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
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
|
\name{liesInSupport}
\alias{liesInSupport}
\alias{liesInSupport-methods}
\alias{liesInSupport,UnivarLebDecDistribution,numeric-method}
\alias{liesInSupport,UnivarMixingDistribution,numeric-method}
\alias{liesInSupport,LatticeDistribution,numeric-method}
\alias{liesInSupport,DiscreteDistribution,numeric-method}
\alias{liesInSupport,Distribution,matrix-method}
\alias{liesInSupport,AbscontDistribution,numeric-method}
\alias{liesInSupport,ExpOrGammaOrChisq,numeric-method}
\alias{liesInSupport,Lnorm,numeric-method}
\alias{liesInSupport,Fd,numeric-method}
\alias{liesInSupport,Norm,numeric-method}
\alias{liesInSupport,DExp,numeric-method}
\alias{liesInSupport,Cauchy,numeric-method}
\alias{liesInSupport,Td,numeric-method}
\alias{liesInSupport,Logis,numeric-method}
\alias{liesInSupport,Weibull,numeric-method}
\alias{liesInSupport,Unif,numeric-method}
\alias{liesInSupport,Beta,numeric-method}
\title{Generic Function for Testing the Support of a Distribution }
\description{
The function tests if \code{x} lies in the support of the
distribution \code{object}.
}
\usage{
liesInSupport(object, x, ...)
\S4method{liesInSupport}{UnivarLebDecDistribution,numeric}(object,x, checkFin = FALSE)
\S4method{liesInSupport}{UnivarMixingDistribution,numeric}(object,x, checkFin = FALSE)
\S4method{liesInSupport}{LatticeDistribution,numeric}(object,x, checkFin = FALSE)
\S4method{liesInSupport}{DiscreteDistribution,numeric}(object,x, checkFin = FALSE)
\S4method{liesInSupport}{AbscontDistribution,numeric}(object,x, checkFin = FALSE)
\S4method{liesInSupport}{Distribution,matrix}(object,x, checkFin = FALSE)
\S4method{liesInSupport}{ExpOrGammaOrChisq,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Lnorm,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Fd,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Norm,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{DExp,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Cauchy,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Td,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Logis,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Weibull,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Unif,numeric}(object,x, checkFin = TRUE)
\S4method{liesInSupport}{Beta,numeric}(object,x, checkFin = TRUE)
}
\arguments{
\item{object}{ object of class \code{"Distribution"} }
\item{x}{ numeric vector or matrix }
\item{checkFin}{ logical: in case \code{FALSE}, we simply check whether
\code{x} lies in the \emph{numerical} (i.e., possibly cut
to relevant quantile range) support; in case \code{TRUE} we try to
check this by more exact techniques (e.g. in case of lattice distributions)
and by using slot \code{.finSupport} / the return values of \code{q.l(object)}
in \code{0} and \code{1}. This is only used on discrete (parts of)
distributions).}
\item{\dots}{ used for specific arguments to particular methods. }
}
%\details{}
\value{logical vector}
\section{Methods}{
\describe{
\item{object = "DiscreteDistribution", x = "numeric":}{
We return a logical vector of the same length as \code{x} with \code{TRUE}
when \code{x} lies in the support of \code{object}.
As support we use the value of \code{support(object)}, so this
is possibly cut to relevant quantile ranges.
In case \code{checkFin} is \code{TRUE}, in addition, we flag those coordinates
to \code{TRUE} where \code{x < min(support(object))} if
\code{is.na(object@.finSupport[1])} or \code{object@.finSupport[1]==FALSE}
or \code{q.l(object)(0)==-Inf}, and similarly, where
\code{x > max(support(object))} if \code{is.na(object@.finSupport[2])}
or \code{object@.finSupport[2]==FALSE} or \code{q.l(object)(1)==Inf}.
In addition we flag those coordinates to \code{TRUE} where
\code{q.l(object)(0)<=x<min(support(object))} if
\code{object@.finSupport[1]==TRUE} and, similarly, where
\code{q.l(object)(1)>=x>max(support(object))} if
\code{object@.finSupport[2]==TRUE}.
}
\item{object = "Distribution", x = "matrix":}{
Argument \code{x} is cast to vector and then the respective
\code{liesInSupport} method for vectors is called. The method throws an
arror when the dispatch mechanism does not find a suitable, applicable
respective vector-method.
}
\item{object = "AbscontDistribution", x = "numeric":}{
We return a logical vector of the same length as \code{x} with \code{TRUE}
where \code{q.l(object)(0)<=x<=q.l(object)(1)} (and replace the boundary
values by \code{q.l(object)(10*.Machine$double.eps)} resp.
\code{q.l(object)(1-10*.Machine$double.eps)} once the return values
for \code{0} or \code{1} return are \code{NaN}.
}
\item{object = "LatticeDistribution", x = "numeric":}{
We return a logical vector of the same length as \code{x} with \code{TRUE}
when \code{x} lies in the support of \code{object}.
As support we use the value of \code{support(object)}, so this
is possibly cut to relevant quantile ranges.
In case \code{checkFin} is \code{TRUE}, we instead use the lattice
information: We check whether all values
\code{(x-pivot(lattice(object))/width(lattice(object))} are non-negative
integers and are non larger than \code{Length(lattice(object))-1}.
In addition, we flag those coordinates to \code{TRUE} where
\code{x < min(support(object))} if
\code{is.na(object@.finSupport[1])} or \code{object@.finSupport[1]==FALSE},
and similarly, where \code{x > max(support(object))} if
\code{is.na(object@.finSupport[2])}
or \code{object@.finSupport[2]==FALSE}.
}
\item{object = "UnivarLebDecDistribution", x = "numeric":}{
We split up \code{object} into discrete and absolutely continuous
part and for each of them apply \code{liesInSupport} separately;
the two return values are combined by a coponentwise logical \code{|}.
}
\item{object = "UnivarMixingDistribution", x = "numeric":}{
We first cast \code{object} to \code{UnivarLebDecDistribution}
by \code{flat.mix} and then apply the respective method.
}
}}
%
%\references{}
\author{Matthias Kohl \email{Matthias.Kohl@stamats.de} and
Peter Ruckdeschel \email{peter.ruckdeschel@uni-oldenburg.de}}
%\note{}
\seealso{\code{\link{Distribution-class}}}
\examples{
liesInSupport(Exp(1), rnorm(10))
# note
x <- rpois(10, lambda = 10)
liesInSupport(Pois(1), x)
# better
liesInSupport(Pois(1), x, checkFin = TRUE)
liesInSupport(Pois(1), 1000*x, checkFin = TRUE)
liesInSupport(-10*Pois(1), -10*x+1, checkFin = TRUE)
xs = c(1000*x,runif(10))
D <- UnivarMixingDistribution(Pois(1),Unif())
liesInSupport(D, xs)
}
\keyword{distribution}
\keyword{methods}
\concept{space}
\concept{S4 space class}
\concept{support}
|
