Jelly Hypertraining

	"'": lambda link, none = None: attrdict(
		arity = link.arity,
		call = lambda x = None, y = None: variadic_link(link, (x, y), flat = True, lflat = True, rflat = True)
	)

user202729

What about for dyads? "spawn"?

caird coinheringaahing

1:41 PM

The default is lflat=False, rflat=False, and the code gets the depth of an atom by doing lflat or ldepth, meaning that the ' quick forces it to only act on the lowest depth

@user202729 No clue. I asked Dennis:

in Jelly, Oct 24 '17 at 12:32, by Dennis

@cairdcoinheringaahing It's buggy for dyads and mostly useless for monads. It's not a no-op in 1,2a'3, but one could just use 1,2ȧ3 instead.

Mr. Xcoder

So we shouldn't worry about it

I should try ƭ, didn't really use it much

caird coinheringaahing

ƙ is the only quick where I don;t understand what it does

user202729

Key?

map monad over group of equal elements.

Mr. Xcoder

Ok ƙ is the easiest IMO. It maps a function over the groups of equal items in a list.

user202729

only if you try it. Documentation is hard to understand.

caird coinheringaahing

1:45 PM

I though it was something like Œg<monad>€, but I was wrong

Mr. Xcoder

[1,2,3,1,2,1,1,2]Lƙ gives [4,3,1] I think (the length of the 1-group, of the 2-group, of the 3-group)

@cairdcoinheringaahing It's just that but they don't have to be adjacent

Anyway gtg]

caird coinheringaahing

o/

user202729

With sort too.

(and then sort groups by first occurences)

caird coinheringaahing

So <link>ƙ is equivalent to ṢŒg<link>$€?

user202729

The latter too. Then sort groups by first occurence.

Mr. Xcoder

2:07 PM

@cairdcoinheringaahing You can get a better hang of it using identity:

Try it online!

(I'm back, had a small house issue)

caird coinheringaahing

That actually makes it a lot clearer, thanks!

Mr. Xcoder

@cairdcoinheringaahing You're welcome!

caird coinheringaahing

I'm confused now. Should I explain tie, or are you good?

Mr. Xcoder

I'm good

Wait I am still confused. It doesn't work as I expected: I tried [1,2,3,4,5,6,7,8,9]’‘H3ƭ

Does it only work for dyads?

caird coinheringaahing

def tie(links, outmost_links, index):
	ret = [attrdict(arity=2 if max(link.arity for link in links) == 2 else 1)]
	n = 2 if links[-1].arity else links[-1].call()
	def _make_tie():
		i = 0
		while True:
			yield links[i]
			i = (i + 1) % n
	cycle = _make_tie()
	ret[0].call = lambda x = None, y = None: variadic_link(next(cycle), (x, y))
	return ret

That's the source code

Mr. Xcoder

2:19 PM

So it should for both monads and dyads.

Where is the parameter outmost_links used?

caird coinheringaahing

It's not. That just refers to the other links in the program, all quicks have that parameter

Mr. Xcoder

Ah ok

I should have a closer look at Jelly's source

Still, can you explain it once more?

caird coinheringaahing

It should take a list a, a nilad n and the previous n links and cycle those links over a. I don't get why it's not working though

Mr. Xcoder

Oh so I chose my code so perfectly that I discovered a bug :S

caird coinheringaahing

This is a good example of how it works with dyads

Mr. Xcoder

2:25 PM

I already analysed that, I am aware of it

n = 2 if links[-1].arity else links[-1].call()

I am just going to ask in the Jelly room I think

caird coinheringaahing

It doesn't seem to like monads, and that line seems to be why

Mr. Xcoder

in Jelly, 1 min ago, by Mr. Xcoder

@Dennis Does the "tie" quick (ƭ) work correctly for monads? I am very confused with the following code: ’‘H3ƭ. If I understand it correctly, it should return [0, 3, 1.5, 3, 6, 3, 6, 9, 4.5] for an input of [1, 2, 3, 4, 5, 6, 7, 8, 9], but it returns [0, 1, 2, 3, 4, 5, 6, 7, 8] instead. – Try it online! If this is, in fact, the expected behaviour, can you please explain how it works?

in Jelly, 43 secs ago, by Dennis

That said, I think you need ’‘H3ƭ€.

