Markov Word Generator

The first is the first program I ever wrote in Haskell. It was initially a Markov speech generator based on Trump's speeches (ugh) circa 2016, but eventually grew to a generator that could train on various public domain sources such as James Joyce, Virginia Woolf, Shakespeare, the Bible, etc. Most of the output was either funny or made absolutely no sense, but occasionally something pretty great came out. The sample here was trained on Shakespeare:

"Falstaff, want of promise in this instant of charity, and by law o'ertake me for our excuse? Paris. I'll follow presently. [Exit]. Cassius. Yet look in their rising so martial! Burgundy. As you of their spleenful sons you like Brutus, none. [exit Imogen] As one bear me love him free? o."

Markov Music Generator

This project, and extension of the previous one, is an experiment in Markov chain generation using midi files from European classical composers from roughly the period from Bach to Debussy. My second project in Haskell, it is still a bit of a learning project. The music these chains generate is mostly unlistenable but it was an interesting experiment using Euterpea, nonetheless. There is a sample below (trained on JS Bach), poor quality because I just recorded it with my phone (sorry hah).