1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
|
library(OpenMx)
if(mxOption(NULL,"Default optimizer")!="SLSQP"){stop("SKIP")}
mxOption(NULL,"Nudge zero starts","No")
mxOption(NULL,"Calculate Hessian","No")
mxOption(NULL,"Standard Errors","No")
mxd <- data.frame(
y=mxFactor(as.numeric(runif(100)<.5), levels=c(0,1)),
k5=runif(100),
k618=runif(100),
age=round(rnorm(100,30,5)),
wc=rbinom(100,1,.5),
hc=rbinom(100,1,.5),
lwg=rnorm(100))
mxd2 <- mxd
mxd2$y <- as.numeric(mxd2$y)-1
mod2b <- mxModel(
"constrLPM",
mxData(observed=mxd2,type="raw"),
mxMatrix(
type="Full",nrow=1,ncol=7,free=F,labels=c("data.k5","data.k618","data.age","data.wc","data.hc","data.lwg",NA),name="X"),
mxMatrix(type="Full",nrow=1,ncol=1,free=T,
values=0,
labels="a",name="A"),
mxMatrix(type="Full",nrow=7,ncol=1,free=T,
values=0,
labels=paste("b",1:7,sep=""),name="B"),
mxAlgebra(A + X%*%B, name="Yhat"),
mxMatrix(type="Full",nrow=1,ncol=1,free=T,values=1,labels="sigma",lbound=0.0001,name="Sigma"),
mxExpectationNormal(covariance="Sigma",means="Yhat",dimnames=c("y")),
mxFitFunctionML(),
mxConstraint(-1%x%Yhat < 0, name="c1"),
mxConstraint(Yhat - 1 < 0, name="c2")
)
mod2b <- omxCheckWarning(
mxRun(mod2b),
"Constraint 'constrLPM.c2' depends on definition variables; This may not do what you expect. See ?mxConstraint")
|
