The APL Orchard

5:11 AM

@LdBeth Yes, it was included in Sharp APL (and thus in J) as monadic dot product. See also dfns.dyalog.com/n_alt.htm

@LdBeth Collapse axes along their shared diagonal. See apl.wiki/Transpose#Duplicates_in_the_left_argument.

8 hours later…

1:08 PM

I'm looking at the Bunda-Gerth parser included in dfns; dfns.dyalog.com/s_Binding.htm

Can someone explain why the parse tree looks so different in these two simple cases:

      xdefs parse 'a←+ ⋄ a'
┌→────────┐
↓      A  │
│    ┌─┴─┐│
│  ┌─┴─┐ a│
│ ┌┴─┐ ⋄  │
│┌┴┐ +    │
│a ←      │
└─────────┘

      xdefs parse 'a←1 ⋄ a'
┌→───────────┐
↓┌───┬────┬─┐│
││ AF│ XLS│A││
││┌┴┐│┌┴┐ │a││
││a ←│1 ⋄ │ ││
│└───┴────┴─┘│
└────────────┘

First one is what I expected.

Second one is not.

Although the first one actually derives the wrong top category, as far as I can tell -- I reckon it should be F, like in the following:

      xdefs parse 'f←+ ⋄ g←f/ ⋄ g'
┌→──────────────────┐
↓              F    │
│           ┌──┴───┐│
│       ┌───┴────┐ g│
│    ┌──┴───┐    ⋄  │
│  ┌─┴─┐  ┌─┴─┐     │
│ ┌┴─┐ ⋄ ┌┴┐ ┌┴┐    │
│┌┴┐ +   g ← f /    │
│f ←                │
└───────────────────┘

1:46 PM

I think the final rule in the grammar is wrong.

AF also needs to bind to XLS

Adding AF:XLS→XLS to the last group yields

      xdefs parse 'a←1 ⋄ a'
┌→────────┐
↓     A   │
│   ┌─┴──┐│
│ ┌─┴─┐  a│
│┌┴┐ ┌┴┐  │
│a ← 1 ⋄  │
└─────────┘

Which is less wrong, possibly even right.

That looks right to me.

@xpqz If you send me a diff/patch file or just a few lines centred on the line that needs changing, I'll have it updated.

But maybe the tests should be amended too?

I am not sure it's correct, and unsure how to check fully; whilst it might fix this particular case, are there effects elsewhere?

Dunno, but you could run the test script.

Last time I tried this, it doesn't play ball on a mac :/

Ah right, because it needs Windows for some things. OK, send me the diff and I'll see.

2:04 PM

Right. Where does the following reside: scripts._dyalog?

I don't follow? #.scripts._dyalog in the dfns workspace?

@user1293156 Hi Gần Em Nhé Anh, if you want to participate here, please email access@apl.chat

@Adám Ok, sent the diff

Stack Exchange Feeds

2:44 PM

Q: Dyalog APL quad(NGET)

I am trying to access a native file using the quad NGET command on a Ubuntu Unix system. This works fine as long as the path/filename string does not contain blanks. If the path contains blanks the file is not recognized. What can I do to make it work?

dyalog

5 hours later…

7:42 PM

This one wasn't too complicated. If i got it right.
{(⍺×⌊0.5+⍺÷⍨⍵×m)+⍵×~m←2|⍵}
But your other Quest++ is.

@Adám This makes it difficult, at least I did not try to do it with Regex.

It is solvable as long as the single quoted text is not at the end or the beginning so I can check wether there is a space before or after the quote

Maybe I just shoudl add a space at the beginning and at the end, do the checks and then remove them

@Richard That only works on integers.

ok, back to the drawing board.

I think I misunderstood the question

I thought "substitute the uneven numbers with the closest multiple of the left argument"

7:58 PM

Oh no, I meant a generalised rounding function.

8:08 PM

So rounding each number to a multple of the left argument after increasing each uneven number with 1?

No, nothing about evens and odds, just rounding.

8:24 PM

{⍺×⌊0.5+⍵÷⍺}