mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Aliquot Sequences (https://www.mersenneforum.org/forumdisplay.php?f=90)
-   -   Some Somewhat Easier n^i Sequences Available for Termination (https://www.mersenneforum.org/showthread.php?t=27659)

EdH 2022-03-16 21:33

Some Somewhat Easier n^i Sequences Available for Termination
 
In the sub-project [URL="https://www.mersenneforum.org/showthread.php?t=23612"]Aliquot sequences that start on the integer powers n^i[/URL], there are some sequences that should terminate with a prime. This thread will list those with a current term that is less than 145 digits* and flagged as unreserved. These sequences are mostly above those of the main project, although some may drop into the main project on their way to termination.**

If you are interested in the excitement of terminating an Aliquot Sequence, although not guaranteed, these are pretty sure bets to do so.

[B]Note:[/B] For anyone, new or old that would like to automate some of their work, please look at the script in [URL="https://www.mersenneforum.org/showpost.php?p=601938&postcount=7"]post 7[/URL] below. The script can be used with Aliqueit to convert the base^exponent value to its decimal and invoke Aliqueit to run the sequence and upload the results.

Please visit the thread mentioned above and its [URL="http://www.aliquotes.com/aliquotes_puissances_entieres/aliquotes_puissances_entieres.html"]associated page[/URL] for more details. You may reserve the available sequences in this thread and see the current status on the project pages, as updates are applied.

As an example of an available sequence, 66^92 is the smallest as of the latest full edit, and has a 136 digit term with a composite cofactor of 136 digits.

It is suggested that if you will take more than a day (or two) to terminate a sequence, you reserve it, so others don't duplicate your work. The following are the current reservations (but, also check the latest posts):[code]
[/code]As of the time of the last edit (fiddling), the following sequences were available:[code]
[/code]The second value is the cofactor size.

Here's a size sorted listing of the above:[code]
[/code]* The current threshold of 145 digits was chosen to ensure the listing has at least a fair number of sequences, with some more challenging.

** Sequences of the type n^i where both n and i are either odd or even (matched parity) nearly always terminate. Also, sequences where n is double a perfect square nearly always terminate. On occasion one will merge with a sequence in the main project and become open-ended.

The following are the terminated sequences that have not yet been updated in the tables. Many have unknown credit for termination (listed as A). If "The Terminator" would like credit, please claim it in this or the other thread:[code]
21^121: Prime - GDB
63^91: Prime - GDB
66^92: Prime - GDB
191^83: Prime - GDB
269^69: Prime - GDB
284^64: Prime - GDB
331^65: Prime - GDB
337^57: Prime - GDB
337^59: Prime - GDB
347^57: Prime - GDB
347^61: Prime - GDB
347^65: Prime - GDB
349^57: Prime - GDB
385^65: Prime - GDB
780^52: Prime - GDB
[/code]

RichD 2022-03-16 23:55

I went through some of my recent initializations and found a few that might be worthy to elevate into the first post.
[CODE]84^66: 128/104
84^68: 132/116
84^70: 136/119
86^66: 128/115
86^68: 131/128
86^70: 136/122
90^68: 134/129
91^65: 127/120
91^67: 131/107
91^69: 135/122
92^62: 122/94
92^64: 126/115
92^66: 130/113
92^68: 134/99
93^65: 128/101
93^67: 131/103
93^69: 136/108
95^63: 124/116
95^65: 129/100
95^67: 133/119
95^69: 136/124
96^66: 132/118
96^68: 136/128[/CODE]

EdH 2022-03-17 00:20

Thanks Rich,

I hadn't planned to make this a new source, but maybe that would work. I'll try to keep up with new available sequences, at least for now.

If we can get some more interest, the newcomers can also initialize some bases and work both terminations and open-ended, too.

VBCurtis 2022-03-17 02:35

I'll help with administration on this thread- updating post 1 with reservations, etc.

kruoli 2022-03-17 12:32

If appropriate, I would like to take these:
[CODE]3^333: 134/123
84^70: 136/119
86^70: 136/122
90^68: 134/129
91^69: 135/122
92^68: 134/99
93^69: 136/108
95^69: 136/124
96^68: 136/128[/CODE]

EdH 2022-03-17 14:07

Perhaps we should discuss which direction to take this thread, and how to minimize confusion with the main thread.

