Clojurians Log v2

My solution https://github.com/zelark/AoC-2020/blob/master/src/zelark/aoc_2020/day_07.clj

Mine: https://github.com/transducer/adventofcode/blob/master/src/adventofcode/2020/day7.clj I don’t know why exactly I had to call set on the tree-seq entries in part 1 nor why I had to do apply + 1 in part 2, but it works 🙂

I forgot about tree-seq … that’s a good idea.

🙂 i now see set is because you have to find amount of different colors and + 1 is because you also have to count the containing bag

(into (set found-bags) (mapcat #(find-outer-bags bags %) found-bags)) nice @zelark

I used this one as an https://github.com/ocisly/advent2020/blob/a96bd7be5e961d737bf33bc552724542e9e65d02/day-7.clj#L19-L41 to learn core.logic... 😬

Wow nice, I tried to use core.logic last year for a somewhat similar problem (https://adventofcode.com/2019/day/14) but I did not manage

@erwinrooijakkers thanks to your solution for part 2, I simplified mine. Now it’s pretty straightforward

(defn count-bags [bags [bag n]]
  (* n (apply + 1 (map #(count-bags bags %) (get bags bag)))))

don't have good intuition for graphs, always struggle https://github.com/nbardiuk/adventofcode/blob/master/2020/src/day07.clj

(loop [bags #{"shiny gold"}
       old-bags #{}]
  (if (= bags old-bags)
    (dec (count bags))
    (let [new-bags (into bags (map first
                                   (filter #(seq (clojure.set/intersection bags (second %))) data)))]
      (recur new-bags bags))))

(defn get-bags-inside [times [color amount]]
  [(mapv (partial get-bags-inside (* times amount)) (data color)) (* times amount)])

(dec (apply + (flatten (get-bags-inside 1 ["shiny gold" 1]))))

https://github.com/lambdaisland/aoc_2020/blob/main/src/lambdaisland/aoc_2020/puzzle07.clj

https://github.com/genmeblog/advent-of-code-2020/blob/master/src/advent_of_code_2020/day07.clj

Nothing very exciting in my solution but I will post anyway 🙂 https://github.com/benfle/advent-of-code-2020/blob/main/day7.clj It took me way too long to figure out I was off by one for part 2 because I was counting the top shiny gold bag :)

using loom: https://github.com/tschady/advent-of-code/blob/master/src/aoc/2020/d07.clj

What am I missing with this regex? Works fine for groups of 1 or two but fails with 3 or more (the second group gets improperly matched)

(\w+ \w+) bags? contain ((\d+) (\w+ \w+).*? bags?)(, (\d+) (\w+ \w+) bags?)*\.

the * doesn’t repeat the capture group like you think it does

i’d consider re-seq

visualization >_< https://i.redd.it/gx6l9oavzp361.jpg

> What am I missing with this regex? @jculp if you repeat a group, this group will be overwritten by the next match. For such cases you need to split a line, or just write 2 regex. Check my solution above for example.

heres mine, O'm definitely interested to see what others came up with today as I'm fairly new to clojure still https://github.com/listba/advent-of-code-2020/blob/master/clojure/src/aoc_2020/days/07.clj

Here's mine: https://github.com/rjray/advent-2020-clojure/blob/master/src/advent_of_code/day07.clj Writing the parsing took longer than I'd like, but I really got stuck in part 1 due to a small logic-error. Part 2 only took about 10 more minutes.

https://redpenguin101.github.io/posts/2020_12_07_aoc2020_day7.html. So impressed at how concise some of the answers here are. That was a fun problem though!

Again, lots of effort spent on processing the raw input into a nice shape. I notice my part 2 recursive function has the same pattern as several others. Took me some grinding to get it working. Still not pretty, oh well.

;; with input processed as:
;;     ["dull aqua" ([4 "dark fuchsia"] [1 "shiny purple"])]
(defn count-nestedbags [inp topbag]
  (letfn [(rcount [bag]
            (let [[_ bt] bag
                  contained (second (first (filter #(= bt (first %)) inp)))]
              (apply + 1
                     (for [[bc bn :as thebag] contained]
                       (if (= bn "no other") bc
                         (* bc (rcount thebag)))))))]
    (dec (rcount [1 topbag]))))

(count-nestedbags input-part2 "shiny gold")

This was so hard for me. It took me a lot of time to realize my first regex approach would not work. Then I had struggles wrapping my head around the recursion I needed. Then I made all sorts of silly mistakes even though the structure was right… And the result aint pretty!

