pcf3est
Pair Correlation Function of a Three-Dimensional Point Pattern
Estimates the pair correlation function from a three-dimensional point pattern.
- Keywords
- spatial, nonparametric
Usage
pcf3est(X, ..., rmax = NULL, nrval = 128, correction = c("translation",
"isotropic"), delta=NULL, adjust=1, biascorrect=TRUE)
Arguments
- X
- Three-dimensional point pattern (object of class
"pp3"
). - ...
- Ignored.
- rmax
- Optional. Maximum value of argument $r$ for which $g_3(r)$ will be estimated.
- nrval
- Optional. Number of values of $r$ for which $g_3(r)$ will be estimated.
- correction
- Optional. Character vector specifying the edge correction(s) to be applied. See Details.
- delta
- Optional. Half-width of the Epanechnikov smoothing kernel.
- adjust
- Optional. Adjustment factor for the default value of
delta
. - biascorrect
- Logical value. Whether to correct for underestimation due to truncation of the kernel near $r=0$.
Details
For a stationary point process $\Phi$ in three-dimensional
space, the pair correlation function is
$$g_3(r) = \frac{K_3'(r)}{4\pi r^2}$$
where $K_3'$ is the derivative of the
three-dimensional $K$-function (see K3est
).
The three-dimensional point pattern X
is assumed to be a
partial realisation of a stationary point process $\Phi$.
The distance between each pair of distinct points is computed.
Kernel smoothing is applied to these distance values (weighted by
an edge correction factor) and the result is
renormalised to give the estimate of $g_3(r)$.
The available edge corrections are: [object Object],[object Object]
Kernel smoothing is performed using the Epanechnikov kernel
with half-width delta
. If delta
is missing, the
default is to use the rule-of-thumb
$\delta = 0.26/\lambda^{1/3}$ where
$\lambda = n/v$ is the estimated intensity, computed
from the number $n$ of data points and the volume $v$ of the
enclosing box. This default value of delta
is multiplied by
the factor adjust
.
The smoothing estimate of the pair correlation $g_3(r)$
is typically an underestimate when $r$ is small, due to
truncation of the kernel at $r=0$.
If biascorrect=TRUE
, the smoothed estimate is
approximately adjusted for this bias. This is advisable whenever
the dataset contains a sufficiently large number of points.
Value
- A function value table (object of class
"fv"
) that can be plotted, printed or coerced to a data frame containing the function values.Additionally the value of
delta
is returned as an attribute of this object.
References
Baddeley, A.J, Moyeed, R.A., Howard, C.V. and Boyde, A. (1993) Analysis of a three-dimensional point pattern with replication. Applied Statistics 42, 641--668.
Ohser, J. (1983) On estimators for the reduced second moment measure of point processes. Mathematische Operationsforschung und Statistik, series Statistics, 14, 63 -- 71.
Ripley, B.D. (1977) Modelling spatial patterns (with discussion). Journal of the Royal Statistical Society, Series B, 39, 172 -- 212.
See Also
Examples
X <- rpoispp3(250)
Z <- pcf3est(X)
Zbias <- pcf3est(X, biascorrect=FALSE)
if(interactive()) {
opa <- par(mfrow=c(1,2))
plot(Z, ylim.covers=c(0, 1.2))
plot(Zbias, ylim.covers=c(0, 1.2))
par(opa)
}
attr(Z, "delta")