Reserved for MF - Sequence 4788 - Page 39

10metreh · 2009-08-29, 11:08

Quote:

Originally Posted by unconnected

Another 2000 curves now in progress...

Thanks, that'll do. Anyone want to do the poly search?

Greebley · 2009-09-01, 15:56

I can try using aliqueit to find a poly. I get:

factoring 16181837014125290739138668566669295398316302912886258004758896093690565962635455503848232760339988586007070042
3649548753039700027221666945941 (141 digits)
searching for 15-digit factors
commencing number field sieve (141-digit input)
commencing number field sieve polynomial selection
time limit set to 64.00 hours
searching leading coefficients from 1 to 341452
deadline: 400 seconds per coefficient

So 64 hours I guess - unless one of the other speedier ppl apply.

10metreh · 2009-09-01, 17:59

Quote:

Originally Posted by Greebley

I can try using aliqueit to find a poly. I get:

<snip>

So 64 hours I guess - unless one of the other speedier ppl apply.

To find a poly with msieve without using aliqueit, run:

Code:

msieve -np -v <the_number>

Greebley · 2009-09-01, 19:52

It looks like that is exactly the command that aliqueit is running, so I should get the same result.

What is the other way? I think I have seen mention of another program to use?

mdettweiler · 2009-09-01, 20:19

Quote:

Originally Posted by Greebley

It looks like that is exactly the command that aliqueit is running, so I should get the same result.

What is the other way? I think I have seen mention of another program to use?

The other way is to use the GGNFS polynomial finder tools, pol51opt and pol51m0b. However, those are somewhat more complex to use, and are not nearly as welf-explanatory. (I myself have never used them except in group efforts where the parameters and command lines to use are laid out ahead of time.

)

Fortunately, for numbers of this size, msieve generally produces better polynomials than pol51. Sometimes pol51 will still be used, say, if msieve has a problem and can't generate a polynomial (which happens every once in a while). And for some numbers much larger than the ones we're dealing with here, pol51 is the only option since msieve's parameters for those ranges are still under development. But, for the purposes of this project, pol51 is mostly not needed.

Batalov · 2009-09-01, 21:53

Quote:

Originally Posted by mdettweiler

... welf-explanatory...

Please don't fix it. This is a great portmanteau. well+self! Loved it.

Note that both pol51 and msieve poly searches can be distributed.
If people just take different ranges, like this:
msieve -v insert_number_here -np 1,12000
msieve -v insert_number_here -np 12001,24000
msieve -v insert_number_here -np 24001,36000
msieve -v insert_number_here -np 36001,48000
#etc. or use different step

P.S. Let's just not call msieve (called from inside aliquiet) aliqueit.
aliqueit is a great program, but here it is simply a shell.

mdettweiler · 2009-09-01, 23:35

Quote:

Originally Posted by Batalov

Please don't fix it. This is a great portmanteau. well+self! Loved it.

Hmm, good point. Indeed I did not notice the boo-boo, but now that you mention it, yes, it does make sense.

Greebley · 2009-09-03, 17:08

My computer rebooted in the middle of finding polygons. It was about 2/3 done (maybe 45 hours out of 64). This means it didn't automatically pick the best and I am not exactly sure of the criteria.

It didn't find anything the first day, but did the second. I am attaching the msieve.dat.p file. Note that there are less than 10 - about 7 or so - all seemed to come all in the same iteration.

Is one of these good enough or do we need to search more?

10metreh · 2009-09-03, 17:12

Quote:

Originally Posted by Greebley

My computer rebooted in the middle of finding polygons.

I didn't think polygons were related to GNFS, except that "polygon" and "poly search" are both related to lost parrots.

BTW, those polys don't look good enough - the closest data point we have is a c140, and the poly we used for this had a Murphy E of 2.21e-11. Could someone do a search with pol51?

Greebley · 2009-09-03, 17:56

I don't know why it came out as polygon. Weird.
Years of programming graphics I guess.

henryzz · 2009-09-08, 17:27

if someone could provide parameters for poly51m0b and poly51opt i will put some effort into poly selection

BTW what is the difference between polym0b and polym0n?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Reserved for MF - Sequence 3366	RichD	Aliquot Sequences	470	2021-04-22 02:17
Reserved for MF - Sequence 3408	RichD	Aliquot Sequences	474	2021-03-07 20:28
Reserved for MF - Sequence 276	kar_bon	Aliquot Sequences	127	2020-12-17 10:05
Assignments are reserved but not showing up	prism019	GPU to 72	6	2020-09-21 22:11
80M to 64 bits ... but not really reserved	petrw1	Lone Mersenne Hunters	82	2010-01-11 01:57

2009-09-01, 15:56	#420
Greebley May 2009 Dedham Massachusetts USA 3·281 Posts	I can try using aliqueit to find a poly. I get: factoring 16181837014125290739138668566669295398316302912886258004758896093690565962635455503848232760339988586007070042 3649548753039700027221666945941 (141 digits) searching for 15-digit factors commencing number field sieve (141-digit input) commencing number field sieve polynomial selection time limit set to 64.00 hours searching leading coefficients from 1 to 341452 deadline: 400 seconds per coefficient So 64 hours I guess - unless one of the other speedier ppl apply.

2009-09-01, 19:52	#422
Greebley May 2009 Dedham Massachusetts USA 843₁₀ Posts	It looks like that is exactly the command that aliqueit is running, so I should get the same result. What is the other way? I think I have seen mention of another program to use?

2009-09-03, 17:56	#428
Greebley May 2009 Dedham Massachusetts USA 34B₁₆ Posts	I don't know why it came out as polygon. Weird. Years of programming graphics I guess.

2009-09-08, 17:27	#429
henryzz Just call me Henry "David" Sep 2007 Cambridge (GMT/BST) 13377₈ Posts	if someone could provide parameters for poly51m0b and poly51opt i will put some effort into poly selection BTW what is the difference between polym0b and polym0n?