Broken aliquot sequences

[schickel]

Originally Posted by schickel

Can't download .elf files at the moment. From the browser there's no response, haven't waited long enough to get a time-out, but I assume it would eventually. When I try to download it, I get just the home page page with nothing following:

OK, it's working again.

[Greebley]

I found some sequences with errors reported by aliqueit:
3785436: ERROR: @index 2021: value != sigma - n
3845784: ERROR: @index 91: value != sigma - n
4159596: ERROR: @index 562: value != sigma - n
4388376: ERROR: @index 1265: value != sigma - n

4751220: ERROR: factor not prime: 443788656239 [@index 1189]
4949376: ERROR: factor not prime: 374606970763 [@index 191]
5202540: ERROR: factor not prime: 55929958013 [@index 299]

For the final 3 there is likely 2 copies of the factor with the aliquot sequence using the wrong one. If that is fixed, it would have an error like the first 4 because the sigma-n would change.

[henryzz]

If you were generating a list of sequences and where they are upto for >1M I would be quite interested.

[ugly2dog]

Originally Posted by henryzz

If you were generating a list of sequences and where they are upto for >1M I would be quite interested.

I am in the process of uploading data on all the open sequences from >1M to <5M [approx 38404 open sequences]. They will have been taken up to 90 digits [or more] when I get done.
It just takes a lot of time to go through each of the files and upload the information into the db.

Right now everything in the 3xxxxxx and 4xxxxxx ranges were taken to at least 90 digits.

I am working on the 2xxxxxx range this week. I have 237 more sequences to input and the entire range will be at the 90 digits or more.

I have only have 7438 sequences in the 1xxxxxx range to go after that to get them all to 90 digits or more, so it will take a while.
I don't know if any of them have terminated or not.

I am in the process of extending the 3xxxxxx range to least 95 digits, the 2xxxxxx range to 100 digits.
It will be a while before I get that data entered.

I don't have a status list that is current from the db. I just have my copies of the sequences and how far each one is. It doesn't have any information on drivers, etc.

[VBCurtis]

If I were interested in running a handful (say, a half-dozen) of the 4-5M sequences up to 110 digits, could you PM or post a list of sequences that have not terminated?

-Curtis

[chris2be8]

When did you start loading them into factordb? I noticed a lot of 70-90+ digit composites appeared at about the end of May, and I've been factoring them since then. At present I'm factoring numbers up to 96 digits, that could be raised if there are not many left below that size.

What proportion of sequences will have terminated or merged? I can't see many composites added that are over about 95 digits.

I'm also factoring SNFS targets from random_composites.txt (as long as faster than GNFS). The main reason for doing small numbers is to clear out enough to ensure I get a reasonable number of SNFS targets in random_composites.txt.

Chris

[ugly2dog]

Originally Posted by chris2be8

When did you start loading them into factordb? I noticed a lot of 70-90+ digit composites appeared at about the end of May, and I've been factoring them since then.

Chris

Those weren't from me as I have been extending the sequences above 1M for quite sometime [several years]. Everything I have entered has been fully factored. The next number in the sequence obviously isn't though.
Those won't show up until someone queries the db for that sequence, to my knowledge.

I finished uploading the the 2xxxxxx sequences yesterday.
So everything from 2M to 5M is has been taken to at least 90 digits.

I have 38402 open sequences in the 1,000,000-4,999,999 range, but I have a bunch [7000+] of in the 1xxxxxxx that I haven't even looked at yet. but I have had 634 sequences terminate in the last eight months.

[ugly2dog]

Originally Posted by VBCurtis

If I were interested in running a handful (say, a half-dozen) of the 4-5M sequences up to 110 digits, could you PM or post a list of sequences that have not terminated?

-Curtis

There are 9688 sequences in the 4xxxxxx range that are still open to the best of my knowledge.
Of those 9521 have been taken to 90 digits, 98 to 95 digits, 62 to 100 digits and seven that are between 111 and 128 digits.

I can send you a list later today.