description of the operation
is correct.
Henry Rich
On 12/26/2023 6:14 PM, Raul Miller wrote:
I should add that the nuvoc has some examples of implied rank, but I
can't figure out why a list of 10 numbers from an infinite ranked v
means rank 0 while a list of five numbers from infinite ranked v
See https://code.jsoftware.com/wiki/System/ReleaseNotes/J902
Henry Rich
On 12/26/2023 3:11 PM, Devon McCormick wrote:
I don't know about the history but in 9.5.5 (under Windows 10):
(0:`1: @. (2&|)) i.4
0 1 0 1
On Tue, Dec 26, 2023 at 7:44 AM Martin Kreuzer wrote:
Dear all -
This
.
For pruning, you are comparing two game positions to decide which one
should be first analyzed further. The comparison may be quite involved.
Henry Rich
On 12/22/2023 12:47 AM, Elijah Stone wrote:
For toposort, though, you'd presumably want a cleverer strategy than
just sorting the nodes, doing a full
Topological sorting was a big topic in graphics back in the day.
I can imagine that during alpha -beta pruning you might run into some
ordering requirements that require more than simple keys.
Henry Rich
On Thu, Dec 21, 2023, 9:56 PM Elijah Stone wrote:
> > The problem
(/: verb-to-put-cards-in-order) table-of-hands
For poker, verb-to-put-cards-in-order is
vtpcio =. '23456789TJQKA'
(/: vtpcio) _5 ]\ 'A22345323452'
23452
23453
2
A
Henry Rich
On 12/21/2023 10:29 AM, 'Viktor Grigorov' via Programming wrote:
The first example is lesser, my
I don't understand the problem statement.
Henry Rich
On 12/21/2023 10:17 AM, 'Viktor Grigorov' via Programming wrote:
Maybe a table, summation and a sort using that:
(/: ([:+/(>/]))) _10 1 10 20 _30 15 25 30 0
Can't imagine that being efficient, but it's better than nothing. The verb
This follows the parsing rules.
f is executed when the pattern
N V V N
is seen on the execution stack. GLOBAL has already been stacked at that
point, as an undefined verb. When later GLOBAL is to be executed, its value
has been changed to a noun. That pun is disallowed.
Henry Rich
On Sun, Dec
It's expected. If you kept the sparse element at 0 the matrix would
become dense.
3 $. -. $. 3 3 $ _4 {. 1
1
3 $. 1 - $. 3 3 $ _4 {. 1
1
3 $. 2 - $. 3 3 $ _4 {. 1
2
3 $. % $. 3 3 $ _4 {. 1
_
Henry Rich
On 12/11/2023 11:53 AM, David Lambert wrote:
$ ijconsole
JVERSION
Engine
($ #: I.@,) y
Henry Rich
On 12/11/2023 5:06 AM, LdBeth wrote:
I've been running to too many situations that I uses 4$.$.y a lot to
get indexes of all non zero elements to an array, I wonder if there
are any alternative ways to do that in J, especially when y is a
boolean array.
Also
2 31 44 53|32 37 42 47 52 57|
| 5 14 23 36 45 54|33 38 43 48 53 58|
| 6 15 28 37 46 55|34 39 44 49 54 59|
| 7 20 29 38 47 60|35 40 45 50 55 60|
|12 21 30 39 52 _|36 41 46 51 56 _|
+-+-----+
Henry Rich
On 12/10/2023 11:37 AM, bill lam wrote:
https://ibb.co/XLn0bxH
T
22j10 (* ^@:j.@:o.&(%&180)) 15
18.6622j15.3533
Make the middle part a named verb if you like.
Henry Rich
On 12/7/2023 11:48 AM, 'PMA' via Programming wrote:
A slightly different approach. Rather than search out all alignments
and then deduce their degrees of rotation, I
Changing the browser size changes by too much. In particular, there is
no way to recover the initial 50% size.
Henry Rich
--
For information about J forums see http://www.jsoftware.com/forums.htm
the top. I refer
to those all the time. Better, allow a user to lock a page pointer into
the left pane.
This is already easier to get around in than the Wiki, just because of
decreased scrolling.
