1.0.1. If you want its binary notation on n = 8 bits you have
to add two 0 in front of it: 0.0.1.0.0.1.0.1.
• So the result of the command integer_to_binary(37,8) must be [0, 0, 1, 0, 0, 1,
0, 1].
• The command integer_to_binary(37,10) returns the binary notation of 37 on 10 bits: [0,
0, 0, 0, 1, 0, 0, 1, 0, 1].

Hints.
• You can use the command bin(p).
• The command list(mystring) returns the list of characters making up mystring.
• Attention! We want a list of integers 0 or 1, not of characters '0' or '1'. The command int('0')
returns 0 and int('1') returns 1.
• mylist = mylist + [element] adds an item at the end of the list, while mylist =
[element] + mylist adds the item at the beginning of the list.

Lesson 3 (Subsets).
Let En = {0, 1, 2, . . . , n − 1} be the set of all integers from 0 to n − 1. The set En therefore contains n
elements. For example E3 = {0, 1, 2}, E4 = {0, 1, 2, 3}. . .

Subsets.
What are the subsets of En ? For example there are 8 subsets of E3 , these are:
• the subset {0} composed of the single element 0;
RAMSEY GRAPHS AND COMBINATORICS 151

• the subset {1} composed of the single element 1;

• the subset {2} composed of the single element 2;
• the subset {0, 1} composed of the element 0 and the element 1;
• the subset {0, 2};
• the subset {1, 2};
• the subset {0, 1, 2} composed of all elements;
• and the empty set ∅ which contains no elements!
Proposition. The set En contains 2n subsets.
For example E4 = {0, 1, 2, 3} has 24 = 16 possible subsets. Have fun finding them all! For E6 there is
26 = 64 possible subsets.

Subsets of fixed cardinal.

We are only looking for subsets with a fixed number k of elements.
Examples:
• For n = 3 and k = 2, the subsets having two elements and contained in E3 = {0, 1, 2} are the three
pairs: {0, 1}, {0, 2}, {1, 2}.
• For n = 5 and k = 3, the subsets having three elements and contained in E5 = {0, 1, 2, 3, 4} are
the 10 triples: {0, 1, 2}, {0, 1, 3}, {0, 2, 3}, {1, 2, 3}, {0, 1, 4}, {0, 2, 4}, {1, 2, 4}, {0, 3, 4}, {1, 3, 4},
{2, 3, 4}.

Activity 4 (Subsets).

Goal: generate all subsets to test all triples of vertices. For this we will use binary notation.

Here is how we associate to each integer p verifying 0 6 p < 2n a subset of En = {0, 1, . . . , n − 1}.
Let’s start with an example, with n = 6 and p = 26:
• the binary notation of p = 26 on n = 6 bits is [0,1,1,0,1,0];
• there are 1 at ranks 1, 2 and 4 (starting at rank 0 on the left);
• the associated subset is then {1, 2, 4}.

p = 26 0 1 1 0 1 0 {1, 2, 4}
(n = 6)
rank : 0 1 2 3 4 5

Initial integer p Binary notation of p on n bits Subset associated to p

Other examples.
• With n = 8 and p = 57 whose binary notation on 8 bits is [0,0,1,1,1,0,0,1], the associated
subset corresponds to the ranks 2, 3, 4, 7, so it is {2, 3, 4, 7}.

p = 57 0 0 1 1 1 0 0 1 {2, 3, 4, 7}
(n = 8)
rank : 0 1 2 3 4 5 6 7
Initial integer p Binary notation of p on n bits Subset associated to p

• With p = 0, the binary notation is only formed of 0, the associated subset is the empty set.
RAMSEY GRAPHS AND COMBINATORICS 152

• With p = 2n − 1, the binary notation is full of 1, the associated subset is the whole set En =
{0, 1, . . . , n − 1} itself.

We model a set as a list of elements. For example:

• The set E4 is for us the list [0,1,2,3].
• A subset of E4 is for example the pair [1,3].
• The empty set is represented by the empty list [].
1. Program a function subsets(n) which returns the list of all possible subsets of En = {0, 1, 2, . . . , n−
1}. For example, with n = 3, subsets(n) returns the list (which itself contains lists):
[ [], [2], [1], [1, 2], [0], [0, 2], [0, 1], [0, 1, 2] ]
That is to say the 8 subsets (starting with the empty set):
∅ {2} {1} {1, 2} {0} {0, 2} {0, 1} {0, 1, 2}.
n
Hint. To test your program, check that the returned list contains 2 subsets.
2. Deduce a function fix_subsets(n,k) from it that returns only subsets of En with k elements.
For example, for n = 3 and k = 2, fix_subsets(n,k) returns the list of pairs:
[ [0, 1], [0, 2], [1, 2] ]
Test your program:
• For n = 4 and k = 3, the list returned by fix_subsets(n,k) contains 4 triples.
• For n = 5 and k = 3, there are 10 triples possible.
• For n = 10 and k = 4, there are 210 possible subsets!
In the following we will use mainly subsets having 3 elements. In particular, for n = 6, there are 20
triples included in {0, 1, 2, 3, 4, 5}:

[[3, 4, 5], [2, 4, 5], [2, 3, 5], [2, 3, 4], [1, 4, 5],
[1, 3, 5], [1, 3, 4], [1, 2, 5], [1, 2, 4], [1, 2, 3],
[0, 4, 5], [0, 3, 5], [0, 3, 4], [0, 2, 5], [0, 2, 4],
[0, 2, 3], [0, 1, 5], [0, 1, 4], [0, 1, 3], [0, 1, 2]]

Activity 5 (Ramsey’s theorem for n = 6).

Goal: check that all graphs with 6 vertices contain three friends or three strangers.

1. Program a function have_3(array) that tests if a graph contains 3 friends or 3 for-

eigners. You must therefore call the functions have_3_fix_friends(array,i,j,k) and
have_3_fix_strangers(array,i,j,k) of the first activity for all possible triples of vertices
(i, j, k).
For the four examples of the first activity, only the fourth (with 6 summits) checks the test.
2. Program a function all_graphs(n) that computes all possible graph arrays with n vertices. There
(n−1)n
are N = 2 possible arrays. You can generate them by a method similar to the one for subsets:
• for each integer p that verifies 0 6 p < 2N ,
• calculate the binary notation of p on N bits,
• fill in the array element by element, with the 0 and 1 of the binary notation.
Hints. To fill an array from a given binary notation of p named binary_notation (that is to say
a list of 0 and 1), you can use a double loop like:
for j in range(0,n):
for i in range(j+1,n):
RAMSEY GRAPHS AND COMBINATORICS 153

b = binary_notation.pop()
array[i][j] = b
array[j][i] = b
Here is the principle of this loop that fills the part above the diagonal (and also the part below it
by symmetry). This loop takes the last bit of the list and places it on the first free box above the
diagonal; then the penultimate bit is placed on the second free box. . . ; the first bit of the list fills
the last free square.

Binary notation of p with N bits

1 0 1 1 0 0 0 1 0 1

0 1 0 1 0
0 0 0 1 N boxes above the diago-
N boxes below the diago- 0 1 0 nal
nal (filled by symetry) 0 1
0

Filling of an array

3. Convert the previous function into a function test_all_graphs(n) which tests the conjecture
“there are three friends or three strangers” for all graphs with n vertices. You must find that:
• for n = 4 and n = 5 the guess is wrong. Give a graph having 4 vertices (then 5 vertices) that
has neither 3 friends, nor 3 foreigners;
5×6
• for n = 6 let the computer check that, for each of the N = 2 2 = 32 768 graphs with 6
vertices, either it has 3 friends or it has 3 foreigners.

Activity 6 (To go further).

Goal: improve your program and prove other guesses. Optional activity.

1. Improve your program so that it checks the guess for n = 6 in less than a second.
Ideas.
• The list of triples must be generated once and for all at the beginning of the program (and
not at each new graph).
• It is not necessary to generate a list of all possible graphs, then test them in a second step. It
is better to generate one and then test it before moving on to the next.
• As soon as you find 3 friends (or 3 foreigners) it’s won! Immediately stop the loop even if it
means using the instruction break and move to the next graph.
• You can only test the graphs that correspond to p between 0 and 2N /2 (because for the next
p it is like exchanging the green segments for red and vice versa).
With these tips here are the calculation times you can expect:
RAMSEY GRAPHS AND COMBINATORICS 154

Number of vertices Number of graphs Approximate calculation time

n=6 32 768 < 1 second
n=7 2 097 152 < 1 minute
n=8 268 435 456 < 1 hour
n=9 68 719 476 476 736 < 10 days

2. There is a more difficult statement. It is a question of finding out at which size n a graph always
contains either 4 friends or 3 foreigners. Being 4 friends means that two by two they are connected
by a green segment, as below:

(a) Find graphs with n = 6 (then n = 7) vertices that do not check this statement.
(b) By searching a little with the machine find graphs with 8 vertices that do not satisfy this
statement.
(c) Prove that any graph having 9 vertices contains 4 friends or 3 foreigners!
Hints. It is necessary to test all the graphs corresponding to integers p between 0 and 2N =
8×9
2 2 = 68 719 476 736. The total calculation time is about 20 days! You can share the
calculations between several computers: one computer does the calculations for 0 6 p 6
1 000 000, a second computer for 1 000 001 6 p 6 2 000 000,. . .
3. • There are arguments to prove with pencil and paper that for n = 6 there is always 3 friends
or 3 foreigners. Look for such reasoning! With a little more effort, we can also prove that it is
n = 9 that answers the problem of 4 friends/3 foreigners.
• We know how to prove that we need n = 18 to always have 4 friends or 4 foreigners.
• However, no one in the world knows what the smallest n is for the 5 friends/5 foreigners
problem!
 Chapter

22
Bitcoin

The bitcoin is a dematerialized and decentralized currency. It is based on two computer principles:
public key cryptography and proof of work. To understand this second principle, you will create a
simple model of bitcoin.

Activity 1 (Proof of work).

