Opinions/Suggestions for Data Collection thread - Page 2

henryzz · 2008-10-22, 17:11

i have always thought that this forum needs a general prime searching subforum like the factoring subforum

Flatlander · 2008-10-22, 22:40

Quote:

Originally Posted by Flatlander

Yes, a good idea. For instance, I've started testing 28 different ks but wasn't sure where to report this. Ok, I'll report it here!

I'm testing k*2^n-1 from n=350,000 upwards for the following ks:
1401, 1403, 1407, 1413, 1419, 1431, 1433, 1435, 1437, 1441, 1445, 1449, 1459, 1467, 1473, 1479, 1483, 1487, 1489, 1491, 1493, 1495, 1497
(Only tested to 351,387.)

3087, 3969, 4851, 5733, 6615
(Tested to 365,300.)

I've just noticed that I included k=1431 by mistake. It is being tested by H.Platzer so I have deleted it from the seive file and am no longer testing it.

IronBits · 2008-10-22, 23:57

I'm not sure what DATA you all are talking about, but if it needs to be in the NPLB or CRUS databases we started for stats, then /waves hand ;)

gd_barnes · 2008-10-23, 00:38

I've reread this thread in more detail and realized I was in error in assuming that it was intended for the collection of results. It's more about collection of DATA regarding prime searching.

I started to compile a list of all k's 300-1001 with the number of primes for each that NPLB has found with some interesting info. about each like:

k=xxx; x # of primes; x # of even primes, x # of odd primes; x primes for n==(0 mod 3); x primes for n==(1 mod 3), etc. The distribution of n-values for primes in specific k's is quite fascinating and relates to conversion to different bases at CRUS.

Ultimately I found I just didn't have time to get into such detail. But anything like that that shows strange patterns and anomolies would be very interesting to compile and would encourage others to search unusual k's or ranges more to see if patterns hold.

I shall definitely vote for this effort.

One more thing: Perhaps we should start a 'choose your own k' thread like RPS has. Not to pilfer what they are doing but to collect info. on searched ranges for random k's that wouldn't necessarily be otherwise reported; exactly like what Flatlander has done here with his k=1400-1500 effort. Perhaps we could call such a thread "report randomly searched k's and n-ranges" or simply "statuses on random prime searches". The thread could be for ANY kind of prime search, including other bases, Proths, PRP's like for k*2^n-3 or +3, etc.

One thing that I would add to such a thread right away: I searched all k<20 up to n=150K for the forms 2^n-k and 2^n+k about a year ago. I reported several PRP's to the top-10000 PRP site at the time including finding a missing PRP in the -3 form. I think I still have my sieved file that goes up to n=200K on it.

Let me know what you guys think.

Gary

MooooMoo · 2008-10-23, 07:26

Quote:

Originally Posted by gd_barnes

The thread could be for ANY kind of prime search, including other bases, Proths, PRP's like for k*2^n-3 or +3, etc.

I don't think PRP's should be included, as they are not primes. IMO, "almost prime" isn't good enough. The exception would be PRPs that have a reasonable chance of being proven prime by ECPP or any other method within a realistic time frame (less than 2-3 years).

Riesel base 2 primes should come first if storage concerns are a major issue. If it's not, then it's perfectly fine to include primes of other bases and Proths.

kar_bon · 2008-10-23, 08:02

the intention of the new forum/thread is to collect primes/ranges of people not working for a special project or different types of primes:
for example there're people searching for prime of the form k*2^n-1 (yeah, Riesel numbers too) with n equal to a Mersenne prime n so called a General Mersenne Prime (example: 755061*2^756839-1 for M32).
i know that's the best way of all is to collect all information about any k/base/n and results(RES64),ranges and stats like Gary mentioned.
i begun to collect data for the Riesel side, i hope there'll be someone to make a page for Sierpinski. other bases collected at CRUS.
but it's too much to do for only one person to handle all.
automatisms needed for those purposes but this needs also man-power and system-resources.
so for the beginning the most searched numbers Riesel and Sierpinski at base 2 are a good choice.
for myself i collect all the results for tested ranges namely the RES64/primes/sieved files for any k for later verification.

so where is the one managing the Sierpinski side?

i'll continue the Riesel ones and that's work/fun enough for one person and there're many many info. i've not included yet!

kar_bon · 2008-11-26, 12:26

this vote is without any new opinions and there're many people whising such a new thread.

i informed the forum admin (Mike) to make a new thread available.

hope this will come the near future!

Karsten

Xyzzy · 2008-11-27, 01:55

Since it is a new "baby" forum, how about we start it here:

http://www.mersenneforum.org/forumdisplay.php?f=89

As time goes on we can evaluate a better place for it.

Does this sound okay?

kar_bon · 2008-11-27, 09:27

that's great and ok for me.

see what the future will bring and then you can put the thread on a higher place

thanks a lot for creating this.

now i have to fill it with some 'life'!

