exercises.md

Exercises for Day 2

Organising, debugging and profiling Python code

1. Creating a Python package

Following the slides from this morning's session, we will create an animals package.

Please note that with Python 3, implicit relative imports are no longer supported https://www.python.org/dev/peps/pep-0328/

a. Create a directory for package

b. Add the modules birds.py and mammals.py into your package directory

c. Add __init__.py module to your package

The __init__.py module should import the Birds and Mammals class from the two modules.

d. Test your animals package

Using a python script or an interactive python session outside the package, test your brand new package:

import animals

m = animals.Mammals()
m.printMembers()

b = animals.Birds()
b.printMembers()

e. Add another module called fish.py amd integrate it into your package

f. Reorganize your package such that you can use it like this:

import animals

harmless_birds = animals.harmless.Birds()
harmless_birds.printMembers()

dangerous_fish = animals.dangerous.Fish()
dangerous_fish.printMembers()

2. Debugging

Investigate buggy code using the pdb or ipdb debugger. Have a look at slides of this mornings's session for help.

a. Find all the bugs in the dicegame

Clone this repo (if you have not already done so), go to buggy and run the main.py with a debug tracer added to the code. Once you have fixed all errors, the game should correctly add up the values of the dice for 6 consecutive turns.

b. If you cannot get enough of debugging

Ask your neighbour to introduce more bugs into the above (or any other) code examples and try to find the bug using the debugger.

3. Profiling

In this section, you should get more familiar with code profiling, in particular with the tools cProfile, line_profiler. Have a look at slides from this morning's session to understand what they are doing and when you should use them. Try out profiling both from the command line and using interactive python (e.g. jupyter notebook). If you get Command not found when running kernprof try searching for it in ~/.local/bin/kernprof. Alternatively install it using Anaconda/conda (e.g. conda install line_profiler).

a. Investigate the performance of the matmult.py script

In which line(s) of the script would you start optimizing for speed?

b. Investigate the performance of the euler72.py script

In which line(s) of the script would you start optimizing for speed? (This is one problem from the euler project: https://projecteuler.net/problem=72)

c. Improve the performance of the matmult.py script

What is the best performance that you achieved with N=250?