'Morning everyone o/ Hey @Adám, are you working on the WPA golf challenge?

@J.Salle No. I had a look, but I'm not going to read long RFC to know what to do. IMHO, the OP should contain everything necessary to solve the problem.

@Adám Yeah, I'm gonna comment something along those lines. Challenge is too language-restrictive, IMO

@J.Salle It isn't really language restrictive. It just uses complex algorithms which few languages have built-ins for. It would be more interesting if the OP prohibited built-ins for the main pieces.

@Adám my point exactly. Every golflang I know would have to create a SHA-1 algorithm, then build an HMAC function using that and then implement a solution to the challenge, it's just too convoluted

@Adám would f←{⍵=3∧⍵=19:doSomething⋄⍵} work as a function? I want to doSomething unless the input is either 3 or 19, in which case I just want to print the input.

Now that I have the mind cleared, I remember why I generated 10 "steps" instead of 9: I'm stupid and I misunderstood how interval index works :D.I'm making good progress again.

@J.Salle Yes, but remember APL's order of execution: Functions are right-associative, so the leftmost = sees 3∧⍵=19 as its right argument. Either use parentheses or use ∊.

@Ven Great to hear!

@Adám how would I use ∊? f←{(3∧19)∊⍵:doSomething⋄⍵}?

@J.Salle f←{⍵∊3 19:doSomething⋄⍵}

∊ is "member of"

Oh okay, so the ∧ is not needed at all? Nice. Thanks!

@J.Salle Correct. You only need ∧ if you have several very different conditions.

And can I nest those conditionals, or I'd need another function? doSomething would have to check whether the argument is a prime or not.

@J.Salle You can write multiple conditions with {(cond1)∧(cond2)∧cond3:true⋄false} but note that all conditional expressions will be evaluated.

@J.Salle You would have to use {} not () and you have to supply the argument(s) to the inner function as well.

{⍵∊3 19:{⍵∊<primes>:doSomethingElse⋄returnSomething}⍵⋄⍵}

@Adám I see! Thanks, I think I've got what I need now

@Adám Is this link up to date? To get the biggest prime factor of ⍵, all I need is to do is ⌈⍵.factors ?

actually I think ⌈⍵.factors will return all the factors anyway, no?

@J.Salle APL isn't JavaScript! ⌈/factors ⍵ or ⊃⌽factors ⍵

@Adám oh, okay! Thanks again hahahah

@J.Salle Btw, ⊃⌽3pco⍵ is shorter.

@Adám nice, I'll use it. Currently my function f←{⍵∊3 19:⍵⋄{⍵∊((⊢~∘.×⍨)1↓⍳):∇((2×⍵)-1)⋄∇(⍵-(⌈/factors ⍵))}⍵} raises a Syntax error on the first bracket after the ⋄, any ideas why?

@J.Salle ⍵∊((⊢~∘.×⍨)1↓⍳) is a function, not a Boolean singleton.

you probably mean ⍵∊((⊢~∘.×⍨)1↓⍳)⍵

Oooooooooh yeah I keep forgetting to pass the arguments there

thanks

@Adám heh in some golfing languages such as jelly (grand-son of apl) the function represents a boolean literal so it's a bit confusing

It's now raising a value error at the function you gave me: f←{⍵∊3 19:⍵ ⋄ {⍵∊((⊢~∘.×⍨)1↓⍳)⍵:∇((2×⍵)-1) ⋄ ∇(⍵-(⊃⌽3 pco ⍵))}⍵} it points at the 3 in ⊃⌽3 pco ⍵

@J.Salle Did you import pco?

but he said it points at the 3

@Adám I did not, is it still shorter than ⌈/factors ⍵ if I do?

@EriktheOutgolfer The caret points at where parsing stopped.

hmm

@J.Salle Where did you get factors from?

possibly since it doesn't know if pco is monadic or dyadic

Apparently it'll be because I also need to import factorseh? I got it from here

How do I import stuff? I'm still getting used to the dyalog ide for windows

...reading license agreement is boring :-(

@J.Salle The language is identical on all platforms.

@Adám nice, thanks!

Hm, I think my function is looping. It exceeded the TIO's time limit and I had to interrupt it in dyalog.

@J.Salle Are you sure the argument eventually reaches 3 or 19?

@J.Salle I would recommend defining the function on multiple lines, and then tracing into it.

Hello buddies!

@Adám the challenge says it does, and it appears to be something with the pco function

@Mr.Xcoder Hello to you, sir

Oh you're solving that challenge?