Goal: understand what “proof of work” means by studying a simple model. This activity is
independent of the rest of the chapter. The idea is to find a problem that is difficult to solve but
easy to check. Like sudokus for example: it only takes ten seconds to check that a grid is filled
correctly, but it took more than ten minutes to solve it.
The mathematical problem is as follows: you are given a prime number p and an integer y; you must
find an integer x such that:
x 2 = y (mod p)
In other words, x is a square root of y modulo p. Be careful, there is not always a solution for x.
Examples.
• For p = 13 and y = 10, then a solution is x = 6: because x 2 = 62 = 36 and 36 = 2 × 13 + 10 so
x 2 = 10 (mod 13).
• The solution is not necessarily unique. For example, check that x = 7 is also a solution.
• There is not always a solution, for example for p = 13 and y = 5, no integer x is a solution.
• Exercise: for p = 13, find by hand two solutions x to the problem x 2 = 9 (mod 13); find a solution
x to the problem x 2 = 3 (mod 13).
The prime number p is fixed for all the activity. For easy examples we will take p = 101, for medium
examples p = 15 486 869 and for difficult examples p = 2 276 856 017. The larger the integer p is, the
more difficult it is to obtain proof of work.
1. Verification (easy). Write a function verification(x,y) that returns “True” if x is the solution
to the problem x 2 = y (mod p) and “False” otherwise.
Check that x = 6 543 210 is the solution when y = 8 371 779 and p = 15 486 869. Display the
calculation time required for this verification. (See the explanations after this activity.)
2. Search for a solution (difficult). To find a solution x, there is really no other choice for us than
to test all x starting with x = 0, x = 1. . . Program a square_root(y) function that returns a
solution x to the problem for a given y (or None if there is no solution).
• For p = 101 and y = 17, find x such as x 2 = y (mod p).
• For p = 15 486 869 and y = 8 371 779, you must find the x of the first question. How long
did the search take?
• For p = 15 486 869 and y = 13 017 204, find x such as x 2 = y (mod p).
Conclusion: we found a problem that was difficult to solve, but for which it is easy to check that a
BITCOIN 156

given solution is suitable. For higher values of p, the search for a solution x can be much too long and
not succeed. We’ll see how we can adjust the difficulty of the problem.
3. Instead of looking for an exact solution to our problem x 2 = y (mod p), which is equivalent to
x 2 − y (mod p) = 0, we are looking for an approximate solution, i.e. one that checks:
x 2 − y (mod p) 6 m.
For example if m = 5, then you can have (modulo p) : x 2 − y = 0, x 2 − y = 1, x 2 − y = 2,. . . or
x 2 − y = 5. Of course there are now many possible solutions x
Program a function approximate_square_root(y,margin) that finds an approximate solution
to our problem x 2 = y (mod p).
How long does it take to find one solution to the problem when p = 15 486 869, y = 8 371 779
and m = 20? Choose a prime number p large enough and find a margin of error m so that finding
a solution to the approximate problem requires about 30 to 60 seconds of calculations (for any y).
Here are some examples of prime numbers you can use for your tests:
101 1097 10651 100109 1300573 15486869
179426321 2276856017 32416187567

Lesson 1 (Stopwatch).
The time module allows you to measure the execution time but also to know the date and time (see also
the timeit module explained in the “Arithmetic – While loop – I”). Here is a small script to measure the
calculation time of an instruction.
import time

start_chrono = time.time()

time.sleep(2)

end_chrono = time.time()

total_chrono = end_chrono - start_chrono

print("Execution time (in seconds):", total_chrono)
Explanations.
• The module is named time.
• The function that returns the time is also named time(), so we call it by time.time()!
• The function time() returns a number that we are not interested in. What interests us is the
difference between two values of this function.
• start_chrono = time.time() is like starting a stopwatch.
• end_chrono = time.time() is like stopping it.
• total_chrono = end_chrono - start_chrono is the total time taken by our script in sec-
onds.
• The time.sleep(duration) function pauses the program for a certain amount of seconds.
BITCOIN 157

Lesson 2 (Bitcoin and blockchain).

The bitcoin is a dematerialized currency. Transactions are recorded in a ledger called blockchain. We will
build a (very simplified) model of such an account book.
Imagine a group of friends who want to share the group’s expenses in the simplest possible way. At the
beginning everyone has 1000 bitcoins and everyone’s expenses and income are recorded as they go along.
At the end of the holidays, they will regularize the situation.
The list of expenses/revenues is noted in the account book, for example:
• “Amir spent 100 bitcoins”
• “Barbara received 45 bitcoins”
• etc.
Just look through the account book to see how much everyone has received or spent since the beginning.

To prevent someone from falsifying the account book, after each transaction a certification based on a
proof of work is added to the book. Here is what is written in the ledger:
1. We start with any initial proof of work. For us it will be [0,0,0,0,0,0].
2. The first transaction is written (for example "Amir -100").
3. A proof of work is calculated and written in the book, which will serve as a certificate. It’s a list
(for example [56,42,10,98,2,34]) obtained after many calculations taking into account the
previous transaction and the previous proof of work.
4. For each new transaction (for example "Barbara +45"), someone calculates a proof of work for
the last transaction associated with the previous proof. We write the transaction, then the proof of
work.

proof

proof transaction 3

proof transaction 2

initial proof transaction 1

The proof of work that is calculated depends on the previous transaction but also on the previous proof
of work, which itself depends on the previous data... Thus, each new proof of work actually depends on
everything that has been written since the beginning (even if the calculation explicitly refers only to the
last two entries).
Someone who would like to fake a past transaction should recalculate all the proofs of work that come
next. This is not possible for a single person: there are several proofs of work to calculate and each proof
requires a lot of calculations.

Activity 2 (Tools for lists).

Goal: build useful functions to manipulate lists of integers for the following activities.
BITCOIN 158

Our lists are lists of integers between 0 and 99.

1. Program a function addition(mylist1,mylist2) that adds term to term
and modulo 100, the elements of two lists of the same length. For example,
addition([1,2,3,4,5,6],[1,1,1,98,98,98]) returns [2,3,4,2,3,4].
2. We will look for lists that start with zeros (or zeros and then rather small numbers). A list mylist
is smaller than the max_list list if each element of mylist is less than or equal to each element
of the same rank of max_list.
For example, the list [0,0,1,2,3,4] is smaller than the list [0,0,5]. This is not the case with
the [0,10,0,1,1] list. Another example: being smaller than the list [0,0,0] means starting
with three zeros. Being smaller than the list [0,0,1] means starting with [0,0,0] or [0,0,1].
Program a function is_smaller(mylist,max_list) that returns “True” when mylist is
smaller than max_list.
3. We will need to transform a sentence into a list of numbers. In addition, we will split our lists into
blocks of size N (with N = 6), hence we add zeros at the beginning of the list so that its length is a
multiple of N .
Write a function sentence_to_list(sentence) that converts a string into a list of integers
between 0 and 99 and if necessary adds leading zeros so that the list has the correct size.
The formula to use to convert a character to an integer strictly less than 100 is:
ord(c) % 100
For example: if sentence = "Be happy!" then the function returns:
[0, 0, 0, 66, 1, 32, 4, 97, 12, 12, 21, 33]
The character "e" has ASCII/unicode code 101 so, modulo 100, the number is 1. Note that the
function adds three 0’s at the beginning of the list to have a length that is a multiple of N = 6.

Activity 3 (Hash function).

Goal: create a hash function. From a long message we calculate a short footprint. It’s hard to find
two different messages with the same footprint.
In this activity, our message is a list of integers (between 0 and 99) of any length that is a multiple of
N = 6. We transform it into a list of length N = 6: its footprint, named hash. Here are some examples of
what our hash function will do:
• the list [1, 2, 3, 4, 5, 6, 1, 2, 3, 4, 5, 6] has a hash:
[10, 0, 58, 28, 0, 90]
• the list [1, 1, 3, 4, 5, 6, 1, 2, 3, 4, 5, 6] has a hash:
[25, 14, 29, 1, 19, 6]
The idea is to mix the numbers per block of N = 6 integers, then combine this block with the next and
start again, until you get a single block.
1. One round.
For a block [b0 , b1 , b2 , b3 , b4 , b5 ] of size N = 6, doing a round consists in making the following
operations:
(a) We add some integers:
[b00 , b10 , b20 , b30 , b40 , b50 ] = [b0 , b1 + b0 , b2 , b3 + b2 , b4 , b5 + b4 ]
(b) We multiply these integers by prime numbers (in order 7, 11, 13, 17, 19, 23) and add 1:
[b000 , b100 , b200 , b300 , b400 , b500 ] = [7× b00 +1, 11× b10 +1, 13× b20 +1, 17× b30 +1, 19× b40 +1, 23× b50 +1]
(c) We make a circular permutation (the last one goes first):
[b0000 , b1000 , b2000 , b3000 , b4000 , b5000 ] = [b500 , b000 , b100 , b200 , b300 , b400 ]
BITCOIN 159

(d) We reduce each integer modulo 100 to get integers between 0 and 99.

b0 b1 b2 b3 b4 b5

+ + +
(a) additions

×11 + 1 ×17 + 1 ×23 + 1

×7 + 1 ×13 + 1 ×19 + 1 (b) multiplications

(c) circular permutation

(d) modulo 100

Starting from the block [0, 1, 2, 3, 4, 5], we have successively:

(a) additions: [0, 1, 2, 5, 4, 9]
(b) multiplications: [7 × 0 + 1, 11 × 1 + 1, 13 × 2 + 1, 17 × 5 + 1, 19 × 4 + 1, 23 × 9 + 1] =
[1, 12, 27, 86, 77, 208]
(c) permutation: [208, 1, 12, 27, 86, 77]
(d) reduction modulo 100: [8, 1, 12, 27, 86, 77]
Program such a function one_round(block) which returns the transformation of the block after
these operations. Verifies that the block [1, 1, 2, 3, 4, 5] is transformed into [8, 8, 23, 27, 86, 77].
2. Ten rounds.
To mix each block well, program a ten_rounds(block) function that iterates ten times the
previous operations. After 10 rounds:
• the block [0, 1, 2, 3, 4, 5] becomes [98, 95, 86, 55, 66, 75],
• the block [1, 1, 2, 3, 4, 5] becomes [18, 74, 4, 42, 77, 42].
Two nearby blocks are transformed into two very different blocks!
3. Hash of a list.
Starting from a list of length a multiple of N = 6, it is split into blocks of length 6 and the hash of
this list is calculated according to the following algorithm:
• The first block of the list is extracted, then mixed by 10 rounds.
• We add term to term (and modulo 100), the result of this mix to the second block.
• We start again from the new second block.
• When there is only one block left, we do 10 rounds, the result is the hash of the list.
Here is the diagram of a situation with three blocks: first there are three blocks (A, B, C); secondly
there are only two blocks (B’ and C) left; at the end there is only one block (C”): this is the hash!
BITCOIN 160

A B C list
block 1 block 2 block 3

10 rounds
addition

A’

B’ C

10 rounds
addition

B”

C’

10 rounds

C” hash

Example with the list [0, 1, 2, 3, 4, 5, 1, 1, 1, 1, 1, 1, 10, 10, 10, 10, 10, 10].
• The first block is [0, 1, 2, 3, 4, 5], its mix after 10 rounds is [98, 95, 86, 55, 66, 75].
• This mix is added to the second block [1, 1, 1, 1, 1, 1] (see function addition() in activity
2).
• The remaining list is now [99, 96, 87, 56, 67, 76, 10, 10, 10, 10, 10, 10].
• Let’s do it again. The new first block is [99, 96, 87, 56, 67, 76], its mix after 10 rounds is
[60, 82, 12, 94, 6, 80], it is added to the last block [10, 10, 10, 10, 10, 10] to get (modulo 100)
[70, 92, 22, 4, 16, 90].
• A last mix is made with 10 rounds to obtain the hash: [77, 91, 5, 91, 89, 99].
Program a bithash(mylist) function that returns the fingerprint of a list. Test it on the examples
given at the beginning of the activity.

