[seqfan] Re: Interesting discovery by Dan Preston concerning a Recaman-like sequence

[Allan Wechsler]

I don't know anything about what algorithms people are using here. Are we
using ordinary bit-vectors, with each bit keeping track of whether the
sequence has hit a particular number? If so, it seems to me (just from
hand-simulation) that run-length encoding would save several orders of
magnitude on space, and probably only cost one order of magnitude in time
(storing the RLE structure as a balanced tree).

On Thu, Mar 21, 2019 at 9:47 AM Hans Havermann <gladhobo at bell.net> wrote:

> I'm running the code (with 11281) as I write, so far without interruption.
> The printouts of how many iterations have been reached are now >
> 1.75*10^12. The only issue I have is that the file that is keeping track of
> the bit vectors is showing at first glance to be > 5.3 TB, which is
> impossible because the storage drive is only 2 TB and a system check
> appears to indicate that it still has 1.25 TB of space available.
>
> > On Mar 20, 2019, at 12:10 PM, hv at crypt.org wrote:
> >
> > I've written some code to attempt this, available via commit b8eef3ea at
> > <https://github.com/hvds/seq>.
> >
> > This keeps a bit vector of the values seen, split into chunks, and spills
> > chunks to disk to limit the memory use.
> >
> > It appears to give correct values for known n, but trying it with 11281
> it
> > gives a bus fault a little after 12,213,813,248 (which takes about 5
> minutes).
> > That's around the time I run out of disk space, so that may be the whole
> > of the problem, but there may be a deeper problem.
>
>
> --
> Seqfan Mailing list - http://list.seqfan.eu/
>