00:03 yewscion left 00:07 yewscion joined 00:12 yewscion left 00:14 yewscion joined 00:19 zenmov left, yewscion left 00:56 yewscion joined 01:52 Manifest0 left 02:25 zenmov joined 02:26 kylese left 02:27 hulk joined 02:33 Aedil joined 03:03 MasterDuke joined 03:15 hulk left, kylese joined 03:40 yewscion left, kjp left 03:41 kjp joined, zetaaaa left 03:48 eseyman left 04:07 yewscion joined 04:18 yewscion left 04:32 guifa joined 05:10 guifa left 05:23 yewscion joined 05:28 yewscion left 05:37 yewscion joined 05:42 yewscion left 06:09 leif joined 06:24 leif left 06:31 Sgeo left 06:58 eseyman joined 07:22 yewscion joined 07:27 yewscion left 07:39 yewscion joined 07:44 yewscion left 08:05 abraxxa-home left 08:28 yewscion joined 08:32 yewscion left 08:48 sena_kun joined 09:02 dakkar joined 09:03 dakkar left 09:05 dakkar joined
patrickb How can I pass through named args in a signature? sub pd(*@pos, *%named) { pretty-dump(|@pos, |%named).put; } <- This seems to pass the nameds as a positional. 09:35
lizmat m: sub a(|c) { dd c }; my %h = :42a, :666b; a |%h 09:55
camelia \(:a(42), :b(666)) 09:56
lizmat m: sub a(|c) { dd c }; my %h = :42a, :666b; my @a = 1,2,3; a |@a, |%h
camelia \(1, 2, 3, :a(42), :b(666))
lizmat I don't see that?
09:57 sena_kun left 10:00 sena_kun joined 10:06 sena_kun left
patrickb So |c does the trick. 10:21
Geth advent/main: a96487c518 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | raku-advent-2024/authors.md
Add Richard's 2nd post
 10:41
10:49 gfldex left 11:25 librasteve_ left 11:52 merp left 12:05 merp joined
lizmat and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2024/12/02/2024-49-advention/ 12:48
xinming m: my @x = (1..5, (8..10).list); @x = |@x; @x.raku.say; for @x -> $t { $t.say } 13:09
camelia [1..5, (8, 9, 10)]
1..5
(8 9 10)
xinming m: my @x = (1..5, 8..10); @x = |@x; @x.raku.say; for @x -> $t { $t.say }
camelia [1..5, 8..10]
1..5
8..10
xinming i know I can use 1..5, or use (8..10).list, But still curious, is there other right way to expand a range? 13:10
When I found (my @x = (1..5, 8..10)) returns an array with 2 range objects, which suprise me. :-) 13:11
lizmat perhaps raku.land/zef:lizmat/OneSeq is what you want?
% raku -e 'use OneSeq; say (1..3) >>> (8..10)' 13:12
(1 2 3 8 9 10)
xinming lizmat: Not Seq, I already achieve that with (1..5).list
13:15 gfldex joined 13:38 yewscion joined 13:50 yewscion left
Geth advent/main: 7b9bbadffd | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | raku-advent-2024/authors.md
Update scheduled / published
 13:58
timo .tell raiph the performance of the fibonacci sequence code in raku is almost guaranteed to be due to the caching in the sequence operator, since the number of calls to generate a fibonacci sequence with the naive recursive approach is quadratic, it's not fair to compare one vs the other 14:01
tellable6 timo, I'll pass your message to raiph
antononcube Any experts on SVG here? 14:09
lizmat I know what it stands for :-)
antononcube šŸ‘ 14:10
I am interested in figuring out how much time to invest in making Raku produce SVG-based graphics for exploratory data analysis. 14:11
lizmat I'm pretty sure librasteve would like it :-) 14:12
antononcube Maybe it is a very good idea. But I have spent too much time "investing" in JavaScript graphcs.
lizmat as would I :-)
antononcube @lizmat Ok. Why would you? šŸ™‚ 14:13
lizmat more modules == better :-)
antononcube Ah, yeah, sure.
