| lgcpestK(InhomCluster) | R Documentation |
Minimum contrast estimation based on K-function for log Gaussian Cox process with exponential covariance function.
lgcpestK=function(t,K,startpar)
t |
vector of distances for which nonparametric estimate of K is evaluated |
K |
vector of corresponding values of nonparametric estimate of K |
startpar |
starting values (sigma^2,alfa) for parameter estimation where sigma^2 is variance of Gaussian field and alfa is correlation scale parameter. |
This function computes minimum contrast estimates of clustering parameters for a log Gaussian Cox process with covariance function sigma^2 exp(-d/alpha). The contrast is the integrated squared distance between the fourth root of a nonparametric estimate of K and the fourth root of the theoretical expression for K.
A list with two components minimum and Ktheoret. The
first components is as provided by the output of optim and the
estimate is given by minimum$par. The second component is the fitted theoretical
K function evaluated at t.
Rasmus Waagepetersen rw@math.aau.dk http://www.math.aau.dk/~rw
Waagepetersen, R. (2006) An estimation function approach to inference for inhomogeneous Neyman-Scott processes, submitted.