f⍣k in the lesson title somewhat looks like f**k as in Brainf**k...

@Bubbler lol. i wonder if any apl programmer ever had to add an element k to a set f, and wrote it as (⊂k)∪⍨f to avoid embarrassment :)

Coincidence, I was trying to tacitise a function, and thinking "there's a pattern here, what's it called, over?" and here it is in the cultivation, "Under". ⌊Under (100∘×) 10.6789 truncate to 2 decimal places

ish. floor under power-of-n

⎕←2{(10*⍺)÷⍨⌊(10*⍺)×⍵}10.6789

@TessellatingHeckler
10.67

trying to make that tacit, while there's that repeated mention of ⍺, looking at the conversion rules.. is this a typo? in dfns.dyalog.com/n_tacit.htm - should it be {⍵}

@TessellatingHeckler Indeed. I spotted that typo the other day too.

@Adám good good, it's enough already without having a special case for if ⍵ is caught in some mismatched brackets :P

@TessellatingHeckler Did you see the webinar on tacitising?

@Adám Yes indeed! Very interested in the extension of '-'(beaten-up-face)'string-here' to partition with multiple split characters, and in the ∨¨ --- ∊¨ --- pattern which I had found, changing to --- (∨/∊)¨ ----, which I hadn't

I can make a tacit version 2((10*⊣)÷⍨(⌊(10*⊣)×⊢))10.6789

@TessellatingHeckler You can swap the arguments of ×, it being commutative.

@TessellatingHeckler And compose ÷⍨∘⌊

You can even assign 10*⊢ a name and reuse it.

but am trying for 2(10*⊣)(÷⍨(⌊×))10.6789

ah I had it as (100×⊢) while working and that only forks with the constant on the left, so that's a good reminder

@TessellatingHeckler You want to "pre-process" the left argument. That's why I want ⍛ which is like ∘ but mirrored.

@Adám I do want to preprocess the left, and even parens don't seem to quite do what I want

@Adám why the underscore? Where else does that imply "mirrored"?

@TessellatingHeckler With the current set of operators, you can only pre-process the right argument, so to preprocess the left, you need main⍨∘pre⍨

@TessellatingHeckler Nowhere. But the ∘ symbol can't be mirrored. If I had the possibility of renaming all the operators, I wouldn't have made ∘ X f g Y

@TessellatingHeckler So, I have p÷⍨∘⌊⊢×p←10*⊣ or (⊢÷⍨∘⌊×)∘(10∘*)⍨ if you don't want the assignment.

@Adám ⌽ and ⍨ say mirroring to me, but neither lends itself to a small circle without confusion. I don't get the point of jot, really

@TessellatingHeckler That's my point. ∘ is a symmetric symbol, ill suited for its assymetric functionality. Iverson actually used the symbol ⍩ for this. If that had cought on, then the it's mirrored twin's symbol would have been given.

@Adám I follow the main and the pre⍨ parts, and was trying to work out the difference between (10∘*)⍨ and 10∘*⍨ and I think I get it

@TessellatingHeckler There's no difference between the two in isolation, but without parens, ∘ will bind the 10 to its left operand.

@Adám just realised I wasn't getting it because I can't treat that bit in isolation because the ()∘() pattern makes a single giant function, which is then commuted

@Adám Why does removing the jot 2 ((⊢÷⍨∘⌊×)(10∘*))⍨ 10.6789 throw a syntax error that the function does not take a left argument? I see without the jot it's no longer a single function, but then what function is the outer commute binding to?

oh!

@TessellatingHeckler The out commute operates on the atop (⊢÷⍨∘⌊×)(10∘*)

I've asked you this before in another form; it is parsing as a train, an atop, so it's doing the wrong thing with the arguments

Which means that first 10∘* is applied, and that is a strictly monadic function.

it's trying to do 10.6789 (10∘*) 2 first, yess got it

It is SO common to confuse dyadic f∘g with an atop. That's what ∘ should have been.

Luckily, in 18.0 we're getting f⍤g which can replace f∘g wherever it actually is an atop, that is, when monadic.

The difference between the ambivalent function f∘g and the ambivalent function f⍤g is that the optional left argument becomes the left argument of f in f∘g and of g in f⍤g.

@Adám beginners confuse everything with everything, eh

Nah, f∘g with f g is notorious.

@Adám - I don't know how long your commute is, but with sunset in less than two hours (for you), let me wish you a Good Shabbos!

@ngn this looks a little too long

@H.PWiz it does

@H.PWiz what's your best?

@JeffZeitlin Thank you. It is about 2 hours, but I went home Thursday. Thanks. To you too?

@ngn ⍴⍴⍳∘≢∘, → ≢⊥¨⍳∘⍴

@Adám - Thank you; I am Jewish, but not Orthodox, and ... shall we say, somewhat lax in observance?

@JeffZeitlin OK, that's important for me to know, actually. Gut shabbos to you then!

Is there a short way to double a string? (eg: 'abc' -> 'abcabc')

@EdgyNerd ,⍨

oh ok

@ngn Same as your's