lizmat seriously, I think I would use it at some point 14:14
especially now since all browsers support them natively (right?)
antononcube Yes. 14:17
That is the main reason I asked -- it seems it is much easier and portable to render SVG graphics in Jupyter notebooks. 14:18
For example, GitHub does not render JavaScript plots in the Jupyter notebooks, but it does render the SVG ones.
lizmat right, and SVG files would be easier to generate, no? 14:19
antononcube I am not sure. That is an interesting general question -- that is why I asked for expert guidance/opinion. šŸ™‚ 14:20
I just made this Jupyter notebook update to show SVG: github.com/antononcube/RakuForPred...et-1.ipynb
Currently, I rarely generate SVG directly, but using the Graphviz framework (for graphs) to generate SVG images is really nice. 14:22
Also, I can make SVG animations with Raku in Jupyter!
Making animations with JavaScript within the Raku kernel is not easy. (To me, at least.) 14:23
14:27 pierrot left, pierrot joined 14:30 zetaaaa joined
xinming Is there a way to get idx with .map method? 14:35
.grep has :k adverb to return idx. is it possible to additional add $idx of element being processed 14:36
holmdunc @antononcube I'm certainly no expert, but was looking at www.pygal.org the other day to produce SVG markup for basic charts server-side (rather than using the Chart.js library in the browser client-side)
xinming m: my @x = (<a b>, <c d>); for @x[]:kv -> $idx, @v { [$idx, @v].raku.say }; @x.map({ .raku.say }, :kv)
camelia [0, ("a", "b")]
[1, ("c", "d")]
$("a", "b")
$("c", "d")
nahita3882 there are options. anonymous state variable is one, or .kv'ing your iterable and then using a function with 2 arity to access index and value is another option 14:37
second one:
m: my @x = <y e s>; @x.kv.map({ say "idx: $^idx, val: $^val" })
evalable6 idx: 0, val: y
idx: 1, val: e
idx: 2, val: s
Raku eval idx: 0, val: y idx: 1, val: e idx: 2, val: s
xinming m: my @x = (<a b>, <c d>); for @x[]:kv -> $idx, @v { [$idx, @v].raku.say }; @x[]:p.map({ .raku.say }, :kv) 14:38
camelia [0, ("a", "b")]
[1, ("c", "d")]
0 => $("a", "b")
1 => $("c", "d")
xinming nahita3882: Thanks. the .kv is what I wanted to ask.
14:41 perlbot left, simcop2387 left
antononcube @holmdunc I experimented with "Chart.js", since I wanted a simpler more pre-canned way to produce graphics than D3.js. Decided to bet / invest on Google Charts. 14:41
xinming I just backlog, And I asked once not long ago. Seems I'm going old and being forgetful. :-(
holmdunc Yeah I tried Google's some years ago, seemed good too. The main thing I wanna avoid with charts is rendering to an actual bitmap image server-side, because then you get stuff like the font rendering looking wrong compared to the rest of the text on screen 14:49
antononcube Also, it seems that LLMs know SVG pretty well. So, "plunging into" SVG graphics via Raku for data analysis might not be that frustrating...
14:50 simcop2387 joined 14:51 MasterDuke left, perlbot joined
For me, the decoration graphics elements are very important. It is easy to make a bar chart (in SVG or whatever.) The hard thing is to have all bar chart options, like, grid lines, ticks, tick labels, axes labels, plot label, callouts, tooltips, etc. That is why I worked with D3.js a lot. (Again, for/within Raku.) 14:53
14:56 euandreh1 joined
holmdunc Yeah it's sometimes hard to even remember the full vocabulary that a given library uses for all these things 14:57
antononcube Right! That is the other reason I went for Google Chart -- old (i.e. good WWW representation), comprehensive documentation, and LLMs seem(ed) to know it well. 14:58
15:07 yewscion joined 15:43 euandreh1 is now known as euandreh 15:59 yewscion left 16:17 yewscion joined 16:21 yewscion left 16:30 yewscion joined 16:31 guifa joined
Geth advent/donaldh-advent: d80d48e36e | (Donald Hunter)++ (committed using GitHub Web editor) | raku-advent-2024/authors.md
Add donaldh for day 24 to authors.md
 16:45
advent: donaldh++ created pull request #123:
Add donaldh for day 24 to authors.md
advent/main: f08104cfa4 | (Donald Hunter)++ (committed using GitHub Web editor) | raku-advent-2024/authors.md
Add donaldh for day 24 to authors.md (#123)
 16:49
[Coke] is there a secret to writing a reduce with a comparator instead of a normal operator? 17:05
I want something like [<] but with a &strictly-monotonic or something.
