<?php
/*
* Simple coword calculation, all in memory
*
* WARNING: Because of PHP's very very very bad UTF-8 support, splitting tweets into words only works reliably for tweets consisting solely of latin chars (i.e. English)
*/
if (0) {
$coword = new Coword();
$coword->setDocuments(array("bla bla bla test test clash", "test test test bla bla", "clash bla test"));
$coword->addDocument("bla, !test 345 cla5sh");
$coword->iterate();
var_export($coword->getCowordsAsCsv());
print "\n\n";
var_export($coword->getWordsAsCsv());
print "\n";
}
class Coword
{
public $documents = array();
public $punctuation = array();
public $hashtags_are_separate_words;
public $extract_only_hashtags;
public $remove_stop_words;
public $min_word_length;
public $min_word_frequency;
public $words = array();
// holds word frequencies
public $cowords = array();
// holds coword frequencies
public $document_word_frequencies = array();
