dse (version 2020.2-1)

minForecastCov: Minimum Forecast Cov Models

Description

Extract the minimum forecastCov at each horizon

Usage

minForecastCov(obj, series=1, verbose=TRUE)

Arguments

obj

An object as returned by stripMine.

series

An indicator of the series which are to be used as the bases for selection.

verbose

If true additional information is printed.

Value

The returned object is a vector indicating the element of forecastCov which was the min at each horizon.

Details

Select the min covariance (for series only!) at each horizon and print. The returned object is a vector indicating the element of forecastCov which was the min at each horizon. It is suitable as an argument to plot eg: tfplot(obj, select.cov=minForecastCov(obj)) The results of this plot are similar to the default results of tfplot(selectForecastCov(obj)). Covariance information and information about the horizon where the model is optimal are given.

See Also

selectForecastCov, excludeForecastCov

Examples

# NOT RUN {
data("eg1.DSE.data.diff", package="dse")
z <- stripMine(eg1.DSE.data.diff, essential.data=c(1,2),
                   estimation.methods=list(estVARXls=list(max.lag=3)))
z <-  minForecastCov(z)
# }