@Mr.Xcoder the frog one

@Adám the challenge lets you assume the tested/proven inputs btw

my code is currently stuck at f←{⍵∊3 19:⍵⋄{⍵∊((⊢~∘.×⍨)1↓⍳)⍵:∇((2×⍵)-1)⋄∇(⍵-(⊃⌽3pco⍵))}⍵}

Yup I know. I have an answer there as well

Pyth does have some convenient built-ins for that challenge but unfortunately other needed ones are quite lengthy

Does APL have a prime factors built-in?

I was wondering, what is the 3 doing in ⊃⌽3pco⍵?

@Mr.Xcoder you need to import it

Oh the pco module

? or something

@J.Salle pco can do a bunch of primes related things. The left argument choses which functionality to use.

@Mr.Xcoder I've found factors and Adam told me about pco

I see I can do ⎕CY'dfns'

To import it?

@Adám I see, but then ⊃⌽3pco26 should return 13, shouldn't it?

And then <number>pco⎕

@Mr.Xcoder NARS APL is an experimental APL which we kind of use as a sandbox for language ideas. It has π which is essentially pco as a primitive.

@J.Salle you need a space after o

I am looking for a CMC I can solve in APL, anyone up for creating one?

@Adám oh good lord I should've realized that

Got an 'Assertion Failure' exception now. Progress!

@J.Salle Huh? What did you do? Please give me the exact code which gave you the ASSERTION FAILURE.

@Adám Is that easy enough so I can solve it? Also must it be formatted exactly like that?

@Adám f←{⍵∊3 19:⍵⋄{⍵∊((⊢~∘.×⍨)1↓⍳)⍵:∇((2×⍵)-1)⋄∇(⍵-(⊃⌽3pco ⍵))}⍵} this is it. The exception is raised at the ( before ⊃⌽3pco ⍵

No, it should give an ⍵-by-⍵ matrix.

I've tested every part of the code by itself and they work separetely

@J.Salle Oh, it is lowercase, phew! Never mind then.

The debugger opens at pco[1] if that means anything

@Adám yeah it's lowercase hahahahah

@J.Salle (Uppercase would mean that the interpreter hit an assertion failure. That's bad.)

@Adám Ok, gonna try this one then.

@Adám I can imagine.

@Adám What is/(are) the "range" function(s)?

@Mr.Xcoder monadic ⍳

Oh right

@Adám my primes function (from the other day, (⊢~∘.×⍨)1↓⍳) doesn't work for some reason. If I do (⊢~∘.×⍨)1↓⍳⍵ I get no result for whatever ⍵ I input

lol nevermind, now it's working. No idea why it wasn't

Okay, I found out what was the problem: ∇ was calling the inner function instead of f, so it was just looping primes over and over. I solved it by making a g←{⍵∊(⊢~∘.×⍨)1↓⍳⍵:f((2×⍵)-1)⋄f(⍵-⊃⌽3pco ⍵)} to call inside f, but now I get a WS FULL exception raised when I try the last test case by calling f 9983

@J.Salle Can you send me a TIO link?

@Adám I managed to fix it in my computer by increasing the workspace size. Here's the link, TIO can handle f up to ~3200

@J.Salle Because your primality checking algorithm is really inefficient. Use 1pco⍵ instead of ⍵∊(⊢~∘.×⍨)1↓⍳⍵

@Adám oooh nice, I had no idea that existed hahahah thanks!

@J.Salle pco manual

@Adám thanks, I'll use it in the future hahahah

@Adám you found the nicer solution?

@Adám I have a little issue with my license

Karen approved it, but instead of a download link I got this:

did I accidentally apply for the commercial version?

I have replied to Karen for further information on that

in the url where I confirmed my email it had something along the lines of ?ltype=pe

@EriktheOutgolfer pe is personal license, so you should be OK. However, mistakes can of course happen. If you emailed Karen, she will get back to you. If you don't get a response tomorrow, let me know, and I'll talk to her. Otherwise, you can always call into the office. Karen is usually there 9-5 UK time.

@Mr.Xcoder Any success?

@Uriel I didn't try. I have too much "work" to do. Next week I'll be talking about code golf on dyalog.tv…

@Adám not yet. TBH I didn't really try it today. I'll tell you if I'm done with it, maybe tomorrow

@Mr.Xcoder No rush.

@Adám oo nice. good luck. I might be able to watch live

@Uriel Thanks. My first webcast ever :-O