Evolutionary Strategies made simple
Project description
evopy
Evolutionary Strategies made simple!
Use evopy to easily optimize a vector of floats in Python.
🏗 Installation
All you need to use evopy is Python 3! Run this command to fetch evopy from PyPI:
pip install evopy
Then you can import EvoPy
like this:
from evopy import EvoPy
⏩ Usage
One-Dimensional Example
Let's say we wanted to find the optimum of a parabola, without using exact methods from calculus! With Evopy, this is as easy as writing the following two lines:
evopy = EvoPy(lambda x: pow(x, 2), 1)
best_coordinates = evopy.run()
The main ingredient here is the fitness function (the lambda). This can also be a normal function reference, just make sure that it accepts a float or an array of floats and outputs a single float. The other ingredient is the 1
at the end of the first line: This is the dimensionality of the inputs that you expect in your fitness function. best_coordinates
will contain an array with a single element, which is the best x
value the algorithm could find in the default number of generations.
Multi-Dimensional Example
If the previous example seemed too simple to you, we can also look at the optimum of a more complex, two-dimensional function, like the Rastrigin function. We don't have to modify much in our previous code snippet to get this to work:
evopy = EvoPy(
lambda X: 5 + sum([(x**2 - 5 * np.cos(2 * np.pi * x)) for x in X]),
2,
generations=1000,
population_size=100
)
best_coordinates = evopy.run()
Compared to the first example, we have interchanged the fitness function for a more complex one, set the dimensionality to 2
, and given the algorithm more time to find an optimum by setting a higher generation and individual count than the default.
Docs
For more detailed information on evopy's functionality, have a look at the docs!
⛏ Development
Clone this repository and fetch all dependencies from within the cloned directory:
pip install -r requirements.dev.txt
Run all tests with:
nosetests
To check your code style, run:
pylint evopy
To measure your code coverage, run:
nosetests --with-coverage --cover-package=evopy --cover-html --cover-branches --cover-erase
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file evopy-0.3.tar.gz
.
File metadata
- Download URL: evopy-0.3.tar.gz
- Upload date:
- Size: 7.1 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/1.13.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/41.0.1 requests-toolbelt/0.9.1 tqdm/4.32.1 CPython/3.7.1
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 |
a3b3c122383a98b3ee9647fb26b07d8cbf656a3ac929cfd2cd98268ea3262fdb
|
|
MD5 |
8ee126c894da70ebe55fd3c9f3ac7f24
|
|
BLAKE2b-256 |
481b9333e30a7930feba44a443be9d06f6e283ea8a1f5df915704a7aba85cfe3
|
File details
Details for the file evopy-0.3-py3-none-any.whl
.
File metadata
- Download URL: evopy-0.3-py3-none-any.whl
- Upload date:
- Size: 8.5 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/1.13.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/41.0.1 requests-toolbelt/0.9.1 tqdm/4.32.1 CPython/3.7.1
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 |
dd1b909cc949f77564d6cf6eb43cf609660e73d6b6b584d16a5937849eac9a81
|
|
MD5 |
659f4ce7ce1d762d67213bd82b25083c
|
|
BLAKE2b-256 |
25c1000d76d1df352ac1353681180d8b778799d4527683781b9447ebe68a15da
|