The APL Orchard

Bubbler

04:09

"Pi times" in action!

Sherlock9

05:29

@dzaima I can't find ⍋ and ⍒ in your Android IDE. Are they somewhere in one of the other menus?

James Heslip

07:27

Not sure if I'm being stupid here- anyone got any example of writing data (matrix, was previously saving in Excel) to pdf? Or converting the Excel worksheet/csv to pdf instead?

Bubbler

Looks like Excel itself can export part of a document to pdf...

James Heslip

@Bubbler was thinking about skipping the Excel step and going straight from memory to producing a PDF.

Sorry, I really am stupid, haven't had a coffee yet! Thanks- will try this out.

Bubbler

Maybe it's also possible in R, JS, or Python; didn't check the details

James Heslip

07:43

Python sounds like more fun. I hate working with Excel, would prefer to grab a library if possible. Will report back.

Bubbler

08:01

TIL J's dyadic bind can lead to some insane golfs

TessellatingHeckler

08:33

⎕←¯2 (⌽) 5 5 ↑ 1

Dyalog APL

@TessellatingHeckler
0 0 1 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0

TessellatingHeckler

a 1, embedded in a 5x5 matrix, rotated two to the right, but if I try to center it in a single move, I can't get anything which works

⎕←¯2 (⌽⊖) 5 5 ↑ 1

Dyalog APL

@TessellatingHeckler
0 0 0 0 0
0 0 0 0 0
0 0 0 0 1
0 0 0 0 0
0 0 0 0 0

Bubbler

⎕←¯2 (⊣⌽⊖) 5 5 ↑ 1

Dyalog APL

Bubbler

08:35

@TessellatingHeckler Do you want this?

TessellatingHeckler

Yes!

@Bubbler and that does work for my use case, which is great, thanks :)

(but .. whyyyyyy)

⎕←{∨⌿(0,¯1↓-+\⍵) (⊣⌽⊖)⍤(0 2) (2/+/⍵)↑⍤2 1 ⍴⍤(1 0)⍨ 2/⍪⍵} 2 4 1 3

Dyalog APL

@TessellatingHeckler
RANK ERROR

TessellatingHeckler

d'oh, you don't speak NARS2000, do you, Dyalog APL bot

nevermind

⎕←{∨⌿(0,¯1↓-+\⍵) (⊣⌽⊖)⍤(0 2) ⊢ (2/+/⍵)↑⍤2 ⊢ 1 ⍴⍤(1 0)⍨ 2/⍪⍵} 2 4 1 3

Dyalog APL

@TessellatingHeckler
1 1 0 0 0 0 0 0 0 0
1 1 0 0 0 0 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 0 0 0 0 1 0 0 0
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 1

TessellatingHeckler

there, a 2x2 and a 4x4 and a 1x1 and a 3x3, in a diagonal line, from the input on the right, rewritten to have no "each" in it; #chuffed

Bubbler

08:57

⎕←{∘.=⍨⍵/⍳≢⍵} 2 4 1 3

Dyalog APL

@Bubbler
1 1 0 0 0 0 0 0 0 0
1 1 0 0 0 0 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 0 0 0 0 1 0 0 0
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 1

Bubbler

⍨

Adám

09:27

@Bubbler Fixed. Thanks!

TessellatingHeckler

09:38

XD

I assumed it could be better, I'm surprised it can be that much better

dzaima

@Sherlock9 I did add them but haven't made an apk. i'll also add ⍎⍕ and on landscape ∆∇ and make one

Sherlock9

Thanks in advance! :D

TessellatingHeckler

⎕←∘.=⍨(/⍨) 2 4 1 3

Dyalog APL

@TessellatingHeckler
1 1 0 0 0 0 0 0 0 0
1 1 0 0 0 0 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 0 0 0 0 1 0 0 0
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 1

TessellatingHeckler

nice approach, @Bubbler

so much simpler

Bubbler

09:50

⎕←∘.=⍨(/⍨) 2 2 1 3

Dyalog APL

@Bubbler
1 1 1 1 0 0 0 0
1 1 1 1 0 0 0 0
1 1 1 1 0 0 0 0
1 1 1 1 0 0 0 0
0 0 0 0 1 0 0 0
0 0 0 0 0 1 1 1
0 0 0 0 0 1 1 1
0 0 0 0 0 1 1 1

Bubbler

That one breaks on inputs with duplicates

TessellatingHeckler

oh ok, that's what the iota was guarding against. Maybe I should stick to my XML encoding problems.

⎕←(⊢/⍳∘⍴) 2 4 1 3

Dyalog APL

@TessellatingHeckler
4

TessellatingHeckler

?_? why doesn't this tacit version make the (2 4 1 3) replica of (1 2 3 4)

dzaima

10:03

@TessellatingHeckler because tacit hates / as it can be both an operator and a function, and it's choosing to be an operator here, hence ((⊢/) ⍳∘⍴)

TessellatingHeckler

@dzaima oh goodness, the right reduce, of course

⎕←∘.=⍨(⍳∘⍴(/⍨)⊢) 2 4 1 3

Dyalog APL

@TessellatingHeckler
1 1 0 0 0 0 0 0 0 0
1 1 0 0 0 0 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 1 1 1 1 0 0 0 0
0 0 0 0 0 0 1 0 0 0
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 1

Bubbler

Btw, I was looking at this challenge, thought ⌹ can be of some use, but then realized that the coefficient matrix is singular so I can't use it

Looks like a nice backtracking problem though

dzaima

10:25

@Sherlock9 released. also now has app.cpy which will copy ⍵ to clipboard so it's finally possible to copy stuff out of the app

Adám

@Bubbler Can you use 8415⌶?

