Estimate the directional distribution of Konig&Schmidt 1992
with (their notation) 0-nearest neighbour i.e. nearest neighbour, the direction
set being a conical wedge of the unit-sphere.
Usage
nn_directional_angle_distribution(x, direction, epsilon, r, antipodal = FALSE)
Arguments
- x
point pattern, $x coords $bbox bounding box
- direction
Unit vector direction
- epsilon
Direction sector's opening half-angle vector.
- r
Maximum range for of nn-distances to consider. If missing, use max(nn-distances)
- antipodal
Make nn-vectors antipodally symmetric?
Details
Compute the probability that a nearest neighbour of a typical point is in direction of the
cone given the distance is less than r.
Input should be one direction and many epsilons OR equal amount of directions and epsilons.
Not using a double cone.