The problem with an error when using "fzero" when solving a nonlinear task

Good afternoon! I can't use the code I have. At the beginning I set the parameters I need, an error occurs during the calculation. What could this be related to? In addition, I have a comment near the function f0 (variable appears to change size on every loop iteration).

clc;
clear all;
% Resonsnt frequency of the Gunn diode
warning off MATLAB:fzero:UndeterminedSyntax
disp('Calculation of DG resonance frequency')
global Rdl
VarName = input('Name of a variable parameter, [R0, Rc, Lc, Cc, C0, Cd]: ', 's' );
if(VarName == 'R0')
    Xaxis=[VarName,', Ohm'];
elseif (VarName=='Rc')
    Xaxis=[VarName,', Ohm'];
elseif (VarName =='Lc')
    Xaxis=[VarName,', nH'];
elseif (VarName=='Cc')
    Xaxis=[VarName,', PF'];
elseif (VarName=='C0')
    Xaxis=[VarName,', PF'];
elseif (VarName=='Cd')
    Xaxis=[VarName,', PF'];
else
    disp('Error in VarName')
end
ParName=input('Name of parametric variable, [R0, Rc, Lc, Cc, C0, Cd]: ','s');
%St = stremp(Varname, Parname)
while (VarName==ParName)
    ParName=input('Input another name of PV','s');
end
LimVar=input('Limits of variable and step [MinVal, [Step], MaxVal] = ');
ParVal=input('Input values of parametric variable [, , ..] = ');
N=101; 
Nv=length(LimVar); 
Vmin=LimVar(1) 
if Nv == 3 
    N=LimVar(2); г
    Vmax=LimVar(3) 
else
    Vmax=LimVar(2) 
end
dVar=(Vmax-Vmin)/(N-1)
VarV=Vmin:dVar:Vmax;
Npar=length(ParVal); 

for n=1:Npar 
    for m=1:N 
        options=optimset('Display','off','TolX',1e-5); 
        [x,fval, exitflag, output]=fzero(@ResFrqNew,35,options,ParVal(n),VarV(m),VarName,ParName); 
        
        f0(m,n)=x;
        Rdom(m,n)=Rdl;
        %Nit(m) = output.iterations
    end
    subplot(1,2,1)
    plot (VarV, f0(:, n), '-k', 'LineWidth', 2)
    text(VarV(5), f0(5, n), int2str(n), 'FontSize', 12)
    xlabel(Xaxis)
    ylabel('Fres, GHz')
    hold on
    subplot(1,2,2)
    plot(VarV, Rdom(:, n) , '-k', 'LineWidth', 2)
    text(VarV(5), Rdom(5, n) , int2str(n), 'FontSize', 12)
    xlabel(Xaxis)
    ylabel('Rd, Ohm')
    hold on
end
function GdNum = DomRes(Rd, f, PV, W, VarName, ParName)
Lcl = 0.1e-9;
R0 = 8;
C0 = 0.6e-12;
Cd = 0.27e-12;
Rc = 0.5;
if(ParName=='R0')
    R0 = PV;
elseif (ParName == 'Rc')
    Rc = PV;
elseif (ParName == 'Lc')
    Lcl = PV*1e-9;
elseif (ParName == 'Cc')
    Cc = PV * 1e-12;
elseif (ParName == 'C0')
    C0 = W * 1e-12;
elseif (ParName == 'Cd')
    Cd = W*1e-12;
else
    disp('Error in defining Parameter name')
end
if(VarName=='R0')
    R0 = W;
elseif (VarName=='Rc')
    Rc = W;
elseif (VarName=='Lc')
    Lcl = W * 1e-9;
elseif (VarName=='Cc')
    Cc = W * 1e-12;
elseif (VarName == 'C0')
    C0 = W * 1e-12;
elseif (VarName == 'Cd')
    Cd = W * 1e-12;
end
Om = 2e9*pi*f;
tau0 = C0*R0;
tauD = Rd*Cd;
F1 = R0/(1 + (Om * tau0)^2);
F2 = Rd/(1 + (Om * tauD)^2);
GdNum = Rc + F1 + F2;
end
function Bd = ResFrq(f, PV, W , VarName, ParName)
% Function calculates Im(Yd) for a given frequency f and parameters Lcl,R0
global Rdl
Rc = 0.5; 
Cc = 0.5e-12; 
Lc = 0.1e-9;
R0 = 8;s
C0 = 0.6e-12;
Cd = 0.27e-12; %емкость домена
if(ParName == 'R0')
    R0=PV;
elseif(ParName=='Rc');
    Rc=W;
elseif(ParName=='Lc')
    Lcl=PV*1e-9;
elseif (ParName == 'Cc')
    Cc = PV * le-12;
elseif (ParName == 'C0')
    C0 = W * le-12;
elseif (ParName == 'Cd')
    Cd = W * le-12;
else
    disp('Error in ParName')
end
if(VarName == 'RO')
    R0 = W;
elseif (VarName == 'Rc')
    Rc = W ;
elseif (VarName == 'Lc')
    Lcl = W * le-9;
elseif (VarName == 'Cc')
    Cc = W * le-12;
elseif (VarName == 'CO')
    C0 = W * le-12;
elseif (VarName == 'Cd')
    Cd = W * le-12;
else
    disp('Error in VarName')
end
Om = 2e9*pi*f;
tau0 = C0 * R0;
optionsR = optimset ('Display', 'off', 'TolX', le-5) ;
[x, fval, exitflag, outp] = fzero(@DomResNew, -10, optionsR, f, PV, W ,VarName, ParName);
Rd = x;
Rdl = Rd;
% Niter = outp.iterations
tauD = Cd * Rd;
F1 = tau0 * R0 /(1 + (Om*tau0)^2);
F2 = tau0 * Rd /(1 + (Om * tauD)^2);
Bd = Cc - (Lcl - F1 - F2) / ((RC + F1 + F1)^2 + Om^2 * (Lcl - F1 - F2)^2);
end

