Of Monads and Men

Sherlock9

03:48

Hello again, I'm having trouble with the following code f a=[a!!i<$[1..(f a)!!i]|i<-[0..]]

Theoretically it should take a[i] from array a, such as [2, 3, 1] and append a[i] to the new list a[i] times, using n<$[1..n]. Is there a way to concatenate n<$[1..n] instead of adding as a list inside a list?

Sherlock9

04:34

Ah nuts. I forget that a!!i in the beginning should be (cycle a)!!i

Laikoni

07:58

@Sherlock9 I'm not exactly sure what you are trying to do, but you can add concat (or id=<<) in front of the list comprehension to get it to type: Try it online!

Sherlock9

@Laikoni Thanks for the help. I'm trying to do this challenge

Laikoni

There was also a $ missing behind take 25.

Sherlock9

Hm, as I expected I do need to initialize the array. That's going to eat a lot of bites

Also how would I put take in f?

Laikoni

Exactly, otherwise the self reference can not be resolved.

Sherlock9

Well, this almost worked

Laikoni

08:04

@Sherlock9 You probably can't because of the self reference, but take 25.f would work. Or is the 25 also an input?

Sherlock9

25 is an input

Sherlock9

08:35

Hm, maybe I can do take n in a function below the main list comprehension instead

@Laikoni Any idea what's going wrong here with f a=(a!!0)<$[1..(a!!0)]++concat[(cycle a)!!i<$[1..(f a)!!i]|i<-[1..]]?

flawr

09:00

I think (cycle a) is the problem, it producs an infinite list

(not sure yet)

Laikoni

@Sherlock9 <$ binds weaker than ++

You need to enclose the initialisation in parentheses.

flawr

And as long you don't have any guards [...|i<-[1..]] can usually be shortened using map or <$>

For some reason I keep thinking about an ice cream cone whenever I see <$

Sherlock9

Huzzah! f a=((a!!0)<$[1..(a!!0)])++concat[(cycle a)!!i<$[1..(f a)!!i]|i<-[1..]]

@flawr I'm having trouble figuring how to use map and <$> because the i variable is used in several different places

I may just need to learn more about lambda calculus reduction methods

flawr

ah right this seems to be getting longer

Sherlock9

I also think about a double-scoop ice cream cone when I see <$

flawr

09:07

@Sherlock9 high five:)

But you could simplify the a!!0 with f a|z<-a!!0=(z<$[1..z])...

Sherlock9

Excellent. 84 bytes now since I have to put take in the main code.

flawr

and I think you don't need the parentheses around f a and cycle a

Sherlock9

@Laikoni Before I forget, thank you very much :D

And we're down to 80 bytes. Is a shorter way to write ++concat?

Laikoni

id=<< is a shorter concat

Sherlock9

Ah, sorry you already said and I forgot to try it

Hm. An error

flawr

09:12

@Laikoni you need to add that to the tips!

Laikoni

@Sherlock9 Oh, ok, it does not work with the ++ in front

Sherlock9

I think you have to put parens because of weaker binding :/ 80 bytes earlier vs 79 bytes now

Laikoni

a#n=take n$f a can be shortened to (.f).take if the arguments are reversed.

Sherlock9

@flawr Apparently it is a tip

n#a=(.f).take is giving me another error

Laikoni

It's point free, you don't need n and a as arguments.

Like this:

[Try it online!](https://tio.run/##y0gszk7Nyfn/P00hsabKRjdRUdHANjk/LzmxREWjykYl2lBPrypW0‌yo6uTI5J1UBKJ0JEUwDs2NrMm10QdzYWK50Ww29NE29ksTs1P@5iZl5CrYKBUWZeSUKKgrpCkamCtFGOs‌Y6hrH/AQ "Haskell – Try It Online")

Sherlock9

09:17

Ah of course. Just need another single-character function name because #=... gives another error

flawr

@Sherlock9 well only as concatMap f, I the concat is probably used more often.

@Sherlock9 you'd have to write (#)=...

Sherlock9

Which brings us to 79 bytes

flawr

@Laikoni how do you come up with solutions like that? I'm getting dizzy by just looking at it!

Laikoni

@flawr pointfree.io

flawr

@Laikoni that was not what I hoped to hear :D

Anonymous

09:23

pointfree.io is how I make myself look like I know way more about Haskell than I do :P

flawr

Does anyone know, is it possible to add links in the room description?

Anonymous

Yep

Laikoni

@Sherlock9 Anonymous functions are allowed, so you don't need to count the (#)=.

Anonymous

See: TNB

flawr

I mean using []()?

Or can you just past URLs?

Laikoni

09:24

@Sherlock9 70 bytes

Anonymous

Just paste URLs

Anonymous

No Markdown in room descriptions, unfortunately

Laikoni

@flawr Well, at least by now I usually understand the transformations and also do them myself in simple cases.

flawr

Hm, I'll have to practice a little bit then:)

@Laikoni Do you mind if I add the tips for golfing in haskell in the room description? (perhpas also pointfree.io?)

