Undefined operator '==' for input arguments of type 'struct'. Error in decic (line 41) free_yp = find(fixed_yp0 == 0); Error in IMWFGD (line 111) [y0, yp0] = decic(f, t0, y0est, yp0est, [], opt) >>

Hi all, I am trying to run a script to solve 10 DAEs, but I get an error message " Undefined operator '==' for input arguments of type 'struct'.

Error in decic (line 41) free_yp = find(fixed_yp0 == 0);

Error in IMWFGD (line 111) [y0, yp0] = decic(f, t0, y0est, yp0est, [], opt)"

Please see the attachment for the script.

[Stephen23]

@Dursman Mchabe: the error message states that the error occurs on line 41 of the function decic, which you have not given us. Without this function we cannot give much advice apart from pointing out what is already written in the error message: clearly fixed_yp0 is a structure, but eq is not defined for structures and so there is a bug in the code.

[Walter Roberson]

decic is a Mathworks supplied function related to ODE.

y0est and yp0est need to be vectors of zeros and ones, but at least one of the two has been passed as a struct.

[Dursman Mchabe]

Thanks a lot, Stephen and Walter. decic is indeed a a function supplied by Mathworks. I don't know where is the bug in the code.

[Walter Roberson]

In your IMWFGD.m what are class(y0est) and class(yp0est)

[Dursman Mchabe]

Double

[Dursman Mchabe]

Is it a good idea to change the operator "==" to "isequalto"?

[Walter Roberson]

"Is it a good idea to change the operator "==" to "isequalto"?"

No.

Answer 1

[Walter Roberson]

The sixth input to decic needs to be fixed_yp0 . You have passed opt there, and your opt is a structure.

It looks to me as if you omitted the fixed_y0 argument.

[Dursman Mchabe]

Over the past hours, I have tried different ways to pass both fixed_y0 and fixed_yp0, but I did not succeed. For example,

"[y0, yp0] = decic(f, t0, y0est,[], yp0est, [], opt)"

gives the error message

"Not enough input arguments.

Error in symengine>@(t,in2,in3,param1,param2,param3...]

Error in decic (line 66) res = feval(odefun,t0,y0,yp0,varargin{:});

Error in IMWFGD (line 110) [y0, yp0] = decic(f, t0, y0est,[], yp0est, [], opt)"

I have also tried

"[y0, yp0] = decic(f, t0, y0est,0, yp0est, 0, opt)"

"Error using decic (line 44) Cannot specify more than 10 components of y0 and yp0.

Error in IMWFGD (line 110) [y0, yp0] = decic(f, t0, y0est,0, yp0est, 0, opt)"

I also tried to omit the opt, but I got error message that says

"Not enough input arguments. Error in symengine>@(t,in2,in3,param1,param2,param3... Error in decic (line 66) res = feval(odefun,t0,y0,yp0,varargin{:});

Error in IMWFGD (line 110) [y0, yp0] = decic(f, t0, y0est,[], yp0est, [])

Am I passing fixed_y0 and fixed_yp0 correctly?

[Guillaume]

The issue is not with the way you pass arguments to decic. It's with your ode function, f. It is the one that throws the error:

Not enough input arguments.
Error in symengine>@(t,in2,in3,param1,param2,param3...

Please, show us how f is defined.

[Dursman Mchabe]

Thanks a lot for your response Guillaume, I followed steps 1 to 6 below:

    %1. Define symbols
    syms a(t) b(t) c(t) d(t) e(t) f(t) g(t) h(t) i(t) j(t) k(t)...
        HSO2 kga DCa2 DSO2 kLa T KSPCaSO3 BETCaSO3 ktot BETCaCO3 MWCaCO3 Kad...
        kCO2 KCO2 HCO2 KSO2 KHSO3 KHCO3 P

