09 → EDA: 30,000 spotify songs
View GitHub Repository Here
This project focuses on an in-depth exploratory analysis of a comprehensive Spotify dataset containing 32,833 tracks with 23 attributes, ranging from basic metadata to detailed audio features. The goal was to uncover insights related to track popularity, audio characteristics, and playlist categorization.
Key Areas of Analysis:
- Track Popularity: Investigating distribution patterns and factors influencing popularity.
- Audio Features: Examining attributes like danceability, energy, and valence to understand their impact on listener preferences.
- Playlist Genre Distribution: Evaluating the diversity of genres and their representation across the dataset.
- Musical Key & Mode Analysis: Assessing how different keys and modes correlate with emotional tone and popularity.
Notable Insights:
- Tracks exhibit a moderate to high danceability and energy level, with key distribution favoring commonly used musical scales such as C sharp/D flat.
- A significant relationship exists between musical keys, modes, and emotional tone (valence), with statistical validation (p-value ~ 5.54E-34).
- Major keys tend to dominate the dataset, often associated with more uplifting and popular tracks.
Technical Approach:
- Employed descriptive statistics and visualizations (pie charts, bar charts, and boxplots) to illustrate key findings.
- Statistical tests were conducted to validate hypotheses related to key-mood relationships.
- Data was analyzed to detect potential trends useful for music recommendation systems or marketing strategies.
This project showcases a strong analytical approach to data exploration, with actionable insights for the music industry, emphasizing the interplay between musical attributes and track performance.