In the session below, the final sentence
valid A
runs "forever". Why?
$ ijconsole
JVERSION
Engine: j803/2014-10-19-11:11:11
Library: 8.04.13
Platform: Linux 64
Installer: unknown
InstallPath: /usr/share/j/8.0.4
LC=: 26}.Alpha_j_
a2v=: _8{.LC&i.
NB. these tests apply to numerical
I think this is a bug as illustrated by your example.
The fix will be in the next base library release. Thanks for
reporting with a clear example.
('^B';'C') rxrplc 'BBB my sentence'
CBB my sentence
('B';'C') rxrplc 'BBB my sentence'
CCC my sentence
Пт, 11 дек 2015, Ryan Eckbo написал(а):
Sorry, no, I meant
lambda=: 3 :
And I do not claim that the definition is equivalent.
I did claim that it is straightforward.
Thanks,
--
Raul
On Thu, Dec 10, 2015 at 5:15 PM, Jose Mario Quintana
wrote:
> Raul writes:
>
> "
> Also straightforward would be this adverb.
>
> lambda= 3 :
> "
NuVoc
(>0 1 )|: y
Don
On 12/10/2015 12:13 AM, Linda A Alvord wrote:
I'm just getting started on these and here is an elementary school solution.
(I couldn't remember a shortcut to get a diagonal from a square matrix.
f=: 13 :'(*/0 1{(/:y){y)+2*+/+/(=/~i.$y)*1|."1*/~(/:y){y'
Linda
-Origi
Ok I see. Unfortunately rxrplc uses rxmatches, so it was messing me up
when I was trying to
replace just the first character:
('^B';'C') rxrplc 'BBB my sentence'
CCC my sentence
On 11 Dec 2015, at 2:52, Alex Shroyer wrote:
If you just want the *first* match, use *rxmatch*. Use *rxmatches* in
Raul writes:
"
Also straightforward would be this adverb.
lambda= 3 :
"
lambda = 3 :
|value error: lambda
| lambda=3 :
Did you mean the following?
lambda=. 3 :
If so,
sf=. [: u0 u1 u2 (u0@[ u1 u2@]) lambda
|domain error: lambda
| sf=.[:u0 u1 u2(u0@[u1 u2@])lambda
I am a
We all had the same solution. My version is
# ,@((# , {.);.1~ (~: |.!.0))^:40 "."0 '1113122113'
Henry Rich
On 12/10/2015 10:16 AM, 'Pascal Jasmin' via Programming wrote:
I did not know fill on |. worked that way. I see it now. neat.
- Original Message -
From: Joe Bogner
To: pro
its pretty much the exact same, T M and T M~.
Its 2 definitions instead of 4 (including the reasonable alternatives you
mentioned). Somewhat addressing your name conflict concerns.
with Y, you can also do this:
2 3 ( 0 Y~) 1 5
2
But then Y is not quite as good a name as M(onad) anymore i
Also straightforward would be this adverb.
lambda= 3 :
This expects a formatted sequence of characters as its argument, but
it's quite a robust (and well documented) implementation.
Thanks,
--
Raul
On Thu, Dec 10, 2015 at 2:21 PM, Jose Mario Quintana
wrote:
> I would rather use strand notati
I agree that M is simple. It reminds me of the definition of X and Y
that has been previously posted. How does it compare to that?
Quick search found a definition like this:
Y =: (&{::)(@:])
X =: (&{::)(@:[)
but it could also be defined as
X=: @[
Y=: @]
3 4 (+:X + -:Y) 2 4
7 10
Regarding
Regarding the first release of Jx see [0].
Personally, regardless of the availability of Jx, as I said, I prefer in
this kind of case, an adverb using strand notation, and one should be able
to produce it using an official J (Pascal might give it a try);
furthermore, one should also be able to pro
took me a few years to realize that dyadic +:is not always a domain error.
(its a boolean op). Nearly all of the builtins are ambivalent.
I think M is simple and clean enough to use, and I like that you can change the
X or Y reference by just removing/adding ~.
That's just me, but I thought
Here is my template for just a straight fork. useful for allowning n v n
constant verb binding.
