Ciphers and Genetic Algorithm

Using Genetic Algorithm (GA) to break ciphers and find the key used to encrypte a message.

Requirements

• Python 3
• Libraries:
  • NumPy

Run

To see a demo, run the following command:

$ python driver.py --dev

To see more outputs, run the following command:

$ python driver.py --dev -v 2

---

Usage

$ python driver.py [operational_mode]

Operational modes:

• --dev - for quick runs, test functionality, see numbers, and debugging.
• --experiment n - run the entire GA for n iterations and collect statistics.

Simple Run (--dev)

$ python driver.py --dev

Expected sample output (with default verbosity of 1):

End of iteration: 1 		 Best Fitness: 31.670175244226233
End of iteration: 2 		 Best Fitness: 31.0
...
End of iteration: 2385 		 Best Fitness: 1.0
End of iteration: 2386 		 Best Fitness: 0.0
Key: quacktim

---

Change the output verbosity with --verbosity (or -v):

$ python driver.py --dev -v 2

Expected output:

=== Initial population ===
['aaaaaaaa', -1]
['bbbbbbbb', -1]
['cccccccc', -1]
...

=== After fitness calculation and sort ===
['ffffffff', 36.55133376499413]
['eeeeeeee', 35.874782229304195]
['dddddddd', 35.7211421989835]
...

=== Parents for crossover ===
aaaaaaaa
bbbbbbbb

=== Children from crossover ===
aaaabbbb
bbbbaaaa

=== Mutated chromosome ===
aaaadbbu

=== Population after repalcing weak chromosome ===
['aaaabbbb', -1]
...
['aaaaaaaa', 31.670175244226233]

End of iteration: 1 		 Best Fitness: 31.670175244226233

---

$ python driver.py --dev -v 0

Expected output is the found key by the GA. E.g.

Key: quacktim

Experiment

Run an experiment using the --experiment (or e) argument. This will run the GA for n iterations. At each iteration, the number of generations is recorded. At the end of the experiment, results are reported. Results include: mean, variance, and standard deviation.

$ python driver.py -e 5

Expected sample output (with default verbosity of 1):

Experiment: 0 		 Number of generation: 1339
Experiment: 1 		 Number of generation: 3147
Experiment: 2 		 Number of generation: 5127
Experiment: 3 		 Number of generation: 1350
Experiment: 4 		 Number of generation: 612
=== Stats Summary ===
n =  5
Mean: 2315.0
Var: 2676715.6
StDev: 1636.067113537828

Verbose level 0 will simply show the Stats Summary at the end of execution. Verbosity of 2 will show the list containing number of generation across all experiments.

Verbose

Specify verbosity level using --verbosity (or -v). Verbosity levels:

0 - None
1 - Summary
2 - All (Debug)

Others

Use --help (or -h) to see a list of all argument and usages.

$ python driver.py -h