Estimating functions for inhomogeneous spatial point processes with incomplete covariate data |
| |
Authors: | Waagepetersen Rasmus |
| |
Institution: | Department of Mathematical Sciences, Aalborg University, Fredrik Bajersvej 7G, DK-9220 Aalborg, Denmark rw{at}math.aau.dk |
| |
Abstract: | The R package spatstat provides a very flexible and useful frameworkfor analysing spatial point patterns. A fundamental featureis a procedure for fitting spatial point process models dependingon covariates. However, in practice one often faces incompleteobservation of the covariates and this leads to parameter estimationerror which is difficult to quantify. In this paper, we introducea Monte Carlo version of the estimating function used in spatstatfor fitting inhomogeneous Poisson processes and certain inhomogeneouscluster processes. For this modified estimating function, itis feasible to obtain the asymptotic distribution of the parameterestimators in the case of incomplete covariate information.This allows a study of the loss of efficiency due to the missingcovariate data. |
| |
Keywords: | Asymptotic normality Cluster process Estimating function Experimental design Inhomogeneous point process Missing covariate data Poisson process |
本文献已被 Oxford 等数据库收录! |
|