Tuning Parameters of the SIMCA Model

dials parameters for the classification rule of the simca() parsnip model (and of rsimca()):

• simca_gamma(), for gamma, the weight of the orthogonal distances against the score distances;

• simca_squared(), for squared, the combination of the squared scaled distances (rule R2) or of the scaled distances (rule R1).

Usage

simca_gamma(range = c(0, 1), trans = NULL)

simca_squared(values = c(TRUE, FALSE))

Arguments

range

A two-element vector with the range of gamma. Default is c(0, 1).

trans

A transformation object from the scales package, or NULL (default) for none.

values

The values of squared. Default is c(TRUE, FALSE).

Value

A quant_param object (simca_gamma()) or a qual_param object (simca_squared()).

See also

simca(), rsimca()

Examples

simca_gamma()
#> Weight of the orthogonal distances (quantitative)
#> Range: [0, 1]
dials::value_seq(simca_gamma(), 5)
#> [1] 0.00 0.25 0.50 0.75 1.00
simca_squared()
#> Squared scaled distances (qualitative)
#> 2 possible values include:
#> TRUE and FALSE