Question

I’ve recently started using new C++11 features like lambda expressions to make my code easier to read.

In this example, I want to generate a vector of numbers [0 to n] sorted randomly. I have some code similar to the following

    #include <vector>
    #include <algorithm>
    #include <iterator>
    using namespace std;
    void _generateRandomIndices(vector<unsigned int> & indices,
                const unsigned int & n) {
        indices.clear();
        unsigned int i = 0;
        generate_n(back_inserter(indices), n , [&i] ()-> unsigned int{ 
                                                              return i++;
                                                                     } );
        random_shuffle(indices.begin(), indices.end());
    } 

However, using g++ 4.6 on windows(with the -std=c++0x flag) threw me the following error:

no matching function for call to 'generate_n(std::back_insert_iterator<std::vector<unsigned int> >, const unsigned int&, _generateRandomIndices(std::vector<unsigned int>&, const unsigned int&)::<lambda()>)'

What is the right way to use a lambda expression in this case ?

Answer 1

Sounds like you are missing an #include <algorithm>, since GCC 4.5.1 aswell as MSVC10 compile the following code without error:

#include <vector>
#include <algorithm>
#include <iterator>
using namespace std;

void _generateRandomIndices(vector<unsigned int> & indices,
        const unsigned int & n) {
    indices.clear();
    unsigned int i = 0;
    generate_n(back_inserter(indices), n , [&i] ()-> unsigned int{ 
                                                          return i++;
                                                                 } );
    random_shuffle(indices.begin(), indices.end());
}

int main(){
}