Also fails in 804 (Windows 10, jqt) and 805 (Windows 10, jqt beta-11)
OK for _46 and _47 but not for _49, other things equal.
The real value is returned easily (for _48) with Newton-Raphson.
Mike
On 30/08/2016 14:21, Andrew Nikitin wrote:
p. _48 1 0 0 0 1
crashes j.
Tried with 803 and earl
I came up with this.
f =: (;@:{~ (i. >./)@:(#&>))@:(<@~.@:(>./\)\.)
NB. The main ingredient is ~.@:(>./\)
~.@:(>./\) l
7 8 13 15 19
~.@:(>./\) }.l
4 5 7 8 13 15 19
f l
4 5 7 8 13 15 19
NB. If there are more than one different subseqs
NB. of equal length, it returns the first.
NB. It f
Checking my attempt with the function in the essay cited, I see that
there's a misprint there.
The value of i should be '1113213211', not '1113122113', assuming i is
meant to be
,@((# , {.);.1~ (~: |.!.0))^:7 "."0 ,'1' NB. numeric result...
1 1 1 3 2 1 3 2 1 1
It's always good to l
Same problem with my version, which was faster but equally wrong!
Mike
On 02/09/2016 14:57, Raul Miller wrote:
Actually, if we try your approach on -8 2 4 3 1 6 we get _8 _2 _1
instead of _8 _4 _3 _1.
Here's a brute force O(2^n) approach that I hacked up earlier - it
obviously only works on
Well, assuming for now that it does work, here's an attempt at a J
version of
the pseudo-code listed at
https://en.wikipedia.org/wiki/Longest_increasing_subsequence#Efficient_algorithms
lis =: 3 : 0 NB. longest increasing subsequence
m =. 0#~>:n =. #x =. y
L =. #p =. ''
that I. works
mx =. ({.x) 1 } (<:<./x), n#>:>./x
for_i. i.n do.
xi=. i{x
lo=. mx I. xi
p =. (m{~<:lo) i } p NB. better than appending to short p for
larger x
m =. i lo } m
mx=. xi lo } mx
end.
|. x{~(p{~])^:(i.L) m{~ L =. <: m i. _1
)
Mik
to avoid clashing.
>
> NB. dyads
> exr=: (1 + {:@[ < ]) {. [ ; ,
> hxr=: [: ; exr&.>
>
> NB. monads
> gxr=: (}.@] ;~ {.@] ; (hxr {.))&>/
> cxr=: ({::~ [: (i. >./) #@>)@>@{.
> dxr=: (<_) ; ]
> lisXR=: ([: cxr gxr^:(#@(>@{:)
Sorry - I've only just seen this.
As usual, I'd forgotten that particular idiom, namely u^:(
I’m trying to rewrite lisM in a readable tacit form but hit something I didn’t
understand.
Supplying i.n to ^: works fine, but
On Sep 5, 2016, at 3:17 AM, 'Mike Day' via Programm
Thanks. No problems with line-wrapping in youe mail here!
Naming your function lisEH, and mine lis, I find:
l
7 4 4 5 2 7 1 8 13 2 10 4 9 1 4 5 5 4 3 10 3 4 5 8 15 7 11 19
;lisEH l
1 2 4 4 4 4 5 7 11 19
lis l
1 2 3 4 5 7 11 19
So lisEH appears to miss the 3
Mike
On 08/09/201
test increasing
subsequence". It is the "longest increasing subsquence" problem we are
solving? /Erling
On 2016-09-08 13:41, 'Mike Day' via Programming wrote:
Thanks. No problems with line-wrapping in youe mail here!
Naming your function lisEH, and mine lis, I find:
> FirstGroupOfMaxLength 3 3
>
> 2 LeftIsNotFirstInGroup (<1 2 3),<2 4 5
>
> 2 LeftIsNotFirstInGroup (<1 2 3),<2 4 5
>
> SelectFirstGroupOfMaxLength (<2),(<2 4 5),<2 6 7
>
> LongestIncreasingSubsequence=: LongestIncreasingSubsequence f.
>
> LongestIncreasingSubse
Thanks.
I'm sure you know the following, but it might be worth pointing out:
longascseq l
6 9 20 21 22 25 26 27
l{~longascseq l
1 2 3 4 5 7 11 19
lisM l
1 2 3 4 5 7 11 19
ie, longascseq returns the indices of the/a subsequence.