Henry Rich
On 11/23/2023 7:56 PM, Ed Gottsman wrote:
load '~addons/gottsman63/jwikiviz
5j1.38675 0.5547j2.7735 0.27735j4.9923
0j1.41421 0j2.82843 _0.707107j4.94975
_1.34164j0.447214 _2.68328j0.894427 _4.91935j0.894427
_1.38675j0.27735 _2.7735j0.5547 _4.9923j0.27735
_1.41421j1.11022e_16 _2.82843j2.22045e_16 _4.94975j_0.707107
H
, / dyad, ~, -. dyad, ^ monad, | dyad, -, /: dyad, *. monad
hhr
On 11/15/2023 6:49 PM, Henry Rich wrote:
I would find the angles from each point to each other point, take
modulus(pi/2), concatenate the list with its negative, and sort.
Rotation by each angle produces a new alignment. Only J primitives
.
Henry Rich
On 11/15/2023 5:52 PM, 'PMA' via Wiki wrote:
Dear J-Wiki,
I want to make a script that, given the X/Y values of a 2D dot-graph,
will rotate the graph around its designated origin, looking for
instances of 2 or more points coming into new horizontal &/or vertical
align
This seems like a useful feature and not difficult to implement. We can
discuss here what should be supported.
Morten's reservations (in the cited article) are reasonable. I note that J
has semiduals, and that it would make sense to support structural under for
dyads using semiduals.
Henry Rich
Something like
(<1 0)&{::@> data
untested.
Henry Rich
On 9/29/2023 9:39 AM, Pablo Landherr wrote:
I have the following
*data=: dec_json {some data source}*
*$data*
4
s the order of axes,
you must then apply the appropriate transpose.
In a scalar language where you have to list the order of axes every time
you access an atom, the transpose is hidden beneath the index order, but
it's still there. J forces you to confront it directly.
Henry Rich
On 9/7/2
Since you want all the slices, what you are looking for is a transpose.
Maybe
~.@(, i.@#) |: ]
Untested.
Henry Rich
On Wed, Sep 6, 2023, 6:10 PM Piet de Jong wrote:
> Here is my “wish"
>
> A dyadic (tacit) verb such that x v y gives all the slices of y along
> dim
This answer put in in mind of Mark Twain fulminations on James Fenimore
Cooper:
[Cooper thought] a man's mouth is a rolling-mill... converting
four-foot pigs
of thought into thirty-foot bars of conversational railroad iron by
attenuation...
Henry Rich
On 8/29/2023 9:13 PM, Devon
I don't know. The last I used it was years ago. I had to make new
wrappers sometimes; it was straightforward.
Henry Rich
On 8/19/2023 5:35 PM, Piet de Jong wrote:
Thank you! Very helpful.
Are there an cgee*.ijs scripts equivalent to dgee*.ijs scripts?
Or do they have to be crafted from eg
You need cgee* for complex.
Henry Rich
On 8/19/2023 4:03 PM, Piet de Jong wrote:
I’ve just updated to J9.4 and Lapack2.
The verb do_dgeev from math/lapack2 appears to work fine for real matrices.
However it gives a “domain error" when applied to a complex matrices.
Is this how it s
Look out, it's a trap!
7 2 is a single word (a 2-item list). You have to separate the numbers.
7 (2)} y
0 1 7 3 4
Henry Rich
On 8/19/2023 10:46 AM, 'PMA' via Programming wrote:
NB. Dear J gurus --
NB. I'm having trouble with x m} y
NB. Given a character array, it works:
y
??? J doesn't use unprintable characters.
The result of 1!:1 is a string, whether the data comes from keyboard or
a file.
Henry Rich
On 8/16/2023 11:29 AM, Ak O wrote:
What are the unprintable characters that set 'n' to string?
Ak
On Wed., Aug. 16, 2023, 09:17 Henry Rich, wrote
9 3 24 10 26 13 15 1 18 8 25 19 0
5 16 14
1. done
2. y is immaterial
3. n is set to a string
3a. string has 'x' appended, is convert to (extended integer) number,
used in A.
Henry Rich
On 8/16/2023 9:00 AM, Ak O wrote:
Hi guys,
I hope you are all well. Thank you for continuing to look at t
owed by an assignment copula =. or =:).
Henry Rich
On 8/15/2023 9:51 AM, Raul Miller wrote:
I think the inconsistency here is that you're thinking of a rather
elaborate parsing system used by some other language.
Brief overview of J's parser: A sentence is parsed first by tokenizing
it (left to r
My advice: learn to live with the quotes.
You need a way to write a string that represents a long value. In J,
that's 'longvaluex'.
