README

This is basic implementation of K-D trees in PHP inspired by Princeton K-D trees assignment and done as a graduation project for Algorithms Course in Projector.

Installation

composer require f1r3starter/kdtree

Usage

Tree construction

Firstly, you have to decide, how many dimensions your tree is going to be used for, after that you can add some points:

<?php

use KDTree\Structure\KDTree;  
use KDTree\ValueObject\Point;

$kdTree = new KDTree(2); // 2 for two-dimensional points, eg cities
$kdTree->put(new Point(35.0844, 106.6504)); 
$kdTree->put(new Point(41.2865, 174.7762));

// if you need somehow connect point to your application, you can use setName method
$point = new Point(46.8117, 33.4902);
$point->setName('Kakhovka');
$kdTree->put($point);
//...
$points = $kdTree->points(); // returns list of all points, which can be iterated through

$kdTree->contains(new Point(46.8117, 33.4902)); // will return "true"

Nearest point search

After tree is constructed, we can try to find nearest point:

<?php

use KDTree\Search\NearestSearch;
use KDTree\ValueObject\Point;

$search = new NearestSearch($kdTree);
$nearestPoint = $search->nearest((new Point(41.2865, 174.7762)));

Range search

Also there is an ability to find points in some particular range, which should have $2^{k}$ points, where k - is number dimensions in your K-D tree.

<?php

use KDTree\Structure\PointsList;
use KDTree\Search\PartitionSearch;
use KDTree\ValueObject\{Partition, Point};

$pointsList = new PointsList(2);  
$pointsList->addPoint(new Point(46.8117, 33.4902));  
$pointsList->addPoint(new Point(31.3142, 42.5245));  
$pointsList->addPoint(new Point(22.2525, 41.3412));  
$pointsList->addPoint(new Point(55.4245, 52.5134));  

$search = new PartitionSearch($kdTree);  
$foundPoints = $search->find(new Partition($pointsList));

Demo

Demo is available in this repository.

Credits

Andrii Filenko

f1r3starter / kdtree

Maintainers

Details