mersenneforum.org Getting others to do the work on exponents I like (was: Trial Factoring Progress)
 Register FAQ Search Today's Posts Mark Forums Read

2022-04-24, 06:05   #540
tuckerkao

"Tucker Kao"
Jan 2020

3×13×19 Posts

Quote:
 Originally Posted by petrw1 In my dream I went to: Account / My Account / Results to see if I've found any factors, expecting a couple; but I see over a dozen and it's so exciting I physically woke up. PS sadly it was only a dream. I did not see a dozen when I wake up.
I better to quote this one before I forget where it locates, if I get a similar dream, I'll write down on my diary and share the details later.

Since the dozenal compound was visible in this dream, it should worth a tribute, best if possible to remember which exponents those 1 dozen factors were linked to. Just 1 digit of number interpreted wrong could cause the user to miss the gem.

However, I still prefer to have a dream about the PRP results of the M168,***,*23 exponents especially those still untested.

 2022-04-24, 07:49 #541 tuckerkao   "Tucker Kao" Jan 2020 Head Base M168202123 3·13·19 Posts Although mersenne.ca still show that M168.9M has 2,001 unfactored exponents, the fresh arrivals of M168961259 and M168961411 marked another subrange I've conquered to under 2,000 - https://www.mersenne.ca/status/tf/0/0/4/16800 Only 20,215 unfactored exponents within the M168M range counting the 2 factors above, just need around half more year to bring the entire range down below 20k. It seems like M168.2M, M168.3M, M168.4M, M168.6M are the 4 more factor-resistant ranges, I should focus on digging those areas instead of other 6. Last fiddled with by tuckerkao on 2022-04-24 at 07:56
2022-04-24, 23:29   #542
tuckerkao

"Tucker Kao"
Jan 2020

3×13×19 Posts

Quote:
 Originally Posted by Zhangrc 30.8 uses a larger FFT in stage 2. That's George's decision. He is so talented, don't put extra pressure on him and he'll give us a surprise! Maybe it will be 30.9b1, because George said that he is planning to implement the new stage 2 algorithm in ECM.
I hope that surprise arrives on July 4th, 2022, wish George can go on a vacation the week before that. If I can complete 1 cycle of ECM of M168M exponent with the same B1 and B2 bounds as P-1 within 3 days, I may try it.

P+1 stage 2 will probably be the great racer to improve the speed too.

3 more new factors found before midnight server time today: M168963161, M168963461, M168963653.

Last fiddled with by tuckerkao on 2022-04-24 at 23:47

2022-05-02, 08:23   #543
tuckerkao

"Tucker Kao"
Jan 2020

3·13·19 Posts

Quote:
 Originally Posted by Zhangrc By the way, why don't you set your account to get smallest exponents and put your computer to work at current wavefront?
I'm only 97 exponents shy of bringing the entire M168.4M subrange to at least 276 - https://www.mersenne.ca/status/tf/0/0/4/16800

It'll be the best that I'm only working on the exponents in a range where other users have already aware of it. I've sometimes typed the exponents with 1 digit wrong and ended up working on a different exponent. It'll be okay if that happens within M168M, but not within the busy front-wave of Cat 0 and Cat 1 assignments.

Last fiddled with by tuckerkao on 2022-05-02 at 08:51

2022-05-02, 23:30   #544
tuckerkao

"Tucker Kao"
Jan 2020

13458 Posts

Quote:
 Originally Posted by Viliam Furik I think that was to factor that specific exponent.
I'm glad I've finally found the specific M168,***,*23 exponent that exactly reflected my exploration interest - M168113423, a factor was found between 278 to 279 and located outside the bounds of the regular recommended P-1.

M168732533 which had Kriesel's TFs from 273 to 278 was factored by my P-1.

Attached screenshot below: My trial factoring efforts are already noticeable with the zooming-in of the Current Primenet Progress on the world map of mersenne.ca planet.
Attached Thumbnails

2022-05-03, 00:01   #545
mathwiz

Mar 2019

277 Posts

Quote:
 Originally Posted by tuckerkao It'll be the best that I'm only working on the exponents in a range where other users have already aware of it.
No, it would be better for the GIMPS project if you worked at the current wavefront. You have repeatedly failed to provide mathematically valid evidence that your cherished 168M range is more likely to yield a Mersenne prime.

 2022-05-03, 15:45 #546 Stargate38     "Daniel Jackson" May 2011 14285714285714285714 22·3·59 Posts I don't think he needs to provide valid evidence for a prime being there. I think he might just be searching a range of his choice that's not heavily used. If I were him, I would do the same to keep from accidentally poaching someone else's exponent (i.e. because of a typo). However, I rarely do Mersenne numbers, and I'm mostly interested in factoring smaller numbers of general or special form (< 140 digits for GNFS, < 190 digits for SNFS, < 500 digits for ECM to t35, < 300 digits for ECM to t40) to help clear the composites on FactorDB. Last fiddled with by Stargate38 on 2022-05-03 at 15:47
2022-05-04, 13:45   #547
Dr Sardonicus

Feb 2017
Nowhere

22·31·47 Posts

Quote:
 Originally Posted by tuckerkao I'm only 97 exponents shy of bringing the entire M168.4M subrange to at least 276 I've sometimes typed the exponents with 1 digit wrong and ended up working on a different exponent. It'll be okay if that happens within M168M, but not within the busy front-wave of Cat 0 and Cat 1 assignments.
This is the most interesting thing you've said in quite some time.

