Post status updates here - Page 25

henryzz · 2009-07-07, 16:01

Quote:

Originally Posted by 10metreh

The same problem once happened to henryzz on sequence 127308.

except for the fact that it was in the reservation thread and it still didnt stop until it reached a c104 a while after i had made it known on the forum

gd_barnes · 2009-07-08, 00:33

This is just plain wrong!!

233324 had a downdriver run all the way from 106 digits to 8 digits and it stopped 2 digits short of merging with a lower sequence! 615 indexes total! This is the 3rd downdriver run of more than 10 digits on this sequence. As of this posting, it is at i=3314 and back to 88 digits.

Now it's hopeless. It's making up for lost time with a ton of 3 and 5 factors and is currently driven by 2^3*3*5. Since the downdriver ended, it's one of the highest slopes I've experienced.

I'll have to admit it was exciting though. The first 2 runs were overnight so I missed them but I got to watch my machine kurchunk through this one. My heart pounded as it dropped below 50 digits. I was out of my seat at 25 like a baseball fan waiting for that final out of the World Series! lol I was then slumped in my chair as it hit 8 digits and rose back above 30. This is one of the few times that running a slower machine was far more fun. I could actually see it scroll quite clearly until it dropped below about 20 digits.

It's nice to feel like these things have a chance to terminate or merge now that I've experienced such a run.

One more thing on a technical note: I'm not sure exactly how these things are defined but the run had only a factor of 2^1 for 336 indexes from 106 to 42 digits (-64 digits). After that, there were 7 different runs of 2^2 and 2^3 mixed in there with the 2^1 runs but 2^1 came back each time until the final 2^2 run added the factor of 3. In all except 1 of the seven 2^2-2^3 runs, it still dropped or stayed the same with the max being a 2-digit drop. I assume it is still counted as one complete downdriver run since the factor 3 never came until it hit just 8 digits.

Gary

axn · 2009-07-08, 01:04

Quote:

Originally Posted by gd_barnes

I assume it is still counted as one complete downdriver run since the factor 3 never came until it hit just 8 digits.

Nope. Nothing to do with "3". Only the presence of the "downdriver" (2^1) qualifies for a "downdriver run."

gd_barnes · 2009-07-08, 02:46

I'm still not clear. So the fact that 2^2 creeped in there ends the downdriver run? If so, there are 8 separate downdriver runs during that stretch for a total of 10 on that one sequence. Would that be some sort of record for one sequence?

Here are the runs of 2^1 during that entire stretch in which there was no factor of 3 as well as higher powers of 2 that occurred in between shown in parens:

i=2454 to 2789; 106 to 42 digits; 336 indexes, -64 digits
(many 2^2 thru 2^7 in between)
i=2814 to 2833; 40 to 36 digits; 20 indexes, -4 digits
(many 2^2 thru 2^5 plus 2^7 in between)
i=2886 to 2946; 34 to 24 digits; 61 indexes, -10 digits
(many 2^2 plus one 2^7 in between)
i=2958 to 3003; 24 to 15 digits; 46 indexes, -9 digits
(all 2^4 in between)
i=3010 to 3012; 15 to 14 digits; 3 indexes, -1 digit
(all 2^2 and 2^4 in between)
i=3028 to 3052; 14 to 9 digits; 25 indexes, -5 digits
(2^3, 2^4, and 2^6 in between)
i=3058 to 3064; 10 to 8 digits; 7 indexes, -2 digits
(2^2 and 2^4 in between)
i=3067 to 3068; 8 to 8 digits; 2 indexes, 0 digits

Followed by 9 straight 2^2, which were then followed by the dreaded 2^2*3^2.

This brings up a topic that was discussed briefly in another thread: That of ever increasing sequences. The low point after the above run was 14650972 at i=3069. It has been straight up ever since then, mainly as a result of the factor of 3 that crept in 9 indexes later.

Gary

Greebley · 2009-07-08, 04:06

You can get a lot more downdriver runs with fewer digits because the odds of a single big prime become much higher. That can start and end down-driver runs. Tough luck on running into one of the worst up drivers (because there are so many low factors it increases very fast).

10metreh · 2009-07-08, 07:25

Quote:

Originally Posted by gd_barnes

If so, there are 8 separate downdriver runs during that stretch for a total of 10 on that one sequence. Would that be some sort of record for one sequence?

