Too many similar sequences

[Robert G. Wilson v]

Hugo,

	When you suspect that there are multiple sequences with slight restrictions, use 
the advanced lookup where you can use other techniques. In this case the core 
sequence appears first.

	Or simply go into the data base through more pages - Database - full database, in 
this particular situation it would be part 19 and scan for the sequence 
1,2,3,5,7,...,.

	As they say there is more than one way to skin a goose.

Bob.

hv at crypt.org wrote:

> As part of my investigation into A33179, I wanted to find the sequence
> of partitions of n. So I calculated a few values by hand, whacked them
> into the advanced look-up page (my usual jump-off point), and waded
> through the 30-odd results looking for the one I wanted.
> 
> No joy, so I calculated more terms - I ended up having to calculate
> the first 9 terms (1, 2, 3, 5, 7, 11, 15, 22, 30) before A41 appeared
> at all in the search results.
> 
> It appears to be getting pushed out of the way by a lot of what I'd
> consider fairly trivial variations - they are mostly variant partition
> functions with minor restrictions - so I wonder if there is any way
> that the "core" sequences could be preferentially shown in such
> situations?
> 
> I appreciate that it may be difficult to reconcile this with the need
> to keep the database search as fast as possible: I suspect that the
> sequences are ordered (somehow) in numerical order, and that this
> particular search was problematic because all the variants tend to
> add a restriction so that - after enough terms - the variant starts
> to take lower values than A41, and so gets sorted earlier than A41.
> 
> Whatever the cause, I suspect this sort of problem will become more
> common as the database continues to grow.
> 
> It might also be useful to consider automatic linking of a selected
> set of key words and phrases - when searching with the first 8 terms
> of A41, none of the first 25 results returned include a reference to
> A41 even though many of them had "partition" as a part of the
> definition. I imagine most of the core sequence have short names
> that it might be useful to link in this manner - even "integer"
> might not be excessive.
> 
> Hugo
>