Spotify Analysis

Authors

• Eduardo Manuel Ceja Cruz lalitoceja@gmaill.com
• Nikolai García Chkourak nikolai.garcia.98@gmail.com

License:

GNU General Public License version 3

Introduction

This project will atempt to give an analysis of songs from Spotify using a library that implements the Spotify API into python called Spotipy. The objective of this analysis is to know which are the most listened songs in spotify by using the "TOP 50" playlist of every country that spotify generates.

Instructions

1. First, you need to download the proyect

   git clone https://github.com/theDeadIns/zombie 
   

2. install the following libraries:

   • Spoitpy library

     An implementation of the spotify API into python.

     You can do this by simply running

     pip install spotipy
     

   • Bokeh

     A library made for data visualization with more tools and flexibility than matplotlib. We use it for generating the gaphs, it's interactivity and teh including them into the HTML.

     Again, you can do it with

     pip install bokeh
     

   • Jinja2

     A way to put funcitonality into a HTML template, and we used it just for that. The web page that is generated uses this library to replace the {{ div }} and {{ script }} from the file called template.html to the actual content

     To install it, type

     pip install Jinja2
     

   • Pandas

     A library made for data processing, analysis and manipulation. We used it to process the data gathered from spotify to get the stats and the data for the graphs.

     Like the ones above, it can be installed with

     pip install pandas
     

     Be sure to use ver 1.0.5 or above

   • MySQL Connector

     As the name suggest, this library is a way to connect a MySQL database with python. We used it to automate the storing of data to the SQL database.

   For all this libraries, we recommend to follow the instructions provided and to check the libraries you have installed. To do this, type pip list or python -m pip list in the terminal.

   Note: If you have an anaconda/conda enviroment, most of this libraries are already included, if you're not sure about which ones are installed, you can check it with

   conda list | grep <packet_name>
   

   Also, anaconda will try to avoid conflicts between the packages and will try to install missing packages too.

3. Then, you need to get the credentials for your spotify account

   1. Go to the dashboard in the spotify developer page

   2. Log in with your spotify account.

   3. Create a new app by clicking "My New App". If you don't know yet for what purpose you want for your app, tick the "I don't know" option. Give it a title and a description and tick all the boxes in the next page.

   4. On the left side you can find your "Client ID" and your "Client Secret". Sometimes the "spotipy" script requiers a redirect page. For that, go to "Edit Settings" and set "https://www.google.com/" (or any other generic and stable page) as your Redirect URI.

4. Set up MySQL Database

   Installation:

   Linux:

   To install MySQL, open a terminal and type the following:

   sudo apt-get install mysql-server mysql-client
   

   for ubuntu and debian.

   Windows:

   Download the instaler from the page of mysql and execute it. For more information, you can visit the MySQL Windows installation guide

   Setup:

   After the installation, you'll need to create a user. In the termial type sudo su for linux or comand line in administrator mode in windows, then, without leaving the mysql pompt type

   mysql_secure_instalation

   This will begin the instalation and will ask for a password for root, after this, without leaving super user, enter to

   CREATE USER 'user'@'localhost' IDENTIFIED BY 'password';

   Replace user and password with the actual values you are going to use

   Then, create the database with

   CREATE DATABASE db_name;

   replacing db_name with the name of the database

   Now, to create the tables, you can copy the contents from spotify_tracks_table.sql or, outside of the mysql promt, in the terminal type:

   mysql -u user -p db_name < spotify_tracks_table.sql
   

   NOTE: Be sure to cahnge db_name for the name of your database in the sql file. If you're copying from the file, be sure to type

   ALTER DATABASE db_name CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci; 

   This is because, some of the values are not in ascii (like chinese, japanese, korean and thai characters) and will cause an issue with the data base

   To finish the setup, you'll need to input your credentials to a json file with the following format:

   {"user": "user", "password": "password", "host": "127.0.0.1", "database": "db_name", "raise_on_warnings": True}

   Replace user, password and db_name with the ones intended

   And replace "db_config.json" with the name of you file in the script whre the code is:

   config = load_config("db_config.json")

Running the proyect

To run the proyect use

python3 main.py <int>

where <int> is the number of elements that will contain the graphs, you can leave it empty. If you want to not request songs, run the run the scripts separately. See how below

Run separately

To run them separately they you need to do it in a particular order, first, run

python3 collect_data_from_spotify.py

this is only if you need to collect new data from spotify, and then, type

python3 -c "import processing_and_graphs; processing_and_graphs.main(<int>)

where <int> is an optional parameter that represents how many songs will be displayed in the graphs, it defaults to 10. If done correctly, it should generate a file named Graficas.html that will contain the webpage.

Graphs:

These screenshots represent a part of the graphs that are generated. Due to the size of the second graph, we will only show one example.

This graph represents the songs and the amount of times it has appeared in the playlist

The second graph, represents the song positions in the countries playlist

This is the data qe used represeted in tables.

Webpage:

The webpage looks like this:

Conclusions:

We noticed that, even though they are the popular songs, in the second graphs you can observe how is lossing popularity, meaning that it'll probably not be ther for much longer of if it has mantained it's popularity.

Resources:

Data stream:

• Spotify

Data processing:

• Python 3
• Pandas

Visualization:

• Bokeh
• jinja2
• HTML and CSS

Data storage:

• json files
• MySQL