Activity 4 (Proof of work - Mining).

Goal: build a proof of work mechanism using our hash function.

We are going to build a complicated problem to solve, for which, if someone gives us the solution, then it
is easy to check that it is correct.
BITCOIN 161

Problem to solve. We are given a list, it is a matter of finding a block such that, when you add it to the
list, it produces a hash starting with zeros. More precisely, given a list mylist and a maximum target of
max_list, we have to find a block proof which, concatenated to the list and then hashed, is smaller
than the list of max_list, i.e.:
bithash(mylist + proof) smaller than max_list
The list is of any length (a multiple of N = 6), the proof is a block of length N , the objective is to find a
list starting with 0 (see activity 2).
For example: let mylist = [0,1,2,3,4,5] and max_list = [0,0,7]. Which block proof can I
concatenate to mylist to solve my problem?
• proof = [12,3,24,72,47,77] is appropriate because concatenated to our list it gives
[0,1,2,3,4,5,12,3,24,72,47,77] and the hash of this whole list gives [0,0,5,47,44,71]
which starts with [0,0,5] smaller than the target.
• proof = [0,0,2,0,61,2] is also suitable because after concatenation we have
[0,1,2,3,4,5,0,0,2,0,61,2] whose hash gives [0,0,3,12,58,92].
• [97,49,93,87,89,47] is not suitable, because after concatenation and hashing we get
[0,0,8,28,6,60] which is greater than the desired objective.

1. Verification (easy).
Program a function verification_proof_of_work(mylist,proof) that returns true if the
proposed solution proof is suitable for mylist. Use the is_smaller() function of activity 2.
2. Search for a solution (difficult).
Program a function proof_of_work(mylist) that looks for a block proof solution to our
problem for the given list.
Hints.
• The easiest method is to take a block proof of random numbers and start again until a
solution is found.
• You can also systematically test all blocks starting with [0,0,0,0,0,0], then
[0,0,0,0,0,1]. . . and stop at the first appropriate one.
• You adjust the difficulty of the problem by changing the objective: easy with Max =
[0,0,50], medium with max_list = [0,0,5], difficult with max_list = [0,0,0], too
difficult with max_list = [0,0,0,0].
• As there are several solutions, you do not necessarily get the same solution for each search.
3. Calculation time.
Compare the calculation time of a simple check against the time of searching for a solution. Choose
the objective max_list so that the search for a proof of work requires about 30 to 60 seconds of
calculations.
For the bitcoin, those who calculate proofs of work are called the minors. The first one to find a proof
wins a reward. The difficulty of the problem is adjusted so that that the calculation time taken by the
winner (among all the minors) to find a solution is about 10 minutes.

Activity 5 (Your bitcoins).

Goal: create an account book (called blockchain for the bitcoin) that records all transactions, this
register is public and certified. It is practically impossible to falsify a transaction that has already
been registered.
BITCOIN 162

proof

proof transaction 3 hash

proof transaction 2 hash

initial proof transaction 1 hash

1. Initialization and addition of a transaction.

(a) Initialize a global variable blockchain which is a list and contains at the beginning a zero
proof: blockchain = [ [0,0,0,0,0,0] ] .
(b) A transaction is a string including a name and the amount to add (or deduct) to your account.
For example "Abel +25" or "Barbara -45".
Program a function add_transaction(transaction) that adds the string transaction
to the list blockchain. For example after initialization add_transaction("Camille
+100"), blockchain is [ [0,0,0,0,0,0], "Camille +100" ]. Attention, to be able
to modify blockchain you must start the function with: global blockchain.
2. As soon as a transaction is added, a proof of work must be calculated and added to the account
book. Program a mining() function, without parameters, that adds a proof of work to the book.
Here’s how to do it:
• We take the last transaction transaction, we transform it into a list of integers by the
function sentence_to_list() of activity 2.
• We also need prev_proof, the previous proof of work just before this transaction.
• We form the list mylist composed first of elements of prev_proof, then of elements of the
list of integers obtained by converting the string transaction.
• A proof of work is calculated from this list.
• This proof is added to the account book.
For example, if the book ends with:
[3,1,4,1,5,9], "Abel +35"
then after calculation of the proof of work the book ends with, for example:
[3,1,4,1,5,9], "Abel +35", [32,17,37,73,52,90]
It should be remembered that proof of work is not unique and that it also depends on the objective
max_list.
Only one person at a time adds a proof of work. However, everyone has the opportunity to verify that
the proposed proof is correct (and should do it). Write a function verification_blockchain(),
without parameters, that checks that the last proof added to the blockchain is valid.
3. Write a blockchain that corresponds to the following data:
• We take max_list = [0,0,5] and start with blockchain = [ [0,0,0,0,0,0] ].
• "Alfred -100" (Alfred owes 100 bitcoins).
• Barnabe receives 150.
• Chloe wins 35 bitcoins.
BITCOIN 163

Conclusion: let’s imagine that Alfred wants to cheat, he wants to change the account book in order to
receive 100 bitcoins instead of having a debit, so he has to change the transaction concerning him to
"Alfred +100" but then he has to recalculate a new proof of work which is complicated, moreover he
has to recalculate the proof of the Barnabe transaction and also that of the Chloe transaction!
Someone who wants to modify a transaction must modify all the following proofs of work. If each
proof requires sufficient computing time this is impossible. For the bitcoin each proof requires a lot of
calculations (too much for a single person) and a new proof is to be calculated every 10 minutes. It is
therefore impossible for a person to modify a past transaction.

The other aspect of the bitcoin that we didn’t address is to make sure that each person involved is who
they are, so that no one can get the money of somebody else. This is made possible by private/public key
cryptography (RSA system). Each account is identified by a public key (two very large integers), which
guarantees anonymity. But above all, only the one who has the private key to the account (a large integer)
can access his bitcoins.
 Chapter

23
Random blocks

You will program two methods to build figures that look like algae or corals. Each figure is made up
of small randomly thrown blocks that stick together.

Lesson 1 (Fall of blocks).

Square blocks are dropped into a grid, on the principle of the game “Connect 4”: after choosing a column,
a block falls from top to bottom. The blocks are placed on the bottom of the grid or on other blocks or
next to other blocks. There is a big difference with the game “Connect 4”, here the blocks are “sticky”, i.e.
a block stays stuck as soon as it meets a neighbor on the left or right.
Here is an example of how to throw blocks:
0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5

1. Block fallen from col. 1 2. Block fallen from col. 4 3. Block fallen from col. 1

0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5

4. Block fallen from col. 2 5. Block fallen from col. 2 6. Block fallen from col. 3

For example, in step 4, the block thrown in the column 2 does not go down to the bottom but remains
hung on its neighbor, so it is permanently suspended.
The random throwing of hundreds of blocks on a large grid produces pretty geometric shapes resembling
algae.
RANDOM BLOCKS 165

Activity 1 (Fall of blocks).

Goal: program the drop of the blocks (without graphic display).

The workspace is modeled by an array of n rows and p columns. At the beginning the array only contains
0’s; then the presence of a block is represented by 1.
Here is how to initialize the table:
array = [[0 for j in range(p)] for i in range(n)]
The table is modified by instructions of the type:
array[i][j] = 1
Here is an example of a table (left) to represent the graphical situation on the right (the block at the top
right is falling).

index j

j=0 j=1 j=2 j=3 j=4 j=5

i=0
0 0 0 0 1 0
i=1
0 0 1 0 0 0
index i
0 0 1 0 0 0 i=2
0 0 1 1 0 0
i=3

An array with 5 blocks (n = 4, p = 6)

1. Program a function can_fall(i,j) that determines if the block in position (i, j) can drop to the
square below or not.
Here are the cases in which the block cannot fall:
• if the block is already on the last line,
• if there is a block just below,
• if there is a block just to the right or just to the left.
2. Program a function drop_one_block(j) which drops a block in the j column until it can no
longer go down. This function modifies the array.
For example, here is the table before (left) and after (right) dropping a block in the j = 3 column.
RANDOM BLOCKS 166

0 0 0 0 0 0 0 0 0 0 0 0
0 0 1 0 0 0 0 0 1 1 0 0
0 0 1 0 0 0 0 0 1 0 0 0
0 0 1 1 0 0 0 0 1 1 0 0

0 1 2 3 4 5 0 1 2 3 4 5

Before After

3. Program a function drop_blocks(k) that launches k blocks one by one, each time choosing a
random column (i.e. an integer j with 0 6 j < p).
Here is an example of a table obtained after throwing 10 blocks:

0 0 0 1 1 0
0 1 1 1 0 0
0 0 1 0 0 1
0 0 1 1 0 1

Throw of 10 blocks

Activity 2 (Falling blocks (continued)).

Goal: program the graphic display of the blocks.

Static display. Program the graphical display of blocks from an array.

Hints.
• Use the module tkinter, see chapter “Statistics – Data visualization”.
• You can add a button that launches a block (or several at once).
RANDOM BLOCKS 167

Dynamic display (optional and difficult). Program the display of falling blocks.

Hints.
• It’s much more complicated to program, but very nice to see!
• For moving the blocks, use the program “Moves with tkinter” at the end of this chapter.
• How to make a “block rain”? On a regular basis (for example every tenth of a second) all existing
blocks are dropped by one square and a new one appears on the top line.

Lesson 2 (Brownian trees).

Here is a slightly different construction, much longer to calculate, but which also draws pretty figures
called “brownian trees”.

The principle is as follows:

• We start from a grid (this time we have to imagine that it is drawn flat on a table). In its center, we
place a first fixed block, the seed.
• A random new block is created on the grid. At each step, this block moves at random to one of the
eight adjacent squares, we are talking about a brownian movement.
• As soon as this block touches another block from one side, it sticks to it and no longer moves.
• If the block leaves the grid, it disintegrates.
• Once the block has been glued or disintegrated, a new block is then restarted from a random point
on the grid.
RANDOM BLOCKS 168

The seed A new block and its

Random move of the block
8 possible moves

A block leaves the grid 10 blocks

Gradually, we obtain a kind of tree that looks like coral. The calculations are very long because many
blocks come out of the grid or take a long time to fix (especially at the beginning). In addition, the blocks
can only be thrown one by one.

