Visualizing Rhymes in Raps with Genomic Algorithms and D3!
How can code help us understand what makes great raps? We built this app to find out.
RapGraph was built in beaker, using python for processing, and d3.js for visualization. Used nltk to translate word to phoneme (with cmudict). Implements Smith-Waterman alignment algorithm (typically used in genomics) to align phonemes from adjacent lines.
TTR: text to rap. Then the entire pipeline, for a rap battling robot.
Visit www.RapGraph.me
git clone https://github.com/el2724/rapper.git rapGraph
pip install virtualenv
virtualenv .
source bin/activate
pip install -r requirements.txt
python webapp/app_rap.py
Beaker Notebook is a program that lets you projects that combine 16+ different programming languages. You can open any of the .bkr
files in the research
folder if you have
It is recommended to install dependencies using pip
, the Python package manager. If you add external libraries, run
pip freeze > requirements.txt