Skip to content
/ nds-py Public

A Python implementation of the non-dominated sorting.

License

Notifications You must be signed in to change notification settings

KernelA/nds-py

Repository files navigation

Non-dominated sorting

Test nds library

Description of the method

You can read about the method in the next article:

Buzdalov M., Shalyto A. A Provably Asymptotically Fast Version of the Generalized Jensen Algorithm for Non-dominated Sorting // Parallel Problem Solving from Nature XIII.- 2015. - P. 528-537. - (Lecture Notes on Computer Science; 8672)

Requirements

  1. Python 3.6 or higher.
  2. Installed setuptools.
  3. Installed wheel.

Installation

PyPI

PyPi version

Local installation

Run pip install ..

Tests

Run command:

python -m unittest discover -v ./tests

How to use

The example:

import random

# Package must be installed.
from nds import ndomsort

seq = [random.sample(range(-10, 11), 5) for i in range(30)]

# It is dictionary.
fronts = ndomsort.non_domin_sort(seq)

# Or we can get values of objectives.
# fronts = ndomsort.non_domin_sort(seq, lambda x: x[:4])

# 'fronts' is a tuple of front's indices, not a dictionary.
# fronts = ndomsort.non_domin_sort(seq, only_front_indices=True)

for front in fronts:
    print("\nFront index is {}".format(front))
    for seq in fronts[front]:
        print("\t{}".format(seq))

Other implementations

Example

Pareto front figure