densityAdaptiveKernel.splitppp {spatstat.explore}

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

## S3 method for class 'splitppp'
densityAdaptiveKernel(X, bw=NULL, ..., weights=NULL)

## S3 method for class 'ppplist'
densityAdaptiveKernel(X, bw=NULL, ..., weights=NULL)

Arguments

X

Split point pattern (object of class "splitppp" created by split.ppp) to be smoothed. Alternatively a list of point patterns, of class "ppplist".

bw

Smoothing bandwidths. See Details.

...

Additional arguments passed to densityAdaptiveKernel.ppp. These may include arguments that will be passed to bw.abram.ppp to compute the smoothing bandwidths if bw is missing, and arguments passed to as.mask to control the spatial resolution of the result.

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 X, using densityAdaptiveKernel.ppp.

The argument bw specifies smoothing bandwidths for the data points. Normally it should be a list, with the same length as x. The entry bw[[i]] will determine the smoothing bandwidths for the pattern x[[i]], and may be given in any format acceptable to densityAdaptiveKernel.ppp. For example, bw[[i]] can be a numeric vector of length equal to npoints(x[[i]]), a single numeric value, a pixel image (object of class "im"), an expression, or a function of class "funxy". For convenience, bw can also be a single expression, or a single pixel image, or a single function. If bw is missing or NULL, the default is to compute bandwidths using bw.abram.ppp.

The argument weights specifies numerical case weights for the data points. Normally it should be a list, with the same length as x. The entry weights[[i]] will determine the case weights for the pattern x[[i]], and may be given in any format acceptable to density.ppp. For example, weights[[i]] can be a numeric vector of length equal to npoints(x[[i]]), a single numeric value, a numeric matrix, a pixel image (object of class "im"), an expression, or a function of class "funxy". For convenience, weights can also be a single expression, or a single pixel image (object of class "im"), or a single function of class "funxy". If weights is missing or NULL, all weights are assumed to be equal to 1.

Value

A list of pixel images (objects of class "im") which can be plotted or printed; or a list of numeric vectors giving the values at specified points.

Author(s)

Adrian Baddeley [e-mail].

See Also

densityAdaptiveKernel.ppp, 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")