Henry Rich
On 8/14/2023 10:57 PM, Ak O wrote:
Hi Henry,
Yes actually all I want is a verb that stting encloses an argument.
After that I can attach ('.@,&am
". string
executes the string. But it is a string, i. e. a literal. If you want
that literal to contain a character string for execution, you must
enclose that string in quotes. You can write a verb to do it.
Henry Rich
On 8/14/2023 9:14 PM, Ak O wrote:
Is there a more simpl
Why is extd a dyad?
Otherwise, this seems to work. What is deficient?
Henry Rich
On 8/14/2023 8:38 PM, Ak O wrote:
extd =: 4 : 0
n=. ((1!:1) 1 )
((".@,&'x' n)) A. i. x
)
--
For information about J forums
You have a mismatched parenthesis in the prompt line of the verb you posted.
Henry Rich
On 8/14/2023 4:42 PM, Ak O wrote:
Hi Raul,
Yes I think I understood Henry's message. This why my thought is to simulte
the keyboard input. Since typing the raw input into the first function
works exactly
x: /number/ is two words. /number/ is evaluated first, and then x: is
applied to its value to give an extended integer.
In that last case, if /number/ has more than 19 digits, it will have
been represented as a float, and the extended integer will have only 16
digits of precision.
Henry R
k to start with
the permutation vector rather than the anagram index.
Henry Rich
On 8/11/2023 8:29 PM, Ak O wrote:
For me,
(x: y)
does not preserve the input.
The result I get is not the same.
( x:180548043269214561950911457875657 )
180548043269214573494164592263168
This doe
(x: value) produces extended version of value.
Henry Rich
On Fri, Aug 11, 2023, 2:44 PM Ak O wrote:
> Hi everybody, I hope you are all well.
>
> I have a question about the Anagram ( A. ) operator.
>
> In a defined function, how do I designate that an input 'x' is treatment
to understand a sentence I wrote, and after 5 minutes
you still can't get it, that would be a good candidate for more words
from me.
Henry Rich
On 8/5/2023 4:20 PM, More Rice wrote:
[Henry] IF YOU ARE NEW TO J, you are the person I need to hear from.
Where do I
need to add explanatory material
I agree with Luke.
Henry Rich
On 8/5/2023 5:12 AM, Luke D wrote:
I suppose it is worth reminding everyone that only the first 100 PE
problems can be shared. Because of this, I don't think spoilers are
necessary. PE only asks that any discussion about said problems is done to
instruct
code, thinking about the whole problem rather than atoms.
These pages are not about showing off a brilliant solution, but about
how you came to the solution.
Henry Rich
On 8/5/2023 3:06 AM, 'Rob Hodgkinson' via Programming wrote:
Yes, of course, please contribute. Like Jan-Pieter you could
I have enjoyed reading your solutions. It's great to see the language
used for somewhat realistic problems.
Will those of you with Project Euler solutions step up and showcase your
J? Just add to the ShareMyScreen page.
Henry Rich
On 8/4/2023 9:31 AM, Jan-Pieter Jacobs wrote:
I've been
They are legion. Example: x -@:{`[`]} y uses special code only when y
is floating point.
Henry Rich
On 7/31/2023 10:40 AM, Ak O wrote:
Can you share examples where the form of an SC is satisfied, but the
interpreter does not perform the SC routine?
AO
On Mon., Jul. 31, 2023, 06:39 Henry
You are right about i.!.0. It is discussed in
https://code.jsoftware.com/wiki/Vocabulary/IFamily but who has read that?
You are also right that the datatypes matter.
Highlighting near-misses would be great, if it could be doone.
Henry Rich
On 7/31/2023 4:18 AM, Jan-Pieter Jacobs wrote
that they have the form for an
SC, the interpreter alters the routine-address from the normal handler
for @:, to proceed instead to the code performing the SC.
Henry Rich
On 7/30/2023 10:13 PM, Ak O wrote:
What system currently governs Syntax Highlighting in the editor?
By example:
If yo
I think the fork/hook idea is excellent, especially as an option.
Is there a way in Jqt to find the match for () ?
Henry Rich
On 7/30/2023 6:02 PM, 'Viktor Grigorov' via Programming wrote:
My twopence:
Worth confirming: did you mean syntax highlighting on wiki pages or for the
various text
gives +0.