@cairdcoinheringaahing

caird coinheringaahing

ƭ must not vectorise then

Mr. Xcoder

Which seems highly non-sensical to me

How would it even work for non-iterables?

blame-miles

caird coinheringaahing

It rangifies integers

Mr. Xcoder

2:39 PM

@cairdcoinheringaahing ಠ_ಠ

caird coinheringaahing

ಠ_ಠ I used the € version

Mr. Xcoder

Strange...

caird coinheringaahing

It only applies the first link then ಠ_ಠ

Mr. Xcoder

ಠ______ಠ

The Jelly Effect: 4 disapprovals in the last 5 messages

caird coinheringaahing

3:18 PM

ಠ_ಠ I've been looking at the source code, and I have no idea how to make tie vectorise

caird coinheringaahing

4:07 PM

Opened a PR to make tie vectorise :P

Mr. Xcoder

5:04 PM

caird coinheringaahing™: I have no idea how to make tie vectorise – One hour later: Opened a PR to make tie vectorise

caird coinheringaahing

5:44 PM

@Mr.Xcoder I just wrapped the call in tie with the code for € in monads :P

Mr. Xcoder

6:00 PM

ಠ_ಠ

Closed by Dennis :P

Mr. Xcoder

8:25 PM

CMC: Given a list L of integer and another integer I that occurs at least twice in L, return the maximum difference between two indexes of I in L.

[1,2,1,2,2,2,1,3,2,3,4,1], 1 -> 5

@Mr.Xcoder Ugh forgot I implemented... You know what... 3 bytes

Erik the Outgolfer

3 bytes

Mr. Xcoder

Mainly just showcasing a new atom

Should get something less trivial

Jonathan Allan

err should't the answer be 11?

rather than 5

Mr. Xcoder

Forgot to mention "Adjacent Occurrences"...

Jonathan Allan

oh

Mr. Xcoder

8:32 PM

CMC: Given a list L and two elements a and b, replace all occurrences of a in L with b (always wondered how that would be done in a golfy manner)

Erik the Outgolfer

1 byte

Mr. Xcoder

[1,2,2,2,2,4,4,4,3,2,4], 2, 4 -> [1,4,4,4,4,4,4,4,3,4,4]

What?

Jonathan Allan

err 1 byte

it's a built in

Mr. Xcoder

Wait are there tryads in Jelly? :O

Jonathan Allan

dyad list

Mr. Xcoder

8:33 PM

Oh... Crap y right?

Erik the Outgolfer

yes

Jonathan Allan

take the a and b as [a,b]

Mr. Xcoder

Ugh Forgot about it

Dang...

In that case...

Jonathan Allan

harder is to replace all a with b AND all existing b with c (I think)

Mr. Xcoder

Yeah that's a good CMC

Jonathan Allan

8:35 PM

maybe?

Erik the Outgolfer

nope

Jonathan Allan

have not done it

Mr. Xcoder

Actually the built-in does that for us, tweaking it a bit

Try it online!

Jonathan Allan

yeah it does, you need to reformat the abc though...

Mr. Xcoder

yes

Jonathan Allan

8:37 PM

question is how to do it golfily, maybe y/ would come into play too?

Erik the Outgolfer

takes input as left argument [a, b, b, c] and right argument L: y (and yes why would this be valid)

why use y/ o_o that's an abomination from hell

Jonathan Allan

takes [[a,b],[b,c]]

well how many bytes is the only question :D

Erik the Outgolfer

you're looking too much into it, of course why should it be valid

but it's actually 3 bytes, yeah

Jonathan Allan

I didn't say it was :/

Mr. Xcoder

Something like ,Ṛy should do it I think?

