Posts

Updated 27 October 2016: Added note about getting a “final” bib file Lately I’ve been putting in a not insignificant amount of effort to make sure that my research is reproducible. Ideally, this means that it’s not just the final product that is reproducible — it should be reproducible at each step along the way. One tool I’ve been increasingly reliant on is GNU make, which is a way to describe dependencies between files.

I’ve been meaning to get everything off of evernote for some time now. I’ve never really liked their service. It always seemed sluggish and the UI/UX wasn’t great. This week, Evernote announced some limitations on free accounts that make it mostly useless now. So I decided to get my data out of there and into org-mode I didn’t have much, really, anyway - mostly recipes. It was over 100 different notes, though, so manually copy/pasting was out of the question.

A few people have told me that my previous blog post has really helped them out. I’ve also noticed that my own website occasionally gets cloned. I’ve done quite a bit with my website, though, so it’s not quite as “clean” as I’d like it to be, were I showing someone how to get their own site set up. To that end, I created a github pages academic starter kit repository on github that you can fork and clone to get started with your own website as quickly as possible.

I co-taught a workshop (with Daniel Weitzel) about how to use R. The workshop was geared towards people who have seen R before, but aren’t comfortable using it. We walked through the basic workflow: importing data, data manipulation, data merging, creating tables, and visualization.

In case others are interested, I’ve put the materials on github.

R’s packaging system is great, but there are a few minor annoyances it has. I think I’ve solved them, and wanted to share my setup so that others can fix these annoyances as well. I use Linux, and this setup works well there. It should work on Mac and Windows as well. If you run into problems, though, feel free to comment on this post. Where R installs packages By default, R installs packages somewhere quite obnoxious.

Update 19-May-2016: broken link removed and link to the github repo added instead I ran across this post (link broken, github repo) announcing a weekly R/python visualization challenge and decided that this was the perfect excuse I needed to brush up on my mapping skills. An R script with all the code in one place is available on my github repo. Here, we’ll graph the lower 48 states and color them depending on how many drone sightings have been reported to the FAA in 2014 and 2015 total.

UPDATE - 12 May 2016 - I have created a github repository that is designed so that you can fork it and get started with your own website as quickly as possible. When you go on the academic job market nowadays, it’s definitely the expectation that you have a personal website that you keep relatively up-to-date. Since this is by no means a painless process, this post is intended to be an guide on how to get started with your own website.

The Southern Political Science Association’s 2016 Annual Meeting just finished up in San Juan, Puerto Rico. Since I’ve been meaning to learn the twitteR package for a little bit, I figured that this was a perfect time to do a fun little analysis. You can find the R code I used to generate the images on Github. I started out by grabbing all the tweets containing the official conference tag (#SPSA2016).

I recently ran a review session for a graduate level statistics course. This isn’t meant to be a comprehensive guide to everything we covered, but I figured that maybe someone would find it useful. It goes through maximum likelihood estimation and method of moments estimation. It also gives some examples for each of those. I also wrote up a bit about statistical significance and power. If you’re interested, you can find the pdf here

Combining code and the document that you’re writing is called “literate programming”. It’s one of the best practices associated with reproducible research, which is a hot topic in political science and other disciplines. This post isn’t about why you should do this, it’s just about how to do it. I’m going to talk about how to combine R code with markdown via rmarkdown and LaTeX via knitr. There are other ways of doing this (such as org-mode in emacs), so this isn’t supposed to be a comprehensive guide to all of the ways of doing this.