How 2 make A* Pathfinding

Resources Community Tutorials
,
#45

Sure, I’ll do it tomorrow since I’m out of town and the laptop’s occupied. Though, I don’t think it’s too hard to grasp (that if you read the part 2). Lmk what’s not clear in the topic since there also will be a part 3 so I could reiterate the draft

#46

Open source would be great. !

I loke reviewing the code that way

Thanks

#47

You should add Heap to increase performance, also you could give option to calculate the path from the last part to increase the speed

#48

There’s already that in part 2

I couldn’t quite get what you mean

#49

instead of Retrace the path, you could just make it so the path follow the one point before the last point.

Like astar(goal, startpoint)

#50

Thats really easy to modify, though I digest. I’ll do that on part 3

#51

Maybe add a vaiable to the node called “Parent” and i used this to retrace
`
while Data.LowestFCost do
task.wait()

	local Ball=Instance.new("Part")
	Ball.Shape=Enum.PartType.Ball
	Ball.Size=Vector3.new(2,2,2)
	Ball.Position=Data.LowestFCost.Position
	Ball.Anchored=true
	Ball.CanCollide=false
	Ball.CanQuery=false
	Ball.Material=Enum.Material.Neon
	Ball.BrickColor=BrickColor.new("Lime green")
	Ball.Parent=workspace
	
	table.insert(FinalNodes,Data.LowestFCost.Position)
	Data.LowestFCost=Data.LowestFCost.ParentNode
end

print(Done Path)
return FinalNodes

`

#52

Best way to teach A* is to teach graph theory and then Dijkstra algorithm as a* is just an extension of Dijkstra

1 Like
#53

I made this tutorial for those that wants to create their own A* pathfinding. I know that everyone doesn’t have the same knowledge at their disposal but I suppose that its pretty easy to grasp if you analyze the pseudocode, if not, I’d just watch Sebastian Lague’s A* video.

With that being said, I will try to explain more in depth on part 3

2 Likes