A is for Auto-Recurrence:

Exploring Song Lyrics with Auto-Recurrence Analysis

This project contains a written walkthrough exploring patterns in song lyrics through the lens of Auto-Recurrence Quantification Analysis (aRQA). A webpage describing the methods and insights from this project can be found at: http://tehrandav.is/a_to_z_project/auto_recurrence.html

Dependencies

This project was written in Julia 1.9.0. The following packages are required to run the code:

using CSV, # For reading CSV files
    DataFrames, # For working with dataframes
    Random, # For generating random numbers
    PrettyTables, # For printing tables
    TextAnalysis, # For text preprocessing
    Statistics, # For statistical analysis
    PyPlot, # For plotting
    Tidier, # For data wrangling
    StatsBase # For statistical functions
    DynamicalSystems # For calculating recurrence quantification analysis

Repo Contents

auto_rqa
├─ README.md
├─ data
│  ├─ lyrics_series.csv (lyrics series for 1 song)
│  └─ playlist_lyrics.csv (sample dataset)
├─ notebooks
│  └─ auto_recurrence.ipynb (notebook walkthrough)
└─ src (Julia environment files)
   ├─ .DS_Store
   ├─ Manifest.toml
   ├─ Project.toml
   └─ auto_rqa.jl

Usage

• auto_recurrence.ipynb: A Jupyter Notebook containing code for analyzing the song lyrics data.

• playlist_lyrics.csv: A CSV file containing song lyrics for a selection of songs. The lyrics were obtained from this dataset on Kaggle.

Project Overview:

• Auto-Recurrence Quantification Analysis (aRQA) is a data analysis technique used to uncover patterns and structures in time series data. In this project, we apply aRQA to song lyrics to reveal and quantify rhythmic and thematic elements.

Contributors:

• Tehran Davis, 2023, tehrandav.is

License:

• This project is open-source and available under the MIT License.