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
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
|
% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/fix.R
\name{intercept}
\alias{intercept}
\alias{intercept<-}
\alias{intercept.lvm}
\alias{intercept<-.lvm}
\alias{intfix}
\alias{intfix<-}
\alias{intfix.lvm}
\alias{intfix<-.lvm}
\title{Fix mean parameters in 'lvm'-object}
\usage{
\method{intercept}{lvm}(object, vars, ...) <- value
}
\arguments{
\item{object}{\code{lvm}-object}
\item{\dots}{Additional arguments}
\item{vars}{character vector of variable names}
\item{value}{Vector (or list) of parameter values or labels (numeric or
character) or a formula defining the linear constraints (see also the
\code{regression} or \code{covariance} methods).}
}
\value{
A \code{lvm}-object
}
\description{
Define linear constraints on intercept parameters in a \code{lvm}-object.
}
\details{
The \code{intercept} function is used to specify linear constraints on the
intercept parameters of a latent variable model. As an example we look at
the multivariate regression model
\deqn{ E(Y_1|X) = \alpha_1 + \beta_1 X} \deqn{ E(Y_2|X) = \alpha_2 + \beta_2
X}
defined by the call
\code{m <- lvm(c(y1,y2) ~ x)}
To fix \eqn{\alpha_1=\alpha_2} we call
\code{intercept(m) <- c(y1,y2) ~ f(mu)}
Fixed parameters can be reset by fixing them to \code{NA}. For instance to
free the parameter restriction of \eqn{Y_1} and at the same time fixing
\eqn{\alpha_2=2}, we call
\code{intercept(m, ~y1+y2) <- list(NA,2)}
Calling \code{intercept} with no additional arguments will return the
current intercept restrictions of the \code{lvm}-object.
}
\note{
Variables will be added to the model if not already present.
}
\examples{
## A multivariate model
m <- lvm(c(y1,y2) ~ f(x1,beta)+x2)
regression(m) <- y3 ~ f(x1,beta)
intercept(m) <- y1 ~ f(mu)
intercept(m, ~y2+y3) <- list(2,"mu")
intercept(m) ## Examine intercepts of model (NA translates to free/unique paramete##r)
}
\seealso{
\code{\link{covariance<-}}, \code{\link{regression<-}},
\code{\link{constrain<-}}, \code{\link{parameter<-}},
\code{\link{latent<-}}, \code{\link{cancel<-}}, \code{\link{kill<-}}
}
\author{
Klaus K. Holst
}
\keyword{models}
\keyword{regression}
|
