OFFSET
1,1
COMMENTS
It appears that a(n+1) - a(n) is in {4,5} for n >= 1.
Lim_{n->infinity} a(n)/n = Pi^2/2 = 4.9348022..., but lim_{n->infinity} (a(n+1) - a(n)) does not exist; Pi^2/2 is only a mean value of these differences. - Vaclav Kotesovec, Oct 09 2014
LINKS
Clark Kimberling and Vaclav Kotesovec, Table of n, a(n) for n = 1..10000 (first 500 terms from Clark Kimberling)
FORMULA
a(n) ~ n*Pi^2/2 = n*A102753. - Vaclav Kotesovec, Oct 09 2014
EXAMPLE
Taking n = 2, we have cos(Pi/9) + 1/(18) = 0.99524... < 1 < 1.0010565... = cos(Pi/10) + 1/(20), so that a(2) = 10, as corroborated for n = 2 in the following list of approximations:
n ... cos(Pi/a(n)) + 1/(n*a(n))
1 ... 1.009016994
2 ... 1.001056516
3 ... 1.000369823
4 ... 1.000188341
5 ... 1.000114701
6 ... 1.000077451
MATHEMATICA
z = 800; f[n_] := f[n] = Select[Range[z], Cos[Pi/#] + 1/(#*n) > 1 &, 1];
u = Flatten[Table[f[n], {n, 1, z}]] (* A248359 *)
Table[Floor[1/(1 - Cos[Pi/n])], {n, 1, z/10}] (* A248360 *)
Table[k=1; While[Cos[Pi/k]+1/(k*n)<=1, k++]; k, {n, 1, 100}] (* Vaclav Kotesovec, Oct 09 2014 *)
CROSSREFS
KEYWORD
nonn,easy
AUTHOR
Clark Kimberling, Oct 07 2014
STATUS
approved