Henry Rich
On 7/30/2023 3:26 PM, Henry Rich wrote:
You have found an error, I think. We were /not/ using Kahan's formula
and branch cuts. (Our version came from Abramowitz & Stegun)
William Kahan is a mighty man. On his authority (and the sources you
cited) I have changed to
What a great idea! I know nothing about Syntax Highlighting but I would
love to see it in the display.
There will be a problem keeping the highlighting up to date with the
evolving support for special combinations.
Henry Rich
On 7/30/2023 12:59 PM, Ak O wrote:
Hi all, I hope you are all
distribution (in
3 dimensions). All tacit code it was. You follow in broad footsteps.
Henry Rich
On 7/15/2023 9:15 AM, Raul Miller wrote:
I think the approaches you described are fine.
That said, another approach would be to think of your dataset as a
table -- one row for each atom, with each column
512 by 3 and 512 by m pieces later.
Thanks for the comments, especially the insights into u;._3 !
On Thu, Jul 13, 2023 at 2:09 PM Henry Rich wrote:
I'm surprised that the result is 512x512 when you use u;._3 on 512x512
argument. Do you pad after the operation?
1. How good the operation
I was just making the point that the verb must return an array of
results. < does not and that fails before the type of the result enters
into the matter. 0: would give the same error.
Henry Rich
On 7/13/2023 6:10 PM, Raul Miller wrote:
+: is double, < is box (which in this example
; was passed an array of cells, but it didn't return a result for each cell.
Hey, you did say 'efficiently'.
Henry Rich
On 7/13/2023 5:50 PM, Marcin Żołek wrote:
Hello,
I need to define a function that will execute function f on the elements of an
input array that meet a certain condition, and exec
how much better 7 worker threads are than 3.
Henry Rich
On 7/13/2023 1:26 PM, Clifford Reiter wrote:
Hi,
I thought I would experiment with t.
I choose an "image" processing problem on a 512 by 512 array.
Local (complex) processing occurs on 3 by 3 cells (u;._3) which results in
a
Look at Vocabulary/Parsing in the wiki. Execution of an adverb produces an
anonymous entity, usually a verb.
Henry Rich
On Sat, Jun 10, 2023, 10:40 PM Raul Miller wrote:
> Adverbs are never dyadic. Nor are they monadic in the sense that verbs
> are monadic. (An adverb takes a singl
and execution proceeds as if the execution had returned 0.
Henry Rich
On Fri, Jun 9, 2023, 5:33 AM Mike Duvos wrote:
> Hi,
>
> Isn’t it a convention that a null character vector behaves the same as a
> null numeric vector in an arithmetic context?
>
> I get the following results in J
coming up with the solution!
Henry Rich
On 6/2/2023 6:51 PM, 'Viktor Grigorov' via Programming wrote:
I'd agree that examples are great, e.g., Rosetta Code is a great compendium of
programming language equiproblem solution comparisons. A comma-delimited
listing of would be too much
as long as you make the solution comprehensible to a
novice.
I know several users on this list solved the AoC2022 problems;
alternative solution pages would be welcome.
Henry Rich
--
For information about J forums see http
See item 1 under 'Common Uses' in
https://code.jsoftware.com/wiki/Vocabulary/dollarco
Henry Rich
On 5/19/2023 9:44 AM, 'Rob B' via Programming wrote:
Apologies I thought I had already sent this message but it got lost….
Thanks Henry, nuvoc $: does say 'tacit' but I missed that, doh
If $: appears in a tacit verb, its value is that entire verb.
It's when $: appears in a phrase in a sentence that it gets more
complicated.
Henry Rich
On 5/19/2023 7:54 AM, David Lambert wrote:
Self reference applies to the verb in which it is defined. This seems
to be the entire tacit
means 'execute myself' with nothing to stop the recursion.
$: is difficult to define in words.
https://code.jsoftware.com/wiki/Vocabulary/dollarco under More
Information is the best I can do.
Henry Rich
On 5/15/2023 8:36 AM, Brian Schott wrote:
Reading in Nuvoc I see the following comment
You can't beat a horse with no horse. What name do you suggest?
Henry Rich
On 5/9/2023 4:56 PM, Richard Donovan wrote:
One of the most frustrating things about J is its name! For example, on Stack
Exchange code review it seems impossible to search for any J articles,
comments, code snippets
Fixed for next release/beta. The error was in formatting a message when
the assignment failed because the value could not be opened. Only
multiple assignment was affected.
Thanks for the very clear report.
