Data Visualization

R2D3 : Project for using D3.js with R


From my last presentation, I have mainly used R packages, such as ggplot2 for visualization.

It does offer nice quality, without any post-processing in photoshop or such. One of the best quality of ggplot2 is customizable (tweak lots of stuff) and ease of use.

However, CS backgrounded colleagues mostly used d3.js, Javascript library for visualization. Compared to ggplot2, it felt like a sandbox game to me, sort of Skyrim vs GTA. We all know freedom comes at costs, and the cost here was difficulty in programming. Simple stacked bar chart requires about 100 lines of code, which us statistician cannot stand (all of us are spoiled by ggplot2 and R). 

I wish to create a R package that helps statistician use D3.js with ease. It will require general script mapping, which I think is doable. The main motivation is from “xtable” package in R, which converts table to latex table. But some of the difficulties can rise from interpreting (or imputing) some of the basic parameters for the plotting, which R users are not aware of, that D3.js requires.

I think this will take at least a month to work on, to make a quality package.

Naming the project was a no-brainer. R2D3! Advanced version of R2D2!

Need to make this project to work before next Star Wars movie by Disney so that I have the naming right for R2D3!