Ranking estimators for binary (and, for bayes, categorical) response
tensors of shape (L, M, N) (L models, M questions, N trials). Every
method returns { ranking, scores }: ranking[l] is model l's rank
(1 = best) and scores[l] the raw method score (larger is better). The
method option selects the tie convention ("competition" by default).
Each method is exported under an idiomatic camelCase name and a snake_case
alias matching the Python/Julia API.
Model ranking methods, imported from
scorio/rank.Ranking estimators for binary (and, for
bayes, categorical) response tensors of shape(L, M, N)(Lmodels,Mquestions,Ntrials). Every method returns{ ranking, scores }:ranking[l]is modell's rank (1 = best) andscores[l]the raw method score (larger is better). Themethodoption selects the tie convention ("competition"by default).Each method is exported under an idiomatic camelCase name and a snake_case alias matching the Python/Julia API.