Erik the Outgolfer

8:41 PM

3 bytes spoiler (yes, you need to click on it)

Jonathan Allan

I have a 3, wonder if it's the same...

oh it's different

Mr. Xcoder

@EriktheOutgolfer ಠ_ಠ why do you take a list with 3 elements as input?

Erik the Outgolfer

unicode MD5: 0E6D2FB0A186CA806DF12CA4FF9A4E51

oh too late it seems

Jonathan Allan

^ is that a two byte one???

Erik the Outgolfer

what no

Jonathan Allan

8:44 PM

oh that's your three

:)

Erik the Outgolfer

I said "unicode" because that's what you should be entering on the MD5 page to check

not JELLY-encoded

Jonathan Allan

never used one

Erik the Outgolfer

(btw Leaky introduced me to it :p)

Jonathan Allan

MD5 e79f9de8a4de9ae9d0e32c9a1d00b5c8

Mr. Xcoder

Ok you can reveal them if you want :P

Erik the Outgolfer

8:47 PM

in a spoiler link though :P

"trying to find another person's approach" is sometimes done that's why

Jonathan Allan

tio.run/##y0rNyan8/1/n2Fwg@d9Qx0TH4n@0oY6RDggbgWljIGkMlDCMBQA

Mr. Xcoder

Ok I reaaaaly wanna move on from this y madness

Jonathan Allan

similar really

Mr. Xcoder

;-; I had ,2\y

Jonathan Allan

I gave a challenge to my colleagues over xmas, but it's easy for Jelly (it was only for Python 2)

Mr. Xcoder

8:49 PM

Forgot "Neighbours" exists

Jonathan Allan

but may as well

CMC: given a list L of three lists of non-empty "strings" (lists of characters) return a "string" (again Jelly - so list of characters) with the TLA of TLAs

I have 3 bytes.

Erik the Outgolfer

@Mr.Xcoder it's a new quick, you'll get used to it ;)

@JonathanAllan "TLA"?

Mr. Xcoder

But... I have used it a lot already (even on Main) Dunno how I missed it

Jonathan Allan

three letter acronmn

Mr. Xcoder

Test case?

Jonathan Allan

8:53 PM

one sec...

Erik the Outgolfer

@JonathanAllan me too (I think)

Jonathan Allan

i.e. [["This","Is","Golf"],["I","Like","Golf"],["This","Was","Dumb"]] -> "TIT"

it's just the 'T's of the two "This"'s with the 'I' of the "I" inbetween.

should be easy in Jelly to get the optimal score

Mr. Xcoder

3 bytes

Jonathan Allan

but in Python 2 with 5 non-golfers the best I saw was 37 bytes (I got 30)

Mr. Xcoder

MD5: 953059D3D9DF82A0659C2037D297F6A7 and spoiler.

Now I am tempted to solve it in Python 2 too

Jonathan Allan

8:56 PM

hmm why is the MD5 different :/

same as mine

I see f3b25d0c942f8959666ff54106084aef

Mr. Xcoder

Eh...

Jonathan Allan

I used miraclesalad.com/webtools/md5.php

is it wrong?

Erik the Outgolfer

try using the one we use here, passwordsgenerator.net/md5-hash-generator

Mr. Xcoder

Oct 17 '17 at 18:56, by Erik the Outgolfer

http://passwordsgenerator.net/md5-hash-generator/

Are lists of characters allowed in Python 2?

Erik the Outgolfer

my MD5 is the same as Mr. Xcoder's

lists of characters? no, they're lists of 1-char strings

Jonathan Allan

8:58 PM

So that one is just wrong?!

Mr. Xcoder

¯\_(ツ)_/¯

@JonathanAllan lambda k:[i[0][0]for i in k]? This is 28... Am I missing something?

