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
|
\name{densityAdaptiveKernel.splitppp}
\alias{densityAdaptiveKernel.splitppp}
\alias{densityAdaptiveKernel.ppplist}
\title{Adaptive Kernel Estimate of Intensity for Split Point Pattern}
\description{
Computes an adaptive estimate of the intensity function
(using a variable-bandwidth smoothing kernel)
for each of the components of a split point pattern,
or each of the point patterns in a list.
}
\usage{
\method{densityAdaptiveKernel}{splitppp}(X, bw=NULL, \dots, weights=NULL)
\method{densityAdaptiveKernel}{ppplist}(X, bw=NULL, \dots, weights=NULL)
}
\arguments{
\item{X}{
Split point pattern (object of class \code{"splitppp"}
created by \code{\link[spatstat.geom]{split.ppp}}) to be smoothed.
Alternatively a list of point patterns,
of class \code{"ppplist"}.
}
\item{bw}{
Smoothing bandwidths. See Details.
}
\item{\dots}{
Additional arguments passed to
\code{\link{densityAdaptiveKernel.ppp}}.
These may include arguments that will be passed to
\code{\link{bw.abram.ppp}} to compute
the smoothing bandwidths if \code{bw} is missing,
and arguments passed to \code{\link[spatstat.geom]{as.mask}}
to control the spatial resolution of the result.
}
\item{weights}{
Numerical weights for the points. See Details.
}
}
\details{
This function computes a spatially-adaptive kernel estimate of the
spatially-varying intensity for each of the point patterns
in the list \code{X}, using \code{\link{densityAdaptiveKernel.ppp}}.
The argument \code{bw} specifies smoothing bandwidths
for the data points.
Normally it should be a list, with the same length as
\code{x}. The entry \code{bw[[i]]} will determine the
smoothing bandwidths for the pattern \code{x[[i]]}, and may be given in
any format acceptable to \code{\link{densityAdaptiveKernel.ppp}}.
For example, \code{bw[[i]]} can be
a numeric vector of length equal to \code{npoints(x[[i]])},
a single numeric value,
a pixel image (object of class \code{"im"}),
an \code{expression}, or a function of class \code{"funxy"}.
For convenience, \code{bw} can also be a single \code{expression},
or a single pixel image, or a single function.
If \code{bw} is missing or \code{NULL}, the default is to compute
bandwidths using \code{\link{bw.abram.ppp}}.
The argument \code{weights} specifies numerical case weights
for the data points.
Normally it should be a list, with the same length as
\code{x}. The entry \code{weights[[i]]} will determine the
case weights for the pattern \code{x[[i]]}, and may be given in
any format acceptable to \code{\link{density.ppp}}.
For example, \code{weights[[i]]} can be
a numeric vector of length equal to \code{npoints(x[[i]])},
a single numeric value, a numeric matrix,
a pixel image (object of class \code{"im"}),
an \code{expression}, or a function of class \code{"funxy"}.
For convenience, \code{weights} can also be a single \code{expression},
or a single pixel image (object of class \code{"im"}),
or a single function of class \code{"funxy"}.
If \code{weights} is missing or \code{NULL}, all weights are assumed
to be equal to 1.
}
\value{
A list of pixel images (objects of class \code{"im"})
which can be plotted or printed;
or a list of numeric vectors giving the values at specified points.
}
\author{
\adrian.
}
\seealso{
\code{\link{densityAdaptiveKernel.ppp}},
\code{\link{bw.abram.ppp}}.
}
\examples{
X <- amacrine
if(!interactive()) X <- X[c(TRUE,FALSE,FALSE,FALSE)]
Z <- densityAdaptiveKernel(split(X), h0=0.15)
plot(Z, main="Adaptive kernel estimate")
}
\keyword{spatial}
\keyword{methods}
\keyword{smooth}
\concept{Adaptive smoothing}
|
