Aliqueit.exe discussion - Page 49

Dubslow · 2018-07-11, 07:26

For general detection of drivers (no special cases, no brute forcing), I'd recommend a translation of my Python code in mfaliquot, which is what detects guides/drivers for the blue page:
https://github.com/MersenneForum/Mer...liquot.py#L165

The math behind it is described densely at https://www.rechenkraft.net//aliquot/analysis.html, or in a longer, fluffier way at https://www.rechenkraft.net//aliquot...is.html#guides.

(Though as far as my Python goes, it would probably be a better design if it was object oriented. I'd been meaning to make that redesign for several months, but haven't gotten around to it.)

wombatman · 2018-07-16, 01:53

I recently re-setup aliqueit, and everything works fine except for when msieve is called to do GNFS. The GNFS will complete successfully (using factmsieve.py), but aliqueit does not detect this correctly, and gives the following:

-> Factorization summary written to g106-test.txt
WARNING: gnfs failed to find a factor. This really shouldn't happen.
I'll just run ecm till the end of time or a factor turns up...
Let's hope you don't run out of disk space before either of those.
c106: running 0 auto-increasing ecm curves...

Is there something I'm forgetting to do or that I need to change to fix this?

ChristianB · 2018-07-16, 11:07

What version of aliqueit are you using? I remember that the msieve output format changed between msieve versions and the original aliqueit 1.12 couldn't hadle that. My version 1.13 should work.

wombatman · 2018-07-16, 12:20

Quote:

Originally Posted by ChristianB

What version of aliqueit are you using? I remember that the msieve output format changed between msieve versions and the original aliqueit 1.12 couldn't hadle that. My version 1.13 should work.

I'm using the 1.12 version, so that might be it. I also had the ini file set incorrectly, so that may have been part of the issue as well. I'll see if that fixes it, and if not, I'll try 1.13. Thanks!

unconnected · 2018-09-27, 12:32

Hello. Several times I saw situations when aliquiet can't parse yafu output in case of squares. Is it known issue?

Code:

[Sep 27 2018, 12:18:29] *** Starting 2604200:377 = 1034639966400072947613253269482811664558130444766493129777733926865043690115635343045352006538844 (97 digits)
[Sep 27 2018, 12:18:29] *** prp1 = 2
[Sep 27 2018, 12:18:29] *** prp1 = 2
[Sep 27 2018, 12:18:29] *** prp1 = 7
[Sep 27 2018, 12:18:29] *** prp3 = 293

[Sep 27 2018, 12:18:29] Cofactor 126114086591915278841205908030571875250869142463005013381001210003052619468019910171300829661 (93 digits)

[Sep 27 2018, 12:18:29] c93: running rho...
[Sep 27 2018, 12:18:29] *** c9 = 187991521

[Sep 27 2018, 12:18:29] c9: running rho...

[Sep 27 2018, 12:18:29] c9: running qs (yafu)...
09/27/18 12:18:29 v1.34.5 @ xxx, prp5 = 13711
09/27/18 12:18:29 v1.34.5 @ xxx, prp5 = 13711

[Sep 27 2018, 12:18:29] c9: running qs (msieve)...
[Sep 27 2018, 12:18:29] *** qs failed to find a factor. Ending.

[Sep 27 2018, 12:18:29] Cofactor 670849865574071709549101993970099721948997595817127376617167803025821677712900201341 (84 digits)

[Sep 27 2018, 12:18:29] c84: running rho...

[Sep 27 2018, 12:18:29] c84: running P-1 at B1=11e4...

[Sep 27 2018, 12:18:29] c84: running P+1 x3 at B1=55e3...

[Sep 27 2018, 12:18:30] c84: running 74 ecm curves at B1=11e3...
Curve 6:
Using B1=11000, B2=1873422, polynomial x^1, sigma=495048787
Step 1 took 20ms
Step 2 took 23ms
********** Factor found in step 2: 26142648139997
[Sep 27 2018, 12:18:30] *** prp14 = 26142648139997

[Sep 27 2018, 12:18:30] Cofactor 25661128971387697104812739795434287636339808790313175826706388993739553 (71 digits)
[Sep 27 2018, 12:18:30] *** prp71 = 25661128971387697104812739795434287636339808790313175826706388993739553
[Sep 27 2018, 12:18:30] ERROR: product(factors) != value

Yafu output:

Code:

09/27/18 12:18:29 v1.34.5 @ xxx, prp5 = 13711
09/27/18 12:18:29 v1.34.5 @ xxx, prp5 = 13711

bsquared · 2018-09-27, 13:57

Quote:

Originally Posted by unconnected

Hello. Several times I saw situations when aliquiet can't parse yafu output in case of squares. Is it known issue?

