Scribd
Upload
3K views2 pages

Coordinate Descent and Golden Selection Search

The document describes a golden selection search algorithm and coordinate descent algorithm. It defines functions for golden selection search and coordinate descent. It then applies the coordinate descent algorithm to minimize a 2D function and plots the results.

Uploaded by

api-285777244
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3K views2 pages

Coordinate Descent and Golden Selection Search

The document describes a golden selection search algorithm and coordinate descent algorithm. It defines functions for golden selection search and coordinate descent. It then applies the coordinate descent algorithm to minimize a 2D function and plots the results.

Uploaded by

api-285777244
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Coordinate Descent and Golden Selection Search

YIK LUN, KEI


golden = function(f, lower.bound, upper.bound, tolerance = 1e-5)
{
golden.ratio = 2/(sqrt(5) + 1)
x1 = lower.bound + golden.ratio*(upper.bound - lower.bound)
x2 = upper.bound - golden.ratio*(upper.bound - lower.bound)
f1 = f(x1)
f2 = f(x2)
while (abs(upper.bound - lower.bound) > tolerance)
{
if (f2 > f1){
lower.bound = x2
x2 = x1
f2 = f1
x1 = lower.bound + golden.ratio*(upper.bound - lower.bound)
f1 = f(x1)
}
else{
upper.bound = x1
x1 = x2
f1 = f2
x2 = upper.bound - golden.ratio*(upper.bound - lower.bound)
f2 = f(x2)
}
}
(lower.bound + upper.bound)/2
}
g<-function(x,y){ 5*x^2 - 6*x*y + 5*y^2}
x<-seq(-1.5,1.5,len=100)
y<-seq(-1.5,1.5,len=100)
z<-outer(x,y,g)
CD<-function(startx,starty,maxx,maxy,iter,tolerance = 1e-5){
new<- matrix(NA,ncol=2,nrow=(iter+1))
new[1,1]<-startx; new[1,2]<-starty
for(i in 2:(iter+1)){
a<-function(x){5*x^2 - 6*x*new[i-1,2] + 5*new[i-1,2]^2}
new[i,1] <- golden(a,new[i-1,1],maxx)
b<-function(y){5*new[i,1]^2 - 6*new[i,1]*y + 5*y^2}
new[i,2]<- golden(b,new[i-1,2],maxy)
if(abs(new[i,1]-new[i-1,1]) < tolerance && abs(new[i,2]-new[i-1,2]) < tolerance){break}
}
return(new)
}

cd<-CD(startx=-1.5,starty=-1.5,maxx=1.5,maxy=1.5,iter=15)

1.5

contour(x,y,z,levels = seq(0.5,5,by=.9))
newx<-sort(cbind(cd[,1],cd[,1]),decreasing=F)[-1]
newy<-sort(cbind(cd[,2],cd[,2]),decreasing=F)[-length(newx)]
cd<-cbind(newx,newy)
points(cd[,1],cd[,2],type="l",col="red")

1.0

4.1

0.5

2.3

0.5

0.5

1.4
3.2

1.5

1.5

1.0

0.5

0.0

0.5

1.0

1.5

You might also like