My suggestions rant for a possible improvement to OEIS
creigh at o2online.de
creigh at o2online.de
Fri Jan 7 14:15:22 CET 2005
As far as my own personal work with OEIS is concerned, the following two
additions to the main look-up page would save me a lot of time
over the long run (I don't claim to be a genious, etc. for thinking these
ideas up; I presume many others have had similar ideas).
First addition: a radio button entitled "include bisections" or "include bisections
in search".
This means: if a user enters the sequence a(0), a(1), a(2), a(3), a(4), a(5),
...
the search engine runs a total of three times: once for the sequence
a(0), a(1), a(2), a(3), a(4), a(5), ...
once for the sequence
a(0), a(2), a(4), a(6), ...
and once for the sequence
then it lumps all the matches together into one big "result" (perhaps with
a small line header: "even indexed terms", "odd indexed terms".
Note: the hints page gives the following advice:
********
Clean up your sequence first.
If every other term is zero (e.g. 1 0 2 0 4 0 8 0 16 0 32 0 64 0 ...),
omit these zeros, and look up 1 2 4 8 16 32 64 instead.
*********
However, in my opinion information will often get overlooked because there is
no such option. Ex.:
1famrokseq[+ .5'i + .5'j + .5i' + .5k' + 'ii']: 1, 3, 12, 60, 360, 2520, 20160,
181440, 1814400,
http://www.research.att.com/projects/OEIS?Anum=A001710
Order of alternating group A_n, or number of even permutations of n
letters.
with no mention of (3, 60, 2520,...):
http://www.research.att.com/projects/OEIS?Anum=A085990
Number of topological types of polygons with 2n different sides.
or of (1, 12, 360, 20160, ...)
http://www.research.att.com/projects/OEIS?Anum=A002674
Right side of the binomial sum n-> sum( (-1)^i * (n-i)^(2*n) *
binomial(2*n, i), i=0..n)
Btw. I only "knew" to search for the sequences because, for ex.
1lesrokseq[+ .5'i + .5'j + .5i' + .5k' + 'ii']: : 1, 0, 12, 0, 360, 0,
20160, 0, 1814400,
(forgive me if I've actually overlooked some cross references relating the
above sequence in the database- this should be seen as merely a particular
example.)
If I remember correctly, I believe Ralf Stephan mentioned not too long ago
something like: "Superseeker should include bisections... really." Perhaps
this has already been implemented but, to me, if you get to the stage where
you need Superseeker to look up bisections, you are practically attempting
to catch a butterfly by pinning it down
with a sledgehammer.
Second addition (??):
If a user looks up a three sequences in a row with the reply
"I am sorry, but the terms
a(0), a(1), ...,
do not match anything in the table."
The fourth reply changes to "I am very sorry, but the terms
a(0), a(1), ..., do not match anything in the table. Would you care to
have a look one of our nice/hot sequences instead?" (otherwise it can get
depressing over time)
Sincerely (o.k., not quite on the last one),
Creighton
More information about the SeqFan
mailing list