Skip to main content

C'est la Z

Transparent Origami - Advent 2021 Day 13

No post so far on day 12. I finished part 1 but my code was pretty messy which turned part 2 into a mess. I still have to go back to get that second star so just like day 10, my day 12 write up is on hold.

That said, I really enjoyed [[https://adventofcode.com/2021/day/13 ][today's]] challenge. Paper foldind. You're given transparent paper with a bunch of marks on it. You have to fold the paper over horizontal or vertical lines and examine the results.

Input was in the form of lines x,y pairs followed by lines specifying the folds.

First decision - data representation. Since we've got a grid, temptation would be to use a 2D structure like a 2D array but there are two problems. One is that each fold will make the paper smaller and smaller which makes a 2D array wasteful if not clumsy. The other issue is that we don't know the ranges for our coordiates. We know that all the values will be positive but we have no idea where they start and end.

Just like with the cellular automaton, I decided to use a dictionary where the key was the x,y pair and the entry was the mark. I could have used a plain list for the points but the dictionary gave me quick lookup. Also, if part 2 had us change the marks based on if they overlap after a fold, I'd be prepared.

For convenience (and as I learned later, part 2) I also wanted a routine that could print the board. Here's that routine in case you want to see some clojure.


  (defn board->string [board]
    (let [maxx (apply max (map first (keys board))) ;; find the largest x
          maxy (apply max (map second (keys board))) ;; find the largest y
          k (keys board)

          ;; The next line makes a vector of maxy vectors
          ;; each of which has maxx spaces
          ;; basically a vector of vectors or 2D matrix if you would
          grid (into [] (repeat (inc maxy) (into [] (repeat (inc maxx) \ ))))

          ;; go through the keys to our board (which is a hash table
          ;; and mark those squares with a #
          filled-grid (reduce (fn [b [y x]] (assoc-in b [x y] \#)) grid k)

          ;; change each line from a vector to a string
          ;; but leave the overall thing a vector since
          ;; it looks fine when I print it.
          string-grid (map #(apply str %) filled-grid)
          ]
      string-grid
      ))

Once we have our "board" we can then deal with folding. Folds had to be over a horizontal (ex: y=7) or vertical (ex: x=5).

It was important to note that since we're folding we only want to take the points with coordinates greater than the folding line and transform those and not just transform all the points.

The actual transformation is pretty straightforward. Given the line's coordinate L the new point value can be calculated using:

newcoord = oldcoord - 2 * abs(oldcoord-L)

We just go through the points and if they have to be transformed, update them.

Part 1 solved.

Part 2 turned out to be far easier than I expected. Part 1 had us just do the first fold. Part 2 had us do all of them at which point, if we printed our board it would show us the 8 secret letters that formed our anwer. Since the example on the problem page was a y fold and the first fold of my data was an x fold I knew both my folds worked. I just ran through all the folds and voila.

Now, I'm not complaining about an easy problem. If it were harder I might not finish and then no write up but I was expecting something more.

There were ways to make part two a bit more complex. I could see maybe placing the data so far off the end so that you had to translate them down closer to the origin so that your result would print or maybe do something with overlapping marks when you fold as I mentioned up top.

Still, nothing wrong with an easy day mixed in and I still think it's a great problem. This could be done by a CS0 or CS1 class. What I particularly love about problems like this is that when you ultimately solve it you "decode" a secret message. Much cooler than just getting some number.

So, over half way done. I still have that part 2 of day 12 to finish but still captured half the total stars for the event. That's my starting goal each year.

If you want to see my full solution, you can check it out here.

Enjoy.

comments powered by Disqus