Advent of Code Progress Tracker

This repository tracks my progress as I solve puzzles from Advent of Code from 2015 to 2024 and beyond. Each puzzle has two parts, and I'll mark them as ✅ when solved or ❌ otherwise. I'll also add notes for interesting learnings or challenges faced during the process.

---

Advent of Code 2024

Day	Part 1	Part 2	Notes
1	✅	✅	Lists and counters for the win!
2	✅	✅	Lists and np.diff operations
3	✅	✅	re.finditer() and re.findall() worked like magic. Could the part2 code be made shorter? Perhaps a task for the future.
4	✅	✅	Search for substring (XMAS) in a grid in all directions. Boundary check for indices.
5	✅	✅	Custom iterative sorting mechanism was needed. Topological sorting failed due to cycles and exhaustive permutation for part 2 failed due to scalability issues. What a way to learn how sorting operations work 😅!
6	✅	✅	Solving bouncing-board-snake-style game on the grid. Part 2 took 50 seconds to brute-force through all alternatives (perhaps there are ways to avoid duplicate searches?). Key lessons: using pretty_print_grids() (in part2.py) aided debugging, and storing the grid as a list of character lists proved more efficient.
7	✅	✅	itertools.product() did the magic of generating all outcomes from the toss of the operator "coin" being two numbers. Took 14.5 seconds for part2, which could be improved!
8	✅	✅	Goal was to find linear alignments of points in a grid. Logic--first collect the coordinates of each unique character. Then use itertools.combinations and coordinate geometry to generate line segment points and extends them in both directions until the grid boundary is reached. Fun puzzle!
9	✅	✅	Goal was to remove white spaces by pushing file IDs from the end to the front. Somewhat long logic using careful edits to list start and end indices.
10	✅	✅	Goal was to count the number of paths in a topologically sorted network with a set of start and end points. I struggled with choosing the right graph algorithm till realizing the network has a topological order.
11	✅	✅	Goal: Apply a defined rule to manipulate numbered stones for a set number of "blinks" (iterations). For 25 blinks, simple recursion worked well; for 75 blinks, using lru_cache was more efficient. The key insight is that we only need the final count of stones, not their specific values.
12	✅	✅	Part 1 solved using networkx.connected_components(). Part 2 counting of sides was done using the logic by couting all top, bottom, left, and right, edges of a region.
13	✅	✅	Use Cramer's rule to solve two linear equations. A fun one!
14	✅	✅	Part 1 was a recursive update to x and y coordinates. Part 2 felt like brute force to find the tree-looking easter egg. There is fortunately a pattern where x and y coordinates update independently and repeat in vertical and horizontal grouped patterns after every max_col and max_row timesteps.
15	✅	❌	Part 1 was a reasonable move logic. Part 2 is involved. It seems like a BFS at every move to determine which blocks will move and then updating it. I wonder if networkx can solve it? A task for the future self.
16	✅	✅	I used the logic of converting the grid into a networkx DiGraph, and then converting to another line graph (using nx.line_graph()) where weights of edges were carefully designed to find the shortest path minimizing weighted sum of number of turns and number of nodes. For part 2, it was a simple nx.all_shortest_path() function and counting the common nodes across all of them.
17	✅	❌	Part 1 was simple if-else. However, I didn't expect output to be reported as a string and wasted a lot of time debugging. Part 2 seems harder; what would be a smart logic that builds on the repetition pattern?
18	✅	✅	networkx shortest path finder made it simple. In particular, I used nx.shortest_path_length(). Part 2 took approximately 6 seconds.
19	✅	✅	Recursive search for a pattern of string in a given design to find whether the design is possible (part1) and the number of ways to make the design possible (part2). Use lru_cache for memoization.
20	❌	❌
21	❌	❌
22	✅	✅	Simple implementation of functions and brute force the logic. Walrus operator (:=) helped and so did np.diff().
23	✅	✅	Magically easy with networkx clique functions. I used enumerate_all_cliques(G) in forward order for the first, and reversed order for the second.
24	✅	❌	Part 1 was easily rule checking. Part 2 sounds some bit adder concept, a task for my future self to understand and solve.
25	✅	❌	Part 1 was character counting. Part 2 cannot be solved until I have solved all the remaining puzzles. Well!

