100 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
ktims	6efb9e0f83	day23: add pivot to bron-kerbosch, use more hashsets for speed	2024-12-23 02:07:49 -08:00
ktims	237ca36381	day23: clippies	2024-12-23 00:57:52 -08:00
ktims	d3bade1bdd	day23: part 2 - bron kerbosch solution	2024-12-23 00:54:23 -08:00
ktims	50a197856a	day23: part 2 working (i think) but too slow	2024-12-23 00:23:36 -08:00
ktims	d3ce12693b	day23: part 1 solution	2024-12-22 21:48:59 -08:00
ktims	d31f9725f5	day22: perf and clippies, down to < 1s	2024-12-22 02:21:24 -08:00
ktims	c5857ed449	day22: improvements	2024-12-21 23:49:07 -08:00
ktims	40d5e820bc	grid: additional impls of Coord2d	2024-12-21 23:49:06 -08:00
ktims	5d518248a8	day20: clippies and performance	2024-12-21 23:49:06 -08:00
ktims	cebdbd1007	day20: part 2 solution (slow)	2024-12-21 23:49:05 -08:00
ktims	32937aaa0e	day22: part 2 solution. 38s runtime but works.	2024-12-21 23:49:04 -08:00
ktims	c681727fb3	day22: part 1 solution	2024-12-21 23:46:08 -08:00
ktims	02fc154547	grid: use coord2d more consistently	2024-12-21 21:01:16 -08:00
ktims	13f61e3a2b	part2: works up to the 4th iteration, then returns too high values ... I think the issue is not testing both possible paths for direction keypads, but doing this naively blows up the compute complexity (and I got the wrong answer)	2024-12-21 02:48:39 -08:00
ktims	8958ba9361	day21: part 1 solution	2024-12-20 23:20:21 -08:00
ktims	e60f6effa3	day20: part 1 solution	2024-12-20 20:56:14 -08:00
ktims	c2e3422544	day19: rayon - ez mode parallelization, speed * 6	2024-12-18 23:09:51 -08:00
ktims	6b6dededc2	day19: clippies	2024-12-18 22:45:10 -08:00
ktims	65d498f168	day19: complete solution	2024-12-18 22:39:32 -08:00
ktims	cdb3a7261a	day19: pre puzzle boilerplate	2024-12-18 20:56:36 -08:00
ktims	dbee7d91b6	day18: clippies	2024-12-18 15:42:23 -08:00
ktims	0fee7c3594	ci: report errors correctly	2024-12-18 15:36:02 -08:00
ktims	824d111b18	ci: don't use '-D warnings' except for clippy run	2024-12-18 15:34:15 -08:00
ktims	a1dceb6ff1	day18: perf: swap dijkstra for bfs for ~30% improvement	2024-12-18 15:31:12 -08:00
ktims	3712b32634	grid: row and column iterators	2024-12-18 15:31:07 -08:00
ktims	2a11e17d92	day18: improve brute for for binary search for big gainz ... also avoid unnecessary path tracking work	2024-12-18 11:35:22 -08:00
ktims	1c254fff93	Revert "cargo: optimize harder" ... This reverts commit 44108c4b35.	2024-12-18 00:53:51 -08:00
ktims	44108c4b35	cargo: optimize harder	2024-12-18 00:41:40 -08:00
ktims	b588837624	day18: small optimization - avoid duplicates ... not sure if the input contains dupes, but avoid doing a second path search if the square already contains a byte, seems like a small performance gain	2024-12-18 00:41:28 -08:00
ktims	9a6ca66059	grid: microoptimizations	2024-12-17 23:52:14 -08:00
ktims	2729799fa2	day18: optimize ... * Implement generic path tracing, so part 1 doesn't store and return the full path, just the count. * Improve part2 impl to only run path search if the block fell on our last path. * Start part2 path search from the best path position immediately before the placed block.	2024-12-17 23:51:09 -08:00
ktims	8d2fbc0fcb	repo: more aggressive optimization	2024-12-17 23:41:48 -08:00
