featured
Multilevel models for social network data
How we structured a mixed-model analysis when observations are nested inside a social network rather than a simple hierarchy.
Jan Wijffels statisticscase-study
Blog
New entries land here first — filter by tag to find what's relevant.
How we structured a mixed-model analysis when observations are nested inside a social network rather than a simple hierarchy.
Notes on the latest udpipe release: tokenisation performance and new language models.
What tends to break first when a scraping pipeline moves from a prototype to a daily production job.