mersenneforum.org  

Go Back   mersenneforum.org > Prime Search Projects > Conjectures 'R Us

Reply
 
Thread Tools
Old 2010-01-14, 16:10   #12
Mini-Geek
Account Deleted
 
Mini-Geek's Avatar
 
"Tim Sorbera"
Aug 2006
San Antonio, TX USA

4,271 Posts
Default

.
Attached Files
File Type: zip unworked-lists.zip (6.3 KB, 301 views)
Mini-Geek is offline   Reply With Quote
Old 2010-01-15, 05:26   #13
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

245458 Posts
Default

Outstanding Tim! If you can run that about once every 3 days or so, I'll continue updating the first 2 posts of each thread, which I'll do now shortly.

Right now, we're in a spate of new bases being worked. I know that Ian will be doing all the reasonably-sized Riesel bases where b==(4 mod 5) due to algebraic factors over the next few weeks. Perhaps after the big push is through, we could go to running it once every 1-2 weeks.

Edit: All 4 posts now updated. I was bumping up against the char. limit in post 1 so I reduced the # of spaces in between the base and conjecture and made the dividing line between the lists at a conjecture of 10K (vs. 100K), which evened things out a little more.

Last fiddled with by gd_barnes on 2010-01-15 at 06:08
gd_barnes is offline   Reply With Quote
Old 2010-01-15, 13:47   #14
rogue
 
rogue's Avatar
 
"Mark"
Apr 2003
Between here and the

13×503 Posts
Default

Quote:
Originally Posted by gd_barnes View Post
Edit: All 4 posts now updated. I was bumping up against the char. limit in post 1 so I reduced the # of spaces in between the base and conjecture and made the dividing line between the lists at a conjecture of 10K (vs. 100K), which evened things out a little more.
Which is fine. I used 100K because a longer list could not be posted in one message. I was forced to post it as two lists. In a couple of years, it might be down to a single list.

Gary, do you have any guidelines regarding how many n should be tested for each k? For example, does it make sense for someone to tackle one of the harder bases, but only test to 10K (or lower) just to get it off of the list? Does it make sense to reserve some bases for "group efforts" only unless they can demonstrate that they have the horsepower to tackle a more difficult base?

Last fiddled with by rogue on 2010-01-15 at 13:47
rogue is online now   Reply With Quote
Old 2010-01-22, 09:10   #15
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

10,597 Posts
Default

Tim,

Can you provide a quick updated list here for both sides? I'll then update the 1st post in this and the Riesel thread. Ian has taken a lot of new bases in the last few days. 1-2 others have taken a couple of them also.


Thanks,
Gary
gd_barnes is offline   Reply With Quote
Old 2010-01-22, 09:27   #16
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

101001011001012 Posts
Default

Quote:
Originally Posted by rogue View Post
Which is fine. I used 100K because a longer list could not be posted in one message. I was forced to post it as two lists. In a couple of years, it might be down to a single list.

Gary, do you have any guidelines regarding how many n should be tested for each k? For example, does it make sense for someone to tackle one of the harder bases, but only test to 10K (or lower) just to get it off of the list? Does it make sense to reserve some bases for "group efforts" only unless they can demonstrate that they have the horsepower to tackle a more difficult base?
For "reasonable sized" bases, I like to see them tested to n=25K. No, no guideline on what is reasonable. For larger conjectured bases (perhaps such as your Riesel base 58), getting a good run of all k's up to n=10K is a sufficient start.

I'd personally prefer that people avoid most bases with > ~1500-2000 k's remaining at n=10K. (Much better is bases with < 200-250 k's remaining at n=10K, especially if you have less than ~2 quads for testing.) Perhaps that is the best guideline that I can come up with. I think Riesel base 39 has something like ~4500 k's remaining at n=15K. Ian put a huge amount of personal and CPU effort into that as well as base 40, I think. I think he got it there within 2-3 months or so. But he has quite a bit of resources (and personal stamina) to be able to stomach so many 1000's of small primes. I would not recommend that to most people.

I can't stop people from reserving bases well out of proportion with their resources [although will properly chastise them if I think they are :-)] and don't really have a guideline for that other than they should be able to get it at least to n=10K within 3-6 months. There are a couple of new bases out there that have been reserved for a long time (>3 months) with what I believe to be very little work done.

For these lists, the bases would be removed if someone simply reserves a new base. It doesn't take actually searching them. But if someone goes AWOL and doesn't respond to requests for a status, then I would unreserve them and we'd put them back on the list here.

As you all well know, I'm pretty anal on how statuses are reported. But when it comes to coordinating a group effort on a single base or several bases, I'm quite happy to let someone else take the lead. I promise I won't be "too" anal about how it is done. :-) I just don't have the time to coordinate large-scale sieving/testing efforts on multiple bases or 100's of thousands or millions of k's. That's why I unstickied the base 3 threads several months ago. I'm happy if people do a little here-and-there on them but it takes too much time when doing group efforts of millions of k's.

