search

guilgautier / DPPy

Python toolbox for sampling Determinantal Point Processes
https://dppy.readthedocs.io
MIT License
219 stars 52 forks source link

Interfacing with the R library spatstat #41

Closed rbardenet closed 2 years ago

rbardenet commented 5 years ago

spatstat has DPP sampling for the stationary kernels that appear in spatial statistics. This includes smart tricks to deal with observation windows, Fourier series that approximate Fourier transforms, estimating functional statistics, etc. Maybe we could use some R API to quickly add their algorithms to DPPy.

Describe the solution you'd like A class for stationary kernels on the plane, where the sampling method would call spatstat behind the scenes.

Describe alternatives you've considered An alternative solution would be to re-implement their algorithms.

Additional context In any case, if we consider this issue at some point, we should discuss with Ege Rubak and the Aalborg group.

guilgautier commented 2 years ago

For-a-few-DPPs-more/spatstat-interface provides a small wrapper for the spatstat R package. An interplay between DPPy and spatstat-interface seems to be a good option.