Day 6

all / 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17 / 18 / 19 / 20 / 21 / 22 / 23 / 24 / 25

Available as an RSS Feed

Prompt / Code / Rendered

Another day that is fairly straightforward in Haskell, I feel! But in other languages that support functional approaches, it should be straightforward as well.

The answer involves lists of groups of responses:

import           Data.List.NonEmpty
import           Data.Set
import qualified Data.List.NonEmpty as NE
import qualified Data.Set           as S

type Response = Set Char
type Group    = NonEmpty Response

parseAnswers :: Set Char -> [Group]
parseAnswers = mapMaybe ((fmap . fmap) S.fromList . NE.nonEmpty . lines)
             . splitOn "\n\n"

And now we just need to decide how to aggregate each group. For part 1, this requires a set union between every Response in a Group:

part1 :: [Group] -> Int
part1 = sum . map (S.size . foldr1 S.union)

(foldr1 here is safe because we have a non-empty container)

And for part 2, this requires a set intersection between every Response in a Group:

part2 :: [Group] -> Int
part2 = sum . map (S.size . foldr1 S.intersection)

That's it!

Back to all reflections for 2020

Day 6 Benchmarks

>> Day 06a
benchmarking...
time                 124.2 μs   (122.7 μs .. 127.3 μs)
                     0.990 R²   (0.970 R² .. 1.000 R²)
mean                 125.7 μs   (123.1 μs .. 130.8 μs)
std dev              13.18 μs   (4.807 μs .. 23.01 μs)
variance introduced by outliers: 82% (severely inflated)

* parsing and formatting times excluded

>> Day 06b
benchmarking...
time                 124.8 μs   (123.9 μs .. 126.4 μs)
                     0.997 R²   (0.991 R² .. 1.000 R²)
mean                 125.4 μs   (124.1 μs .. 127.8 μs)
std dev              6.333 μs   (790.0 ns .. 11.65 μs)
variance introduced by outliers: 51% (severely inflated)

* parsing and formatting times excluded

advent-of-code-2020/reflections-out/day06.md

Day 6

Day 6 Benchmarks