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
|
\name{respModule-class}
\Rdversion{1.1}
\docType{class}
\alias{respModule-class}
\alias{glmRespMod-class}
\alias{nlsRespMod-class}
\alias{nglmRespMod-class}
\alias{fitted,respModule-method}
\alias{residuals,respModule-method}
\alias{residuals,glmRespMod-method}
\title{"respModule" and derived classes}
\description{
The \code{"respModule"} class is the virtual base class of
response modules for \code{\linkS4class{glpModel}} model objects.
Classes that inherit from \code{"respModule"} include
\code{\linkS4class{glmRespMod}}, for
generalized linear models, \code{\linkS4class{nlsRespMod}}, for
nonlinear models and \code{\linkS4class{nglmRespMod}} for generalized
nonlinear models.
}
\section{Objects from the Class}{
Objects from these classes are usually created with
\code{\link{mkRespMod}} as part of an \code{\linkS4class{glpModel}}
object returned by model-fitting functions such as the hidden function
\code{glm4}.
}
\section{Slots}{
\describe{
\item{\code{mu}:}{Fitted mean response.}
\item{\code{offset}:}{offset in the linear predictor -- always
present even if it is a vector of zeros. In an
\code{\linkS4class{nlsRespMod}} object the length of the offset can be
a multiple of the length of the response.}
\item{\code{sqrtXwt}:}{the matrix of
weights for the model matrices, derived from the \code{sqrtrwt} slot.}
\item{\code{sqrtrwt}:}{Numeric vector of the square roots
of the weights for the residuals. For
\code{respModule} and \code{\linkS4class{nlsRespMod}}
objects these are constant. For
\code{\linkS4class{glmRespMod}} and \code{\linkS4class{nglmRespMod}}
objects these are updated at each iteration of the iteratively
reweighted least squares algorithm.}
\item{\code{weights}:}{Prior
weights -- always present even when it is a vector of ones.}
\item{\code{y}:}{Numeric response vector.}
\item{\code{family}:}{a glm family, see \code{\link{family}} for
details - \code{glmRespMod} objects only.}
\item{\code{eta}:}{numeric vector, the linear predictor that is
transformed to the conditional mean via the link function -
\code{glmRespMod} objects only.}
\item{\code{n}:}{a numeric vector used for calculation of the aic family
function (it is really only used with the binomial family but
we need to include it everywhere) - \code{glmRespMod} objects only.}
\item{\code{nlenv}:}{an environment in which to evaluate the
nonlinear model function - \code{nlsRespMod} objects only.}
\item{\code{nlmod}:}{an unevaluated call to the
nonlinear model function - \code{nlsRespMod} objects only.}
\item{\code{pnames}:}{a character vector of parameter names -
\code{nlsRespMod} objects only.}
}
}
\section{Methods}{
\describe{
\item{fitted}{\code{signature(object = "respModule")}: fitted values;
there may be several types, corresponding to the residuals, see there
(below).}
\item{residuals}{\code{signature(object = "respModule")}: residuals,
depending on the type of the model, there are several types of
residuals and correspondingly residuals,
see \code{\link[stats:glm.summaries]{residuals.glm}} from the
\pkg{stats} package. Because many of these types of residuals are
identical except for objects that inherit from "glmRespMod", a
separate method is defined for this subclass.}
}
}
\seealso{\code{\link{mkRespMod}}}
\examples{
showClass("respModule")
showClass("glmRespMod")
showClass("nlsRespMod")
}
\keyword{classes}
|
