Combining separate files

??? Subscripted assignment dimension mismatch.

Error in ==> fminsearch at 205
fv(:,1) = funfcn(x,varargin{:});

Error in ==> p5final at 95
sig4=fminsearch('P5F',1.5);% minimse the cost function using 'fminsearch' giving
a predicted value for sigma4

i get this error for the coding

sig4=fminsearch('P5F',1.5);% minimse the cost function using 'fminsearch' giving a predicted value for sigma4
c44=exp(-z.^2/(2*(sig4)^2));% Use global value for 'z' and predicted value for sigma4 to find a forth predicted value for average plume concentration ratio 'c'
e4=c-c44;% calculate prediction error between global c (average plume concentration ratio) and the forth predicted plume concentration

>> P5F is a another file codes are

function J=P5F(Sig)
global c2 z
J=0;
for k=1:201;
s1=c2*(k);
s2=exp(-z(k)^2/(2*(Sig)^2));
J=J+(s1-s2)^2;
end

Answer 1

[Walter Roberson]

The argument to fminsearch must be a function handle, not a string. Use @P5F instead of 'P5F'

[buxZED]

still gives the same error

?? Subscripted assignment dimension mismatch.

Error in ==> fminsearch at 205

fv(:,1) = funfcn(x,varargin{:});

Error in ==> Q5 at 99

sig4=fminsearch(@P5Function ,1.5);% minimse the cost function using

'fminsearch' giving a predicted value for sigma4

[Matt Tearle]

Actually, you *can* use a string... but you *should* use a function handle.

Answer 2

[Matt Tearle]

What are your globals c2 and z? Not having access to them, I hard-coded in a scalar for c2 and a 201-element vector for z, and it worked fine for me.

While we're here, don't use globals. Do this instead:

function J=P5F(Sig,c2,z)
  [etc]

Then, in p5final

c2 = <something>;
z = <something>;
f = @(sig) P5F(sig,c2,z)
sig4=fminsearch(f,1.5)

See the doc for more information on this use of anonymous function handles.