Note that 1e6 ?@# 1e3 and 1e6 ?@$ 1e3 are both optimised compared to ? 1e6
# 1e3
10 timespacex '?100#1000'
0.0230934 1.67786e7
10 timespacex '100?@#1000'
0.0176543 8.39002e6
10 timespacex '100?@$1000'
0.0186202 8.39002e6
On Fri, Jul 28, 2017 at 10:36 AM, 'Mike Day' vi
Just back from doing a BBC Prom! Very loud... Still some
voice left.
If I'm doing a simulation, I often do ? on a large sample,
eg ?100#1000 or the like, and then "draw" from that
sample in sequence. Usually for independent trials.
For example, for this donation problem, it _might_ b
The default result from 9!:42 is 2, which is the Mersenne Twister
algorithm for roll, which is pretty fast.
FYI,
--
Raul
On Thu, Jul 27, 2017 at 4:27 PM, Jimmy Gauvin wrote:
> I was wondering if making rg tacit would boost performance.
> Thanks to Raul for answering my unvoiced question.
>
>
I was wondering if making rg tacit would boost performance.
Thanks to Raul for answering my unvoiced question.
I guess key, nub and amend are well optimized in the J interpreter.
Jimmy
PS Is roll also optimized for speed? If not should we look into ISAAC or
other fast PRNGs?
On Thu, Jul 27, 2
Ah, i was testing it on too small of a data set.
Thanks,
--
Raul
On Thu, Jul 27, 2017 at 4:33 AM, 'Mike Day' via Programming
wrote:
> Jimmy's original rg outperforms both tock and the best of
> my various attempts at "new" for "larger" problems, eg
>
>10 ts 'rg^:1 #~400'
> 0.633403 3
Jimmy's original rg outperforms both tock and the best of
my various attempts at "new" for "larger" problems, eg
10 ts 'rg^:1 #~400'
0.633403 38272
10 ts 'tock^:1 #~400'
0.751826 34816
10 ts 'new^:1 #~400'
0.793394 39424
NB. less good on smaller ones
30 ts 'rg^:1 #~1
Oh, well, you could make it tacit. For example:
rg=:(((#/.~@])`(~.@])`(0*[)} +/?@##)+-) *
That does not speed it up, though.
Anyways, good luck,
--
Raul
On Thu, Jul 27, 2017 at 12:29 AM, Jimmy Gauvin wrote:
> I was hoping to get rid of some or all of the temporary vars so I guess
> I'll
I was hoping to get rid of some or all of the temporary vars so I guess
I'll try option 2 and see if I can come up with something.
I like the *y instead of y>0
and the #/.~
the }~ was a bit harder to grasp until I realized that } is an adverb
and then the equivalence of "what where} array" and "ar
When going for speed, there's two basic approaches:
(1) try minor variations on the theme - rearrange to see if you can
eliminate anything, or to incorporate any known special code,
(2) rewrite from scratch and see if that's faster.
Personally, I don't see any way to make that one any faster, th
Hi,
this function seems reasonably fast but I think it could be re-arranged to
be a bit more compact,
Would anyone have some tips as how to go about this?
rg =: 3 : '(y-g) + (r #/. r) (~. r =. ? (+/ g =. y>0)$ $y) } 0 $~ $y'
ts 'rg ^:1 ] 400$100'
0.209447 38272
ts 'rg ^:10 ]
fwiw, I had spotted a slip in my version. I wasn't going to bother the
forum, but
might as well mention it in view of your comparisons.
" &. " should have been " &.: " - I put my carelessness down to the
chaleur en France,
unfamiliarity with the iPad and
The correction improves "new"
I like your new version very much Raul. Hard to top that. Hat's off for the
i.@# #/.~@,
part.
Bravo!