[Walter Roberson]

Do not use == to compare character vectors -- use strcmp() or strcmpi()

Or use == to compare against a string such as VarName == "Cc" instead of VarName == 'Cc'

[Dyuman Joshi]

You could also use switch - switch, case, otherwise

[Torsten]

Please include executable code in which the error is reproducable.

[Albert]

(I attach a file to the message)

Calculation of DG resonance frequency

Name of a variable parameter, [R0, Rc, Lc, Cc, C0, Cd]: R0

Name of parametric variable, [R0, Rc, Lc, Cc, C0, Cd]: Rc

Limits of variable and step [MinVal, [Step], MaxVal] = [1:1:102]

Input values of parametric variable [, , ..] = [1,2]

Error using fzero>localFirstFcnEval

FZERO cannot continue because user-supplied function_handle ==> ResFrqNew failed with the error below.

Undefined function 'ResFrqNew' for input arguments of type 'double'.

Error in fzero (line 305)

fx = localFirstFcnEval(FunFcn,FunFcnIn,x,varargin{:});

Error in hop (line 45)

[x,fval, exitflag, output]=fzero(@ResFrqNew,35,options,ParVal(n),VarV(m),VarName,ParName);

[Torsten]

The attached code cannot be simply executed. We have to reproduce the error.

And I don't see a function with name "ResFrqNew" that "fzero" could call.

Answer 1

[Walter Roberson]

        [x,fval, exitflag, output]=fzero(@ResFrqNew,35,options,ParVal(n),VarV(m),VarName,ParName); 

You ask for ResFrqNew to be invoked

        [x, fval, exitflag, outp] = fzero(@DomResNew, -10, optionsR, f, PV, W ,VarName, ParName);

You ask for DomResNew to be invoked

        function GdNum = DomRes(Rd, f, PV, W, VarName, ParName)

You define function DomRes not DomResNew

        function Bd = ResFrq(f, PV, W , VarName, ParName)

You define function ResFrq, not ResFrqNew

So you are missing definitions of two functions, and you define two functions that you are not using.

[Albert]

Thank you!

[Albert]

but it still didn't solve the problem:(

[John D'Errico]

@Albert

Walter tried to help you. BUT, you did not actually supply sufficient information to allow a solution. Should we be able to see inside your computer, to know what those functions are?

[Albert]

I think that the code as a whole needs to be improved, I am grateful to everyone who responded to my request for help!

[Walter Roberson]

I suspect that you need

        [x,fval, exitflag, output]=fzero(@ResFrq,35,options,ParVal(n),VarV(m),VarName,ParName);

and

         [x, fval, exitflag, outp] = fzero(@DomRes, -10, optionsR, f, PV, W ,VarName, ParName);

[Albert]

Yes, I renamed these functions, but for some reason the error still remained

[Albert]

It still sends the error

Error in VarName

Error using fzero>localFirstFcnEval

FZERO cannot continue because user-supplied function_handle ==> ResFrq failed with the error below.

Not enough input arguments.

Error in fzero (line 305)

fx = localFirstFcnEval(FunFcn,FunFcnIn,x,varargin{:});

Error in main (line 52)

[x,fval, exitflag, output]=fzero(@ResFrq,35,options,ParVal(n),VarV(m),VarName,ParName);

[Walter Roberson]

See http://www.mathworks.com/help/matlab/math/parameterizing-functions.html

[Albert]

Okay, I think I've solved this problem, but I have a new one:

Unrecognized function or variable 'f0'.

Error in main (line 64)

plot (VarV, f0(:, n), '-k', 'LineWidth', 2)

[Walter Roberson]

You appear to be assigning to f0 inside of

for n=1:Npar 
    for m=1:N 

If Npar is less than 1, or N is less than 1, then the assignment to f0 would not be executed and f0 would be undefined after the loop.

You should consider pre-allocating f0 .

[Albert]

I tried using syms(), but it didn't help

syms f0(m,n);

syms Rdom(m, n);

f0(m,n)=x; %элемент [m,n] равен x

Rdom(m, n)=Rdl;

%Nit(m) = output.iterations

Unrecognized function or variable 'f0'.

Error in main (line 65)

plot (VarV, f0(:, n), '-k', 'LineWidth', 2)