My initial vision was to have a few smaller sequences available to introduce newcomers to the project at a level they could work with a single machine.

As it now looks, we could create a large set of available sequences, much larger than my original thoughts. This could easily spiral into a mass of confusion for us. We need to keep this coordinated with Jean-Luc and not task him too heavily.

We need also to consider yoyo in this, since he'll be needing <140 work for his hungry project.

Let's step back momentarily to prioritize project goals. We'll need Jean-Luc to help with this. Advancing the tables is going to be more intensive due to how fast the terms now get large. How does table advancement, vs. same parity termination, vs. new table additions work toward the goals that provide the data for the questions that drive this project?

My proposals, for now:
- We hold only a very few to attract newcomers and see if we do. (we need to decide how few, etc.)
- We should go ahead and terminate the rest among ourselves as we would normally do.
- - kruoli has asked for some. I'm OK with that and they aren't reserved in the tables, but I would also like input from RichD, since he provided the bulk of them.*
- - VBCurtis also expressed interest in the ones I'm bringing below 140 digits. Let's go ahead and let our members reserve and work these as we have been.*
- I'm hesitant due to workload and confusion, but we may want to use the first post as a reference to smaller, same parity, available sequences. I would accept all help in that upkeep, but again, I'd like to minimize confusion with the main thread and Jean-Luc, so the table workings don't get too complicated. Keeping up with reservations could become duplication of effort and confusion if it isn't timely.

* We still need to use the other thread for reservations so the tables get updated and I'll move any reservation posts from this one over once we've discussed this a little more. We could be more timely showing reservation status here, but would it conflict with those on the main table pages?

All comments welcome. . .

kruoli 2022-03-17 14:51

If we want to attract new personnel, I would suggest we take their hand a bit (at least give the possibility) and give some guiding on how to execute this work. For example, I prepared a small script for this thread:
[CODE]export BC_LINE_LENGTH=0; # disable line breaks in bc
bc < list.txt > list.bc;
line_count_input=$(wc -l < list.bc);
base_dir=../terminations;

rm -f *.log siqs.dat nfs.*;
for i in $(seq 1 $line_count_input);
do
number=$(sed "${i}q;d" list.bc); # use this instead of read line (etc.) to prevent a misdeteciton of file redirection in YAFU, which would enter batch mode and cause problems
alq_file=alq_${number}.elf;
wget -O $alq_file "http://factordb.com/elf.php?seq=${number}&type=1";
line_count_elf=$(wc -l < $alq_file);
./aliqueit -y $number | tee execution.log; # use tee to see the progress while still logging to a file
./aliqueit -s $(($line_count_elf-1)) $number > upload.log; # maybe check if upload limit was reached here
dir=$base_dir/$(sed "${i}q;d" list.txt);
mkdir $dir;
mv -t $dir aliqueit.log execution.log upload.log $alq_file;
done;[/CODE]
Put your work in a file named [C]list.txt[/C], one entry per line in the form [C]x^y[/C]. It is assumed that you have an aliqueit executable in the same directory as the script (optimally with aliqueit.ini and yafu.ini if you have configured aliqueit to use YAFU as it would be recommended). Additionally, you would need [C]wc[/C], [C]wget[/C], [C]sed[/C] and [C]bc[/C] (these do not come with every Linux distribution by default). It will get the current ELF files and upload the results immediately after a sequence has terminated. The results will be stored as condfigured by [C]base_dir[/C]. (One could add another parameter to aliqueit to prevent getting in the rare case of a sequence not ending trivially.)

We could add links to threads (e.g. EdH's) on how to set up and compile YAFU(2) and aliqueit.

[SIZE="1"]As an aside, how do you pronounce aliqueit? Like ah-lee-kweet?[/SIZE]

EdH 2022-03-17 15:43

[QUOTE=kruoli;601938]If we want to attract new personnel, I would suggest we take their hand a bit (at least give the possibility) and give some guiding on how to execute this work. For example, I prepared a small script for this thread:
. . .
We could add links to threads (e.g. EdH's) on how to set up and compile YAFU(2) and aliqueit.

