<"_1"_1 i.2 3
gets parsed as
(<"_1"_1) (i.2 3)
- Original Message -
From: Linda Alvord
To: programm...@jsoftware.com
Cc:
Sent: Friday, May 8, 2015 11:54 PM
Subject: Re: [Jprogramming] Puzzle: nontransitive rank
Pascal, You show:
<"
7, 2015 6:48 PM
To: programm...@jsoftware.com
Subject: Re: [Jprogramming] Puzzle: nontransitive rank
an example of your 2nd and 3rd reason
<"_1 i.2 3
┌─┬─┐
│0 1 2│3 4 5│
└─┴─┘
<"_1"_1 i.2 3
┌─┬─┬─┐
│0│1│2│
├─┼─┼─┤
│3│4│5│
└─┴─┴─┘
- Original Message --
min' via Programming
wrote:
> is there a reason to call
>
> (+" 0 _1) b. 0
> _ 0 _
>
> a feature?
>
> why should it not be
>
> _ 0 _1
>
> ?
>
>
>
> - Original Message -
> From: Marc Simpson
> To: Programming forum
> Cc:
>
a reason to call
>
> (+" 0 _1) b. 0
> _ 0 _
>
> a feature?
>
> why should it not be
>
> _ 0 _1
>
> ?
>
>
>
> - Original Message -
> From: Marc Simpson
> To: Programming forum
> Cc:
> Sent: Thursday, May 7, 2015 4:33 PM
> Sub
rogramming wrote:
> is there a reason to call
>
> (+" 0 _1) b. 0
> _ 0 _
>
> a feature?
>
> why should it not be
>
> _ 0 _1
>
> ?
>
>
>
> - Original Message -
> From: Marc Simpson
> To: Programming forum
> Cc:
> Se
is there a reason to call
(+" 0 _1) b. 0
_ 0 _
a feature?
why should it not be
_ 0 _1
?
- Original Message -
From: Marc Simpson
To: Programming forum
Cc:
Sent: Thursday, May 7, 2015 4:33 PM
Subject: Re: [Jprogramming] Puzzle: nontransitive rank
Or indeed
enigma=. +
com/help/dictionary/d600xv.htm
> >>
> >> the property of enigma is that when operated at its "native rank" it
> >> returns the first result, but
> >>
> >> enigma"enigma applies enigma at enigma's rank, and so how could this be
esult, but
>>
>> enigma"enigma applies enigma at enigma's rank, and so how could this be
>> different?
>>
>>
>> ----- Original Message -
>> From: Henry Rich
>> To: Programming forum
>> Cc:
>> Sent: Thursday, May 7, 2015 3:25 PM
>&
: Henry Rich
> To: Programming forum
> Cc:
> Sent: Thursday, May 7, 2015 3:25 PM
> Subject: [Jprogramming] Puzzle: nontransitive rank
>
> 1 2 enigma 1 2
> 2 4
> 1 2 enigma"enigma 1 2
> 2 3
> 3 4
>
> enigma is a short verb. What might it be?
>
&
t?
>
>
> - Original Message -
> From: Henry Rich
> To: Programming forum
> Cc:
> Sent: Thursday, May 7, 2015 3:25 PM
> Subject: [Jprogramming] Puzzle: nontransitive rank
>
> 1 2 enigma 1 2
&g
nt?
- Original Message -
From: Henry Rich
To: Programming forum
Cc:
Sent: Thursday, May 7, 2015 3:25 PM
Subject: [Jprogramming] Puzzle: nontransitive rank
1 2 enigma 1 2
2 4
1 2 enigma"enigma 1 2
2 3
3 4
enigma is a short verb. What
oops..overlooked that.
On Thu, May 7, 2015 at 3:40 PM, Henry Rich wrote:
> That doesn't explain the first line:
>
> 1 2 enigma 1 2
> 2 4
>
> Henry Rich
>
> On 5/7/2015 3:34 PM, Kenneth Lettow wrote:
>
>> enigma=: +/
>>
>> 1 2 enigma"enigma 1 2
>>
>> 2 3
>>
>> 3 4
>>
>>
>>
>>
>> On Thu, M
That doesn't explain the first line:
1 2 enigma 1 2
2 4
Henry Rich
On 5/7/2015 3:34 PM, Kenneth Lettow wrote:
enigma=: +/
1 2 enigma"enigma 1 2
2 3
3 4
On Thu, May 7, 2015 at 3:25 PM, Henry Rich wrote:
1 2 enigma 1 2
2 4
1 2 enigma"enigma 1 2
2 3
3 4
enigma is a shor
enigma=: +/
1 2 enigma"enigma 1 2
2 3
3 4
On Thu, May 7, 2015 at 3:25 PM, Henry Rich wrote:
>1 2 enigma 1 2
> 2 4
>1 2 enigma"enigma 1 2
> 2 3
> 3 4
>
> enigma is a short verb. What might it be?
>
> Henry Rich
> --
1 2 enigma 1 2
2 4
1 2 enigma"enigma 1 2
2 3
3 4
enigma is a short verb. What might it be?