---

Advent of Code 2023

I have solved some of these in a different notebook, but yet to include the attempts on Github.

Day	Part 1	Part 2	Notes
1	❌	❌
2	❌	❌
3	❌	❌
4	❌	❌
5	❌	❌
6	❌	❌
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Advent of Code 2022

Day	Part 1	Part 2	Notes
1	✅	✅	Addition over lists.
2	✅	✅	Rock-paper-scissor!
3	✅	✅	Learned about ord() string function.
4	✅	✅	Finding overlap of two lists
5	❌	❌
6	❌	❌
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Advent of Code 2021

Day	Part 1	Part 2	Notes
1	✅	✅	Take difference of lists and count the positives. Use numpy though.
2	✅	✅	Simple logic after input splitting.
3	✅	✅	(1) Good use of counters, zip to swap columns as rows, reversed, and boolean logic, (2) Good use of functions with changing list copies.
4	✅	✅	Bingo analysis
5	✅	✅	Draw lines and their meeting points
6	✅	✅	Recursion and LRU cache for the win! 🎉
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Advent of Code 2020

Day	Part 1	Part 2	Notes
1	✅	✅	For now a blind nested loops to a tuple of numbers T such that: T \subseteq numbers,mod(T) = k, and sum(T) = d though recursion can work.
2	✅	✅	One line input splitting; Counter and xor operations.
3	✅	✅	Cycle through column indices using itertools.
4	❌	❌
5	❌	❌
6	❌	❌
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Advent of Code 2019

Day	Part 1	Part 2	Notes
1	✅	✅	Recursive function with a twist!
2	✅	✅	List indexing carefully executed. There might be a smarter approach for part 2 than brute-force checking all nouns and verbs, but oh well. It took less than a second on my machine!
3	❌	❌
4	❌	❌
5	❌	❌
6	❌	❌
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Advent of Code 2018

Day	Part 1	Part 2	Notes
1	✅	✅	First part was a simple sum. However, itertools saved the logic for part 2. Probably there might be more efficient ways!
2	❌	❌
3	❌	❌
4	❌	❌
5	❌	❌
6	❌	❌
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Advent of Code 2017

Day	Part 1	Part 2	Notes
1	✅	✅	List indexing magic. Could there be a better logic for cyclicity?
2	✅	✅	List combinatorics. Helpful tools: itertools combination and re.findall()
3	❌	❌
4	❌	❌
5	❌	❌
6	❌	❌
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Advent of Code 2016

Day	Part 1	Part 2	Notes
1	✅	❌	Found a way to track coordinates. Still pondering how to estimate the crossing point!
2	❌	❌
3	❌	❌
4	❌	❌
5	❌	❌
6	❌	❌
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Advent of Code 2015

Day	Part 1	Part 2	Notes
1	✅	✅	Character loops for the win.
2	✅	✅	Lambda functions and Regex help simplify.
3	❌	❌
4	❌	❌
5	❌	❌
6	❌	❌
7	❌	❌
8	❌	❌
9	❌	❌
10	❌	❌
11	❌	❌
12	❌	❌
13	❌	❌
14	❌	❌
15	❌	❌
16	❌	❌
17	❌	❌
18	❌	❌
19	❌	❌
20	❌	❌
21	❌	❌
22	❌	❌
23	❌	❌
24	❌	❌
25	❌	❌

---

Notes

• ✅: Puzzle solved successfully.
• ❌: Puzzle not solved yet.
• Notes will include interesting learnings, challenges, or algorithms explored during each day's puzzle.

Stay tuned for updates as I continue solving puzzles year by year!