sel.control {changepointsVar}

Auxiliary function for controlling model selection

Description

Auxiliary function as user interface for model selection. Typically only used when calling 'jumpointsVar'

Usage

sel.control(type=c("bic", "rss"), S=1, Cn="2*log(log(n))",
            alg=c("lasso", "stepwise"), edf.psi=TRUE)

Arguments

type

the criterion to be used to perform model selection.

S

if type="rss" the optimal model is selected when the residual sum of squares decreases by the threshold S.

Cn

if type="bic" a character string (as a function of 'n') to specify to generalized BIC. If Cn=1 the standard BIC is used.

alg

which procedure should be used to perform model selection? The value of alg is passed to the argument 'type' of lars.

edf.psi

logical indicating if the number of changepoints should be computed in the model df.

Details

This function specifies how to perform model seletion, namely how many change points should be selected.

Value

A list with the arguments as components to be used by 'jumpointsVar' and in turn by 'lars'.

Author(s)

Gianluca Sottile Maintainer: Gianluca Sottile <[e-mail]>

See Also

jumpointsVar, lars