ktims/aoc2024
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
53 Commits 1 Branch 0 Tags
ebf5a0a489065d20697a8c8f93111cedc0d7ca30
Commit Graph
53 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
ktims ebf5a0a489 day13: complete solution 2024-12-13 02:12:18 -08:00
ktims 4aa7e9f43c day3: perf - branchless 2024-12-12 17:56:28 -08:00
ktims a5439062a4 day3: performance - avoid converting &[u8] to &str 2024-12-12 17:38:42 -08:00
ktims 3658183deb day7: performance. DFS with pruning. 2024-12-12 16:02:13 -08:00
ktims 38cba37b06 day5: perf: switch to sort-based implementation 2024-12-12 14:50:00 -08:00
ktims 6022d2cc39 multiple: Use FxHashMap for s p e e d 2024-12-12 14:07:22 -08:00
ktims 447ff5c62c grid: improve ergonomics with more trait impls and other improvements 
update puzzles to pass tests, some performance gains
 2024-12-12 02:14:29 -08:00
ktims c213bbbc27 day12: optimization and cleanup 2024-12-12 01:24:14 -08:00
ktims a56fc933c9 day12: complete solution 2024-12-11 22:27:59 -08:00
ktims 31eb500832 day7: slight performance improvement from skipping ThreadLoacls 2024-12-11 19:40:45 -08:00
ktims 4c14c6092e bugfix: count off by one, thought ^ was left in the map, but it gets 
replaced by X
 2024-12-11 18:59:58 -08:00
ktims de535303d4 day11: don't use aoc_generator 2024-12-11 18:43:57 -08:00
ktims 50b6d045e7 multiple: genericize Grid over coordinate type, parallelism for day 11 2024-12-11 18:43:13 -08:00
ktims d2defae8a2 chore: make day modules public 2024-12-11 18:40:53 -08:00
ktims cd8900d936 chore: enable LTO for release builds 2024-12-11 18:40:29 -08:00
ktims 11b29a4d57 chore: refactor for cargo-aoc / codspeed 2024-12-11 15:45:52 -08:00
ktims 3bfde9fd9b day11: some cleanup, slight performance 2024-12-11 03:11:17 -08:00
ktims 2e239681ce day11: performance 2024-12-11 00:55:47 -08:00
ktims a7354b6ed7 day11: part 2 solution 2024-12-10 23:40:03 -08:00
ktims 462918b382 day11: part 1 solution 2024-12-10 21:55:00 -08:00
ktims 8af11a6092 day10: performance. slight gain from memoizing visited positions in p1. 2024-12-10 17:07:00 -08:00
ktims d9d55b069f day10: optimization 
Part 2 doesn't actually require tracking state at all, it's enough to
just follow the rules and count every DFS that lands on a target square.

Part 1 optimized by only tracking which targets have been previously
visited, instead of all squares.

Both parts, pass position reference to recursive calls instead of by
value for tiny improvement.
 2024-12-10 16:20:15 -08:00
ktims 0716dde8b1 day10: complete solution 2024-12-09 21:59:39 -08:00
ktims 206c1fca85 day9: optimize part 2 2024-12-09 00:23:40 -08:00
ktims f7cf4f1e9f day9: optimize part 1 2024-12-08 23:58:47 -08:00
ktims b08c8fbd80 day9: use part 2 'inodes' table to speed up part 1 2024-12-08 23:51:51 -08:00
ktims 2bc751dd0d day9: complete solution 
previous commit should actually work for part 1 if you don't truncate
the input.
 2024-12-08 23:34:15 -08:00
ktims a6ea5b4155 day9: second rewrite, still failing 2024-12-08 22:57:20 -08:00
ktims 622877843e day8: refactor and cleanup 2024-12-07 22:49:57 -08:00
ktims 47e40942e2 day8: complete solution 2024-12-07 21:43:56 -08:00
ktims 3e0bc0d5cc day7: parallel solution, < 100ms 
This is definitely a problem that calls for DFS, but I couldn't be
bothered.
 2024-12-06 22:52:00 -08:00
ktims 868c3e56fc day7: complete solution 2024-12-06 21:40:25 -08:00
ktims fee37aebd0 day6: further performance improvement, ~55ms total runtime 
We only need to run the guard forward from the step before the obstacle
we placed, instead of starting at the original starting position.

The code starts to look ugly...
 2024-12-06 17:05:36 -08:00
ktims b1918bbebf day6: performance - only check obstacles on the original path 
700ms -> 150ms
 2024-12-06 10:59:43 -08:00
ktims 1cd535c2aa chore: clippies 2024-12-06 10:52:31 -08:00
ktims 1dd3ce5862 day6: performance improvement 2024-12-06 10:40:28 -08:00
ktims 145d779e83 day6: complete solution + grid lib (used in solution) 2024-12-05 22:41:47 -08:00
ktims 49c37800a0 day5: complete solution 2024-12-04 22:41:33 -08:00
ktims e2cd0fe9cc day3: code format 2024-12-04 16:41:39 -08:00
ktims 6a619d756f gitignore 2024-12-04 16:41:16 -08:00
ktims dc67b38420 day3: disable unicode matching, don't recompile RE for every iteration 2024-12-04 16:41:00 -08:00
ktims e5faee3470 day4: do a lot fewer allocations, performance improvement x10 2024-12-03 22:09:13 -08:00
ktims a81904f2eb day4: complete solution 2024-12-03 22:02:49 -08:00
ktims ec55b07c61 boilerplate: fix it 2024-12-02 21:22:05 -08:00
ktims 7a7fb4d02a day3: full solution 2024-12-02 21:18:37 -08:00
ktims b24593a469 day2: performance and adopt new duration print 2024-12-01 23:02:00 -08:00
ktims ec2651c3e4 boilerplate: improve duration printing 2024-12-01 21:32:17 -08:00
ktims 33c0c876c3 day2: full solution 2024-12-01 21:31:29 -08:00
ktims 968d795720 day 1: cleanup and refactor 
* I misunderstood the problem and thought we were asked for the distance
  between the numbers' positions, not their values. Removed remnants of
  that.
* Move to struct-oriented solution
* Fix clippys and unnecessary code
 2024-11-30 22:03:19 -08:00
ktims 8cab780ff3 day 1: problem 2 solution 2024-11-30 21:46:10 -08:00
ktims 892d59f21b day 1: problem 1 solution 2024-11-30 21:38:31 -08:00
ktims 76cacf4704 gitignore 2024-11-30 21:37:02 -08:00
ktims fd24918c7a 2024 AoC is here! 2024-11-30 21:36:05 -08:00