@Adám will "depth" always be equal to "tally shape"?

@Uriel No, ≢⍴ is rank, not depth.

'Morning everyone o/

⍞←'Good morning, J. Sallé!'

@Adám Good morning, J. Sallé!

Hey @Adám! Would you mind explaining to me how does the code you wrote for the QuadR question I asked yesterday works? I know \d+. is the regex expression, but what is ¯1((⍎↓)⍴↑)⍵M doing?

I also know ⍵M is the match

@J.Sallé It is a fork (3-train), equivalent to ¯1 {(⍎⍺↓⍵) ⍴ ⍺↑⍵} ⍵M

@Adám So it's dropping the last part of the match to get the numeric value and transforming it to a number n, then it reshapes the last part of the match, which is the symbol, to repeat it n times, correct?

@J.Sallé Yes, exactly. Try e.g.

⍞←¯1((⍎↓)⍴↑)'10v'

@Adám vvvvvvvvvv

I see, that's good. Do you mind if I post it as an answer? I was trying to solve it yesterday but I couldn't figure QuadR out at all :p

I was trying to use groupings and \1, \2, etc.

@J.Sallé I don't mind at all.

Thanks!

You can use groups too: (\d+)(.) and then digging them out from ⍵B, but it will be much longer.

@Adám could you give this answer a quick check? I'm just wondering if I got lucky using the logical NAND operator or if that's what it's supposed to do

@J.Sallé function

@J.Sallé No, you have the inverse of the desired result. Just use ∧.

@Adám I tried using ∧ but it wasn't working! For some reason it was returning 1 on some of the falsy test cases

I just checked and it's working now on TIO. That's weird

I'm sure it wasn't working because I tested every one of the logical functions and NAND was the closest to the actual result.

@J.Sallé dfns originally meant dynamic functions. See footnote here.

@Adám oh, that's cool. The 24-byter you posted is a tradfn, right? Does it count the ∇f and ∇?

@J.Sallé Yes, it is a tradfn. The ∇ certainly don't count, as they are not part of the function at all, but rather escape characters for TIO's interface. I don't count the function header if the function takes care of I/O (like this one does), as the header doesn't do anything, and could be omitted if the code was instead piped into APL (the Input field in TIO).

@Adám I see. I'll update my answer there c:

#TIO permissions

@ThomasWard
Room Owners and Moderators:
10145 (Thomas Ward)
130368 (Adám)

User Granted Permission:
226059 (Socratic Phoenix)

@Adám ^ for context

@SocraticPhoenix ^ also

@ThomasWard Right now I'm auto-granted permission in every room... I'm going to change that, but it's nice for testing :)

@SocraticPhoenix indeed. But for production... :P

That's a great bot, btw guys. What's it written in?

@J.Sallé Java :-)

and I'm working on a Python fork. :P

Nice, very nice.

@Adám why does that work? I think I still can't get ∘ figured out

@J.Sallé Yeah, dyadic application of functions derived with ∘ are wonderful beasts.

⎕←×∘?∘0

@Adám
┌───────┬─┬─┐
│┌─┬─┬─┐│∘│0│
││×│∘│?││ │ │
│└─┴─┴─┘│ │ │
└───────┴─┴─┘

@J.Sallé So you can see that the leftmost ∘ binds first (of course, since operators do that). It makes a function which is equivalent to {⍺×?⍵}.

@Adám that makes it surprisingly easier to understand, actually

Now that function is dyadic, but gets a right argument bound to it with ∘0 so now you have {⍺×?⍵}∘0 which is the same as {⍵×?0}

@Adám oooooooooooooooh so this is the catch

I like it.

@J.Sallé :-D You have to remember that ∘ is asymmetric. Iverson originally gave it the symbol ⍩ and called it "with" which maybe makes more sense: (times with roll) with 0

@Adám indeed it does make more sense

@Adám TIOBot v 0.0.3

TIOBot logged in!

#tio version

@Adám TIOBot v 0.0.4

#tio alias view

@Adám
Command Aliases:
python -> #TIO run python3 %args%
⍞← -> #TIO do apl-dyalog ⎕←%args%
⎕← -> #TIO run apl-dyalog (⎕NS⍬).⍎'⎕CY''salt''⋄⎕SE.UCMD''box on -fns=on''⊣enableSALT'⋄⎕←%args%

Language Aliases:
apl-dyalog -> [apl]

#tio alias rcommand python

@Adám removed alias for "python"