Head/tail breaks is a clustering algorithm scheme for data with a heavy-tailed distribution such as power laws and lognormal distributions.
This does not set a cutoff based on the percentage of data the head represents but rather you set the number of bins you would like (like all the others), this is similar to CartoDB's Postgresql implementation.
limit : Basics.Int -> Chroma.Limits.Analyze.Scale -> List.Nonempty.Nonempty Basics.Float
Create bins number of results using the given scale.