Re: jeden generator liczb losowych do testowania drugiego?

eGospodarka.pl › Grupy › pl.comp.programming › jeden generator liczb losowych do testowania drugiego? › Re: jeden generator liczb losowych do testowania drugiego?

X-Received: by 10.31.2.144 with SMTP id 138mr30484vkc.5.1513615049713; Mon, 18 Dec
2017 08:37:29 -0800 (PST)
X-Received: by 10.31.2.144 with SMTP id 138mr30484vkc.5.1513615049713; Mon, 18 Dec
2017 08:37:29 -0800 (PST)
Path: news-archive.icm.edu.pl!news.icm.edu.pl!news.nask.pl!news.nask.org.pl!news.unit
0.net!weretis.net!feeder6.news.weretis.net!feeder.usenetexpress.com!feeder-in1.
iad1.usenetexpress.com!border1.nntp.dca1.giganews.com!nntp.giganews.com!g35no33
44217qtk.1!news-out.google.com!v55ni1296qtc.0!nntp.google.com!g35no3344213qtk.1
!postnews.google.com!glegroupsg2000goo.googlegroups.com!not-for-mail
Newsgroups: pl.comp.programming
Date: Mon, 18 Dec 2017 08:37:29 -0800 (PST)
In-Reply-To: <1...@g...com>
Complaints-To: g...@g...com
Injection-Info: glegroupsg2000goo.googlegroups.com; posting-host=77.254.46.188;
posting-account=xjvq9QoAAAATMPC2X3btlHd_LkaJo_rj
NNTP-Posting-Host: 77.254.46.188
References: <1...@g...com>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <a...@g...com>
Subject: Re: jeden generator liczb losowych do testowania drugiego?
From: "M.M." <m...@g...com>
Injection-Date: Mon, 18 Dec 2017 16:37:29 +0000
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Lines: 69
Xref: news-archive.icm.edu.pl pl.comp.programming:211923
[ ukryj nagłówki ]
On Monday, December 18, 2017 at 5:33:04 AM UTC+1, M.M. wrote:
> [...]
> Ciekawe jak się zachowa std::mt19937_64 w takim teście.

Wygląd na to, że std::mt19937_64 dwa testy przeszedł, jeden
co prawda z p-value równym 0.9984, ale przeszedł. Liczb w
obu testach było około 38mld.

Pozdrawiam

P.S.

Pierwszy test:
time ./frnd
rnd seed1: 1048344733
rnd seed2: 707250714
--------------------------------------
loops: 302514177
// kubełek std::mt19937_64 std::ranlux48 chi-square
0 183946412 183955260 0.425576871246
1 92094858 92080380 2.27640767773
2 22458053 22465964 2.785721592
3 3561951 3558532 3.28493912658
4 412856 413504 1.01547748027
5 37076 37688 9.93801740607
6 2788 2676 4.68759342302
7 172 162 0.617283950617
8 11 11 0
chi-square: 25.03101752753471722
chi-dist: 0.000599442989703091512
chi-dist-cum: 0.9984771823024356957
--------------------------------------

^C

real 518m3.047s
user 517m44.904s
sys 0m6.612s

Drugi test:
time ./frnd
rnd seed1: 1800537340
rnd seed2: 2132837258
--------------------------------------
loops: 301268993
// kubełek std::mt19937_64 std::ranlux48 chi-square
0 183192536 183166993 3.56202194683
1 91701620 91722424 4.7186543609
2 22376562 22380635 0.741235849653
3 3547901 3548389 0.0671132730938
4 410212 410556 0.288233517474
5 37260 37082 0.854430721105
6 2726 2730 0.00586080586081
7 163 176 0.960227272727
8 10 8 0.5
chi-square: 11.69777774763957802
chi-dist: 0.04807272058814565609
chi-dist-cum: 0.8347998915301052147
--------------------------------------

^C

real 510m57.016s
user 510m39.752s
sys 0m6.032s