Activity 3 (Brownian trees).

Goal: program the creation of a brownian tree.

Part 1.
1. Model the workspace again with an array of n rows and p columns containing 0 or 1. Initialize all
values to 0, except 1 in the center of the table.
2. Program a function is_inside(i,j) which determines if the position (i, j) is in the grid (other-
wise the block is coming out).
3. Program a function can_move(i,j) that determines if the block in position (i, j) can move (the
function returns “True”) or if it is hung on (the function returns “False”).
4. Program a function launch_one_block(), without parameters, that simulates the creation of a
block and its random movement, until it sticks or leaves the grid.
Hints.
• The block is created at a random position (i, j) of the grid.
• As long as the block is in the grid and free to move:
– you choose a horizontal move by randomly drawing an integer from {−1, 0, +1}, the
same for a vertical move;
– you move the block according to the combination of these two movements.
• Then modify the table.
5. End with a function launch_blocks(k) that launches k blocks.
Second part.
RANDOM BLOCKS 169

Program the graphic display using tkinter. You can add a button that launches 10 blocks at once.

Lesson 3 (Moves with “tkinter”).

Here is a program that moves a small square and bounces it off the edges of the window.

Here are the main points:

• An object rect is defined, it is a global variable, as well as its coordinates x0, y0.
• This object is (a little bit) moved by the function mymove() which shifts the rectangle by (dx,dy).
• The key point is that this function will be executed again after a short period of time. The command:
canvas.after(50,mymove)
requests a new execution of the function mymove() after a short delay (here 50 milliseconds).
• The repetition of small shifts simulates movement.
from tkinter import *

the_width = 400
the_height = 200

root = Tk()
canvas = Canvas(root, width=the_width, height=the_height, background="white")
canvas.pack(fill="both", expand=True)
RANDOM BLOCKS 170

# Coordinates and speed

x0, y0 = 100,100
dx = +5 # Horizontal speed
dy = +2 # Vertical speed

# The rectangle to move

rect = canvas.create_rectangle(x0,y0,x0+20,y0+20,width=2,fill="red")

# Main function
def mymove():
global x0, y0, dx, dy

x0 = x0 + dx # New abscissa
y0 = y0 + dy # New ordinate

canvas.coords(rect,x0,y0,x0+20,y0+20) # Move

if x0 < 0 or x0 > the_width:

dx = -dx # Change of horizontal direction
if y0 < 0 or y0 > the_height:
dy = -dy # Change of vertical direction

canvas.after(50,mymove) # Call after 50 milliseconds

return

# Function for the button

def action_move():
mymove()
return

# Buttons
button_move = Button(root,text="Move", width=20, command=action_move)
button_move.pack(pady=10)

button_quit = Button(root,text="Quit", width=20, command=root.quit)

button_quit.pack(side=BOTTOM, pady=10)

root.mainloop()
PA R T V

GUIDES

171
 Chapter

Python survival guide

24

1. Test and loops

1.1. If ... then ...

keyword "if"

a condition

colon

if condition :
instruction_1 indented block
instruction_2 will be executed only
... if the condition is true
other instructions
program continuation

1.2. If ... then ... else ...

if condition :
instruction this block will be executed
instruction if the condition is true
...
else:
instruction this block will be executed
...
if the condition is not true
other instructions
PYTHON SURVIVAL GUIDE 173

1.3. Loop for

keywords "for" and "in"

a variable
list to browse, here 0, 1, 2, . . . , n − 1
colon
for i in range(n) :
instruction_1 indented block of instructions
instruction_2 will be repeated n times
... for i = 0, then i = 1. . . up to i = n − 1
other instructions
indentation
program continuation

1.4. Loop while

keyword "while"

a condition

colon

while condition :
instruction_1 indented block will
instruction_2 be executed as long as
... the condition is true
other instructions
program continuation

1.5. Quit a loop

The command break immediately leaves a loop “while” or a loop “for”.

2. Data type
• int Integer. Examples: 123 or -15.
• float Floating point (or decimal) number. Examples: 4.56, -0.001, 6.022e23 (for 6.022×1023 ),
4e-3 (for 0.004 = 4 × 10−3 ).
• str Character or string. Examples: 'Y', "k", 'Hello', "World!".
• bool Boolean. True or False.
• list List. Example: [1,2,3,4].
PYTHON SURVIVAL GUIDE 174

3. Define functions

3.1. Definition of a function

keyword "def"
well-chosen name
parentheses
colon
def my_function () :
instruction_1
instruction_2 instructions
...
return
end of the function

3.2. Function with parameter

Functions achieve their full potential with:
• an input, which groups variables that serve as parameters,
• an output, which is a result returned by the function (and which will often depend on the input
parameters).

# Definition of the function

def my_function (param) :
instruction_1
instruction_2 a parameter
...
return result
returns the result

# Call of the function

x = 7
val = my_function (x)
argument

call of the function

returned result

3.3. Function with several parameters

There can be several input parameters, there can be several output results.
PYTHON SURVIVAL GUIDE 175

one or several parameters

def my_function (param1,param2) :

""" My function does this and that. """
if param1 == param2: immediate termination of the
return 0,0,0
function in this case
other instructions...
return value1, value2, value3

returns several values

function documentation

Here is an example of a function with two parameters and two outputs.

def sum_product(a,b):
""" Computes the sum and product of two numbers. """
s = a + b
p = a * b
return s, p

# Call of the function

mysum, myprod = sum_product(6,7) # Results
print("The sum is:",mysum) # Display
print("The product is:",myprod) # Display
• Very important! Do not confuse displaying and returning a value. The display (by the command
print()) just displays something on the screen. Most functions do not display anything, but return
one (or more) value. This is much more useful because this value can be used elsewhere in the
program.
• As soon as the program meets the instruction return, the function stops and returns the result. There
may be several times the instruction return in a function but only one will be executed. It is also
possible not to put an instruction return if the function returns nothing.
• In the instructions of a function, you can of course use other functions.

3.4. Comments and docstring

• Comment. Anything following the hash sign # is a comment and is ignored by Python. For example:
# Main loop
while r != 0: # While this number is not zero
r = r - 1 # Decrease it by one
• Docstring. You can describe what a function does by starting it with a docstring, i.e. a description in
English, surrounded by three quotation marks. For example:
def product(x,y):
""" Compute the product of two numbers
Input: two numbers x and y
Output: the product of x and y """
PYTHON SURVIVAL GUIDE 176

p = x * y
return p

3.5. Local variable

Here is a very simple function that takes a number as an input and returns the number increased by one.

def my_function(x):
x = x + 1
return x

• Of course my_function(3) returns 4.

• If I define a variable by y = 5 then my_function(y) returns 6. And the value of y has not changed,
it is still equal to 5.
• Here is the delicate situation that you must understand:

x = 7
print(my_function(x))
print(x)

– The variable x is initialized to 7.

– The call of the function my_function(x) is therefore the same as my_function(7) and
logically returns 8.
– What is the value of x at the end? The variable x is unchanged and is still equal to 7! Even if in
the meantime there has been an instruction x = x + 1. This instruction changed the x inside
the function, but not the x outside the function.

• Variables defined within a function are called local variables. They do not exist outside the function.
• If there is a variable in a function that has the same name as a variable in the program (like the x in
the example above), it is as if there were two distinct variables; the local variable only exists inside
the function.
To understand the scope of the variables, you can color the global variables of a function in red, and the
local variables with one color per function. The following small program defines a function that adds one
and another that calculates the double.
PYTHON SURVIVAL GUIDE 177

global variable
x = 7
def add_one( x ):
x = x +1
local variable
return x

def double( x ):
x = 2 * x
another local variable
return x

print( x )
print(add_one( x ))
print(double( x ))
print( x )

The program first displays the value of x, so 7, then it increases it by 1, so it displays 8, then it displays
twice as much as x, so 14. The global variable x has never changed, so the last display of x is still 7.

3.6. Global variable

A global variable is a variable that is defined for the entire program. It is generally not recommended to
use such variables but it may be useful in some cases. Let us look at an example.
The global variable, here the gravitational constant, is declared at the beginning of the program as a classic
variable:
gravitation = 9.81
The content of the variable gravitation is now available everywhere. On the other hand, if you want to
change the value of this variable in a function, you must specify to Python that you are aware of modifying
a global variable.
For example, for calculations on the Moon, it is necessary to change the gravitational constant, which is
much lower there.
def on_the_moon():
global gravitation # Yes, I really want to modify this variable!
gravitation = 1.625 # New value for the entire program
...

3.7. Optional arguments

It is possible to give optional arguments. Here is how to define a function (here that would draw a line) by
giving default values:
def draw(length, width=5, color="blue"):
• The command draw(100) traces my line, and as I only specified the length, the arguments width
and color take the default values (5 and blue).
• The command draw(100, width=10) traces my line with a new thickness (the color is the default
one).
PYTHON SURVIVAL GUIDE 178

• The command draw(100, color="red") traces my line with a new color (the thickness is the
default one).
• The command draw(100, width=10, color="red") traces my line with a new thickness and a
new color.
• Here is also what you can use:
– draw(100, 10, "red"): do not specify the names of the options if you pay attention to the
order.
– draw(color="red", width=10, length=100): you can name any variable, the named
variables can pass as arguments in any order.

4. Modules

4.1. Use a module

• from math import * Imports all functions from the module math. You are now able to use the
sine function for example, by sin(0). This is the simplest method and it is the one we use in this
book.
• import math Allows you to use the functions of the module math. You can then access the sine
function with math.sin(0). This is the officially recommended method to avoid conflicts between
modules.

4.2. Main modules

• math contains the main mathematical functions.
• random simulates the random draw.
• turtle the Python turtle, it is some equivalent of Scratch.
• matplotlib allows you to draw graphs and visualize data.
• tkinter allows you to display graphic windows.
• time for date, time and length of time.
• timeit to measure the execution time of a function.
There are many other modules!

