Blog

Blog about statistics related things.

A new take on documentation sites for R packages

pkgdown is the default way to build a documentation site for an R package, and it’s a great tool. But the support for Quarto is lacking. It does support .qmd vignettes (see v…

Data wrangling in R, Python, and Julia

This post is a replication of Rasmus Bååth’s excellent Why pandas feels clunky when coming from R (licensed CC BY 4.0), which compares tidyverse with pandas.

A response to: A Rant (about R)

Recently, I stumbled across this blog post: A Rant. This rant is about all the things R is bad at according to the author, and I will comment on (some of) the points the…

Hierarchical Bayesian Multinomial Models for Partial Pooling

Consider the problem of finding the best player or deck in a game based on match outcomes. A naive approach is to rank by their win rates, without taking anything else…

Metrics for comparing models

In machine learning and deep learning for classification, models are often evaluated based on metrics such as accuracy, F1-score, and others. This is for example done at pape…