Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Rmpk – Mixed Integer Linear and Quadratic Programming in R (github.com/r-opt)
11 points by watwatwat123 on Nov 27, 2021 | hide | past | favorite | 9 comments



The inspiration from JuMP is obvious however I couldn't find a list of supported solvers.


I need to document that better, but rmpk ships with a "ROI optimizer". ROI is an R package that provides bindings to most of the popular solvers: https://roi.r-forge.r-project.org

But I am also working on providing custom bindings to different solvers, to be able to have more control over them. E.g. GLPK: https://github.com/r-opt/GLPKoptimizer


The list of supported solvers is impressive. There are some which I don't know MOI can talk too.

Are there any solvers written in R and not just written in another programming language and wrapped?


I just added a list of the supported solvers to the README. Around 20.


I am not aware of any.


Originally posted in Jan 2020 [0], albeit without much traction then.

[0] https://news.ycombinator.com/item?id=22025704


I recently restarted development, so I thought I'd repost and see if anyone from HN is interested/has feedback or ideas.


Thanks for your effort. You've already covered some ground in the examples, at this stage you're trying to fire up network effects and attract users/analysts/developers.

Treating this like an optimization problem, it seems that there's more potential for attracting users if you broaden the scope of vignettes and let Duckduckgo/Bing/Google do its magic for queries like 'optimum scheduling in R' or 'cutting stock in R' etc etc.


Thanks!




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: