mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Aliquot Sequences (https://www.mersenneforum.org/forumdisplay.php?f=90)
-   -   Aliquot ranges and status (https://www.mersenneforum.org/showthread.php?t=11625)

schickel 2009-03-20 07:25

Aliquot ranges and status
 
As of [COLOR="Blue"][B]August 2011[/B][/COLOR], all sequences with a starting number below [B]1000000[/B] have been calculated to a minimum height of 110 digits (some may have dropped back below that size due to recent downdriver activity), and there are [B]9253[/B] open sequences in that range.

We are, I believe, the largest project right now that is actively pursuing sequences throughout the entire range. Any open ended sequence is available for reservation, with the following possible exceptions:

The [B][I]Lehmer Five[/I][/B], [B]276, 552, 564, 660, & 966[/B], along with [B]1074, 1134, 19560 & 204828[/B] are currently maintained by [URL="http://www.loria.fr/~zimmerma/records/aliquot.html"]Paul Zimmermann[/URL]. [B]Please co-ordinate any work with him[/B].

The 74 sequences from [B]1200-10000[/B], with the exception of [B]4788[/B], are maintained by [URL="http://christophe.clavier.free.fr/Aliquot/site/Aliquot.html"]Christophe Clavier[/URL] and associates. [b]Email Christophe Clavier to reserve these.[/b]

[b]4788[/b] is currently reserved as a team project by the members of Mersenneforum. Its primary interest is due to work originally done by [URL="http://www.math.ru.nl/~bosma/Projects/ali.html"]Wieb Bosma[/URL]. His work showed a merge occuring after an extensive run. Further investigation by [URL="http://www.lafn.org/~ax810/aliquot.htm"]Clifford Stern[/URL] showed that it actually merges with [B]4788[/B] rather than [B]16100[/B]. Since the merge occurs after 6400+ lines, the combined sequence pair [B]314718/4788[/B] is the longest open sequence at 9000+ lines. Fortunately, stability was achieved at line 2470 and 171 digits with the loss of the '3' factor, eventually followed by a short downdriver run. Stability has now been reached again at 166 digits. Check the [URL="http://www.mersenneforum.org/showthread.php?t=11615"]4788[/URL] thread for the latest info.

Sequences in the range [B]10116-99840[/B] are of particular interest to [URL="http://www.lafn.org/~ax810/aliquot.htm"]Clifford Stern[/URL]. This range has been completed to a minimum of 110 digits by the forum, with selected sequences being pursued higher by Clifford. [b]If you wish to work in this range, check with Clifford, he may have completed work that has not been uploaded to the Factoring Database yet.[/b] Most sequences that he is working will be called out in the reservation thread here.

[SIZE="1"][b]1000000-4000000[/b] has been completed up to 71 digits by axn. [b]Currently there is no active project working above 1000000.[/b][/SIZE]

Everything else is open for pursuit. Check the reservation [URL="http://www.mersenneforum.org/showthread.php?t=11588"]thread[/URL] for current reservations before embarking on any large-scale effort. Links to sites with more in-depth information are available [URL="http://www.mersenneforum.org/showthread.php?t=11613"]here[/URL].

Advice and help on getting started with Aliquot Sequences are available here, just post any questions or problems you might have.

Advice and help on getting started with large-scale factoring efforts are available over in the [URL="http://www.mersenneforum.org/forumdisplay.php?f=19"]Factoring Forum[/URL].

akruppa 2009-03-24 14:28

I just asked Paul about the sequences he maintains. He mentioned that some time ago, he was looking for someone that could run a web-based interface to give access to these sequences with the option of submitting new factors so the sequence would be updated automatically. Does someone operate a web page like this? Maybe even with ECMNet server attached or some such? He'd transfer these 8 sequences then.

Alex

frmky 2009-03-24 16:16

Except for the ECMNet part, that describes Syd's database, [URL="http://factorization.ath.cx/search.php"]http://factorization.ath.cx/search.php[/URL]. The database understands ELF format for the bulk submission of factors, not Paul's fmt format, but I'll be happy to convert them and submit the sequences.

Greg

Andi47 2009-03-24 16:48

[QUOTE=frmky;166512]Except for the ECMNet part, that describes Syd's database, [URL="http://factorization.ath.cx/search.php"]http://factorization.ath.cx/search.php[/URL]. The database understands ELF format for the bulk submission of factors, not Paul's fmt format, but I'll be happy to convert them and submit the sequences.

Greg[/QUOTE]

Does the database check for integrity? (sometimes aliquot.ub finds nonsense factors due to a bug either in ubasic or in the script.)

Syd 2009-03-24 16:55

[quote=Andi47;166514]Does the database check for integrity? (sometimes aliquot.ub finds nonsense factors due to a bug either in ubasic or in the script.)[/quote]

It does, nonsense factors are simply ignored.

Andi47 2009-03-24 17:48

[QUOTE=Syd;166515]It does, nonsense factors are simply ignored.[/QUOTE]

Are lines which follow nonsense factors (and are therefore nonsense too) ignored too?

Syd 2009-03-24 18:02

[quote=Andi47;166525]Are lines which follow nonsense factors (and are therefore nonsense too) ignored too?[/quote]

No they are imported, but taken as another sequence starting with the first nonsense result, therefore it wont show up unless you request it.
There is also no need to submit the sequences in order, you can start for example with lines 2000 to x and submit the first 2000 lines later, even mix the whole sequence up, mix it with another one, leave out the index numbers, etc.
In most cases it recovers the sequences from it.

mklasson 2009-03-24 19:02

[QUOTE=Syd;166530]No they are imported, but taken as another sequence starting with the first nonsense result, therefore it wont show up unless you request it.
There is also no need to submit the sequences in order, you can start for example with lines 2000 to x and submit the first 2000 lines later, even mix the whole sequence up, mix it with another one, leave out the index numbers, etc.
In most cases it recovers the sequences from it.[/QUOTE]

This is very neat Syd! Now when are you going to add a "reserve sequence" feature? :razz:

schickel 2009-03-25 08:44

[QUOTE=akruppa;166500]I just asked Paul about the sequences he maintains. He mentioned that some time ago, he was looking for someone that could run a web-based interface to give access to these sequences with the option of submitting new factors so the sequence would be updated automatically. Does someone operate a web page like this? Maybe even with ECMNet server attached or some such? He'd transfer these 8 sequences then.

Alex[/QUOTE]I have some ideas I'm considering on building that exact type of system. I'll have to see how much spare time I can come up with this year. With the economy in the state it's in, this year looks to prove very, um, interesting at work.....

With the upsurge in interest here lately, I'd like to get something built in the not too distant future to take advantage of all the energy.

RichD 2009-07-13 02:03

Updating database
 
[QUOTE=schickel;166068]
[B]10,000-100,000[/B] maintained by [URL="http://www.lafn.org/~ax810/aliquot.htm"]Clifford Stern[/URL]. He maintains records of downdriver runs among other type of records.[/QUOTE]

How do we get this info into the database?

Do we have to recreate the work for each index?

I am willing to volunteer to help import the data if it is freely available.

RichD.

Greebley 2009-07-13 02:22

The database does have toubles with squares over 2000, so I have had troubles with terms like 3701^2 for example (if 3701 is prime that is).

If you find your sequence you just added has a small term of 7-12 digits that didn't factor, then you may have to manually find the square root. Because of the feature Syd mentioned, once you fix the square root issue, your whole sequence shows up.


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

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.