If someone does take the lead in such a group effort, I'll just make sure the web pages are kept properly updated and do the balancing of k's remaining as needed. I just want to make sure the math is right in all aspects whenever possible so that nothing gets missed and extra CPU/personal time is avoided.

And finally and you've all heard this numerous times before: It's still my preference that we extend the testing of already tested bases and/or fill in the "base holes" for bases < ~150-200. I took a few new smaller-conjectured bases <= 150 recently to knock out some of them but generally stick with checking or double checking others work or pushing "under" tested existing bases a little higher. Within a few weeks, I'll probably reserve a few already-tested bases <= 150 that are only at n=10K or 15K to push them up to n=25K or higher. Once all "reasonable" bases <= 150 are at n=25K, I might find a few "unreasonable" bases <= 150 to take to n=25K (> ~100 k's remaining at n=25K) or will do some work on the lower bases to n=50K or 100K. As an interesting side note on that: I recently reserved new Riesel base 91 with a conjecture of k>220K. Although it has 235 k's remaining at n=2500, due to its high weight, I estimate that it will have 90-100 k's remaining at n=25K. So it is what I would call a "borderline case" as to whether it is "reasonable" or not. Also, I kind of have a personal interest in the high-weight k==(1 mod 30) bases, hence my year-long effort to get Sierp base 31 up to n=25K on ONE core. lol (It's currently at n=24K and will have 1100-1105 k's remaining when done. That's actually very low for a conjecture of k>6.3M.)


Gary

Last fiddled with by gd_barnes on 2010-01-22 at 09:48
gd_barnes is offline   Reply With Quote
Old 2010-01-22, 12:10   #17
Mini-Geek
Account Deleted
 
Mini-Geek's Avatar
 
"Tim Sorbera"
Aug 2006
San Antonio, TX USA

4,271 Posts
Default

Quote:
Originally Posted by gd_barnes View Post
Tim,

Can you provide a quick updated list here for both sides? I'll then update the 1st post in this and the Riesel thread. Ian has taken a lot of new bases in the last few days. 1-2 others have taken a couple of them also.


Thanks,
Gary
Updated lists are attached.
By the way, is it at all necessary to pull from the Riesel/Sierp reservations/statuses lists, or does just pulling from the 'conjectures and proofs' list suffice? I've been doing the latter, but recently realized that perhaps some new bases that are reserved but with nothing reported yet might only be on the former list.
Attached Files
File Type: zip lists.zip (6.2 KB, 303 views)

Last fiddled with by Mini-Geek on 2010-01-22 at 12:12
Mini-Geek is offline   Reply With Quote
Old 2010-02-04, 11:10   #18
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

10,597 Posts
Default

Quote:
Originally Posted by Mini-Geek View Post
Updated lists are attached.
By the way, is it at all necessary to pull from the Riesel/Sierp reservations/statuses lists, or does just pulling from the 'conjectures and proofs' list suffice? I've been doing the latter, but recently realized that perhaps some new bases that are reserved but with nothing reported yet might only be on the former list.
I've now updated the lists in the first 2 posts here and on the Riesel side as of this moment. With so many new bases done since your posting, I went ahead and manually keyed in the bases done from the pages. I then used VLOOKUP tables that had all bases and conjectures in Excel to remove the bases already started.

Bases that have been reserved should be removed from the lists, regardless of whether they have had actual work done on them or not. If the person decides to abandom them, then we can add them back to the lists.

For clarification, newer reservations on bases shown as "testing just started" on the pages would not be shown here but bases that show "no testing done" should be shown here because they have not been reserved. I'm just choosing to show the latter for one reason or another...generally because they are "kind of" a low base -or- they were started at one time and later unreserved with no work done -or- they are just a base of interest.


Gary
gd_barnes is offline   Reply With Quote
Old 2010-05-10, 14:26   #19
MyDogBuster
 
MyDogBuster's Avatar
 
May 2008
Wilmington, DE

54448 Posts
Default

Audit complete on posts 1 and 2. No problems
MyDogBuster is offline   Reply With Quote
Old 2010-11-10, 14:04   #20
rogue
 
rogue's Avatar
 
"Mark"
Apr 2003
Between here and the

13·503 Posts
Default

I took each of the conjectures where the conjectured k>10000 and ran a range of 1000 up to n=500 with the script to get an idea as to the difficulty to proving them. The following list is an extrapolation of that data, showing the number of expected remaining k per conjecture at n=500. This list only shows those for which the expected remaining k < 10000.

