Skip Navigation

πŸ›Ά - 2023 DAY 18 SOLUTIONS -πŸ›Ά

Day 18: Lavaduct Lagoon

Megathread guidelines

  • Keep top level comments as only solutions, if you want to say something other than a solution put it in a new post. (replies to comments can be whatever)
  • You can send code in code blocks by using three backticks, the code, and then three backticks or use something such as https://topaz.github.io/paste/ if you prefer sending it through a URL

FAQ

14 comments
  • Rust

  • Nim

    I am not making good time on these anymore.

    For part 1, I walked through the dig plan instructions, keeping track of the highest and lowest x and y values reached, and used those to create a character grid, with an extra 1 tile border around it. Walked the instructions again to plot out the trench with #, flood-filled the exterior with O, and then counted the non-O tiles. Sort of similar to the pipe maze problem.

    This approach wouldn't have been viable for part 2, due to the scale of the numbers involved. Instead I counted the number of left and right turns in the trench to determine whether it was being dug in a clockwise or counterclockwise direction, and assumed that there were no intersections. I then made a polygon that followed the outer edge of the trench. Wherever there was a run of 3 inward turns in a row, that meant there was a rectangular protrusion that could be chopped off of the main polygon. Repeatedly chopping these off eventually turns the polygon into a rectangle, so it's just a matter of adding up the area of each. This worked great for the example input.

    Unfortunately when I ran it on the actual input, I ran out of sets of inward turns early, leaving an "inside out" polygon. I thought this meant that the input must have intersections in it that I would have to untwist somehow. To keep this short, after a long debugging process I figured out that I was introducing intersections during the chopping process. The chopped regions can have additional trench inside of them, which results in those parts ending up outside of the reduced polygon. I solved this by chopping off the narrowest protrusions first.

  • C

    Fun and interesting puzzle! In part 1 I fumbled a bit trying to implement even/odd outside/inside tracking before realizing that wouldn't work for this shape and just did the flood fill.

    For part 2 I correctly guessed that like the intersecting cuboids (2021 day 22) it would be about finding a better representation for the grid or avoiding representing it entirely. Long story shorter:

     c
        
    /*
     * Conceptually: the raw map, which is too large to fit directly in
     * memory for part 2, is made much smaller by collapsing (and counting)
     * identical rows and columns. Another way to look it at is that a grid
     * is fitted to make 'opaque' cells.
     *                                           |   |#|##|#
     * For example:                             -+---+-+--+-
     *                                          #|###|#|  |#
     *       ####               ### 1           -+---+-+--+-
     *   #####  #             ### # 1           #|   | |  |#
     *   #      #   becomes   #   # 2     or:   #|   | |  |#
     *   #      #             ##### 1           -+---+-+--+-
     *   ########             13121             #|###|#|##|#
     *
     * To avoid a lot of complex work, instead of actually collapsing and
     * splitting rows and columns, we first generate the wall rectangles and
     * collect the unique X and Y coordinates. Those are locations of our
     * virtual grid lines.
     */
    
      

    Despite being quite happy with this solution, I couldn't help but notice the brevity and simplicity of the other solutions here. Gonna have a look what's happening there and see if I can try that approach too.

    (Got bitten by a nasty overflow btw, the list of unique X coordinates was overwriting the list of unique Y coordinates. Oh well, such is the life of a C programmer.)

    https://github.com/sjmulder/aoc/blob/master/2023/c/day18.c

    • Oh, just like day 11! I hadn't thought of that. I was initially about to try something similar by separating into rectangular regions, as in ear-clipping triangulation. But that would require a lot of iterating, and something about "polygon" and "walking the edges" went ping in my memory...

  • Nim

    Decided to go for a polygon approach for part 1 using the Shoelace formula to calculate the area. This meant part 2 only resulted in larger values, no additional computation.

    Code runs in <1ms for part 1 and 2 combined

    Source

    • Shoelace formula

      This would have been really useful to know about. I've committed to a certain level of wheel-reinvention for this event unless I get really stuck, but I'm sure it'll come up again in the future.

