Reduce the numbers of inputs in a numerical function

Hi all, if I have a numerical function f(x,y,z), and given that in my program most of the time x and y are fixed, I would like to evaluate f for some values x_0 and y_0, retaining the z dependence. At the end I would like to have a numerical funcion g(z)=f(x_0,y_0,z).

Is this possible?

Thanks!

Answer 1

[Andrei Bobrov]

eg:

f = @(x,y,z)x+y+z
x_0 = 3; y_0 = 6;
g = @(z)f(x_0 ,y_0 ,z)

EDIT

>> g(5),g(9),g(13)
ans =
    12
ans =
    16
ans =
    20
>>

ADD use SymbolicToolbox

syms x y z
f = x*y*z
f1 = subs(f,{x,y},{3,4})
g = matlabFunction(f1)

>> g(5),g(9),g(13)
ans =
    60
ans =
   108
ans =
   156
>>

[Antonio]

I'm sorry, I have not specified that I do not want to call f each time!

I think what i want could be done with symbolic function. e.g.

syms z

g=f(x_0,y_0,z)

then call g as a function of z. But I want to know if one can do this without the use of symbolic functions.

The point is that the function f is very demanding, so I want to avoid calling it every time.

[Walter Roberson]

I would wrap a simplify() around the subs(). That gives the opportunity for MuPAD to recognize any special cases and to resolve values that have effectively become constants (e.g., for cos(0) emit 1).

Answer 2

[Sargondjani]

use passing of parameters: function value=my_fun(x,y,z); end

then you make an anonymous function: x0=... y0=... my_fun_z=@(z)my_fun(x0,y0,z);

now you can evaluate by just giving the value for z. (note: if you change the value of x0 or y0 you have to do the "my_fun_z=... " again)

       however if some calculations inside your function stay the same (independent of z) and these are very demanding then you should split the function in a part with and without z to save calculation time

[Antonio]

Yeah, the point is exactly this one, that I cannot separate the two parts into parts with and without z.

And your solution again calls the whole function my_fun again.

Well, I guess this cannot be done

[Antonio]

Of course I could separate them, but this is so complicated that cannot be done in practice.

[Sargondjani]

if you can not or do not want to separate the parts, then it is impossible indeed... it might still be worth the effort to try.... sometimes things look more complicated than they are