sim.partially.matched: Simulate Paired, Independent, or Partially Matched Two-Sample...

In robustrank: Robust Rank-Based Tests

Description

sim.partially.matched generates partially matched two-sample data. for Monte Carlo studies. r2sample is a wrapper for sim.partially.matched and generates indepenent two-sample data.

Usage

sim.partially.matched(m, n.x, n.y, 
distr = c("normal","logistic","student","mixnormal","gamma","lognormal","beta",
    "uniform","hybrid1","hybrid2","doublexp"), params, seed)

r2sample(m, n, 
distr = c("normal", "logistic", "student", "mixnormal"), params, seed)

sim.paired.with.replicates(m, meanRatio, sdRatio, within.sd, type, hyp, distr, seed)

Arguments

m	Number of pairs.
n	Number of Ys.
n.x	Number of extra Xs.
n.y	Number of extra Ys.
distr	Distributions.
params	Named vector. See details.
seed	Seed for random number generator.
meanRatio
sdRatio
within.sd
type
hyp

Details

If the distribution is in c("normal","student","logistic"), params should have three fields: loc.2, rho and scale.2; loc.1 is set to 0 and scale.1 is set to 1.

If the distribution is mixnormal, params should have three fields: p.1, p.2 and sd.n.

If the distribution is gamma, params should have fix fields: loc.2, shape.1, shape.2, rate.1, rate.2 and rho.

For details on bivariate logistic distribution, see rbilogistic

Value

sim.partially.matched return a list with the following components:

X	m sample 1 that pair with Y
Y	m sample 2 that pair with X
Xprime	n.x sample 1
Yprime	n.y sample 2

r2sample returns a list with the following components:

X	m sample 1 that are independent of Y
Y	n sample 2 that are independent of X

Examples

dat=sim.partially.matched(m=10,n.x=5,n.y=4,distr="normal",
    params=c("loc.2"=0,"rho"=0,"scale.2"=1),seed=1)
X=dat$X; Y=dat$Y; Yprime=dat$Yprime    

#dat=sim.partially.matched(m=10,n.x=5,n.y=4,distr="logistic",
#    params=c("loc.2"=0,"rho"=0,"scale.2"=1),seed=1)
#X=dat$X; Y=dat$Y; Yprime=dat$Yprime    

robustrank documentation built on July 15, 2020, 5:08 p.m.