ml docs

Split, fit, evaluate, assess. Same API in Python and R.

import ml

s = ml.split(data, "target", seed=42)
model = ml.fit(s.train, "target", seed=42)
ml.evaluate(model, s.valid)
ml.assess(model, test=s.test)

library(ml)

s <- ml_split(data, "target", seed = 42)
model <- ml_fit(s$train, "target", seed = 42)
ml_evaluate(model, s$valid)
ml_assess(model, test = s$test)

The boundary

evaluate uses validation data — call it freely. assess uses test data and locks after one call. The types enforce it.

Verbs