how can you pass values to variables in an an array of functions?

I am trying to make a movie of the solutions to the heat equation for various values of a parameter in the coefficients . I want to create an array for the fourier coefficients of the solution. How can I define an 'array function' ? does this idea exist.

apologies if theis is a duplicate question. couldn't find this answer anywhere.

[Jan]

What does "array function" mean? What have you tried already and which explicit problems occur?

[dpb]

I was puzzled, too, Jan. I can't envision why there would be a problem in doing what OP asks about with just "ordinary" arrays which are passed to functions by simply placing the array name in the argument list...

[Leonard]

I could do this another way.... but i think its an interesting lesson to learn how to do..... and it is prettier i think..... lets see if I can produce the idea here.

a;
x = [a+2, a+3]

for i = 1:10
      a = i
      print x
end

maybe I am missing something simple. I don't code matlab or any other language. I have found that the only way an array or vector will hold a variable is if I define them syms (symbollicaly) ...

but it seems once i define a variable as syms, i cant give it any value.

by function array or vector I mean.... like a function f(x) = x^2 can receive a value for x is there a way to define an array like array(x) = [x^2, x^3] such that if I type array(2) it outputs [4,8]?

obviously x indexes the array, but is there a way to perform this idea?

[David Young]

Yes, I think you're right that you're missing the simple basis of arrays, functions and variables. A run through the initial tutorials will help a lot. For example you can write

x = 2;
x.^[2 3]

to compute your example. I suspect that using the symbolic toolbox is highly confusing at this stage - it's not needed for numerical computations.

[Leonard]

what if the function which defined each cell was more complicated than an exponent?

so basically passing a value to a parameter (variable) in multiple cells in an array is something everyone should avoid?

[Jan]

@Leonard: If you have a more complicated function, you create a function file, which can be as complicated as you want.

Answer 1

[per isakson]

Playing with anonymous functions in combination with cellfun

    fh{1}=@(x) x.*x;
    fh{2}=@(x) x.*x.*x;
    fh{3}=@(x) x.*x.*x.*x;
    fh{4}=@(x) sin(x);
    fh{5}=@(x) cos(x);
    foo = @(z) cellfun( @(f,x) f(x), fh, repmat({z},size(fh)), 'uni', false);
    foo(2)

returns

    ans = 
        [4]    [8]    [16]    [0.9093]    [-0.4161]

However, those are cell arrays. Matlab doesn't support ordinary arrays of anonymous functions.

.

Or a little better

    foo = @(z) cellfun( @(f,x) f(x), fh, repmat({z},size(fh)), 'uni', true);
    foo(2)

returns

    ans =
        4.0000    8.0000   16.0000    0.9093   -0.4161

which is an array of doubles.

.

I guess that what you want is a Functional programming languages

[Leonard]

Thanks. I appreciate the guidance. This teaches me a little something. My design was unnecessary for sure. So, what I was describing to do would be something like this:

for n = 1:10
    fh{n}=@x x.*x;
end

foo = @z cellfun(@(f,x) f(x), fh, repmat({z},size(fh)), 'uni', true);

for i = 1:5
    for p = 0:size(foo)
coeffecientsarray(p)=foo(i)
    end
end

But i don't think that makes sense at all to do. I should just make a single 1D array where every next n entries has the parameter take on the next new value.

perhaps this thread should be deleted so no one gets confused by my nonsense haha any votes on that idea?

[per isakson]

An array with ten identical functions might not be useful.

However, anonymous functions are powerful and it takes some experiments to learn how to use them. "[...] any votes on that idea?" I vote for keeping the post. I contributed an experiment.

[David Young]

Yes, building the 10 anonymous functions is being overcomplicated. Almost certainly you can do what you want just by passing vector arguments to ordinary functions. I'd say keep the post - there are many less interesting and less useful threads here.