Very limited testing suggests that it's a little bit sl
s, so have avoided long names!
Thanks,
Mike
On 10/09/2016 00:39, Henry Rich wrote:
They're about the same speed on my machine. I don't think the
if-then-else is material. They're essentially the same algorithm.
(I'm testing on J8.05 beta, where the long names don'
Thanks.
It might help to point out that the main loop in Henry Rich's version,
as shown in the wiki page, can be simplified slightly with negligible
effect on run time, some savings in space, and doing without the
"sorted value" array.
I'm rather surprised - I'd introduced the array which I'd
Actually, the hit on run time gets more significant as the length of the
sequence and/or the optimal subsequence increase, so it's not negligible
for, say, ?100#50, but fine for ?1#5000, and not too bad
for ?10#5
Mike
On 13/09/2016 19:22, 'Mike Day' via Pr
longascseq2 { ]) 10?10'
0.468622 6.41946e6
ts'(longascseq1 { ]) ?10$10'
0.447244 3.28205e6
ts'(longascseq2 { ]) ?10$10'
0.458428 3.27373e6
ts'(longascseq1 { ]) 100?100'
4.55264 5.07078e7
ts'(longascseq2 { ]) 100?100&#x
I suspect it's pretty near optimal. The wikipedia article points at
Fredman's paper:
http://www.sciencedirect.com/science/article/pii/0012365X7590103X
analysing an algorithm of Knuth's. He shows that it performs better than
n log2(n) - n log2 log2 (n) + O(n)
rather than the O(n logn) cited
/download/3btahsyzax4w7xp/Vol%2C3+SortingAndSearching-Donald+Knuth.pdf
Youngs tableau is on page 47.
/Erling
On 2016-09-14 23:27, 'Mike Day' via Programming wrote:
I suspect it's pretty near optimal. The wikipedia article points at
Fredman's paper:
http://www.scie
09-15 11:52, Erling Hellenäs wrote:
The Fredman paper is about calculation of the length L of the longest
increasing subsequence, not the calculation of the longest increasing
subsequence as such. The algorithm only computes "valueofminendvalue"
I think. /Erling
On 2016-09-14 23:2
ending to an array requires a copy; so does tacit
amending. Thus, any tacit version is going to have quadratic
performance.
But I am working on making (x , y) and (x m} y) operate in-place even
in tacit forms. It will be a couple of months before that gets into a
beta.
Henry Rich
rmo's definition?
Anyway, here's an approach which works on the indices.
Please note that I haven't tested it thoroughly for boundary problems.
I just consider the rising diagonals as extra columns. Top & bottom
indices will be reduplicated, but can be removed with nub.
eg for LHA = 5, we get
e colon on the first line of cv?
Thanks,
Jon
----
On Tue, 9/27/16, 'Mike Day' via Programming wrote:
Subject: Re: [Jprogramming] Zig Zag problem
To: programm...@jsoftware.com
Date: Tuesday, September 27, 2016, 10:12 PM
rmo's definiti
nd dyad sections
I did not know that, thanks.
Your solution is great. I'll have to spend a little time figuring out what's
going on, but I can't find any flaws yet. And much more J-like than mine.
Regards,
Jon
----
On Tue, 9/27/16, 'M
You might enjoy this tacit version:
start =: i.@(<. #)
nstep =: (0 >. (-~ #)) >.@% <:@[
diffs =: (,:~ |.)@:+:@start
tconvert =: (]{~(#@])([-.~~.@:<.) ,@|:@(+/\)@ (start, nstep $diffs))
M
On 27/09/2016 15:27, 'Mike Day' via Programming wrote:
Thanks. I
ls for the case of number of
rows = 1:
1 tconvert 'SOMETEXT'
----------------
On Wed, 9/28/16, 'Mike Day' via Programming wrote:
Subject: Re: [Jprogramming] Zig Zag problem
To: programm...@jsoftware.com
Date: Wednesday, September 28, 201
Whoosh!
It even works ok with larg = 1!
Good stuff,
Mike
On 28/09/2016 11:58, Erling Hellenäs wrote:
ZigZagE=:([: /: ] (([: $ [) $ ]) [: (], [: |. [: }: }.) [: i. [) { ]
--
For information about J forums see http://www.
It's easier to see if we stick to indices rather than an arbitrary string.
