The following error occurred converting from sym to double: Unable to convert expression into double array.

Hi!

I am testing a function in MATLAB: ode45.

I created some functions in a class to run this testing, because I am development a software in App Designer and I need to realize this test in this organization.

Well, when I run this code, an error appears about converting from sym to double in my parameter P. I put the command double() or vpa(), but the error always apperars.

Can you help me, please? Thanks!

Next, the code:

classdef Teste < handle
    methods (Static)
        function parameters = eq()
            syms t
            P = t - exp(t);
            K1 = 2;
            K3 = 3;
            parameters = [P, K1, K3];
        end

        function dydt = eq1(y)
            parameters = Teste.eq();
            P = parameters(1);
            K1 = parameters(2);
            K3 = parameters(3);

            dydt = zeros(2,1);
            dydt(1) = y(2);
            dydt(2) = P - K1 * y(1) - K3*(y(1)^3);
        end
        
        function graph
            
            tspan = [0 5];
            y0 = [0 0.02];
            [t,y] = ode45(@(t,y) Teste.eq1(y), tspan, y0);

            plot(t,y(:,1),'-o',t,y(:,2),'-.')
        end
    end
end

Answer 1

[Walter Roberson]

zeroes creates a numeric array. Your parameters are constructed using symbolic t so the second assignment is trying to assign a symbolic expression with unresolved variable into a numeric slot in the array.

[Ana Reis]

Yeah, I understand. But, how can I resolve this? Because I put a syms array and the function ode45 not receive a symbolic parameter.

[Walter Roberson]

            [t,y] = ode45(@(t,y) Teste.eq1(y), tspan, y0);

Make that

            [t,y] = ode45(Teste.eq1, tspan, y0);

And make eq1 return a function handle that takes two parameters. It can construct the function handle by using matlabFunction with the 'vars' parameter.