mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Conjectures 'R Us (https://www.mersenneforum.org/forumdisplay.php?f=81)
-   -   Prime testing software suggestions please. (https://www.mersenneforum.org/showthread.php?t=17911)

ishkibibble 2013-03-10 23:08

Prime testing software suggestions please.
 
I'm trying to determine if (30^78857-1)/29 is prime.
Three types of software I used have crashed testing this.
Thank you for any pointers.

firejuggler 2013-03-10 23:17

you want to look for pfgw.
(30^78857-1)/29 is 3-PRP! (51.9943s+0.0034s)
so, most probably prime.

henryzz 2013-03-11 00:24

[QUOTE=ishkibibble;332720]I'm trying to determine if (30^78857-1)/29 is prime.
Three types of software I used have crashed testing this.
Thank you for any pointers.[/QUOTE]

(30^78857-1)/29 -1 = 30*(30^78856-1)/29

(30^78856-1) has algebraic factors since 78856 isn't prime.
You might be able to use this to make a primality prove with a N-1 test.
You would need to factor I think 33% of N-1 and then put them in a helper file for PFGW.

firejuggler 2013-03-11 00:44

henryzz i'm afraid you made a mistake there...
30*(30^78856-1)/29 is composite: RES64: [D3B186476037718B] (304.1575s+0.0032s)
maybe you meant
((30*30^78856)-1)/29?

henryzz 2013-03-11 01:24

[QUOTE=firejuggler;332736]henryzz i'm afraid you made a mistake there...
30*(30^78856-1)/29 is composite: RES64: [D3B186476037718B] (304.1575s+0.0032s)
maybe you meant
((30*30^78856)-1)/29?[/QUOTE]

(30^78857-1)/29 [COLOR="Red"]-1[/COLOR] = 30*(30^78856-1)/29
Notice the -1.
It is interesting how the lack of special modulus slowed the testing down by a factor of 6.

firejuggler 2013-03-11 01:54

my bad.

ishkibibble 2013-03-11 20:14

primetest
 
I was able to complete one test where the value is shown as prime. I will test the same value on different software and platforms.
PFGW is on the list for benchmarking.
ty!

firejuggler 2013-03-11 20:34

If you have small prime - below 8000 digits- and want *definite* proof ,there is a program called 'primo' [URL="http://www.ellipsa.eu/"]here[/URL]
certification take from a few second for a 300 digits prime to a few day for a 8000 digits prime

Batalov 2013-03-11 21:58

[QUOTE=ishkibibble;332875]I was able to complete [B]one test where the value is shown as prime[/B]. I will test the same value on different software and platforms.
PFGW is on the list for benchmarking.
ty![/QUOTE]
And what test would that be?

How far have you factored (30^78856-1)/29 (which will produce helpers for the N-1 test)?

Batalov 2013-03-11 22:03

[QUOTE=ishkibibble;332720]I'm trying to determine if (30^78857-1)/29 is prime.
Three types of software I used have crashed testing this.
Thank you for any pointers.[/QUOTE]
It is a [URL="http://www.primenumbers.net/prptop/searchform.php?form=%2830^78857-1%29%2F29&action=Search"]known PRP[/URL], found by R.Price.

firejuggler 2013-03-11 22:03

[URL="http://factordb.com/index.php?query=%2830%5E78857-1%29%2F29-1"]here[/URL] had a free core, ran a few curve/pm1 still a looooooooooooong way before using N-1

Puzzle-Peter 2013-03-12 16:13

This question probably proves my ignorance when it comes to factoring: How did you find factors with thousands of digits?

henryzz 2013-03-12 16:18

[QUOTE=Puzzle-Peter;332975]This question probably proves my ignorance when it comes to factoring: How did you find factors with thousands of digits?[/QUOTE]

Algebraic factors.
2^(2n)-1 = (2^n-1)*(2^n+1)

Puzzle-Peter 2013-03-12 16:36

[QUOTE=henryzz;332977]Algebraic factors.
2^(2n)-1 = (2^n-1)*(2^n+1)[/QUOTE]

Haha, give me the right side of the equation and I give you the left side in no time at all. Doing it in reverse never crossed my mind... thanks!

ishkibibble 2013-03-13 21:37

(Re: It is a [URL="http://www.primenumbers.net/prptop/searchform.php?form=%2830%5E78857-1%29%2F29&action=Search"]known PRP[/URL], found by R.Price. )
Thank you Batalov. I looked for but couldn't find this kind of info so I had to roll it.
Many computer cycles saved!

gd_barnes 2013-03-14 08:41

[QUOTE=ishkibibble;333242](Re: It is a [URL="http://www.primenumbers.net/prptop/searchform.php?form=%2830%5E78857-1%29%2F29&action=Search"]known PRP[/URL], found by R.Price. )
Thank you Batalov. I looked for but couldn't find this kind of info so I had to roll it.
Many computer cycles saved![/QUOTE]

I think you are confused. You still have not proven its primality so no computer cycles have been saved. Just because it is a PRP does not mean that it is prime. There is not currently an easy way to test a > 100,000 digit number for primality unless n-1 or n+1 can be factored to 33%.

When you made the bold and clearly incorrect statement that "I was able to complete one test where the value is shown as prime.", Batalov was being sarcastic when he said "And what test would that be?" knowing that it is not easily possible to prove its primality.


All times are UTC. The time now is 10:18.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.