I'm clustering a set of points using DBScan algorithm. I have a set of IDs for a set of points and I have a set of clusters, where each cluster has a set of points. I would like to correlate between the clusters and the points ID.

For example I have a set of ids { 1,2,3,4}, Now if I have two cluster and two clusters has two points, then those two points of the first cluster should have the ids 1,2 and for the second one 3,4. Also If I have 4 clusters and each cluster has one point, then the Ids of the points should be 1,2,3, and 4. Furthermore, If I have two clusters but one cluster has 3 points and the other one has one point, then the Ids of the points should be 1,2,3 for the first cluster's points and for the second cluster's point is 4.

I tried to code it, but I'm stopped at calculating the formula to achieve that scenario.

std::vector<int>_IDs;
// for each cluster
for( int j = 0; j<clusters.size();j++ )
{
// for each point in that cluster
for ( int i=0; i < clusters[j].m_Points.size(); i++)
{
// assign its ID from the _IDs array based and save it in Clusters Vector
clusters[j].m_IDs.push_back(_IDs[j+ i*clusters[j].m_Points.size()]);

}
}

Agreed. I'm unclear what the clustering criteria are. The IDs, 2D co-ordinates, 3D co-ordinates, other? Are the clusters known ahead of time, or are you trying to dynamically create clusters from the points?