Fork =: 1 : ' 2 : (''u '' , ''('', u lrA , '')'' , '' v"_'')'
+/ % Fork #
+/ % #"_
FxHy =: 1 : ' 2 : (''u@[ '' , ''('', u lrA , '')'' , '' v@]'')'
+: % FxHy -:
+:@[ % -:@]
These are adv
shorter:
sfork=:(1 : '(4 0 1 5 6 2 3) C. (;: ''@[ @]'') , (1{:: 0{:: u`[)') 1 : 'u`:6'
3 4 (+: + -:) sfork 2 4
Here's a version that will check to see if the tines are in a list of
monads (proof of concept)
buildTrain is a mess to look at
NB. creates and documents a function
NB. could be impro
On Thu, Dec 10, 2015 at 2:21 PM, Jose Mario Quintana
wrote:
> I would rather use strand notation (a multiple adverb). Using Jx and the
> adverb lambda mentioned in [0] is straightforward:
Personally, I don't even remember where to download Jx - nor do I know
where to find docs on how it differs
I would rather use strand notation (a multiple adverb). Using Jx and the
adverb lambda mentioned in [0] is straightforward:
sf=. [: u0 u1 u2 (u0@[ u1 u2@]) lambda
+: + -:sf
+:@[ + -:@]
3 4 +: + -:sf 2 4
7 10
type'sf'
┌──┐
│adverb│
└──┘
[0] [Jprogramming] Can whatever be wr
Oh, that is clever but not too mysterious. I also went down the path
of trying to get the linear representation of u but got stuck because
5!:5 expects a name. I didn't think to assign a local name.
This was also cool and the piece I was missing to gerundify u
(+: + -:) (1 : '0{:: u`[')
┌─┬──
Here's a draft implementation of what I think you mean by a "smart fork":
NB. draft
sfork=:1 :0
impl=. 0 {:: u`-
if. (<,'3')-:{.impl do.
d=. 5!:1@<
e=. 5!:0
'`a b c'=. 1 {:: impl
(d'a')e@[ (d'b')e (d'c')e@]
end.
)
Example use:
(+-*)sfork
+@[ - *@]
Hopefully that's not to
Interesting. I'm not sure if M is worth it over @] due to potential
naming conflicts and ambiguity in the code .@] is very clear
It got me thinking that it might be interesting (albeit probably not
overly useful) to have a 'smart fork' adverb that could figure out
that its arguments are monadic an
On 03/12/2015 9:10 AM, bill lam wrote:
Right, 9!:36/37 is needed but not sufficient for Jconsole, since
it also depends on the scrolling capability of the terminal emulator.
Thanks for this additional info, which i was not aware of.
I am using jconsole under windows 7 command prompt.
Scrolling
Here is another approach to simplifying J and meeting dyadic J goals without
verb definitions:
M =: @]
3 4 (+:M~ + -:M) 2 4
7 10
(+: M~ + -: M) 2 4
5 10
~ (swap) basically works like @[ on a monad that is defined to ignore x rather
than return error.
M or M~ is easier to type than @
Hi!,
Sorry for the late response.
I was subscribed with digest mode only.
Took me some time to find how to respond to individual messages.
So thanks for your help.
Which is, of course, exactly what i needed.
--
For information
Thanks Raul, I got to that section of phrases but didn't scroll down far
enough.
I like the fork in the tree better.
m7=:(<0 1)&|:
f=: 13 :'(<0 1) |:y'
5!:4 <'m7'
-- <0 1
-- & -+- |:
5!:4 <'f'
-- <0 1
--+- |:
L- ]
m7
(<0 1)&|:
f
(<0 1) |: ]
> On Thu, Dec 10, 2015 at 3:52 AM, Ryan Eckbo wrote:
Also someone could probably write a clever verb to generate the table.
>>
I don't know if it's clever, but here's one that matches Raul's input matrix
(|@((_0.1)&+)@{. , ] ) |: (10 10 $ _0.2,(10 # 0)) + (i.10) +/ (10 # 0.2)
0.1 1.1 2.1 3.1 4
If you just want the *first* match, use *rxmatch*. Use *rxmatches* in when
you want to keep doing *rxmatch* on the rest of the string after finding a
match, until the end of the string.
