I previously wrote about collecting and graphing my keystrokes. That system worked well for a while, but the graphs started looking old and lame, especially after I started learning how to use ggplot2 in a data visualization course I’m currently taking.

At the same time, I’ve also wanted to learn how to use git/GitHub for a while.1 I happened to decide to migrate the personal analytics graphing to using ggplot2, and start learning git seriously at the same time.

So I created the project and jumped right in.

You can find the persanalytics GitHub page here. Right now, there is one .Rmd R markdown file that I wrote in RStudio that crunches all the data and makes all the plots. But I’ve realized that as this project grows, one file won’t cut it. I will refactor the project soon and split the script into smaller scripts each of which handles one type of data.

For now, these are a couple of the plots it gives me:

polar plot of all-time data split by day

plot of when during the day I am typing, over time (I am especially proud of this one)

Right now, I have serious plans to start crunching the following:

  • Cycling data.
  • last.fm data. (I’ve already requested my archive and I can’t wait.)
  • Chat log data (currently in .xml format and goes back many years).

I’m excited.

  1. Seriously you guys. git is amazing. And not just for programming. It’s great for keeping track of revisions of your posts, manuscripts, or any files that are in plain text format. When Octopress 3.0 is done, I’ll redesign this site and start tracking it with git.