HiGHS solver bindings

[dirkschumacher]

I create a small, experimental package to use HiGHS with ompr. Still a bit WIP, but already useable.

https://github.com/dirkschumacher/ompr.highs

HiGHS is a great open-source solver. More on HiGHS here: https://highs.dev

[datadrivensupplychain]

Would you recommend this package over ROI.plugin.highs? I was able to solve a toy model with ROI.plugin.highs but I couldn't get it to take even a MIPGap argument.

[dirkschumacher]

Since ROI.plugin.highs is on CRAN it may be better suited to depend on - depends on your use case. Not sure what arguments the ROI plugin registered. If you have a specific use case in mind maybe it's worth it to engage with the maintainer and see if you can contribute something.

[sbmack]

Dirk, thanks for the effort with HiGHS. Unfortunately:

Installing from: https://github.com/dirkschumacher/ompr.highs

Throws a Warning:

• installing source package 'ompr.highs' ...
  ** using staged installation
  ** R
  ** byte-compile and prepare package for lazy loading
  Note: possible error in 'highs_solve(L = as.numeric(obj$solution), ': unused argument (dry_run = FALSE)
  ** help
  *** installing help indices
  ** building package indices
  ** testing if installed package can be loaded from temporary location
  ** testing if installed package can be loaded from final location
  ** testing if installed package keeps a record of temporary installation path
• DONE (ompr.highs)

Executing the BIN Packing problem throws an error:

Error in highs_solve(L = as.numeric(obj$solution), A = constraints$matrix,  : 
  unused argument (dry_run = FALSE)

[datadrivensupplychain]

@sbmack I forked the repo, commented out dry_run=FALSE in the code, and installed from my github. Works fine now. I also made a pull request to Dirk.