(comment ; step 1
  (def input (util/fetch-input 7))
  (->> (clojure.string/split-lines input)
       (map #(clojure.string/split % #"[ ,]"))
       (map #(->> %
                  (partition 2 5)
                  (map vec)
                  (map (partial apply str))))
       ((fn dig [look-for found rules]
          (let [directs (reduce (fn [acc [bag & bags]]
                                  (if (seq (clojure.set/intersection (set bags) look-for))
                                    (conj acc bag)
                                    acc))
                                #{}
                                rules)]
            (if (seq directs)
              (dig directs (conj found directs) rules)
              found)))
        #{"shinygold"} #{})
       (apply clojure.set/union)
       (count)))

(comment ; step 2
  (def input (util/fetch-input 7))
  (->> (clojure.string/split-lines input)
       (map #(clojure.string/split (str "1 " %) #"[ ,]"))
       (remove #((set %) "no"))
       (map #(->> %
                  (partition 3 5)
                  (map (fn [[n c1 c2]]
                         [(Integer/parseInt n) (str c1 c2)]))))
       ((fn dig [look-for found rules]
          (let [directs (reduce (fn [acc [[_n bag-color] & bags]]
                                  (->> (for [color look-for
                                             :when (= color bag-color)]
                                         (map #(repeat (first %) %) bags))
                                       (apply concat)
                                       (apply concat)
                                       (into acc)))
                                []
                                rules)]
            (if (seq directs)
              (dig (map second directs) (into found directs) rules)
              found)))
        ["shinygold"] [])
       (count)))

when I have troubles naming things, I know I don’t know what I am doing, but I pressed on on sheer intuition. ¯\(ツ)/¯

@zelark is my hero. his solution in this thread blows me away.

I didn’t finish Part II in this style yet, but I decided to try and do day7 in a rules engine (#clara). There’s a better way to write this, but this gets the right answer:

(defrecord ContainerBag [type contained-bags])
(defrecord IntermediateBag [parent-type type])
(defrecord GoldBag [parent-type])

(defrule produce-intermediate-bags
  [ContainerBag
   (= ?type type)
   (= ?contained-bags contained-bags)]
  =>
  (when (seq ?contained-bags)
    (insert-all!
     (mapv (fn [[kind num]]
             (if (= kind "shiny gold")
               (map->GoldBag {:parent-type ?type})
               (map->IntermediateBag {:parent-type ?type
                                      :type kind})))
           ?contained-bags))))

(defrule indirect-gold-bags
  [ContainerBag
   (= ?type type)]
  [IntermediateBag
   (= ?type parent-type)
   (= ?intermediate-type type)]
  [GoldBag
   (= ?intermediate-type parent-type)]
  =>
  (insert! (map->GoldBag {:parent-type ?type})))

(defquery gold-bags []
  [?gold-bags <- (acc/distinct :parent-type) :from [GoldBag]])

(defn run-rules []
  (-> (mk-session :cache false)
      (insert-all (doall (for [[k vs] (parse input)]
                           (map->ContainerBag {:type k
                                               :contained-bags vs}))))
      (fire-rules)))
;; Part I
(-> (run-rules)
    (query gold-bags)
    first
    :?gold-bags
    count)
;; => 179

It would be cool to dynamically generate defrules for each line

https://github.com/green-coder/advent-of-code-2020/blob/master/src/aoc/day_8.clj

https://github.com/nbardiuk/adventofcode/blob/7a65f3ea75e9fc33fa31dbcd4e75e2ab837090f1/2020/src/day08.clj#L35-L44

Todays challenge was super fun! https://github.com/oxalorg/aoc/blob/main/src/day8.clj

It reminds me my old emulator https://github.com/green-coder/girlfriend-advance

(defn run-code [code]
  (let [history (atom #{})]
    (loop [acc 0
           ip  0]
      (let [[op arg] (nth code ip [:hlt 0])]
        (if (contains? @history ip)
          {:inf-loop acc}
          (do (swap! history conj ip)
              (case op
                :hlt {:halt acc}
                :nop (recur acc (inc ip))
                :acc (recur (+ acc arg) (inc ip))
                :jmp (recur acc (+ ip arg)))))))))

And my final solution: https://github.com/zelark/AoC-2020/blob/master/src/zelark/aoc_2020/day_08.clj in the Part 2 firstly I used case , but re-implement it with a hash-map as @plexus showed on the live stream.