I guess the secret is: not actually a straight reduction like it would be with the &plus example. 17:11
guifa with a traditiona comparator? .min(&strictly-monotonic) 17:12
or .max(&strictly-monotonic)
err
no
sorry that's not te same thing ignore my thinking outloud
antononcube But why specifying reduce in that way would work? Also, maybe &strictly-monotonic has to be made listable? 17:14
[Coke] m: say [<] <1 2 3 4>; say [+] <1 2 3 4> # these seem to work differently 17:15
camelia True
10
[Coke] and I can implement the latter with a custom routine, but not the former. 17:16
guifa I've decided I need to port Mozilla's Readability
antononcube To be clear, I think you are trying to do [and] |@a.rotor(2=>-1).map({ &strictly-monotic(|$_) }) 17:17
[Coke] yup, just was hoping for the better syntax. 17:22
17:34 dakkar left
ab5tract "and I can implement the latter with a custom routine, but not the former." <-- [Coke], do you mean that you can't create a custom infix that behaves like the comparator? 17:37
Geth advent/main: 0c33485ebb | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | raku-advent-2024/authors.md
Schedule Donald Hunter's post
 17:38
ab5tract just wanted to make sure that you are implementing a slurpy candidate for such a custom infix routine, if so
ISTR having trouble with reduce unless I provided a slurpy version. 17:40
[Coke] trying... 17:49
guifa oh yeah
it does do better with slurpy because of the zero case I think
17:50 yewscion left
[Coke] if I do slurpy, then I can get get the first two args... then I can get the comparator... but then I'm getting 1,2, then True and 3, then True and 4... 17:51
antononcube I should have said "slurpy" instead of "listable" ...
[Coke] so I'm still not able to do the comparison on each two.
so I think I still have to write this with the rotor so I can compare pairwise.
antononcube @Coke Right, that is why I said that I do no expect reduce to work. 17:52
[Coke] Yup.
18:06 sorenson left 18:24 yewscion joined
tbrowder antononcube: re svg i once had an animated svg website that was pretty cool: a model steam loco pulling a couple of cars, working wheels and a blinking road crossing. i built the svg stuff with perl. on my TODO list is to get it running again at raku-rr.net. 18:42
right now that address is just a redirect... 18:43
i should put that code on github... 18:44
guifa [Coke] did you set assoc? 18:46
m: compare
camelia ===SORRY!=== Error while compiling <tmp>
Undeclared routine:
compare used at line 1
guifa err
Compare the following
m: sub infix:<`> (*@args) is assoc<left> { say @args }; 1 ` 2 ` 3 ` 4 ` 5
camelia [1 2]
[True 3]
[True 4]
[True 5]
guifa m: sub infix:<`> (*@args) is assoc<list> { say @args }; 1 ` 2 ` 3 ` 4 ` 5
camelia [1 2 3 4 5]
guifa m: sub infix:<`> (*@args) is assoc<chain> { say @args }; 1 ` 2 ` 3 ` 4 ` 5
camelia [1 2]
[2 3]
[3 4]
[4 5]
[Coke] it's a sub, not an operator. 18:48
though I guess I could make it an operator.
m: sub infix:<WAT>(*@args) is assoc<chain> { @args[0] < @args[1] }; say [[&infix:<WAT>]] <1 2 3 4 3> 18:50
camelia True
[Coke] ^^ doesn't seem to help there.