Henry Rich
--
For information about J forums see http://www.jsoftware.com/forums.htm
) (L:0)) :. brxinv
- Original Message -
From: 'Pascal Jasmin' via Programming
To: Programming forum
Cc:
Sent: Tuesday, April 21, 2015 4:16 PM
Subject: [Jprogramming] :. puzzle (bug?)
brxinv =: ([: (x: inv :: ] (L:0)^:IF64) (3!:2))
brx =: 1 (3!:1) x: :: ] (L:0) :. brxinv
brxinv =: ([: (x: inv :: ] (L:0)^:IF64) (3!:2))
brx =: 1 (3!:1) x: :: ] (L:0) :. brxinv
intent is to get portable 32bit representation of non-extended data
on 64 bit machine, brx looks right (ie :. rightside looks the same)
brx
(1) 3!:1 x: ::]L:0 :.([: (x:^:_1 ::]L:0^:1) 3!:2)
brxinv
[:
You and I have different definitions of cheesy.
Anyway, I wrote:
> Here's a silly variation.
>3 ([ p:^:_1&|: p:\) i.7
> BTW, if the bug in [1] were fixed, we could've written:
>3 ([ p:^:_1 p:\) i.7
And if we fixed the bug and further extended the definition of p: in a
consistent and com
If you want cheesier, you might try:
ps3p=: 30 105 385 1001 2431 4199 7429"1
--
Raul
On Sat, Dec 7, 2013 at 5:22 PM, Dan Bron wrote:
> I wrote:
>> Here's a silly variation.
>> 3 ([ p:^:_1&|: p:\) i.7
>
> BTW, if the bug in [1] were fixed, we could've written:
>
>3 ([ p:^:_1 p:\) i.7
>
>
I wrote:
> Here's a silly variation.
> 3 ([ p:^:_1&|: p:\) i.7
BTW, if the bug in [1] were fixed, we could've written:
3 ([ p:^:_1 p:\) i.7
Is there any nobler reason to improve the language than to make jokes even
cheesier?
-Dan
[1] http://jsoftware.com/jwiki/System/Interpreter/Requests#r
Here's a silly variation.
3 ([ p:^:_1&|: p:\) i.7
30 105 385 1001 2431
[Hint: the middle tine can be greatly simplified :]
-Dan
PS: I'll gladly cede the title of Stupidest J Pun to anyone who can say it
cleverly using &. . Not sure it'll improve your resume, but you can always
break it out
Nice!
Here's a minor variation:
*/@:p:@(0 1 2 +/ ]) i.7
30 105 385 1001 2431 4199 7429
:)
--
Raul
On Sat, Dec 7, 2013 at 2:03 PM, Sebastiano Tronto
wrote:
> Hi, I'm new to the forums.
> I love this kind of puzzles (in fact, that's kind of why I learnt J) :-)
>
> Solution:
> p3sp =: */@:p
Hi, I'm new to the forums.
I love this kind of puzzles (in fact, that's kind of why I learnt J) :-)
Solution:
p3sp =: */@:p:@(],>:,2&+)"0
Didn't use oeis :-)
Inviato con AquaMail per Android
http://www.aqua-mail.com
Il 07 Dicembre 2013 19:36:49 km ha scritto:
Figure out what verb p3sp does a
Figure out what verb p3sp does and write it. The name is a hint. Extra credit
for not using
https://oeis.org/
--Kip Murray
p3sp i. 7
30 105 385 1001 2431 4199 7429
Sent from my iPad
--
For information about J forums see ht
Note that ?. might give different results in different versions of J.
It's designed to be a convenience, nothing more.
Thanks,
--
Raul
On Sat, Feb 16, 2013 at 12:06 PM, Mike Day wrote:
> Yes! Similar results here, though I'm a bit puzzled that graph3 doesn't
> seem to yield quite the same co
55.3 GB sounds like disk space
I am not even certain that memory pages have a fixed size in iOS, so I
cannot interpret the numbers about pages.
I often use the 6!: foreigns to deal with memory issues.
But I'm comfortable at this point, with the idea that I have
overreached myself with t his proj
Yes! Similar results here, though I'm a bit puzzled that graph3
doesn't seem to yield quite the same counts:
graph3=:0=?.2000 2000$1000 NB. try to reproduce Raul's latest
graph, but with results inconsistent with his!?
timer'(,:#@(] graph3 wrap & 0)"0) ]>:i.6'NB. using Raul's wrap.
Raul,
There is a time and space report below for "four moves", failure came on "five
moves".
My iPad reports 55.3 GB available memory. A crash report for the "five moves"
failure says reserved pages for the J process were 73234, "recent max" was
73253. There may have been an interruption fro
Actually, thinking about this, a path length of 100 is probably
unreasonable to measure precisely.
