Classification of sequences

[Leroy Quet]

I do agree that, if the bandwidth is available so that nothing more 
important is excluded, the more information given about any sequence, the 
better.
However, many sequences that originate from one branch of mathematics may 
very well apply to other branchs in ways that mathematicians, for the 
most part or as a whole, do not yet realize.
And I would not want anyone to discard an interesting conjecture about a 
sequence simply because of any additional info in the database leading 
them astray.

(But otherwise, the idea seems like a good one.)

Thanks,
Leroy Quet


>As the Encyclopedia of Integer sequences grows larger and larger,
>it could be usefull to have a kind of (rough) classification 
>of sequences according to the area where they naturally arise
>(something in the spirit of Math classification): eg.
>
>Number theoretic  (Euler functions etc.)
>Enumerative combinatorics (Catalan numbers etc)
>Recreational Maths (palindromic primes etc)
>Elementary algebraic functions
>Elementary transcendental functions 
>etc.
>with the idea that a given sequence, say Catalans, can belong to 
>several classes.
>
>This could be used to accelerate search (by proposing an option for
>searching only over sequences belonging to some class)
>and/or eliminating uninteresting coincidences:
>
>eg. after having computed (often with considerable) pain five terms
>of a sequence giving cardinalities of a sequence of finite sets,
>a result announcing palindromic primes can be generally discarded 
>at once as a coincidence.
>
>Of course, carefull thought, especially in the definition of classes,
>would be needed before starting such a project which surely amounts
>to much work.
>
>Roland Bacher