mersenneforum.org New feature in my ECM applet
 Register FAQ Search Today's Posts Mark Forums Read

 2010-05-23, 11:35 #67 Merfighters     Mar 2010 On front of my laptop 7716 Posts A bug? When I enter a factor during a Rabin-Miller PRP test, then the applet will search for perfect power for a long time, and it will say the factorization is complete. I tried Phi(9816,10)*Phi(24,10). The applet found the factor 9817 easily, and I entered 1059411433 as a factor(the known factor), then the applet searched for perfect power for a long time, and it said the factorization is complete. It says that Phi(9816,10)*Phi(24,10) = 9817*1059411433*(a large prime cofactor). But that cofactor is certainly not a prime. It has a prime factor: Phi(24,10) = 99990001. It is certainly a bug.
 2010-08-20, 16:49 #68 henryzz Just call me Henry     "David" Sep 2007 Cambridge (GMT/BST) 2×5×587 Posts I think I have found a bug in the applet. When I enter x=447213595500007;x=x-1;i-21;(x*x-200000000000034400000000000999) in the box at the bottom it produces some negative numbers as well as positive numbers. The positive numbers are factored correctly but the negative numbers aren't. I realize that factorizing negative numbers wasn't really your plan but it would be helpful sometimes if it was fixed although a factor of -1 would need to be added. When I found this I was searching for a suspected bug in my quadratic sieve program.
2010-08-20, 17:43   #69
alpertron

Aug 2002
Buenos Aires, Argentina

22·3·113 Posts

Quote:
 Originally Posted by henryzz I realize that factorizing negative numbers wasn't really your plan but it would be helpful sometimes if it was fixed although a factor of -1 would need to be added.
I'm busy with "real life" now but I will change the applet as soon as possible in order to add this request.

2010-08-20, 20:11   #70
henryzz
Just call me Henry

"David"
Sep 2007
Cambridge (GMT/BST)

2·5·587 Posts

Quote:
 Originally Posted by alpertron I'm busy with "real life" now but I will change the applet as soon as possible in order to add this request.
Brilliant thanks.

2010-08-21, 06:45   #71
S485122

Sep 2006
Brussels, Belgium

2×5×167 Posts

Quote:
 Originally Posted by henryzz I realize that factorizing negative numbers wasn't really your plan but it would be helpful sometimes if it was fixed although a factor of -1 would need to be added.
Nitpicking : Adding unity factors (even negative) is not a good idea. It would be better if the number is negative to output a minus in front of the result : -6 = - (2 * 3) Otherwise there is no reason not to add pairs of -1 factors to any positive number to be factored... You only have a unique factorisation if the input number is positive, which is why the applet currently gives "Number too low (less than 2)." as answer when you try to factor negative numbers ... except when you use a loop (and that is the real bug ;-)

Jacob

2010-08-21, 08:27   #72
xilman
Bamboozled!

"πΊππ·π·π­"
May 2003
Down not across

2·72·109 Posts

Quote:
 Originally Posted by S485122 Nitpicking : Adding unity factors (even negative) is not a good idea. It would be better if the number is negative to output a minus in front of the result : -6 = - (2 * 3) Otherwise there is no reason not to add pairs of -1 factors to any positive number to be factored... You only have a unique factorisation if the input number is positive, which is why the applet currently gives "Number too low (less than 2)." as answer when you try to factor negative numbers ... except when you use a loop (and that is the real bug ;-) Jacob
Nit-picking: unique factorization is usually interpreted as "unique up to ordering of factors and the presence of units". Personally, I'd not store negative integers in the DB at all.

Paul

2010-08-23, 02:25   #73
alpertron

Aug 2002
Buenos Aires, Argentina

22·3·113 Posts

Quote:
 Originally Posted by henryzz I think I have found a bug in the applet. When I enter x=447213595500007;x=x-1;i-21;(x*x-200000000000034400000000000999) in the box at the bottom it produces some negative numbers as well as positive numbers. The positive numbers are factored correctly but the negative numbers aren't. I realize that factorizing negative numbers wasn't really your plan but it would be helpful sometimes if it was fixed although a factor of -1 would need to be added. When I found this I was searching for a suspected bug in my quadratic sieve program.
I think it is ready now. I also changed some parts of the code in order to require less memory by using arrays of integers instead of arrays of longs (32-bit instead of 64-bit data). Please let me know if you find any error.

2010-08-23, 09:02   #74
henryzz
Just call me Henry

"David"
Sep 2007
Cambridge (GMT/BST)

2×5×587 Posts

Quote:
 Originally Posted by alpertron I think it is ready now. I also changed some parts of the code in order to require less memory by using arrays of integers instead of arrays of longs (32-bit instead of 64-bit data). Please let me know if you find any error.
Seems to be working thanks.

 2010-09-05, 17:05 #75 ThiloHarich     Nov 2005 1458 Posts I did some timings for different JDK or Java compilers. Times in sec. jdk 6_21 .588 -> 100% jdk 7b108 .591 -> 100% jdk 6_16 .61 -> 104% jrokit4.0.1 .974 -> 166% jet7.2 .455 -> 77% Windows XP Intel dual core T9400 2.53 Ghz 32 bit Jet is a Java compiler like gcj. I tried gcj some years ago on windows with cygwin, but had to fight to get an executable (on Windows). http://www.excelsior-usa.com/downloa...tdlevalaw.html http://dlc.sun.com.edgesuite.net/jdk...ies/index.html Last fiddled with by ThiloHarich on 2010-09-05 at 17:05
 2010-09-17, 12:27 #76 alpertron     Aug 2002 Buenos Aires, Argentina 22·3·113 Posts New ECM record when running the applet Robert J DuChateau was running ECM using the applet on 24+29^81 and found the factorization: 347 x 34503928686970842350938851377044353167566697578436905273 (Curve 210438) x 2377058158103859925326651218033244145946260234462377898431463 So he found a 56-digit factor using the applet (a new record). Of course this could have been found faster with SIQS or GNFS (the latter is not implemented in the applet and it will never be done).
 2010-09-17, 13:25 #77 warut   Dec 2009 89 Posts And fastest with SNFS.

 Similar Threads Thread Thread Starter Forum Replies Last Post a1call Programming 19 2019-11-08 22:31 ET_ Lone Mersenne Hunters 69 2014-06-01 17:34 3.14159 Miscellaneous Math 7 2010-06-01 01:29 alpertron Factoring 14 2006-01-01 04:00 jinydu Lounge 2 2004-05-05 08:33

All times are UTC. The time now is 14:04.

Fri May 14 14:04:10 UTC 2021 up 36 days, 8:45, 0 users, load averages: 1.94, 1.91, 1.86