README.md

In cwliu007/minConf: R code of the minConf_TMP used for minimization.

minConf_TMP for minimization in R

R code of the minConf_TMP (converted from minConf package in MATLAB).

How To Install in R:

library(devtools)

install_github("cwliu007/minConf")

Example:

fw <- function(x) (1-x[1])^2 + 100*(x[2]-x[1]^2)^2 # Rosenbrock function where c(1,1) is minimum

x = c(0,0)

m1 <- optim(x, fw, method = "L-BFGS-B") # for comparison purpose

m1$par

0.9998007, 0.9996013

m2 <- minConf_TMP(x, fw, method="lbfgs")

m2$par

0.9999954, 0.9999901

m3 <- minConf_TMP(x, fw, method="newton")

m3$par

0.9999458, 0.9998900

More Examples

See https://www.cs.ubc.ca/~schmidtm/Software/TMP/examples.html

Citations

M. Schmidt. minConf: projection methods for optimization with simple constraints in Matlab. http://www.cs.ubc.ca/~schmidtm/Software/minConf.html, 2008.

cwliu007/minConf documentation built on May 31, 2019, 10:01 a.m.