Rules: no spoilers.

The other rules are made up as we go along.

Share code by link to a forge, home page, pastebin (Eric Wastl has one here) or code section in a comment.

  • swlabr@awful.systemsEnglish
    3·
    11 months ago
    spoiler

    DP to me is when you use memoisation and sometimes recursion and you want to feel smarter about what you did.

    I also struggle to think of the need for DP, even in a more “elegant” approach. Maybe if you wanted to do an O(n) memory solution instead of n^2, or something. Not saying this out of derision. I do like looking at elegant code, sometimes you learn something.

    I feel like there’s an unreadable Perl one line solution to this problem, wanna give that a go, @gerikson?

    • zogwarg@awful.systemsEnglish
      3·
      11 months ago
      spoiler

      Part 2 only, but Part 1 is very similar.

      #!/usr/bin/env jq -n -R -f
[
  # For each line, get numbers eg: [ [1,2,3] ]
  inputs / " " | map(tonumber) | [ . ] |

  # Until latest row is all zeroes
  until (.[-1] | [ .[] == 0] | all;
   . += [
     # Add next row, where for element(i) = prev(i+1) - prev(i)
     [ .[-1][1:] , .[-1][0:-1] ] | transpose | map(.[0] - .[1])
    ]
  )
  # Get extrapolated previous element for first row
  |  [ .[][0] ] | reverse | reduce .[] as $i (0; $i - . )
]

# Output sum of extapolations for all lines
| add

      I’m pretty sure you could make this one line and unreadable ^^.

      • swlabr@awful.systemsEnglish
        3·
        11 months ago

        Here’s where I landed in dart

        no comments 
        d9(bool s) {
  print(getLines().fold(0, (p, e) {
    int pre(List h, bool s) {
      return h.every((e) => e == 0)
          ? 0
          : (pre(List.generate(h.length - 1, (i) => h[i + 1] - h[i]), s)) *
                  (s ? -1 : 1) +
              (s ? h.first : h.last);
    }

    return p + pre(stois(e), s);
  }));
}