Trouble with double complex numbers in C++

AlexBB · 01-23-2016, 03:06 PM

It is Ubuntu 14.04. The compiler is g++.

I need to operate with complex<double> values. When I include

Code:

#include <cmath>
#include <complex>

I expect to get full range of complex numbers in my code. Instead this definition

Code:

typedef complex<double> dcomp;

gives me an error on compilation: complex is not a template.

This website, however, shows that it is a correct usage:

http://stackoverflow.com/questions/1...-number-i-in-c

Also this website

http://stackoverflow.com/questions/1...t-defined-in-c

suggests this construction, that allegedly allows to use complex<double>:

template <typename T, typename U>
inline std::complex<T> operator*(std::complex<T> lhs, const U& rhs)
{
return lhs *= rhs;
}

It does not even compile. It gives me two errors:

complex in namespace std does not name a type

If I remove std::

I get another error:

Complex is not a template

Why is it so difficult to use double in complex numbers in C++?

Thanks, - Alex

AlexBB · 01-23-2016, 04:56 PM

Also this nuisance:

In the website
http://en.cppreference.com/w/cpp/numeric/complex

Code:

#include <iostream>
#include <iomanip>
#include <complex>
#include <cmath>
 
int main()
{
    using namespace std::literals;
    std::cout << std::fixed << std::setprecision(1);
 
    std::complex<double> z1 = 1i * 1i;     // imaginary unit squared
    std::cout << "i * i = " << z1 << '\n';
 
    std::complex<double> z2 = std::pow(1i, 2); // imaginary unit squared
    std::cout << "pow(i, 2) = " << z2 << '\n';
 
    double PI = std::acos(-1);
    std::complex<double> z3 = std::exp(1i * PI); // Euler's formula
    std::cout << "exp(i, pi) = " << z3 << '\n';
 
    std::complex<double> z4 = 1. + 2i, z5 = 1. - 2i; // conjugates
    std::cout << "(1+2i)*(1-2i) = " << z4*z5 << '\n';
}

Trying to compile with g++ I get a pageful of errors:

Quote:

g++ -std=c++11 test1.cpp
test1.cpp: In function ‘int main()’:
test1.cpp:8:26: error: ‘literals’ is not a namespace-name
using namespace std::literals;
^
test1.cpp:8:34: error: expected namespace-name before ‘;’ token
using namespace std::literals;
^
test1.cpp:11:31: error: unable to find numeric literal operator ‘operator"" i’
std::complex<double> z1 = 1i * 1i; // imaginary unit squared
^
test1.cpp:11:36: error: unable to find numeric literal operator ‘operator"" i’
std::complex<double> z1 = 1i * 1i; // imaginary unit squared
^
test1.cpp:14:40: error: unable to find numeric literal operator ‘operator"" i’
std::complex<double> z2 = std:

ow(1i, 2); // imaginary unit squared
^
test1.cpp:18:40: error: unable to find numeric literal operator ‘operator"" i’
std::complex<double> z3 = std::exp(1i * PI); // Euler's formula
^
test1.cpp:21:36: error: unable to find numeric literal operator ‘operator"" i’
std::complex<double> z4 = 1. + 2i, z5 = 1. - 2i; // conjugates
^
test1.cpp:22:43: error: ‘z5’ was not declared in this scope
std::cout << "(1+2i)*(1-2i) = " << z4*z5 << '\n';
^
alex@alex-HP-ProBook-6555b:~/GFORTRAN-APPS/SPHERE-DESIGN/SPHERE-CPP$

Why do I have so much trouble with complex numbers in C++?

Thanks.

AlexBB · 01-23-2016, 05:05 PM

And how about this? I have the following function header:

Code:

public : void legendreTransform (int NN,complex TotFourierTransf[][NN],complex GaussLegTransf[][NN],PhiLimits phiBounds)
  {

This is what I get on compilation:

Quote:

g++ -std=c++11 sphere_perspective.cpp
sphere_perspective.cpp:551:72: error: use of parameter outside function body before ] token
public : void legendreTransform (int NN,complex TotFourierTransf[][NN],complex GaussLegTransf[][NN],PhiLimits phiBounds)
^
sphere_perspective.cpp:551:73: error: expected ) before , token
public : void legendreTransform (int NN,complex TotFourierTransf[][NN],complex GaussLegTransf[][NN],PhiLimits phiBounds)
^
sphere_perspective.cpp:551:74: error: variable or field complex declared void
public : void legendreTransform (int NN,complex TotFourierTransf[][NN],complex GaussLegTransf[][NN],PhiLimits phiBounds)
^
sphere_perspective.cpp:551:74: error: expected ; at end of member declaration
sphere_perspective.cpp:551:82: error: GaussLegTransf does not name a type
public : void legendreTransform (int NN,complex TotFourierTransf[][NN],complex GaussLegTransf[][NN],PhiLimits phiBounds)

What is wrong with my compiler? Am I making mistakes?

Thanks.

norobro · 01-23-2016, 10:21 PM

Hi Alex,

Post #1: Use:

Code:

typedef std::complex<double> dcomp

Post #2: http://en.cppreference.com/w/cpp/num...mplex#Literals Notice that operator""i is a c++14 feature. (-std=c++14)

Post #3: The compiler expects a member declaration to be terminated with a semicolon. You have an opening brace.

