Advent of Code 2024 Day 2 (BaseA Style)

MeganBowers

Discussion thread for day 2 of the Advent of Code - https://adventofcode.com/2024/day/2

Advent of Code

Accepted answers

All comments

Gawa

It's time to use Macro but not Iterative one yet.

Spoiler

Part1: Put flag as per this expression, and concat flag for each report.

We need to know the number of records which is '+++++' or '-----' so apply Regex_Match() function, and count the True record.
*we need to escape meta character like '+' by '\'

Part2: Basically the same logic as Part1 was applied. Difficult point was how to prepare all pattern that miss just one level.
I go with Batch Macro; 1st batch=remove 1st level, 2nd batch=remove 2 level....
Of course, the outcome contains duplicated Report so count distinct by Summarize tool.

2024D2_gawa.yxzp

geoff_zath

Messy solution without a macro!

Spoiler

AoC_2024_day2.yxzp

CoG

Updated x2 (Tool Golf/Optimized version V4):

Spoiler

To accomplish a 6 Tool solution, the Generate Rows Tool is quite the monster, utilizing a 2D matrix loop to accomplish what the Text to Columns/RegEx Tool would have. Not quite as elegant, but a stroke is a stroke. A very elegant 7 tool solution can be achieved with very reasonable formulae.

Updated (Tool Golf/Optimized version V3 - elegant)

Spoiler

I am quite pleased with the elegance and simplicity of this solution.

Main V3 Formula.png

Updated (Tool Golf/Optimized version V2):

Spoiler

Achieves both Part 1 and Part 2 using 9 tools (per part). Multi Row Formula Tool requires some extra logic to make this work:
Main V2 Formula.png

(My Original Solution) No fancy macros from me yet:

Spoiler

For part 2, I joined the output of part 1 with main data to identify "Unsafe" reports that could benefit from the dampener, then I applied the exact same logic as Part 1 after dampening (Generate rows to brute force)

Main V2.yxmd

Main.yxmd

Main V3 (Elegant 7 Tool).yxmd

Main V4 (6 Tool).yxmd

ScottLewis

Day 2 was a nice case of ugly solutions that work. Considered a batch macro but chose to brute force it rather than fiddle with inputs.

The trick is

Spoiler

Creating all of the possible iterations of set - 1 row. You can do this for the maximum length set (8 in my case) since you're going to have to sum away duplicates anyway.
My version had a small hiccup in that I was throwing out row 1 (which is always safe but has an extra case on the change calculation.
In Part 2 that is only right if you aren't working on the iteration that doesn't have row 1, so you need to do Min(Row) instead of Row1.

Day2Scott.yxmd

Hub119

Why sleep when you can earn AoC stars after midnight? 😀

Spoiler

AoC 2024 Day 2.yxzp

AkimasaKajitani

My solution!

Is today really the second day?

Spoiler

Day02.yxzp

kelsey_kincaid

This one definitely challenged me more than I expected for Day 2. The concepts felt straightforward, but my execution was... not. I found myself reworking the logic several times. I have the feeling there's a more straightforward approach, but my tired brain couldn't find it today. Hoping to revisit this one!

Spoiler

One of my gotcha's was the way I handled the change in value for the first value in the multi-row formula. I also mishandled situations where the change was neither an increase or a decrease. Lots of little pieces of logic to think through in this one!

2024_Day2_KelseyKincaid.yxzp

kelsey_kincaid

@gawa Your solution is so elegant! Thank you for sharing

DaisukeTsuchiya

Spoiler

DT2024_2_R1.yxzp

AncientPandaman

no macro. let see how many days challenge can solve without macro

Spoiler

cgoodman

I went down the rabbit hole with part 2!

Spoiler

Does it count not using a macro when you hard code a batch 😜

Screenshot 2024-12-02 063349.png
Ignoring the bad res.

Part 1 was simple, but I got caught out where the total number of levels could be different for each record in the real data.
Part 2, I spend so long trying something based on what I had done in part one of identifying if each row was an increase or decrease on the previous, going with complex multirows. For example IF Decrease but previous row was an increase, change to increase if the next row - previous row is actually an increase <=3... But my brain wasn't working.
In the end went down the batch macro route, but also couldn't work out if I need a batch or iterative, so just hardcoded the logic of skipping layers and calculating if all increasing or all decreasing.

LHolmes

Well I feel like the heat turned up a little bit... 😅

Spoiler

2024 Day 2 Part 1

2024 Day 2 Part 2

2024 Advent of Code Day 2 _ Solution LMH.yxzp

Tokimatsu

I solved it, but it is an uncool workflow.