Code:

[Sep 27 2018, 12:18:29] *** Starting 2604200:377 = 1034639966400072947613253269482811664558130444766493129777733926865043690115635343045352006538844 (97 digits)
[Sep 27 2018, 12:18:29] *** prp1 = 2
[Sep 27 2018, 12:18:29] *** prp1 = 2
[Sep 27 2018, 12:18:29] *** prp1 = 7
[Sep 27 2018, 12:18:29] *** prp3 = 293

[Sep 27 2018, 12:18:29] Cofactor 126114086591915278841205908030571875250869142463005013381001210003052619468019910171300829661 (93 digits)

[Sep 27 2018, 12:18:29] c93: running rho...
[Sep 27 2018, 12:18:29] *** c9 = 187991521

[Sep 27 2018, 12:18:29] c9: running rho...

[Sep 27 2018, 12:18:29] c9: running qs (yafu)...
09/27/18 12:18:29 v1.34.5 @ xxx, prp5 = 13711
09/27/18 12:18:29 v1.34.5 @ xxx, prp5 = 13711

[Sep 27 2018, 12:18:29] c9: running qs (msieve)...
[Sep 27 2018, 12:18:29] *** qs failed to find a factor. Ending.

[Sep 27 2018, 12:18:29] Cofactor 670849865574071709549101993970099721948997595817127376617167803025821677712900201341 (84 digits)

[Sep 27 2018, 12:18:29] c84: running rho...

[Sep 27 2018, 12:18:29] c84: running P-1 at B1=11e4...

[Sep 27 2018, 12:18:29] c84: running P+1 x3 at B1=55e3...

[Sep 27 2018, 12:18:30] c84: running 74 ecm curves at B1=11e3...
Curve 6:
Using B1=11000, B2=1873422, polynomial x^1, sigma=495048787
Step 1 took 20ms
Step 2 took 23ms
********** Factor found in step 2: 26142648139997
[Sep 27 2018, 12:18:30] *** prp14 = 26142648139997

[Sep 27 2018, 12:18:30] Cofactor 25661128971387697104812739795434287636339808790313175826706388993739553 (71 digits)
[Sep 27 2018, 12:18:30] *** prp71 = 25661128971387697104812739795434287636339808790313175826706388993739553
[Sep 27 2018, 12:18:30] ERROR: product(factors) != value

Yafu output:

Code:

09/27/18 12:18:29 v1.34.5 @ xxx, prp5 = 13711
09/27/18 12:18:29 v1.34.5 @ xxx, prp5 = 13711

That's puzzling. I can see why it can't find the yafu factors. It's because the log parser in aliqueit first looks for the input number before trying to parse the factors and in this case, yafu doesn't print the input number (it is a special case: feeding a square to siqs, which apparently bypasses the normal factor-writing routine).

But as far as I can tell it should be able to find the factors with msieve.

ChristianB · 2018-09-28, 11:41

I'm looking into why the msieve factors can't be found. I don't have an easy solution for the yafu log parsing other than changing yafu to print the input number in this case.

I don't see this particular problem because I set

Code:

trial_cutoff = 1000000

which equates to 78,498 primes that only take up ~306kB of memory.

ChristianB · 2018-09-28, 12:02

I'm not sure my msieve is working properly. Could you (unconnected) please post the output from your msieve executable?

Code:

msieve -v 187991521

Mine looks like this but this also happens with other numbers so there seems to be something wrong with my msieve build.

Code:

Msieve v. 1.53 (SVN 984)
Fri Sep 28 13:57:45 2018
random seeds: a0f36fe6 5af69a46
factoring 187991521 (9 digits)
commencing number field sieve (9-digit input)
error generating or reading NFS polynomials
elapsed time 00:00:00

unconnected · 2018-09-28, 13:53

Quote:

Originally Posted by ChristianB

I'm not sure my msieve is working properly. Could you (unconnected) please post the output from your msieve executable?

It was v. 1.48 or v. 1.50, can't reproduce problem because I've upgraded my factoring tools recently on that box. Latest svn-build works OK. BTW, aliquiet is also newest version 1.13

Code:

Msieve v. 1.54 (SVN 1025)
Fri Sep 28 13:33:22 2018
random seeds: c534409f 4385af03
factoring 187991521 (9 digits)
p5 factor: 13711
p5 factor: 13711
elapsed time 00:00:00

ET_ · 2018-09-28, 13:56

Quote:

Originally Posted by ChristianB

I'm not sure my msieve is working properly. Could you (unconnected) please post the output from your msieve executable?

