When data visualization goes beautifully wrong.
@kara_woo and @ErikaMudrak
My attempt to create a better color ramp produced Art instead of a useful figure. The final, informative version is Fig. S2C here: http://www.sciencedirect.com/science/article/pii/S0960982214006630.
via Elinor Lichtenberg
Miscoding for predicted values and incidental traceplot lead to colourful abstract line art.
Making a map of North America with some points on it.points(mods$lon,mods$lat,col=mods$Module, cex=10)cex was the culprit, but looks cool!
Not sure how this one happened…but kinda glad it happened.
This is what happens when you use plot(y~x) and, due to low caffeine levels, omit the ~x. Surprising, inscrutable, and strangely beautiful— in a word, aRt.
via Susan Letcher
Messed up optimization procedure of a dimensionality reduction technique (a standalone tool, not in R). I have no idea, where does the symmetry come from;)
Plotting a scatterplot (using R/d3.js) with way too big point radiuses (on mtcars R dataset).
I was plotting a bigger dataset with ggplot as stacked graphs, but did some wrong ordering of the data…
No mistake here, just unexpectedly beautiful. This is the modeled %cover of desert annuals as a function of distance from creosotebush, grouped by shrubs in our plot, if it had rained the normal amount.