Operator overloading for a class containing boost::numeric::ublas::matrix

Question

I have a class template that has a boost matrix as a private member variable. The matrix data type is determined by the class type when constructed. This class has a member function that is supposed to add a constant to the member matrix. The constant is consistent with the matrix data type. I am having trouble writing an overloaded operator that will return an updated member matrix for arbitrary constant values. Currently, another member function does this addition. My code is as follows;

/*content from main.cpp compiled with `g++ -o main main.cpp` */
#include <iostream>
#include <boost/numeric/ublas/matrix.hpp>
#include <boost/numeric/ublas/io.hpp>

using namespace boost::numeric::ublas;
using namespace std;

template <typename T>
class MTool
{
private:
matrix<T> m_ME;

public: 
MTool(int N, int M) { m_ME.resize(M, N); } // constructor
~MTool() { m_ME.clear(); } // destructor
void PutMatValues(); // insert values into member matrix
void AddConst(const T &k) { m_ME+=k; } // add a constant to member matrix 
void PrintMat() { cout << "The ME " << endl << m_ME << endl; } // print the member matrix

// overloaded operator function
matrix<T> operator+ <> (const T& kvalue) { return (m_ME+kvalue); }  
};

template<typename T>
void MTool<T>::PutMatValues()
{
    for (unsigned row=0; row<m_ME.size1(); row++)
        for (unsigned col=0; col<m_ME.size2(); col++)   
            m_ME(row,col)=static_cast<T> (row*col);
} 

int main()
{
    MTool<double> mymat(5, 3); 
    double dk=123.67001;
    
    mymat.PutMatValues();   
    mymat.PrintMat();
    mymat.AddConst(dk);
    mymat.PrintMat();
     
    return 0; 
} 

Some of the compiler errors I get are

error: template-id ‘operator+<>’ in declaration of primary template

error: no match for ‘operator+=’ in ‘((MTool*)this)->MTool::m_ME += k’

I am rather new to C++ templates and classes and am sure there is something fundamental missing from my approach. Any suggestions would highly appreciated.

Answer 1

The first is just a syntax error, this member operator+ would be written as

MTool<T> operator+ (const T& kvalue) const { ...

although it's somewhat unusual to see a binary plus operator as a member function: it's almost always implemented as a non-member, so that you could write expressions c + M as well as M + c.

The second error is simply pointing out that boost::numeric::ublas::matrix has no operator+= that takes a scalar argument. There is also no operator+ that can add a matrix and a scalar, so the expression m_ME+kvalue in your operator+ won't compile either.

Addition is only defined between matrices of equal shape. If you're adding the scalar to every element of the matrix, you're going to need to write that, something like:

void AddConst(const T &k) {
    for(auto& element : m_ME.data())
           element += k;
}