Henry Rich
On 4/20/2023 8:44 PM, Gilles Kirouac wrote:
Say I have
]d=: _2
In the next beta u m. n means 'u(mod n)' and will work for + - * % ^ .
%. coming soon.
The only specials I see with m&| are m&|@^ and m&|@(n&^) which can be
replaced by [n] (^ m. m) y.
Example: 5 % m. 7 (9) gives 5 *(mod 7) (inverse of 9(mod 7))
Henry Rich
On 4/17/2023 2
I think I have figured out a way to return float when the result has been
made inaccurate.
We had no choice about m&|@^ for negative y: the behavior of that is
defined by the language, and it isn't modular.
u m. n is a much cleaner solution, and faster. m&|@^ is deprecated.
Henry Rich
The info you are talking about at the end of your message is the block
header. I can tell you all about that (so can Raul) but I am away from my
machine until Monday.
Look at jtype.h in the J source and find the definition of AD.
Henry Rich
On Sat, Apr 15, 2023, 6:36 PM Michal Wallace
wrote
, the way the internal interfaces are, it's
difficult to leave the value as floating-point, so you cannot use the
fact that an integer was returned as a guarantee of accuracy.
Henry Rich
On 4/13/2023 11:34 AM, 'Michael Day' via Programming wrote:
Yet again I found myself resorting to Pari GP
liable to get roundoff,
because you are at the limit of the IEEE-754 mantissa.
These numbers are just big enough that you need to be using extended
integers if you want accurate results.
Henry Rich
On 4/13/2023 11:34 AM, 'Michael Day' via Programming wrote:
Yet again I found myself resorting t
.
Henry Rich
On 4/7/2023 3:09 AM, Thomas McGuire wrote:
In my haste to produce an email after finally figuring out how Pepe’s elegant J
implementation of the maximal sum of the subsequences worked ,I can see how you
might interpret this as I was disappointed in the implementation as not being
ute questions of instruction
ordering.
Henry Rich
On 4/6/2023 8:12 PM, Thomas McGuire wrote:
On Apr 3, 2023, at 1:28 PM, Jose Mario Quintana
wrote:
For what it is worth...
kad=. >./@:(([ >. +)/\.)
OK that is worth a lot. But it’s not really the kadane algorithm. It ins
Domain error.
Henry Rich
On Tue, Apr 4, 2023, 6:56 PM 'Mike Day' via Programming <
programm...@jsoftware.com> wrote:
> Lovely. Hopefully including negative powers! How will you deal with
> ill-defined results,
> though?
>
> Thanks,
>
> Mike
>
> Sent from my i
Expect primitive support for modular inverse in an upcoming beta.
Henry Rich
On Tue, Apr 4, 2023, 1:49 PM 'Michael Day' via Programming <
programm...@jsoftware.com> wrote:
> Thanks, Chris, and Cliff too.
>
> Yes, a mod inverse helps a lot. Once you've got an inverse, it's e
When is linebreak used? That is, when is it superior to [?
Henry Rich
On 4/1/2023 11:29 AM, Eric Iverson wrote:
I am in favor of linebreak as well as a linejoin. I'd very much like to
have an easier to type NB.
.. NB.
... linejoin (followed by comment)
linebreak
I used to think
; then executing each row of the array. The
characters are Unicode and give spelling error. Use
;@:("."0@":&.>) 40 5 9008 90
Henry Rich
On 3/30/2023 2:52 PM, Richard Donovan wrote:
Hi Henry
I deleted all the spaces and replaced with “real” ones. Still in error
Ran a.i. Wh
My guess is you pasted in a nonbreaking space. The caret points to the
error.
Wouldn't you like release 9.4, which would give more error information?
Henry Rich
On 3/30/2023 12:54 PM, Richard Donovan wrote:
JVERSION
Engine: j902/j64avx2/windows
Release-a: commercial/2020-12-05T13:36:01
I think it's Roger's variation on "fie upon thee".
Henry Rich
On Sat, Mar 18, 2023, 7:46 AM Elijah Stone wrote:
> The phrase 'foo upon thee' appears many times in the j test suite.
>
> A google search reveals a forum post by roger from 1994
> (http://computer-progr
n there.
There are a few comments.
Henry Rich
On 3/14/2023 1:04 AM, Ak O wrote:
lol a few, and I expect many more.
What command restores each in their correct position
Store is?
(3!:1) (0 s: 3) 1!:2<...'
(3!:1) (2 {."1(0) s: 2) 1!:2 <...'
