mersenneforum.org "I am experimenting with heuristic sieves."
 Register FAQ Search Today's Posts Mark Forums Read

2021-08-28, 23:15   #12
Dobri

May 2018

C216 Posts

Quote:
 Originally Posted by Dr Sardonicus The question of whether factoring Mp - 1 would help in factoring Mp was discussed on this very forum about a year ago, here. It specifically mentions the exponent 1277. And, as of this posting, the thread title is still visible on the Mersenne Forum home page. The short answer to the question is "No."
Thank you for providing a link to a previous thread, of which I was unaware about, discussing this matter.
I did not intend to create a separate thread or provide extensive details on a work in progress but rather to post a few hints and see if there are similar attempts for factorization of composite Mersenne numbers.
One heuristic scheme I am currently testing and modifying is to use the factors of 2p-1-1 (which are obviously coprime with Mp) as a factor base in a version of the rational sieve and proceed in a series of iterations in potentially finding factors of Mp. As pointed out in a previous post, the factors of 2p+1-1 are also known. This gives me an idea to attempt a hybrid rational sieve using a combined factor base.

2021-08-28, 23:32   #13
Dobri

May 2018

3028 Posts

Quote:
 Originally Posted by tuckerkao x11-1 = didn't get any Google Search results, but since there are answers for x5-1 and x5+1, the same patterns should be observable.
x11 - 1 = (x - 1)(x10 + x9 + x8 + x7 + x6 + x5 + x4 + x3 + x2 + x + 1)
https://mathworld.wolfram.com/GeometricSeries.html
https://www.wolframalpha.com/input/?i=x%5E11-1

 2021-08-29, 00:33 #14 tuckerkao   "Tucker Kao" Jan 2020 Head Base M168202123 7518 Posts 22^n + 1 have the higher chance to be a prime. I always wondering why nobody is testing 3n - 2 or 3n + 2. Last fiddled with by tuckerkao on 2021-08-29 at 00:34
2021-08-29, 00:36   #15
Viliam Furik

"Viliam Furík"
Jul 2018
Martin, Slovakia

23·5·17 Posts

Quote:
 Originally Posted by tuckerkao ... x2p±1-1 = Always composite
Nope. 22*2-1 - 1 is 23-1, prime. 22*3-1 - 1 is 25-1, prime. 22*3+1 - 1 is 27-1, prime. 22*7-1 - 1 is 213-1, prime.

2021-08-29, 00:40   #16
tuckerkao

"Tucker Kao"
Jan 2020

1E916 Posts

Quote:
 Originally Posted by Viliam Furik Nope. 22*2-1 - 1 is 23-1, prime. 22*3-1 - 1 is 25-1, prime. 22*3+1 - 1 is 27-1, prime. 22*7-1 - 1 is 213-1, prime.
I think I've made the error by not knowing how to put the powers in 2 layers. For some reason SUP & /SUP doesn't work inside another SUP & /SUP bracket.

x2^(p+1)-1 = Maybe a Prime
x2^(p-1)-1 = Always composite, p > 3
x2^p+1-1 = Always composite, p ≥ 3
x2^p-1-1 = Maybe a Prime

Last fiddled with by tuckerkao on 2021-08-29 at 01:10

2021-08-29, 05:52   #17
retina
Undefined

"The unspeakable one"
Jun 2006
My evil lair

2×32×349 Posts

Quote:
 Originally Posted by tuckerkao For some reason SUP & /SUP doesn't work inside another SUP & /SUP bracket.
Nesting BB code used to work in the past, some older posts do that. But an "upgrade" sometime ago broke it.

2021-08-29, 13:57   #18
Dr Sardonicus

Feb 2017
Nowhere

10011011110112 Posts

Quote:
 Originally Posted by tuckerkao x2p+1 = Maybe a prime
Did you mean x2^p + 1?

2021-08-29, 16:10   #19
Dr Sardonicus

Feb 2017
Nowhere

4,987 Posts

Quote:
 Originally Posted by Dobri Thank you for providing a link to a previous thread, of which I was unaware about, discussing this matter.
Quote:
 One heuristic scheme I am currently testing and modifying is to use the factors of 2p-1-1 (which are obviously coprime with Mp) as a factor base in a version of the rational sieve
Good luck with that.

 2021-08-29, 18:29 #20 a1call     "Rashid Naimi" Oct 2015 Remote to Here/There 32×239 Posts FWIW, for odd positive integers a, b & n = a*b, there are relationships between the factors of n-1 and factors of n which can be used to make the solution-space for factors of n smaller. However this reduction may or may not be marginal. Unfortunately for large composites it often (but not always) is. For example: GCD(a-1,n-1) == GCD(b-1,n-1) always holds true. Example: 65341= 361*181 => gcd(65341-1, 361-1) == gcd(65341-1, 181-1) == 180 => 65341 == (2*180+1) * (1*180+1) Pari-GP code: Code: \\EDE-100-A Relevance of factors of n-1 to factors of n by Rashid Naimi 2021/8/29 \\ The following Pari-GP code shows that for odd positive integers a, b & n= a*b, GCD(a-1,n-1) == GCD(b-1,n-1) runningN = 1 runningGcd = 1 theExp = 2 \\determines the upper range for both a & b forstep(a=3,19^theExp ,2,{ forstep(b=3,19^theExp ,2, n=a*b; ag=gcd(a-1,n-1); bg=gcd(b-1,n-1); if(ag > runningGcd && !ispower(n), runningN =n; runningGcd =ag; ); print("** ",n," >> ",ag); if(ag != bg, print("**** Counterexample ****"); next(19); \\\\ Halt RUN if a counterexample is found ); ); }) print("***** ",runningN ," >> ",runningGcd ); factor(runningN) Last fiddled with by a1call on 2021-08-29 at 18:32
2021-08-31, 01:19   #21
tuckerkao

"Tucker Kao"
Jan 2020

48910 Posts

Quote:
 Originally Posted by Dr Sardonicus Did you mean x2^p + 1?
That was what I meant to type in, but the script coding seemed to take the inner layer out, thus the entire expression looked like another equation.

2021-08-31, 01:42   #22
Dr Sardonicus

Feb 2017
Nowhere

137B16 Posts

Quote:
Originally Posted by tuckerkao
Quote:
 Originally Posted by Dr Sardonicus Did you mean x2^p + 1?
That was what I meant to type in, but the script coding seemed to take the inner layer out, thus the entire expression looked like another equation.
Yes, some bbcode tags inside of bbcode tags don't work.

I advise you to hit "Preview Post," and check the preview every time before you post a message. That way, you won't have to bother trying to remember which nested bbcode tags work and which ones don't. If any nested bbcodes are squirrelly, you'll see it in the preview window. You can then edit the message in the Message window (NOT the Preview window). You can check "Preview Message" repeatedly and edit repeatedly until it looks satisfactory, before you hit "Submit Reply."

 Similar Threads Thread Thread Starter Forum Replies Last Post diep Probability & Probabilistic Number Theory 0 2020-11-21 17:14 rogue And now for something completely different 4 2017-07-31 19:36 Cruelty Riesel Prime Search 18 2006-06-25 03:44

All times are UTC. The time now is 19:09.

Thu Oct 21 19:09:33 UTC 2021 up 90 days, 13:38, 1 user, load averages: 1.41, 1.47, 1.52