5. Errors

5.1. Indentation errors

a = 3
b = 2
Python returns the error message IndentationError: unexpected indent. It also indicates the line number
where the indentation error is located, it even points using the symbol “^” to the exact location of the error.
5.2. Syntax errors

• while x >= 0
x = x - 1
PYTHON SURVIVAL GUIDE 179

Python returns the error message SyntaxError: invalid syntax because the colon is missing after the
condition. It should be while x >= 0 :
• string = Hello world! returns an error because the quotation marks to define the string are
missing.
• print("Hi there" Python returns the error message SyntaxError: unexpected EOF while parsing
because the expression is incorrectly parenthesized.
• if val = 1: Another syntax error, because you would have to write if val == 1:

5.3. Type errors

• Integer
n = 7.0
for i in range(n):
print(i)
Python returns the error message TypeError: ’float’ object cannot be interpreted as an integer. Indeed
7.0 is not an integer, but a floating point number.
• Floating point number
x = "9"
sqrt(x)
Python returns the error message TypeError: a float is required, because "9" is a string and not a
number.
• Wrong number of arguments
gcd(12) Python returns the error message TypeError: gcd() takes exactly 2 arguments (1 given)
because the gcd() function of the math module requires two arguments, such as gcd(12,18).

5.4. Name errors

• if y != 0: y = y - 1 Python returns the message NameError: name ’y’ is not defined if the
variable y has not yet been defined.
• This error can also occur if upper/lower case letters are not scrupulously respected. variable,
Variable and VARIABLE are three different variable names.
• x = sqrt(2) Python returns the message NameError: name ’sqrt’ is not defined, you must import
the module math to be able to use the function sqrt().
• Function not yet defined
product(6,7)

def product(a,b):
return a*b
Returns an error NameError: name ’product’ is not defined because a function must be defined before it
can be used.
PYTHON SURVIVAL GUIDE 180

5.5. Exercise
Fix the code! Python should display 7 5 9.
a == 7
if (a = 2) or (a >= 5)
b = a - 2
c = a + 2
else
b = a // 2
c = 2 * a
print(a b c)

5.6. Other problems

The program starts but stops along the way or doesn’t do what you want? That’s where the trouble starts,
you have to debug the code! There are no general solutions but only a few tips:
• A clean, well structured, well commented code, with well chosen variable and function names, is
easier to read.
• Test your algorithm by hand with paper and pencil for easy cases.
• Do not hesitate to display the values of the variables, to see their evolution over time. For example
print(i,mylist[i]) in a loop.
• A better way to inspect the code is to view the values associated with the variables using the features
debug of your favorite Python editor. It is also possible to make a step by step execution.
• Does the program work with some values and not others? Have you thought about extreme cases? Is
n zero but was not allowed to? Is the list empty, while the program does not handle this case? etc.
Here are some examples.
• I want to display the squares of integers from 1 to 10. The following program does not return any
errors but does not do what I want.
for i in range(10):
print(i ** 2)
The loop iterates on integers from 0 to 9. You have to write range(1,11).
• I want to display the last item in my list.
mylist = [1,2,3,4]
print(mylist[4])
Python returns the error message IndexError: list index out of range because the last element is the
one of rank 3.
• I want to do a countdown. The next program never stops.
n = 10
while n != "0":
n = n - 1
print(n)
With a loop “while” you have to take great care to write the condition well and check that it ends up
being wrong. Here, it is poorly formulated, it should be while n!= 0:
 Chapter

25
Main functions

1. Mathematics
Classical operations
• a + b, a - b, a * b classic operations
• a / b “real” division (returns a floating point number)
• a // b Euclidean division quotient (returns an integer)
• a % b remainder of the Euclidean division, called a modulo b
• abs(x) absolute value
• x ** n power x n
• 4.56e12 for 4.56 × 1012

Module “math”
The use of other mathematical functions requires the module math which is called by the command:
from math import *
p
• sqrt(x) square root x
• cos(x), sin(x), tan(x) trigonometric functions cos x, sin x, tan x in radians
• pi approximate value of π = 3.14159265 . . .
• floor(x) integer just below x
• ceil(x) integer just above x
• gcd(a,b) gcd of a and b

Module “random”
The random module generates numbers in a pseudo-random way. It is called by the command:
from random import *
• random() on each call, returns a floating number x at random satisfying 0 6 x < 1.
• randint(a,b) for each call, returns an integer n at random satisfying a 6 n 6 b.
• choice(mylist) on each call, randomly draws an item from the list.
• mylist.shuffle() mixes the list (the list is modified).

Binary notation
• bin(n) returns the binary notation of the integer n as a string. Example: bin(17) returns
'0b10001'.
• To write a number directly in binary notation, simply write the number starting with 0b (without
quotation marks). For example 0b11011 is equal to 27.
MAIN FUNCTIONS 182

2. Booleans
A boolean is a data that takes either the value True or the value False.

Comparisons
The following comparison tests return a boolean.
• a == b equality test
• a < b strict lower test
• a <= b large lower test
• a > b or a >= b higher test
• a != b non-equality test
Do not confuse “a = b” (assignment) and “a == b” (equality test).

Boolean operations
• P and Q logical “and”
• P or Q logical “or”
• not P negation

3. Strings I
Chains
• "A" or 'A' one character
• "Python" or 'Python' a string
• len(string) the string length. Example: len("Python") returns 6.
• string1 + string2 concatenation.
Example: "I love" + "Python" returns "I lovePython".
• string[i] returns the i-th character of string (numbering starts at 0).
Example with string = "Python", string[1] is equal to "y". See the table below.

Letter P y t h o n
Rank 0 1 2 3 4 5

Number/string conversion
• Chain. str(number) converts a number (integer or floating point number) into a string. Examples:
str(7) returns the string "7"; str(1.234) returns the string "1.234".
• Integer. int(string) returns the integer corresponding to the string. Example: int("45") re-
turns the integer 45.
• Floating point number. float(string) returns the floating point number corresponding to the
string. Example: float("3.14") returns the number 3.14.

Substrings
• string[i:j] returns the substring of characters with a rank of i to j − 1 from string.
Example: with string = "This is a string", string[2:7] returns "is is".
• string[i:] returns characters with a rank of i until the end of string.
Example : string[5:] returns "is a string".
• string[:j] returns characters from the beginning to the rank j − 1 of string. Example:
string[:4] returns "This".
MAIN FUNCTIONS 183

Format
The format() method allows you to format text or numbers. This function returns a string.
• Text

Test Test Test

– '{:10}'.format('Test') left aligned (on 10 characters)
– '{:>10}'.format('Test') right alignment
– '{:^10}'.format('Test') centered
• Integer

456 456 000456

– '{:d}'.format(456) integer
– '{:6d}'.format(456) right aligned (on 6 characters)
– '{:06d}'.format(456) adding leading zeros (on 6 characters)
• Floating point number

3.141593 3.14159265 3.1416 003.1416

– '{:f}'.format(3.14159265653589793) floating point number
– '{:.8f}'.format(3.14159265653589793) 8 decimal places
– '{:8.4f}'.format(3.14159265653589793) on 8 characters with 4 numbers after the
decimal point
– '{:08.4f}'.format(3.141592653589793) adding leading zeros

4. Strings II
Encoding
• chr(n) returns the character associated with the ASCII/unicode code number n. Example: chr(65)
returns "A"; chr(97) returns "a".
• ord(c) returns the ASCII/unicode code number associated with the character c. Example:
ord("A") returns 65; ord("a") returns 97.
The beginning of the ASCII/unicode table is given below.
MAIN FUNCTIONS 184

33 ! 43 + 53 5 63 ? 73 I 83 S 93 ] 103 g 113 q 123 {

34 " 44 , 54 6 64 @ 74 J 84 T 94 ^ 104 h 114 r 124 |

35 # 45 - 55 7 65 A 75 K 85 U 95 _ 105 i 115 s 125 }

36 $ 46 . 56 8 66 B 76 L 86 V 96 ‘ 106 j 116 t 126 ~

37 % 47 / 57 9 67 C 77 M 87 W 97 a 107 k 117 u 127 -

38 & 48 0 58 : 68 D 78 N 88 X 98 b 108 l 118 v

39 ’ 49 1 59 ; 69 E 79 O 89 Y 99 c 109 m 119 w

40 ( 50 2 60 < 70 F 80 P 90 Z 100 d 110 n 120 x

41 ) 51 3 61 = 71 G 81 Q 91 [ 101 e 111 o 121 y

42 * 52 4 62 > 72 H 82 R 92 \ 102 f 112 p 122 z

Upper/lower-case
• string.upper() returns a string in upper case.
• string.lower() returns a string in lowercase.

Search/replace
• substring in string returns “true” or “false” depending on if substring appears in string.
Example : "NOT" in "TO BE OR NOT TO BE" returns True.
• string.find(substring) returns the rank at which the substring was found (and -1 otherwise).
Example: with string = "ABCDE", string.find("CD") returns 2.
• string.replace(substring,new_substring) replaces each occurrence of the substring by
the new substring.
Example: with string = "ABCDE", string.replace("CD","XY") returns "ABXYE".

Split/join
• string.split(separator) separates the string into a list of substrings (by default the separator
is the space).
Examples:
– "To be or not to be.".split() returns ['To', 'be', 'or', 'not', 'to', 'be.']
– "12.5;17.5;18".split(";") returns ['12.5', '17.5', '18']
• separator.join(mylist) groups the substrings into a single string by adding the separator
between each.
Examples:
– "".join(["To", "be", "or", "not", "to", "be."]) returns 'Tobeornottobe.'
Spaces are missing.
– " ".join(["To", "be", "or", "not", "to", "be."]) returns 'To be or not to
be.' It’s better when the separator is a space.
– "--".join(["To", "be", "or", "not", "to", "be."]) returns 'To--be--or--not--to--be.'
MAIN FUNCTIONS 185

5. Lists I
Construction of a list
Examples:
• mylist1 = [5,4,3,2,1] a list of integers 5.
• mylist2 = ["Friday","Saturday","Sunday"] a list of 3 strings.
• mylist3 = [] the empty list.
• list(range(n)) list of integers from 0 to n − 1.
• list(range(a,b)) list of integers from a to b − 1.
• list(range(a,b,step)) list of integers from a to b − 1, with a step given by the integer step.

Get an item
• mylist[i] returns the element of rank i. Be careful, the rank starts at 0.
Example: mylist = ["A","B","C","D","E","F"] then mylist[2] returns "C".

Letter "A" "B" "C" "D" "E" "F"

Rank 0 1 2 3 4 5

• mylist[-1] returns the last element, mylist[-2] returns the second last element. . .
• mylist.pop() removes the last item from the list and returns it.

