Year-end appeal: Please make a donation to the OEIS Foundation to support ongoing development and maintenance of the OEIS. We are now in our 61st year, we have over 378,000 sequences, and we’ve reached 11,000 citations (which often say “discovered thanks to the OEIS”).
%I #18 Jun 28 2019 21:15:57
%S 1,2,2,3,3,3,5,4,2,4,5,4,5,6,4,5,8,5,6,7,3,5,10,7,3,7,8,5,9,8,4,8,9,6,
%T 5,11,9,4,9,10,6,6,9,10,8,8,11,8,9,13,7,5,13,11,5,13,13,7,9,12,8,9,16,
%U 9,10,13,9,15,11,5,15,13,8,15,12,5,14,13,8
%N a(1) = 1, a(2) = 2, and for any n > 2, a(n) = o(n-1) + o(n-2) where o(k) is the number of occurrences of a(k) among a(1), ..., a(k).
%C The sequence o corresponds to the ordinal transform of the sequence a.
%C This sequence has a taste of Fibonacci, as each pair of consecutive terms gives rise to the next term.
%C This sequence is unbounded:
%C - if the sequence was bounded, say a(n) <= M for any n > 0, then some value in the range 1..M, say t, would appear at least M times among the first M^2 terms, and we would have a value > M after the M-th occurrence of t, a contradiction.
%C This sequence has interesting graphical features (see Links section).
%H Rémy Sigrist, <a href="/A306246/b306246.txt">Table of n, a(n) for n = 1..10000</a>
%H Rémy Sigrist, <a href="/A306246/a306246.png">Density plot of the first 10000000 terms</a>
%e The first terms, alongside o, are:
%e n a(n) o(n)
%e -- ---- ----
%e 1 1 1
%e 2 2 1
%e 3 2 2
%e 4 3 1
%e 5 3 2
%e 6 3 3
%e 7 5 1
%e 8 4 1
%e 9 2 3
%e 10 4 2
%e 11 5 2
%e 12 4 3
%e 13 5 3
%e 14 6 1
%e 15 4 4
%e 16 5 4
%e 17 8 1
%e 18 5 5
%e 19 6 2
%e 20 7 1
%o (PARI) o = vector(18); for (n=1, 79, if (n<=2, v=n, v=pp+p); print1 (v ", "); [pp,p]=[p,o[1+v]++])
%Y See A306251 for the ordinal transform of this sequence.
%K nonn
%O 1,2
%A _Rémy Sigrist_, Jan 31 2019