🎤 Rap Quant Research

Quantitative consumer insights analysis for Hip-Hop music data
Built with Python · Pandas · Scikit-learn · Plotly · Streamlit

---

📌 Project Overview

This project applies quantitative research methods to Hip-Hop music data to extract consumer insights — combining audio feature analysis, NLP sentiment scoring, and machine learning to understand what drives track popularity.

---

📊 Key Results

Metric	Value
ML Model R² (test set)	~0.70
CV R² (5-fold)	~0.68 ± 0.04
Top popularity driver	Danceability
Sentiment range	-0.43 (Dark) → +0.75 (Positive)
Tracks analysed	500+

---

📈 Charts

🌡️ Feature Correlation Heatmap

💃 Danceability vs Popularity

🏆 Popularity by Artist

🤖 ML Feature Importance

🗺️ Consumer Segment Map (PCA)

💬 Lyric Sentiment Analysis

---

🧰 Tech Stack

Layer	Tools	Status
Data processing	pandas numpy scikit-learn	✅
Statistical analysis	scipy statsmodels	✅
NLP	vaderSentiment	✅
Machine Learning	GradientBoostingRegressor PCA	✅
Visualisation	plotly	✅
Dashboard	streamlit	✅
Data source	Kaggle Spotify Hip-Hop Dataset	✅

With a Streamlit app

To Get Your Real Streamlit app Run the appstreamlit.py along with the data that we shared.