Maps

Map objects in BayesX

In BayesX, map objects are used to handle, store and visualize geographical information. In the current version, map objects serve as auxiliary objects for regression analyses, where the effect of spatial covariates on a dependent variable can be modelled via Markov random field priors or other spatial effects. In addition, they allow the visualization of any geographical information associated with the map.

Map objects can be created from boundary files or graph files, compare Section 5 of the manual for a description of these file formats.

Creating boundary files and graph files

In principle, boundary and graph files can be created from scratch by applying the structure described in Section 5 of the reference manual, provided that the boundary information is available. The BayesX R package provides some functionality to ease this task. For example, function shp2bnd allows to convert shapefiles to boundary files. There are also functions to convert boundary files to graph files and to manipulate boundary and graph files. For example, additional neighbors can be added to an existing graph specification.

Available boundary and graph files

On this page we provide boundary files as well as graph files for some maps that have already been used with BayesX. We also request users of BayesX to provide us their maps so that we can make them available on this page.

Germany:

For the map of Germany we provide several maps, since between 1993 and 1996 the boundaries and number of districts changed in the eastern part of Germany. In 2001 the two districts of Hanover were united. Moreover the island Rügen causes the map to be disconnected resulting in problems when estimating structured spatial effects and therefore we provide one version including Rügen and one version without Rügen. The regioncodes are given by the key numbers of the administrative districts.

Munich:

The regioncodes are given by the key numbers of the administrative districts.

Nigeria:

The regioncodes are defined arbitrarily. A picture of the map containing the regioncodes can be found here.

Tanzania:

The regioncodes are defined arbitrarily. A picture of the map containing the regioncodes can be found here.

Zambia:

The regioncodes are defined arbitrarily. A picture of the map containing the regioncodes can be found here.