Python 3.x
matplotlib
./genetic.py --help
This will display all the relevant arguments for this Python script
To run with the exact arguments from the associated paper (Population Size = 10000, Number of Generation = 5000, Elitism Factor = 3, Number of non-elite individuals to select from parent population = 3, Mutation Chance = 0.5, Number of swaps per mutation = 1, Number of participants = 27, Group size = 3), run
./genetic.py -p 10000 -g 5000 -el 3 -rest 3 -mchance 0.5 -mswaps 1 -n 27 -s 3
You will need a file called rankings.csv
with a n x n
matrix of preference values starting at the first row and first column. The value in row i
and column j
will refer to the rating given by the i
th participant to the j
th participant.
There is some support for running multiple GA instances in parallel. Running:
./genetic-parallel.py
will execute a number of ./genetic.py
instances in parallel. This has no command-line arguments - all changes need to be made in the code itself currently. The call to run_genetic(...)
and the variable cpus
are the main points of modification.