Issue #6183 has been reported by gregolsen (Innokenty Mikhailov). ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183 Author: gregolsen (Innokenty Mikhailov) Status: Open Priority: Normal Assignee: Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-03-21 17:59
on 2012-03-21 17:59
Issue #6183 has been updated by mame (Yusuke Endoh).
Status changed from Open to Rejected
Hello,
Enumerator::Lazy is not a silver bullet; it removes the overhead for
creating an intermediate array, but brings the drawback for calling
a block. Unfortunately, the latter is much bigger than the former.
Thus, in general, Lazy does bring performance drawback.
The worth of Enumerator::Lazy is to extract first some elements from
big array, especially, infinite sequence. For example:
Prime.lazy.select {|x| x % 4 == 3 }.take(10).to_a
The code becomes much complex without Lazy:
a = []
Prime.each do |x|
next if x % 4 != 3
a << x
break if a.size == 10
end
Anyway, this is not a bug. If you have any concrete idea to "fix"
this issue, please reopen the ticket.
Thank you,
--
Yusuke Endoh <mame@tsg.ne.jp>
----------------------------------------
Bug #6183: Enumerator::Lazy performance issue
https://bugs.ruby-lang.org/issues/6183#change-25017
Author: gregolsen (Innokenty Mikhailov)
Status: Rejected
Priority: Normal
Assignee:
Category:
Target version:
ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux]
I benchmarked Enumerator::Lazy and that's what I got:
user system total real
Lazy: 0.690000 0.010000 0.700000 ( 0.733160)
Normal: 0.160000 0.010000 0.170000 ( 0.186695)
It seems like even with 4 chain links and 3000 elements in initial
array, Lazy enumerator is almost 4(!) times slower than the normal case.
Instead of performance benefit we've got 4 times performance drawback.
See test file attached.
on 2012-04-03 10:43
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). File bench.rb added File lazy_enumerator.diff added File code added =begin Finally come up with a concrete idea how to "fix" (based on my first PR https://github.com/ruby/ruby/pull/100). The idea is to keep all blocks (passed with lazy methods like map or select) as (({Proc})) objects inside the enumerator and apply them one by one when value requested ((({to_a})), (({next})), etc) This strategy avoids enumerator chaining on each lazy method call and eliminates fair amount of 'calling the block' with (({rb_block_call})) operations. Here's benchmark results: 2.0.0| ~/projects/ruby(trunk)$ rvm ruby-head 2.0.0| ~/projects/ruby(trunk)$ ruby bench.rb user system total real Lazy enumerator 1.460000 0.000000 1.460000 ( 1.465739) Simple array 0.420000 0.000000 0.420000 ( 0.421446) 0.287671 NaN NaN ( 0.287531) 2.0.0| ~/projects/ruby(trunk)$ rvm system 2.0.0| ~/projects/ruby(trunk)$ ruby bench.rb user system total real Lazy enumerator 0.770000 0.000000 0.770000 ( 0.764750) Simple array 0.370000 0.000000 0.370000 ( 0.382653) 0.480519 NaN NaN ( 0.500364) ruby-head is current trunk compiled, and system ruby - is the same trunk but with my patch applied. Last row in results is ratio between 'Simple array' time and 'Lazy Enumerator' time. So, as you can see, with this patch lazy enumerator becomes almost 2 times faster. It's a 'proof of concept' patch (only map and select added) - let me know if it makes sense. I believe that using this approach and with your help lazy enumerator performance can be improved significantly. I'm attaching the diff along with the main part of the source code just in case it's hard to follow the diff. =end ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-25630 Author: gregolsen (Innokenty Mikhailov) Status: Rejected Priority: Normal Assignee: Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-04-03 10:48
Issue #6183 has been updated by matz (Yukihiro Matsumoto). Status changed from Rejected to Assigned Assignee set to nobu (Nobuyoshi Nakada) Priority changed from Normal to Low Nobu, could you review the patch? And if you don't see any problem, check it in? Matz. ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-25631 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-04-03 11:29
Issue #6183 has been updated by mame (Yusuke Endoh). Hello, matz (Yukihiro Matsumoto) wrote: > Nobu, could you review the patch? And if you don't see any problem, check it in? I didn't read the patch, but I found a weird behavior. I think the approach is interesting, though. $ cat t.rb a = (1..10).lazy p a.map {|x| x + 1 }.to_a p a.map {|x| x + 1 }.to_a p a.map {|x| x + 1 }.to_a # without the patch $ ./miniruby.org t.rb [2, 3, 4, 5, 6, 7, 8, 9, 10, 11] [2, 3, 4, 5, 6, 7, 8, 9, 10, 11] [2, 3, 4, 5, 6, 7, 8, 9, 10, 11] # with the patch applied $ ./miniruby.new t.rb [2, 3, 4, 5, 6, 7, 8, 9, 10, 11] [3, 4, 5, 6, 7, 8, 9, 10, 11, 12] [4, 5, 6, 7, 8, 9, 10, 11, 12, 13] -- Yusuke Endoh <mame@tsg.ne.jp> ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-25633 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-04-03 13:14
