mat_cost_dist: Compute cost distances between points on a raster

In graph4lg: Build Graphs for Landscape Genetics Analysis

Compute cost distances between points on a raster

Description

The function computes cost-distances associated to least cost paths between point pairs on a raster with specified cost values.

Usage

mat_cost_dist(
  raster,
  pts,
  cost,
  method = "gdistance",
  return = "mat",
  direction = 8,
  parallel.java = 1,
  alloc_ram = NULL
)

Arguments

raster	A parameter indicating the raster file on which cost distances are computed. It can be: A character string indicating the path to a raster file in format .tif or .asc. A RasterLayer object already loaded in R environment All the raster cell values must be present in the column 'code' from cost argument.
pts	A parameter indicating the points between which cost distances are computed. It can be either: A character string indicating the path to a .csv file. It must have three columns: ID: The ID of the points. x: A numeric or integer indicating the longitude of the points. y: A numeric or integer indicating the latitude of the points. A data.frame with the spatial coordinates of the points. It must have three columns: ID: The ID of the points. x: A numeric or integer indicating the longitude of the points. y: A numeric or integer indicating the latitude of the points. A SpatialPointsDataFrame with at least an attribute column named "ID" with the point IDs. The point coordinates must be in the same spatial coordinate reference system as the raster file.
cost	A data.frame indicating the cost values associated to each raster value. It must have two columns: 'code': raster cell values 'cost': corresponding cost values
method	A character string indicating the method used to compute the cost distances. It must be: 'gdistance': uses the functions from the package gdistance assuming that movement is possible in 8 directions from each cell, that a geo-correction is applied to correct for diagonal movement lengths and that raster cell values correspond to resistance (and not conductance). 'java': uses a .jar file which is downloaded on the user's machine if necessary and if java is installed. This option substantially reduces computation times and makes possible the parallelisation.
return	A character string indicating whether the returned object is a data.frame (return="df") or a pairwise matrix (return="mat").
direction	An integer (4, 8, 16) indicating the directions in which movement can take place from a cell. Only used when method="gdistance". By default, direction=8.
parallel.java	An integer indicating how many computer cores are used to run the .jar file. By default, parallel.java=1.
alloc_ram	(optional, default = NULL) Integer or numeric value indicating RAM gigabytes allocated to the java process when used. Increasing this value can speed up the computations. Too large values may not be compatible with your machine settings.

Value

The function returns:

• If return="mat", a pairwise matrix with cost-distance values between points.

• If return="df", an object of type data.frame with three columns:

  • from: A character string indicating the ID of the point of origin.

  • to: A character string indicating the ID of the point of destination.

  • cost_dist: A numeric indicating the accumulated cost-distance along the least-cost path between point ID1 and point ID2.

Author(s)

P. Savary

Examples

## Not run: 
x <- raster::raster(ncol=10, nrow=10, xmn=0, xmx=100, ymn=0, ymx=100)
raster::values(x) <- sample(c(1,2,3,4), size = 100, replace = TRUE)
pts <- data.frame(ID = 1:4,
                  x = c(10, 90, 10, 90),
                  y = c(90, 10, 10, 90))
cost <- data.frame(code = 1:4,
                   cost = c(1, 10, 100, 1000))
mat_cost_dist(raster = x,
              pts = pts, cost = cost,
              method = "gdistance")

## End(Not run)

graph4lg documentation built on Feb. 16, 2023, 5:43 p.m.