Advent of Code 2025 Day 7 (BaseA Style)

[AlteryxCommunityTeam]

Discussion thread for day 7 of the Advent of Code - https://adventofcode.com/2025/day/7

[JeffF]

Spoiler

[DataNath]

Day 7 done! Originally tried to find an elegant solution to part 1 which I have a feeling can be done with some Generate Rows + Multi-Row Formula magic, but soon gave up on that and decided to just get it solved, resulting in the iterative approach as is usual for these grids. Part 2 took me a while to figure out how I could track all possibilities, then I remembered previous years where... (spoiler)

Spoiler

... you can just keep track of the sum of routes that get you to a certain point i.e. how many possibilities of left/right. So we just need to take the total of these at the bottom of the grid.

[mmontgomery]

Day 7

Spoiler

P1 was rather straight forward, per usual. Just do a count records each time a split was hit while making sure to create new beacons and add a new row
P2 took me a real long time. Eventually leaned on @DaisukeTsuchiya 's approach to sum the number of unique paths that made it to the final row

[Erin]

I understood Part 1 and was able to recreate the last picture in the example. I'm a visual learner. 

But, I guess Day 7 part 2 is the end of my AoC journey. I worked on it for an embarrassingly long time without success.  I think I might need a private tutor, a DeLorean and/or a friend named "Doc". 

Good luck to those who are continuing on! 

[Stosh]

Day 7 was insane! That being said, I was able to solve this in pretty good time and we are back to needing no assistance! I am excited to see what comes of the rest of the second half of AOC.

My Solutions:

Spoiler

Part 1 and 2: 

The macro (works for both parts!)

My Reflection:

Spoiler

This problem was one that freaked me out at first. It just is so far gone from what Alteryx was made to do that I honestly didn't know where to start, besides knowing I would have to do the classic RecordID -> RegEx -> RecordID that I have become so familiar with this past week. That being said, for Part 1 I was pretty quickly able to come up with a routine that I could repeat in an iterative macro. The plan was simple: Examine two rows at a time, sliding down one row per iteration. For each slice of two rows, I would first go top to bottom, continuing uninterrupted beams and identifying ^'s that result in splits. Then, I would go left to right, adding new beams to the left and right of splits. Once the iterations are complete, all I had to do was count my identified splits. This worked like a charm, and I was actually correct on my first attempt!

Part 2 was a test of my mental patience, as I really really just wanted to pack it up and call it a day😂I tried a couple of gimmicks (Splits * 2, Splits ^ 2) before heading back to the description to look for inspiration. While I was literally counting out all the possibilities in the example input one by one, I realized that the key is just to count how many beams are in a certain position, as opposed to treating cells as binary (beam or no beam) and then take the last row for each column and sum up the counts of each beam. My macro from part 1 was already primed to handle this logic switch, so I decided to just upgrade it as opposed to starting from scratch. Instead of |'s, I started with 1 and used digits to track beams, adding up joining beams as I traversed down the list. Once I completed the logic switch, It was just about grabbing the final row for each column, removing columns that didn't have beams, and summing instead of counting!

Overall, this was probably my favorite challenge so far! It was a super tough problem at first, but with some clever moves and key discoveries, it ended up being one of my most elegant solutions. See you all tomorrow!

[AncientPandaman]

wonder why most ppl is generate row is use instead of 2 formula tool. 

for me it give direct instinct to split to 2. where it hard to tell by generate row. 

Spoiler

remove "." to speed up.
use 2 formula tool to indicate split to 2.
check last row of "^" to identify whether to stop iteration

add count for part 2.
where it classic method to solve these question.
workflow:

macro:

[DanFlint]

Part 2 defeated me today, so I'll have to take it away to figure out what I'm missing. 

Part 1 solution:

Spoiler

Ended up being much simpler when I realised I didn't need to draw out the matrix first.

[martinson]

Spoiler

Part 1 Macro

Spoiler

Part 2 Macro

Spoiler

[lwolfie]

Pt 2 took a while.

[danboll_life]

It took a long time to figure out Part 2.

Spoiler

Main

Macro