Here's an example case:
graph=: 0=?.2000 2000$1000
$I.1 graph wrap 0
1 1
$I.2 graph wrap 0
4 2
I'm being experimental here, reissuing lines, and it's inconvenient to
edit in the middle of the
It's hard for me to understand this "out of memory" state without
knowing how much memory your machine has, or how much gets used in the
successful cases.
Thanks,
--
Raul
On Sat, Feb 16, 2013 at 1:03 AM, km wrote:
> iPad time and space results for Kip's code. Brian Schott tells me correctly
iPad time and space results for Kip's code. Brian Schott tells me correctly
that verb open can be replaced by verb Raze ;
ones =: ; pairs 1 = ?. 200 200 $ 100 NB. paths with one move
# ones
390
5 {. ones NB. each pair represents one move
1 8
1 9
2 41
2 44
2 84
ts ';@:next^:3 one
OK, I think I've got it, thanks. I was using -- for a bidirectional
connection. --Kip
Sent from my iPad
On Feb 15, 2013, at 7:47 PM, Raul Miller wrote:
> On Fri, Feb 15, 2013 at 7:23 PM, km wrote:
>> My code is for directed graphs as defined in Graphs in Computer Science:
>>
>> http://w
On Fri, Feb 15, 2013 at 7:23 PM, km wrote:
> My code is for directed graphs as defined in Graphs in Computer Science:
>
> http://web.cecs.pdx.edu/~sheard/course/Cs163/Doc/Graphs.html
This is consistent with my intent when I used the phrase.
> However, suppose our undirected graph is
>
> 0--1--2-
My code is for directed graphs as defined in Graphs in Computer Science:
http://web.cecs.pdx.edu/~sheard/course/Cs163/Doc/Graphs.html
However, suppose our undirected graph is
0--1--2--3 (the only edges are those shown).
Am I right that every set of two nodes is visitable from node 2? These ar
wrapc does look plausible - to avoid memory problems we could break
the processing of the problematic line so it chews on smaller chunks
of data.
And, I am expecting that one step of iteration have a computational
cost which is a multiple of the square of the number of distinct nodes
reached so fa
Yes: for me, since I was just interested in knowing the unique sets of
visitable nodes with a population of N, I would treat 1 8 6 and 1 6 8
as equivalent (or, more precisely, in the context of your code:
interchangeable), and I would treat a 7 7 path as visiting only a
single node.
How does your
Raul,
I have carefully ignored your later posts and may be on a different track than
you intended. Here is where I am. Path 1 8 6 means there is a directed edge
from 1 to 8 and a directed edge from 8 to 6 and is different from path 1 6 8 if
such a path exists. You can even have a one edge
Yes, "tree" is not the best word, though it does hint at the existence
of branches - however its structure doesn't reveal the links. Each row
is a sorted list, I suppose. Whatever we call it, I think you've got a
big problem in finding all such "things" each with 100 unique members,
in a 2
Thank you, again.
I have not yet had time to read through wrapb, but a quick glance
suggests that the "too many paths (or 'trees', maybe )" issue could
probably be resolved by sorting each "path" before determining
uniqueness.
That said, I am more uncomfortable labelling these "visitable sets"
wi
Thanks. I think my verb "wrapb" _is_ doing what you want, even though
it doesn't get your answer of 5945 tree-paths for
10 graph2 wrap 0
I've just seen your later posting.
Sorry about 1 1$y - I agree, but had thought you only wanted a single
starting node.
You then say
"
But this is a prob
After reading this code, I see two issues.
First, this line:
p=. 1 1 $ y
should be
p=. ,. y
The original supported multiple start nodes, and I do not yet see any
simplicity advantages to considering only a single starting node. We
already need to deal with multiple paths, so why not allow
I'll have to take some time to read your code (there's a lot of it).
But perhaps this can illustrate why I initially called them "paths"
(though they can branch):
Here's a numeric representation of the connection graph. First column
is index, second column represents what that node connects to:
Not very sophisticated, and the first verb (wrapa) below doesn't even
reproduce your answer for N=6 , but it might be a starting point for
further work. The second verb (wrapb) does work for N=6, but apparently
greatly overestimates the number for your larger 2000 node graph!
wrapa=: 1 : 0
Since I messed up the description, I'll just pose an example solution
and ask for improvements.
graph=: 1= ?.20 20$10
start=: ,:0=i.#graph
gather=: 1 :'[: ~.@; <@u'
wrap=: 1 :0
:
graph=. m
start=. ,:(i.#graph) e. y
N=. x
paths=. (+./ .* # (+."1 =@i.@#))&graph"1 gather^:(N-1) start
(N=
I described this poorly:
A "path" implies a linear sequence, but all that's really required
here is that the nodes be connected.
So for any one path, there's a frontier which is all of the nodes
adjacent to any node in that path which has not yet been visited on
that path. This frontier is depen
Is there anything wrong with just walking the graph and setting each node
visited to 0 while keeping track of where we've been? Once our walk ends,
start again on any remaining node until they're all gone.
On Wed, Feb 13, 2013 at 8:29 AM, Raul Miller wrote:
> Let's say that we have a directed,
46 matches
Mail list logo
