EXPLORATORY METHODS FOR SPATIAL AND TEMPORAL DATA

点击翻译按钮获取中文摘要

The objective of our research is the development of an S-PLUS module S+SpaceTime as an environment for exploratory data analysis of spatial- temporal data. Data with space and time attributes are very common in epidemiology and biometry. Special exploratory methods are desired in order to study these kinds of data for the simultaneous assessment of environmental exposures and health outcomes observed over time. Geographical Information Systems (GIS) are useful in visualizing spatial data. To simultaneously assess health outcomes with GIS requires further research on interface, graphical methods, and spatial-temporal modeling. Dynamic graphics is one of the best approaches to visualize spatial-temporal data. Dynamic graphics with GIS are computing intensive research. S-PLUS is an object-oriented programming language and is capable to link with Java and ArcView GIS. The S+SpaceTime will provide a visualization and modeling environment. User will be able to link statistical graphical windows and a GIS map in the processes of space-time modeling. A comprehensive case study guidebook will be developed to demonstrate the advantages of the environment. The resulting module is useful for studies on longitudinal data, spatial data, and spatial-temporal data. PROPOSED COMMERCIAL APPLICATIONS: No commercial statistical software is available to visualize and explore disease incidence maps and repeated measures simultaneously in the processes of modeling. S+SpaceTime provides dynamic graphics with GIS and space-time modeling based on mixed effects models and models for space-time processes. With S+SpatialStat, S+EnvironmentalStat, and S-PLUS for ArcView, the proposed software R&D is a continuation of efforts on these substantial areas and markets.