rrp.dist {rrp}R Documentation

RRP dissimilarity matrix

Description

Main piece of the RRP algorithm which generates the RRP dissimilarity matrix.

Usage

rrp.dist(X, treated = NULL, msplit = 10, Rep = 250, cut.in = 15, 
 check.bal = FALSE, plot = FALSE)

Arguments

X a data.frame object
treated optional class indicator variable
msplit minimum split parameter in the rpart algorithm
Rep number of RRP replications
cut.in number of breaks in which to cut continuous variables
check.bal indicator function. If TRUE balance check using hyper-rectangles will be used inside leaves
plot wheter to plot the porximity matrix as image

Details

This algorithm allows for missing data in X.

Value

an object of class dist

Author(s)

S.M. Iacus

References

Iacus, S.M., Porro, G. (2006) Random recursive partitioning: a matching method for the estimation of the average treatment effect

See Also

rpart


[Package rrp version 1.4 Index]