The parametric version of fsolve, another enhancement

[Craig Earls]

Heber Farnsworths asked if the additional argument in the new version of
fsolve could be matrices, and it turned out that I had forced the
additional arguments into being real scalar or vectors.

To allow ANY type of octave value to be passed as an additional argument
change the following (at line 171)

for(int i=2;i<nargin;i++){
          extras(i-2)=args(i).vector_value();
      }

to:

for(int i=2;i<nargin;i++){
          extras(i-2)=args(i);
      }

Only the second line is changed, it just removes the vector_value()
call.

I hereby promise to release this thing only through JWE from now on.
This really should be under his configuration control since it replaces
a builtin (of course only if he likes it).
-- 
-----------------------------------------------------------------
Craig P Earls                                 address@hidden
LT US Navy, MIT Ocean Engineering             address@hidden
-----------------------------------------------------------------