Help me please.
Tell me how to sum each vector in a two-dimensional vector and write the result into a one-dimensional vector. On C .
Input:

std::vector<std::vector<double>> a = { { 1, 2 }, { 3, 4 }, { 5, 6 } };

Required Output:

b = { 9, 12 }

I found how to sum all the elements in a 2D vector, but I don't know how to sum the vectors in an existing 2D vector and write the result to a 1D vector

#include <iostream>
#include <vector>

int main() 
{
    std::vector<std::vector<double>> a = { { 1, 2 }, { 3, 4 }, { 5, 6 } };
    std::vector<double> b;
    std::vector<int> v;

    int sum = 0;
    for (int i = 0; i < a.size(); i  )
        for (int j = 0; j < a[i].size(); j  )
        {
            sum  = a[i][j];
        }
    std::cout << sum;
    return 0;
}

CodePudding user response：

From your example I assume that in the sums vector that you require, element in index i, is the sum of all elements in index i in the vectors in a. This requires that all elements in a have the same size.

You can do it in the following way:
Traverse the elements in a. For each element: traverse the element (which is a vector itself) and sum element i into the ith bin of the output vector.

Something like:

#include <iostream>
#include <vector>
#include <assert.h>

int main()
{
    std::vector<std::vector<double>> a = { { 1, 2 }, { 3, 4 }, { 5, 6 } };
    size_t num_of_sums = a[0].size();   // we assume all elements in a have the same size
    std::vector<double> sums(num_of_sums);
    for (auto const & v : a)
    {
        assert(v.size() == num_of_sums);
        for (int j = 0; j < num_of_sums; j  )
        {
            sums[j]  = v[j];
        }
    }
    for (auto d : sums)
    {
        std::cout << d << ", ";
    }
    std::cout << std::endl;
    return 0;
}

Output:

9, 12,

Note that my sums vector contains doubles, like your input data (unlike your sum which is an int and seems improper to sum double values).

CodePudding user response：

This would do:

• Create a vector b of 2 elements.
• Walk each vector v in a, adding each number to b.
• fmt is a utility library that lets you easily print vectors.

[Demo]

#include <fmt/ranges.h>
#include <vector>

int main() {
    std::vector<std::vector<double>> a = { { 1, 2 }, { 3, 4 }, { 5, 6 } };
    std::vector<double> b(2);

    for (const auto& v : a) {
        b[0]  = v[0];
        b[1]  = v[1];
    }
    fmt::print("{}", b);
}

// Output: [9, 12]

CodePudding user response：

You can use std::accumulate:

#include <iostream>
#include <numeric>
#include <vector>


int main() {
    std::vector<std::vector<double>> a = { { 1, 2 }, { 3, 4 }, { 5, 6 } };

    auto res = std::accumulate(a.begin(),a.end(),std::vector<double>{0.0,0.0},[](auto a,const auto& b){        
        a[0]  = b[0];
        a[1]  = b[1];
        return a;
    });
    for (const auto& r : res ) { std::cout << r << " ";}
}

Live Demo

The elements of a have to have all same size for this to work. I assumed they all have two elements, if not the adding of the two elements has to be replaced by a loop.