Colin

SongSim. React web app for drawing self-similarity matrices from text

char-rbm. Character-level RBMs for short text

pejorative-compounds. Analysing patterns in English noun-noun pejorative compounds on Reddit

reddit-dubious-spelling. Analyzing words Redditors aren't sure how to spell

lalala. Investigating repetitiveness in pop music using compression algorithms

instacart-basket-prediction. RNNs (and other stuff) for Kaggle's Instacart order prediction competition

wiki-pageview-floor. Finding and analysing the least viewed articles on English Wikipedia

tour-of-heroes. some crummy incremental game made with angular 2

atypicality. Jupyter Notebook

reponames-dataset. Dataset of 4.6m GitHub repository names

sketch-rnn-experiments. Experiments with Sketch-RNN and the Quick Draw dataset

crawl-coroner. Analyzing completed games ('morgue files') of Dungeon Crawl Stone Soup

favicon-scraper. Scraping a dataset of favicons.

colinmorris.github.com. HTML

lm1b-notebook. Various scripts used while playing around with Google Brain's billion word language model

song-repetition. Are Pop Lyrics Getting More Repetitive?

snl-notebooks. Some ipython notebooks analyzing SNL data

csc236. Course website for CSC236 winter 2020

reddit-username-suffixes. Little one-off experiment analysing the frequency of numerical suffixes in reddit usernames

circles-of-hell-ngrams. An ngram experiment

reddit-misspelling-trends. Trends in the frequency of common misspellings on Reddit over time

size-of-an-x. Looking at the changing frequencies of different size analogies in Google Books ngram data.

Tone.js. A Web Audio framework for making interactive music in the browser.

unique-country-prefixes. Infographic ranking countries by the length of the shortest unique prefix that identifies them

moz-graphs. Python

testify. Unit test generator proof of concept (intended to be used in Exercism)

openai-gym-sandbox. Experiments with OpenAI gym environments (https://gym.openai.com/)

lm-sentences. Web app for visualizing how a language model assigns probabilities to sentences

wiki-controversial-titles. Analyzing Wikipedia articles with the most debated titles

xscreensaver-text-toys. Shell

wiki-listicle-generation. Semi-automatically writing Wikipedia list articles using Wikidata/Wikipedia APIs

tldr. :books: Simplified and community-driven man pages

rm_stats. Web app for visualizing rename discussions on Wikipedia