Add one element (or more)

• mylist.append(element) adds the item at the end of the list. Example: if mylist =
[5,6,7,8] then mylist.append(9) adds 9 to the list, mylist is now [5,6,7,8,9].
• new_mylist = mylist + [element] provides a new list with an extra element at the end.
Example: [1,2,3,4] + [5] is [1,2,3,4,5].
• [element] + mylist returns a list where the item is added at the beginning. Example: [5] +
[1,2,3,4] is [5,1,2,3,4].
• mylist1 + mylist2 concatenates the two lists. Example: with mylist1 = [4,5,6] and
mylist2 = [7,8,9] then mylist1 + mylist2 is [4,5,6,7,8,9].

Example of construction. Here is how to build the list that contains the first squares:

list_squares = [] # We start from an empty list

for i in range(10):
list_squares.append(i**2) # We add squares one by one

At the end list_squares is:

[0, 1, 4, 9, 16, 25, 36, 49, 64, 81]

Browse a list
• len(mylist) returns the length of the list. Example: len([5,4,3,2,1]) returns 5.
• Just browse a list (and here display each item):
for element in mylist:
print(element)
• Browse a list using the rank.
MAIN FUNCTIONS 186

n = len(mylist)
for i in range(n):
print(i,mylist[i])

6. Lists II
Mathematics
• max(mylist) returns the largest element. Example: max([10,16,13,14]) returns 16.
• min(mylist) returns the smallest element. Example: min([10,16,13,14]) returns 10.
• sum(mylist) returns the sum of all elements. Example: sum([10,16,13,14]) returns 53.

Slicing lists
• mylist[a:b] returns the sublist of elements from the rank a to the rank b − 1.
• mylist[a:] returns the list of elements of rank a until the end.
• mylist[:b] returns the list of items from the beginning to the rank b − 1.

Letter "A" "B" "C" "D" "E" "F" "G"

Rank 0 1 2 3 4 5 6

For example if mylist = ["A","B","C","D","E","F","G"] then:

• mylist[1:4] returns ["B","C","D"].
• mylist[:2] is like mylist[0:2] and returns ["A","B"].
• mylist[4:] returns ["E","F","G"]. It’s the same thing than mylist[4:n] where n =
len(mylist).

Find the rank of an element

• mylist.index(element) returns the first position at which the item was found. Example: with
mylist = [12, 30, 5, 9, 5, 21], mylist.index(5) returns 2.
• If you just want to know if an item belongs to a list, then the statement :
element in mylist
returns True or False. Example: with mylist = [12, 30, 5, 9, 5, 21], “9 in mylist” is
true, while “8 in mylist” is false.

Order
• sorted(mylist) returns the ordered list of items.
Example: sorted([13,11,7,4,6,8,12,6]) returns the list [4,6,6,7,8,11,12,13].
• mylist.sort() does not return anything but the list mylist is now ordered.

Invert a list
Here are three methods:
• mylist.reverse() modifies the list in place;
• list(reversed(mylist)) returns a new list;
• mylist[::-1] returns a new list.

Delete an item
Three methods.
MAIN FUNCTIONS 187

• mylist.remove(element) deletes the first occurrence found.

Example: mylist = [2,5,3,8,5], the instruction mylist.remove(5) modifies the list which is
now [2,3,8,5] (the first 5 has disappeared).
• del mylist[i] deletes element at rank i (the list is modified).
• element = mylist.pop() removes the last item from the list and returns it.

List comprehension
• Let’s start from a list, for example mylist = [1,2,3,4,5,6,7,6,5,4,3,2,1].
• list_doubles = [ 2*x for x in mylist ] returns a list that contains the doubles of the
items of mylist. So this is the list [2,4,6,8,...].
• liste_squares = [ x**2 for x in mylist ] returns the list of squares of the items in the
list mylist. So this is the list [1,4,9,16,...].
• partial_list = [ x for x in mylist if x > 2 ] extracts from the list only the elements
greater than 2. So this is the list [3,4,5,6,7,6,5,4,3].

List of lists
Example:
array = [ [2,14,5], [3,5,7], [15,19,4], [8,6,5] ]
corresponds to the table:

index j

j=0 j=1 j=2

i=0 2 14 5

i=1 3 5 7
index i
i=2 15 19 4

i=3 8 6 5

Then array[i] returns the sublist of rank i, and array[i][j] returns the element located in the sublist
number i, at rank j of this sublist. For example:
• array[0] returns the sublist [2,14,5].
• array[1] returns the sublist [3,5,7].
• array[0][0] returns the integer 2.
• array[0][1] returns the integer 14.
• array[2][1] returns the integer 19.
A table of n rows and p columns.
• array = [[0 for j in range(p)] for i in range(n)] initializes an array and fills it
with 0.
• array[i][j] = 1 modifies a value in the table (the one at the location (i, j)).
MAIN FUNCTIONS 188

7. Input/output
Display
• print(string1,string2,string3,...) displays strings or objects. Example: print("Value
=",14) displays Value = 14. Example: print("Line 1 \n Line 2") displays on two lines.
• Separator. print(...,sep="...") changes the separator (by default the separator is the space
character). Example: print("Bob",17,13,16,sep="; ") displays Bob; 17; 13; 16.
• End of line. print(...,end="...") changes the character placed at the end (by default it is
the line break \n). Example print(17,end="") then print(76) displays 1776 on a single line.

Keyboard entry
input() pauses the program and waits for the user to send a message on the keyboard (end by pressing
the “Enter” key). The message is a string.
Here is a small program that asks for the user’s first name and age and displays a message like “Hello Kevin”
then “You are a minor/adult” according to age.
first_name = input ("What's your name? ")
print("Hello",first_name)

age_str = input("How old are you? ")

age = int(age_str)

if age >= 18:

print("You're an adult!")
else:
print("You're a minor!")

8. Files
Order
• fi = open("my_file.txt","r") opening in reading ("r" = read).
• fi = open("my_file.txt","w") opening in writing ("w" = write). The file is created if it does
not exist, if it existed the previous content is first deleted.
• fi = open("my_file.txt","a") opening for writing, the data will be written at the end of the
current data ("a" = append).
• fi.write("one line") write to the file.
• fi.read() reads the whole file (see below for another method).
• fi.readlines() reads all the lines (see below for another method).
• fi.close() file closing.

Write lines to a file

fi = open("my_file.txt","w")

fi.write("Hello world!\n")

line = "Hi there.\n"

fi.write(line)
MAIN FUNCTIONS 189

fi.close()

Read lines from a file

fi = open("my_file.txt","r")

for line in fi:

print(line)

fi.close()

Read a file (official method)

with open("my_file.txt","r") as fi:
for line in fi:
print(line)

9. Turtle
The turtle module is called by the command:
from turtle import *

Main commands
• forward(length) advances a number of steps
• backward(length) goes backwards
• right(angle) turns to the right (without advancing) at a given angle in degrees
• left(angle) turns left
• setheading(direction) points in one direction (0 = right, 90 = top, −90 = bottom, 180 = left)
• goto(x,y) moves to the point (x, y)
• setx(newx) changes the value of the abscissa
• sety(newy) changes the value of the ordinate
• down() sets the pen down
• up() sets the pen up
• width(size) changes the thickness of the line
• color(col) changes the color: "red", "green", "blue", "orange", "purple". . .
• position() returns the (x, y) position of the turtle
• heading() returns the direction angle to which the turtle is pointing
• towards(x,y) returns the angle between the horizontal and the segment starting at the turtle and
ending at the point (x, y)
• speed("fastest") maximum travel speed
• exitonclick() ends the program as soon as you click

Several turtles
Here is an example of a program with two turtles.
MAIN FUNCTIONS 190

turtle1 = Turtle() # with capital 'T'!

turtle2 = Turtle()

turtle1.color('red')
turtle2.color('blue')

turtle1.forward(100)
turtle2.left(90)
turtle2.forward(100)

10. Matplotlib
With the matplotlib module it is very easy to draw a list. Here is an example.

import matplotlib.pyplot as plt

mylist1 = [3,5,9,8,0,3]
mylist2 = [4,7,7,2,8,9]

plt.plot(mylist1,color="red")
plt.plot(mylist2,color="blue")
plt.grid()
plt.show()
Main functions.
• plt.plot(mylist) traces the points of a list (in the form (i, `i )) that are linked by segments.
• plt.plot(list_x,list_y) traces the points of a list (in the form of (x i , yi ) where x i browses the
first list and yi the second).
• plt.scatter(x,y,color='red',s=100) displays a point in (x, y) (of a size s).
• plt.grid() draws a grid.
• plt.show() displays everything.
• plt.close() exits the display.
MAIN FUNCTIONS 191

• plt.xlim(xmin,xmax) defines the interval for the x.

• plt.ylim(ymin,ymax) defines the interval for the y.
• plt.axis('equal') imposes an orthonormal basis.

11. Tkinter

11.1. Graphics
To display this:

The code is:

# tkinter window
root = Tk()

canvas = Canvas(root, width=800, height=600, background="white")

canvas.pack(fill="both", expand=True)

# A rectangle
canvas.create_rectangle(50,50,150,100,width=2)

# A rectangle with thick blue edges

canvas.create_rectangle(200,50,300,150,width=5,outline="blue")

# A rectangle filled with purple

canvas.create_rectangle(350,100,500,150,fill="purple")

# An ellipse
canvas.create_oval(50,110,180,160,width=4)

# Some text
canvas.create_text(400,75,text="Bla bla bla bla",fill="blue")

# Launch of the window

root.mainloop()
Some explanations:
• The tkinter module allows us to define variables root and canvas that determine a graphic
window (here width 800 and height 600 pixels). Then describe everything you want to add to the
window. And finally the window is displayed by the command root.mainloop() (at the very end).
MAIN FUNCTIONS 192

• Attention! The window’s graphic marker has its y-axis pointing downwards. The origin (0, 0) is the
top left corner (see figure below).
• Command to draw a rectangle: create_rectangle(x1,y1,x2,y2); just specify the coordinates
(x 1 , y1 ), (x 2 , y2 ) of two opposite vertices. The option width adjusts the thickness of the line, outline
defines the color of this line, fill defines the filling color.
• An ellipse is traced by the command create_oval(x1,y1,x2,y2), where (x 1 , y1 ), (x 2 , y2 ) are the
coordinates of two opposite vertices of a rectangle framing the desired ellipse (see figure). A circle is
obtained when the corresponding rectangle is a square!
• Text is displayed by the command canvas.create_text(x,y,text="My text") specifying the
coordinates (x, y) of the point from which you want to display the text.

(0, 0) x

(x 1 , y1 )

(x 2 , y2 )
(x 1 , y1 )

(x 2 , y2 )
y

11.2. Buttons
It is more ergonomic to display windows where actions are performed by clicking on buttons. Here is the
window of a small program with two buttons. The first button changes the color of the rectangle, the second
button ends the program.

