Single variable optimization

January 1, 2011
By

(This article was first published on YGC » R, and kindly contributed to R-bloggers)

Optimization means to seek minima or maxima of a funtion within a given defined domain.

If a function reach its maxima or minima, the derivative at that point is approaching to 0. If we apply Newton-Raphson method for root finding to f’, we can get the optimizing f.

?View Code RSPLUS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
f2df <- function(fun) {
	fun.list = as.list(fun)
	var <- names(fun.list[1])
	fun.exp = fun.list[[2]] 
	diff.fun = D(fun.exp, var) 
	df = list(x=0, diff.fun) 
	df = as.function(df) 
	return(df)
}
 
newton <- function(fun, x0, tol=1e-7, niter=100) { 
	df = f2df(fun) 
	for (i in 1:niter) { 
		x = x0 - fun(x0)/df(x0) 
		if (abs(fun(x)) < tol) 
			return(x) 
		x0 = x      
	} 
	stop("exceeded allowed number of iterations") 
}
 
newton_optimize <- function(fun, x0, tol=1e-7, niter=100) {
	df <- f2df(fun)
	x = newton(df, x0, tol, niter)
	ddf <- f2df(df)
	if (ddf(x) > 0) {
		cat ("minima:\t", x, "\n")
	} else {
		cat ("maxima:\t", x, "\n")
	}
	return(x)
}


The golden-section method does not need f’. And it is similar to the root-bracketing technique for root finding.

?View Code RSPLUS
1
2
3
4
5
6
7
8
9
10
11
12
gSection <- function(f, x1, x2, x3, tol=1e-7) {
	r <- 2 - (1+sqrt(5))/2 
	x4 <- x2 + (x3-x2)*r
	if ( abs(x3-x1) < tol ){
		return(x2)
	}
	if (f(x4) < f(x2)) {
		gSection(f, x2, x4, x3, tol)
	} else {
		gSection(f, x4, x2, x1, tol)
	}
}
> f <- function(x) (x-1/3)^2
> newton_optimize(f, 0, tol=1e-7)
minima:  0.3333333
[1] 0.3333333
> gSection(f, 0,0.5,1)
[1] 0.3333333
> optimize(f, c(0,1), tol=1e-7)
$minimum
[1] 0.3333333

$objective
[1] 0

Related Posts

To leave a comment for the author, please follow the link and comment on their blog: YGC » R.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Tags: , , ,

Comments are closed.

Sponsors

Mango solutions



RStudio homepage



Zero Inflated Models and Generalized Linear Mixed Models with R

Quantide: statistical consulting and training



http://www.eoda.de







ODSC

ODSC

CRC R books series











Contact us if you wish to help support R-bloggers, and place your banner here.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)