| Title: | Boltzmann Entropy for Spatial Data |
|---|---|
| Description: | Calculates several entropy metrics for spatial data inspired by Boltzmann's entropy formula. It includes metrics introduced by Cushman for landscape mosaics (Cushman (2015) <doi:10.1007/s10980-015-0305-2>), and landscape gradients and point patterns (Cushman (2021) <doi:10.3390/e23121616>); by Zhao and Zhang for landscape mosaics (Zhao and Zhang (2019) <doi:10.1007/s10980-019-00876-x>); and by Gao et al. for landscape gradients (Gao et al. (2018) <doi:10.1111/tgis.12315>; Gao and Li (2019) <doi:10.1007/s10980-019-00854-3>). |
| Authors: | Jakub Nowosad [aut, cre] |
| Maintainer: | Jakub Nowosad <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 0.1.2 |
| Built: | 2024-11-12 06:12:25 UTC |
| Source: | https://github.com/nowosad/bespatial |
Calculates Cushman's configurational entropy for surfaces (2021)
bes_g_cushman(x, nr_of_permutations = 1000, independent = FALSE)bes_g_cushman(x, nr_of_permutations = 1000, independent = FALSE)
x |
SpatRaster, stars, RasterLayer, RasterStack, RasterBrick, matrix, or array containing one or more continuous rasters |
nr_of_permutations |
Number of permutations performed on each input raster to calculate possible distribution of "slope" values |
independent |
Should an independent set of permutations be performed for each input raster?
|
A tibble
Cushman, S. A. (2021). Generalizing Boltzmann Configurational Entropy to Surfaces, Point Patterns and Landscape Mosaics. In Entropy (Vol. 23, Issue 12, p. 1616). MDPI AG. https://doi.org/10.3390/e23121616
library(bespatial) library(terra) gradient = rast(system.file("raster/gradient.tif", package = "bespatial"), lyrs = 1) ce2 = bes_g_cushman(gradient, 100) plot(gradient, main = round(ce2$value, 2)) bes_g_cushman(gradient, 1000, independent = TRUE)library(bespatial) library(terra) gradient = rast(system.file("raster/gradient.tif", package = "bespatial"), lyrs = 1) ce2 = bes_g_cushman(gradient, 100) plot(gradient, main = round(ce2$value, 2)) bes_g_cushman(gradient, 1000, independent = TRUE)
Calculates the Boltzmann entropy of a landscape gradient by Gao (2017, 2019)
bes_g_gao( x, method = "aggregation", na_adjust = TRUE, base = "log10", relative = FALSE )bes_g_gao( x, method = "aggregation", na_adjust = TRUE, base = "log10", relative = FALSE )
x |
SpatRaster, stars, RasterLayer, RasterStack, RasterBrick, matrix, or array. |
method |
A method used. Either "hierarchy" for the hierarchy-based method (Gao et al., 2017) or "aggregation" (default) for the aggregation-based method (Gao et al., 2019). |
na_adjust |
Should the output value be adjusted to the proportion of not missing cells? Either TRUE (default) or FALSE |
base |
A logarithm base ("log", "log2" or "log10"). |
relative |
Should a relative or absolute entropy be calculated? TRUE or FALSE (default). |
The method for computing the Boltzmann entropy of a landscape gradient works on integer values that are either positive or equals to zero. This function automatically rounds values to the nearest integer value (rounding halfway cases away from zero) and negative values are shifted to positive values.
A tibble
Gao, Peichao, Hong Zhang, and Zhilin Li. "A hierarchy-based solution to calculate the configurational entropy of landscape gradients." Landscape Ecology 32.6 (2017): 1133-1146.
Gao, Peichao, Hong Zhang, and Zhilin Li. "An efficient analytical method for computing the Boltzmann entropy of a landscape gradient." Transactions in GIS (2018).
Gao, Peichao and Zhilin Li. "Aggregation-based method for computing absolute Boltzmann entropy of landscape gradient with full thermodynamic consistency" Landscape Ecology (2019)
library(terra) library(bespatial) gradient = rast(system.file("raster/gradient.tif", package = "bespatial")) gg1 = bes_g_gao(gradient) plot(gradient, main = round(gg1$value, 2))library(terra) library(bespatial) gradient = rast(system.file("raster/gradient.tif", package = "bespatial")) gg1 = bes_g_gao(gradient) plot(gradient, main = round(gg1$value, 2))
Calculates Cushman's configurational entropy for landscape mosaics (2015)
bes_m_cushman(x, nr_of_permutations, independent = FALSE)bes_m_cushman(x, nr_of_permutations, independent = FALSE)
x |
SpatRaster, stars, RasterLayer, RasterStack, RasterBrick, matrix, or array containing one or more categorical rasters |
nr_of_permutations |
Number of permutations performed on each input raster to calculate possible distribution of total edge values |
independent |
Should an independent set of permutations be performed for each input raster?
