n-queens problem

N queens problem implementation using Genetic Algorithm (or something like it)

Working best on 4x4 tables

Also made using Vue.js

basic usage

• OPEN BROWSER CONSOLE
• first generation
• minimal fitness is used to select best generated queen sets
• call app.init(iterations, minimal_fitness)

app.init(100, 4);

A message to call iterate will be showed

Making new generations

• call app.iterate(iterations, minimal_fitness)
• the iteration can be a larger number
• will be finished if a valid set is finded (fitness = 0)

app.iterate(100000, 4)

tunning and changes

The following parameters can be changed:

• validSetButton -> show button to place a valid set (also can call 'validSet(0 or 1)')
• dark -> dark/light mode
• population -> population size (how many queens, rows and cols will have)
• mutation -> mutation rate (0-1), change how many changes the queen array will have
• crossover -> crossover rate (0-1), change where the parents arrays will be splitted/spliced
• minFit -> (0-10) minimal fitness to selection after crossover
• maxIterations -> (0-?) max iterations to aply in algorithm

data:{
    validSetButton: false,
    dark: true,
    population: 4,
    tuning:{
        mutation: 0.5,
        crossover: 0.3,
        minFit: 4,
        maxIterations: 100
    },
}

Known issues

• sometimes the generations will be finished with a sequence on diagonal ( / )
• 8x8 or higher populations will take longer time to solve and has at least 1 queen matching

license

• MIT

Changelog

2019/10/09

• Added themes (dark/ligh)
• General functions
• generation, mutation and crossover

2019/10/10

• Verifyers and general things
• iterators and finish
• finished first version

2019/10/11

• added running flag
• added inputs in html
• changelog
• minor fixes
• fixed maxIterations and mininal fitness in wrong place