Indices of elements in descending vertical columns in the zig-zag layout
for m rows belong to
(i.m) + 2 * (m-1) * i. n ,
where n is sufficient to exhaust the array,
and indices of the RISING diagonals are, for m>1,
easy way to do this. You Mike
solved it but the solution seemed a little complicated. /Erling
On 2016-09-30 00:27, 'Mike Day' via Programming wrote:
It's easier to see if we stick to indices rather than an arbitrary
string.
Indices of elements in descending vertical columns in t
What's the position with J for I-Pad?
The literal "VERSION" has value "1.3 5" on mine.
I don't use it very much as it's so tedious to transfer scripts to or
from a conventional
laptop or desktop, and writing scripts from scratch is difficult, but
it's interesting
from time to time to see how
Is it recommended to download the zip files and
a) just pick out the new executables, j.dll and jconsole.exe (in
windows at least)
b) overwrite one's existing installation with all the zip contents
or is it better to do a complete installation with the much larger
j805_... executable?
Tha
Sorry - should have made clear, the question applies when one
already has a full J installation.
Is it recommended to download the zip files and
a) just pick out the new executables, j.dll and jconsole.exe (in
windows at least)
b) overwrite one's existing installation with all the zip conte
While I'm also not too clear what you're really trying to achieve,
it might be worth offering this solution to your particular example
knowing your required solution:
x ((0~:(17 b.))&(#.@|:)) y
1 1 0 0 1 1
It uses the boolean and operation (17 b.) on bits at word level.
0~: could be 0< if yo
Well, this is pretty clunky, but it seems to work ok on
your data. I noticed that p may be derived from an unboxed
vector pp here, but that might not be relevant, and it's
easier to process p than pp:
pp =: 1 2 2 4 1 5 3 4 4 4 2 3 3
[p =: 2<\pp
+---+---+---+---+---+---+---+---+---+---+--
4 4|4 4|4 2|2 3|6 |
+---+---+---+---+---+---+---+---+---+---+---+---+
|1 2|2 2|2 4|4 1|1 5|5 3|3 4|4 4|4 4|4 2|2 3|3 3|
+---+---+---+---+---+---+---+---+---+---+---+---+
I tried working on the predecessor to p, without much
success.
Mike
On 07/11/2016 09:55, 'Mike Day' via Programming wr
Thanks - I'm not too surprised. v2 is still clunky. My earlier
offering is even slower and consumes
much more space!
But, FWIW, Skip's original example result had 5 rows, the last being
the unaltered input data, which is
why I added it - that's the one you then remove with }:@:. I now
Apologies for posting the previous message (below) earlier on Beta.
Anyway, I have now seen that there's a part two to the problem,
presumably hidden
to me as I hadn't got part one correct, and discussion of it (in
reddit.com/Advent...)
suggests that recursion IS required, or, at least, rele
. (>:e) }. y
end.
o , ',' , ": # y
)
R =: 2 : ' m * # n'
RB =: 2 : ' m * (1 +/@:".@:}.@:p1 n)'
echo +/ ". }. 0 p1 LF -.~ a NB. part1
echo +/ ". }. 1 p1 LF -.~ a NB. part2
- Original Message -
From: 'Mike Day' via Progra
Dec 10, 2016 at 1:19 PM, 'Mike Day' via Programming <
programm...@jsoftware.com> wrote:
[snip]
On 10/12/2016 16:47, 'Pascal Jasmin' via Programming wrote:
posted a slightly different solution on wiki, but the original did get
correct answer, even though it took
Trying the patience of the forum here, but may I ask: were both parts
of the problem
visible all the time? After getting part one correct I started looking
for and found part
two below part one's submission area - either it was already there, or
else it was
revealed by the correct entr
submitted a correct entry for Part 1.
Upon successful submission, it gives you the option to “Return to Problem n”,
and re-displays Part 1, along with your “correct” answer, then below that the
Part 2 extension.
Cheers Mike, Rob
On 11 Dec. 2016, at 7:18 pm, 'Mike Day' via Programmi
It looks to me as if the two left hooks form one big left hook,
not that I'm into boxing:
mn =: -"1 mean
mp1=: mp~|:
((mp1 mn)%#)dataNB. big "hook"
504 360 180
360 360 0
180 0 720
((mp1 @: mn)%#)data NB. mp1 after mn using @:
504 360 180
360 360 0
180 0 720
(([: mp
FWIW, here are some results running one million tests of bestof3N
compared with a
very slightly more J way of doing it.
(Your correspondence with Joe & Pascal while I was drafting this might
render the
following redundant.)
