Stats Update

[mgpower0]

Can someone supply me the link to sieve stats, the page I keep getting to hasn't been updated for about 5-6 Days. Am I using the wrong page or is there a problem with stats update at the moment???

[Joe O]

The stats are updated on an irregular basis by MikeH. He is a volunteer and has done a great job throughout the years.
The link the "all users" stats page is:
http://www.henleyclan.co.uk/sobsieve...e/ui/19999.htm

[engracio]

Hey Joe any idea when he is coming back and update the stats. Havent updated the site in about four months or so.

e

[Joe O]

Hey Joe any idea when he is coming back and update the stats. Havent updated the site in about four months or so.

e

I'm not sure what you mean. The last update was yesterday:

SB Sieving and factoring scores for engracio (id=7192)
Last Update: Wed 21-Mar-2007 18:30 (GMT)

Current 'complete' window:        0 < n <  6958662
Current 'active' windows:   6958663 < n <  7158663
                           12667294 < n < 12867294
Current 'n upper bound':    7324797 

Current 90% sieve point: 1330.613T (max bias: 33.27)

Score breakdown
                                 Total   New factors (    %)  Score change (    %)         Other (    %)
Daily change last  1 day :     4201.42       1217.61 ( 29.0)          0.00 (  0.0)       2983.81 ( 71.0)
Daily change last 15 days:     1454.52        585.90 ( 40.3)          0.00 (  0.0)        868.63 ( 59.7)

[KriZp]

The thing I have noticed is that the current (main) active window hasn't been updated since december. The DC window is updated with every stat update as far as I can see. I have been curious for some time about why the main window isn't updated regularly.

[shauge]

When started with SOB-Sieving the main window was updated every day, then approximately from last summer it started to be updated more rarely and now it has stopped, as you also have noticed. The way this stats funtions, it is important that the windows are up to date.
We can only hope that there is some background routine that keeps track.

[engracio]

When started with SOB-Sieving the main window was updated every day, then approximately from last summer it started to be updated more rarely and now it has stopped, as you also have noticed. The way this stats funtions, it is important that the windows are up to date.
We can only hope that there is some background routine that keeps track.

Shauge I agree with you, I also noticed that it has not been updated in a long time. I do not know if below is the cause of the stats update problem or what but it seems no answer from anyone is available. Joe does not think there is any issue. What can we do? Unless MikeH comes around and look around his website. it would be nice if he visit anytime soon. Like all of us, we're all busy.

OH well, much easier to just prp anyway, start it up let it run. Good change of scenery, no fuss or hassle.



"Mike's stats page has been stuck in its current position since the new year. Is there any reason why it is on
Current 'active' windows: 6972221 < n < 7172221
12667294 < n < 12867294
where is should be around 13300000-13400000. thanks."

e

[vjs]

I always thought it had to do with the next.txt page.


Alien88 fixed the link we will have to wait for a cycle and see if it works.

[MikeH]

Hi all,

Long time no speak. Work, life, etc.

I hope I've sorted everything out with the sieve stats (except for a couple of little things like the 60%: -0.001T 50%: -0.001T)

I've started a 2007 set of stats (this does go back to 1st Jan correctly), and I'll be stopping the 2006 (which has turned into a 2006 + 1H2007) in a couple of days. I'll also be archiving all the pre-2007 stuff onto another server at some point, but I'll leave some links so they won't be lost.

Now a little request. Most of the updates just work automatically, but I really don't get much time to look at issues, so could I ask anyone who submits factors without logging in (and they show up as user [unknown]) to post something somewhere so the mapping becomes obvious. This is most import for people that don't use the same ID for forum name and SB name - in this case please give me a mapping. So that's people like

dgordon20794 // dgordon in SoB proper
KWSN_Dagger // On SB he's just Dagger
XS_THE_MACHINE // On SB he's just XSTheMachine

Might look obvious enough, but a nightmare for me to figure out.

For the record, can so one let me know the SB names of the following (thanks):
Forum 0: 200255 - 200265 Lupus1647 (unknown)
Forum 0: 978500 - 978700 xian (unknown)
Forum 0: 1209000 - 1209010 torfau (unknown)
Forum 1: 63250 - 63320 Lupus1647 (unknown)
Forum 1: 63400 - 63475 Lupus1647 (unknown)
Forum 1: 74000 - 74200 Lupus1647 (unknown)

Great news on the latest factor find. Everyone keep up the good work.

Bye for now.

Mike.

[glennpat]

There is one unknown 1221.340T 67607 2404211 that is mine (glennpat). I'm using sobistrator to summit results. It looked like it was set up ok, but I could not tell if password was correct so I entered password again.

[engracio]

MikeH,

Thanks for the stats update. I know how hectic it can get. Thank for your support.


e

[Matt]

978500 - 978700 xian (unknown) - That's my friend Xian, I think he only crunched that one range. Can be credited to him as just Xian. Cheers

[balachmar]

How do you create the stats?
From which webpage? Because I might be able to host them as well, or write something myself, and put it on my server with a cron job. So that it is updated regularly.

[Death]

Hey Mike, nice to see you again. Hope you'll find somehow more time to chat here.

[glennpat]

The stats have not been updated for a week. Just thought I would mention it.

[glennpat]

The stats have not been updated for a week. Just thought I would mention it.

There were updated today. Thanks

[MikeH]

Nice to see the stats are still useful.

Vacation last week, so PC that does update was off. Advanced warning - next week I have a further two weeks vacation, so stats will probably not be updated between Fri 8th Aug and Sun 24th Aug.

PrimeGrid really has added some serious weight to the SoB project as a whole, removing 4% of all outstanding candidates so far this year. By contrast throughout the whole of last year 1% were removed. Brilliant, can only make those primes show themselves quicker.

[MikeH]

I've modified the sieve stats to show that there is life after n=20M. Only gone as far as 25M, but I now have all the data all the way to n=50M.

http://www.henleyclan.co.uk/sobsieve/2008/scores_p.htm

Total outstanding factors removed this year now up to about 6%. PrimeGrid you are a star

...and I just had to use this one....

[MikeH]

So 2008 ends with about 6.5% of all remaining candidates having been removed throughout the year. Truly incredible.

A realistic target for next year is for the number of candidates per million n to drop below 15000. Of course a lot will depend on where PrimeGrid recommends efforts should be directed.

Finding a new prime would make life so much easier.

BTW 2009 stats are now up and running.

http://www.henleyclan.co.uk/sobsieve/2009/scores_p.htm

[vjs]

Pretty amazing...

We are currently testing about half the tests per 1M from here on out compared to the number of tests per 1M we were testing at n=6M.