SpatEntropy (version 0.1.0)

batty: Batty's entropy.

Description

This function computes Batty's spatial entropy, following Batty (1976), see also Altieri et al (2017) (references are under the topic SpatEntropy).

Usage

batty(data, data.assign, is.pointdata = FALSE, category, win = NULL,
  G.coords)

Arguments

data

A data matrix or vector, can be numeric, factor, character, ... If the dataset is a point pattern, data is the mark vector.

data.assign

A three column matrix, containing the data coordinates (centroids when pixels) and the id of the corresponding sub-area. Provided by user or returned by areapart().

is.pointdata

Logical: T if data are a point pattern, F if they are pixels.

category

A character string, the exact name of the category for which Batty's spatial entropy is computed, as in data.

win

An owin object (see package spatstat), the observation area. Only needed for lattice data.

G.coords

A point pattern (an object of class ppp see package spatstat), or a two column matrix with the area centroids' coordinates. Provided by user or returned by areapart().

Value

Batty's spatial entropy value, as well as a table with information about each sub-area:

• area.id the sub-area id

• abs.freq the number of points/pixels presenting the category of interest for each sub-area

• rel.freq the proportion of points/pixels presenting the category of interest in each sub-area, with regard to the total number of points/pixels with the category of interest

• Tg the sub-area size.

Details

Batty's spatial entropy measures the heterogeneity in the spatial distribution of a phenomenon of interest, with regard to an area partition. It is high when the phenomenon is equally intense over the sub-areas, and low when it concentrates in one or few sub-areas. This function starts from the output of areapart() and allows to compute Batty's entropy as $$H=\sum p_g \log(T_g/p_g)$$ where \(p_g\) is the probability of occurrence of the phenomenon over sub-area \(g\), and \(T_g\) is the sub-area size. When data are categorical, the phenomenon of interest corresponds to one category, which must be specified. If data are an unmarked point pattern, a fake mark vector must be created with the same category for all points.

Examples

# NOT RUN {
#LATTICE DATA
data.lat=matrix(sample(c("a","b","c"), 100, replace=TRUE), nrow=10)
ccc=coords_pix(area=square(10), nrow=10, ncol=10)
partition=areapart(square(10), G=5, data.coords=ccc)
batty(data.lat, partition$data.assign, category="a",
win=square(10), G.coords=partition$G.coords)
plot_areapart(partition$data.assign, square(10), is.pointdata=FALSE,
add.data=TRUE, data.bin=TRUE, category="a",
data=data.lat, G.coords=partition$G.coords, main="")

#POINT DATA
data.pp=runifpoint(100, win=square(10))
marks(data.pp)=sample(c("a","b","c"), 100, replace=TRUE)
ccc=coords(data.pp)
partition=areapart(square(10), G=4, data.coords=ccc)
batty(marks(data.pp), partition$data.assign, is.pointdata=TRUE,
category="b", G.coords=partition$G.coords)
plot_areapart(partition$data.assign, square(10), is.pointdata=TRUE,
add.data=TRUE, data.bin=TRUE, category="b",
data=data.pp, G.coords=partition$G.coords, main="")

# }