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
|
\name{evaluateLogConDens}
\alias{evaluateLogConDens}
\title{Evaluates the Log-Density MLE and Smoothed Estimator at Arbitrary Real Numbers xs}
\description{Based on a \code{"dlc"} object generated by \code{\link{logConDens}}, this function computes the values of
\deqn{\widehat \phi_m(t)}{\hat \phi_m(t)}
\deqn{\widehat f_m(t) = \exp(\widehat \phi_m(t))}{\hat f_m(t) = exp(\hat \phi_m(t))}
\deqn{\widehat F_m(t) = \int_{x_1}^t \exp(\widehat \phi_m(x)) dx}{\hat F_m(t) = int_{x_1}^t exp(\hat \phi_m(x)) dx}
\deqn{\widehat f_m^*(t) = \exp(\widehat \phi_m^*(t))}{\hat f_m^*(t) = exp(\hat \phi_m^*(t))}
\deqn{\widehat F_m^*(t) = \int_{x_1}^t \exp(\widehat \phi_m^*(x)) dx}{\hat F_m^*(t) = int_{x_1}^t \exp(\hat \phi_m^*(x)) dx}
at all real number \eqn{t} in \code{xs}. The exact formula for \eqn{\widehat F_m}{\hat F_m} and \eqn{t \in [x_j,x_{j+1}]} is
\deqn{\widehat F_m(t) = \widehat F_m(x_j) + (x_{j+1}-x_j) J\Big(\widehat \phi_j, \widehat \phi_{j+1}, \frac{t-x_j}{x_{j+1}-x_j} \Big)}{\hat F_m(t) = \hat F_m(x_j) + (x_{j+1}-x_j) J(\hat \phi_j, \hat \phi_{j+1}, (t-x_j)/(x_{j+1}-x_j))}
for the function \eqn{J} introduced in \code{\link{Jfunctions}}. Closed formulas can also be given for \eqn{\widehat f_m^*(t)}{\hat f_m^*(t)}
and \eqn{\widehat F_m^*(t)}{\hat F_m^*(t)}.
}
\usage{evaluateLogConDens(xs, res, which = 1:5, gam = NULL, print = FALSE)}
\arguments{
\item{xs}{Vector of real numbers where the functions should be evaluated at.}
\item{res}{An object of class \code{"dlc"}, usually a result of a call to \code{logConDens}.}
\item{which}{A (sub-)vector of \code{1:5} specifying which of the above quantities should be computed.}
\item{gam}{Only necessary if \code{smoothed = TRUE}. The standard deviation of the normal kernel. If equal to
\code{NULL}, \code{gam} is chosen such that the variances of the original sample \eqn{x_1, \ldots, x_n}
and \eqn{\widehat f_n^*}{\hat f_n^*} coincide. See \code{\link{logConDens}} for details.}
\item{print}{Progress in computation of smooth estimates is shown.}
}
\value{Matrix with rows \eqn{(x_{0, i}, \widehat \phi_m(x_{0, i}), \widehat f_m(x_{0, i}), \widehat F_m(x_{0, i}), \widehat f_m^*(x_{0, i}), \widehat F_m^*(x_{0, i}))}{(x_{0, i}, \hat \phi_m(x_{0, i}), \hat f_m(x_{0, i}), \hat F_m(x_{0, i}), \hat f_m^*(x_{0, i}), \hat F_m^*(x_{0, i}))}
where \eqn{x_{0,i}} is the \eqn{i}-th entry of \code{xs}.}
\author{
Kaspar Rufibach, \email{kaspar.rufibach@gmail.com}, \cr \url{http://www.kasparrufibach.ch}
Lutz Duembgen, \email{duembgen@stat.unibe.ch}, \cr \url{https://www.imsv.unibe.ch/about_us/staff/prof_dr_duembgen_lutz/index_eng.html}}
\examples{
## estimate gamma density
set.seed(1977)
x <- rgamma(200, 2, 1)
res <- logConDens(x, smoothed = TRUE, print = FALSE)
## compute function values at an arbitrary point
xs <- (res$x[100] + res$x[101]) / 2
evaluateLogConDens(xs, res)
## only compute function values for non-smooth estimates
evaluateLogConDens(xs, res, which = 1:3)
}
\keyword{htest}
\keyword{nonparametric}
|
