imnotteixeira / cal_proj_17-18 Goto Github PK

create a calcWeight so that the updateQueue both in Dijkstra and A* is abstracted from that calculation and the code is more modular

This feature can be implemented with just a loop outside the chosen algorithm, but instead of using pre-calculated weights for the edges, calculate them as they are needed in the algorithm (by letting the edges have weight calc functions). This will avoid calculating values for all the edges (even the ones that will not be used) before each path calculation.

Ideas for applications:
Adapt to traffic;
Avoid red lights (adding weight if the user is expected to catch a red light through a certain path, based on the road's speed and the distance);

Develop a brute force solution for path finding to be able to compare its performance to the other, more advanced, algorithms

Ability to generate Samples for testing the algorithms of pathfinding

Create a method that returns a vector with pairs of <Node,Edge> that connect with the node with 'nodeId'

Make A* class extend Dijkstra and modify calcOptimalSolution