mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Programming (https://www.mersenneforum.org/forumdisplay.php?f=29)
-   -   How to use LLR? (https://www.mersenneforum.org/showthread.php?t=25542)

 sweety439 2020-05-13 17:26

How to use LLR?

1 Attachment(s)
Can someone help me? I want to use LLR to find the smallest (probable) prime of the form (185^n-1)/184 (base 185 repunit, I already know that this is composite for all n<=60000, and this can be prime only for prime n), thus I download the program cllr.exe, and I type this in the sieve file:

[CODE]
ABC (185^\$a-1)/184
a: primes from 60000 to 100000
[/CODE]

but the llr program result is

[CODE]
Invalid ABC format, next data lines will be flushed...
[/CODE]

 paulunderwood 2020-05-13 17:38

[QUOTE=sweety439;545273]Can someone help me? I want to use LLR to find the smallest (probable) prime of the form (185^n-1)/184 (base 185 repunit, I already know that this is composite for all n<=60000, and this can be prime only for prime n), thus I download the program cllr.exe, and I type this in the sieve file:

[CODE]
ABC (185^\$a-1)/184
a: primes from 60000 to 100000
[/CODE]

but the llr program result is

[CODE]
Invalid ABC format, next data lines will be flushed...
[/CODE][/QUOTE]

Run it through PFGW downloaded from sourceforge and use the -f switch for trial factoring if necessary.

From pfgwdoc.txt:

[QUOTE]-f{801} uses only primes which are of the form k*801+1[/QUOTE]

 sweety439 2020-05-13 17:49

[QUOTE=paulunderwood;545276]Run it through PFGW downloaded from sourceforge and use the -f switch for trial factoring if necessary.

From pfgwdoc.txt:[/QUOTE]

So if I want to use LLR to test all numbers of the form (b^n-1)/(b-1) with b=185, 269, 281 and 60000<=n<=100000 with prime n, what should I type?

 paulunderwood 2020-05-13 17:51

[QUOTE=sweety439;545278]So if I want to use LLR to test all numbers of the form (b^n-1)/(b-1) with b=185, 269, 281 and 60000<=n<=100000 with prime n, what should I type?[/QUOTE]

You need to generate a list to input to LLR.

Why bother? PFGW is much better at this sort of thing.

More from the documentation:

[QUOTE]Phi(x,y) Cyclotomic number[/QUOTE]

 Kebbaj 2020-05-22 20:54

[QUOTE=paulunderwood;545280]You need to generate a list to input to LLR.

Why bother? PFGW is much better at this sort of thing.

More from the documentation:[/QUOTE]

The great advantage of pfgw over LLR is that pfgw test any form while LLR is limited to some form. But multithreading on LLR is faster than on the new version of pfgw64 with the -T switch. I don't know if I don't know how to use the -T switch on pfgw.

 paulunderwood 2020-05-22 21:09

[QUOTE=Kebbaj;546235]The great advantage of pfgw over LLR is that pfgw test any form while LLR is limited to some form. But multithreading on LLR is faster than on the new version of pfgw64 with the -T switch. I don't know if I don't know how to use the -T switch on pfgw.[/QUOTE]

We need to put it to the test with a (b^n-1)/(b-1) number of considerable size not necessarily prime.

[code]./llr64 -q"(10^49081-1)/9" -t4
Invalid expression in command line.[/code]

[code]
./pfgw64 -q"Phi(49081,10)" -T4
PFGW Version 4.0.1.64BIT.20191203.x86_Dev [GWNUM 29.8]

Phi(49081,10) is 3-PRP! (7.5663s+0.0012s)
[/code]

 All times are UTC. The time now is 06:26.