|
A tibble
Cushman, S. A. (2015). Calculating the configurational entropy of a landscape mosaic. In Landscape Ecology (Vol. 31, Issue 3, pp. 481–489). Springer Science and Business Media LLC. https://doi.org/10.1007/s10980-015-0305-2
library(terra) library(bespatial) mosaic = rast(system.file("raster/mosaic.tif", package = "bespatial")) ce1 = bes_m_cushman(mosaic, 1000) plot(mosaic, main = round(ce1$value, 2)) bes_m_cushman(mosaic, 1000, independent = TRUE)library(terra) library(bespatial) mosaic = rast(system.file("raster/mosaic.tif", package = "bespatial")) ce1 = bes_m_cushman(mosaic, 1000) plot(mosaic, main = round(ce1$value, 2)) bes_m_cushman(mosaic, 1000, independent = TRUE)
Calculates Zhao's entropy for landscape mosaics based on the Wasserstein metric (2019)
bes_m_zhao(x, neighbourhood = 4)bes_m_zhao(x, neighbourhood = 4)
x |
SpatRaster, stars, RasterLayer, RasterStack, RasterBrick, matrix, or array containing one or more categorical rasters |
neighbourhood |
The number of directions in which cell adjacencies are considered as neighbours: 4 (rook's case), 8 (queen's case) |
A tibble
Zhao, Y., & Zhang, X. (2019). Calculating spatial configurational entropy of a landscape mosaic based on the Wasserstein metric. Landscape Ecology, 34(8), 1849-1858. https://doi.org/10.1007/s10980-019-00876-x
library(terra) library(bespatial) mosaic = rast(system.file("raster/mosaic.tif", package = "bespatial")) w_dists1 = bes_m_zhao(mosaic) plot(mosaic, main = round(w_dists1$value, 2))library(terra) library(bespatial) mosaic = rast(system.file("raster/mosaic.tif", package = "bespatial")) w_dists1 = bes_m_zhao(mosaic) plot(mosaic, main = round(w_dists1$value, 2))
Calculates Cushman's configurational entropy for point patterns (2021)
bes_p_cushman(x, nr_of_permutations, independent = FALSE)bes_p_cushman(x, nr_of_permutations, independent = FALSE)
x |
SpatRaster, stars, RasterLayer, RasterStack, RasterBrick, matrix, or array containing one or more rasters with one value and NAs |
nr_of_permutations |
Number of permutations performed on each input raster to calculate possible distribution of the number of nearest neighbors |
independent |
Should an independent set of permutations be performed for each input raster?
|
A tibble
Cushman, S. A. (2021). Generalizing Boltzmann Configurational Entropy to Surfaces, Point Patterns and Landscape Mosaics. In Entropy (Vol. 23, Issue 12, p. 1616). MDPI AG. https://doi.org/10.3390/e23121616
library(terra) library(bespatial) point_pattern = rast(system.file("raster/point_pattern.tif", package = "bespatial")) ce3 = bes_p_cushman(point_pattern, 100) plot(point_pattern, main = round(ce3$value, 2)) ce3b = bes_p_cushman(point_pattern, 100, independent = TRUE) plot(point_pattern, main = round(ce3b$value, 2))library(terra) library(bespatial) point_pattern = rast(system.file("raster/point_pattern.tif", package = "bespatial")) ce3 = bes_p_cushman(point_pattern, 100) plot(point_pattern, main = round(ce3$value, 2)) ce3b = bes_p_cushman(point_pattern, 100, independent = TRUE) plot(point_pattern, main = round(ce3b$value, 2))
Calculates an average distance between non-NA cells
get_distance(p, x)get_distance(p, x)
p |
A matrix |
x |
A SpatRaster with proper metadata (e.g., extent and CRS) |
It converts permuted matrix into a vector dataset, and calculates an average distance between the points
An average distance between points
Calculate a slope
get_slope(x, neighbourhood = matrix(4))get_slope(x, neighbourhood = matrix(4))
x |
A matrix |
neighbourhood |
The number of directions in which cell adjacencies are considered as neighbours: 4 (rook's case), 8 (queen's case) or a binary matrix where the ones define the neighbourhood. The default is 4. |
"Slope" is calculated as follows:
For each cell, the algorithm looks at its 4 neighbors and calculates the absolute difference between the main cell and its neighbors.
Next, it sums these four values.
After repeating this operation for every cell, it calculates an average of the sum of the absolute differences for the whole raster.
A slope value
Calculate total edge based on the input matrix
get_total_edge(x, resolution, neighbourhood = as.matrix(4))get_total_edge(x, resolution, neighbourhood = as.matrix(4))
x |
A matrix |
resolution |
A numeric vector with two values representing the input matrix resolution on the x and y axis |
neighbourhood |
The number of directions in which cell adjacencies are considered as neighbours: 4 (rook's case), 8 (queen's case) or a binary matrix where the ones define the neighbourhood. The default is 4. |
A total edge value
Permute values in the input raster
permute_raster(x, nr_of_permutations)permute_raster(x, nr_of_permutations)
x |
SpatRaster object ( |
nr_of_permutations |
Number of permutations performed on each input raster |
A list of matrices