[SIZE=1]As an aside, how do you pronounce aliqueit? Like ah-lee-kweet?[/SIZE][/QUOTE]This could be a great idea, especially if we get anyone at that beginning of a level. I would hope the script wouldn't look too complicated to them. We can add this to the first post once we have a clear direction.

[SIZE=1]My pronunciation, which is actually rarely vocal, is more ah-leh-cue-it, but the ah still isn't quite right. I don't know the author's version.[/SIZE]

RichD 2022-03-17 16:36

Perhaps it is I that has the misunderstanding. For a newbie to look at the main status table is a bit overwhelming. For the rest of us that grew up with it, it is easy to understand.

I mostly do initialization work. Take sequences up to C100. If they start bigger, I take the (expected) terminating ones from C118-C120 to termination. I leave several in the C120-C140 range and above.

garambois 2022-03-17 18:44

Thank you very much Edwin for taking care of this and thank you very much to all the other people who are taking part in this new venture concerning n^i sequences with n and i of the same parity (matched parity) and with i large enough to require a good computing power.
I don't know at the moment if I have a role to play in this thread ?
I think the easiest way to avoid any confusion is indeed for you to let me know about your reservations on the main project thread, that's what I understand ?
Then I will update the project page according to these reservations.

RichD 2022-03-17 22:24

My (two cents) thoughts are, we would have a list of terminating candidates in post #1. People would speak up here to reserve a few. No need to flag them in post #1, simply remove them from the availability list. As more are spotted they can be added. We should never deplete the list, always leave a few.

Since the termination runs usually last less than a day (or so), no need to flag them here. As they complete they should be reported in the main thread for proper credit. This thread is for add and subtract.

More thoughts welcome.

EdH 2022-03-17 22:58

My initial expectations were for those who already were working sequences, but were not involved in this sub-project and would like some easier runs that would actually end. I think everyone likes to see a sequence they're working on terminate.

But, I would like to make sure we do cover someone who may be starting out.

I like the latest suggestions from RichD, as well, although I wonder about the reservation issue. Some of us can terminate these smaller sequences rather quickly, but some of our group may take a bit longer starting from the larger 13x terms and that may well include newcomers if we do attract them. As to the potential newcomers, I think we should just drive toward our goals and accept and assist if any do appear. we may find all the interested parties are already participating.

As to adding and removing from the first post, that sounds like a good idea, but should they be flagged elsewhere, in case they aren't finished quickly? And, yes, I would welcome all the help in keeping it updated. Maybe a compromise would be to once in a while add a post to the main thread with a list of the ones we've flagged over here. Not necessarily a reservation, but that those sequences are being worked. I just don't know how that would work out.

Although we haven't actually come up with the final operations, let's go ahead and add/remove sequences and start letting them be worked, by acknowledging Oliver's request and inviting more members to join. I will remove his requested sequences from the first post.

Sound good, so far?

RichD 2022-03-18 16:58

A few more smallish sequences.
[CODE]82^66: 127/121
82^68: 130/123
82^70: 134/127
82^72: 138/121[/CODE]

garambois 2022-03-18 18:52

[QUOTE=EdH;601965]
Sound good, so far?[/QUOTE]

OK, great !
Let's work like that, it should be fine !
My job will be to follow the events to make the updates with the right attributions.
Will you report all new completed sequences to me on the main project thread or will there be information here on this thread that will not be on the main Thread ?

EdH 2022-03-18 19:16

[QUOTE=garambois;602003]. . .
Will you report all new completed sequences to me on the main project thread or will there be information here on this thread that will not be on the main Thread ?[/QUOTE]This will have to be figured out. The best route would probably be having them reserved in the main thread. Then, when they are shown terminated via your update scripts, you can attribute the credit. The reservation post would also be available to anyone checking that thread for reservations. If someone reserves and immediate terminates sequences, it shouldn't really matter, since your scripts will pick that up and credit can still be awarded in the next update.

Does the above work for everyone?

RichD 2022-03-18 19:27

I'm thinking to keep it simple. What has been done on the first go around seems to work OK. Oliver grabbed a handful of sequences and posted the terminations in the main thread. If it effects the main table, put it in the main thread. These short reservations don't need posting -- reserving -- flagging -- completing -- removing -- updating.
KISS :smile:

More thoughts welcome.

