Choose the challenges in any order, the first exercises will be easier than the exercises at the end.

Challenge: RNA translation

Write a script which translates a FASTA file containing RNA sequences to a FASTA file with corresponding amino acid sequences.


Challenge: Formula generator

Write a function which takes two mass values and prints all mass formulas consisting of zero or more C, H and / or O where the mass is in the given range. The exact masses are: mass_C = 12.0, mass_H = 1.0078250319 and mass_O = 15.994915.

So for example for the mass range 100.0 to 100.1 the formulas C4H4O3, C5H8O2, C6H12O and C8H4 are printed.

Hint: you need three nested for loops for this to iterate over all combinations of numbers of C, H and O. The inner body computes the mass of the current combination and checks if this fits to the prescribed range. You can estimate the upper bounds of the loops from the upper limit of the mass range.

If you solved this introduce some "pretty printing" so that a combination 8 C, 1 H, zero O is displayed as C8H.

  • Restrict your program to only look up carbo hydrates having the general formula $C_m(H_2O)_n$. You can either filter the results from the existing solution or implement a new version with two nested loops for m and n.

Challenge: Simple Numerical Computations

These exercises assume some familiarity with basic calculus. If this is not you cup of tea skip the challenge. The exercises try to give you an idea what numerical computation is about.

Challenge: fit data to a given function

The following snippet shows how you can determine parameters of a given function to fit to given data:

import numpy as np
import scipy.optimize
import matplotlib.pyplot as pyplot

n_points = 10

x_data = np.arange(n_points)

# we generate artificial data by applying the function f(x) = -0.1 * x^2 + 2 * x + 1
exact_y_data = -.1 * x_data ** 2 + 2 * x_data + 1

# we add some noise with std dev 0.5:
measured_y_data = exact_y_data + 0.5 * np.random.randn(n_points)

def f(x, a, b, c):
    argument x:  vector of values
    argument a:  first parameter to fit
    argument b:  second parameter to fit
    argument c:  third parameter to fit
    returns:  vector of y values corresponding to parameters a and b
    return a * x * x + b * x + c

# we start with the assumption a = 0, b = 0 and c = 0: 
p_start = np.array([0, 0, 0])

parameters, cov_matrix = scipy.optimize.curve_fit(f, x_data, measured_y_data, p_start)

pyplot.plot(x_data, exact_y_data, "b", label="exact")
pyplot.plot(x_data, measured_y_data, "b*", label="measured")
pyplot.plot(x_data, f(x_data, parameters[0], parameters[1], parameters[2]), "green", label="fitted")
pyplot.legend(loc=2)  # upper left corner
[-0.09965958  1.95233482  1.38766372]
  • Install numpy, scipy and matplotlib (if not done yet) and reproduce this example

Logistic growth describes the evolution of a single population in an environment of limited capacity. See for details.

Challenge: Hangman game

Implement the Hangman-Game:

  • the computer knows a secret word
  • you have to guess letters from the word until the word is disclosed
  • after every wrong guess the computer informs you about the number of wrong guesses up to now, you loose after 10 wrong guesses
  • before you enter a guess the computer tells what you disclosed up to now by showing the secret word where a * is displayed for not yet guessed letters.

Below is a template to start. The bodies of the functions are empty, fill them up so that the final program works. The multi line strings you see serve as comments and describe how the functions should work.

If you solved this here are two advanced but very good code katas !!!!

First code kata: only keep the functions from you solutions and try to reimplement play_game without cheating.

Second code kata: implement the full solution.

def print_guessed(word, user_inputs):
    word:  a string holding the secret word
    use_inputs:  a list of letters, collecting the user inputs
    returns: None
    this function prints the word letter by letter showing
    unguessed letters as "*"
    So the function might show "P Y * * O N" if the word is "PYTHON" and
    the user_input is ["O", "N", "P", "X", "E", "Y"]
def count_guessed(word, user_inputs):
    word:  a string holding the secret word
    use_inputs:  a list of letters, collecting the user inputs
    returns: the number of letters in "word" appearing in "user_inputs".
def ask_new_guess(user_inputs):
    user_inputs: list of letters already input by user
    returns: new guess from user in upper case
    asks until the user provids a new guess, if he repeats a
    previous guess a message is displayed and the user is
    asked again. Same happens if user inputs empty string
    or more than one character

def play_game(secret_word):
    user_inputs = []
    guesses_left = 5

    while True: 
        print_guessed(secret_word, user_inputs)

        guess = ask_new_guess(user_inputs)
        if guess not in secret_word:
            guesses_left -= 1
            print("wrong %d guesses left !!!" %  guesses_left)
            if guesses_left == 0:
                print("GAME OVER ! THE WORD IS", secret_word)

        letters_guessed = count_guessed(secret_word, user_inputs)
        if letters_guessed == len(secret_word):
            print("YOU GOT IT !")
import random
secret_words = ["notebook", "pythoncourse"]

secret_word = secret_words[random.randint(0, len(secret_words) - 1)].upper()