My version: https://github.com/lambdaisland/aoc_2020/blob/main/src/lambdaisland/aoc_2020/puzzle08.clj

I always feel filthy when I use atoms…. This one is pretty dirty, but I got it done: https://github.com/jreighley/aoc2020/blob/master/src/day8.clj

https://github.com/transducer/adventofcode/blob/master/src/adventofcode/2020/day8.clj

https://github.com/akovantsev/adventofcode/blob/master/src/adventofcode/y2020/day08.cljc#L45-L54

definitely a bit engineered, but I’m not getting caught off-guard by a part 2: intcode boogaloo https://github.com/motform/advent-of-clojure/blob/master/src/advent-of-clojure/2020/eight.clj

Part 2 was nice. https://github.com/genmeblog/advent-of-code-2020/blob/master/src/advent_of_code_2020/day08.clj#L29-L36

Looks like most of us have similar approaches. https://github.com/benfle/advent-of-code-2020/blob/main/day8.clj

Day 8 makes me appreciate immutable data structures and defmulti. I am fairly happy with my code for a change. https://github.com/chamaeleon/aoc2020-clj/blob/master/src/aoc2020/day08.clj

https://github.com/tschady/advent-of-code/blob/master/src/aoc/2020/d08.clj

https://github.com/bradlucas/advent-of-code-2020/blob/master/src/advent/day08.clj

https://redpenguin101.github.io/posts/2020_12_08_aoc2020_day8.html - this one did give me some flashbacks to intcode 😨

I really liked the intcode ones last year so I hope there's some future days that build this one up Anyways my solution for day 08 https://github.com/listba/advent-of-code-2020/blob/master/clojure/src/aoc_2020/days/08.clj

https://github.com/kfirmanty/advent-of-code-2020/blob/main/src/advent_of_code_2020/day8.clj

I tried for so long do step 2 in a tree-ish way. Gave up. The brute force was with me though. Copying the program and modifying each one differently worked. I want my gold stars! Glad I learned about cond-> the other day.

(defn parse [input]
  (->> (clojure.string/split-lines input)
       (map #(clojure.string/split % #" "))
       (mapv (fn [[op arg]] [op (Integer/parseInt arg)]))))

(defn run [program]
  (let [length (count program)]
    (reduce (fn [cpu _]
              (let [{:keys [pc pcs]} cpu]
                (if (< pc length)
                  (let [[op arg]         (nth program pc)]
                    (if (pcs pc)
                      (reduced cpu)
                      (cond-> cpu
                        (= op "acc")        (update :acc #(+ % arg))
                        (#{"acc" "nop"} op) (update :pc inc)
                        (= op "jmp")        (update :pc #(+ % arg))
                        :always             (update :pcs conj pc))))
                  (reduced cpu))))
            {:acc    0
             :pc     0
             :length (count program)
             :pcs    #{}}
            (range))))

(comment
  (def input (util/fetch-input 8))
  ; step 1
  (->> input
       (parse)
       (run)
       :acc)

  ; step 2
  (->> input
       (parse)
       (#(repeat (count %) %))
       (keep-indexed (fn [i program]
                       (let [[op val] (program i)]
                         (assoc program i [(cond
                                             (= "nop" op) "jmp"
                                             (= "jmp" op) "nop"
                                             :else        op)
                                           val]))))
       (map run)
       (filter #(= (:pc %) (:length %)))
       (first)
       :acc))

I’m officially a day behind now and late to the party https://gist.github.com/KennyMonster/9b9db6daa056e41413112d0fb31a5e47

feels like part 2 could be way nice even brute forcing it… but I dunno how

A reason part 2 took me so long is that I was looking at it as a tree problem. So I was able to build a structure something like so (simplyfied):

[[nop jmp]
 [acc]
 [acc]
 [jmp nop]
 [acc]]

[[nop acc acc jmp acc]
 [jmp acc acc jmp acc]
 [nop acc acc nop acc]
 [jmp acc acc nop acc]]

@pez Here is what I would do to optimize this problem: • Collect all the switch-indexes , indexes of the nop and jmp where the original program go before an infinite loop is detected, because the solution is to flip one of those indexes. • Build the reverse-jmp-graph, where edges are from jmp-targets to jmp-sources. • In this graph, collect all land-indexes which are reachable from the end of the program. • Find which change on instructions at switch-indexes let the program go into instructions at land-indexes .