Code:

sphere_perspective.cpp:551:74: error: expected ; at end of member declaration

AlexBB · 01-24-2016, 05:03 PM

Thank you norobro. I did pay attention to c++14 but the compiler rejected it. I tried this:

Code:

g++ -std=c++14 sphere_perspective.cpp

Where and how can I get C++ 2014 version?

Also, I am sure I had a semicolon at the end of that line. I removed this line of code after posting. I now reentered it and got an error anyway:

Code:

sphere_perspective.cpp:14:9: error: complex in namespace std does not name a type
 typedef std::complex<double> dcomp;

The line of code is now:

Code:

typedef std::complex<double> dcomp;

^
How can I advance my compiler to C++14?

Thanks, - Alex

AlexBB · 01-24-2016, 05:21 PM

I tried clang-3.5. It gave me a copious output but many lines had a proclamation: "note." So, there were errors, warnings and notes. What is the note in this context?

Thanks, - Alex

norobro · 01-24-2016, 06:29 PM

The only thing that requires c++14 is operator""i. In lieu of using it in that program, you can declare/define a complex variable.

The following compiles on my machine using g++-4.7. Give it a try:

Code:

// g++ -std=c++11 -o test test.cpp

#include <iostream>
#include <iomanip>
#include <complex>
#include <cmath>
 
int main()
{
    typedef std::complex<double> dcomp;    // your typedef
    //using namespace std::literals;
    std::cout << std::fixed << std::setprecision(1);
    
    dcomp one_i(0.0, 1.0);
    dcomp z1 = one_i * one_i;     // imaginary unit squared
    std::cout << "i * i = " << z1 << '\n';
	
    dcomp z2 = std::pow(one_i, 2); // imaginary unit squared
    std::cout << "pow(i, 2) = " << z2 << '\n';
	
    double PI = std::acos(-1);
    dcomp z3 = std::exp(one_i * PI); // Euler's formula
    std::cout << "exp(i, pi) = " << z3 << '\n';
	
    dcomp two_i(0.0, 2.0);					     
    dcomp z4 = 1. + two_i, z5 = 1. - two_i; // conjugates
    std::cout << "(1+2i)*(1-2i) = " << z4*z5 << '\n';
}

If you want to post your code, I'll be happy to take a look.

NevemTeve · 01-24-2016, 11:33 PM

@OP: Are you you just experiencing, or do you actually want to work with complex numbers? If it is the latter, then use C instead of C++.

a4z · 01-25-2016, 12:51 AM

Quote:

Originally Posted by NevemTeve

@OP: Are you you just experiencing, or do you actually want to work with complex numbers? If it is the latter, then use C instead of C++.

why do you think so?

I see no advantage in the C version, it's pretty similar

C++

Code:

#include <iostream>
#include <iomanip>
#include <complex>
#include <cmath>
 
int main()
{
    using namespace std::literals;
    std::cout << std::fixed << std::setprecision(1);
 
    std::complex<double> z1 = 1i * 1i;     // imaginary unit squared
    std::cout << "i * i = " << z1 << '\n';
 
    std::complex<double> z2 = std::pow(1i, 2); // imaginary unit squared
    std::cout << "pow(i, 2) = " << z2 << '\n';
 
    double PI = std::acos(-1);
    std::complex<double> z3 = std::exp(1i * PI); // Euler's formula
    std::cout << "exp(i, pi) = " << z3 << '\n';
 
    std::complex<double> z4 = 1. + 2i, z5 = 1. - 2i; // conjugates
    std::cout << "(1+2i)*(1-2i) = " << z4*z5 << '\n';
}

C

Code:

#include <stdio.h>
#include <complex.h>
#include <tgmath.h>
 
int main(void)
{
    double complex z1 = I * I;     // imaginary unit squared
    printf("I * I = %.1f%+.1fi\n", creal(z1), cimag(z1));
 
    double complex z2 = pow(I, 2); // imaginary unit squared
    printf("pow(I, 2) = %.1f%+.1fi\n", creal(z2), cimag(z2));
 
    double PI = acos(-1);
    double complex z3 = exp(I * PI); // Euler's formula
    printf("exp(I*PI) = %.1f%+.1fi\n", creal(z3), cimag(z3));
 
    double complex z4 = 1+2*I, z5 = 1-2*I; // conjugates
    printf("(1+2i)*(1-2i) = %.1f%+.1fi\n", creal(z4*z5), cimag(z4*z5));
}

AlexBB · 01-26-2016, 08:49 AM

@NevemTeve, I need to work with complex numbers. I can invoke (use) complex numbers all right, it is the double complex that gives me trouble. I need higher precision than float complex. Spherical Harmonics that I have to use are complex functions by definition, I mean all values are complex: a + ib.

NevemTeve · 01-26-2016, 10:00 AM

Sorry for the delay; actually, I wouldn't use C99 either, instead I'd manually declare a struct of two double (re and im; or perhaps r and phi).

Well, back to your case, please provide a minimal complete example, that shows the problem.