On Thu, Dec 10, 2015 at 6:33 AM, bill lam wrote:
> I think this is the way that rxmatches works, it iterates
I would be tempted to express your S like this:
SM=: 0 10#: 10* 1+ ".;._2 noun define
0.1 1.1 2.1 3.1 4.1 5.1 6.1 7.1 8.1 9.1NB. initial
0.0 1.2 2.2 3.2 4.2 5.2 6.2 7.2 8.2 9.2NB. 0
0.2 1.0 2.2 3.2 4.2 5.2 6.2 7.2 8.2 9.2NB. 1
0.2 1.2 2.
I did not know fill on |. worked that way. I see it now. neat.
- Original Message -
From: Joe Bogner
To: programm...@jsoftware.com
Sent: Thursday, December 10, 2015 9:40 AM
Subject: Re: [Jprogramming] advent 10
Thanks, the fill is needed otherwise it would return the wrong result
if
On Thu, Dec 10, 2015 at 3:13 AM, Linda A Alvord wrote:
> (I couldn't remember a shortcut to get a diagonal from a square matrix.
m7, m8, m9 and m10 from
http://www.jsoftware.com/help/phrases/special_matrices.htm get the
main diagonal (leading from the upper left corner to the lower right
of the t
Thanks, the fill is needed otherwise it would return the wrong result
if the last number is the same as the first. I don't think my version
is susceptible to that but would be interested to find an example if
it is
correct:
(,@((#,{.);.1~ _1&(|.!._) ~: ])) (1 1 1 3 2 2 2 1 1 1)
3 1 1 3 3 2 3 1
You may have gotten lucky with _1&|. (fill not needed afaiu)
an improvement to mine, after seeing yours.
# ,@( ((# , {. );. 1)~ 1 , 2 ~:/\ ])^:(40) 1 1 1 3 2 2 2 1 1 3
- Original Message -
From: Joe Bogner
To: programm...@jsoftware.com
Sent: Thursday, December 10, 2015 8:47 AM
Subj
Clever use of the state machine. I enjoy seeing those...
Here's mine:
NB. part 1
# (,@((#,{.);.1~ _1&(|.!._) ~: ]))^:40 input=:(1 1 1 3 2 2 2 1 1 3)
252594
NB. part 2
# (,@((#,{.);.1~ _1&(|.!._) ~: ]))^:50 input=:(1 1 1 3 2 2 2 1 1 3)
3579328
To explain, we are going to use cut to split the
I think this is the way that rxmatches works, it iterates by
beheading the right argument so that ^B match 3 times.
Not sure this is a bug or a feature.
Чт, 10 дек 2015, Ryan Eckbo написал(а):
> I believe there should only be one match. Why are there 3?
>
> load'regex'
> '^B' rxmatches 'BBB my s
I believe there should only be one match. Why are there 3?
load'regex'
'^B' rxmatches 'BBB my sentence'
0 1
1 1
2 1
--
For information about J forums see http://www.jsoftware.com/forums.htm
Here's a way of doing Advent 1 in Grade.
Find all the first kind. Count them. Do the same with the second. Then
subtract. Discuss a possible meaning when "It can't be done"
s=: 13 :'(?y#2){''()'''
[S=:s 10
)()))(
f=: 13 :'(+/''(''=y)-+/'')''=y'
f S
_6
Linda
-Original Message
Oops, typo in the title - this is for day 10, not 8.
On 10 Dec 2015, at 19:52, Ryan Eckbo wrote:
A previous advent answer got me thinking about state machines, so I
wrote
one for this problem. The extra initial state ruins the natural
mapping
between states and numbers, but I think it's unav
A previous advent answer got me thinking about state machines, so I
wrote
one for this problem. The extra initial state ruins the natural
mapping
between states and numbers, but I think it's unavoidable? Also someone
could probably write a clever verb to generate the table.
NB. state machin
I'm just getting started on these and here is an elementary school solution.
(I couldn't remember a shortcut to get a diagonal from a square matrix.
f=: 13 :'(*/0 1{(/:y){y)+2*+/+/(=/~i.$y)*1|."1*/~(/:y){y'
Linda
-Original Message-
From: programming-boun...@forums.jsoftware.com
[mailto
37 matches
Mail list logo