guifa Sometimes the list can be the easier one -- then you can do more complicated logic since you'll access to every single member 18:52
but you could also maybe do
18:55 sena_kun joined
guifa m: sub foo { say "custom monotonic sub" }; sub infix:<`> (*@args, :$with) is assoc<list> { $with(); say @args }; 1 ` 2 ` 3 ` 4 ` 5 :with(&foo) 18:56
camelia custom monotonic sub
[1 2 3 4 5]
guifa basically, you now have an adverb'ed listy operator, so you can do a customized reduction with whatever function (obviously changing the interior of those subs to match the behavior you want)
[Coke] the list could have hundreds of elements, I'm not going to do individual operators between 19:02
botato Coke: I assume it's for today's advent of code, mscha had a clever way: github.com/mscha/aoc/blob/master/aoc2024/aoc02 19:08
[Coke] mmm. 19:11
19:22 librasteve_ joined
librasteve @antononcube on SVG --- yes please --- take a look at / steal from github.com/librasteve/raku-Math-Po...ns.rakumod for some ideas on objects -> svg 19:41
19:57 abraxxa-home joined
antononcube @librasteve Ok, but why? šŸ™‚ 20:14
Is better / easier to maintain? Is it more learnable than JavaScript? 20:15
[Coke] I could potentially 3d print an SVG? 20:16
librasteve i hear that its very good with LLMs and that you can easily view it in a Chatbook ;-) 20:20
20:28 Aedil left, sjn left
antononcube @Coke I was doing a search earlier to today about 3D plots with/via SVG. 20:31
20:49 sena_kun left 21:05 destroycomputer- is now known as destroycomputers
xinming m: my @x = [1, [2], [], [3]]; @x.grep({ $_ ~~ Positional and $_.elems > 0 }).raku.say; 21:07
camelia ($[2], $[3]).Seq
xinming Is there a better version for $'_ ~~ Positional' please? 21:08
ab5tract xinming: best I could think of was to leverage signatures 21:20
m: multi predicate(@ where * > 0 --> True) {}; multi predicate($ --> False) {}; my @x = [1, [2], [], [3]]; @x.grep(&predicate).raku.say;
camelia ($[2], $[3]).Seq
21:20 eseyman left
nahita3882 we can combine that idea with duckmap 21:28
m: my @x = [1, [2], [], [3]]; say gather @x.duckmap(-> @a where ?* { take @a } )
evalable6 ([2] [3])
ab5tract nahita3882: nice one :) 21:29
nahita3882 signature can also be @a [$, |] by using destructuring 21:30
which says I want at least one element (name of which is not imperative)
thank you ab5tract! 21:34
21:39 Manifest0 joined
ab5tract m: my @x = [1, [2], [], [3]]; say gather @x.duckmap(-> @a [$,|] { take @a } ) 21:45
camelia ([2] [3])
ab5tract wicked!
I dig that that this version doesn't itemize. In practice it probably doesn't matter much when itemization has occurred, but after implementing `is item`, I have come to appreciate the distinction 21:46
nahita3882 oh yeah, i guess that's due to accepting it as @a, so either we can take $@a or for the latter, we can do $a [$, |] 21:50
22:00 zetaaaa left
i've been bitten by itemization mostly when iterating over Hash of Arrays/Hashes of ... when these guys keep them singly, then I put <> here and there until it works :y 22:01
22:15 ashfield left 22:21 yewscion left 22:34 ecocode left 22:36 ecocode joined 22:45 Sgeo joined 23:13 kst`` joined 23:17 kst left
coleman antononcube: the docs site has a build step that requires graphviz to make svgs. they look nice, but a custom alternative would be welcome 23:24
23:27 donaldh_ joined 23:28 donaldh left 23:32 abraxxa-home left
antononcube Yeah, of course. It is interesting point. BTW, PlantUML also uses Graphviz DOT. 23:59

Powered by Cro and the RakuĀ® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!