There is an error in call the function file

function V = fluxfunction(a,b)
% Godunov
V = max(f(max(a,0)),f(min(b,0)));
end
%%%%%%%%%%%%%%%%%%
% Godunov Scheme
% ut + (u^2/2)x = eps*uxx;
x0 = 0;
xf = 5;
N = 100;
h = (xf-x0)/N;
eps = 0.01;
delt = 0.5*h;
lambda = delt/h;
mu = delt/(h^2);
t0 = 0;
tf = 5;
x = zeros(N-1,1);
u0 = zeros(N-1,1);
for j=1:N-1
    x(j) = x0+j*h;
    if(x(j) < 1)
        u0(j) = sin(pi*x(j));
    end
    if(x(j) > 1)
        u0(j) = 0;
    end
end
M = fix((tf-t0)/delt);
unew = zeros(N-1,1);
t = t0;
for k=1:M
    unew(1) = u0(1) - lambda*(fluxfunction(u0(1),u0(2)) - fluxfunction(0,u0(1))) + mu*eps*(u0(2) - 2*u0(1) + 0);
    unew(N-1) = u0(N-1) - lambda*(fluxfunction(u0(N-1),0) - fluxfunction(u0(N-2),u0(N-1))) + mu*eps*(u0(N-2) - 2*u0(N-1) + 0);
    for j=2:N-2
        unew(j) = u0(j) - lambda*(fluxfunction(u0(j),u0(j+1)) - fluxfunction(u0(j-1),u0(j))) + mu*eps*(u0(j+1) - 2*u0(j) + u0(j-1));
    end
    plot(x,unew)
    title('Godunov Scheme');
    ylim([0,1]);
    pause(0.01);
    u0 = unew;
    t = t+delt;
end
ugod = unew;

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%Error %%%%%%%%%%%%%%%%%%%%%

Not enough input arguments.

Error in f (line 3)

V = exp(-t)*sin(pi*x)*(pi^2-1);

Error in fluxfunction

V = max(f(max(a,0)),f(min(b,0)));

Error in Godnov

[KALYAN ACHARJYA]

f in the function?

Answer 1

[Star Strider]

You need to call ‘fluxfunction’ from a script with specific values for ‘a’ and ‘b’ that already existing in the script workspace. You cannot run it by clicking on the green ‘Run’ arrow in the Editor or by running it from the Command Window as you would a script, without supplying apppropriate values for the arguments.

[Walter Roberson]

For once that is not the case. The error message is inside f, which is being sucessfully passed something -- but f expects at least two parameters. Probably f expects both x and t

[Star Strider]

I was thinking that in the ‘V’ assignment (that’s throwing the error), ‘f’ is being passed ‘max(a,0)’ and ‘min(b,0)’ so that if ‘a’ and ‘b’ don’t exist, those arguments don’t exist and ‘f’ (that apparently takes only one argument) throws the error (although I would expect max or min to throw it instead in that instance).

I was actually able to find that reference as the MATLAB code for Shape-from-Shading, with the same function names although the code is quite different. I don’t know what’s going on with this.

[Walter Roberson]

f would not get invoked until all of the parameters to the call to f are ready -- so if the problem were that a or b were missing, that would be detected as an error in fuxfunction not as an error in f

You can only get error reports with line numbers inside f if f has started execution.

[Walter Roberson]

fluxfunction([8 1 5],[9 3 -1])

Not enough input arguments.

Error in solution>f (line 7)
V = exp(-t)*sin(pi*x)*(pi^2-1);

Error in solution>fluxfunction (line 4)
V = max(f(max(a,0)),f(min(b,0)));

function V = fluxfunction(a,b)
  V = max(f(max(a,0)),f(min(b,0)));
end
function V = f(x,t)
    V = exp(-t)*sin(pi*x)*(pi^2-1);
end

[Walter Roberson]

fluxfunction()

Not enough input arguments.

Error in solution>fluxfunction (line 4)
V = max(f(max(a,0)),f(min(b,0)));

function V = fluxfunction(a,b)
  V = max(f(max(a,0)),f(min(b,0)));
end
function V = f(x,t)
    V = exp(-t)*sin(pi*x)*(pi^2-1);
end

Observe that if not enough inputs are passed to fuxfunction so the parameters to f cannot be evaluated, then the error is reported in fluxfunction, whereas if enough parameters are passed to fluxfunction but not enough are passed to f, then the error is reported in f.