is it possible to run spatstat functions on multiple processors

Question

I am curios to know if spastat functions like envelope or MAD tests can be run on multiple processors on a machine to speed up calculations? Is there any document or tutorial to do this?

Thank you.

Answer 1

Unfortunately, parallelisation is not an integrated part of spatstat, but rather left to the user. For envelopes and MAD tests the easiest option is probably to run envelope with a smaller number of realisations on each core and then combine the results using pool.envelope. How to run envelope in parallel may depend on your setup. A simple possibility is to use parallel::mclapply which I know works out of the box on linux, but much better cross platform alternatives are surely available in packages on CRAN:

library(spatstat)
ppplist <- replicate(4, cells, simplify = FALSE)
envlist <- parallel::mclapply(ppplist, spatstat::envelope, savefuns = TRUE, nsim = 10)
envfinal <- do.call(pool, envlist)
envfinal
#> Pointwise critical envelopes for K(r)
#> and observed value for 'X[[i]]'
#> Obtained from 40 simulations of CSR
#> Alternative: two.sided
#> Significance level of pointwise Monte Carlo test: 2/41 = 0.0488
#> .....................................................................
#>      Math.label     Description                                      
#> r    r              distance argument r                              
#> obs  hat(K)[obs](r) observed value of K(r) for data pattern          
#> theo K[theo](r)     theoretical value of K(r) for CSR                
#> lo   hat(K)[lo](r)  lower pointwise envelope of K(r) from simulations
#> hi   hat(K)[hi](r)  upper pointwise envelope of K(r) from simulations
#> .....................................................................
#> Default plot formula:  .~r
#> where "." stands for 'obs', 'theo', 'hi', 'lo'
#> Columns 'lo' and 'hi' will be plotted as shading (by default)
#> Recommended range of argument r: [0, 0.25]
#> Available range of argument r: [0, 0.25]