C++ Reference: class SavingsFilteredDecisionBuilder
This documentation is automatically generated.Filter-based decision builder which builds a solution by using Clarke & Wright's Savings heuristic. For each pair of nodes, the savings value is the difference between the cost of two routes visiting one node each and one route visiting both nodes. Routes are built sequentially, each route being initialized from the pair with the best avalaible savings value then extended by selecting the nodes with best savings on both ends of the partial route. Cost is based on the arc cost function of the routing model and cost classes are taken into account.