changing expression to function and then optimize the function

for the follwing expresson:

expr = expr + w(r2) * (sqrt((x(r1)-a(i))^2 + (y(r1)-b(i))^2)) - w(r1) * (sqrt((x(r2)-a(i))^2 + (y(r2)-b(i))^2));

I converted it to function by

syms expr x y w 
ht = matlabFunction(expr);

Please suggest hw to minimize the above expression to find the values of multiple values of x,y and w.

I have data for the variable a,b,r1 and r2

Answer 1

[Matt J]

Do not use syms or matlabFunction for a situation like this. Just write a plain function to compute the expression,

function expr=ht(xyw, a,b,r1,r2)

   x=xyw(1);
   y=xyw(2);
   w=xyw(3)

   expr=...

end

and then optimize it with, for example,

x0=... %initial guesses
y0=...
w0=...

xyw=fminunc(@(xyw)  ht(xyw, a,b,r1,r2)   ,  [x0,y0,w0])

[Akhil]

is the following way is correct:

f = 0;

for i = 1:numel(a)

r1 = reg1(i);

r2 = reg2(i);

f = f + w(r2) * sqrt((x(r1)-a(i))^2 + (y(r1)-b(i))^2) - w(r1) * sqrt((x(r2)-a(i))^2 + (y(r2)-b(i))^2);

end

x0=0;

y0=0;

w0=10;

function f=ht(x,y,w)

x=zeros(1,26);

y=zeros(1,26);

w=zeros(1,26);

end

and kindly suggest how to perform the minimization

[Matt J]

Now that I see the whole loop, I don't quite understand what x(r1) refers to. Please give the sizes/dimensions of all the variables. In particular, aren't x,y,w scalar unknowns? If so, what does it mean to index x with r1?