      • This was actually something I learned for my job, it was nice to be able to apply it here.

        I like your commitment to wheel-reinvention, it can be a lot more fun than going for an existing or 'intended' approach.

  • Haskell

    Wasn't able to start on time today, but this was a fun one! Got to apply the two theorems I learned from somebody else's solution to Day 10.

  • C++

    No scala today

     cpp
        
    #include 
    #include 
    #include <map>
    #include 
    
    #include 
    #include 
    #include 
    #include 
    #include 
    #include 
    #include 
    
    struct HorizontalEdge { boost::icl::discrete_interval x; long y; };
    
    long area(std::vector he) {
        if(he.empty())
            return 0;
    
        boost::icl::interval_set intervals;
        std::ranges::sort(he, std::less{}, &amp;HorizontalEdge::y);
        long area{};
        long y = he.front().y;
    
        for(auto const&amp; e : he) {
            area += intervals.size() * (e.y - std::exchange(y, e.y));
            if(intervals.find(e.x) != intervals.end())
                intervals.erase(e.x);
            else 
                intervals.add(e.x);
        }
    
        return area;
    }
    
    struct Instruction {
        long l;
        int d;
        std::string color;
    };
    
    enum Dir { R=0, U=1, L=2, D=3 };
    std::unordered_map char_to_dir = {{'R', R}, {'U', U}, {'L', L}, {'D', D}};
    
    auto transcode(std::vector const&amp; is) {
        return flux::from(std::move(is)).map([](Instruction in) {
            long v = std::stoul(in.color.substr(0, 5), nullptr, 16);
            return Instruction{.l = v, .d = (4 - (in.color.at(5) - '0')) % 4, .color=""};
        }).to>();
    }
    
    std::vector read(std::string path) {
        std::ifstream in(std::move(path));
        return flux::getlines(in).map([](std::string const&amp; s) {
            Instruction i;
            char dir;
            if(auto r = scn::scan(s, "{} {} (#{:6})", dir, i.l, i.color)) {
                i.d = char_to_dir[dir];
                return i;
            } else {
                throw std::runtime_error{r.error().msg()};
            }
        }).to>();
    }
    
    auto turns(std::vector is) {
        if(is.empty()) throw std::runtime_error{"Too few elements"};
        is.push_back(is.front());
        return flux::from(std::move(is)).pairwise_map([](auto const&amp; lhs, auto const&amp; rhs) { return (rhs.d - lhs.d + 4) % 4 == 1; });
    }
    
    std::vector toEdges(std::vector is, bool left) {
        std::vector res;
        long x{}, y{};
    
        auto t = turns(is).to>();
    
        // some magic required to turn the ### path into its outer edge
        // (which is the actual object we want to compute the area for)
        for(size_t j = 0; j &lt; is.size(); ++j) {
            auto const&amp; i = is.at(j);
            bool s1 = t.at((j + t.size() - 1) % t.size()) == left;
            bool s2 = t.at(j) == left;
            long sign = (i.d == U || i.d == L) ? -1 : 1;
            long old_x = x;
            if(i.d == R || i.d == L) {
                x += i.l * sign;
                auto [l, r] = old_x &lt; x ? std::tuple{old_x + !s1, x + s2} : std::tuple{x + !s2, old_x + s1};
                res.push_back(HorizontalEdge{.x = {l, r, boost::icl::interval_bounds::right_open()}, .y = y});
            } else {
                y += (i.l + s1 + s2 - 1) * sign;
            }
        }
    
        return res;
    }
    
    long solve(std::vector is) {
        auto tn = turns(is).sum() - ssize(is);
        return area(toEdges(std::move(is), tn > 0));
    }
    
    int main(int argc, char* argv[]) {
        auto instructions = read(argc > 1 ? argv[1] : "../task1.txt");
        auto start = std::chrono::steady_clock::now();
        fmt::print("task1={}\ntask2={}\n", solve(instructions), solve(transcode(std::move(instructions))));
        fmt::print("took {}\n", std::chrono::steady_clock::now() - start);
    }
    </map>
      
14 comments