yafu.pl Usage Currently

[EdH]

I am apparently getting a lot of collisions, maybe from my own machines. I changed the "mindig=" to 98, since there are currently a couple pages worth, but the ratio of 0 accepted factors is quite high.

Would it be better to set each machine to a different "mindig=" value, as I am about to try, or is there a better way?

Any thoughts?

[pinhodecarlos]

I'm on Windows 64 bit. I'll look here on the forum for the Yafu version. Thank you.

[yoyo]

Originally Posted by EdH

I am apparently getting a lot of collisions, maybe from my own machines. I changed the "mindig=" to 98, since there are currently a couple pages worth, but the ratio of 0 accepted factors is quite high.

Would it be better to set each machine to a different "mindig=" value, as I am about to try, or is there a better way?

Any thoughts?

Can you post some yafu.pl output here?

[Antonio]

Originally Posted by EdH

I am apparently getting a lot of collisions, maybe from my own machines. I changed the "mindig=" to 98, since there are currently a couple pages worth, but the ratio of 0 accepted factors is quite high.

Would it be better to set each machine to a different "mindig=" value, as I am about to try, or is there a better way?

Any thoughts?

You could try increasing the random offset used to reduce the possibility of collision i.e.

while(1){
    print "get composites\n";
    my $rand=int(rand(10)); # increase this if collision rate is high <------
  my $contents = get("http://factordb.com/listtype.php?t=3&mindig=10&perpage=1&start=$rand&download=1");

[henryzz]

Originally Posted by Antonio

You could try increasing the random offset used to reduce the possibility of collision i.e.

while(1){
    print "get composites\n";
    my $rand=int(rand(10)); # increase this if collision rate is high <------
  my $contents = get("http://factordb.com/listtype.php?t=3&mindig=10&perpage=1&start=$rand&download=1");

If the default is 10 then collisions would be expected.

[Antonio]

Originally Posted by henryzz

If the default is 10 then collisions would be expected.

That appears to be the case for the code at https://www.rechenkraft.net/wiki/ind...ctordb#yafu.pl, and I agree - which is why my code adjusts the range of the random offset depending on the number of collisions detected.

[EdH]

Originally Posted by yoyo

Can you post some yafu.pl output here?

This machine is set at mindig=70:

get composites
Factoring 82 digits: 8824723598733329480027149634177382412509478707463229244601576697159989183192516231


fac: factoring 8824723598733329480027149634177382412509478707463229244601576697159989183192516231
fac: using pretesting plan: normal
fac: using tune info for qs/gnfs crossover
div: primes less than 10000
rho: x^2 + 3, starting 1000 iterations on C82 
rho: x^2 + 2, starting 1000 iterations on C82 
rho: x^2 + 1, starting 1000 iterations on C82 
pm1: starting B1 = 150K, B2 = gmp-ecm default on C82
ecm: 30/30 curves on C82, B1=2K, B2=gmp-ecm default
ecm: 67/67 curves on C82, B1=11K, B2=gmp-ecm default

starting SIQS on c82: 8824723598733329480027149634177382412509478707463229244601576697159989183192516231

==== sieving in progress ( 2 threads):   53984 relations needed ====
====            Press ctrl-c to abort and save state            ====
54258 rels found: 22388 full + 31870 from 409689 partial, (864.76 rels/sec)

failure to equate relation
Q = 33295584411235026159695317064955691174848901402045760033248703719070968513367774688, RHS = 332473758670508514024936687884407699743321722609172692364888471226272027282380256
SIQS elapsed time = 514.4458 seconds.
Total factoring time = 522.4666 seconds


***factors found***

P36 = 197026036775478773525199325149414661
*****
P47 = 44789631579452377127192895097525441928684057371
*****
1
===========================================================================
report factors
    New factors added:        0 / 112
    Factors already known:    2 / 465
    Small factors:            0 / 746
    Errors (does not divide): 0 / 0
===========================================================================

Originally Posted by Antonio

You could try increasing the random offset used to reduce the possibility of collision i.e.

I must confess to having thought of that, but not looking for the very obvious line of code, yet.

Originally Posted by henryzz

If the default is 10 then collisions would be expected.

Indeed! I really should have looked...

Originally Posted by Antonio

That appears to be the case for the code at https://www.rechenkraft.net/wiki/ind...ctordb#yafu.pl, and I agree - which is why my code adjusts the range of the random offset depending on the number of collisions detected.

I still have to try out your code, instead of stubbornly sticking with the other...

[henryzz]

There are 100k 60 digit comps that have just been added

[pinhodecarlos]

Originally Posted by henryzz

There are 100k 60 digit comps that have just been added

Were 120k but not for long. I've switched on a few cores to run yafu.pl.

[EdH]

Originally Posted by pinhodecarlos

Were 120k but not for long. I've switched on a few cores to run yafu.pl.

I will run there as well. I thought we were to leave <70 to the db. Sorry for my misunderstanding...

[henryzz]

Got a core doing the C60s. Shouldn't take that long.