Uses This

A collection of nerdy interviews asking people from all walks of life what they use to get the job done.

A picture of Yihui Xie

Yihui Xie

PhD statistician, R developer

Posted in data, developer, linux

Who are you, and what do you do?

I'm Yihui Xie, a PhD student in statistics at Iowa State University. I blog in both Chinese and English on statistical computing, graphics and programming. I founded an online community for statistics in China in 2006 called Capital of Statistics, which is also the most popular community for R in China. I'm very active on Github. At the moment, I'm working on a book on reproducible research with R and knitr.

What hardware do you use?

I do not know or care a lot about hardware. I use a Thinkpad T510 (15.6-inch screen, 4GB RAM, Intel Core i5 CPU). I have to use a mouse since I'm very old-fashioned and uncomfortable with the touchpad; I have a wireless mouse for my laptop. I won a Kindle from the Strata conference in 2011 by chance.

And what software?

I work under Ubuntu for the most of time with open source software packages. I dislike the default Unity interface like others.

For programming and data analysis, I primarily use R since I'm a statistician. I have created a bunch of R packages including animation, knitr, formatR, Rd2roxygen, R2SWF, fun and cranvas, etc. I use other R packages like ggplot2, gWidgets, roxygen2, Shiny and XML. Emacs was my editor before I switched to RStudio; for small tasks, I just use the default editor gedit. I also use sed, awk, grep and shell scripts frequently.

For typesetting I use LyX which is a wrapper for LaTeX, and I have argued with other people for many times that writing raw LaTeX is not helpful after you have mastered LaTeX, although I absolutely love LaTeX and avoid Word. I'm visually tired of beamer slides (mainly because too many people use the same style), and I use pandoc to create HTML5 slides.

I love the web. Believe it or not, I learned building websites with Macromedia Dreamweaver, ASP and MS Access when I was a Windows user back in 2003. Later I switched to PHP and MySQL (of course). I stuck with WordPress for about 4 years. Finally I came to Jekyll, a blog engine for hackers. I'm pretty happy with it since I will never worry about databases; I just write simple Markdown posts and push them to Github via Git. I learned a tiny bit of Ruby during this process.

Occasionally I write a little bit JavaScript (often with jQuery), e.g. with D3. The other amazing JS application that I use is MathJax, which allows me to write LaTeX math in web pages. I combine the SciAnimator library with my animation package to create animations on the web, which demonstrates concepts and methods in statistics. I also use the Shiny package to develop online and interactive stats demos like ANOVA. For debugging purposes, I use Chrome when I develop web applications.

I use my knitr package to generate reports directly from code (literate programming); the "reports" may include homework, blog posts, websites, papers and books. I build interactive statistical graphics based on Qt and R.

What would be your dream setup?

I do not have a big dream. I guess a screen with a higher resolution will make me happier. By the way, I hope the Internet connection can be cut off for two months every year so I can focus on writing stuff other than coding. Then if you want to file a bug report to me, write me a mail with a return envelope.