Tried using std::vector to avoid having to implement DeviceType into...

#include <TNL/Matrices/SparseRow.h>

#include <TNL/ParallelFor.h>

// Following includes are here to enable usage of std::vector and std::cout. To avoid having to include Device type (HOW would this be done anyway)

#include <iostream>

#include <vector>

namespace TNL {

namespace Matrices {   

getNonZeroElementsCount() const

{

    using NonConstIndex = typename std::remove_const< Index >::type;

    // using DeviceType = typename TNL::Matrices::Matrix::DeviceType;

    NonConstIndex elementCount ( 0 );

//    ParallelFor< Device >::exec( ( NonConstIndex ) 0, length, computeNonzeros );

//    The ParallelFor::exec() function needs a < DeviceType >, how to get this into SparseRow?

    /*

    for( NonConstIndex i = 0; i < length; i++ )

        if( getElementValue( i ) != 0 ) // This returns the same amount of elements in a row as does getRowLength(). WHY?

     */

    // std::vector< Real > vls = values; // Size of values should be something like: (sizeof(this->values)/sizeof(*this->values)) from https://stackoverflow.com/questions/4108313/how-do-i-find-the-length-of-an-array

    for( NonConstIndex i = 0; i < length; i++ ) // this->values doesn't have anything similar to getSize().

        if( this->values[ i * step ] != 0.0 ) // This returns the same amount of elements in a row as does getRowLength(). WHY?

            elementCount++;

    return elementCount;