Code:

msieve -v 187991521

Mine looks like this but this also happens with other numbers so there seems to be something wrong with my msieve build.

Code:

Msieve v. 1.53 (SVN 984)
Fri Sep 28 13:57:45 2018
random seeds: a0f36fe6 5af69a46
factoring 187991521 (9 digits)
commencing number field sieve (9-digit input)
error generating or reading NFS polynomials
elapsed time 00:00:00

Is it normal trying to factorize a 9-digit input using NFS?

EdH · 2018-09-28, 14:18

Here's a 984 build from an i7:

Code:

Msieve v. 1.53 (SVN 984)
Fri Sep 28 10:10:21 2018
random seeds: d67be873 78c4c89b
factoring 187991521 (9 digits)
commencing number field sieve (9-digit input)
error generating or reading NFS polynomials
elapsed time 00:00:00

And, a 1022 build from the same machine:

Code:

Msieve v. 1.54 (SVN 1022)
Fri Sep 28 09:54:55 2018
random seeds: 03a13a26 f9dd7f25
factoring 187991521 (9 digits)
p5 factor: 13711
p5 factor: 13711
 elapsed time 00:00:00

I would guess a more recent revision is needed.

2018-07-11, 07:26	#529
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 3×29×83 Posts	For general detection of drivers (no special cases, no brute forcing), I'd recommend a translation of my Python code in mfaliquot, which is what detects guides/drivers for the blue page: https://github.com/MersenneForum/Mer...liquot.py#L165 The math behind it is described densely at https://www.rechenkraft.net//aliquot/analysis.html, or in a longer, fluffier way at https://www.rechenkraft.net//aliquot...is.html#guides. (Though as far as my Python goes, it would probably be a better design if it was object oriented. I'd been meaning to make that redesign for several months, but haven't gotten around to it.) Last fiddled with by Dubslow on 2018-07-11 at 07:29

2018-07-16, 01:53	#530
wombatman I moo ablest echo power! May 2013 2²·449 Posts	Not picking up msieve GNFS results I recently re-setup aliqueit, and everything works fine except for when msieve is called to do GNFS. The GNFS will complete successfully (using factmsieve.py), but aliqueit does not detect this correctly, and gives the following: -> Factorization summary written to g106-test.txt WARNING: gnfs failed to find a factor. This really shouldn't happen. I'll just run ecm till the end of time or a factor turns up... Let's hope you don't run out of disk space before either of those. c106: running 0 auto-increasing ecm curves... Is there something I'm forgetting to do or that I need to change to fix this?

2018-09-28, 11:41	#535
ChristianB Apr 2013 Germany 3²·5·7 Posts	I'm looking into why the msieve factors can't be found. I don't have an easy solution for the yafu log parsing other than changing yafu to print the input number in this case. I don't see this particular problem because I set Code: trial_cutoff = 1000000 which equates to 78,498 primes that only take up ~306kB of memory.

2018-09-28, 12:02	#536
ChristianB Apr 2013 Germany 13B₁₆ Posts	I'm not sure my msieve is working properly. Could you (unconnected) please post the output from your msieve executable? Code: msieve -v 187991521 Mine looks like this but this also happens with other numbers so there seems to be something wrong with my msieve build. Code: Msieve v. 1.53 (SVN 984) Fri Sep 28 13:57:45 2018 random seeds: a0f36fe6 5af69a46 factoring 187991521 (9 digits) commencing number field sieve (9-digit input) error generating or reading NFS polynomials elapsed time 00:00:00

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Resuming aliqueit	johnadam74	Aliquot Sequences	4	2016-03-28 12:32
Apparent aliqueit issue with specifying factors	pakaran	Aliquot Sequences	2	2015-09-12 23:10
Using Several Instances of Aliqueit for a large gnfs job	EdH	Aliquot Sequences	6	2011-12-13 18:58
Setting up aliqueit	science_man_88	Aliquot Sequences	185	2011-11-08 12:18
Tried out aliqueit.exe: ggnfs failing	Greebley	Aliquot Sequences	35	2010-02-13 15:23

2018-07-16, 11:07	#531
ChristianB Apr 2013 Germany 3²·5·7 Posts	What version of aliqueit are you using? I remember that the msieve output format changed between msieve versions and the original aliqueit 1.12 couldn't hadle that. My version 1.13 should work.