Nope. Take 130396 (probably nowhere near the record):

Lines 185-198; size 36-33
Lines 231-315; size 35-20
Lines 474-525; size 36-25
Lines 539-570; size 25-19
Lines 596-604; size 19-17
Lines 647-649; size 17-17
Lines 1311-1367; size 88-76
Lines 1439-1534; size 81-59
Lines 2045-2082; size 97-90
Lines 2235-2315; size 96-83
Lines 2431-2460; size 91-85
Lines 2588-2738; size 81-53
Lines 2827-2887; size 52-41
Lines 2926-2967; size 39-29
Lines 2979-3005; size 28-23
Lines 3250-3263; size 51-48
Lines 3340-3430; size 54-36
Lines 3731-3850; size 81-57
Lines 3900-3932; size 56-51
Lines 3974-3990; size 49-46
Lines 4049-4053; size 51-50
Lines 4082-4150; size 51-37
Lines 4161-4162; size 36-36
Lines 4170-4215; size 36-27

To give you an idea how those 24 downdriver runs fit in, here's the graph:

gd_barnes · 2009-07-08, 07:35

Hum. Very good...and that answers my question about what specifically constitutes a downdriver run.

I just realized that 233324 has far more downdriver runs than just 10. 10 is the number since I've taken it over. There are at least 2 more and I'll have to look in the file to see if those are broken up with 2^x (where x>1) runs in between them like my long run was. If so, there could likely be 20 or more...probably not as much as your amazing 24 but not too bad.

In the mean time, 233324 has made a bee line to 103 digits and its rapidly approaching a hard C>=100 where I will be done with it. (not there yet) It's never let go of the 2^3*3*5 driver since 9 indexes after it reached 8 digits...most remarkable right after a huge downdriver run. The graph at the end is almost a perfect V formation.

10metreh · 2009-07-08, 07:44

Just to report that 233324 did not acquire a side-sequence under 1 million. (Well, it probably didn't - C9C30 isn't fully up to date.)

kar_bon · 2009-07-08, 09:14

the end (or beginning) of a downdriver can be quite sure noticed, when a factorization with 2*p occurs:

Gary, to your 233324 seqs:
see the 'downdriver' runs just before it hits the uptime:
index 3064: 2*p -> next index is 2^2*5*p (not so bad)
index 3068: 2*p -> next index is driver 2^2*7 (bad)

the same for a driver:
index 2453: 2^3*p -> next index downdriver with 2!

there're many other examples with <driver>*p changing the driver. mostly bad to good, or good to bad.

so if you got a downdriver, the chance the enter a factorization of 2*p becomes quite more possible when the digitsize is decreasing.
and mostly that is the return to a driver! and, boy, goooooooo to the sky!

gd_barnes · 2009-07-08, 13:24

Quote:

Originally Posted by 10metreh

Just to report that 233324 did not acquire a side-sequence under 1 million. (Well, it probably didn't - C9C30 isn't fully up to date.)

What is a side sequence?

kar_bon · 2009-07-08, 13:31

Quote:

Originally Posted by gd_barnes

What is a side sequence?

sometimes a sequences get the same number of another sequence and therefore one of those seqs can be omitted.

for example: 18480 goes from index 368 in to the seq 11408, so you could stop 18480!
if you compare the 2 graphs it's best to see!
only the lower seq will be calculated further!

BTW: in the summary they coloured green with the note: Merges...

2009-07-08, 00:33	#266
gd_barnes May 2007 Kansas; USA 2×41×127 Posts	This is just plain wrong!! 233324 had a downdriver run all the way from 106 digits to 8 digits and it stopped 2 digits short of merging with a lower sequence! 615 indexes total! This is the 3rd downdriver run of more than 10 digits on this sequence. As of this posting, it is at i=3314 and back to 88 digits. Now it's hopeless. It's making up for lost time with a ton of 3 and 5 factors and is currently driven by 2^335. Since the downdriver ended, it's one of the highest slopes I've experienced. I'll have to admit it was exciting though. The first 2 runs were overnight so I missed them but I got to watch my machine kurchunk through this one. My heart pounded as it dropped below 50 digits. I was out of my seat at 25 like a baseball fan waiting for that final out of the World Series! lol I was then slumped in my chair as it hit 8 digits and rose back above 30. This is one of the few times that running a slower machine was far more fun. I could actually see it scroll quite clearly until it dropped below about 20 digits. It's nice to feel like these things have a chance to terminate or merge now that I've experienced such a run. One more thing on a technical note: I'm not sure exactly how these things are defined but the run had only a factor of 2^1 for 336 indexes from 106 to 42 digits (-64 digits). After that, there were 7 different runs of 2^2 and 2^3 mixed in there with the 2^1 runs but 2^1 came back each time until the final 2^2 run added the factor of 3. In all except 1 of the seven 2^2-2^3 runs, it still dropped or stayed the same with the max being a 2-digit drop. I assume it is still counted as one complete downdriver run since the factor 3 never came until it hit just 8 digits. Gary Last fiddled with by gd_barnes on 2009-07-08 at 00:45