Code:
Base   Conjectured k   Expected Remaining k
 490        15123             136
 697        14308             172
 403        11412             194
 150        49074             196
 627        12354             222
 382        11491             230
 445        14986             285
 640        11925             286
 523        10872             294
 663        10042             321
 880        25282             329
 355        23586             330
 282        10807             346
 658        20428             347
 598        18568             371
 642        10932             372
 810        30008             420
 957        19638             452
 465        78056             468
1012        16207             470
 943        15636             485
 733        14314             501
 235        15706             518
 277        19578             529
 772        23191             557
 348        26523             557
 808        24271             680
 738        12767             702
 592        23721             735
 270        62060             745
 625        17428             767
 215        19924             777
 555        32388             777
 858        35218             845
 522        32644             849
 667        26218             891
 927        28624             945
 247        71392            1071
 513        45828            1100
 460        37803            1134
1005        54610            1147
 430        22413            1278
 568        23328            1330
 660        74031            1333
 292        40393            1333
 591        16242            1364
 918        24812            1365
 843        28486            1367
 432        46765            1403
 997        36048            1406
 971        14876            1473
 708        28361            1475
 982        39640            1506
 381        18526            1519
 742        30462            1554
 871        21676            1561
 735       174778            1573
 621        19592            1607
 243        40078            1643
 852        34974            1644
 223        57814            1792
 442        36768            1949
 431        20138            2155
 262       110724            2214
 820        30378            2218
 283       106714            2241
 388        90249            2346
 807        53428            2404
 841        22312            2410
1018        77443            2556
 291        33232            2559
 763       151462            2575
 576        30651            2667
 448       139191            2923
 768        55367            2934
 931        37978            2962
 757        47376            2985
 751        41032            3159
 310       268392            3221
 457        84958            3228
 393        58608            3282
 847       150678            3315
 732        81364            3336
 297       133654            3341
 856        39457            3591
 205       138330            3735
 411        46246            3746
 616        53061            4086
 711        49572            4214
 646        52701            4216
 717       179678            4312
 303       174742            4369
 675       293812            4407
 975       375364            4504
 606        50380            4786
 313       111312            5454
 421        53806            5703
 225       117406            5753
 477        78152            5783
 537       176734            5832
 357       456628            5936
 366        79231            5942
 745       334816            6027
 336        92000            6348
 612       162446            6985
 451        97068            6989
 822       278173            9458
 595       301128            9636
rogue is online now   Reply With Quote
Old 2011-01-26, 18:29   #21
rogue
 
rogue's Avatar
 
"Mark"
Apr 2003
Between here and the

198B16 Posts
Default

After all conjectures for conjectured k < 10000 have been started, I suggest moving the conjectures for conjectured k < 1000000 into the first post.
rogue is online now   Reply With Quote
Old 2011-01-27, 02:27   #22
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

296516 Posts
Default

Quote:
Originally Posted by rogue View Post
After all conjectures for conjectured k < 10000 have been started, I suggest moving the conjectures for conjectured k < 1000000 into the first post.
Better yet: I've combined them all into one post now.
gd_barnes is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Untested Riesel conjectures sorted by conjecture rogue Conjectures 'R Us 36 2018-01-03 19:53
I took a New assignment untested Prime Number manual testing &TF ONeil Information & Answers 4 2017-12-23 05:30
Conjectures with one k remaining rogue Conjectures 'R Us 109 2017-04-29 01:28
decomp sorted <61 moo Lone Mersenne Hunters 0 2005-03-24 22:38
Smallest untested number? wirthi Math 10 2003-10-05 13:02

All times are UTC. The time now is 21:04.


Thu Jan 27 21:04:23 UTC 2022 up 188 days, 15:33, 2 users, load averages: 3.06, 4.70, 3.30

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

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.

≠ ± ∓ ÷ × · − √ ‰ ⊗ ⊕ ⊖ ⊘ ⊙ ≤ ≥ ≦ ≧ ≨ ≩ ≺ ≻ ≼ ≽ ⊏ ⊐ ⊑ ⊒ ² ³ °
∠ ∟ ° ≅ ~ ‖ ⟂ ⫛
≡ ≜ ≈ ∝ ∞ ≪ ≫ ⌊⌋ ⌈⌉ ∘ ∏ ∐ ∑ ∧ ∨ ∩ ∪ ⨀ ⊕ ⊗ 𝖕 𝖖 𝖗 ⊲ ⊳
∅ ∖ ∁ ↦ ↣ ∩ ∪ ⊆ ⊂ ⊄ ⊊ ⊇ ⊃ ⊅ ⊋ ⊖ ∈ ∉ ∋ ∌ ℕ ℤ ℚ ℝ ℂ ℵ ℶ ℷ ℸ 𝓟
¬ ∨ ∧ ⊕ → ← ⇒ ⇐ ⇔ ∀ ∃ ∄ ∴ ∵ ⊤ ⊥ ⊢ ⊨ ⫤ ⊣ … ⋯ ⋮ ⋰ ⋱
∫ ∬ ∭ ∮ ∯ ∰ ∇ ∆ δ ∂ ℱ ℒ ℓ
𝛢𝛼 𝛣𝛽 𝛤𝛾 𝛥𝛿 𝛦𝜀𝜖 𝛧𝜁 𝛨𝜂 𝛩𝜃𝜗 𝛪𝜄 𝛫𝜅 𝛬𝜆 𝛭𝜇 𝛮𝜈 𝛯𝜉 𝛰𝜊 𝛱𝜋 𝛲𝜌 𝛴𝜎 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