MATLAB: Curve fit for multiple variables

curve fitmultiple variable inputs

I have an equation of form y=a*(x1^b)*(x2^c)*(x3^d)*(x4^e)

where x1, x2, x3, x4 are variables and y is output. I have all these values experimentally.

I need to fit this equation to calculate optimum values of a, b, c, d and e.

Please provide a suitable matlab code.

x = [x1(:) x2(:) x3(:) x4(:)];
abcde = [ones(numel(y),1),log(x)]\log(y);
abcde(1) = exp(abcde(1));

your function

yourfun = @(x,a)a(1)*prod(bsxfun(@power,x,reshape(a(2:end),1,[])),2);

Please read about use '\'

Perhaps I'm misunderstanding but if

 x = [1.0000         0
      0.5000    1.0000
      0.5000    1.5000
           0   -1.0000
     -1.0000    1.0000
      0.5000   -1.0000];

which is a 6x2 array, then wouldn't

 x*x'

be the desired 6x6 array?

%Suppose
x1=x
x2=x'
x3=y
%we get the system
dx1=x2
dx2=x2/a*(f/c - b*(1-x2*x3/d)/x3)  % You have to check your brackets 
dx3=1-x2*x3/d