RichD 2022-03-18 19:40

[QUOTE=RichD;602009]KISS :smile:[/QUOTE]
Oops, Jean-Luc may not be familiar with this American acronym.

KISS stands for - Keep It Simple Stupid

By no means am I referring to anyone here as stupid. That is why I added the smiley to the end.

garambois 2022-03-18 20:16

[QUOTE=RichD;602014]Oops, Jean-Luc may not be familiar with this American acronym.

KISS stands for - Keep It Simple Stupid

By no means am I referring to anyone here as stupid. That is why I added the smiley to the end.[/QUOTE]

It is certain that a French person will understand something else if you write "KISS" !
I didn't know this other meaning of this acronym !

;-)

EdH 2022-03-18 22:15

Sounds like we'll go with it as it stands currently, then.

We should probably add posts when adding sequences. That will keep the thread current and visible as new. Simply editing the first post won't do either.

If you'd just like to add a post, that's fine and I'll try to keep up. For those that can, feel free to edit the first post as well, but don't feel obligated.

birtwistlecaleb 2022-03-19 12:34

Taking 29^6

EdH 2022-03-19 13:17

[QUOTE=birtwistlecaleb;602078]Taking 29^6[/QUOTE]Welcome, . . .but, . . .29^6 merges with 2441868, which is within the main project. It should be reserved in the main project thread using that number.

richs 2022-03-19 18:26

Taking 92^62

richs 2022-03-20 16:15

92^62 terminated. Taking 92^64

EdH 2022-03-20 19:09

Looks like we have a few being terminated. Great!

I just added a few more for bases up through 50. Some are quite small. Maybe those are the ones to leave for new factorers, but not too long.

richs 2022-03-22 13:36

92^64 and 92^66 terminated. 92^68 already terminated by someone else, so taking 92^70 next.

EdH 2022-03-22 20:05

I just did a full sweep of all the unreserved matched parity sequences <base 100 and relisted in post 1, all the ones with the main term <140 digits. I tried to make sure none were reserved, but let me know if I did include any.

RichD 2022-03-24 09:43

The following can be added.
[CODE]101^69: 135/118
101^71: 138/137
101^73: 135/128[/CODE]

EdH 2022-03-24 13:07

Thanks! I hadn't started searching above base 100 yet and I try to stay away from any new initializations for a while. It looks like we're over 60 in our list. VBCurtis will have a bit of work if they're all left for him. . .

I haven't done so yet, but would a sorting by term size be advantageous. I have the creation of the list up to base 100 automated via scripts. I'm sure I could add in a sorting. Probably not today, but maybe I'll play at that later on.

firejuggler 2022-03-24 15:46

Taking 76^60 to it's end unless something very unhappy happen.

kar_bon 2022-03-24 15:59

95^63 terminates

firejuggler 2022-03-24 17:35

76^60 terminate in a P9

richs 2022-03-25 20:55

92^70 terminated. Taking:

5^225: 94/89
5^231: 138/129
5^239: 139/136

EdH 2022-03-28 19:10

I added a size sorted listing. Let me now if something is amiss.

EdH 2022-03-29 02:19

I'm going to list any new sequences I add to keep the thread fresh and allow for anyone checking new posts to catch them.

Newly added sequences:[code]12^146: 139/134
28^96: 139/108
48^86: 74/60 The db elves might finish this one while we're not looking.
76^74: 139/129
95^71: 139/118[/code]We're getting quite a collection. VBCurtis is going to be busy when he starts trying to run these down. If we collect too many, maybe yoyo will jump in again.

VBCurtis 2022-03-29 06:59

The c176 I'm running locally will finish sieving overnight tomorrow night, with LA expected to take ~6 days (I'm letting CADO do it for a full timing run at that size). So I'm a week away from starting on these. I'll be doing the biggest ones first, as I want more CADO timing data on 135-145 digit numbers.

