excluded/duplicate factors - what's the difference?

[Silverfish]

I've looked at my sieving stats and there are 3 types of Factors listed, U, D and E which I gather to correspond to Unique, Duplicate and Excluded. Unique Factors I gather are those that eliminate a test, but what's the difference between Duplicate and Excluded Factors?

[ceselb]

duplicate is those that fall within the n range of the dat (they were once there, but a factor removed them)

Excluded are beyond the n range (never present in the dat at all, but if the n range was greater it would have been a hit).
These are found sometimes, depending on how the bitmap is used. Sometimes it goes beyond the edge so to speak. (it's faster to do it like that in some cases in the current code)

[Mystwalker]

Caselb, I guess you mixed up excluded factors and those out of range...

Now that I think about it, I don't know where duplicates differ from excluded factors - seeing they are both present in the stats.
I can only guess:

- duplicate factors are factors of a k/n pair which already has a factor reported

- excluded factors are factors of a k/n pair which is not in the database anymore (basically, it flew out due to some sort of sieving as well...)


Any definite answer to this question?

[Death]

Originally posted by Silverfish
I've looked at my sieving stats and there are 3 types of Factors listed, U, D and E which I gather to correspond to Unique, Duplicate and Excluded. Unique Factors I gather are those that eliminate a test, but what's the difference between Duplicate and Excluded Factors?

there's also suggestion to submit ALL factors from fact.txt factexcl.txt and factrange.txt.

[vjs]

Yes, I submit all of the factors for each of my files.

It's sort of funny, one will see some holes in the gap finder a couple small holes etc.

But once I finish or every couple weeks or so I'll submit all the factors and transfer the >txt's into another folder.

Once I do this there are very little to no gaps at all. Actually I'm a little worried about one of my ranges it's still showing a 4G gap??? Had I not submitted all factors the gap probably would have been 20-50G.

[Joe O]

Originally posted by Mystwalker

- duplicate factors are factors of a k/n pair which already has a factor reported

- excluded factors are factors of a k/n pair which is not in the database anymore (basically, it flew out due to some sort of sieving as well...)

- excluded factors are factors of a k/n pair which never was in the sob.dat file at all. It was excluded because a factor less than 1G was found for it, during the process of creating the original sob.dat file.

[Mystwalker]

That confirms my assumption - thanks for the info!