dist.obstacle: Calculate 2D distances from points to points avoiding...
In statnmap/GeoDist: Constrained distance calculation and associated geotools

Description Usage Arguments Value Examples

View source: R/Calculate_distances_functions.R

Calculate 2D distances from points to points avoiding obstacles

dist.obstacle(
  from,
  to,
  r.ref,
  dref.r,
  n.cores,
  closest = FALSE,
  step.ang = 5,
  r.ref3D = FALSE,
  filename = paste0(tempdir(), "/dref.r.RData"),
  small.dist = TRUE,
  keep.path = FALSE,
  longlat = raster::isLonLat(r.ref),
  tol = 1e-05,
  igraph = FALSE
)

`from`	Point data set from which to calculate distances. May be a 2-col matrix of (x,y) coordinates or a SpatialPointDataFrame. Should be in same projection than r.ref. This can also be an object of class distref.data.
`to`	Point data set from which to calculate distances. May be a 2-col matrix of (x,y) coordinates or a SpatialPointDataFrame. Should be in same projection than r.ref. If not set to = from. This can also be an object of class distref.data. If "to" is provided, it should be the bigger dataset.
`r.ref`	reference raster used for distance without obstacles calculation
`dref.r`	object of class distref.raster. Reference raster already processed using distref.raster (e.g. usefull if previously saved using filename.)
`n.cores`	number of cores to run parallel calculation. Default to all cores using parallel::detectCores()
`closest`	if a point is over an obstacle, set to TRUE to find the closest non-obstacle r.ref cell
`step.ang`	Step for angles classes breaks (in degrees). This means that a distance error is allowed in future calculations. Distance error (%) = 100 * (1 - cos((step.ang/2)*pi/180)). With step.ang = 5 degrees, error < 0.1%. Default set to 5.
`r.ref3D`	Logical. Wether the reference raster should be used as a 3D surface to calculate distance accounting for elevation.
`filename`	Path where to save the dref.r issued from reference raster. This may be usefull for re-using the raster for different datasets. Default to paste0(tempdir(),"/dref.r.RData")
`small.dist`	if TRUE, distances <= 1 paths are directly calculated.
`keep.path`	Logical. Whether to keep SpatialLines of paths. This allows real distances when 2 points are in the same raster cell or close.
`longlat`	Logical. if FALSE, Euclidean distance, if TRUE Great Circle distance. Defined from r.ref.
`tol`	numeric Tolerance to define distances between points to be null.
`igraph`	Logical. Wether to calculate all distances through igraph (TRUE) or use `Pt2Pts.wo.obstacle` to use igraph only when necessary. Using igraph maybe less precise but a little quicker. Default to FALSE.

If keep.path is false, returns a matrix of distances with rows = "from" and cols = "to". If keep.path is true, returns a list where dist.mat is the matrix of distance and listLines is a list of SpatialLines. Each list is a SpatialLinesDataFrame corresponding to a starting point in "from".

## Not run: 
# For "from" and "to" being SpatialPoints
allpath <- dist.obstacle(from, to, r.ref, keep.path = TRUE)
plot(from)
points(to)
# All path from "from[1,]" to all "to"
lines(allpath$allLines[[1]])

## End(Not run)