advent-of-code-2022 VS advent-of-code-julia

The code is horrible, but performance is great. The basic idea of the algorithm is to consider all the valves we can move to. We compute how long it will take to move to a valve, and we also precompute how much pressure will be released in total by this valve using the remaining time. Then, if there are paths that have an equal distance, we only have to consider the path that has the highest potential pressure release. This drastically reduces the iterations needed, hence the low runtime. Also, the distances are not precomputed, but they are cached. This is more efficient, since we don't need all paths (for example we can skip all paths that have 0 flow rate, as we will never move to them, only pass them).

Julia using linked list

I used your solution and implement it in Julia. It takes 58.756 ms for part 1 and 60.697 ms for part 2.

Julia simple solution simulating the sand. Too much code duplication but nothing better comes to my mind.

Julia

Julia using ResumableFunctions for parsing, DataStructures for queue and OrderedCollections for Dict.

Julia. For part 1 there is no need to save the value for each cycle but this is necessary for part 2.

Julia without black magic.

Julia using the same solution for both problem.

Julia using DataStructures Stack

Julia using modular arithmetic