Jonathan Allan

who host a site doing incorrect MD5?!

Erik the Outgolfer

I don't know which is wrong, but if we all use the same tool we all get the same results as appropriate

Jonathan Allan

onlinemd5.com agrees with the one I posted

Mr. Xcoder

Perhaps ours is wrong... But who cares, as long as we use the same service?

Jonathan Allan

9:01 PM

@Mr.Xcoder yeah it was a requirement for output to be a string (or printed version thereof) in my Python 2 question

Mr. Xcoder

Oh... Ouch

That will cost me many bytes

Erik the Outgolfer

@JonathanAllan actually miraclesalad agrees with passwordsgenerator and onlinemd5

Jonathan Allan

well I have a 30 :)

Erik the Outgolfer

I think your code is different

Jonathan Allan

:/

Erik the Outgolfer

9:02 PM

(that is, for MY md5)

Mr. Xcoder

Or did you just add 1 more space or something?

Erik the Outgolfer

and also a fourth service, md5.cz

Jonathan Allan

when I paste my code into password generator I also get the same again - err so it's how I copy from TIO to the generator.

(cos I know the code is the same from the fact that Mr. Xcoder posted a spoiler link :))

Erik the Outgolfer

perhaps you copied encoded code and not UTF-8?

@JonathanAllan was the 37-byter lambda(a,b,c):a[0][0]+b[0][0]+c[0][0]?

Jonathan Allan

nope but it's cool

If I knew how to MD5 properly I'd give the one for my 30 :p

Mr. Xcoder

9:06 PM

Just spoil it :P

Jonathan Allan

need to rewrite from memory first...

Mr. Xcoder

Also there is this: lambda c:zip(*c)[0]. Now getting the heads ain't short

Jonathan Allan

spoiler

Mr. Xcoder

I won't look at it immediately, and that was a joke BTW...

Jonathan Allan

:p

I feel it's optimal but if I posted on main xnor would post a 22 byter or something :p

Mr. Xcoder

9:09 PM

It might be possible in under ~ 28 using recursion, perhaps

Jonathan Allan

I got 35 using recursion myself

I had loads of 35 byte solutions actually

first submission from the non-golfers was 50 bytes :p

Erik the Outgolfer

@JonathanAllan that looks like 35 bytes to me...

Jonathan Allan

?

Mr. Xcoder

@JonathanAllan That's longer than just typing [0][0][0]+... ...

Jonathan Allan

yep

looks like 30 to me :/

Mr. Xcoder

9:11 PM

This is clearly 35 Jonathan

Jonathan Allan

yeah I think that's one I had :)

Erik the Outgolfer

@JonathanAllan w...what? even purposefully ungolfing, this is 48 bytes, maybe excessive spaces and variable name lengths?

Mr. Xcoder

That's what you linked, and it's not 30

Jonathan Allan

first solution used join and indexing and a for loop in a lambda was 50.

Mr. Xcoder

@EriktheOutgolfer perhaps no lambdas

Erik the Outgolfer

9:13 PM

but 48 bytes is the full program

Jonathan Allan

how does it look like 35?

ಠ_ಠ

but...it's 35 bytes

^

prob had redundant [] in the join

Erik the Outgolfer

9:14 PM

no join in the 48-byte version...

a=input()
print a[0][0][0]+a[1][0][0]+a[2][0][0]

maybe they spaced the =

Jonathan Allan

that is not the normal solution though

Erik the Outgolfer

oh

Jonathan Allan

for a non-golfer

Erik the Outgolfer

@JonathanAllan 38 bytes then?

Jonathan Allan

no it was way more, pretty sure it was 50

Erik the Outgolfer

9:17 PM

actually 37 bytes

lambda l:''.join([i[0][0]for i in l])

Jonathan Allan

prob two for loops in the list comprehension

Mr. Xcoder

My ungolfed-yet attempt with recursion: Try it online!

