old bug ?

Aillas · 2022-03-02, 11:02

Hi,

I'm still using yafu 1.34.5 (I know it's bad, I should update to v2) and I found a strange bug this morning (not sure if it's related to yafu or alequeit.exe?)
Due to an error of my part, I launched 2 times the same factorization, but got 2 different results :
When I launched aliqueit, I've got the same header lines about the first prp:

Code:

[Mar 02 2022, 11:12:08] *** prp1 = 2
[Mar 02 2022, 11:12:08] *** prp1 = 2
[Mar 02 2022, 11:12:08] *** prp1 = 7
[Mar 02 2022, 11:12:08] *** prp3 = 131
[Mar 02 2022, 11:12:08] *** prp3 = 521
[Mar 02 2022, 11:12:08] *** prp5 = 22307
[Mar 02 2022, 11:12:08] *** prp28 = 2223998111369977306688342171

[Mar 02 2022, 11:12:08] Cofactor 2403073886903957117399148772530191181140223846314929788314328684513911507787234809444659989493643 (97 digits)

[Mar 02 2022, 11:12:08] c97: running rho...

Then yafu is launched: But I've got 2 different results. The first time it didn't find any factors, the second times, it found factors:

Code:

****************************
[Mar 02 2022, 11:12:09] c97: running factor (yafu)...
Starting factorization of 2403073886903957117399148772530191181140223846314929788314328684513911507787234809444659989493643
using pretesting plan: normal
no tune info: using qs/gnfs crossover of 95 digits
****************************
rho: x^2 + 3, starting 1000 iterations on C97
rho: x^2 + 2, starting 1000 iterations on C97
rho: x^2 + 1, starting 1000 iterations on C97
pm1: starting B1 = 150K, B2 = gmp-ecm default on C97
current ECM pretesting depth: 0.00
scheduled 30 curves at B1=2000 toward target pretesting depth of 29.85
Finished 30 curves using Lenstra ECM method on C97 input, B1=2K, B2=gmp-ecm default
current ECM pretesting depth: 15.18
scheduled 74 curves at B1=11000 toward target pretesting depth of 29.85
Finished 74 curves using Lenstra ECM method on C97 input, B1=11K, B2=gmp-ecm default
current ECM pretesting depth: 20.24
scheduled 214 curves at B1=50000 toward target pretesting depth of 29.85
Finished 216 curves using Lenstra ECM method on C97 input, B1=50K, B2=gmp-ecm default
pm1: starting B1 = 3750K, B2 = gmp-ecm default on C97
current ECM pretesting depth: 25.34
scheduled 388 curves at B1=250000 toward target pretesting depth of 29.85
Finished 396 curves using Lenstra ECM method on C97 input, B1=250K, B2=gmp-ecm default
final ECM pretested depth: 29.94
c97 cofactor = 2403073886903957117399148772530191181140223846314929788314328684513911507787234809444659989493643
Total factoring time = 158.1769 seconds
Max cofactor found... 2403073886903957117399148772530191181140223846314929788314328684513911507787234809444659989493643

Code:

****************************
[Mar 02 2022, 11:18:14] c97: running factor (yafu)...
Starting factorization of 2403073886903957117399148772530191181140223846314929788314328684513911507787234809444659989493643
using pretesting plan: normal
no tune info: using qs/gnfs crossover of 95 digits
****************************
rho: x^2 + 3, starting 1000 iterations on C97
rho: x^2 + 2, starting 1000 iterations on C97
rho: x^2 + 1, starting 1000 iterations on C97
pm1: starting B1 = 150K, B2 = gmp-ecm default on C97
current ECM pretesting depth: 0.00
scheduled 30 curves at B1=2000 toward target pretesting depth of 29.85
Finished 30 curves using Lenstra ECM method on C97 input, B1=2K, B2=gmp-ecm default
current ECM pretesting depth: 15.18
scheduled 74 curves at B1=11000 toward target pretesting depth of 29.85
Finished 74 curves using Lenstra ECM method on C97 input, B1=11K, B2=gmp-ecm default
current ECM pretesting depth: 20.24
scheduled 214 curves at B1=50000 toward target pretesting depth of 29.85
Finished 216 curves using Lenstra ECM method on C97 input, B1=50K, B2=gmp-ecm default
pm1: starting B1 = 3750K, B2 = gmp-ecm default on C97
current ECM pretesting depth: 25.34
scheduled 388 curves at B1=250000 toward target pretesting depth of 29.85
prp32 = 98334484010724528439277600187127 (curve 6 stg2 B1=250000 sigma=8313459 thread=5)
Finished 72 curves using Lenstra ECM method on C97 input, B1=250K, B2=gmp-ecm default
final ECM pretested depth: 26.17
scheduler: switching to sieve method
prp65 = 24437753562035000197194779984203921061096086814222883868215498509
Total factoring time = 61.0330 seconds
*** prp32 = 98334484010724528439277600187127
*** prp65 = 24437753562035000197194779984203921061096086814222883868215498509

Is it a known issue? Something I do wrong?

Thanks...

bsquared · 2022-03-02, 14:23

Quote:

Originally Posted by Aillas

Hi,

I'm still using yafu 1.34.5 (I know it's bad, I should update to v2) and I found a strange bug this morning (not sure if it's related to yafu or alequeit.exe?)
Due to an error of my part, I launched 2 times the same factorization, but got 2 different results :
When I launched aliqueit, I've got the same header lines about the first prp:
...

Then yafu is launched: But I've got 2 different results. The first time it didn't find any factors, the second times, it found factors:

...

Is it a known issue? Something I do wrong?

Thanks...

It's not a bug... ECM is a probabilistic algorithm, meaning there is only a chance that a random curve will find a factor. The more you run it the better chance you have. So you just didn't get lucky the 1st time and did get lucky the 2nd.

Aillas · 2022-03-03, 07:39

Thanks for the explanation.

2022-03-03, 07:39	#3
Aillas Oct 2002 France 2·83 Posts	Thanks for the explanation.