Freq =: ~.,:#/.~ NB. 2-row frequency table
297407
285714 240173 157035 71081
Mike
On 21/02/2017 14:24, 'Mike Day' via Programming wrote:
FWIW, here are some results running one million tests of bestof3N
compared with a
very slightly more J way of doing it.
(Your correspondence with Joe & Pascal while I was dra
New behavoiur this evening. Running jqt 64 vn 8.05 for windows
('type contour')plot(<"1]2#,:i:q),The plot demo does work partially. A plot is shown, but there's also
a dialogue
window with the same error message, preceded by "error in
plotdemo_default"
I ran package manager earlier
Thanks, Bill.
It works now, after running BOTH updatejqt64 805 AND package manager to
update ide/qt .
But did I actually need to do both updates? (updatejqt64 is slow here)
Thanks again,
Mike
On 24/02/2017 02:51, bill lam wrote:
Now fixed. Thank you for the report.
---
This email h
Sorry - I wasn't making myself clear. What I meant was, I ran both
updates because
I didn't know what does what. It's not obvious to someone who doesn't
write J software
where different bits reside. I didn't know whether your fixes were in
addons or in qtlib.
No harm done!
Thanks for such
What is z, please?
M
On 26/02/2017 08:57, Linda A Alvord wrote:
Here's a little demo to show you that there are really two circles in the
final plot.
The second plot erases the previous plot, so look at the code line by line.
load 'plot'
do=: 13 :'({.y)+(i.{:>:y)*(--/}:y)%{:y'
t
It's not clear what you had in mind for the dyadic function "softmax" .
Anyway, is this what you require?
I=: =/~@:i.@#NB. Identity matrix (Kronecker delta)
pdsm =: (*"1 (-~I))@:sm NB. Partial derivative (matrix) of sm
pdsm a
0.186192 _0.0676431 _0.0467234 _0.0718259
_0.06764
That's the exact approach, as you know.
Meanwhile, I'd got the impression that Linda wished to present the
quarter-circle approximations, as shown in, eg,
https://en.wikipedia.org/wiki/Golden_spiral#/media/File:Fibonacci_spiral_34.svg
So I've amused myself for a while developing the followin
t helix in some such fashion,
than to develop the fractal squares or cubes that I'd started on.
Mike
On 01/03/2017 08:18, Linda A Alvord wrote:
Mike,
Here's an error that I am getting.
Linda
-Original Message-
From: Programming [mailto:programming-boun...@forums.jsoftwa
ke,
Here's an error that I am getting.
Linda
-Original Message-
From: Programming [mailto:programming-boun...@forums.jsoftware.com] On Behalf
Of 'Mike Day' via Programming
Sent: Tuesday, February 28, 2017 11:03 AM
To: programm...@jsoftware.com
Subject: Re: [Jprogramming] Exp
FWIW, Eric, I don't find the distinction between "file" and "files"
obvious -
they're evidently different but I don't easily remember which is recent
files and
which is a file browser.
Why not rename "files" as "recent files" and have done?
But I'm (still) not a frequent user of JHS, so shou
I didn't know I had avx available on this machine, an AMD A10-7300,
running Windows 10. Anyway, J806 JVERSION says I do!
I'd have a go at running the benchmarks, but wonder if there's a script
available
to save doing them "by hand"
JVERSION
Engine: j806/j64avx/windows
Beta-1: commerci
This has arisen in solving Euler Project problem 593.
https://projecteuler.net/problem=593
It is basically quite easy, especially for the examples with given
answers, ie
for F(100,10) and even for F(10, 1) but as usual with these
problems,
there's a sting in the tail, in the computing
ast.
> But may be interesting as a reference.
>
> Jon
>
> ------------
> On Wed, 3/15/17, 'Mike Day' via Programming wrote:
>
> Subject: [Jprogramming] Moving median - any good methods?
> To: programm...@jsoftware.com
> Date: Wednesday, March 15, 2017,
Only just up - having breakfast! I evidently solved it, as the
project's discussion thread
is open to me, but I don't have a record of my method.
Early-ish PE problems were amenable to brute force, and I expect that's
how I would have
done it.
AFAIrecall, later problems involving phi requ
ly to mike_liz@tiscali.co.uk.