Challenge: Phone number data base

Write a phone book application. If you start the program the user if asked if he wants to add an new name + phone number or if he wants to lookup an existing phone number. Store the data provided by the user in a csv file so if you restart the program the already entered data will not be lost. So the interaction with the program could look like this:


 Already 3 entries in the phone book

 Please choose:

     1: add a new entry
     2: lookup a number
     3: show all entries
     0: exit

 Your choice: 1

 Ok, please enter a name: Julian
 And his phone number: +41 2342 2342


 Already 4 entries in the phone book

 Please choose:

     1: add a new entry
     2: lookup a number
     3: show all entries
     0: exit

Write functions like def main_menu(phone_book_entries):, def submenu_add_entry(phone_book_entries) and def submenu_lookup_entry(phone_book_entries) etc.

Discard invalid input.

Suggestions: Either you use a dictionary for lookup or lists. The second approach would allow you to make a lookup which ignores special characters like spaces or . in a name and would ignore the case of the charaters.

If you start and the phone book is empty you can use import os and then os.path.exists(..) (use google for further details) to check if there is already an phone book file or not.

Challenge: Tic Tac Toe

Users should place X and O in alternating order by providing coordinates like A2 (for the first row and second column).

A user is not allowed to overwrite a taken field and the computer should detect if X or O wins.

You can represent a 3 times 3 board by nested list like board = [[" ", " ", " "], [" ", " ", " "], [" ", " ", " "]] for the empty board at the beginning.

To get started:

  • Write a function which asks the user to place a X or O to given coordinates until he chooses a free field. So this could be def place(symbol, board): which then returns the updated board.
  • Write a function which prints the board nicely as def print_board(board):.
  • Write a function which checks if a given field is free. def is_empty(coordinates, board):.
  • Write a function which checks if a given board represents a winning state like def check_winner(board) which returns X or O or None.

For check_winner you could for example check if you find a row where all values are the same, than the first value of this row would be the result. Same for columns and the two diagonals.

Challenge: Fetch CHF exchange rates from a web service

The following snippet demonstrates how to fetch exchange rate EUR to X for a given date using a so called "web service".

import json
import urllib.request

url = ""
response_bytes = urllib.request.urlopen(url).read()

So the web service returns a (so called) "byte string" which looks like a dictionary (the format is called JSON). To get a "real" dictionary we can use:

import pprint
data_as_dict = json.loads(str(response_bytes, "ascii"))
pprint.pprint(data_as_dict)  # nicer output than a plain print()

Take the code as it is, talking to web services is a topic on its own. Another example for a web-service is PUG to query data from Pubchem:

  • Start with this snippet to write a function which determines the CHF to EUR exchange rate for a given year, month and day.
  • Then create a plot for the EUR to CHF exchange for the first of every month starting with 1st of January 2015 until today.
  • Write a function which determines the average exchange rate for a given month. (You remember the leap year computation exercise ?)
  • Create a plot for the monthly averages.

Challenge: A simple sorting algorithm

Implement selection sort to sort a list of numbers. This is the strategy of this algorithm:

  • Start with index 0 and find the position of a minimal value in the list starting with index 0. Swap the values at both positions.
  • Go on with index 1 and find the position of a minimal value in the list starting with index 1. Swap the values at both positions.
  • Go on with index 2 and find the position of a minimal value in the list starting with index 2. ...
  • .. and so on

So for a list [2, 5, 1, 4, 3] the result of step 0 is [1, 5, 2, 4, 3], the result of step 1 is [1, 2, 5, 4, 3], then [1, 2, 3, 4, 5].

  1. Reproduce and exercise this procedure with pen and paper.
  2. Start with a function min_position(li, start_at) which computes the position of the minimal value staring at the given position start_at.
  3. Then implement a function step(i, li) which performs step i for the given list (look up the position with min_position(i + 1, li) then swap values at i and the computed position)
  4. Write a loop which performs step(0, li) up to step(len(li) - 1, li)
  5. Test your program for many inputs including empty lists and lists with only one element.

Challenge: Decrypt a substitution cipher

Download the encrypted file The text was encoded by a "simple substitution cipher" as described at The cipher kept space characters as space characters.

We assume that the letters in English texts ordered by their frequency are etaonhisrldwugfycbmkpvjqxz. So e is the most common character, t the second common character and so on. (This ordering is not fixed but depends on the text corpus used, so you will find slightly modified version on the Internet).

So if a simple substitution cipher maps e to a we should see that a occurs most often in the encrypted text (given that the original text was in English and that the encrypted text is long enough).

To decode the text:

  • first order the letters from the encoded text by their frequency
  • then construct a dictionary to map the characters from the given text to their most probable counterpart
  • and finally use this to decode the text.

The decoding will not be perfect but you should recognize English words and might refine the constructed dictionary manually to improve the result.