PDE toolbox, specifying coefficients, error telling me "Function specifying a coefficient must accept two input arguments and return one output argument."

Hi all,

I keep running into the error while I am specifying coefficients, namely c, a, and f, and the problem is I do not know why.

It would be very much appreciated if somebody kindly pointed me into the right direction.

• Below is my code:

Mypde = createpde(1);
geometryFromEdges(Mypde, dl);
mesh = generateMesh(Mypde, 'Hmax', 1e-6);

-----------------------

function amatrix = acoeffunction(region,state)
  %Given
gv=-2.7460e-6;
I_tkns=7e-8;
F=9.6485e4;
  %Now define variables
dP = @(state.u) 30*(state.u^4)-60*(state.u)^3+30*(state.u)^2;
E_chem = gv*dP;
E_over = -E_chem/F;
ED_rate = -I_tkns*E_over;
  %Build a matrix for the a coefficient.
n1=2; %length of the vector, representing coefficients 
nr = numel(region.x*region.y) %number of columns, region x and y are multiplied because it is a 2D problem.
amatrix = zeros(n1,nr); %allocate a
amatrix(1,:) = ED_rate*V_li*state.ux/state.u
amatrix(2,:) = ED_rate*V_li*state.uy/state.u

• In order for me to test whether the function is correct, I simply put a constant for all other coefficients, and the handle function for a indicated above.

specifyCoefficients(Mypde, 'm', 0, 'd', 1, 'c', 1, 'a', @acoeffunction, 'f', 1)

• The error I am getting is shown below:

Error using pde.CoefficientAssignment/checkCoefFcnHdlArgCounts (line
500)
Function specifying a coefficient must accept two
input arguments and return one output argument.

Error in pde.CoefficientAssignment/checkFcnHdlArgCounts (line 273)
            self.checkCoefFcnHdlArgCounts(self.f, systemsize, ndims);
            
Error in pde.CoefficientAssignment (line 105)
      obj.checkFcnHdlArgCounts(systemsize, numdims);

Error in pde.PDEModel/specifyCoefficients (line 139)
        coef =
        pde.CoefficientAssignment(self.EquationCoefficients,argsToPass{:});

Thank you so much in advance.

If you find any other problems in my script, please let me know.

Have a great day.

[Charlotte Rougier]

Hi, have you managed to solve this? I have the same issue. Thanks!

[ADSW121365]

Also seeking a solution to this.

[Minjun Bae]

Hi guys,

I was able to solve this problem.

When you specify a handle function for a coefficient value, if your code includes some functions or variables that are not recognized by the system, this error comes up.

Double check if your code has some contents undefined.

For example,

function fmatrix = fcoeffunction(location,state)

fmatrix=zeros(3,nr)

fmatrix(1,:)=0;

fmatrix(2,:)=0;

fmatrix(3,:)=0;

This also will cause the same error because 'nr' has not been defined in the function handle.

Therefore, nr = numel(location.x) has to be defined before you call nr for defininf fmatrix.

[MEHDI Heydari]

Hello Minjun,

I did the same steps as you state, but still getting the same error.

I really appriciate if you can comment on my code below.

%c coefficient Matrix
function cMatrix=cCoeff(location, state)
s=0.000610655417342708;
n1=6;
n2=numel(location.x);
cMatrix=zerose(n1,n2);
cMatrix(1,:)= s^2*45337.*exp(-3887./state.u(2,:))/3600;
cMatrix(3,:)=cMatrix(1,:);
cMatrix(4,:)=0.0004*(state.u(1,:)./(1+state.u(1,:))).^2-0.0066*...
    (state.u(1,:)./(1+state.u(1,:)))+0.2224;
end

[Ravi Kumar]

You seem to have a typo in assiging zeros, zerose(n1,n2) must be zeros(n1,n2).

[MEHDI Heydari]

Thanks so much for your attention.

You are right! It works after the correction

[mar tav]

Dear MEHDI Heydari

would you please guide me? I have the same problem as you, but my function is

function [cmatrix] = ccoeffunction(location, B_HMM)

n1 = 4;

nr = numel(location.x);

cmatrix = zeros(n1,nr);

cmatrix(1,:) = B_HMM(1,:);

cmatrix(2,:) = B_HMM(2,:);

cmatrix(3,:) = B_HMM(3,:);

cmatrix(4,:) = B_HMM(4,:);

end

I want to assign the B_HMM matrix as the c coefficient.

[Ravi Kumar]

The ccoeffunction should take two argument, location, state. You are using second argument to pass additional arguments. You can pass additional argument using a wrapper.

In your main scrtip:

B_HMM = ...

ccoefWrapper = @(location, state) ccoeffunction(location, state, B_HMM)

specifyCoefficients(model,'m',...,'c', ccoefWrapper,'f',...)

Then write the function:

function [cmatrix] = ccoeffunction(location, state,B_HMM)

n1 = 4;

nr = numel(location.x);

cmatrix = zeros(n1,nr);

cmatrix(1,:) = B_HMM(1,:);

cmatrix(2,:) = B_HMM(2,:);

cmatrix(3,:) = B_HMM(3,:);

cmatrix(4,:) = B_HMM(4,:);

end

You need to fillin appropriately for ... in the example I have shown above.

Regards,

Ravi

Answer 1

[Jothi Saravanan]

Hi, I am getting the same error. Can someone say what is the mistake I made?

clc; tic;                                              % Clears the screen
clear all;
e=0.5;
mu=0.00689;
V=265.988178;
omega=1047.197;
C=0.0000508;
R=0.508/2;
phi=45;
L=0.127;
%v=12*mu/h^3;
phi=45;
model = createpde();
rect1 = [3, 
    4,
    -1,
    1,
    1,
    -1,
    0.3,
    0.3,
    -0.3,
    -0.3];
r=decsg(rect1);
geometryFromEdges(model,r);
applyBoundaryCondition(model,'dirichlet','Edge',1:4,'u',1*10^5);
specifyCoefficients(model,'m',0,...
                          'd',0,...
                          'c',1,...
                          'a',0,...
                          'f',@fcoeffunction);
generateMesh(model,'Hmax',0.083);
results = solvepde(model);
p= results.NodalSolution;
function f = fcoeffunction(location,state)
N = 1; % Number of equations
nr = length(location.x); % Number of columns
f = zeros(1,nr); % Allocate f
dh=-e*cosd(phi)*sind(location.x+phi)+e*sind(phi)*cosd(location.x+phi);
dt=0.1*omega(e*cosd(phi)*cosd(location.x+phi)+e*sind(phi)*sind(location.x+phi);
o=location.y^2-(L^2/4);
h=C+e*cosd(location.x);
r=(omega*dh+2*dt);
f(1,:)=3*mu*r*o/h^3;
end

[MEHDI Heydari]

Hi,

There are some errors in your fcoeffunction. Creat new function save it with the name fcoeffunction and then it will run!

function f = fcoeffunction(location,~)
eA=0.5;
mu=0.00689;
omega=1047.197;
C=0.0000508;
L=0.127;
%v=12*mu/h^3;
phi=45;
nr = length(location.x); % Number of columns
f = zeros(1,nr); % Allocate f
dh=-eA*cosd(phi)*sind(location.x+phi)+eA*sind(phi)*cosd(location.x+phi);
dt=0.1*omega*eA*cosd(phi)*cosd(location.x+phi)+eA*sind(phi)*sind(location.x+phi);
o=location.y.^2-(L^2/4);
h=C+eA*cosd(location.x);
r=(omega*dh+2*dt);
f(1,:)=3*mu*r.*o./(h.^3);
end