fold_funs: Cross-Validation Schemes

Description Usage Arguments Value See Also

Description

These functions represent different cross-validation schemes that can be used with origami. They should be used as options for the fold_fun argument to the make_folds function in this package. make_folds will call the requested function specify n, based on its arguments, and pass any remaining arguments (e.g. V or pvalidation) on.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
folds_vfold(n, V = 10)

folds_resubstitution(n)

folds_loo(n)

folds_montecarlo(n, V = 1000, pvalidation = 0.2)

folds_bootstrap(n, V = 1000)

folds_rolling_origin(n, first_window, validation_size, gap = 0, batch = 1)

folds_rolling_window(n, window_size, validation_size, gap = 0, batch = 1)

Arguments

n

(integer) - number of observations.

V

(integer) - number of folds.

pvalidation

(double) - proportion of observation to be in validation fold.

first_window

(integer) - number of observations in the first training sample.

validation_size

(integer) - number of points in the validation samples; should be equal to the largest forecast horizon.

gap

(integer) - number of points not included in the training or validation samples; Default is 0.

batch

(integer) - Increases the number of time-points added to the training set each CV iteration. Applicable for larger time-series. Default is 1.

window_size

(integer) - number of observations in each training sample.

Value

A list of Folds.

See Also

Other fold generation functions: fold_from_foldvec, folds2foldvec, make_folds, make_repeated_folds


origami documentation built on March 18, 2018, 1:25 p.m.