Evgeni Chasnovski

My approach and struggles during counting number of bounces in @TweeTTDaily's 5h21m4s world record for the longest duration to control a table tennis ball with a bat.

#rstats overview of a very important topic of statistical uncertainty and inference. In fact, my internal understanding of this was one of the factors that shaped purpose and functional API of #pdqr package.

TIL that #rstats 4.0.0 will change default color palette used in base R. The giveaway was that my 'vdiffr' test with color passed as `col = 2` failed only on R-devel.
Link to news: developer.r-project.org/blosxom.cgi/R-…
Actual palettes: github.com/wch/r-source/b…

This describes a trick I learned when writing an #rstats function, that uses randomness but has non-random nature. It constrains side effects of random generation process to take place only inside a function environment.

In this post I illustrate with #gganimate how changing arguments of `density()` function from 'stats' package affects the output curve.
This is also a very important topic for creating distribution pdqr-functions in my #pdqr #rstats package (echasnovski.github.io/pdqr/).