Jonathan Allan

think like a non-golfer

someone who writes clear code

Erik the Outgolfer

41 bytes:

lambda l:''.join(map(lambda x:x[0][0],l))

Jonathan Allan

@Mr.Xcoder LOL what? That is not ungolfed

Mr. Xcoder

9:19 PM

40 bytes is ungolfed

Jonathan Allan

do you guys write real-world code ever?

Mr. Xcoder

[0][0][0]...

Jonathan Allan

designed to be read easily

Mr. Xcoder

Yeah, but not Python

Erik the Outgolfer

well, an 74-byte version:

lambda l : l [ 0 ] [ 0 ] [ 0 ] + l [ 1 ] [ 0 ] [ 0 ] + l [ 2 ] [ 0 ] [ 0 ]

Mr. Xcoder

9:20 PM

Pfff

Jonathan Allan

we are not code-bowling

Mr. Xcoder

7 mins ago, by Jonathan Allan

first solution used join and indexing and a for loop in a lambda was 50.

@Mr.Xcoder I didn't see this

Erik the Outgolfer

@JonathanAllan that would be banned in code-bowling

Jonathan Allan

hmm I got 48, maybe there were a couple of redundant spaces too!

Mr. Xcoder

How about f=

Perhaps they had it?

Jonathan Allan

9:22 PM

I think not as I explicitly mentioned that f= was only necessary for a recursive lambda

Mr. Xcoder

Why are we discussing Python in JHT :S

Jonathan Allan

^^

Mr. Xcoder

@EriktheOutgolfer Can you please move the messages about Python to:

Jonathan Allan

I think there were two redundant spaces...

Mr. Xcoder

Python Byte Shavers

A room dedicated to Python golfing founded by FlipTack and Mr....

Jonathan Allan

9:23 PM

like this

Mr. Xcoder

And perhaps we shall move there

Jonathan Allan

we are ungolfing though :p

Mr. Xcoder

Yeah ^^^ makes sense

Erik the Outgolfer

@JonathanAllan oh we can't assume excessive spacing, that's something arbitrary

Jonathan Allan

But when they saw ... they all said "what is that???"

Erik the Outgolfer

9:24 PM

@Mr.Xcoder that'd be offensive there :P

Jonathan Allan

^back ticks :p

Mr. Xcoder

@EriktheOutgolfer No it wouldn't, I'm a RO :P

Jonathan Allan

` ... `

ugh formatting

Mr. Xcoder

Got it...

CMC: Compute the Nth non-square number (the 1st / 0th is 2).

2, 3, 5, 6, 7, 8, 10, 11, 12, 13, 14, 15, 17, 18, 19, 20, 21, 22, 23, 24, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, …

Jonathan Allan

naive solution in 7

oh make that 6

Mr. Xcoder

9:31 PM

I think I have 6 too

Make that 5 bytes. Returning as a float.

Spoiler or Spoiler

Jonathan Allan

So RE the MD5 thing, yes it was due to copying from TIO it seems

as a float - is it guaranteed to work for any valid input?

(I guess not)

Mr. Xcoder

Yes. Guaranteed.

It just returns .0

The second approach doesn't return as a float, but works with floats

Jonathan Allan

even with a very large n it will find the right answer?

Mr. Xcoder

I guess yes. OEIS seems to agree :P

Jonathan Allan

OEIS probably does not go very high

I meant if n is huge

Mr. Xcoder

9:41 PM

I am not sure

(BTW the links open a blank tio window)

To view the code, just hover over them

Anyway, I have to go get some sleep now. It's been a pleasure golfing with you. Mind if I see your solution now @JonathanAllan?

Jonathan Allan

moved away from it um... I think I have another 6, one sec.

oops now I have another 7... one sec I'll recreate my six

Mr. Xcoder

No rush :)

Jonathan Allan

six

hover over what?

