Problem porting Matlab code containing feval

[V P]

Hi All:
I am in the process of porting my old Matlab code to Octave since I no longer have access to Matlab.

I recreated the problem in the following snippet:

function test(fn)
yN = [1.2, 0.8];
i=0
while i < 3,
feval(fn,yN)
i=i+1;
end;
endfunction;

function d=dfns2(y)
d1=-1 + 2./2.^(16 *(-3.75 + y(:,1)).^2);
d2=-1 + 2./ 2.^((-8 + y(:,2)).^2);
d = [d1, d2];
endfunction

in Octave (2.1.71) I do the following:
>test(dfns2)

will get errors I do not understand.

The two m files are attached.
Any help would be appreciated.
Thanks
Vijaya Poudyal

test.m
Description: Binary data

dfns2.m
Description: Binary data