Well said. Why don't you put that into the NuVoc page for Key at an
appropriate place?
For 9.01 and later, prefer u. to u .
You might need some examples of sc.
Henry Rich
On 10/14/2019 5:35 PM, Louis de Forcrand wrote:
If this makes it any clearer, here is an implementation of /. key:
key=: 1 : 0
=@[ u@# ]
)
where = self-classify can itself be defined as
sc=: ~. -:"_ _1"_1 _ ]
Cheers,
Louis
On 14 Oct 2019, at 21:26, Raul Miller <[email protected]> wrote:
I'm not sure what you mean, but consider:
'abcba' </. 'abcde'
+--+--+-+
|ae|bd|c|
+--+--+-+
Basically, any monadic verb that can handle lists of items from the
right argument will work with /. (but keep in mind that if the results
have differing lengths, the short results will be padded when
assembling the result).
Thanks,
--
Raul
On Mon, Oct 14, 2019 at 3:18 PM 'Jim Russell' via Programming
<[email protected]> wrote:
Yes, much. That seems not the case for a character array left argument, thus my
confusion. Thank you!
What other arguments/verbs play nice with /, ?
On Oct 14, 2019, at 3:05 PM, Raul Miller <[email protected]> wrote:
Ok, try this:
1 2 3 2 1 </. 'abcde'
+--+--+-+
|ae|bd|c|
+--+--+-+
The nub of the left argument determines the number of items in the
result, but it's not sufficient to determine what's in each box.
Is this clearer?
Thanks,
--
Raul
On Mon, Oct 14, 2019 at 1:24 PM 'Jim Russell' via Programming
<[email protected]> wrote:
Sorry, I'm still stuck on u=#. (Or "," or "<" or "[" ). I'll try to understand
some harder working verbs, perhaps + or ! or I. or e?
On Oct 14, 2019, at 12:18 PM, Raul Miller <[email protected]> wrote:
On Mon, Oct 14, 2019 at 12:06 PM 'Jim Russell' via Programming
<[email protected]> wrote:
Thank you Raul! I thought I was agreeing, until I got to the following:
On Oct 14, 2019, at 11:40 AM, Raul Miller <[email protected]> wrote:
So ... it's the "only the nub" which is relevant here, if you do not
care about the calculation which produces the result.
What calculation? What would be different if (I was allowed to) just supply
(~.x) as a left argument? (Assuming I didn't care to waste the cycles
calculating the nub when it, at least its size, is going fall out as a result
of looking up each of the eight argument rows).
But I'll accept that I'm missing something; the alternate that Roger suggested
looks like it should do exactly what I was looking for.
The X U/.Y calculation.
The order and values of X matter for that calculation, since they control it.
Thanks,
--
Raul
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
---
This email has been checked for viruses by AVG.
https://www.avg.com
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
