Open emsr opened 7 years ago
So we would have, say,
template<typename _Tp>
struct __cyl_bessel_t
{
struct param_type
{
_Tp __nu;
_Tp __z;
// Return true Wronskian for the cylindrical Bessel functions.
_Tp __wronskian() const;
};
_Tp _J_value;
_Tp _J_deriv;
_Tp _N_value;
_Tp _N_deriv;
// Return Wronskian of the obtained solutions.
_Tp __wronskian();
};
I think I like this.
Would it be a good idea for specfun states to have a separate param type that would have the (true) wronskian?