ktims	4f48d839b2	day18: part 2 solution + some optimizations	2024-12-17 22:13:49 -08:00
ktims	414569537e	day18: part 2 solution	2024-12-17 22:00:10 -08:00
ktims	5036866663	day18: part 1 solution	2024-12-17 21:52:41 -08:00
ktims	9be86e2cc2	day17: clippies	2024-12-17 01:03:06 -08:00
ktims	5bcead2691	day17: part 2 solution	2024-12-17 00:46:18 -08:00
ktims	c99d8a400a	day17: part 1 solution	2024-12-16 22:29:55 -08:00
ktims	5e8b974137	clippies!	2024-12-16 14:54:48 -08:00
ktims	28a88e1aa7	day14: replace RE parser with nom consuming parser ... Almost 100x speedup on parsing phase avoiding RE compile, saves ~60ms	2024-12-16 14:47:48 -08:00
ktims	33615b015f	cargo: update and add nom, misc	2024-12-16 14:44:30 -08:00
ktims	b7a1f05b1e	utils: add misc and implement CustomBounded	2024-12-16 14:41:32 -08:00
ktims	de7ee8f0f6	day14: don't print robots map	2024-12-16 11:40:10 -08:00
ktims	c31d653612	day16: minor performance and refactoring	2024-12-16 01:33:02 -08:00
ktims	20e6889572	day16: improve perf again by going to i16 positions ... map dimensions fit in i16, so use those in data structures make grid support generic position type as long as they implement to i64	2024-12-16 01:06:39 -08:00
ktims	755fbbc53d	day16: refactor, optimize ... split path recording and best cost functions for big gainz use i32 instead of i64 positions to shrink data structures for some gainz	2024-12-16 00:58:15 -08:00
ktims	4b85a90635	day16: part 2 solution	2024-12-15 23:22:14 -08:00
ktims	2cf8527c4f	grid: add add_mut()	2024-12-15 23:21:59 -08:00
ktims	6283ff37f9	day16: part 1 solution	2024-12-15 22:06:21 -08:00
ktims	f2186d18d3	ci: allow test failures, fix input getting script mistake	2024-12-15 01:06:09 -08:00
ktims	c261ee56fe	day15: cleanup, clippies	2024-12-15 01:03:08 -08:00
ktims	411d6aa26d	day15: cleanup & use the same implementation for both parts	2024-12-15 00:56:07 -08:00
ktims	c2c0145219	day15: part 2 submitted solution	2024-12-15 00:50:19 -08:00
ktims	4dfdaca58c	day15: part 1 solution + beginning of part 2	2024-12-15 00:49:44 -08:00
ktims	003bc3212d	pre-commit: check clippies and format in pre-commit to avoid CI fail	2024-12-13 23:51:50 -08:00
ktims	b060de20c7	day11: clippies	2024-12-13 23:44:46 -08:00
ktims	bc7ec50c94	CI: get real inputs, and do a full run	2024-12-13 23:44:29 -08:00
ktims	8ae2115b52	CI: create fake inputs so tests can run	2024-12-13 23:06:23 -08:00
ktims	74a6b16924	chore: rust fmt	2024-12-13 23:06:07 -08:00
ktims	e8a38e7b24	day14: pretty print a tree	2024-12-13 23:05:58 -08:00
ktims	c6153663b5	day14: complete solution	2024-12-13 22:53:07 -08:00
ktims	8b011941c4	CI: make a workspace	2024-12-13 18:09:22 -08:00
ktims	d88f907c03	clippies	2024-12-13 17:55:28 -08:00
ktims	d6d81a0c29	CI: run tests	2024-12-13 17:52:42 -08:00
ktims	ed184fc92c	day12/day13: codspeed compat (&str input)	2024-12-13 17:42:44 -08:00
ktims	a1774d1f73	grid: impl FromStr for Grid<u8>	2024-12-13 17:28:45 -08:00
ktims	3bbf05b30c	pre-commit	2024-12-13 02:24:18 -08:00
ktims	6a8a7a9ad1	pre-commit update	2024-12-13 02:22:55 -08:00
ktims	1a6d37f4f3	clippies	2024-12-13 02:17:57 -08:00
ktims	35637cece1	readme	2024-12-13 02:15:41 -08:00
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