EdH 2022-03-29 13:08

[QUOTE=VBCurtis;602783]The c176 I'm running locally will finish sieving overnight tomorrow night, with LA expected to take ~6 days (I'm letting CADO do it for a full timing run at that size). So I'm a week away from starting on these. I'll be doing the biggest ones first, as I want more CADO timing data on 135-145 digit numbers.[/QUOTE]That's good that you'll be working with the larger ones. I had wondewred if 140 wasn't a good break point. I chose it at first because that was yoyo's threshold. I'm still hoping some more newcomers will notice the smaller ones.

If you'd like a listing of matched parity for some in the 14x region (or other), let me know. I can easily generate that now. In fact, I'm using such a list right now to move some of the sequences from c141 down.

I'm also working some sequences that are likely to take huge drops for the next term. That's where 48^86: 74/60 came from. It dropped from a c141 term. I notice it's prime now. I expect the elves ran it. Maybe someone helped.

unconnected 2022-03-30 01:11

76^58 and 78^58 done
Taking
[CODE]33^97: 139/138
33^99: 138/131
37^99: 137/136
[/CODE]

EdH 2022-03-30 03:11

Added the following:[code]7^181: 139/129
39^91: 106/95
59^83: 93/93
59^91: 104/98
74^86: 117/108[/code]

Luminescence 2022-03-30 17:56

Grabbing these:

[CODE]6^192: 119/110
7^171: 139/130
[STRIKE]7^177: 91/88[/STRIKE]
7^181: 139/129
7^185: 125/124[/CODE]

Edit: 7^177 was already terminated

EdH 2022-03-31 03:21

Added the following:[code]10^148: 139/132
11^145: 139/121
12^134: 138/137
12^142: 136/117
28^98: 138/138[/code]

kar_bon 2022-03-31 14:57

78^62 and 78^64 terminated.

richs 2022-03-31 22:05

Terminated:

5^225: 94/89 done
5^231: 138/129 done
5^239: 139/136 done
7^177: 91/88 done (sorry, Ed, this one was so quick that I didn't have time to reserve it)

Taking:

59^83: 93/93 (already terminated by someone else)
59^91: 104/98
39^91: 106/95
76^58: 110/104
78^58: 111/94
39^93: 111/103

richs 2022-04-01 01:44

Terminated:

59^83: 93/93 (already terminated by someone else)
59^91: 104/98 done
39^91: 106/95 done
76^58: 110/104 (already terminated by someone else)
78^58: 111/94 (already terminated by someone else)
39^93: 111/103 done

Taking:

78^60: 114/103
74^86: 117/108
76^62: 117/116
76^64: 120/115
28^86: 123/111
17^123: 123/121
76^66: 123/122

EdH 2022-04-01 02:39

Great work, everyone! I'm afraid I might be missing some timely updates. Others who can, please feel free to edit the first post.

I'll look for some highlights tomorrow. For now, here are some additions:[code]51^83: 139/133
78^74: 139/124
90^72: 139/129
99^71: 139/138
101^77: 139/134[/code]

garambois 2022-04-01 17:01

Thank you all for these wonderful advances !
I'll update this weekend with all your announcements from the week.

kruoli 2022-04-01 17:36

Oh, I saw you inserted bases 76 and 78 here. I am currently working on the whole bases to extend my initial work. I might have not stated that in the original thread, sorry.

EdH 2022-04-01 18:03

[QUOTE=kruoli;603023]Oh, I saw you inserted bases 76 and 78 here. I am currently working on the whole bases to extend my initial work. I might have not stated that in the original thread, sorry.[/QUOTE]I didn't search the other thread for reservations. I only looked at the tables. Sorry about that!

EdH 2022-04-01 18:08

@all: I missed some reservations by kruoli for bases 76 and 78. Please release any of these you have reserved here, back to Oliver. Sorry for the double work and any confusion.

unconnected 2022-04-01 21:34

33^99 done

EdH 2022-04-02 01:53

Added the following:[code]6^204: 98/89
17^127: 92/66
19^127: 139/128
44^86: 139/111
45^87: 139/135
54^92: 112/101
82^74: 139/126
91^73: 139/139
99^77: 130/130[/code]

richs 2022-04-02 04:36

Terminated:

78^60: 114/103 done
74^86: 117/108 done
76^62: 117/116 (already done)
76^64: 120/115 (already done)
28^86: 123/111 done
17^123: 123/121 done
76^66: 123/122 (already done)

EdH 2022-04-02 13:00

[QUOTE=richs;603084]Terminated:

78^60: 114/103 done
74^86: 117/108 done
76^62: 117/116 (already done)
76^64: 120/115 (already done)
28^86: 123/111 done
17^123: 123/121 done
76^66: 123/122 (already done)[/QUOTE]If you're still working any 76 or 78 base numbers, please stop. I missed a reservation in the other thread and the table doesn't show them. That's why you were finding some already done.

