NPLB Free-DC Stats

[IronBits]

Gary/Max, where are your results.txt files located at so I can download them, if you want me to import them into the database.
Also, which file do I start with? Can you mark them with a red * or something similar to what Carlos does to indicate that they have already been processed and imported?

[IronBits]

results_20081115_1909_CC_nplb_443.zip *
results_20081120_1846_CC_nplb_443.zip *
results_20081123_1932_CC_nplb_443.zip *
Those were already done, fortunately I figured it out before they got processed twice.

Mark this one done to please - it was just imported into the db.
results_20081124_2149_CC_nplb_443.zip *

[em99010pepe]

Originally Posted by IronBits

results_20081115_1909_CC_nplb_443.zip *
results_20081120_1846_CC_nplb_443.zip *
results_20081123_1932_CC_nplb_443.zip *
Those were already done, fortunately I figured it out before they got processed twice.

I don't understand, last import was on Nov 13 so those are missing.

Carlos

[IronBits]

Below is what I show as being processed for your Server, thanks to the _CC_ part in the name
After the perl script imports them, it moves each processed .csv to a backup location, in case a restore to the database is required...

results_20080907_0658_CC_nplb_443.csv
results_20080912_0658_CC_nplb_443.csv
results_20080916_0658_CC_nplb_443.csv
results_20080917_0658_CC_nplb_443.csv
results_20080919_0658_CC_nplb_443.csv
results_20080923_0658_CC_nplb_443.csv
results_20080924_0658_CC_nplb_443.csv
results_20080925_0658_CC_nplb_443.csv
results_20080926_0658_CC_nplb_443.csv
results_20080927_0658_CC_nplb_443.csv
results_20080930_0658_CC_nplb_443.csv
results_20081001_0658_CC_nplb_443.csv
results_20081002_0658_CC_nplb_443.csv
results_20081003_0658_CC_nplb_443.csv
results_20081005_0658_CC_nplb_443.csv
results_20081006_0658_CC_nplb_443.csv
results_20081009_0658_CC_nplb_443.csv
results_20081015_0658_CC_nplb_443.csv
results_20081017_0658_CC_nplb_443.csv
results_20081024_0658_CC_nplb_443.csv
results_20081115_1909_CC_nplb_443.csv
results_20081120_1846_CC_nplb_443.csv
results_20081123_1932_CC_nplb_443.csv
results_20081124_2149_CC_nplb_443.csv

[em99010pepe]

Just looked at the stats, the results have been imported thanks. My last message was refered to the fact that the results were processed but not yet imported to the stats.

From now on I'll be concentrated on doing some manual ranges so I think it's better to find a way to import results from manual work....

Carlos

[IronBits]

Looks like Bok updates the stats now at 00:58 (-7 GMT)
I let him know that I started processing the .csv files at 00:00, so he adjusted accordingly.

http://stats4.free-dc.org/stats.php?...b&team=Free-DC

Getting better all the time

ha! You posted while I was composing mine

I can do that, send me a sample output file and I'll make a .txt to .csv converter for it.
Just need to standardize that Date/Time Format to yyyy-mm-dd HH:MM and it will be good to go.
If the output files resemble what is in results.txt, I could import everything under the sun, since the beginning of time, just need someone to rustle up all the files :LOL:

[IronBits]

This really shows how everyone is doing... Looks like Bok has put some time into fixing up the database to :cheers: to him for that.

I like all the graph work on the beta site

http://stats4.free-dc.org/stats.php?...sers&proj=nplb

[mdettweiler]

Originally Posted by IronBits

Gary/Max, where are your results.txt files located at so I can download them, if you want me to import them into the database.
Also, which file do I start with? Can you mark them with a red * or something similar to what Carlos does to indicate that they have already been processed and imported?

Our csv files are at http://nplb-gb1.no-ip.org/llrnet/csv/.

As to whether we can mark them with something that shows they've been imported: well, I'm not so sure about that. You see, the index of that page isn't a fancy HTML file like Carlos is using--just the listing of the contents of the /llrnet/csv/ directory sort of like what you do for your /results/ directories. Thus, I don't know of any way to add anything that denotes what's been imported and what hasn't.

