mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Data

Reply
 
Thread Tools
Old 2021-05-11, 01:39   #89
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

2×3,769 Posts
Default

Quote:
Originally Posted by keisentraut View Post
My script above tries to consider all of this, but it was "hacked" together so please use with caution and review the worktodo.txt manually if it is sane..
Bug report: Exponent 314423. The worktodo entries include a composite factor that was reported by one of the P-1 runs.
Prime95 is offline   Reply With Quote
Old 2021-05-11, 06:18   #90
pinhodecarlos
 
pinhodecarlos's Avatar
 
"Carlos Pinho"
Oct 2011
Milton Keynes, UK

3×17×97 Posts
Default

Prime95 crashed here first time and every time I start it just switches off so quickly that I don’t have time to see if any message appears. Which files should I send across to troubleshoot? My run is stopped for now.
pinhodecarlos is offline   Reply With Quote
Old 2021-05-11, 08:38   #91
keisentraut
 
Jul 2020

23·3 Posts
Default

Quote:
Originally Posted by Prime95 View Post
Bug report: Exponent 314423. The worktodo entries include a composite factor that was reported by one of the P-1 runs.
Should be fixed. For this exponent, the server will not only report the composite factors but also the actual prime factors. Therefore, no factorization is necessary. If the server would only report the composite factor, but none of the prime factors, then my script will attempt to factor it with a crappy Pollard-Rho implementation. This will likely result in a hang, but it should not happen often and I don't want to implement more fancy factoring algorithms.

I am currently doing 503k and 504k range, but I'm very slow and only doing it irregularly with my laptop.

Last fiddled with by keisentraut on 2021-05-11 at 08:39
keisentraut is offline   Reply With Quote
Old 2021-05-11, 09:48   #92
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

2·3,769 Posts
Default

Quote:
Originally Posted by pinhodecarlos View Post
Prime95 crashed here first time and every time I start it just switches off so quickly that I don’t have time to see if any message appears. Which files should I send across to troubleshoot? My run is stopped for now.
worktodo.txt and any save files. Thanks.
Prime95 is offline   Reply With Quote
Old 2021-05-11, 09:49   #93
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

2·3,769 Posts
Default

Quote:
Originally Posted by ATH View Post
Not sure how this "P+1" factor was found in stage1 with B1=550,000 ?
8676163 : P=538825884739595715103

B1 is larger than 5 of 6 factors of P-1 and larger than 4 of 6 factors of P+1.

P+1=25 * 19 * 43 * 15973 * 568091 * 2271287
P-1= 2 *3 * 2953 * 12163 * 288181 * 8676163

Maybe stage2 ran but was not reported?
Stage 1 P-1 hit. 8676163 was included in stage 1.
Prime95 is offline   Reply With Quote
Old 2021-05-11, 11:10   #94
keisentraut
 
Jul 2020

23×3 Posts
Default

I just had an idea how to improve P+1 factoring in case there are already known factors.

One example is 504359. This has a factor 819812563073067041 which one can find with B2=216M and I have done an unsuccessful P+1 run with B1=5M, B2=365M. If I would not have told Prime95 that factor before, than two things could have happened:
  1. P+1 factoring attempt finds the factor. Then we know that we had bad luck and it was not P+1 but rather a slow P-1 attempt. So we should rerun P+1 with a different start value!
  2. P+1 factoring attempt does not find the factor. Then, it was an actual P+1 run, so we don't need to run another.
So, if we do the P+1 on numbers with known factors, Prime95 should report the found factors to the server. Then, we would have an indication if the run was a "proper" P+1 run. Should I change my script in order to not include known factors in the worktodo.txt?
keisentraut is offline   Reply With Quote
Old 2021-05-11, 11:11   #95
axn
 
axn's Avatar
 
Jun 2003

508210 Posts
Default

Quote:
Originally Posted by Prime95 View Post
Stage 1 P-1 hit. 8676163 was included in stage 1.
The 6/5 seed would also have found this factor (as a true P+1) if the B1 were ever-so-slightly bigger!!
axn is offline   Reply With Quote
Old 2021-05-11, 11:17   #96
axn
 
axn's Avatar
 
Jun 2003

2×3×7×112 Posts
Default

Quote:
Originally Posted by keisentraut View Post
P+1 factoring attempt does not find the factor. Then, it was an actual P+1 run, so we don't need to run another.
No, that's not how this works. I try to explain it here, but basically, a "run" is not P+1 or P-1 per se, but rather, for a specific (unknown) factor and a given seed, the run will function as P+1 or P-1. A run can simultaneously be both a P+1 run and a P-1 run if there are multiple factors.

Last fiddled with by axn on 2021-05-11 at 11:18
axn is offline   Reply With Quote
Old 2021-05-11, 11:23   #97
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

2×3,769 Posts
Default

Quote:
Originally Posted by keisentraut View Post
Should I change my script in order to not include known factors in the worktodo.txt?
No.

There is no such thing as a "proper" P+1 run. There is such a thing as a "lucky" P+1 run. Every seed finds half of the P+1 smooth factors. If a P+1 seed on M504359 finds the known factor 819812563073067041 then the seed has only a 50% chance of finding any unknown factors.
Prime95 is offline   Reply With Quote
Old 2021-05-11, 11:28   #98
keisentraut
 
Jul 2020

23×3 Posts
Default

Both of you are right. P+1 can actually find both P-1 and P+1 smooth factors in one and the same run... Forget about this then :)
keisentraut is offline   Reply With Quote
Old 2021-05-11, 14:35   #99
ATH
Einyen
 
ATH's Avatar
 
Dec 2003
Denmark

1100010101102 Posts
Default

Quote:
Originally Posted by Prime95 View Post
Stage 1 P-1 hit. 8676163 was included in stage 1.
How does that work?
ATH is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
New TF assignment strategy Prime95 PrimeNet 103 2012-04-09 07:39
TF strategy davieddy Lounge 34 2012-03-17 02:03
Assignment Strategy diamonddave GPU to 72 18 2011-12-06 19:56
Strategy game Kees Puzzles 4 2006-04-07 07:17
New Strategy Citrix Prime Sierpinski Project 5 2004-10-31 12:25

All times are UTC. The time now is 21:26.


Mon Aug 2 21:26:32 UTC 2021 up 10 days, 15:55, 0 users, load averages: 1.54, 1.47, 1.48

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

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.