Skip to content

Latest commit

 

History

History
20 lines (11 loc) · 986 Bytes

week3.md

File metadata and controls

20 lines (11 loc) · 986 Bytes

Week 3

What I have done this week and program progression

I made a minimum heap for holding the nodes, increased the test coverage and started implementing IDA*.

What I have learned this week

I have learned that IDA* is a mixture of BFS and DFS search with heuristic acting as a bound. I also noticed that A* performs fine overall, but JPS and IDA* outperforms A* depending on the complexity of the map. For example, IDA* is the fastest of the three when there are no obstacles/few obstacles, but JPS is the fastest of the three when the map has corridors.

What is unclear or has proven to be difficult?

I've done IDA* according to the pseudo code. It runs well when there are no obstacles, but on a path with obstacles it usually timeouts.

Edit: I added a tiebreaker for IDA*, now it finds a path when there is an obstacle, but is still rather slow when facing an obstacle.

What should I do next

I'll start to create the other documents for the project.