Add multicore/parallel computing option

[pinusm]

This adds quite a few dependencies, but here's a small anecdotal benchmark, using 1000 iterations, on a 5-node network, with 'ggmModSelect':

and here's the same 5-node network, using 5000 iterations, with 'EBICglasso':

I've tested this on a Windows, Ubuntu (in WSL2), and Red Hat Enterprise Linux.

[cvborkulo]

This looks interesting! Will look into it. I did some updates that I haven’t uploaded here yet, so will do that first and then see if this merge still works. Thanks!

[pinusm]

I see the maintainers of the package changed. Please consider the PR. It is quite straight forward to use multiple cores for parallel / independent resampling procedures.

[vandenman]

Hi @pinusm!

Yes, @KarolineHuth and I are taking over the maintenance. Roughly, our idea is the following:

1. do some basic maintenance (e.g., use roxygen2, unit tests, etc.).
2. push to CRAN.
3. take a look at this PR.

So it will be a little longer before we revisit this PR (but probably not that long).

[pinusm]

Sounds like a great plan!
Adding unit testing before this change is definitely a must, and would help evaluate this PR more confidently