analyrics

About

This node package is for easily retrieving and analyzing song lyrics. Currently, the metrics available for songs are the frequency of words, and the overall sentiment (positivity/negativity). However, the end goal is to discover and visualize all sorts of patterns found in music over the years.

Getting Started

This package relies on Genius for the lyrics it analyzes, meaning that you'll have to grab a Genius API Key.

Then just load the package into your program and set the key as shown below.

var analyrics = require("analyrics");
analyrics.setToken("your_api_key");

Usage

Fetch Song (analyrics.getSong(searchQuery, callback))

This returns a song object with a title, artist, source URL, lyrics, word frequency, and sentiment. Here's an example.

analyrics.getSong("Can't Take My Eyes off You", function(song) {
  console.log(song.title);
  console.log(song.artist);
  console.log(song.url);
  console.log(song.lyrics);
  console.log(song.frequency);
  console.log(song.sentiment);
});

This is the output from the function above:

The output for the title, artist, and source of lyrics (url):

Can't Take My Eyes Off You
Lauryn Hill
http://genius.com/Lauryn-hill-cant-take-my-eyes-off-you-lyrics

Song lyrics are stored in a single string

You're just too good to be true
Can't take my eyes off of you
You'd be like heaven to touch
I wanna hold you so much
At long last love has arrived
And I thank God I'm alive
You're just too good to be true
Can't take my eyes off of you
...

Word frequency is stored in an array

[ { w: 'baby', c: 7 },
  { w: 'love', c: 4 },
  { w: 'good', c: 3 },
  { w: 'eyes', c: 3 },
  { w: 'true', c: 3 },
  { w: 'feel', c: 2 },
  { w: 'pretty', c: 2 },
  { w: 'long', c: 1 },
           ...       ]

The sentiment of lyrics is stored in an object that contains the score and arrays of positive and negative words. The sign of the score correlates to the overall sentiment of the lyrics. (Score = Positive Words - Negative Words).

{ score: 14,
  comparative: 0.0958904109589041,
  positive:
   { score: 15,
     comparative: 0.10273972602739725,
     words:
      [ 'good',
        'like',
        'heaven',
        'love',
        'thank',
        'good',
        'like',
        'good',
        'warm',
        'love',
        'trust',
        'pretty',
        'pretty',
        'love',
        'love' ] },
  negative:
   { score: 1,
     comparative: 0.00684931506849315,
     words: [ 'lonely' ] } }

Fetch Billboard Chart (analyrics.getBillboard(billboardURL, callback))

This returns an array of song objects with the following fields: Rank, Title, Artist, Image URL, Spotify URL (if available). Here is an example.

analyrics.getBillboard("http://www.billboard.com/charts/hot-100", function(chart) {
  console.log(chart);
});

This is the expected output:

[ { rank: 1,
    title: 'Closer',
    artist: 'The Chainsmokers Featuring Halsey',
    image: 'http://www.billboard.com/images/pref_images/q56523k8p5q.jpg',
    spotify: 'http://open.spotify.com/track/7BKLCZ1jbUBVqRi2FVlTVw' },
  { rank: 2,
    title: 'Starboy',
    artist: 'The Weeknd Featuring Daft Punk',
    image: 'http://www.billboard.com/images/pref_images/q59123tine7.jpg',
    spotify: 'http://open.spotify.com/track/2IY7eOUDjw2ArKYxKa2jXc' },
  { rank: 3,
    title: 'Heathens',
    artist: 'twenty one pilots',
    image: 'http://www.billboard.com/images/pref_images/q39056czrzc.jpg',
    spotify: 'http://open.spotify.com/track/6i0V12jOa3mr6uu4WYhUBr' },
  ... ]

Billboard charts can be pulled from several URLs, here are some that I've tested.

| Chart Name | URL | | ------------- |:-------------:| | Hot 100 | http://www.billboard.com/charts/hot-100 | | R&B/Hip-Hop | http://www.billboard.com/charts/r-b-hip-hop-songs |
| Year End 2015 (Year End charts start from 2006) | http://www.billboard.com/charts/year-end/2015/hot-100-songs | | Greatest 100 All Time | http://www.billboard.com/charts/greatest-hot-100-singles |

Coming Soon

• Classifiers for things such as slang, themes, material objects
• Uniqueness of lyrics
• Function to fetch lyrics given an album name

License

MIT