kperm is an R package for efficiently estimating Ripley’s K under complete spatial randomness using a robust statistical approach called KAMP (K adjusted by Analytical Moments of the Permutation distribution).
KAMP avoids the need for computationally expensive permutations while accounting for spatial inhomogeneity, making it suitable for large-scale spatial analyses such as those encountered in spatial proteomics multiplex imaging datasets.
Overview
This package provides functions to compute both univariate and bivariate KAMP expectation and variance (spatstat and matrix-based implementation both included). At this time, the only edge correction methods supported are translational (trans) and isotropic (iso).
Basic Functionality and Examples
Univariate
library(kperm)
# Simulate a point pattern
pp <- sim_pp_data(lambda_n = 500, abundance = 0.3)
# Compute KAMP expectation
kamp_expec_univ <- kamp_expectation(pp, markvar = "immune")
print(kamp_expec_univ)
# Compute KAMP variance
kamp_var <- kamp_variance(pp, markvar = "immune")
print(kamp_var)Bivariate
library(kperm)
# Simulate a point pattern
pp <- sim_pp_data_biv(lambda_n = 500, abundance = 0.3)
# Compute KAMP expectation
kamp_expec_biv <- kamp_expectation_biv(pp)
print(kamp_expec_biv)
# Compute KAMP variance
kamp_var_biv <- kamp_variance_biv(pp)
print(kamp_var_biv)