Sent from my iPad
> On 15 Mar 2017, at 07:48, 'Mike Day' via Programming
> wrote:
>
> Only just up - having breakfast! I evidently solved it, as the project's
> discussion thread
> is open to me, but I don't ha
Thanks .to Jon Hough for the link to priority queues. I had a look. I doubt
they're directly relevant as a moving median needs to be able to remove a
particular element rather than that chosen by the remove method used there -
if I understand it right. However, it does show how to set up a
Thanks to Jon Hough for the priority queue oo example.
It works quite well as an example method, but I'm rather stumped as to how to
remove an item other than the priority one. I.e. I need removenum rather than
remove''
I have to go out now, so sorry for no more detail.
Thanks,
Mike
Please r
Using extended integers is the obvious answer to this, but I'd like to
avoid the
memory and performance hits.
I'm using J806 (with avx if that's relevant) under Windows 10 on a
64-bit machine.
I wish to multiply integers modulo 2^48 . We can assume that inputs are
< 2^48.
Set
M =: 2 <.@^
ay be useful is that when either x or y not prime,
M&|@*/ ,&q:/ M | x , y
equivalent to
M&|@*/ x , y
From: 'Mike Day' via Programming
To: programm...@jsoftware.com
Sent: Monday, May 8, 2017 4:44 AM
Subject: [Jprogramming] Precision o
Thanks, Raul
Mike
On 08/05/2017 14:19, Raul Miller wrote:
On Mon, May 8, 2017 at 4:43 AM, 'Mike Day' via Programming
wrote:
Is there a better way to do modulus multiply with largish arguments?
Hypothetically speaking, you might be able to do something based on
http://code.jsoftware
I appear to have solved PE 159, & have even found the verb that seemed to do
the work, but have forgotten how it works!
Anyway, it might be worth considering a constructive approach, whereby you
consider possible combinations of primes generating all numbers in the domain
rather than anal
I've now got my p159 verb to work again. I'd written it back in J4 or J5 days,
and some chars had got mashed! Anyway, although it doesn't break the sound
barrier, it does limp home with the correct answer in under a minute for the
given problem size, running in J806 under Windows 10.
Mike
My verb is pedestrian, not J806 itself!
M
Please reply to mike_liz@tiscali.co.uk.
Sent from my iPad
> On 11 May 2017, at 17:36, Istvan Kadar wrote:
>
> I have Windows 7 and J602. It is not limp in PDF or DOC. Thanks!
>
> 2017-05-11 18:10 GMT+02:00 'Mike
metic on the two operands only.
I wouldn't need to use extended multiplication if multiply of <63-bit
integers modulo
M< ~ 2^63 was guaranteed correct! Too much to ask?
Any better ways to achieve the result of cumoddprod?
Thanks,
Mike
On 08/05/2017 15:08, 'Mike Day'
Thanks again, Raul.
Well - I'd actually meant ~2^24 odd numbers, 1... _1+2^25 - trust me
to make a mistake in trying to simplify the presentation! Let's call it
a typo.
I don't think my post was a spoiler, as the problem needs quite a bit
of insight elsewhere; this array helps in getting t
It would be nice if it worked.
Unfortunately (again!), for the actual size required,
timer'_10{. |. 281474976710656&|@*/\.<.1+2*i.-2^24'
+--+---+
|32.001|0 0 0 0 0 0 0 0 0 0|
+--+---+
whereas the correct values are all > 0
This is why I needed specia
Yep - a blocky approach addresses the virtues of a vector approach and
minimises the hit with extended. Could fine tune it but why bother!?
I quite often end up using blocks in this sort of fashion, though
overlooked
it here.
Note to self, though - must look into why /\. with reverse is bet
It still seems to run under 10.3.1, with a warning re future incompatibility.
I expect new iOS vn's benefits outweigh cost of losing J. Any views?
Thanks,
Mike
Please reply to mike_liz@tiscali.co.uk.
Sent from my iPad
> On 26 May 2017, at 10:43, Kip Murray wrote:
>
> My iPad Air
Yes - it's a bit embarrassing, but my message didn't get sent until Eric's
had arrived, 'cos iPad out of wifi range when I posted, I think.
Thanks, Mike
On 26/05/2017 19:09, Kip Murray wrote:
See Eric's recent note "iOS/iPad/iPhone". --Kip
On F
This problem has just re-surfaced for me, J806 under Windows 10,
details below.
Today's inaccessible fif has
Fif=-1383 154 2394 529
in the configqt file, winpos.dat
Difficult to say what caused it, as I tend to keep J sessions going
until they
fail for some reason.
