README

Solution for Knapsack problem in PHP applications.

See description of the problem.

Available algorithms

Example of use

<?php

use Writ3it\LibAlgo\KnapsackProblem\Impl\Item;
use Writ3it\LibAlgo\KnapsackProblem\Impl\Bag;
use Writ3it\LibAlgo\KnapsackProblem\Algorithm\DynamicKnapsackSolver;

$items = [
    // new Item(<weight>, <value f.e. price>)
    new Item(2, 4), /** Item 0 **/
    new Item(1, 3), /** Item 1 **/
    new Item(4, 6), /** Item 2 **/
    new Item(4, 8)  /** Item 3 **/
];

// new Bag(<capacity>)
$bag = new Bag(8); 

$algo = new DynamicKnapsackSolver();
$value = $algo->solve($items, $bag);

var_dump([

     // Item 0,1,3 which are optimal content of the bag.
    'content'=>$bag->getItems(), 

     // Summary value of items in bag.
    'value' =>$value 

]);

Custom Item

A item can be any PHP object. Sometimes, you may need to model your logic with more complex objects than Item class. You can do that by implementing ItemInterface on your own:

<?php

use Writ3it\LibAlgo\KnapsackProblem\ItemInterface;

class CustomItem implements ItemInterface
{

    
    /**
     * {@inheritdoc}
     */
    public function getWeight(): int
    {
        // Custom logic to compute a weight.
    }
    
    /**
     * {@inheritdoc}
     */
    public function getValue(): int
    {
        // Custom logic to compute a value.
    }

    /**
     * More your own methods
     */

}

Efficient algorithms for solving the knapsack problem require that weight and value be integers and constant while solver runtime.

Custom Bag

A bag can be any PHP object. Sometimes, you may need to model your logic with more complex objects than Bag class. You can do that by implementing BagInterface on your own:

<?php

use Writ3it\LibAlgo\KnapsackProblem\ItemInterface;

class CustomBag implements BagInterface
{

    
    public function getCapacity(): int
    {
       // Custom logic to compute a capacity.
    }

    public function addItem(ItemInterface $item): void
    {
        // Custom logic to receive a items which are a solution.
    }

    /**
     * More your own methods
     */

}

Efficient algorithms for solving the knapsack problem require that weight and value be integers and constant while solver runtime.