Web development tools for interactive data visualization. These are notes originally from a previous post.

Cool gist hack. A bl.ocks.org is a viewer for github gists if you have it titled index.html. People have used this to render beautiful plots and graphs using D3.

D3 uses vocabulary that comes from webstandards of html, css, and svg. An alternative project is Processing.js, which can do alot, but is a whole different set of methods.

D3 basically uses html, css, and svg to produce nice graphics for quantitative data.

Check out this beautiful matrix with different ordering options rendered using D3 on a Les miserables co-occurence dataset from Donald Knuth.

Mike Bostock is the author github/mbostock

boxplot visualizations using d3

Web visualization with R bindings

plotly library for interactive graphs in webbrowser. With python, R, matlab.

Contour plots of the volcano dataset using R --> json --> D3.

Why we don't yet have interactive graphics. Comments about Hadley Wickham working on interactive versions of ggplot. D3 is the best option right now, but javascript based.


Simple examples combing R, gridSVG package and d3.js to make interactive graphics (tool tips on a scatterplot) from any grid based graphics in R (lattice or ggplot2)

Shiny package by RStudio might be the way to go for now with R plots. e.g. here is a matrix with hierarchical clustering on the sides. Could have toggles for different orders. Same with the graphs.