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). That's because each time you mapping lazy enumerator another proc objected added to procs array, so in your example you effectively mapping 3 times. I should return new enumerator object rather than modifying existing one while calling lazy map or select (or whatever lazy method). A lot of work should be done to finish this patch: all other lazy methods should be added. Also I'm getting an error while working with big arrays (> 10^4). But if you are all positive about the approach I'll happily proceed and do my best to make this fully work. ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-25638 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-04-03 15:26
Issue #6183 has been updated by trans (Thomas Sawyer). In response to #6250, wouldn't the simplest implementation be to return a new Lazy Enumerator for each lazy call? See https://github.com/rubyworks/facets/blob/master/li... for the basic idea. Maybe storing each block internally in an Enumerator might be a tad more efficient, but I would think the added complexity to Enumerator state would make it not worth it. ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-25640 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-04-04 12:42
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). File new_lazy_enumerator.diff added Here's the new patch attached - problem, mentioned by Yusuke Endoh, fixed - now I'm creating a new copy of enumerator on each lazy method call. Also I fixed an error for big arrays - forgot to gc_mark procs array. Thomas, that's the point - current implementation is very simple and hence very inefficient. It mimics ruby implementations but as soon as we are in the C sources already - we can come up with something more efficient. ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-25651 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-04-04 14:00
Hello,
2012/4/4 gregolsen (Innokenty Mikhailov) <anotheroneman@yahoo.com>:
> Here's the new patch attached - problem, mentioned by Yusuke Endoh, fixed - now
I'm creating a new copy of enumerator on each lazy method call.
Okay, the next problem :-)
(1..10).lazy.select {|x| false }.map {|x| p x }.to_a
should print nothing, but it actually prints 1, 2, ..., 10
with your patch applied. It can be fixed easily, though.
I glanced your patch. It will degrade functional modularity
in enumerator.c. Currently, it not so big problem because
it only implements #map and #select. But I guess implementing
other methods, especially, #cycle and #zip, will make some
functions (process_element and lazy_iterator_block) complex
and hard to maintain.
Thus, until you create the final patch, it is hard to say
whether we can import your patch or not.
on 2012-04-05 10:16
Hi, (12/04/04 20:59), Yusuke Endoh wrote: > I glanced your patch. It will degrade functional modularity > in enumerator.c. Currently, it not so big problem because > it only implements #map and #select. But I guess implementing > other methods, especially, #cycle and #zip, will make some > functions (process_element and lazy_iterator_block) complex > and hard to maintain. Agree. Naturally, this approach which chains lazy enumerator processes directly should be faster than current one. So I want to see this being merged in an extensible way. > Thus, until you create the final patch, it is hard to say > whether we can import your patch or not. And, do not comment out existing code with //. It unnecessarily increases noise in the patch.
on 2012-04-08 11:32
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). File lazy_enumerator_hybrid.diff added (Yusuke Endoh) wrote: > (1..10).lazy.select {|x| false }.map {|x| p x }.to_a > should print nothing, but it actually prints 1, 2, ..., 10 Fixed, thanks. > But I guess implementing > other methods, especially, #cycle and #zip, will make some > functions (process_element and lazy_iterator_block) complex > and hard to maintain. Agree, those methods (especially #cycle) will be hard to implement in terms of procs chaining approach. (Nobu Nakada) wrote: > So I want to see this being merged in an extensible way. I came up with a new hybrid patch. It uses procs chaining to handle lazy #map and #select, and current enumerator chaining approach for other methods. I believe this is an extensible way. We can move forward step by step and we can stop any time. If those tricky #zip and #cycle methods optimization won't worth the code complexity added - we can leave it as it is: based on enumerator chaining. See new lazy_enumerator_hybrid.diff (tests are green except test_inspect). Please, let me know you thoughts about this. ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-25714 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-04-16 17:59
> It's a 'proof of concept' patch (only map and select added) - let me know if it makes sense. > I believe that using this approach and with your help lazy enumerator performance can be improved significantly. That would be awesome. -r
on 2012-05-16 10:16
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov).
