[c] How to work with complex numbers in C?

How can I work with complex numbers in C? I see there is a complex.h header file, but it doesn't give me much information about how to use it. How to access real and imaginary parts in an efficient way? Is there native functions to get module and phase?

This question is related to c complex-numbers

The answer is


The notion of complex numbers was introduced in mathematics, from the need of calculating negative quadratic roots. Complex number concept was taken by a variety of engineering fields.

Today that complex numbers are widely used in advanced engineering domains such as physics, electronics, mechanics, astronomy, etc...

Real and imaginary part, of a negative square root example:

#include <stdio.h>   
#include <complex.h>

int main() 
{
    int negNum;

    printf("Calculate negative square roots:\n"
           "Enter negative number:");

    scanf("%d", &negNum);

    double complex negSqrt = csqrt(negNum);

    double pReal = creal(negSqrt);
    double pImag = cimag(negSqrt);

    printf("\nReal part %f, imaginary part %f"
           ", for negative square root.(%d)",
           pReal, pImag, negNum);

    return 0;
}

For convenience, one may include tgmath.h library for the type generate macros. It creates the same function name as the double version for all type of variable. For example, For example, it defines a sqrt() macro that expands to the sqrtf() , sqrt() , or sqrtl() function, depending on the type of argument provided.

So one don't need to remember the corresponding function name for different type of variables!

#include <stdio.h>
#include <tgmath.h>//for the type generate macros. 
#include <complex.h>//for easier declare complex variables and complex unit I

int main(void)
{
    double complex z1=1./4.*M_PI+1./4.*M_PI*I;//M_PI is just pi=3.1415...
    double complex z2, z3, z4, z5; 

    z2=exp(z1);
    z3=sin(z1);
    z4=sqrt(z1);
    z5=log(z1);

    printf("exp(z1)=%lf + %lf I\n", creal(z2),cimag(z2));
    printf("sin(z1)=%lf + %lf I\n", creal(z3),cimag(z3));
    printf("sqrt(z1)=%lf + %lf I\n", creal(z4),cimag(z4));
    printf("log(z1)=%lf + %lf I\n", creal(z5),cimag(z5));

    return 0;
}

To extract the real part of a complex-valued expression z, use the notation as __real__ z. Similarly, use __imag__ attribute on the z to extract the imaginary part.

For example;

__complex__ float z;
float r;
float i;
r = __real__ z;
i = __imag__ z;

r is the real part of the complex number "z" i is the imaginary part of the complex number "z"


Complex types are in the C language since C99 standard (-std=c99 option of GCC). Some compilers may implement complex types even in more earlier modes, but this is non-standard and non-portable extension (e.g. IBM XL, GCC, may be intel,... ).

You can start from http://en.wikipedia.org/wiki/Complex.h - it gives a description of functions from complex.h

This manual http://pubs.opengroup.org/onlinepubs/009604499/basedefs/complex.h.html also gives some info about macros.

To declare a complex variable, use

  double _Complex  a;        // use c* functions without suffix

or

  float _Complex   b;        // use c*f functions - with f suffix
  long double _Complex c;    // use c*l functions - with l suffix

To give a value into complex, use _Complex_I macro from complex.h:

  float _Complex d = 2.0f + 2.0f*_Complex_I;

(actually there can be some problems here with (0,-0i) numbers and NaNs in single half of complex)

Module is cabs(a)/cabsl(c)/cabsf(b); Real part is creal(a), Imaginary is cimag(a). carg(a) is for complex argument.

To directly access (read/write) real an imag part you may use this unportable GCC-extension:

 __real__ a = 1.4;
 __imag__ a = 2.0;
 float b = __real__ a;