Point pattern analysis on spatially aggregated data

Point pattern analysis is a basic but essential analysis in geography and other fields. Point pattern analysis evaluates the spatial pattern of points using their locational data. Locational data, however, are not always available, especially when points represent individuals. Spatial units aggregate the information of individuals to keep their confidentiality, and existing methods of point pattern analysis cannot fully evaluate the spatial point pattern on spatially aggregated data. To fill the research gap, we propose a new method of point pattern analysis on spatially aggregated data. We consider the spatial patterns of points and labels, the latter of which are often referred to as “marked” points in spatial statistics. We propose two statistics to evaluate these patterns, defined based on spatially aggregated data. We test the validity of the statistics through computational experiments. The results indicate the effectiveness of the statistics in a wide variety of situations.