How to extract real part from complex number?

[alexBB]

I'll tell you c++ is phenomenally difficult for a beginner. I need to extract the real part from a complex number. My compiler is g++ in Ubuntu 12.04.

Nothing I have tried worked so far and I googled and googled. It is just ridiculous. Let say the number is complC, then complC.real() does not work, std::real (complC) does not work either. Needless to day I tried numerous other variants. This is a simple operation. Why is it made so difficult?

Thanks, - Alex

[alexBB]

OK, I have this code:

 
#include "fft.h"
#include <math.h>
#include <fstream>      // std::ofstream
#include <complex>

  int main ()
{
  std::ofstream outFileInput,outFileOutput;
  complex Input[1024]=0, Output[1024]=0;
  complex valueC;
  int N=1024;
  double xx;
  double welshWindow(int n, int N);
  outFileInput.open ("cppFFTinp.dat");
  outFileOutput.open("cppFFTout.dat");
  for (int jj = 0; jj<N-1; jj++)
    {
      xx = welshWindow(jj,N);
      valueC = complex(xx,0);
      Input[jj] = valueC;
      outFileInput << jj << "  " << xx << std::endl;
    }
  CFFT::Forward (Input,Output,N);
  for (int jj = 0; jj<N-1; jj++)
    {
      valueC = Output[jj];
      xx = std::real(valueC);   // this statement does not compile
      outFileOutput << jj << "  " << xx << std::endl;
    }
}            // main  

If I remove the statement marked // This statement does not compile, the file compiles and I can run it. As I said, I need to extract the real part of the complex number in question. In fortran it is elementary. Here it is a monumental problem. It is simply gross.

If I leave the statement in question and to g++ fft.cpp I get these errors:

 fft.cpp: In function ‘int main()’:
fft.cpp:33:28: error: no matching function for call to ‘real(complex&)’
fft.cpp:33:28: note: candidates are:
/usr/include/c++/4.6/complex:545:5: note: template<class _Tp> _Tp& std::real(std::complex<_Tp>&)
/usr/include/c++/4.6/complex:550:5: note: template<class _Tp> const _Tp& std::real(const std::complex<_Tp>&) 

33 is the number line of that statement. What can I do about the "candidates?" I think my code already complies with this template.

[D80DSL3A]

I found this by following a link you posted in your previous thread on the proprietary complex class you're using:

 //   Basic operations - taking parts
   double re() const { return m_re; }
   double im() const { return m_im; }

Here's where the class functions are described
http://www.librow.com/articles/article-10/appendix-b-1

[alexBB]

Thanks, but that leads to terrible problems. In the end if I include complex.h I will get five terminals worth of errors. I've tried it before. When I included these definitions you posted in the file I got a bunch of errors also. I am not sure it is worth even trying to post them unless requested.

There should be a SIMPLE way to get the real part of a complex number. WHAT IS GOING ON? Anybody knows how to do it?

[D80DSL3A]

I'm sorry you're having such problems. I see your other thread on the compilation problem you're having.
I copy/pasted the code for the complex.h and complex.cpp files to my Code::Blocks compiler
and tried this:

#include<iostream>
#include "complex.h"

int main()
{
    complex c1( 3.0, 5.0 );
    std::cout << c1.re() << std::endl;

    return 0;
}

It built and ran 1st try.

Are you naming the files something other than .h and .cpp ?

[norm b]

The following compiles on my box. Of course, since you didn't provide the definition for double welshWindow(), I had to comment out line 18 and only get zeros as output.

g++ -std=c++11 main.cpp fft.cpp

#include "fft.h"
#include <math.h>
#include <fstream>      // std::ofstream
#include <complex>

  int main ()
{
  std::ofstream outFileInput,outFileOutput;
  complex Input[1024]={0}, Output[1024]={0};
  complex valueC;
  int N=1024;
  double xx;
  double welshWindow(int n, int N);
  outFileInput.open ("cppFFTinp.dat");
  outFileOutput.open("cppFFTout.dat");
  for (int jj = 0; jj<N-1; jj++)
    {
//      xx = welshWindow(jj,N);
      valueC = (xx,0);
      Input[jj] = valueC;
      outFileInput << jj << "  " << xx << std::endl;
    }
  CFFT::Forward (Input,Output,N);
  for (int jj = 0; jj<N-1; jj++)
    {
      valueC = Output[jj];
      xx = valueC.re(); //std::real(valueC);   // this statement does not compile
      outFileOutput << jj << "  " << xx << std::endl;
    }
}            // main 

[alexBB]

norm b, thank you. The welshWindow is here:

 double welshWindow (int n, int N)
    {
      double res;
      res = (N-1)/2.0;
      res = 1.0 - pow ((n - res)/res,2);
      return res;
    } 

I got it from Wikipedia article "Window function." You can find it there among others.

Thank you much. WOW!!! It compiled. It works. Thank you very much!

[norm b]

You're welcome. I was just reiterating what fun2code pointed out.

Take heed of the warning on line 19:
valueC = (xx,0);

[alexBB]

Everything, everything is working now. Thank you very much for your help. I took care of the warning also.

- Alex