K Means Clustering Implemented On Codeigniter To Find Strategic Market

Installation

1 .You need to install composer, Read how to install composer

2. Install git, Read how to install git

3. Clone repository with git

git clone https://github.com/AlifAbhiesa/Codeigniter-K-Means-Clistering.git

And Run Composer Update

composer update

4. Create your own Google API

You need 3 google API, google maps API to show routes, geocoding for translating address into Geographic Coordinates, and Distance matrix to provides distance between two place. Read how to get the API

4.1. Read how to setup Google Maps API

4.1. Read how to setup Geocoding API

4.1. Read how to setup Distance Matrix API

5. Configure Google API in your code

5.1. Placing Maps API

Open file /application/views/default/index.php and place your API key here

 <script src="https://maps.googleapis.com/maps/api/js?key= YOUR API KEY" type="text/javascript"></script>

5.2. Placing Geocoding API

Open file /application/controller/Data.php and place your API key here

public function getGeo($location){
 $httpClient = new \Http\Adapter\Guzzle6\Client();
 $provider = new \Geocoder\Provider\GoogleMaps\GoogleMaps($httpClient, null, ' YOUR API KEY ');
 $geocoder = new \Geocoder\StatefulGeocoder($provider, 'en');
 $result = $geocoder->geocodeQuery(GeocodeQuery::create($location));
 $Longitude = $result->first()->getCoordinates()->getLongitude();
 $Latitude = $result->first()->getCoordinates()->getLatitude();
 $coordinates = array(
	 'Longitude' => $Longitude,
 	'Latitude' => $Latitude
 );
 return $coordinates;
}

5.2. Placing Distance Matrix API

Open file /application/controller/Data.php and place your API key here

public function distance($LatOri,$LongOri,$LatDes,$LongDes){
 $distance_data = file_get_contents('https://maps.googleapis.com/maps/api/distancematrix/json?&origins='.
 $LatOri.','.$LongOri.'&destinations='.$LatDes.','.$LongDes.'&key= YOUR API KEY');
 $distance_arr = json_decode($distance_data);
 $result = $distance_arr->rows;
 //result is a distance on Meter
 return $result[0]->elements[0]->distance->value;
}

Methode

In this application i'm using 2 methode. K Means clustering to create the cluster of data, and Silhouette Coefficient to test the result cluster of K Means clustering. If you want to understand the flow of that method, you need to read this paper

1. Paper of K Means Clustering

2. Paper of Silhouette Coefficient

Read Me

Read Me on Medium, Story about this application by Alif Abhiesa