About

I’m Calvin. My interests are in Bayesian machine learning (deep and “shallow”), functional programming, and mathematics. In my past life, I was a graduate student studying algebraic topology and homotopy theory. This background gave me a natural interest in statically-typed languages and category theory, but I haven’t been able to find a way to combine them with my day job yet.

> why haskell?

When I first started learning how to program last year, I started with Python like any good data scientist and quickly got frustrated by runtime errors, mistakes in inputs, and a frame of thinking that I wasn’t entirely used to. Imperative languages force you to think in terms of mutable state, which meant that the way I reason with it is different than what my mathematical training tells me to do. Although now I am fairly competent with it, I don’t think I have ever really fallen in love with Python the way other data scientists say they do.

Studying homotopy theory, when everyone started raving about the newstart theory of homotopy type theory I was curious in learning about it foundationally, as it promised this intimate link between computer science and algebraic topology. Since then, even though I don’t really understand the hype behind HoTT, it gave me an appreciation for functional languages and type systems, of which Haskell has one of the best. (That, and no one ever actually uses Idris, right?!)

> how did you build this site?

This site is built on Hakyll and is deployed automatically with CircleCI. I recommend anyone who doesn’t know how these tools work to try it out– it’ll take several hours if you don’t know what you’re doing, but you will learn so much from it (as I did).