As to what to start with, though, I'm afraid I really have no idea what you've imported from the various servers yet. First of all, have you imported anything at all from G4000 yet? If not, just grab all of them and you should be good. If, on the contrary, you *have* imported stuff already from G4000...then I would recommend looking at the time/date of the last imported result for G4000 and manually locating the results file ending with that pair. That's what I do when I'm processing the LLRnet results, to determine how far back I have to download and concatenate onto the end of the "big results file" for the respective server that I store on my hard drive.

Originally Posted by IronBits

Looks like Bok updates the stats now at 00:58 (-7 GMT)
I let him know that I started processing the .csv files at 00:00, so he adjusted accordingly.

http://stats4.free-dc.org/stats.php?...b&team=Free-DC

Getting better all the time

ha! You posted while I was composing mine

I can do that, send me a sample output file and I'll make a .txt to .csv converter for it.
Just need to standardize that Date/Time Format to yyyy-mm-dd HH:MM and it will be good to go.
If the output files resemble what is in results.txt, I could import everything under the sun, since the beginning of time, just need someone to rustle up all the files :LOL:

Hmm...let's hold off on importing any manual stuff until we get Gary's approval first. The database is already huge enough the way it is--and believe me, adding all the manual ranges would add millions of additional results to the DB, possibly even tens of millions. And on top of that, manual LLR doesn't write down the date and time that each result is generated--it just pops the results in a file like this:

851*2^569214-1 is not prime.  LLR Res64: 595AE39D6CE2501F  Time : 489.236 sec.
873*2^569214-1 is not prime.  LLR Res64: ED643B1A150B2FFE  Time : 484.171 sec.
917*2^569214-1 is not prime.  LLR Res64: 9F5C73545055B697  Time : 489.560 sec.
927*2^569214-1 is not prime.  LLR Res64: DEFBF51F8884AF3F  Time : 483.263 sec.
429*2^569215-1 is not prime.  LLR Res64: 8515DA83BB2FD309  Time : 508.224 sec.
433*2^569215-1 is not prime.  LLR Res64: 020DDF7FDB59214B  Time : 514.832 sec.

...i.e., no user or date/time line. Granted, the user line would be easy to figure out, since each manual range is assigned to just one person (unlike an LLRnet server range which is split up among multiple users)--but the best thing we could hope for as far as date/time goes is, say, the date/time that the user posted their forum message reporting the range complete.

Thus, I'm not so sure that importing manual results would be a good idea. The final decision, of course, is up to Gary--but, anyway, just my $0.02.

Max

P.S.: That's not to say, however, that we couldn't do a separate database for manual results. A separate database might be ideal, so that we don't have a discrepancy where more than half of the results in our main database essentially have "dummy" dates and times.

Even with a separate database, of course, we could still, say, have one main stats page that ranks users/teams/etc. based on the combined LLRnet and manual stats. That would be quite feasible, I'd think--and much easier/better than having to muck around with putting everything in one database.

[IronBits]

OK, my head hurts, forget about me doing anything with or mentioning anything about manual reservation stats.

Sounds like the perfect job for you

[IronBits]

You could make those stats file names with a GB instead of just one G, else I'll rename them once I get them here for importing.
I'll have to create a directory for Gary's files so I can compare what I have to what's up there. Hopefully we'll be able to do just an wget type thing to snag them and place them into the import directory, how to keep from getting dupes is another matter.

[mdettweiler]

Originally Posted by IronBits

You could make those stats file names with a GB instead of just one G, else I'll rename them once I get them here for importing.
I'll have to create a directory for Gary's files so I can compare what I have to what's up there. Hopefully we'll be able to do just an wget type thing to snag them and place them into the import directory, how to keep from getting dupes is another matter.

Okay, I've changed my script to put "GB" in the filenames instead of "G", and I've also renamed all the old CSV and results files to the new naming convention.