    %2. Define equations
    eqs = [diff(a(t),t) + diff(b(t),t) + diff(c(t),t) == (P-HSO2...
            *a(t))/(1/kga)+(HSO2/(1+(DCa2 *g(t))/(DSO2*(a(t) + b(t) + c(t))))...
            *kLa) - 0.162*exp(-5153/T)*(((g(t)*c(t))/KSPCaSO3) - 1)^3 ...
            * (BETCaSO3/((g(t)*c(t))/KSPCaSO3)),diff(d(t),t) + diff(e(t),t)...
            + diff(f(t),t)+ diff(g(t),t) == 0.162*exp(-5153/T)*(((g(t)*c(t))...
            /KSPCaSO3) - 1)^3 * (BETCaSO3/((g(t)*c(t))/KSPCaSO3)),diff(h(t),t)...
            + 2*diff(g(t),t) - diff(b(t),t)- 2* diff(c(t),t) - diff(e(t),t)...
            - 2* diff(f(t),t) - diff(i(t),t) == 0,diff(d(t),t) + diff(e(t),t)...
            + diff(f(t),t) + diff(a(t),t) + diff(b(t),t) + diff(c(t),t) == ...
            -ktot*BETCaCO3*MWCaCO3*j(t)*h(t) *(1-(Kad*h(t))/(1+Kad*h(t)))-...
            0.162*exp(-5153/T)*(((g(t)*c(t))/KSPCaSO3) - 1)^3 ...
            *(BETCaSO3/((g(t)*c(t))/KSPCaSO3)),diff(d(t),t) == (kCO2/KCO2)...
            *e(t) - kCO2*HCO2,h(t)*diff(b(t),t)...
            + b(t)*diff(h(t),t) - KSO2*diff(a(t),t) == 0,h(t)*diff(c(t),t)...
            + c(t)*diff(h(t),t) - KHSO3*diff(b(t),t) == 0,h(t)*diff(e(t),t)...
            + e(t)*diff(h(t),t) - KCO2*diff(d(t),t) == 0,h(t)*diff(f(t),t)...
            + f(t)*diff(h(t),t) - KHCO3*diff(e(t),t) == 0,h(t)*diff(i(t),t)...
            + i(t)*diff(h(t),t)==0];

     %3. Define variables
     vars = [a(t); b(t); c(t); d(t); e(t); f(t); g(t); h(t); i(t); j(t)];

     %4. Rewrite this DAE system to a system of first-order differential algebraic equations.
     [eqs, vars, newVars] = reduceDifferentialOrder(eqs, vars)

     %5. Convert a DAE system to a MATLAB function handle, using daeFunction
     F = daeFunction(eqs, vars, [HSO2, kga, DCa2, DSO2, kLa, T, KSPCaSO3, BETCaSO3, ktot, BETCaCO3, MWCaCO3, Kad,...
        kCO2, KCO2, HCO2, KSO2, KHSO3, KHCO3, P])

     %6. Assign numerical values to the symbolic parameters: HSO2 = 149;
       %kga = 4.14e-6;
       %DCa2 = 1.39e-9;
       %DSO2 = 2.89e-9;
       %kLa = 8.4e-4;
       %T = 323.15;
       %KSPCaSO3 = 1.07e-7;
       %BETCaSO3 = 10;
       %ktot = 8.825e-3;
       %BETCaCO3 = 12.54;
       %MWCaCO3 = 100.0869;
       %Kad = 0.84;
       %kCO2 = 0.1242;
       %KCO2 = 1.7e-3;
       %HCO2 = 5.15e+3;
       %KSO2 = 6.24;
       %KHSO3 = 5.68e-5;
       %KHCO3 = 6.55e-8;
       %P = 47.28325;
     f = @(t, y, yp)  F(t, y, yp, [149,4.14e-6,1.39e-9,2.89e-9,8.4e-4,323.15,1.07e-7,10,8.825e-3,12.54,100.0869,0.84,0.1242,1.7e-3,5.15e+3,6.24,5.68e-5,6.55e-8,1])

[Dursman Mchabe]

I have also attached an m-file with all the steps clearly numbered.

[Guillaume]

I don't have the symbolic toolbox so can't do any testing and I don't know how daeFunction works.

Something called by f is not happy about the number of arguments it receives. So first checked, that you've passed the right number of arguments to F. Possibly, use the 'File' option of daeFunction to generate an m file instead. It may aid in debugging as it will error on a particular line of the file rather than in an anonymous function.

[Dursman Mchabe]

Thanks a lot for your efforts Guillaume.