2009-07-08, 07:44	#272
10metreh Nov 2008 912₁₆ Posts	Just to report that 233324 did not acquire a side-sequence under 1 million. (Well, it probably didn't - C9C30 isn't fully up to date.) Last fiddled with by 10metreh on 2009-07-08 at 07:45

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Any updates please ?	Aillas	XYYXF Project	11	2020-04-05 12:38
Updates	Baumann Eduard	Information & Answers	5	2012-10-25 18:44
result updates??	timespy	Information & Answers	3	2009-04-28 20:26
K=13236795 Status Post	ValerieVonck	15k Search	33	2005-10-31 22:54
Stats updates	SandStar	NFSNET Discussion	7	2004-07-03 22:59

2009-07-08, 02:46	#268
gd_barnes May 2007 Kansas; USA 2·41·127 Posts	I'm still not clear. So the fact that 2^2 creeped in there ends the downdriver run? If so, there are 8 separate downdriver runs during that stretch for a total of 10 on that one sequence. Would that be some sort of record for one sequence? Here are the runs of 2^1 during that entire stretch in which there was no factor of 3 as well as higher powers of 2 that occurred in between shown in parens: i=2454 to 2789; 106 to 42 digits; 336 indexes, -64 digits (many 2^2 thru 2^7 in between) i=2814 to 2833; 40 to 36 digits; 20 indexes, -4 digits (many 2^2 thru 2^5 plus 2^7 in between) i=2886 to 2946; 34 to 24 digits; 61 indexes, -10 digits (many 2^2 plus one 2^7 in between) i=2958 to 3003; 24 to 15 digits; 46 indexes, -9 digits (all 2^4 in between) i=3010 to 3012; 15 to 14 digits; 3 indexes, -1 digit (all 2^2 and 2^4 in between) i=3028 to 3052; 14 to 9 digits; 25 indexes, -5 digits (2^3, 2^4, and 2^6 in between) i=3058 to 3064; 10 to 8 digits; 7 indexes, -2 digits (2^2 and 2^4 in between) i=3067 to 3068; 8 to 8 digits; 2 indexes, 0 digits Followed by 9 straight 2^2, which were then followed by the dreaded 2^2*3^2. This brings up a topic that was discussed briefly in another thread: That of ever increasing sequences. The low point after the above run was 14650972 at i=3069. It has been straight up ever since then, mainly as a result of the factor of 3 that crept in 9 indexes later. Gary

2009-07-08, 04:06	#269
Greebley May 2009 Dedham Massachusetts USA 3·281 Posts	You can get a lot more downdriver runs with fewer digits because the odds of a single big prime become much higher. That can start and end down-driver runs. Tough luck on running into one of the worst up drivers (because there are so many low factors it increases very fast).

2009-07-08, 07:35	#271
gd_barnes May 2007 Kansas; USA 10100010101110₂ Posts	Hum. Very good...and that answers my question about what specifically constitutes a downdriver run. I just realized that 233324 has far more downdriver runs than just 10. 10 is the number since I've taken it over. There are at least 2 more and I'll have to look in the file to see if those are broken up with 2^x (where x>1) runs in between them like my long run was. If so, there could likely be 20 or more...probably not as much as your amazing 24 but not too bad. In the mean time, 233324 has made a bee line to 103 digits and its rapidly approaching a hard C>=100 where I will be done with it. (not there yet) It's never let go of the 2^335 driver since 9 indexes after it reached 8 digits...most remarkable right after a huge downdriver run. The graph at the end is almost a perfect V formation.

2009-07-08, 09:14	#273
kar_bon Mar 2006 Germany 5534₈ Posts	the end (or beginning) of a downdriver can be quite sure noticed, when a factorization with 2p occurs: Gary, to your 233324 seqs: see the 'downdriver' runs just before it hits the uptime: index 3064: 2p -> next index is 2^25p (not so bad) index 3068: 2p -> next index is driver 2^27 (bad) the same for a driver: index 2453: 2^3p -> next index downdriver with 2! there're many other examples with <driver>p changing the driver. mostly bad to good, or good to bad. so if you got a downdriver, the chance the enter a factorization of 2*p becomes quite more possible when the digitsize is decreasing. and mostly that is the return to a driver! and, boy, goooooooo to the sky!