Learn R Programming

StratifiedSampling (version 0.4.1)

landingRM: Landing by suppression of variables

Description

This function performs the landing phase of the cube method using suppression of variables proposed by Chauvet and Tillé (2006).

Usage

landingRM(A, pikstar, EPS = 1e-07)

Value

A vector with elements equal to 0 or 1. The value 1 indicates that the unit is selected while the value 0 is for rejected units.

Arguments

A

matrix of auxiliary variables on which the sample must be balanced. (The matrix should be divided by the original inclusion probabilities.)

pikstar

vector of updated inclusion probabilities by the flight phase. See ffphase

EPS

epsilon value

Author

Raphaël Jauslin raphael.jauslin@unine.ch

References

Chauvet, G. and Tillé, Y. (2006). A fast algorithm of balanced sampling. Computational Statistics, 21/1:53-62

See Also

fbs, balstrat.

Examples

Run this code
N <- 1000
n <- 10
p <- 4
pik <- rep(n/N,N)
X <- cbind(pik,matrix(rgamma(N*p,4,25),ncol= p))
pikstar <- ffphase(X,pik) 
s <- landingRM(X/pik*pikstar,pikstar)
sum(s)
t(X/pik)%*%pik
t(X/pik)%*%pikstar
t(X/pik)%*%s

Run the code above in your browser using DataLab