I am writing my own CFD code using c language. I want to create a function which can deal with both single and double precision arrays such that

grad_cal1(fi1) where fi1 is an array of float

grad_cal2(fi2) where fi2 is an array of double

void grad_cal1(float *fi){a.h}

void grad_cal2(double *fi){a.h}

Inside the function, the programme is exactly the same. Is there any way to call the same function instead of two?

Why not simplify things by making the whole code double precision? Why do you really need to mix variable types?

Hello John,

in C it is possible to compile the function with appropriate preprocessor macros, e.g.

#define REAL float

or

#define REAL double

and the corresponding definitions in the source, e.g.

void grad_cal1(REAL *fi){a.h}

Another (and much better) possibility would be using C++ templates.

Regards, Markus

Markus' preprocessor approach lets you create either version independently at compile-time. To have both options simultaneously at run-time you are basically talking about object oriented programming: This is a case for C++, not C.