This site is supported by donations to The OEIS Foundation.

User:Charles R Greathouse IV/Imp

From OeisWiki
Jump to: navigation, search

The Imp: A fable

There's an imp loose in the OEIS, wrecking havoc. I'm trying to catch it.

It has many tricks, but let me describe its favorite. First, it picks a small number and generates a linear recurrence of that length, picking the values by some random process favoring small numbers. Then it generates initial terms, also at random—at least as many as in the signature unless the signature is very large. It then populates the sequence to about the right number of characters. It's a smart creature, there's no denying that. It understands generating functions, links to the OEIS index, Binet and asymptotic formulas, a number of common programming languages, and so forth. Sometimes it fills those fields in, and other times it leaves them out (perhaps to be sporting, so it's not so easy to detect).

Other times it takes a random sequence in the OEIS and applies either some random transformation or composes it with another sequence. Since there's no connection between the two, the sequence feels disjointed: the imp's just sewing together Things That Were Not Meant To Be. (On a practical level, there are over ways to compose two sequences, and we don't have room for that many... see here for more of my thoughts on the matter.)

At the end of the day, the sequences the imp generates aren't meaningful additions to the OEIS—after all, it's just picking numbers out of a hat. I try very hard to stop the imp from hurting this encyclopedia!

What can you do? I'd ask you to catch it, but I think it's just too quick. If you see one of its sequences, please leave a 'pink box' comment so that the editors can deal with it. More importantly, if you're submitting a sequence, don't let it look like an imp-sequence lest we reject it by mistake!

Call to arms

When you submit a sequence, make sure that it doesn't look like other sequences with the numbers changed around unless there's good reason for it. Always provide some kind of motivating information, when possible, that helps distinguish your sequence from others. This not only lets the Editorial Board know how great your sequence is, but it helps the next person who comes across the sequence. (Isn't it annoying to find just the sequence you were looking for, only to find nothing but a name and the terms you had already calculated?)

Quality over quantity, please. I strive to raise the quality of this encyclopedia, and I hope that you, the reader, will try to keep high the quality of the sequences that you submit.

One rule of thumb that I use is that a sequence should take an hour to submit, between researching, reading, writing programs, and typing it up. If you're talking significantly less, that's a red flag. Are you making this sequence the best it could be? Or if there's really that little to say about it, should it really be a sequence rather than just a comment on some other sequence? Food for thought.

See also

External link