Thanks everyone for all the work and please holler at me if I misstep.

richs 2022-04-02 13:29

I stopped on the 76 and 78 bases when I saw your post, Ed.

Taking:

17^127: 92/66
6^204: 98/89
54^92: 112/101
7^185: 125/124
91^65: 127/120

Luminescence 2022-04-02 21:01

[QUOTE=Luminescence;602887]Grabbing these:

[CODE]6^192: 119/110
7^171: 139/130
[STRIKE]7^177: 91/88[/STRIKE]
7^181: 139/129
7^185: 125/124[/CODE]

Edit: 7^177 was already terminated[/QUOTE]

All done.

EdH 2022-04-02 21:32

I'm seeing some duplication of effort (7^185 is one). Maybe I'm not keeping up with the first post enough. I've been trying to remove any sequence as soon as I see a reservation, but maybe I'm not on top of it enough.

Again, this has affected richs!

I apologize for this. (If I do poorly enough, perhaps I'll be fired. . . )

Let me know of any way we can minimize the duplication.

Maybe I need to quit trying to keep up with two versions. I see I didn't remove 7^185 from the sorted list (or I put it back at some time). I will try to come up with a better method of keeping up. Maybe both lists should be checked prior to work.

ETA: I have written a script to use the first set in the first post to build the second set. Hopefully that will help me stay a little closer to "on top of it." Still, all suggestions are welcome.

richs 2022-04-02 23:09

Ed, no worries. I check the status before I start aliqueit. Fyi, 6^204 and 17^127 are already terminated.

Terminated:

17^127: 92/66 (already done by another)
6^204: 98/89 done
54^92: 112/101 done
7^185: 125/124 (already done by another)
91^65: 127/120 done

Taking:

10^148: 139/132
11^145: 139/121
12^134: 138/137
12^142: 136/117
12^146: 139/134

EdH 2022-04-03 02:24

Thanks Rich!

Added a few more, but they are slowing down a bit:[code]14^126: 139/133
19^115: 110/97
77^75: 137/117
88^72: 139/127[/code]

garambois 2022-04-03 08:51

[QUOTE=richs;603084]Terminated:
17^123: 123/121 done
[/QUOTE]

There must be an error : 17^123 is not terminated on FactorDB.

garambois 2022-04-03 11:12

[QUOTE=richs;603084]Terminated:
76^66: 123/122 (already done)[/QUOTE]

There must be another error : 76^66 is not terminated on FactorDB.

garambois 2022-04-03 11:26

Please, if someone wants to reserve a base, can they request it on the main thread ?
For example, Oliver, should I reserve all the sequences of bases 76 and 78 for you ?

birtwistlecaleb 2022-04-03 11:47

19^115 to it's end
Complete! Length: 53
[URL="http://www.factordb.com/sequences.php?se=1&aq=19%5E115&action=all&fr=0&to=100"]link to sequence[/URL]
Now taking 93^65 and 95^65
95^65 has c107, moving on
93^65 has c99 that's will also take a long time

kruoli 2022-04-03 12:57

[QUOTE=garambois;603182]Please, if someone wants to reserve a base, can they request it on the main thread ?
For example, Oliver, should I reserve all the sequences of bases 76 and 78 for you ?[/QUOTE]

Currently, I am extending bases 76, 78, 94 and 127 (the ones I initialised). I am going to try to terminate all possible sequences up to 135 or maybe 140 digits. There are already progresses you may include in your update if you like. I will state in the main thread when I have finished my work.

