---

Pathfinding

---

How pathfinding works

Although details vary a bit depending on the selected ⊚ Search algorithm, the basic gist is, for each path & cluster:

1. A suitable Seed and Goal point (Vtx) are picked within the cluster.
2. The Search Algorithm will then find the best path that goes from Seed to Goal, accounting for its internal maths, relying on 🝰 Heuristics to determine whether one connection is better than another.
3. Once a path is found between a given Seed and its Goal, traversed points are then added to a dataset, in order from Seed (start) to Goal (end).

Note: The Seed and Goal node are picked based on closest distance to input positions only (point bounds are ignored)

Starting from the seed, each “next step” is weighted according to the V Vertex weight and the E Edge weight that connects to it.
The search returns the path found with the lowest possible weight, or score.

While the selected search algorithm is important, 🝰 Heuristics are more critical to the operation, as user-defined weights can drastically change and shape the path deemed best by the search.

Note: The Plot nodes variations don’t have a goal picker and instead process each point Dataset as a list of points to go through from start to finish. The first point is the initial seed, the last point is the final goal, and then a path is found that goes through each point in-between, in order.

---

Pathfinding Nodes

Note that there is also two hidden node that enable pathfinding using the existing navmesh: Navmesh Pathfinding and Plot Navmesh. They’re not part of the main pool because they’re very legacy. They can still be very useful if you do level blocking with a navmesh, not so much for open world.

---

Available 🝰 Heuristics modules

---

Available ⊚ Search modules