%I #8 Jul 12 2015 15:44:50
%S 5,8,12,11,11,15,14,14,14,18,18,17,17,17,17,17,21,20,21,21,21,20,20,
%T 20,20,20,20,20,20,20,20,20,24,24,24,24,25,23,23,23,23,23,23,23,23,23,
%U 23,23,23,23,23,23,23,23,23,28,27,27,27,27,27,27,27,28,26,26,26,26,26,26
%N Let b(1)=1/n, b(2)=1, b(k+1)=abs(b(k))-b(k-1)^2; then b(k) is >0 for k>a(n).
%C Conjecture: lim k->infinity (b(k)-k)/log(k) = f(n), constant depending on n. f(n) seems erratic: f(2)=2.9..., f(3)=2.5..., f(4)=3.2..., f(5)=2.25..., f(6)=4.0...
%F a(n) seems to be asymptotic to c*log(n) with c=5.63...
%e If n=4, b(11)<0 and b(k)>0 for any k>11 hence a(4)=11.
%K nonn
%O 1,1
%A _Benoit Cloitre_, Oct 23 2002