Sherlock9

@Laikoni Many thanks :D I'm going to write up my answer. Is it alright if in my answer I write the following: Many thanks to Laikoni and flawr for their help in debugging and golfing this answer in the PPCG Haskell chatroom, Of Monads and Men

Laikoni

09:27

@flawr Sure, go ahead.

Anonymous

Credit and advertisement, nice double-whammy

2

flawr

room topic changed to Of Monads and Men: For discussion about, learning of and golfing in Haskell codegolf.stackexchange.com/questions/19255/… pointfree.io [haskell]

too bad it doesn't just show the title of the question =/

Anonymous

I thought you actually did it there, and was confused why it didn't change on the side

Laikoni

The Guide to Golfing Rules in Haskell is also a good place to link to.

flawr

Oh yeah, is there something else to add?

One could also link these in the haskell description

ok this is strange: I recently pinned this CMS you can still see on the starboard, without actually starring it - now that I've unpinned it it looks like I starred my own message (but cannot unstar it)

and if I try to unstar it it says "you cannot star your own message" as if I haven't starred it yet.

Laikoni

09:32

@Sherlock9 I like that. Also 68 bytes

Anonymous

@flawr Yep, known bug.

Anonymous

As with all chat bugs, it has existed for quite a while, and SE has made no visible effort to fix it or even acknowledge it

Sherlock9

@Laikoni Well, I was starting to explain the 70-byte one, but that's gone and confused me

Oh wait

flawr

@Mego ok at least it was not me doing something wrong:)

Sherlock9

a@(z:_) is just defining a and defining z as the head

flawr

09:33

Also shouldn't ROs be able to clear stars?

Anonymous

They should be able to

Anonymous

You use the cancel stars button from the starboard, not unstarring

flawr

Ah that was what I was looking for! Thanks a lot!

"Remove the stars against this post."

Anonymous

Glad to help :P

flawr

I just added some links to the haskell tag description: codegolf.stackexchange.com/tags/haskell/info

Sherlock9

09:39

Answer is posted

Anonymous

I got to figure all this stuff out by myself when I was added as a TNB RO :P At least you're not having to try to find the button during a star spam wave

Sherlock9

Ah, trial by fire

Anonymous

Trial by fire with gasoline

Sherlock9

Or since it's stars, perhaps, trial by meteor shower

Anonymous

That's pretty accurate :P

Anonymous

09:41

I can't remember the last time I had to cancel stars, though, so things have gotten better

Anonymous

Either that or I've just gotten luckier and not been around for it

Feeds

flawr has made a change to the feeds posted into this room

Laikoni

@Feeds Oh, nice.

flawr

32

A: Point Free problems in Haskell

Converting to pointfree style can be done entirely mechanically, though it's hard without being comfortable with the fundamentals of Haskell syntax like left-associative function application and x + y being the same as (+) x y. I will assume you are comfortable with Haskell syntax; if not, I sug...

@Laikoni @Mego ^ provides an algorithm for making thins pointless

Anonymous

09:59

@flawr Oh, I'm aware :) pointfree.io (and the pointfree package it uses) uses that algorithm, I believe

Sherlock9

Thanks for the link. Reading it now

Did we actually get 5 Haskell questions at once, or is this Feeds trying to catch up?

Anonymous

Probably the latter

flawr

Oh no, can I delete these messages somehow?

Anonymous

You can move them to Trash

Anonymous

You might need to get access, though, since it's a gallery room

Anonymous

10:04

I think ArtOfCode is the one to ping

Anonymous

Apparently this guy is the sole RO of Trash, but I think Art used his diamond powers to give the TNB ROs write access so we could move trash there

flawr

^ this is what I feel like :D

Anonymous

Too much RO power? :P

Feeds

10:28

0

A: Tips for golfing in Haskell

Converting to Pointfree Als @Laikoni already mentioned there is pointfree.IO, but if you want to do this conversation manually you can check out this answer on stackoverflow by @luqui, which provides an algorithm: You need the following combinators, which are in the standard library. I hav...

Sherlock9

@flawr May edit your post in Haskell tips for clarity?

flawr

@Sherlock9 sure!

Sherlock9

Done!

wizzwizz4

11:56

→ 14 messages moved to Trashcan

@flawr I'm not reading the contents of that room this time... It's scary.

flawr

Is it possible to trash things as RO?

wizzwizz4

@flawr It's possible to move things from any room you are RO in to any room you have access to, iirc.

Try moving this message to the Sandbox.

flawr

→ 1 message moved to Sandbox

Ah now I see, thanks a lot!

wizzwizz4

But that should never be done.

flawr

Best if we don't tell anyone :)

Thanks a lot for your help!

wizzwizz4

12:04

It's one of the less important abuses because it's obvious when it's being done.

Transcript for