Maximum-likelihood spatial models, point patterns and local privacy
morie_spatial_models.RdThe spatial lag and spatial error models fitted by maximum likelihood, Ripley's K with edge correction, ordinary co-kriging, and k-ary randomised response under local differential privacy.
Value
A list: the spatial parameter with beta, sigma2 and
loglik for the regressions; r, K and L
for the point pattern; prediction with its kriging
variance and weights; and the perturbed reports with a
debiased estimate for the privacy mechanism.
Details
The two regressions are checked against spatialreg::lagsarlm
and spatialreg::errorsarlm – those functions live in
spatialreg, not spdep, since spdep was split. The same
methods exist in the Python package as morie.fn._robust_core.
References
Ord, J. K. (1975) Estimation methods for models of spatial interaction. JASA 70, 120-126.
Ripley, B. D. (1976) The second-order analysis of stationary point processes. Journal of Applied Probability 13, 255-266.
Warner, S. L. (1965) Randomized response. JASA 60, 63-69.