- Perform derivative-free optimization algorithms in R using C++.
- A wrapper interface is provided to call
Cfunction of the 'bobyqa' implementation (See https://github.com/emmt/Algorithms/tree/master/bobyqa).
Get the development version from github:
install.packages("devtools")
library(devtools)
devtools::install_github("ypan1988/rminqa", dependencies=TRUE)Or the released version from CRAN:
install.packages("rminqa")An example of using minqa::bobyqa() in R environment:
fr <- function(x) { ## Rosenbrock Banana function
x1 <- x[1]
x2 <- x[2]
100 * (x2 - x1 * x1)^2 + (1 - x1)^2
}
(x1 <- minqa::bobyqa(c(1, 2), fr, lower = c(0, 0), upper = c(4, 4)))
## => optimum at c(1, 1) with fval = 0
str(x1) # see that the error code and msg are returnedCorresponding code written in C++ using package rminqa (file demo.cpp):
#include <cmath> // std::pow
#include "RcppArmadillo.h"
// [[Rcpp::depends(RcppArmadillo)]]
#include "rbobyqa.h"
// [[Rcpp::depends(rminqa)]]
using namespace rminqa;
class Rosen : public Functor {
public:
double operator()(const arma::vec &x) override {
double x1 = x(0);
double x2 = x(1);
return 100 * std::pow((x2 - x1 * x1), 2) + std::pow(1 - x1, 2);
}
};
// [[Rcpp::export]]
void bobyqa_rosen() {
Rosen rb;
Rbobyqa<Rosen> opt;
arma::vec x = {1, 2};
opt.minimize(rb, x);
Rcpp::Rcout << "-------------------------" << std::endl;
Rcpp::Rcout << "par = \n" << opt.par() << std::endl;
Rcpp::Rcout << "fval = " << opt.fval() << std::endl;
Rcpp::Rcout << "feval = " << opt.feval() << std::endl;
Rcpp::Rcout << "msg = " << opt.msg() << std::endl;
Rcpp::Rcout << "-------------------------" << std::endl;
}Compile and run the function in R:
library(Rcpp)
sourceCpp("~/demo.cpp") # you may need to change the directory
bobyqa_rosen()Then you will get expected output as follows:
-------------------------
par =
1.0000
1.0000
fval = 9.98796e-15
feval = 341
msg = Normal exit from bobyqa
-------------------------
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent