distmatGen: Compute cost-based distances

In famuvie/geoRcb: An Extension of Package geoR that Works with Cost-Based Distances

Description

Generate a cost-based distance matrix among observations and/or between observations and prediction locations.

Usage

distmatGen(pts, condsurf, ret = c("both", "obs", "loc"), directions = 16,
  silent = FALSE)

Arguments

pts	A SpatialPoints[DataFrame] with a defined projection or a two-column data.frame/matrix of coordinates
condsurf	raster. Conductivity surface.
ret	specify whether to return distances between obs-obs ("obs"), obs-loc ("loc"), or both ("both") (default)
directions	See transition.
silent	logical. If TRUE avoids any warnings or messages.

Details

Use the centroids of the conductivity or conductivity surface raster cells as prediction locations.

Examples

data(noise)
if (require(raster)) {
  r <- raster(extent(c(0,3,0,3)), nrows = 3, ncols = 3)
  wall.idx <- 4:5
  values(r)[-wall.idx] <- 1
  obs <- coordinates(r)[-wall.idx, ]

  ddm <- distmatGen(obs, r, ret = "obs", directions = 8)

  par(mfrow = c(2, 1))
  plot(r)
  text(obs[, 1], obs[, 2],
       col = c('red', rep('black', 6)))

  plot(dist(obs)[1:6], ddm[-1, 1],
       type = 'n',
       main = 'Distances to point 1',
       xlab = 'Euclidean distance',
       ylab = 'Cost-based distance')
  text(dist(obs)[1:6], ddm[-1, 1],
       labels = 2:7)
  abline(0, 1)
}

famuvie/geoRcb documentation built on May 16, 2019, 10:04 a.m.