I'll try to reset the va
e x meant measured from a viewport on the right hand side.
On 6 Jun, 2017 1:17 am, "'Mike Day' via Programming" <
programm...@jsoftware.com> wrote:
This problem has just re-surfaced for me, J806 under Windows 10,
details below.
Today's inaccessible fif has
FWIW - not much, here, as it isn't relevant to factorial, and on
investigation, not
something used in Mr Rice's LISP example, but, as Raul's posted a verb
for "union",
I've sometimes wondered why J doesn't include primitives for union and
intersection.
As I sometimes need intersection, here'
, Raul Miller wrote:
On Tue, Jun 6, 2017 at 4:09 AM, 'Mike Day' via Programming
wrote:
if I were developing a solver for solitaire, I'd include a variable as a
parameter for
the size of problem, eg the number of rows, 1 2 3 etc, or the ravel-size,
eg 1 3 6 etc.
You had not spe
Thanks, Eric.
I've just bitten the bullet and upgraded my iPad Air to IOS 10.3.2(14F89)
(I don't know if the bracketed details are significant)
I then downloaded the J64 "app" - it started up ok, and the initial demos
worked fine.
What surprised me is that the old J32 app still appears to wor
ok, thanks
On 08/06/2017 04:37, Eric Iverson wrote:
Mike,
The old j32 will stop running on some future iOS release. Perhaps before
October.
On Wed, Jun 7, 2017 at 6:53 PM, 'Mike Day' via Programming <
programm...@jsoftware.com> wrote:
Thanks, Eric.
I've just
Raul's listing at
http://www.jsoftware.com/pipermail/programming/2017-June/047522.html
works fine for me!
Mike (Day, not Rice!)
On 08/06/2017 15:07, Michael Rice wrote:
move_table=:
<:(,|."1)(#~0<*/"1)~.,/,/_3&{.\@/:~"1(,|:"2)>|:((;|.)@(1++)+/)\i.5
flip=: ~: (i.15)&e.
legal_moves=
I've only just seen this.
I expect it's obvious, but you ARE looking only for rectangles with "vertical"
and "horizontal" sides, i.e. no rotations other than multiples of pi/2 involved?
Thanks,
Mike
Please reply to mike_liz@tiscali.co.uk.
Sent from my iPad
> On 20 Jun 2017, at 16:29
sides, and slanted "rectangles" would thus include empty
> space on the edges.
>
> Thanks,
>
> --
> Raul
>
> On Tue, Jun 20, 2017 at 4:49 PM, 'Mike Day' via Programming
> wrote:
>> I've only just seen this.
>> I expect it'
Having at last had a go at solving this, post breakfast (if not Brexit).
I don't think it's a new approach, but it's fairly nifty.
It considers run lengths times heights.
a =: >./^:2@:(>:@i.@# ([ * <./\) ])
qq=:1000?.100
a qq
7030779
ts'a qq' NB. time & space
0.0328879
q as previous post, below
0.0316207 2.24246e7
ts'c qq'
0.0149281 28544
Thanks,
Mike
On 24/06/2017 10:09, 'Mike Day' via Programming wrote:
Having at last had a go at solving this, post breakfast (if not Brexit).
I don't think it's a new approach, but it's
Oops - that diagnostic "q" was unnecessary:
c =: 3 : 0
m =. >./y
for_i. }.>:i.#y do.
m =. m >. >./ i ([ * <./\) y
end.
)
ts'c qq'
0.0143082 21632
M
On 24/06/2017 10:53, 'Mike Day' via Programming wrote:
This explicit version is faster and
It's super fast, but doesn't seem to produce the correct result!
Where's the typo?
(C,c) qq
152 7030779
ts'c qq'
0.0159982 21632
ts'C qq'
0.0107839 289920
Sorry,
Mike
On 24/06/2017 14:04, Raul Miller wrote:
Or,
C=: <./@(>:@i.@# <./@([ * <./\)"0 1 ])
Thanks,
---
T
Ah! I'd spotted the outer, left-most >. for <. , but had missed the
inner, rightmost...
The speed-up and space-saving (cf my verb "a", say) appears to come from
using ranks
"0 1 rather than the placement of the >./ fragments.
My c verb remains slower (because it loops!?) but uses less spac
There must be something about breakfast.
