Create an explicit leave-future-out validation plan

The plan defines sequential training cut-points within a single ordered series. Execution is deliberately separate because it requires refitting.

Usage

create_pupil_lfo_plan(
  fit,
  initial_fraction = 0.6,
  horizon = 5L,
  step = 5L,
  max_refits = 8L
)

Arguments

fit

An advanced fitted model.

initial_fraction

Initial fraction of each series available for the earliest training set.

horizon

Number of future samples scored at each refit.

step

Number of samples by which the training cut moves.

max_refits

Maximum refits per model.

Value

A gp3bayes_pupil_lfo_plan object.