idf3.js

function inverseDocumentFrequency(document){
    // calculates the inverse document frequency of every sentence
    const words_without_stopwords = prettify(document);
    const unique_words_set = uniqueWords(words_without_stopwords);

    const sentences = document.split(".").map(item => item.trim());
    sentences[0] = sentences[0].substring(146);

    const lengthOfDocuments = sentences.length;
    // prettifys each sentence so it doesn't have stopwords

    const wordCountAll = countWords(words_without_stopwords);

    // counts words of each sentence
    // as each sentence is a document
    wordCountSentences = [];
    for (let i = 0; i <= lengthOfDocuments - 1; i ++){
        wordCountSentences.push(countWords(prettify(sentences[i])));
    }

    // calculate TF values of all documents
    let IDFVals = {};

    // how many times that word appears in all sentences (documents)
    wordCountSentencesLength = wordCountSentences.length;
    // for every unique word
    for (let i = 0; i <= unique_words_set.length - 1; i++){
        let temp_add = 0;
        // count how many times unique word appears in all sentences
        for (let x = 0; x <= wordCountSentencesLength - 1; x++){
            if (unique_words_set[i] in wordCountSentences[x]){
                temp_add =+ 1;
            }
        }
        IDFVals[unique_words_set[i]] = Math.log10(wordCountAll[unique_words_set[i]] / temp_add);
    }

    let IDFSentences = {};
    // for every sentence
    for (let i = 0; i <= lengthOfDocuments - 1; i ++){
        // for every word in that sentence
        let sentence_split_words = sentences[i].split(" ");
        // get the assiocated IDF values of each word
        // temp.add is the "IDF" value of a sentence, we need to divide it at the end
        let temp_add = 0.0;
        let words_no_stop_words_length = prettify(sentences[i]).length;
        for (let x = 0; x <= sentence_split_words.length - 1; x++){
            // if the word is not a stopword, get the assiocated IDF value and add it to temp_add
            if (sentence_split_words[x].toLowerCase() in IDFVals){
                // adds all the IDF values up
                temp_add = temp_add + IDFVals[sentence_split_words[x].toLowerCase()];
            }
            else{
                // nothing, since it's a stop word.
            }
        }
        IDFSentences[sentences[i]] = temp_add / words_no_stop_words_length;
    }
    return IDFSentences;
}