File 16_may.diff added
File new_bench.rb added
=begin
Here's an update.
All methods except (({Lazy#cycle})), (({Lazy#zip})) and
(({Lazy#flat_map})) are optimized.
Benchmark results shown below.
I was working in ((<this
branch|URL:https://github.com/gregolsen/ruby/tree/lazy_enumer...)).
Cycle, zip and flat_map are really tough to convert to procs chaining,
however they are working fine in this hybrid solution and can be leave
as it is.
All tests pass except test_inspect.
If this implementation is acceptable then the next step will be to fix
inspect and add more tests to cover different types of chaining:
like chaining of enumerator chained (cycle, zip, flat_map) methods with
procs chained optimized methods.
Please, let me know your thoughts.
Thanks.
======================Lazy#map
user system total real
Trunk 1.420000 0.010000 1.430000 ( 1.461111)
Optimized 0.830000 0.000000 0.830000 ( 0.833560)
======================Lazy#select
user system total real
Trunk 1.270000 0.010000 1.280000 ( 1.274074)
Optimized 0.780000 0.000000 0.780000 ( 0.785303)
======================Lazy#grep
user system total real
Trunk 1.540000 0.010000 1.550000 ( 1.552651)
Optimized 0.780000 0.000000 0.780000 ( 0.783526)
======================Lazy#take_while
user system total real
Trunk 1.260000 0.000000 1.260000 ( 1.257465)
Optimized 0.780000 0.010000 0.790000 ( 0.784682)
======================Lazy#drop_while
user system total real
Trunk 1.030000 0.000000 1.030000 ( 1.030987)
Optimized 0.400000 0.000000 0.400000 ( 0.394112)
======================Lazy#reject
user system total real
Trunk 1.240000 0.000000 1.240000 ( 1.243565)
Optimized 0.780000 0.000000 0.780000 ( 0.781825)
======================Lazy#drop
user system total real
Trunk 4.150000 0.000000 4.150000 ( 4.159758)
Optimized 1.590000 0.000000 1.590000 ( 1.598785)
======================Lazy#take
user system total real
Trunk 0.520000 0.000000 0.520000 ( 0.517902)
Optimized 0.010000 0.000000 0.010000 ( 0.003274)
=end
----------------------------------------
Bug #6183: Enumerator::Lazy performance issue
https://bugs.ruby-lang.org/issues/6183#change-26654
Author: gregolsen (Innokenty Mikhailov)
Status: Assigned
Priority: Low
Assignee: nobu (Nobuyoshi Nakada)
Category:
Target version:
ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux]
I benchmarked Enumerator::Lazy and that's what I got:
user system total real
Lazy: 0.690000 0.010000 0.700000 ( 0.733160)
Normal: 0.160000 0.010000 0.170000 ( 0.186695)
It seems like even with 4 chain links and 3000 elements in initial
array, Lazy enumerator is almost 4(!) times slower than the normal case.
Instead of performance benefit we've got 4 times performance drawback.
See test file attached.
