bnspatial (version 0.9)

ConwyData: Land use change data

Description

Data derived from the Conwy catchment in North Wales (UK), widely modified for demonstration purposes. Once loaded, the data consist of several objects:

• LandUseChange An object of class grain. The Bayesian network.
• currentLU An object of class RasterLayer. A simplified version of the current land use map from the Conwy catchment (Wales, UK). It includes three classes: arable (raster value 3), forest (2), other (1).
• slope An object of class RasterLayer. A raster of slope derived from a digital elevation model at 50 meters resolution, units are degrees.
• status An object of class RasterLayer. The land ownership type (dummy data), divided into three possible classes: public (raster value 4), private (3), protected (1).
• evidence A matrix. The collection of available spatial data (see above) as extracted from each location (i.e. cell) in the catchment, where the latter is represented by the raster object currentLU. Each value from the spatial data was discretized through dataDiscretize or bulkDiscretize functions, then assigned to the corresponding state from the Bayesian network (LandUseChange).
• LUclasses A list with the classification of input spatial data (its corresponding states and values). The list is formatted accordingly to bnspatial functions requirement and as returned by functions importClasses and setClasses.

Usage

data(ConwyData)

Format

A dataset in native RData format.

Examples

library(bnspatial)
data(ConwyData)
ls()

## The network nodes and states
LandUseChange$universe$levels

## Lookup list relating raster values and network nodes
LUclasses

## Table of evidence extracted from input spatial data
head(evidence)

## The input spatial data
par(mfrow=c(2,2))
raster::plot(currentLU)
raster::plot(slope)
raster::plot(status)