Estimate the type-specific probabilities for a multivariate Poisson point process with independent component processes of each type.
phat(gpts, pts, marks, h)
matrix containing the x,y
-coordinates of the
point locations at which type-specific probabilities are estimated.
matrix containing the x,y
-coordinates of the data points.
numeric/character vector of the types of the point in the data.
numeric value of the bandwidth used in the kernel regression.
A list with components
matrix of the type-specific probabilities for all types, with the type marks as the matrix row names.
copy of the arguments pts, dpts, marks, h
.
The type-specific probabilities for data \((x_i, m_i)\), where \(x_i\) are the spatial point locations and \(m_i\) are the categorical mark sequence numbers, \(m_i=1,2,\ldots\), are estimated using the kernel smoothing methodology \(\hat p_k(x)=\sum_{i=1}^nw_{ik}(x)I(m_i=k)\), where \(w_{ik}(x)=w_k(x-x_i)/\sum_{j=1}^n w_k(x-x_j)\), \(w_k(.)\) is the kernel function with bandwidth \(h_k>0\), \(w_k(x)=w_0(x/h_k)/h_k^2\), and \(w_0(\cdot)\) is the standardised form of the kernel function.
The default kernel is the Gaussian. Different kernels can be
selected by calling setkernel
.
Diggle, P. J. and Zheng, P. and Durr, P. A. (2005) Nonparametric estimation of spatial segregation in a multivariate point process: bovine tuberculosis in Cornwall, UK. J. R. Stat. Soc. C, 54, 3, 645--658.
cvloglk
, mcseg.test
, and
setkernel