View Poll Results: Do we need a thread for Prime data collecting?
Yes	12	100.00%
No	0	0%
What ever.	0	0%
Voters: 12. You may not vote on this poll

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
'All Your Data ❝Я❞ Belong To Us' Thread	ewmayer	Tales From the Crypt(o)	283	2021-05-25 16:23
Your opinions on "x times less" and other errors	Jayder	Miscellaneous Math	12	2014-10-26 14:41
Opinions wanted on direction of drive 3	gd_barnes	No Prime Left Behind	26	2008-03-07 05:40
Opinions on team drive #3	gd_barnes	Conjectures 'R Us	8	2008-01-24 01:01
Collection of Proofs?	Orgasmic Troll	Math	1	2004-12-30 15:10

2008-10-22, 17:11	#12
henryzz Just call me Henry "David" Sep 2007 Cambridge (GMT/BST) 2³×3×5×7² Posts	i have always thought that this forum needs a general prime searching subforum like the factoring subforum

2008-10-22, 23:57	#14
IronBits I ♥ BOINC! Oct 2002 Glendale, AZ. (USA) 2131₈ Posts	I'm not sure what DATA you all are talking about, but if it needs to be in the NPLB or CRUS databases we started for stats, then /waves hand ;)

2008-10-23, 00:38	#15
gd_barnes May 2007 Kansas; USA 3³·5·7·11 Posts	I've reread this thread in more detail and realized I was in error in assuming that it was intended for the collection of results. It's more about collection of DATA regarding prime searching. I started to compile a list of all k's 300-1001 with the number of primes for each that NPLB has found with some interesting info. about each like: k=xxx; x # of primes; x # of even primes, x # of odd primes; x primes for n==(0 mod 3); x primes for n==(1 mod 3), etc. The distribution of n-values for primes in specific k's is quite fascinating and relates to conversion to different bases at CRUS. Ultimately I found I just didn't have time to get into such detail. But anything like that that shows strange patterns and anomolies would be very interesting to compile and would encourage others to search unusual k's or ranges more to see if patterns hold. I shall definitely vote for this effort. One more thing: Perhaps we should start a 'choose your own k' thread like RPS has. Not to pilfer what they are doing but to collect info. on searched ranges for random k's that wouldn't necessarily be otherwise reported; exactly like what Flatlander has done here with his k=1400-1500 effort. Perhaps we could call such a thread "report randomly searched k's and n-ranges" or simply "statuses on random prime searches". The thread could be for ANY kind of prime search, including other bases, Proths, PRP's like for k*2^n-3 or +3, etc. One thing that I would add to such a thread right away: I searched all k<20 up to n=150K for the forms 2^n-k and 2^n+k about a year ago. I reported several PRP's to the top-10000 PRP site at the time including finding a missing PRP in the -3 form. I think I still have my sieved file that goes up to n=200K on it. Let me know what you guys think. Gary

2008-10-23, 08:02	#17
kar_bon Mar 2006 Germany 5530₈ Posts	the intention of the new forum/thread is to collect primes/ranges of people not working for a special project or different types of primes: for example there're people searching for prime of the form k2^n-1 (yeah, Riesel numbers too) with n equal to a Mersenne prime n so called a General Mersenne Prime (example: 7550612^756839-1 for M32). i know that's the best way of all is to collect all information about any k/base/n and results(RES64),ranges and stats like Gary mentioned. i begun to collect data for the Riesel side, i hope there'll be someone to make a page for Sierpinski. other bases collected at CRUS. but it's too much to do for only one person to handle all. automatisms needed for those purposes but this needs also man-power and system-resources. so for the beginning the most searched numbers Riesel and Sierpinski at base 2 are a good choice. for myself i collect all the results for tested ranges namely the RES64/primes/sieved files for any k for later verification. so where is the one managing the Sierpinski side? i'll continue the Riesel ones and that's work/fun enough for one person and there're many many info. i've not included yet!

2008-11-26, 12:26	#18
kar_bon Mar 2006 Germany 101101011000₂ Posts	this vote is without any new opinions and there're many people whising such a new thread. i informed the forum admin (Mike) to make a new thread available. hope this will come the near future! Karsten

2008-11-27, 01:55	#19
Xyzzy "Mike" Aug 2002 2⁵×257 Posts	Since it is a new "baby" forum, how about we start it here: http://www.mersenneforum.org/forumdisplay.php?f=89 As time goes on we can evaluate a better place for it. Does this sound okay?

2008-11-27, 09:27	#20
kar_bon Mar 2006 Germany 2³·3·11² Posts	that's great and ok for me. see what the future will bring and then you can put the thread on a higher place thanks a lot for creating this. now i have to fill it with some 'life'!