Error "using symengine>@()0.0" in ODE Solver

Hello,

I restarted everything I did from scratch and in the most simpliest way possible to recreate the same error I get in my much larger code. This error occurs when I make the B-field components functions instead of constants as the E parts are currently defined in this specfiic example. However, in my much larger code this error occurs when I just change the syntax of the ode set so if this error can be resolved here, perhaps then I can figure out why it's happening in the larger code set.

Error:

Error using symengine>@()0.0
Too many input arguments.

Error in reffun (line 14)
ode1 = Ex + s(2).*Bz(s(4),s(5),s(6)) - s(3).*By(s(4),s(5),s(6));

Error in odearguments (line 90)
f0 = feval(ode,t0,y0,args{:});   % ODE15I sets args{1} to yp0.

Error in ode15s (line 150)
    odearguments(FcnHandlesUsed, solver_name, ode, tspan, y0, options, varargin);

Error in diffiqtest (line 7)
[t,S] = ode15s(@reffun,tspan,s0);

Here is the main code:

v0 = [0 0 0];
p0 = [0 0 0];
s0 = [v0 p0];
tspan = [0 5];

[t,S] = ode15s(@reffun,tspan,s0);

Function where ODEs are defined:

function refsolve = reffun(t,s)

Ex = 0;
Ey = 0;
Ez = 0;

persistent Bx By Bz
%Used so that the B-field function is only run once
if isempty(Bx)
    [Bx, By, Bz] = B_test();
end

%Reference: s(1) = vx, s(2) = vy, s(3) = vz, s(4) = x, s(5) = y, s(6) = z
ode1 = Ex + s(2).*Bz(s(4),s(5),s(6)) - s(3).*By(s(4),s(5),s(6));
ode2 = Ey + s(3).*Bx(s(4),s(5),s(6)) - s(1).*Bz(s(4),s(5),s(6));
ode3 = Ez + s(1).*By(s(4),s(5),s(6)) - s(2).*Bx(s(4),s(5),s(6));
ode4 = s(1);
ode5 = s(2);
ode6 = s(3);

refsolve = [ode1; ode2; ode3; ode4; ode5; ode6];

end

B_test function:

function [Bx, By, Bz] = B_test()

%Bfieldstrength = 0.64; %In (Teslas)
Bfieldstrength = 0;
magvol = 3.218E-6; %In (m)
mu0 = (4*pi)*10^-7;
magnetization = (Bfieldstrength*magvol)/mu0;

syms x y z
m = [0,0,magnetization];
r = [x, y, z];
B = mu0*(((dot(m,r)*r*3)/norm(r)^5) - m/norm(r)^3);
Bx = matlabFunction(B(1));
By = matlabFunction(B(2));
Bz = matlabFunction(B(3));

end

Answer 1

[Walter Roberson]

When you call matlabFunction pass

'vars', r

[Tom Keaton]

Sorry, could you clarify? Like adding r to syms?

syms x y z r

Or?

Bx = matlabFunction(B(1),'vars',r);
By = matlabFunction(B(2),'vars',r);
Bz = matlabFunction(B(3),'vars',r);

Either way I get the same error.

[Walter Roberson]

Bx = matlabFunction(B(1), 'vars', r);
By = matlabFunction(B(2), 'vars', r);
Bz = matlabFunction(B(3), 'vars', r);

Without this, the expression is examined with symvar to generate the list of parameters -- so for example if you had an expression that happened to come out as x^2+y then matlabFunction would by default create a function with parameters (x,y) but not z, and if the expression happened to come out as x^2+z then matlabFunction would by default create a function with parameters (x,z) without y. Looking through the expression to find the variables is normally the Right Thing To Do -- but it fails if you have a fixed calling sequence like you do in your case.

[Walter Roberson]

You need to

clear reffun

in order to clear the persistent results that you memorized on an earlier run.

[Tom Keaton]

I see, but I still get the exact same error when I change it to this and clear reffun.

[Walter Roberson]

I have attached the exact files I used.

[Tom Keaton]

It is odd. Your files work, but mine do no even with the same definitions. Why is this the case? Also, when I try plotting the solution, it is incorrect, because I know the solution should not look the way it does. If the ode solver is not solving the equations correctly, and it is not outputting an error, what should I do in this sitatuon?