VBCurtis 2022-04-03 16:03

[QUOTE=birtwistlecaleb;603183]
95^65 has c107, moving on
93^65 has c99 that's will also take a long time[/QUOTE]

A C99 is maybe a 10-minute job on a modern desktop CPU, less than an hour on many laptops. C107 maybe 3 times as hard. If you think these take a long time, I think your factoring software is not set up correctly, or you're not using the right tools.
If you're in windows, yafu can do these quite easily.
If you run linux, get CADO set up.

garambois 2022-04-03 17:00

[QUOTE=kruoli;603188]Currently, I am extending bases 76, 78, 94 and 127 (the ones I initialised). I am going to try to terminate all possible sequences up to 135 or maybe 140 digits. There are already progresses you may include in your update if you like. I will state in the main thread when I have finished my work.[/QUOTE]

OK, I'll reserve these 4 bases for you (76, 78, 94 and 127).

birtwistlecaleb 2022-04-03 17:17

1 Attachment(s)
[QUOTE=VBCurtis;603200]A C99 is maybe a 10-minute job on a modern desktop CPU, less than an hour on many laptops. C107 maybe 3 times as hard. If you think these take a long time, [B]I think your factoring software is not set up correctly, or you're not using the right tools.[/B]
If you're in windows, yafu can do these quite easily.
If you run linux, get CADO set up.[/QUOTE]
Correct: I officially have no idea how to find yafu's post-nfs app.
Reference image attached.
Found it!
(also returning to 95^65)

birtwistlecaleb 2022-04-03 18:11

95^65 complete! (25 steps)
Interesting: ends in a 77 digit prime!
Now to continue 93^65

birtwistlecaleb 2022-04-03 18:29

1 Attachment(s)
second problem:

VBCurtis 2022-04-03 20:58

The program output tells you it cannot find the sievers. While I agree that is a problem, you didn't tell us whether the siever exe's are there in your folder (meaning a path error of some sort), or that you don't have / don't know where to get the lasieve executables.

I haven't done any factoring on windows in many years, so I can't direct you to the siever executables- sorry. Someone surely will soon in this thread!

kruoli 2022-04-03 21:01

[URL="http://download.mersenne.ca/GGNFS"]Here[/URL] are some. Thanks, James! :smile:

birtwistlecaleb 2022-04-03 21:13

I do have that, i found a sourceforge link in msieve. The problem is on yafu's end. It uses a default link for gnfs-lasieve4I12e.exe. does anyone know how to fix that?

kar_bon 2022-04-04 11:31

Read the docfile.txt in the yafu.zip and the parameter "ggnfs_dir".

richs 2022-04-04 15:26

[QUOTE=garambois;603176]There must be an error : 17^123 is not terminated on FactorDB.[/QUOTE]

I'll get back to 17^123. I am unsure how I erred.

76^66 is kruoli's reservation now.

EdH 2022-04-04 15:28

I have expanded my scripts to cover all the existing tables below a base of 2^32. The scripts are supposed to skip any sequences flagged as reserved in the tables. I will try to also keep track of any more recent reservations that haven't made it into the tables yet. If I fail on either of these, please post here, noting my failure.

Since I'm now generating the lists via script, rather than manually adding new sequences I've brought below 140 digits, I will "probably not" post additions as I have in the past.

birtwistlecaleb 2022-04-04 17:18

reserving 54^96 until it goes up or terminates
Terminates after 60 terms (prime)

birtwistlecaleb 2022-04-04 17:28

reserving 101^53 until it goes up or terminates
93^65 is also unreserved

birtwistlecaleb 2022-04-04 18:34

101^53 Terminates after 74 terms (prime)

kar_bon 2022-04-07 10:04

93^65 terminates in P21
52^94 terminates in P17
439^55 terminates in P7

firejuggler 2022-04-07 15:55

439^55 terminate as a P7
52^94 terminate as a P17


Not by me, BTW
taking 211^71

richs 2022-04-08 05:12

10^148 finally terminated after losing the 3.

EdH 2022-04-08 12:03

[QUOTE=richs;603508]10^148 finally terminated after losing the 3.[/QUOTE]Yes, I have found some to be pretty slow in their decline. Then, on the other side, there are some, with no small factors, that fall a few digits all at once. I'm running a 168 digit right now that should be pretty deficient, but even across 40 threads, LA alone is taking a while.

firejuggler 2022-04-08 14:32

211^71 terminate as a P5


Big fall from 118 digits to 86 at i12


edit taking 720^40

firejuggler 2022-04-08 15:34

and it end in a P23. I'll take another low hanging fruit in a few hours

firejuggler 2022-04-08 23:47

taking 660^50




edit : and it end in a prp88!