The code is:

from tkinter import *
from random import *
MAIN FUNCTIONS 193

root = Tk()
canvas = Canvas(root, width=400, height=200, background="white")
canvas.pack(fill="both", expand=True)

def action_button():
canvas.delete("all") # Clear all
colors = ["red","orange","yellow","green","cyan","blue","violet","purple"]
col = choice(colors) # Random color
canvas.create_rectangle(100,50,300,150,width=5,fill=col)
return

button_color = Button(root,text="View", width=20, command=action_button)

button_color.pack(pady=10)

button_quit = Button(root,text="Quit", width=20, command=root.quit)

button_quit.pack(side=BOTTOM, pady=10)

root.mainloop()
Some explanations:
• A button is created by the command Button. The text option customizes the text displayed on the
button. The button created is added to the window by the method pack.
• The most important thing is the action associated with the button! It is the option command that
receives the name of the function to be executed when the button is clicked. For our example
command=action_button, associate the click on the button with a change of color.
• Attention! You have to give the name of the function without brackets: commande=my_function
and not command=my_function().
• To associate the button with “Quit” and close the window, the argument is command=root.quit.
• The instruction canvas.delete("all") deletes all drawings from our graphic window.

11.3. Text
Here’s how to display text with Python and the graphics window module tkinter.

The code is:

from tkinter import *
from tkinter.font import Font
# tkinter window
root = Tk()
canvas = Canvas(root, width=800, height=600, background="white")
canvas.pack(fill="both", expand=True)
# Font
myfont = Font(family="Times", size=30)
# Some text
canvas.create_text(100,100, text="Text with Python!",
MAIN FUNCTIONS 194

anchor=SW, font=myfont, fill="blue")

# Launch the window
root.mainloop()
Some explanations:
• root and canvas are the variables that define a graphic window (here of width 800 and height 600
pixels). This window is launched by the last command: root.mainloop().
• We remind you that for the graphic coordinates, the y-axis is directed downwards. To define a
rectangle, simply specify the coordinates (x 1 , y1 ) and (x 2 , y2 ) from two opposite vertices (see figure
below).
• The text is displayed by the command canvas.create_text(). It is necessary to specify the coor-
dinates (x, y) of the point from which you want to display the text.
• The text option allows you to pass the string to display.
• The anchor option allows you to specify the text anchor point, anchor=SW means that the text box
is anchored to the Southwest point (SW) (see figure below).
• The fill option allows you to specify the text color.
• The option font allows you to define the font (i.e. the style and size of the characters). Here are
some examples of fonts, it’s up to you to test them:
– Font(family="Times", size=20)
– Font(family="Courier", size=16, weight="bold") in bold
– Font(family="Helvetica", size=16, slant="italic") in italic

(0, 0) x

(x 1 , y1 )

South West anchor

(x 2 , y2 )
y

11.4. Mouse click

Here is a small program that displays a graphic window. Each time the user clicks (with the left mouse
button) the program displays a small square (on the window) and displays “Click at x = . . ., y = . . .” (on
the console).
from tkinter import *

# Window
root = Tk()
canvas = Canvas(root, width=800, height=600, background="white")
canvas.pack(side=LEFT, padx=5, pady=5)

# Catch mouse click

def action_mouse_click(event):
canvas.focus_set()
MAIN FUNCTIONS 195

x = event.x
y = event.y
canvas.create_rectangle(x,y,x+10,y+10,fill="red")
print("Click at x =",x,", y =",y)
return

# Association click/action
canvas.bind("<Button-1>", action_mouse_click)

# Launch
root.mainloop()
Here are some explanations:
• The creation of the window is usual. The program ends with the launch by the command mainloop().
• The first key point is to associate a mouse click to an action, that’s what the line does:
canvas.bind("<Button-1>", action_mouse_click)
Each time the left mouse button is clicked, Python executes the function action_mouse_click.
(Note that there are no brackets for the call to the function.)
• Second key point: the action_mouse_click function retrieves the click coordinates and then does
two things here: it displays a small rectangle at the click location and displays the (x, y) coordinates
in the terminal window.
• The coordinates x and y are expressed in pixels; (0, 0) refers to the upper left corner of the window
(the area delimited by canvas).

11.5. Movement
Here is a program that moves a small square and bounces it off the edges of the window.

Here are the main points:

• An object rect is defined, it is a global variable, as well as its coordinates x0, y0.
• This object is (a little bit) moved by the function mymove() which shifts the rectangle by (dx,dy).
• The key point is that this function will be executed again after a short period of time. The command:
canvas.after(50,mymove)
requests a new execution of the function mymove() after a short delay (here 50 milliseconds).
• The repetition of small shifts simulates movement.
from tkinter import *
MAIN FUNCTIONS 196

the_width = 400
the_height = 200

root = Tk()
canvas = Canvas(root, width=the_width, height=the_height, background="white")
canvas.pack(fill="both", expand=True)

# Coordinates and speed

x0, y0 = 100,100
dx = +5 # Horizontal speed
dy = +2 # Vertical speed

# The rectangle to move

rect = canvas.create_rectangle(x0,y0,x0+20,y0+20,width=2,fill="red")

# Main function
def mymove():
global x0, y0, dx, dy

x0 = x0 + dx # New abscissa
y0 = y0 + dy # New ordinate

canvas.coords(rect,x0,y0,x0+20,y0+20) # Move

if x0 < 0 or x0 > the_width:

dx = -dx # Change of horizontal direction
if y0 < 0 or y0 > the_height:
dy = -dy # Change of vertical direction

canvas.after(50,mymove) # Call after 50 milliseconds

return

# Function for the button

def action_move():
mymove()
return

# Buttons
button_move = Button(root,text="Move", width=20, command=action_move)
button_move.pack(pady=10)

button_quit = Button(root,text="Quit", width=20, command=root.quit)

button_quit.pack(side=BOTTOM, pady=10)

root.mainloop()
 Chapter

26
Notes and references

You will find here comments and readings on each of the activities.

General resources
• The official documentation of Python contains tutorials and explanations of each of the functions.
docs.python.org/3/
• Wikipedia is a reliable source for learning more about some concepts (mainly projects) but the level is
not always suitable for a high school student.
• Internet and in particular the forums often have the answer to the questions you ask yourself!
• The most experienced among you can participate in the Euler project which offers a list of mathematics-
with-computer puzzles. Hang on! Hang on!
projecteuler.net

1. Hello world
Learning a programming language can be very difficult. It’s hard enough to learn alone in your own corner.
It is not uncommon to stay blocked for several hours for a stupid syntax error. You have to start small, don’t
hesitate to copy code already written by others, be persistent and ask for help quickly!

2. Turtle (Scratch with Python)

The ideal is to master Scratch before attacking Python. Python offers a module Turtle that works on the
same principle as Scratch. Of course instead of moving blocks, you have to write the code! It is a good
exercise to transcribe into Python all the activities you can do with Scratch.
NOTES AND REFERENCES 198

3. If ... then ...

We really start the programming with the test “if/else”. The computer therefore acts in one way or another
depending on the situation. It is no longer an automaton that always does the same thing. In addition to
the keyboard entry, we can start having interactive programs.
The classic syntax errors are:
• forget the colon after if condition: or else:,
• incorrectly indent the blocks.
These errors will be reported by Python with the line number where there is a problem (normally your
editor places the cursor on the wrong line). On the other hand, if the program starts but does not do the
right thing, it is surely the condition that is incorrectly formulated. It is more complicated to understand the
right condition: a little logic and reflection with paper and pencil are welcome. Some editors also allow a
step-by-step execution of the Python program.
We will then come back to the “true” and the “false”. There is also the test
if ... elif ... elif ... else ...
which allows you to run the tests in sequence and which is not covered here. Understanding if ...
else ... is enough.

4. Functions
Quite quickly it is necessary to understand the structure of a computer program: the program is broken
down into blocks of definitions and simple actions. Simple actions are grouped into intermediate actions.
And at the end the main program is just to perform some good functions.
The arguments/parameters of a function are a delicate learning process. You can define a function by def
func(x): and call it by func(y). The x corresponds to a dummy mathematical variable. In computing,
we prefer to talk about the scope of the variable which can be local or global.
What we can remember is that nothing that happens within a function is accessible outside the function. You
just have to use what the function returns. It is necessary to avoid the use of global in the first instance.
Last but not least, it is really important to comment extensively on your code and explain what your functions
are doing. If you define a function func(x), plan three lines of comments to (a) say what this function
does, (b) say what input is expected (x must be an integer? a floating point number ? positive?. . . ), (c) say
what the function returns.
It is also necessary to comment on the main points of the code, give well-chosen names to the variables.
You’ll be happy to have a readable code when you come back to your code again later!
A good computer scientist should check that the parameter passed as an argument verifies the expected
hypothesis. For example if func(x) is defined for an integer x and the user transmits a string, then a nice
warning message should be sent to the user without causing the whole program to fail. The commands
assert, try/except allow to manage this kind of problem. For our part, we will refrain from these checks
assuming that the user/programmer uses the functions and variables in good intelligence.
NOTES AND REFERENCES 199

5. Arithmetic – While loop – I

We will only use Python3. If you don’t know which version you have, type 7/2: Python3 returns 3.5
while Python2 returns 3 (in version 2, Python considered that the division of two integers should return
an integer).
With Python3 it is clearer, a / b is the usual division (real numbers) while a // b is the Euclidean
division between two integers and returns the quotient.

6. Strings – Analysis of a text

Handling the strings allows you to do fun activities and leave the mathematical world a little bit. You can
code quizzes, programs that chat with the user. . . Strings are especially a good introduction to the notion of
list, which is an essential tool later on.

7. Lists I
Python is particularly flexible and agile for using lists. The “old” languages often only allowed lists containing
a single type of element, and to browse a list you always had to do this:
for i in range(len(mylist)):
print(mylist[i])
While:
for element in mylist:
print(element)
is much more natural.
Sorting a list is a fundamental operation in computer science. Would you imagine a dictionary containing
60 000 words, but not sorted in alphabetical order? Sorting a list is a difficult operation, there are many sorting
algorithms. The bubble sort algorithm presented here is one of the simplest. Programming faster algorithms
in high school is a great challenge: you have to understand recursivity and the notion of complexity.

