'ROI' Optimization Problems Based on 'globalOptTests'

A collection of non-linear optimization problems with box bounds transformed into 'ROI' optimization problems. This package provides a wrapper around the 'globalOptTests' which provides a collection of global optimization problems. More information can be found in the 'README' file.

ROI.models.globalOptTests

globalOptTests

library(ROI)
library(ROI.models.globalOptTests)

globalOptTests is a R package which contains several non-linear programming problems with box constraints. More information about globalOptTests can be found in Mullen (2014).

Using ROI.models.globalOptTests

globopt()

lists all the names of the globalOptTests instances.

globopt(x = "all")

gives all the globalOptTests instances converted into a ROI optimization problem.

globopt(x = "metainfo")

gives the available meta information.

globopt(x = "Rastrigin")

returns the Rastrigin optimization problem as an ROI optimization problem.

z <- ROI_solve( globopt(x = "Rastrigin"), start =  double(10))
abs(solution(z, "objval") - globopt("metainfo")["Rastrigin", "optimum"])

References

• Mullen, K. (2014). Continuous Global Optimization in R. Journal of Statistical Software, 60(6), 1 - 45. doi:http://dx.doi.org/10.18637/jss.v060.i06