I missed your point, didn't I, Brian? I was thinking of 5!:2 when I answered.
Atomic representation (5!:1) does at least offer some sort of diagnosis of baa, for any chosen (baa). Maybe not enough for my purpose, but better than nothing. On Fri, Jul 20, 2012 at 4:01 PM, Ian Clark <[email protected]> wrote: > Yes I have. But I was hoping to exclude constants and primitives > (whose type can be recognised at face value) by looking at phrases > which 5:!6 explicitly parenthesizes. Nodes of the nesting tree in > boxed format don't have this property. My conjecture is that the only > non-verbs which get parenthesized by 5!:6 are constants specified > using (i.) or ({a.). > > BTW I was fatigued and talking nonsense about (-~) because that's not > a non-verb. > > Since you could argue this problem is largely solved in Art's and > Ambrus's scripts anyway, not to mention trace.ijs, I need to explain > that I'm feeling my way towards a far simpler algorithm for > explication (ie tacit-->explicit). > > On Fri, Jul 20, 2012 at 12:42 PM, Brian Schott <[email protected]> wrote: >> Have you tried this? >> >> 5!:1 <'foo' >> ┌──────────────────────────────────────┐ >> │┌─┬──────────────────────────────────┐│ >> ││3│┌─────────────┬───────────────┬──┐││ >> ││ ││┌─┬─────────┐│┌─┬───────────┐│*:│││ >> ││ │││0│0 1 2 3│││.│┌───────┬─┐││ │││ >> ││ │││ │4 5 6 7│││ ││┌─┬───┐│*│││ │││ >> ││ │││ │8 9 10 11│││ │││/│┌─┐││ │││ │││ >> ││ ││└─┴─────────┘││ │││ ││+│││ │││ │││ >> ││ ││ ││ │││ │└─┘││ │││ │││ >> ││ ││ ││ ││└─┴───┘│ │││ │││ >> ││ ││ ││ │└───────┴─┘││ │││ >> ││ ││ │└─┴───────────┘│ │││ >> ││ │└─────────────┴───────────────┴──┘││ >> │└─┴──────────────────────────────────┘│ >> └──────────────────────────────────────┘ >> >> >> -- >> (B=) >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