TessellatingHeckler

(I made it into =/~(]#i.&#) 2 4 1 3 in J)

Bubbler

10:45

@Adám I don't know much about SVD, but the task is related to finding an inverse modulo 2 (not just an inverse).

rcabaco

13:04

⎕←(⊢⊆⍨∘~∘⊢∊∘'=&')'a=1&c=1-232589&c=1-232435&c=3-2317078'

Dyalog APL

@rcabaco
┌─┬─┬─┬────────┬─┬────────┬─┬─────────┐
│a│1│c│1-232589│c│1-232435│c│3-2317078│
└─┴─┴─┴────────┴─┴────────┴─┴─────────┘

rcabaco

can i use ⌸ to group the arguments here, or do i need to do a reshape?

Adám

@rcabaco What do you mean by "group"?

rcabaco

i am trying to find all the values for the key 'c', for example

or i could find the indexes of occurrences of 'c' and do a +1 do get the values

Adám

@rcabaco Maybe something like this?

⎕←⊃{⍺⍵}⌸/↓⍉↑'='(≠⊆⊢)¨'&'(≠⊆⊢) 'a=1&c=1-232589&c=1-232435&c=3-2317078'

Dyalog APL

13:07

@Adám
┌───┬─────────────────────────────┐
│┌─┐│┌─┐                          │
││a│││1│                          │
│└─┘│└─┘                          │
├───┼─────────────────────────────┤
│┌─┐│┌────────┬────────┬─────────┐│
││c│││1-232589│1-232435│3-2317078││
│└─┘│└────────┴────────┴─────────┘│
└───┴─────────────────────────────┘

rcabaco

@Adám yes! i need to dissect this. thank you

Adám

@rcabaco Do you know that all parameters will be single-letter?

rcabaco

yes

i control that

Adám

13:49

@rcabaco So:

⎕←(⊃¨{⍺⍵}⌸2∘↓¨)'&'(≠⊆⊢) 'a=1&c=1-232589&c=1-232435&c=3-2317078'

Dyalog APL

@Adám
┌─┬─────────────────────────────┐
│a│┌─┐                          │
│ ││1│                          │
│ │└─┘                          │
├─┼─────────────────────────────┤
│c│┌────────┬────────┬─────────┐│
│ ││1-232589│1-232435│3-2317078││
│ │└────────┴────────┴─────────┘│
└─┴─────────────────────────────┘

Adám

Announcing: The new help system (feedback welcome)

J. Sallé

@Adám Looks very Dyalog-y. I like it already, will use in the near future for sure.

I only took a peek at it, but I already like that it's very easy to find all the UCMDs

Adám

@J.Sallé I think the biggest benefit is that a link to a specific page now includes navigational controls.

@J.Sallé User commands? How?

dzaima

@Adám yay permalinks in the url :D

Adám

13:56

@dzaima Yes, much easier than having to open the frame.

Also, it appears (I don't have a mobile device) more mobile friendly too.

dzaima

@Adám indeed it does seem so

rcabaco

@Adám The help looks good. For small screens it might help for the breadcrumbs to be links

J. Sallé

@Adám I meant System Commands

All those obscure quads I never heard of

Commands and Functions, actually

Adám

I think the content is pretty much as before, but if I understand right, more content is coming to the new system, with pdf documentation being ported to HTML.

dzaima

also damn html/css

J. Sallé

14:04

Both the quads and the )s

Adám

@dzaima I already complained about that. We should give Fiona (she has been working on the new system in the background for years) one of these:

rcabaco

14:20

@Adám in the last solution you gave i understand what each part is doing but i am confused about something

(⊃¨{⍺⍵}⌸2∘↓¨)'&'(≠⊆⊢) parameters

2∘↓ receives as the right argument the partition

ahhh

⊃¨ also receives the partition as right argument

yes?

Adám

@rcabaco Yes, that parenthesis is simply an fgh.

rcabaco

⎕←(⊃¨{⍺⍵}⌸2∘↓¨)

Dyalog APL

@rcabaco
  ┌───┼────┐
  ¨   ⌸    ¨
┌─┘ ┌─┘  ┌─┘
⊃   {⍺⍵} ∘
        ┌┴┐
        2 ↓

Adám

@Ven Done and done. (CC: @dzaima)

Ven

@Adám yay!

@Adám search works again! woo

ngn

15:07

@TessellatingHeckler =/~&I.

Bubbler

23:51

For me, the contents tab doesn't load, and seems to fail with:

Uncaught Error: Load timeout for modules: Data/Tocs/Dyalog_Chunk5.js
http://requirejs.org/docs/errors.html#timeout
at C (require.min.js:13)
at D (require.min.js:36)
at require.min.js:37

Transcript for