Finding maximum to a symbolic array inside an m-file

Hi everyone,

I'm using MATLAB to graphically display a model with three variables, L1, L2 and Psi. I have created the following function:

function d = dafstand(L1,L2)
%do a lot of stuff, with the result being a 1- or 2-dimensional complex vector, called x, containing a symbolic Psi

Now, for this variable x, I want to elementwise maximize x given a Psi between 0 and pi/2 (for example, using fminbnd with -x). When such a Psi is given, the complex part of the vector results to 0, and I would like to drop that using something like real(). Finally, I want to get the maximum element in x, such that the last line of the m-file would be:

d = max(x)

So that I can call from the command window:

ezsurf(@(L1,L2)(dafstand(L1,L2)),[0,1,0,1]

Can someone help me?

Answer 1

[Walter Roberson]

If the symbolic formula are sufficiently complex, then fminbnd would not always be able to find the correct maximum, especially if there are discontinuities.

There is a possibility that you might find it more productive to solve() or fsolve() for the imaginary part being 0, since that is what you care about anyhow.

[Sander Hermanussen]

Actually, the answer never has a complex result, no matter what the input is. However, due to rounding errors, it always gives a +0.000i or 0.000i part.

Also, the answer is continous at least on the given interval of (0,pi/2).

I want to do something like the following:

x = matlabFunction(x);

[Psi,x]=fminbnd(@(Psi)(real(x(Psi))),0,pi/2);

This gives the following error:

??? Operands to the || and && operators must be convertible to

logical scalar values.

Error in ==> fminbnd at 315

if ( (fu <= fw) || (w == xf) )

Error in ==> dafstand at 27

[Psi,x]=fminbnd(@(Psi)(real(x(Psi))),0,pi/2)