oh if someone is already looking at it, i forgot to commint the latest version to github. so sorry for that

@BrianBED f&Y is f Y but in a background thread.

@Adám which apls have that?

Only Dyalog, afaik.

oh

oh nice thanks

hey, i'm looking for a way to extend this train

      2 3 (⊣×⊂⍤⊢) 4 5
┌────┬─────┐
│8 10│12 15│
└────┴─────┘

@justin2004 Press uparrow, Ctrl+k, Enter now.

to do this

      +/¨2 3 (⊣×⊂⍤⊢) 4 5
18 27

OK, what's the problem with +/¨⊣×⊂⍤⊢?

Btw, ⊣×⊂⍤⊢ can be written as ×∘⊂

i am going to use this expression in a stencil and i think i need the train instead of the expression with the each

No, I meant +/¨⊣×⊂⍤⊢ as a train: ⋄ 2 3 (+/¨⊣×⊂⍤⊢) 4 5

@Adám 18 27

oh, wow

However, this is just for the structure. For this particular computation +/∘.× is much better.

i hadn't been thinking of each as a train element for some reason. let me see if that works in my stencil.

Each isn't a "train element" here. +/¨ is a single monadic function.

ok. i think i haven't been thinking of each like i do the other operators.

Another alternative formulation is +.×⍤0 1 which isn't even a train.

@Adám is it called a derived function?

It is.

@justin2004 ∘ is a dyadic operator. f∘g is exactly like f except the right argument is pre-processed by g before f is applied.

ah, ok.

a f∘g x = a f g x

One of the many names for ∘ is therefore "after", because f∘g applies f after g.

@justin2004 if it were ×⍤⊂ then it will be used dyadically

And another name for it is "beside" for the reason rak1507 stated: It is equivalent to having f beside g in an explicit expression.

@PyGamer0 ah, that is helpful.

I like BQN's glyphs ⊸⟜ for before/after and ∘ for atop

I can never remember which one is in which.

gotcha. i think this illustrates that i have been call it "bind" and not fully understanding it. :P

wdym

I have a hard time remembering which of ⊸ and ⟜ to use in each situation.

While I agree that f∘g and f⍤g should have been swapped in Dyalog, now that it is what it is, ⍛ is a good symbol for "before".

And I like having the same symbol for binding arguments: A∘f and f∘A

dyalog optimises some special forms iirc

Yeah, but not this one, I think. However, try comparing {⍵×kernel}⌺3 3 with {kernel×⍵}⌺3 3

@Adám I guess you can think of ∘ in ⊸⟜ as the function and the leg - goes toward the argument

oh it's the other way round

@Adám regardless they're better than ∘⍛, beting at least horizontally asymmetric

@PyGamer0 But when there are two functions…

dyalog.com/blog/2020/06/towards-improvements-to-stencil @justin2004 have you read this?

@Adám i guess you are too used to APL \o/

@rak1507 i have not. this looks interesting! thanks

@PyGamer0 yeah. I really don't think that for a fresh learning experience there's anything better than ⊸⟜ possible

yeah i am not very used to APL so i like the BQN glyphs better

@Adám the dash side always goes to the function called monadically on its side argument, and the round side is always the dyadic function called with preprocessed arguments. Why would two functions change anything?

@dzaima I meant in contrast to function-with-array/array-with-function.

@Adám BQNs ⊸⟜ act exactly the same with function or value operands

Ah, right so I have to think of 4⊸F X as X {4}⊸F X

well, x {𝕊: 4}⊸F x because immediate blocks :)

⍨

@rak1507 Added to Stencil wiki page.