Interpolate by fitting a plane within each tessellation facet

Superseded. Use grid_facet_lm() for the Delaunay case and grid_voronoi() for the Dirichlet one, both of which take the same arguments as everything else here and return a grid rather than a point pattern.

Usage

facets(
  X,
  nx,
  ny,
  x = NULL,
  y = NULL,
  na.v = 0,
  method = c("dirichlet", "delaunay")
)

Arguments

X

spatstat object

nx

number of x coords

ny

number of y coords

x

option input x values

y

optional input y values

na.v

na value

method

dirichlet or delaunay

Value

ppp object

Details

Tessellate a marked point pattern into Dirichlet (Voronoi) cells or Delaunay triangles, fit a linear trend in x and y to the marks falling within each facet, and predict that trend at a set of grid locations.

Worth knowing what the two methods were: a Dirichlet tile contains exactly one point, so method = "dirichlet" fits an intercept and nothing else, and predicts that one point's value across its whole tile. It is nearest neighbour, by way of a linear model per tile in an R loop. grid_voronoi() is the same numbers, and says so.