complexity: Calculates complexity of a prediction rule ensemble.

In marjoleinF/premixed: Derives mixed-effects prediction rule ensembles

Description

complexity returns the complexity (total number of variables in the ensemble, number of terms and mean number of variables per term) of a prediction rule ensembles (i.e., an object of class 'pre' or 'premixed').

Usage

complexity(object, penalty.par.val = "lambda.1se", ...)

Arguments

object	an object of class 'pre' or premixed'.
penalty.par.val	As usual.
...	not currently used.

Value

Returns a vector with the total number of variables in the ensemble, the total number of terms (i.e., baselearners with a non-zero coefficient), and the mean number of variables per term) in the ensemble.

Examples

set.seed(42)
airq <- airquality[complete.cases(airquality),]
airq.ens1 <- premixed(Ozone ~ 1 | Month | Solar.R + Wind + Temp + Day, data = airq, ntrees = 10)
complexity(airq.ens1)

marjoleinF/premixed documentation built on May 27, 2019, 4:50 a.m.