[R] simulation data in SEM

[PO SU]

PLZ mke sure the package installed which contains "mvrnorm" function.




--

PO SU
mail: desolator88 at 163.com 
Majored in Statistics from SJTU




At 2014-09-14 09:56:34, "thanoon younis" <thanoon.younis80 at gmail.com> wrote:
>Dear R members
>I want to simulate data depending on SEM and when i applied the code below
>i found some errors and i still cannot run it.
>many thanks in advance
>
>
>Thanoon
>
>#Do simulation for 100 replications
>N<-1000; P<-10
>
>phi<-matrix(data=c(1.0,0.3,0.3,1.0),ncol=2) #The covariance matrix of xi
>Ro<-matrix(data=c(7.0,2.1,2.1,7.0), ncol=2)
>yo<-matrix(data=NA,nrow=N,ncol=P) p<-numeric(P); v<-numeric(P)
>
>for (t in 1:100) {
>    #Generate the data for the simulation study
>    for (i in 1:N) {
>        #Generate xi
>        xi<-mvrnorm(1,mu=c(0,0),phi)
>        #Generate the fixed covariates
>        co<-rnorm(1,0,1)
>        #Generate error term is structural equation
>        delta<-rnorm(1,0,sqrt(0.3))
>        #Generate eta1 according to the structural equation
>        eta<-0.8*co[i]+0.6*xi[1]+0.6*xi[2]+0.8*xi[1]*xi[2]+delta
>        #Generate error terms in measurement equations
>        eps<-rnorm(3,0,1)
>
>        #Generate theta according to measurement equations
>        v1[1]<-1.0+eta+eps[1]; v1[2]<-1.0+0.7*eta+eps[2]
>        v1[3]<-1.0+0.7*eta+eps[3]
>        v1[4]<-1.0+xi[1]; v1[5]<-1.0+0.8*xi[1]; v1[6]<-1.0+0.8*xi[1]
>        v1[7]<-1.0+xi[2]; v1[8]<-1.0+0.7*xi[2]; v1[9]<-1.0+0.7*xi[2];
>        v1[10]<-1.0+0.7*xi1[2]
>
>
>        #transform theta to orinal variables
>        for (j in 1:10) { if (v[j]>0) yo[i,j]<-1 else yo[i,j]<-0 }
>
>
>    #Input data set for WinBUGS
>    data<-list(N=200,P=10,R=Ro,z=yo)
>
>}   #end
>
>	[[alternative HTML version deleted]]
>
>______________________________________________
>R-help at r-project.org mailing list
>https://stat.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>and provide commented, minimal, self-contained, reproducible code.