Restore?
???(3!:2) 1!:1<
s: 2) .
Henry Rich
On 3/13/2023 9:02 PM, Ak O wrote:
I hope you are all well.
Is there a mechanism for storing the Global Symbols Data in parts? Also
restoring from those parts?
The documented methods of:
((3!:1) 0 s: 10) 1!:2<'symb_tbl_dat.dat'
10 s: (3!:2)1!:1<'symb_tbl_dat.dat
Please, help us by updating the wiki. At this moment you know more
about this misdescription than anyone else.
Henry Rich
On 3/12/2023 10:01 PM, More Rice wrote:
I still couldn't find the equivalent of xticklabels
Ok. Figured it out. The description of xlabel in jwiki is wrong
is mine.]
Henry Rich
On 3/12/2023 1:07 AM, More Rice wrote:
Hi,
some questions, if I may ...
1. plot can't handle asymptote (or "not in domain" issue) right at the
boundary in its domain?
install 'all'
load 'plot trig numeric'
plot _5 5 ; '%' NB. ok
plot _5 0 ; '%' NB. Bad
What we have agreed to do is to treat a #! line as a comment ONLY when
#! are the first 2 characters of a file loaded by 0!:n (which is used by
the load command).
Henry Rich
On 3/10/2023 7:29 PM, Raul Miller wrote:
Byte order mark is another can of worms, as is wide encoding.
But, yes
"Ancillary Pages" is on the main NuVoc page; there is also a pointer to
the details for !:, with a banner, in the NuVoc page for !: . Why don't
you add a reference to it in the places you looked?
Henry Rich
On 3/10/2023 6:42 PM, Don Guinn wrote:
I used the link to the "Ea
Yes it is: https://code.jsoftware.com/wiki/Vocabulary/Foreigns#m9
Henry Rich
On 3/10/2023 11:06 AM, Don Guinn wrote:
What is 9!:55 ?
Not shown in the !: conjunction.
On Fri, Mar 10, 2023 at 8:46 AM Henry Rich wrote:
If we do that, should we revisit the question of ignoring the BOM (Byte
the shebang line explicitly, rather
than putting our faith in garbage processing?
Henry Rich
On 3/10/2023 12:04 AM, Elijah Stone wrote:
Perhaps it's simply time to tell the interpreter to start ignoring the
first line of a script if it starts with #!...
On Thu, 9 Mar 2023, Raul Miller wrote
The mouseMove handler would have the form selected. Perhaps you need to
select the form, and glselect the isigraph, inside your timer?
paint_base_ ''
is the preferable way to change locales.
Henry Rich
On 3/8/2023 2:45 PM, Ed Gottsman wrote:
Hello. I’ve got a timer firing a callback
sys_timer used to be called in base locale. Now it is called in z, as you
can see from the message. Please fix the wiki.
Henry Rich
On Mon, Feb 27, 2023, 3:34 PM David Lambert wrote:
>
> https://code.jsoftware.com/wiki/Guides/Window_Driver/Window_Driver_Overview#System_Events
>
>
Now that gerund"n applies gerund cyclically, the need for the oblique
trick is reduced.
Henry Rich
On 2/25/2023 3:26 PM, neit...@gaertner.de wrote:
I want to convert the second one into numerical data
Can you simplify the above expression?
Applying a gerund cyclically, as
I agree about lines 164-166 and have no opinion about the rank of constants.
Henry Rich
On 2/24/2023 10:57 AM, Jan-Pieter Jacobs wrote:
I was looking at Raul's pull request for math/calculus here:
https://github.com/jsoftware/math_calculus/pull/5 . This would simplify a
lot of derivatives
ci =. 'abc';'13.2'
(datatype L:0) 0 1 (0&".@])^:[ L:0"0 ci
---++
literal|floating|
---+----+
Henry Rich
On 2/21/2023 11:08 PM, Gilles Kirouac wrote:
I have two character strings :
datatype each 'abc';'13.2'
┌───┬───┐
│l
Rank.
The rank of *: is 0.
Thus ... @ *: has rank 0.
The rank of fork is _ .
*: @: fork (equivalent to [: *: fork) has rank _ , as does (*: @ fork)
Henry Rich
On 2/12/2023 4:46 PM, More Rice wrote:
Masters,
I understand that there is an excellent reference implementation of varp
Good find & fix. (i>>3)&8 was an attempt to fix earlier similar
problems, and it did fix the ones that were reported! The 8 was a
blunder, I meant 7. You are right that i&7 suffices.
