Advent of Code 2024 Day 8 (BaseA Style)

Discussion thread for day 8 of the Advent of Code - https://adventofcode.com/2024/day/8

Accepted answers

All comments

Eric is treating us right for the weekend! Very refreshing (Scored top 1000 for both parts today). Reading the problem felt harder than solving it:

Spoiler

Brute force all the way. Small map size makes this problem trivial to solve!

Main.yxmd

ScottLewis

A second straightforward day in a row. Probably too much to hope for one more in advance of Monday Morning Meetings but we'll find out in 24 hours.

This marks the only time I've ever been happy that joins are case sensitive.

Spoiler

This is pretty much an intermediate data tricks lesson. Filter out junk, join with yourself to generate all possible combinations, throw out self joins, calculate the two antinodes from the two points (or the 50 possible for part 2) for each pair, unique the thing and count the rows. Bing. I missed that the possible count is N, rather than N/2 for part 2, which kept me out of the rarified triple digit air but otherwise this is a very good one for Alteryx.

Day 8 Scott.yxmd

Gawa

Difficult part was to understand what we are asked to do rather than configuring WF today.

I might have violated Base-A...I used a pencil and paper. I love @CoG 's neat and wise WF. It's elegant.

Spoiler

Part1: Create all pair of the same antennas, and calc X-Y of antinodes at opposite side. Once understanding the concept, it's not difficult.
Part2: I used linear algebra; Y = aX + b. By using intermediate result of part1, we can find two parameters 'a' and 'b' easily by the Formula tool. Cross join all of lines and points, and check points are on it. As 'a' and/or 'b' can be decimal, result of expression 'Y - (aX + b)' is not always exactly equal to 0 (0.0001234 etc). I applied fuzzy threshold like ABS([Y]-([a]*[X]+[b])) < 0.01, and it worked well in this case.

(Edit) For part2, if linear algebra method is applied, it's successful because of:

I was just luck to have solved it without knowing this nature of data.

Spoiler

Span-X and Span-Y is relatively prime(互いに素) for any pair of antenna. Line between two antenna does not intersect any points within that range.
That means the extended line outside of two antenna will Intersect points where they locate at multiple of Span-X/Y, that are we are asked to find.

2024D8_gawa.yxzp

DataNath

Wow - didn't expect this after some of the challenges we've had this week! Did almost have an absolute meltdown trying to parse the question originally though. Finished ~5k for both parts on the global leaderboard which is pretty sweet. Went overboard generating rows for P2 but hey ho, covering bases = stars apparently.

Spoiler

kelsey_kincaid

I'm sure someday I'll feel confident with grid problems, but I won't get there without practice! I spent a decent amount of time trying a spatial solution, but I didn't get very far with that.

I'm 100% sure I overcomplicated this one, but I was happy that Part 2 didn't take toooo long to get to from Part 1.

Spoiler

I nested iterative macros within a batch macro, which is why I'm certain I overcomplicated it, but it still runs quickly which is nice.

Batch Macro:
All this is doing is filtering down to antennae that have the same transmission value. I used REGEX_Match with case sensitivity to ensure the difference between 'a' and 'A' would be captured here.

Iterative Macro(s):
My part 1 and part 2 macros are identical except for the Generate Rows tool & a small tweak to the formulas.
I take each row of data and join it with every other row of data that has the same transmission value. Then it computes the location of each antinode.

The generate rows tool for Part 2 provides a multiplier. This lets me compute all of the points along the line that could possibly be within the specified size of grid.

EDIT: Seeing the other (much simpler) solutions is a good lesson for me. Sometimes my mind jumps straight to macros for AoC and it doesn't need to!

2024_Day8_Alteryx_KelseyKincaid.yxzp

Yoshiro_Fujimori

For Part 2, at first I tried "Generate Rows" or "Append Fields".

As it did not seem to work well, I switched to Iterative Macro and it worked.

Spoiler

Workflow

Macro for Part2

workflow.yxzp

DaisukeTsuchiya

Done. I can sleep tonight.

Spoiler

DT2024_Day8_Rev0.yxmd

AkimasaKajitani

Solved!

Not elegant, but work!

Spoiler

I was able to use the Arrange tool to stack multiple x,y elements vertically!

Day08.yxzp

mmontgomery

Day8. More in spoiler

Spoiler

P1: Spent most of the time drawing the relationships by hand and then creating the formulas around the hand calcs to make it easier to see
P2: Take P1 framework but apply it to generate rows while setting boundaries relative to the map size

Day 8_MM.yxzp

Hub119

Either Day 8 was just WAY easier (once you decipher what is being asked) than the last few problems OR I am quickly acknowledging the benefits of doing these in the middle of the day on a good night's rest 😀.

Spoiler

So nice when solutions can run in under a second...

AoC 2024 Day 8.yxzp

leozhang2work

Spoiler

Almost went for macro again, turns out no need, just need a correct join does the job.

day 8.yxmd

geoff_zath

Workflow

Part 1

Spoiler

Part 2

Spoiler

Batch Macro

Spoiler

AoC_2024_day8.yxzp

AoC_2024_day8_part2_v2.yxzp

peter_gb

Struggle with the logic for this at first but it wasn't so horrendous in the end! Very glad to get the second star. Solved the entire day 8 whilst my day 9 part two was running...!

Spoiler

AoC_8.yxmd

Tokimatsu

Done without macro.

Spoiler

AoC2024_Day08.yxzp

ntakeda

My solution.

Spoiler

AoC2024-part8.yxmd

AncientPandaman

not sure why it is working. but it worked 😆

Spoiler

use generate rows to have multiple to guess the limit, (about 50 for my input, lazy to put in macro to dynamic the multiple)

Screenshot 2024-12-10 145215.png

Day 08.yxzp

ccanterman

Solved! I went with brute force for Part 2 since I needed that square for the bingo game

Spoiler

For Part 2, I tried to remember my linear algebra and come up with a proper line formula but after playing with pen and paper without luck, I said "screw it" and just generated all 50 x 50 points and saw which ones were on the same line as my antennae pairs.

Day 8.yxmd

aiahwieder

Whee!

Resonant Collinearity.yxzp

Spainey

I originally skipped this one as I didn't have enough time at the weekend, but came back to it today and was able to figure out how to approach the question.

Spoiler

The first part was quite simple, only requiring 1 antinode each at the positive and negative end of the antenna pairing.
For Part 2, I used Generate rows within the bounds of 1 and max_row to ensure that new antinodes were generated within a restricted range (which was filtered further to remove erroneous antinodes once column positions were calculated also). Wasn't entirely sure it would work so was pleasantly surprised when the answer was correct