A Cox process is a Poisson process with a random intensity; given the
realised intensity it is an inhomogeneous Poisson process.
CoxProcess treats intensity_field (an ny by nx lattice over
window, row iy the iy-th band from ymin) as that realisation and
draws the conditional Poisson process: a Poisson count with mean
max(lambda, 0) times the cell area per cell, placed uniformly (Philox
draws as LgcpSimulate, streams 1000 and up). The
unconditional log-Gaussian Cox process is LgcpSimulate. Identical to
the Python arm morie.fn.sgcox.cox_process.
Usage
CoxProcess(intensity_field, window, seed = 1)
Arguments
- intensity_field
Matrix of intensities (ny rows, nx columns).
- window
Rectangle c(xmin, xmax, ymin, ymax).
- seed
Philox seed.
Value
List: points (two-column matrix), n_points, value, mean_intensity,
expected_count.
References
Cox, D. R. (1955). Some statistical methods connected with
series of events. Journal of the Royal Statistical Society B 17, 129-164.
Moller, J., Syversveen, A. R. and Waagepetersen, R. P. (1998). Log
Gaussian Cox processes. Scandinavian Journal of Statistics 25, 451-482.
Examples
CoxProcess(rbind(c(2, 5), c(1, 8)), c(0, 2, 0, 2), seed = 3)$n_points
#> [1] 15