Mr. Xcoder

Yeah, that's what thought you had

Jonathan Allan

Oh I see

Mr. Xcoder

9:46 PM

12 mins ago, by Mr. Xcoder

Spoiler or Spoiler

Jonathan Allan

they prob both fail for large n

but not usually a problem I think

I never quite know

Mr. Xcoder

I don't think it's a prob

Jonathan Allan

sleep well!

Mr. Xcoder

Thanks! You too (soon) I guess (if you're in the UK)!

Jonathan Allan

in a while :p

Mr. Xcoder

9:49 PM

Yeah of course who's sleeping at 10 PM :P

caird coinheringaahing

It's Saturday, no-one sleeps before 1AM here :P

Mr. Xcoder

Yeah, I'm just tired so I won't go after 12

Actually might be back later from mobile

caird coinheringaahing

Never mind, I googled it

Mr. Xcoder

Out of curiosity, what did you google? :p

caird coinheringaahing

Time in Romania

Mr. Xcoder

9:55 PM

Offset of 2 from the GMT

Or +3 during summer, I think

caird coinheringaahing

I know that now :P

Mr. Xcoder

I wondered if I am right actually :p

caird coinheringaahing

So, Erik and Xcoder are on the same times, and me and Jonathan are on the same time. Then Hyper's just by himself on Canadian time :P

Mr. Xcoder

Tfw you accidentally drop a recently-bought phone and you’re afraid to pick it up and look at its screen...

Jonathan Allan

aggggh

caird coinheringaahing

9:59 PM

Hmm, the discussion about KOTHs in TNB reminds me of my idea for a Jelly KOTH :P

Mr. Xcoder

ooo yeah I’d really like one of these

But I am not sure how non-Jellyers would receive it

caird coinheringaahing

The problem is creating a suitable problem to do

@Mr.Xcoder Same way non-JS users took the recent JS KOTH: ignore it

Mr. Xcoder

Yeah should be something a bit difficult

caird coinheringaahing

ಠ_ಠ TFW a music compilation calls a song from 2015 "Early 2010s"

Corrected

Mr. Xcoder

Best part is, I can now do Jelly on mobile :P

Jonathan Allan

10:04 PM

yeah but writing explanations is still hard on mobile, right?

caird coinheringaahing

Have you finished the replacements app?

Mr. Xcoder

Yeah, but the problem is that I can’t install it on my device. I have finished the Smart Keyboard, though

caird coinheringaahing

Have you added escapes?

Mr. Xcoder

@JonathanAllan yeah, but I was talking about CMCs

@cairdcoinheringaahing Not yet, but I finished the design part and the basic functionality

I should

make it more mobile-friendly

Jonathan Allan

you mean "if only..."?

CMC: given n output the product of the divisors of the nth composite number.

Mr. Xcoder

10:10 PM

@JonathanAllan ... If only SE used monospace fonts

Jonathan Allan

yep

@JonathanAllan starting with n=1 that would be [8, 36, 64, 27, 100, 1728, 196, 225, 1024, 5832, 8000...]

Mr. Xcoder

Can we start with 0 and 1 as composites?

Jonathan Allan

0 and 1 are not composite

Mr. Xcoder

I know, just asking :P

Jonathan Allan

composite is not defined as "not prime" :)

Mr. Xcoder

10:13 PM

Just a try :P

Jonathan Allan

part of what makes it worth a CMC I think

Mr. Xcoder

@JonathanAllan First try, dumb approach: Try it online! (10 bytes)

Again, 10 bytes

caird coinheringaahing

10 bytes here

Mr. Xcoder

Try it online! (10 bytes, AGAIN)

@JonathanAllan What’s your current byte count?

Jonathan Allan

um I had 10 but am cooking :)

Mr. Xcoder

10:24 PM

Also 10 bytes :/

Mr. Xcoder

10:37 PM

It's probably do-able in under 10, though.

Jonathan Allan

11:00 PM

TFW u think you got 9

Python Byte Shavers

Transcript for