Spoiler

AoC2024_Day02.yxzp

Gawa

Tool Golf for Part2 without Macro. Another fun of AoC is refining WF which I believed it was the best at that time.

Spoiler

Trick here is to create all patterns that miss one level by combo of the Generate Row tool and the Regex_Replace function.
By doing so, the logic is almost same as part1. Difference is only to exclude duplication of report.

Overall WF for Part2

2024D2_Rev1_ToolGolf_gawa.yxzp

rbgruwel

Used an iterative macro, does not yet deserve the beauty prize. But it works.

Spoiler

Day 2.yxmd

danboll_life

lacks versatility

Spoiler

Continue Filter...

AoC_2024_02_DL.yxzp

cgoodman

With a bit of thinking time I've come up with my attempt at tool golf.

Spoiler

In the bottom half I create a look up table for all substitution sequences.

e.g.
1,2,3,4,5
1,3,4,5
1,2,4,5
1,2,3,5
1,2,3,4

Then transpose this like with the initial dataset, then with a join I get all possible outcomes so only need to evaluate once. Then it's a case of filtering only for the first pattern for part 1 and including all for part 2.

Screenshot 2024-12-02 093010.png

DataNath

Certainly won't be winning any awards for elegance but work calls and 2 stars = 2 stars. Definitely lured into a false sense of security yesterday!

Spoiler

Macro is just a copy and paste of the main workflow, removing one record at a time from sequences that failed the initial part 1 ask and checking for compliance against the new allowance.

Spainey

Took a lot longer than yesterday, but I got there after I gave up refusing to use a macro on Day 2.

Spoiler

My workflow:
AoC 2024 Day 2.png

I used the RegEx tool to Tokenize the data to Rows (which initially tripped me up on Part 1 as I assumed that all numbers in the sequence were a single digit). An Order column was added so I could easily identify and remove the first number in the sequence and could determine the State (increasing/decreasing) of the other numbers. I concatenated the State and then used a formula to remove duplicate strings in order to then identify those that were deemed SAFE.

The formula below identifies and removes duplicate strings in a comma separated string array.

IF !IsEmpty([Concat_State]) && Contains([Concat_State], ",")
THEN REGEX_Replace([Concat_State], '(\b[^,]+)(?=.*, *\1(?:,|$)), *', '')
ELSE [Concat_State]
ENDIF&#8203;

For Part 2 - I wrapped the above logic into an iterative macro, where I removed the row corresponding to the current iteration (+1 as it is 0-based index) and redefined the Order column.

and the macro:
AoC 2024 Day 2 - Macro.png

2024_AoC_mark-spain_Day_02.yxzp

Samantha_Jayne

Part 1 done. Working on Part 2. Boy that's a corker!

Spoiler

Star 2... coming soon...

Kenda

Spoiler

copy and paste method to get this one done lol if there were more numbers in each row, I would have definitely found a different route

AoC 2024 D2.yxmd

Sebastian_Chaieb2

My Solution is a little bit chaotic and I used Batch macro for Start 2

Spoiler

Spoiler

no shame in copy/paste with AoC imo

Spoiler

Day 2.yxmd

mmontgomery

Day2

Spoiler

Had a bug in P1 that took too long to figure out and ended up doing a batch approach for p2

Day 2_MM.yxzp

Yoshiro_Fujimori

My solution.

Spoiler

Part 1

Part 2

Day 2. Struggled without the batch macro for part2 but finally got it.

Spoiler

Day2.yxzp

AkimasaKajitani

I have gotten rid of no use tools and I have reorganized my messy workflow. As a result, tools are decreased from 33 to 18.

Spoiler

My 1st workflow.

My reorganized workflow

Day02_1.2.yxzp

se_alex

grossal

I was missing a true tool golf solution ... so I had to create one. Here it is:

Spoiler

2024-12-02_15-42-47_AlteryxGui_Alteryx_Designer_x64_-_New_Workflow2.png

Formulas are a little too short and the bar is a little to high for Obscura, but they look nice anyway.

JS1989

Solution

AOC_Day2.yxmd

Raj

part 1
kept it simple and working.

AOC_2024_2_1.yxmd

SeanAdams

Your V2 is pretty slick @CoG - nicely done!

(I split the asc and desc - but with your carrying over the direction flag you can avoid this)

CailinS

Thanks to @gawa for your solution. While I took a different approach, I used yours to check my input and determine which specific records were causing me to not get the correct count. I found my issue and am so grateful to have final solved part 2!

It is messy but my solution is attached. Here are screenshots too:

Spoiler

Day 2.yxzp

Quick Links

This months top contributors

atcodedog05 19458

Qiu 15866

binu_acs 15708

MarqueeCrew 13708

apathetichell 13703