1. Frontier
The frontier is the algorithm's to-do list. These are cells it has discovered but has not fully explored yet.
Pathfinding Lab
Draw walls, move nodes, then watch algorithms search in real time.
How to read the search
2. Visited
Visited cells have already been processed. Once a cell is visited, the algorithm usually does not need to check it again.
3. Previous pointers
When a cell is discovered, it remembers which cell found it. At the end, the yellow path is built by following those breadcrumbs backward.
4. Heuristics
A heuristic is an educated guess. A* and Greedy use Manhattan distance: how many grid steps away the target looks if there were no walls.
5. Guarantees
BFS, Dijkstra, and A* can find shortest paths in this grid. DFS and Greedy can be fast or interesting, but they can choose poor routes.
Try this
- Generate a maze and compare A* vs DFS.
- Draw a wall trap and watch Greedy get fooled.
- Lower speed delay to see the full search shape.