on 2012-07-16 09:03
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). File 2012-07-14.diff added Here's an update: Finally I've fixed the last test for Enumerator::Lazy#inspect - now it supports procs chaining too. Nobuyoshi Nakada, please, see diff attached and let me know your thoughts about this. Thanks. ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-28140 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-07-26 14:38
Issue #6183 has been updated by mame (Yusuke Endoh). Your patch makes the following code stuck. p [1,2,3].to_enum.lazy.cycle.take(10).to_a It should print [1, 2, 3, 1, 2, 3, 1, 2, 3, 1]. -- Yusuke Endoh <mame@tsg.ne.jp> ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-28457 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-07-31 12:15
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). File 31_july.diff added Yusuke Endoh, thanks a lot for pointing out on this issue. Fixed, please see new diff attached. ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-28573 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: nobu (Nobuyoshi Nakada) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-11-19 03:38
Issue #6183 has been updated by zzak (Zachary Scott). Assignee changed from nobu (Nobuyoshi Nakada) to mame (Yusuke Endoh) Yusuke-san seems to be last to review this, what is your opinion? ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-33074 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: mame (Yusuke Endoh) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-12-03 14:28
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). File December_3.diff added I've merged the patch with latest trunk (see latest diff attached), specifically with Enumerator lazy size feature. Also I've removed the ugly case switch: proc entry stores pointer to a function that is executed when iterating over the elements. So now it even resembles the current implementation a bit. Please, let me know your thoughts about all this. Thanks in advance. ---------------------------------------- Bug #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-34361 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: mame (Yusuke Endoh) Category: Target version: ruby -v: ruby 2.0.0dev (2012-03-17 trunk 35075) [x86_64-linux] I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2012-12-09 12:58
Issue #6183 has been updated by mame (Yusuke Endoh). Sorry, maybe I have no time to review your patch. Anyone interested? -- Yusuke Endoh <mame@tsg.ne.jp> ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-34552 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: mame (Yusuke Endoh) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2013-04-28 16:38
Issue #6183 has been updated by zzak (Zachary Scott). Assignee changed from mame (Yusuke Endoh) to zzak (Zachary Scott) This patch is big, but I hope to review it soon as I will be working on Lazy soon ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-39004 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: zzak (Zachary Scott) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2013-04-29 12:10
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). Zachary Scott, thanks for your interest. I'm afraid it doesn't merge into trunk cleanly anymore after introducing lazy #size. I'll try to fix it and let you know when ready. ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-39036 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: zzak (Zachary Scott) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2013-04-29 15:07
Issue #6183 has been updated by zzak (Zachary Scott). @gregolsen Sure, np! ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-39039 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: zzak (Zachary Scott) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2013-05-16 22:50
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). File 16.05.2013.diff added Finally managed to merge. Please see latest diff attached. ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-39364 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: zzak (Zachary Scott) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2013-05-17 14:42
Issue #6183 has been updated by zzak (Zachary Scott). @gregolsen Thank you! I will try to review this soon, before you have to rebase again ;) Mind if I delete the old patches? It might confuse someone looking at the ticket. ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-39372 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: zzak (Zachary Scott) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2013-05-17 20:13
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). Sure, feel free to clean old files. Thanks! ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-39376 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: zzak (Zachary Scott) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2013-05-31 09:23
Issue #6183 has been updated by zzak (Zachary Scott). @gregolsen Looks like some of this was applied, sorry I haven't had a chance to check it yet. Could you check? It may need a rebase, the patch didn't apply cleanly and some of the functions already exist (like append_method()) ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-39580 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: zzak (Zachary Scott) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
on 2013-06-02 21:21
Issue #6183 has been updated by gregolsen (Innokenty Mikhailov). Indeed append_method was exctracted by nobu 2 weeks ago as a refactoring of enumerator_inspect. But that's it, nothing was merged yet. I'm not sure I'll be able to rebase patch in next few weeks - got only android tablet with me. I'll let you know when ready. Thanks a lot for your interest. ---------------------------------------- Feature #6183: Enumerator::Lazy performance issue https://bugs.ruby-lang.org/issues/6183#change-39654 Author: gregolsen (Innokenty Mikhailov) Status: Assigned Priority: Low Assignee: zzak (Zachary Scott) Category: Target version: I benchmarked Enumerator::Lazy and that's what I got: user system total real Lazy: 0.690000 0.010000 0.700000 ( 0.733160) Normal: 0.160000 0.010000 0.170000 ( 0.186695) It seems like even with 4 chain links and 3000 elements in initial array, Lazy enumerator is almost 4(!) times slower than the normal case. Instead of performance benefit we've got 4 times performance drawback. See test file attached.
Enable email notification
Enable multi-page viewPlease log in before posting. Registration is free and takes only a minute.
Existing account
(Switch to SSL-encrypted connection)
NEW: Do you have a Google/GoogleMail or Yahoo account? No registration required!
Log in with Google account | Log in with Yahoo account
Log in with Google account | Log in with Yahoo account
No account? Register here.