[seqfan] Re: 2D extension to A327474

[nitai.s at campus.technion.ac.il]

Thank you :)

I am not familiar with PARI so I glanced over that part of your previous message. Upon closer inspection - nope, I still can't read that! :P

I had forgotten to post this to the mailing list, but after some more help from Zachary, I've submitted the sequence:

https://oeis.org/draft/A360192 - nonempty subsets
https://oeis.org/draft/A360193 - all subsets, with empty set having a unique mean (principally (0,0) but it doesn't matter so long as it's distinct from any other subset's mean)

You are all of course welcome to make or suggest edits.

There is debate whether two variants that just differ by 1 should be included. See the draft discussion.

I'll leave it to Zachary to share his C++ code - he shared it with me, it's quite good. You can see in the drafts the extra elements.

Nitai Sasson

________________________________
From: SeqFan <seqfan-bounces at list.seqfan.eu> on behalf of M. F. Hasler <oeis at hasler.fr>
Sent: Tuesday, 31 January 2023, 3:03
To: Sequence Fanatics Discussion list <seqfan at list.seqfan.eu>
Subject: [seqfan] Re: 2D extension to A327474

If you multiply the values {1...n} right from the beginning by LCM(1...n)
as I suggested in the PARI code, you can use integers only, no more
fractions, which will certainly speed up the code by a significant factor.
But yes, it might not be enough to go more than 1-2 n's further...

FWIW, i think the sequence is interesting and should be in OEIS, as well
the diagonal n×n as the triangle m×n, 1 ≤ n ≤ m .

- Maximilian