A loopy stopping condition has just presented itself. Others might see
a tacitisation:
cs =: 3 : 0
m =. >./y
for_i. }.>:i.n =. #y do.
m =. m >. q =. >./ i ([ * <./\) y
if. (m * i) >: n * q do. break. end.
end.
m
)
q holds the maximum area at "wi
j] ), min( V[i+1] … V[j+1] ) ).
This version is based on that:
ms=: #\ >./ . * [: >./@> # 2&(<./\&.>)&< ]
1e3 ts 'ms qq'
0.00491435 5.74234e6
Incorporating this in Mike's explicit version with an early stopping condition
could lead to even faster
Thanks.
I haven't heard of paste & match - I like to use Thunderbird to support
the sharewarers,
but fight against some of its features. With time & patience, I copy &
paste into a notepad
file, and then re-copy and paste into T/bird. I hadn't bothered in my
recent posts.
I do (and did in t
Chat really...
Breakfast-time again! Noticing my name (below) I assumed you meant my
recent discussions
with Raul et al on "the largest rectangle problem" were relevant! On
checking, I see that I
had contributed something years ago, on colouring contiguous areas in
rectangles.
So perhaps I'
Linda, bear in mind that Raul was (I think) deriving a tacit improvement
of my explicit verb "c" which had incorporated an explicit loop.
Here it is, again, possibly without gratuitous line-throws:
c =: 3 : 0
m =. >./y
for_i. }.>:i.#y do.
m =. m >. >./ i ([ * <./\) y
end.
)
It happens to
I believe this is the wrong thread, Linda.
I commented yesterday on a similar message from you under the heading,
"Largest Rectangles."
Mike
On 29/06/2017 02:07, Linda Alvord wrote:
Here is how I wrote your verb. How would you write it as an explicit verb?
f=: [:>./([:([:>:i.)#)([: >./[ * <
Linda, yet again, I believe you're looking at treatments of Raul's
largest
rectangle problem, as discussed in the thread "Largest Rectangle"
qq is a randomised example of a list of widths or heights, depending on
your
view, of what might be regarded as a histogram or bar-diagram. Raul soug
>>> m=. >./y
>>> for_i. }.>:i.n=. #y do.
>>>m=. m>.q=. i*>./y=. 2<./\y
>>>if. (m*i)>:n*q do. break. end.
>>> end.
>>> m
>>> )
>>
>>
>>> 1e3 ts'cs1 qq'
>>> 0.00107603 29440
Interesting! I was puzzled as to why ldrs1/2 also each consumes less space
than lrds or finalrev etc, but then realised that the source of the
difference
is probably the expression " i.n " in the do-loop version.
ts'i.1000' NB. time & space for i.1000
1.24189e_5 8960
supports the suppo
Not multiple assignment, but copula-related.
I've often wondered why J doesn't have the C family feature of assignments
such as
t =: 1 2 3
[t+=: 3
4 5 6
I don't know what it's called, and I realise it's only an arguably
attractive
feature, not an essential one, but would be useful, espe
Early try-out on iPad, iOS 10.3.2 :-
the "?" "Button" appears not to work, not with my finger-jabbing, at least...
Thanks,
Mike
Please reply to mike_liz@tiscali.co.uk.
Sent from my iPad
> On 5 Jul 2017, at 01:42, Eric Iverson wrote:
>
> A significant new release of J iOS is avai
pe work too?
Thanks again,
Mike
Please reply to mike_liz@tiscali.co.uk.
Sent from my iPad
> On 5 Jul 2017, at 07:54, 'Mike Day' via Programming
> wrote:
>
> Early try-out on iPad, iOS 10.3.2 :-
>
> the "?" "Button" appears not
keyboard in
landscape mode so that you would see the html view, but that can be a
nuisance for general use.
Please confirm that J works properly (as described here) on your device or
if you still think there is a problem.
(The problem is partly because the J kb is so big and it is big because my
fi
[Jprogramming] might not be right - is there an iPad beta group?
I've just discovered the "keyboard" includes _both_ a delete left arrow
_and_ cursor
movement left (and right) arrows. It seems to work in both terminal and
edit sessions.
That's a great improvement in itself, especially for fa
While drafting this, I've just seen Joey Tuttle's post. But I'll
persist!n (And
now, Eric's!)
There's the official method for the previous (32 bit) iOS versions,
described
under http://code.jsoftware.com/wiki/Guides/iOS/32bit#FileTransfers
It involves emailing to yourself. Possibly simpler
1 - 100 of 583 matches
Mail list logo
