Hello everyone.
i have a friend of mine that we intend to test few systems on Java that he knows to code. I will tell him how the systems work to check out and provide him with few millions numbers to test.
I wanted to get my spins from random.org but i couldn't generate more than 10.000 integers per time so i would had to generate and copy lots of times to get 1million+ numbers. I instead found a site that was randomly generating up to 100.000 integers so I did my job up to 2-3 millions much easier.
Is there though any way I can check the true randomness of my results before we start testing with them? I guess a way would be to copy them to excel and filter them per number and see if they are equally split (total numbers/37). Is there any faster or better way/or any site to check out the randomness of my numbers/spins so I can start testing systems with them?
PS: Is there any known site that I can get guaranteed Random numbers in bulk sizes (2-3 millions+) to test them in systems?
thanks!
Hi BellagioOwner,
I don't use Java but you can find the Mersenne Twister algorithm for it here (link:://maths.uncommons.org/).
QuoteMersenneTwisterRNG
A Java port of the fast and reliable Mersenne Twister RNG originally developed by Makoto Matsumoto and Takuji Nishimura. It is faster than java.util.Random, does not have the same statistical flaws as that RNG and also has a long period (219937).
The Mersenne Twister is an excellent general purpose RNG.
There's also a Java version of the 'Die Hard" randomness test suite here (link:s://sourceforge.net/projects/jrandtest/).
You can run chi square test in RX
Set RX to Turbo mode.
Thanks for the replies and sorry for late response. Had some pressing issues these days and i didn't have that much time.
I confirmed with an online tool i found that the random numbers/spins i have, are producing right on the expected average so this means that they are fairly randomized. I can use them for testings on systems i was intending to as an RNG spins.
I will test the MersenneTwisterRNG and the chi square test though. They sound intresting.
Now about the RX... Can the free sample be used for creation and testing of systems for many spins (1MIL+) or is the full version needed for more complicated systems?
Thanks for interest. If anyone would like to have the numbers I will post them here. They are randomly generated and right on the expeted average.
They are for European Roulette 1 million spins. I wanted to post also 2 millions that i have but they excedd 5000KB limit and can't post. sorry :)
Quote from: BellagioOwner on Jun 22, 02:51 PM 2016
I wanted to post also 2 millions that i have but they excedd 5000KB limit and can't post. sorry :)
zip make text file very small.
haha. had totally forgot the zip option. lol :twisted:
you are right!
there...