mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Factoring (https://www.mersenneforum.org/forumdisplay.php?f=19)
-   -   Some help needed with GGNFS syntax. (https://www.mersenneforum.org/showthread.php?t=13451)

3.14159 2010-05-27 22:56

Some help needed with GGNFS syntax.
 
Syntax = <number file : polynomial file : msieve polynomial file>

I already have the number file, I just need the other two files.
Any suggestions? (That do not include ad hominems?)

fivemack 2010-05-27 23:01

To get the msieve polynomial, put the number you want in a file called worktodo.ini and run "msieve -v -np"; after a little while (eight hours for a 125-digit number, eighty hours for 135 digits, 300 hours for larger) it will give you a polynomial.

3.14159 2010-05-27 23:08

worktodo.ini is a text file, correct? [b]yes[/b]

axn 2010-05-27 23:09

you only need the .n file. just running factmsieve.pl/py should do the rest.

3.14159 2010-05-27 23:15

"you only need the .n file. just running factmsieve.pl/py should do the rest."

Don't you mean, polyselect?

fivemack 2010-05-27 23:26

No, factmsieve.pl/.py runs all the other tools for you in the right order.

Mini-Geek 2010-05-27 23:47

Msieve is the best GNFS polynomial finder, the GGNFS tools are the best for sieving, and Msieve is the best for filtering, LA, and square root. (at least for numbers of the sort of size you're likely to be trying to run, and assuming a common CPU architecture)

If this is a GNFS factorization (as opposed to SNFS), then you don't need to generate the polynomial separately. Just put the n in the right file and run factmsieve.py (I'm not sure if the .pl version will run msieve instead of the slower polyselect). By "right file", I mean that it works like this:
You run "python factmsieve.py example" (or replace "example" throughout this post with the name you want all this to run as) with a file "example.n" in the current folder with the file contents being:
[code]n: TheDecimalDigitsOfYourNumber[/code]e.g.
[code]n: 5393165364156033842058422235206821326759903094697482605822376776383843195054519804159
[/code]And to be safe, include a blank line after that one.
Assuming it's set up right, factmsieve.py will then read the n, use msieve to find a GNFS poly for it, then continue with the NFS (sieving and then finishing).

Alternately, you can put the n in worktodo.ini (I think just the n on a single line, with no "n: " or similar before it), use msieve to generate a poly as mentioned previously, then save it as "example.fb" and run "python factmsieve.py example" like above.

3.14159 2010-05-27 23:48

The number I'm attempting to factor is actually just a meager 100 digits or so.

Syntax = <example.n : polynomial file : worktodo.ini> [1] ?

Mini-Geek 2010-05-27 23:56

[quote=3.14159;216405]The number I'm attempting to factor is actually just a meager 100 digits or so.[/quote]
Ok, so use factmsieve.py. GNFS is right for numbers that are about 100 digits. I recommend using a .n file.
Have you read [url]http://gilchrist.ca/jeff/factoring/nfs_beginners_guide.html[/url] ?

One way to make this all much easier, assuming you can get it set up without too much trouble, is to run it through [URL="http://www.mersenneforum.org/forumdisplay.php?f=90"]aliqueit[/URL]. With aliqueit set up to be able to run GNFS using factmsieve.py, run "aliqueit -e -q [your number, without these brackets]" (optionally with a -p at the start to say you want it to run at a low priority. Aliqueit will then see that it's large enough to do GNFS and make factmsieve.py start doing that.

3.14159 2010-05-28 00:00

"Ok, so use factmsieve.py. GNFS is right for numbers that are about 100 digits. I recommend using a .n file.
Have you read [url]http://gilchrist.ca/jeff/factoring/n...ers_guide.html[/url] ?"

Yeah, that's how I got the app running in the first place :P

3.14159 2010-05-28 00:02

I just need to get past the damn syntax, and misc setup to begin factoring anything

The number I want to begin factoring is 4612788901088241110489338309885610592729008898152693541020459649741762187222672923167872911223544489 (100 digits)

And on that note: Please refrain from posting the prime factors of the above. -.-


All times are UTC. The time now is 00:32.

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