8. Statistics – Data visualization

If all readers can program the calculations of sum, mean, standard deviation, median. . . and their visu-
alization, then the objective of this book is achieved! This demonstrates a good understanding of basic
mathematical and computer tools.
The tkinter module allows a graphic display. To begin with, you have to copy the lines of a code that
works without asking yourself too many questions, then adapt it to your needs.
NOTES AND REFERENCES 200

9. Files
The files make it possible to communicate Python with the outside world: for example, you can retrieve a
file of grades to calculate the averages and produce a report card for each student.
The activities on the images are nice and these image formats will be used later, even if this format has the
big disadvantage of producing large files. Nevertheless they are standard and recognized by image software
(Gimp for example). Note that some software writes files with only one data per line (or all data on a single
line). It is a good exercise to implement the reading of all possible formats.

10. Arithmetic – While loop – II

Arithmetic in general and prime numbers in particular are very important in computer science. They are the
foundation of modern cryptography and ensure the security of transactions on the Internet.
The algorithms presented here are of course elementary. There are sophisticated techniques to say in a few
seconds if a number of several hundred digits is prime or not. However, factoring a large integer remains a
difficult problem.
A computer shows its power when it handles a large quantity of numbers or very large numbers. With
arithmetic, we have both at the same time!

11. Binary I
The first difficulty with binary notation is to make a good distinction between a number and writing the
number. We are so used to decimal numeral system that we forgot its origin, 1234 is just 1 × 1000 + 2 ×
100 + 3 × 10 + 4 × 1. As computers work with 0 and 1 you have to be comfortable with binary numeral
system. The transition to binary notation is not very difficult, but it is still better to make a few examples by
hand before starting the programming.
Binary writing will be used for other problems on the following principle: you have 4 switches so 1.0.0.1
means that you activate the first and last switch and not the others.

12. Lists II
Lists are so useful that Python has a whole efficient syntax to manage them: list slicing and list comprehen-
sion. We could of course do without it (which is the case for most other languages) but it would be a pity.
We will also need lists of lists and in particular arrays to display beautiful images.

13. Binary II
There are 10 kinds of people, those who understand binary and others!
NOTES AND REFERENCES 201

14. Probabilities - Parrondo’s paradox

Here is a project with some probabilities and a nice paradox, it’s surprising (like all paradoxes) and have
been recently discovered. This activity is based on the article “Parrondo’s paradox” by Hélène Davaux (La
gazette des mathématiciens, 2017).

15. Find and replace

Searching and replacing are two actions so common that you don’t always realize how strong they are.
First of all, it is a good exercise to program the search and replacement operations yourself. There is a
sophisticated version of these operations called regular expressions (regex). It is a language in its own right,
very powerful, but a little esoteric.
There is also the mathematics of “find/replace”! The proposed activities are examples that illustrate the
main theorem of the article A complete characterization of termination of 0 p 1q → 1 r 0s , by H. Zantema and
A. Geser (AAECC, 2000).

16. Polish calculator – Stacks

Polish notation (the exact name is “reverse Polish notation”) is another way to write algebraic operations.
Once again, the hardest part is to adapt your brain to this change in writing. This allows operations (and
their priorities) to be seen in a new light. Another interesting vision is to see an operation in the form of a
binary tree:

5 4

+ 3

which represents (5 + 4) × 3 or in Polish notation 5 4 + 3 ×.

We have tried to postpone as much as possible the change of a global variable in a function, but here it is
natural. The use of global is to be avoided in general.
The notion of a stack is a very simple way to structure and access data. However, this is the right way to
manage an expression with brackets. The analogy with the sorting station should be illuminating. We will
find the stacks again in the activity on L-systems.
A stack works on the principle of “first in, last out” (filo). Another possible data management is on the
principle “first in, first out” (fifo) as in a queue.

17. Text viewer – Markdown

The purpose of this chapter is twofold: to discover the Markdown which is a very practical language for
formatting a text, but also to understand the justification of a paragraph.
Obviously the Markdown language has more tags than those presented here. We could continue the project
by doing the justification with fonts of different sizes and shapes, or even create a complete small word
processor.
NOTES AND REFERENCES 202

18. L-systems
We come back to the theme “find/replace” but this time with a geometric vision. The figures obtained
are beautiful, easy to program using the turtle, but the most beautiful is to see in live the layout from the
L-systems. For L-systems defined by expressions containing brackets, we find again the notion of stacks.
The formulas are taken from the book The algorithmic beauty of plants, by P. Prusinkiewicz and A. Lindenmayer
(Springer-Verlag, 2004) with free access here: The algorithmic beauty of plants (pdf).
The illustrations that begin each part of this book are iterations of the L-system called the Hilbert curve and
defined by:
start = "X" rule1 = ("X","lYArXAXrAYl") rule2 = ("Y","rXAlYAYlAXr")

19. Dynamic images

This activity is based on the article “Blurred images, recovered images” by Jean-Paul Delahaye and Philippe
Mathieu (Pour la science, 1997). This article also looks at the calculation of the number of iterations
required to find the original image. The underlying mathematical notion is that of permutation: a one-to-one
transformation of a finite set (here the set of pixels) into itself.

20. Game of life

A great classic of fun computing! There are dozens of sites on the Internet with buildings having incredible
properties and many other ideas. But the most fascinating thing is that extremely simple rules lead to
complex behaviors that resemble the life and death of cells.

21. Ramsey graphs and combinatorics

Graphs are very common objects in mathematics and computer science. The problem presented here is
simple and fun. But what we may have to remember from this chapter is how the difficulty increases with
the number of vertices. Here we do the calculations up to 6 vertices and we can’t go much further with our
exhaustive verification method.
A great mathematician Paul Erdös stated that if aliens landed on Earth threatening to destroy our planet
unless we could solve the problem of 5 friends/5 foreigners, then by mobilizing all the computers and
mathematicians in the world we would manage to get by (we know that the answer is between 43 and 48
people). On the other hand, if the aliens asked us to solve the problem for 6 friends/6 foreigners, then the
easiest thing would be to prepare for war!

22. Bitcoin
Before investing all your savings in bitcoins it is better to understand how it works! The activities presented
here are intended to present a (very) simplified version of the blockchain that is the basis of this virtual
currency. The principle of the blockchain and the proof of work are not so complicated.
NOTES AND REFERENCES 203

23. Random blocks

These random constructions are first of all computer recreations that produce pretty figures, all similar but
all different. But they are also the subject of modern and difficult mathematical work. Martin Hairer won
the Fields Medal in 2014 for studying the upper boundary of our falling blocks, whose shape is governed
by an equation, called “KPZ equation”. A good bonus activity would be to drop blocks of the game Tetris
instead of small squares.
Thanks
I would like to thank Stéphanie Bodin for her encouragement and for testing the activities of the first parts.
I would like to thank Michel Bodin very much for having tested all the first activities and his proofreading.
Thanks to François Recher for his enthusiasm and careful review of the book. Thanks to Éric Wegrzynowski
for his review and ideas for the activities on the images. Thanks to Philippe Marquet for his review of the
first chapters. Thanks to Kroum Tzanev for the model of the book.

You can retrieve all the activity codes Python and all the source files on the Exo7 GitHub page: “GitHub :
Python in high school”.

Exo7

This book is distributed under license Creative Commons – BY-NC-SA – 4.0 .

On the Exo7 website you can download the book in color for free.
 Index

π, 5 chr, 45
**, 2 click, 144
+=, 36 close, 68
=, 3, 21 comments, 3
#, 3 concatenation, 26, 40, 51
!=, 21 cos, 5
//, 33 csv, 70
<=, 21
def, 24
==, 21
degrees, 76
>=, 21
del, 53
%, 33
digits, 21
abs, 6 distance, 42, 74
absolute value, 6 divisor, 36
and, 21, 96 DNA, 43
angle, 76 docstring, 27
append, 50 down, 10
argument, 27, 127
else, 18
array, 90
entry, 19
ascii, 44
Euclidean division, 2, 33
assignment, 3, 21
eval, 118
average, 59, 66
expected value, 99
ballistics, 56
False, 21
base 10, 82
file, 68
bin, 86
find, 102
binary, 82, 95, 150
first number, 36
bitcoin, 155
float, 69
bits, 83
floating point number, 2
blockchain, 157
floor, 5
boolean, 21
font, 120
break, 80
for, 6, 42, 51
button, 63
forward, 9
calculation time, 37, 156 function, 24
character, 40 argument, 27
character encoding, 44 docstring, 27
choice, 69 hash, 158
 optional argument, 127 while, 34
parameter, 24
return, 27, 175 magic square, 91
markdown, 120
game of life, 140 math, 5
gcd, 5 matplotlib, 55
global, 108 max, 59
Goldbach, 77 median, 64
goto, 10 min, 59
graph, 146 mining, 160
graphic, 9, 60 module, 5
math, 5
hash, 158 matplotlib, 55
if, 18 random, 19
if/then, 18
re, 103
image, 71, 134
time, 156
import, 5 timeit, 37
in, 6, 42, 89, 102 tkinter, 60, 119, 144, 169
indentation, 6
turtle, 9
modulo, 2, 33
index, 89, 102
Morgan’s laws, 97
input, 19
mouse, 144
int, 19, 69
interest, 51 None, 44
isdigit, 110 not, 21, 96
join, 110 not in, 42

Koch’s snowflake, 128

open, 68
or, 21, 96
L-systems, 126 ord, 45
lcm, 5 otherwise, 18
left, 9
palindrome, 42, 95
len, 40, 51
parameter, 24
list, 50, 89
pbm/pgm/ppm, 71, 136
add, 50, 52
comprehension, 89
pi, 5
delete, 53
plot, 55
Polish notation, 112
invert, 52
length, 51
pop, 107, 131
power, 2
merge, 51
power of 2, 83
of lists, 90
prime number, 54
slice, 89
sort, 54
print, 3
proof of work, 155
sublist, 52
push, 107
list, 7, 150
local variable, 31, 176 quartiles, 65
logic operation, 21, 95 quotient, 33
loop
for, 6, 42, 51 radians, 76
quit, 80 randint, 19
random, 19, 99 string, 26, 40, 150
range, 7 join, 110
regex, 103 split, 109
regular expression, 103 sum, 59
remainder, 2, 33
remove, 53 ceil, 5
replace, 102, 126 time, 156
return, 24, 27, 175 timeit, 37
reverse/reversed, 52 tkinter, 60, 119, 144, 169
rgb, 73, 74 triangle, 22
right, 9 True, 21
round, 5 try/except, 81
turtle, 28, 126
shuffle, 92 turtle, 9
Sierpinski’s triangle, 12, 129
unicode, 44, 122
sieve of Eratosthenes, 54
sin, 5 up, 10
sort/sorted, 54 upper, 26
split, 109 variable, 3, 108
sqrt, 5 variance, 60
square root, 5, 79
stack, 131 while, 34
standard deviation, 60 window, 60
str, 19, 40, 68 write, 68

Version 0.01 – May 2019