I'm sorry for the spamming.
next time i'll take 3 or 4 in one post

unconnected 2022-04-09 19:32

Taking[CODE]99^65: 130/127
99^67: 134/122
99^69: 138/130
99^77: 130/114[/CODE]

firejuggler 2022-04-09 21:40

taking 91^73

unconnected 2022-04-09 21:59

99^77 quickly ended with P104.

firejuggler 2022-04-10 00:43

and as usual, factmsieve crash on the last step
[code]
-> Running square root step ...
-> msieve150_gpu -s ..\ggnfs_86953847287548242793486845908133249296614592756211423212379954848484723528350642653180365724764413\test.dat -l ..\ggnfs_86953847287548242793486845908133249296614592756211423212379954848484723528350642653180365724764413\test.log -i ..\ggnfs_86953847287548242793486845908133249296614592756211423212379954848484723528350642653180365724764413\test.ini -nf ..\ggnfs_86953847287548242793486845908133249296614592756211423212379954848484723528350642653180365724764413\test.fb -t 3 -nc3
-> Computing 1.64955e+09 scale for this machine...
-> procrels -speedtest> PIPE
Traceback (most recent call last):
File "D:\Download\aliqueit\factmsieve.py", line 2104, in <module>
output_summary(NAME, fact_p, pols_p, poly_p, lats_p)
File "D:\Download\aliqueit\factmsieve.py", line 1789, in output_summary
timescale = float(re.sub('timeunit:\s*', '', tmp[0]))
IndexError: list index out of range
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.
[/code]

EdH 2022-04-10 02:15

It's been a long time since I ran factmsieve.py and much longer since I ran Windows. Can you try just the command for the Square Root phase in a terminal and get the factors?

Luminescence 2022-04-10 05:20

[QUOTE=EdH;603638]It's been a long time since I ran factmsieve.py and much longer since I ran Windows. Can you try just the command for the Square Root phase in a terminal and get the factors?[/QUOTE]

It's not just on Windows. I got a new machine and installed Ubuntu, but I am getting the same error. For what Python version was this script made? It throws me that error with Python 2.7

EdH 2022-04-10 13:29

What do the Msieve .log files say about the Square Root phase? I think the error shown is because the time couldn't be calculated, but is that because no factors have shown up or something else?

Brian Gladman would be the one to contact, but I don't know if he still looks at anything here.

firejuggler 2022-04-10 13:42

I do have the factors. it is just that factmsieve/ecm.py/aliqueit don't retrieve them.

EdH 2022-04-10 15:00

[QUOTE=firejuggler;603668]I do have the factors. it is just that factmsieve/ecm.py/aliqueit don't retrieve them.[/QUOTE]I will try later today to run some sessions with factmsieve.py to see if I can duplicate and track down what's going on. (It will give me a chance to see if my "How I. . ." is still valid.)

richs 2022-04-10 16:35

Update on terminations:

17^123
11^145
12^142

EdH 2022-04-10 18:05

[QUOTE=EdH;603672]I will try later today to run some sessions with factmsieve.py to see if I can duplicate and track down what's going on. (It will give me a chance to see if my "How I. . ." is still valid.)[/QUOTE]Odd! I'm not able to get factmsieve.py to work at all. I'm going to have to really look into this, by first starting all over from scratch. This will be a bit later.

firejuggler 2022-04-10 19:05

No worries. It take 2 min to get the factor out, add it to my elf file and restart aliqueit.

EdH 2022-04-10 20:16

The best I've been able to come up with for my troubles with factmsieve.py, is that Msieve is failing to use test, which Aliqueit is trying to assign, as its work name. It's insisting on a worktodo.ini file that doesn't exist. I'm probably not going to pursue this much more, at least not for now.

firejuggler 2022-04-13 17:44

91^73 end in a P6
Lost the 3 at i:48 , re-acquired at i:53
went from 131 digits to 89. At this point, it was quick and easy.


edit : taking 55^85


All times are UTC. The time now is 15:59.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.