Henry Rich
On 2/4/2023 11:07 PM, Elijah Stone wrote:
I think the j implementation doe
-alt.pdf which
talks about forcing newton to converge, which seems promising.
On Sat, 4 Feb 2023, Henry Rich wrote:
Somehow I lost the original post.
For some reason Laguerre's method doesn't converge for this
polynomial. There seem to be large regions of non-convergence.
Changes of up to 1e_5
a random
float, ?0 in J (Julia has a table to avoid computing one).
Removing this, I see a small region of non-convergence around 0, radius
between 0.1 and 0.2. I suppose this isn't what you get?
Marshall
On Sat, Feb 04, 2023 at 08:34:46PM -0500, Henry Rich wrote:
Somehow I lost the original post
toward any
solution. In the implementation the initial guess is always 0j0.
I don't see why this would fail. If anyone on this list can help, or
knows someone who can help, I'd appreciate suggestions.
Henry Rich
These crash for me too. I have fixed the crash, which now produces a
stack error instead.
Sorry I can't give you the right answer. I'll look again after the 9.04
final release is out. Or maybe Elijah will come up with something.
Henry Rich
On 2/3/2023 11:34 PM, Schmidt-Gröttrup, Markus
a sizable niche for small-range integer sort at all
lengths, but the range decreases with length.
Henry Rich
On 2/3/2023 9:59 PM, Marshall Lochbaum wrote:
In case it helps, I took the following measurement with bencharray.
Looks like the parts other than those plateaus in the middle have gotten
with the detail would be
helpful.
I just noticed that some kind person put edit handles into the Foreigns
page so you don't have to scroll through the whole page. Thanks!
Henry Rich
On 2/4/2023 2:38 AM, Raul Miller wrote:
I guess you might want
https://code.jsoftware.com/wiki/Vocabulary
Integer sort chooses between 4 algorithms based on timings we make every
now and then. It's been a few years since we tuned it.
10 integers or more uses mergesort; 5-9 uses radixsort.
It looks like I'd better retune that. Thanks for the tip.
Henry Rich
On 2/3/2023 5:01 PM
J9.04 is frozen for the release, but J9.05 will follow. This thread has
convinced me that sorting large arrays can be greatly improved by
multithreading, and also that I can reduce the threading overhead.
Henry Rich
On 1/27/2023 6:07 AM, vadim wrote:
What result do you get when you use
question is if you could confirm that it (slower
than I expected speed with numerical sort in threads) is neither J issue,
nor 't.', nor '/:~'. Sorry if I wasted your time.
Best regards,
Vadim
On Wed, Jan 25, 2023 at 8:02 PM Henry Rich wrote:
Fixed for the release. Thanks for the clear report.
ts+result. +/ . * is a perfect example. On large matrices the
arguments are cycled through repeatedly.
Henry Rich
On 1/25/2023 7:08 AM, vadim wrote:
Hi, please consider this:
((0,:2),:(2,:2)) (< @: +: @: ]);.0 i. 4
+---+---+
|0 1|2 3|
+---+---+
((0,:2),:(2,:2)) (< @: (+/) @: ]);.0 i. 4
+--
, while ignoring the fact
that if not used in a restricted way it will crash the system: a serious
defect.
Henry Rich
On 1/22/2023 8:51 PM, Raul Miller wrote:
After thinking about this (and sleeping on it), I think this is a good idea.
Thanks
(privately).
Henry Rich
On 1/19/2023 12:16 AM, Elijah Stone wrote:
I don't see why name_. is preferable to lexical closures. It seems
more difficult to use and complicated to describe; not significantly
easier to implement; and it doesn't nest.
Leaving aside performance (which can be resolved
Read- and write-locks as you describe are what we use for public
locales. Each lock requires a RFO cycle, plus more if the lock is
contended.
Private namespaces require no lock, since nothing can ever contend. We
would have to add something to those paths.
Those paths are carefully timed
gging
purposes. (This also relates to my comment in github issue #153.)
What say?
-E
On Tue, 17 Jan 2023, Henry Rich wrote:
Is it about namespaces then? That is indeed a vexed question. I
will take this under advisement. I will need more help I'm sure.
Henry Rich
On 1/17/2023 10:15 PM, Elijah
1 - 100 of 2010 matches
Mail list logo