Alias Method Sampling
Build an alias table for weighted categorical draws and compare exact probabilities with reproducible samples.
Free theory articleBuild an alias table for weighted categorical draws and compare exact probabilities with reproducible samples.
Concept & interpretation
A categorical distribution chooses one of finitely many outcomes with specified probabilities. The alias method represents this distribution using equally likely columns. Each column contains at most two outcome labels: its own label and an alias. After preprocessing, a draw needs one uniformly selected column and one threshold comparison, regardless of the number of outcomes.