Louis
Original Message
From : rauldmil...@gmail.com
Date : 20/07/2017 - 17:45 (EDT)
To : programm...@jsoftware.com
Subject : Re: [Jprogramming] Everyone giving dollars to random o
So... here's a collection of some of the definitions seen here, with
some timing runs:
NB. Xiao-Yong Jin's implementation as revised by Pascal Jasmin
NB. http://jsoftware.com/pipermail/programming/2017-July/047881.html
revised=: ''1 :(0 :0-.LF)
/:~@(
-&1`]@.(=&0)"0
(#/.~@] + ~.@]{[)
My explicit, looping version:
NB. y is number of people and dollars
NB. x is iteration count
hist =: <: @ (#/.~) @ (i.@#@[ , I.) NB. Histogram
iter =: 4 :0
nx =: y$y NB. initialize money
while. x>0
do. sg =: +/nx>0 NB. How many still giving?
nx =: nx-nx~:0NB. Th
7;stats'
>>>>> stddev"1 tick^:(i.10) 45#45
>>>>> 0 0.977008 1.18705 1.73205 2.01133 2.15322 2.46798 2.86832 2.97719 3.1334
>>>>> tick^:1]20#20
>>>>> 3 20 1 15 10 38 6 8 23 24 11 90 33 67 5 13 7 20 1 5
>>>>>
&
You win, Louis! Yes, it looks as if outer-product methods suffer,
naturally enough, with larger problem size. Presumably they're
~ O(n^2) in time and space.
NB. in J806 with avx, fairly pedestrian Lenovo
ts'new ^:5000 #~200' NB. the one I posted eariler
0.758277 20992
ts'run ^:5000 #~2
>>
>>>> On Mon, Jul 10, 2017 at 7:50 PM, 'Pascal Jasmin' via Programming
>>>> wrote:
>>>>>
>>>>> The number of $ available each round is the number of people with >&0.
>>>> The number of potential recipients is
t;>
>>>> The number of $ available each round is the number of people with >&0.
>>> The number of potential recipients is the population. This simplification
>>> though means its possible for someone to pay himself.
>>>>
>>>> A correction to your ve
doesn't seem to happen.
As soon as some have 0, the rest have a "negative return" expectation
each turn.
From: Xiao-Yong Jin
To: "programm...@jsoftware.com"
Sent: Monday, July 10, 2017 3:43 PM
Subject: [Jprogramming] Everyone givi
that doesn't seem to happen.
> >
> > As soon as some have 0, the rest have a "negative return" expectation
> each turn.
> >
> >
> >
> >
> >
> > From: Xiao-Yong Jin
> >
> > To: "programm...@
________
>
>
> From: Xiao-Yong Jin
>
> To: "programm...@jsoftware.com"
>
> Sent: Monday, July 10, 2017 3:43 PM
>
> Subject: [Jprogramming] Everyone giving dollars to random others
>
>
>
>
> I thought this is a good lunch break exercise
__
From: Xiao-Yong Jin
To: "programm...@jsoftware.com"
Sent: Monday, July 10, 2017 3:43 PM
Subject: [Jprogramming] Everyone giving dollars to random others
I thought this is a good lunch break exercise.
http://www.decisionsciencenews.com/2017/06/
His narration states
Quote: “Imagine a room full of 100 people with 100 dollars each. With every
tick of the clock, every person with money gives a dollar to one randomly
chosen other person. After some time progresses, how will the money be
distributed?”
And I came up with this simulation. (45 p
Why do you think that?
Or were you referring to Xiao-Yong Jin's narrative?
Thanks,
--
Raul
On Mon, Jul 10, 2017 at 5:17 PM, Devon McCormick wrote:
> So, you start with $10,000 (=100*$100) but end up with $2,025 (=45*$45),
> how?
>
> On Mon, Jul 10, 2017 at 3:59 PM, Raul Miller wrote:
>
>> I
So, you start with $10,000 (=100*$100) but end up with $2,025 (=45*$45),
how?
On Mon, Jul 10, 2017 at 3:59 PM, Raul Miller wrote:
> I don't really understand what you mean by "mean expectations" - could
> you describe the intent there?
>
> That said, here's how far I've gotten in implementing th
I don't really understand what you mean by "mean expectations" - could
you describe the intent there?
That said, here's how far I've gotten in implementing this concept:
tick=: (0 >. <:) + i.@# +/@:(=/)~ +/@:* ?@# #
(mean,stddev)"1 tick^:(i.10)]100#100
1000
100 0.932034
100 1.41421
I thought this is a good lunch break exercise.
http://www.decisionsciencenews.com/2017/06/19/counterintuitive-problem-everyone-room-keeps-giving-dollars-random-others-youll-never-guess-happens-next/
Quote: “Imagine a room full of 100 people with 100 dollars each. With every
tick of the clock, ev
27 matches
Mail list logo
