Computation and art
Normally I wouldn’t be linking favourably to this scatterplot, which has an ill-defined sampling scheme, and where at least the y-axis data are objectively wrong. On the other hand, normally the scatterplot would be there to convey information. In this case it’s just an index to some beautiful animated triangular art
The point, and the relevance to this blog, is the way Matt Daniels has written software to make these pictures (relatively) easy to create.
Incidentally, before anyone starts complaining that sharks and fish are separate, that bit is exactly correct. Fish (typical fish with bones, such as the swordfish in the animation) have a more recent common ancestor with sheep than with sharks.
Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »