https://github.com/jbytecode/randomkeyga.py
Random Key Genetic Algorithm for Permutation Based Optimization Problems in Python
https://github.com/jbytecode/randomkeyga.py
Last synced: 4 months ago
JSON representation
Random Key Genetic Algorithm for Permutation Based Optimization Problems in Python
- Host: GitHub
- URL: https://github.com/jbytecode/randomkeyga.py
- Owner: jbytecode
- License: mit
- Created: 2024-08-07T13:37:38.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2024-08-09T10:16:45.000Z (11 months ago)
- Last Synced: 2025-01-29T05:34:09.532Z (6 months ago)
- Language: Python
- Size: 8.79 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# randomkeyga.py
Random Key Genetic Algorithm for Permutation Based Optimization Problems in Python
## Example
Here is a simple traveling salesman problem.
First things first:
```python
from randomkeyga import optimizer
```
Coordinates of the points to visit:
```python
points = [
[0, 0], [0, 1], [0, 2],
[1, 2], [2, 2], [3, 2],
[3, 1], [3, 0], [2, 0], [1, 0]
]
```
The Euclidean distance function (Numpy can be used instead for faster calculation)
```python
def distance(point1, point2):
return ((point1[0] - point2[0])**2 + (point1[1] - point2[1])**2)**0.5
```
The objective function. The total distance will be minimized:
```python
def total_distance(path):
total = 0
for i in range(len(path)-1):
total += distance(points[path[i]], points[path[i+1]])
# Add the distance from the last point to the first point
total += distance(points[path[-1]], points[path[0]])
return total
```
Optimize:
```python
result = optimizer.optimize(
popsize=100,
chsize=10,
costfunction=total_distance,
selection_operator=optimizer.TournamentSelection(2),
crossover_operator=optimizer.OnePointCrossOver(),
mutation_operator=optimizer.RandomMutation(0.01),
crossover_rate=crossover_rate,
mutation_rate=0.01,
elitism=2,
generations=200)
best = result.chromosomes[0]
# Expected result is Chromosome([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], cost = 10)
# Note that the solution Chromosome([3, 4, 5, 6, 7, 8, 9, 0, 1, 2]), cost = 10) has the same meaning.
```
Currently only the `TournamentSelection` is implemented as the selection operator.
`OnePointCrossOver`, `TwoPointCrossOver`, and `UniformCrossOver` are alternative recombination operators.
`RandomMutation` is the implemented mutation operator.