alex / frecklebars

Ryven - The Data Science Package

Ryven is a simple flow-based visual scripting runtime environment for Python. I wrote a package that glues Scikit-Learn to Ryven. The goal is to make data science and machine learning more aproachable and easier to use by people with little to no coding experience.

As a proof of concept I solved the Titanic problem on Kaggle using this fork. The link to the notebook explaining the working process is available below.

< GitHub > / < Kaggle Notebook >

Color Neural Net .

Demo neural network.
This was my hello world project in the field of AI, specifically neural networks. Made to learn and learned to make. Simple to explain.
Written in python3 originally, then again a couple months later in JavaScript for an interactive demo.

< Live Demo > / < GitHub >

Sweatshirt .

“To my son Thebe, whose growth and insights inspire me, a thousand kisses”

Markov chain.
Trained on lyrics from various artists such as Earl Sweatshirt, Kanye West, MF DOOM, and many others.
Comes with a simple lyrics scraper used to get the training data from Genius.

< GitHub >

Brobotto .

Discord bot.
< !offline >
Runs on my raspi. Unavailable at this point in time due to hardware issues, and me not caring enough to replace it.

< Invite > / < GitHub >