It occurred to me to wonder about how mistyping a single digit might occur, and how likely such a mistyping would be to yield another prime.

One obvious kind of mistyping is "fat-fingering." This is typing a key adjacent to the one intended.

Assuming the digits keys used are the ones laid out in the top row as 1234567890, I used Pari-GP to work up some numbers. Note that 2 through 9 can be fat-fingered into either of two adjacent digits, while 0 and 1 can only be fat-fingered to another digit in one way.

There are 5251 primes p, 168400000 < p < 168500000. There are 83828 possible fat-fingerings of these primes. Assuming the first 4 digits 1684 are typed correctly, there are 47071 fat-fingerings of the remaining five digits.

There are 3904 fat-fingerings of a single digit of a prime p, 168400000 < p < 168500000 which result in another prime in the interval. Thus, assuming the first 4 digits are typed correctly, about 8.3% of fat-fingerings result in other primes.

There are 2932 primes in the interval which can be changed into other primes in the interval by fat-fingering a single digit. Fat-fingering any digit of any of the 2319 other primes results in a composite number.

Some of the primes can be changed by fat-fingering a single digit into more than one other prime in the interval. The largest number of such fat-fingerings is 4. There are eight primes in the interval which can be changed into four other primes in the interval. I give the pairs [p,q] where q is obtained by fat-fingering a single digit of p, for the smallest and largest p which can be changed into 4 other primes in the interval.
Code:
[168400697, 168409697]
[168400697, 168490697]
[168400697, 168400597]
[168400697, 168400607]

[168496463, 168406463]
[168496463, 168486463]
[168496463, 168495463]
[168496463, 168496453]

 2022-05-04, 14:51 #548 slandrum   Jan 2021 California 2×211 Posts The OP has said that they have vision trouble, so there's other types of transcription errors that are possible, 3/8 confusion, 0/8 confusion, 5/6 confusion (well this is covered in the fat-fingering case), 2/7 confusion and possibly others depending on the degree of vision impairment and the font used.
2022-05-04, 15:56   #549
Dr Sardonicus

Feb 2017
Nowhere

582810 Posts

Quote:
 Originally Posted by slandrum The OP has said that they have vision trouble, so there's other types of transcription errors that are possible, 3/8 confusion, 0/8 confusion, 5/6 confusion (well this is covered in the fat-fingering case), 2/7 confusion and possibly others depending on the degree of vision impairment and the font used.
I was intrigued by the OP's specificity in claiming to have mistyped a single digit. Why only one - and how did he know? And, the OP has expressed interest in digit patterns (both in decimal and base twelve) in candidate exponents, so, one would assume, want to be as sure as possible of entering exponents correctly into the software.

I made the simplifying assumption that in each instance the OP knew exactly what exponent he intended to test, and made the simplest kind of mistake in mistyping it. I also wanted to keep the number of cases to check under control.

In most circumstances, transcription errors due to mistyping can easily be avoided, by using copy-paste instead of retyping. If the desired number can't be copy-pasted directly (e.g. if it's hand-written on a piece of paper), it can be typed into a text editor, with the size cranked up as large as necessary to make it clearly readable, and then copy-pasted from the text editor to wherever it needs to go.

If there is a way to mistype something, my dim old eyes and stumbling fingers have probably found it. In my experience, the key to reducing typing mistakes is to avoid typing whenever possible.

 2022-05-05, 00:00 #550 tuckerkao   "Tucker Kao" Jan 2020 Head Base M168202123 13458 Posts There were 3 cases which I finished the trial factoring or P-1 or PRP of another exponent first than originally intended. 1. M168173323, M168174323 (P-1 was completed on the 2nd exponent earlier instead of the 1st exponent) 2. M168607123, M168607223 3. M168828523, M168829523, M168928523 There was also 1 case of swapping between the 2 adjacent digits: M168468523, M168648523 I didn't enter the wrong exponent between M168155623 and M168156523, but still mixed up between the 2 at times. There maybe other cases but I don't remember at this moment. Last fiddled with by tuckerkao on 2022-05-05 at 00:42

 Similar Threads Thread Thread Starter Forum Replies Last Post LaurV GPU to 72 81 2020-12-02 05:17 jfamestad PrimeNet 3 2016-11-06 20:32 lidocorc PrimeNet 4 2008-11-06 18:48 jocelynl 15k Search 0 2003-07-11 14:23 ThomRuley Software 5 2003-05-30 20:34

All times are UTC. The time now is 00:29.

Wed Jun 29 00:29:01 UTC 2022 up 75 days, 22:30, 1 user, load averages: 2.05, 1.96, 1.92