login

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”).

A006340
An "eta-sequence": [ (n+1)*tau + 1/2 ] - [ n*tau + 1/2 ], tau = (1 + sqrt(5))/2.
(Formerly M0100)
5
2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 2, 2, 1, 2, 1
OFFSET
0,1
COMMENTS
Equals its own "derivative", which is formed by counting the strings of 1's that lie between 2's.
Conjecture: A006340 = continued fraction expansion of (2.729967741... = sup{f(n,1)}), where f(1,x) = x + 1 and thereafter f(n,x) = x + 1 if n is in the lower Wythoff sequence (A000201), else f(n,x) = 1/x. The first 12 values of f(n,1) are given in Example at A245216. - Clark Kimberling, Jul 14 2014
From Michel Dekking, Mar 05 2018: (Start)
The description of this sequence is not correct, since the derivative of a equals
a' = 1,2,2,1,2,2,1,2,1,2,2,1,2,1,2,2,...
The claim by Hofstadter in formula (4) in the 1977 letter to Sloane is also not correct, since the second derivative of a is equal to
a'' = 2,2,1,2,1,2,2,1,2,1,2,2,1,...
so a is not equal to its own second derivative.
Nevertheless, this sequence has a self-similarity property: if one replaces every chunk 212 with 1 and every chunk 21212 with 2, then one obtains back the original sequence. In other words, (a(n)) is the unique fixed point of the morphism sigma given by sigma: 1->212, 2->21212.
This can be proved following the ideas of Chapter 2 in Lothaire's book and Section 4 of my paper "Substitution invariant Sturmian words and binary trees".
To comply with these references change the alphabet to {0,1}. This changes sigma into the morphism 0->101, 1->10101.
The fractional part {tau} of tau is larger than 1/2; as it is convenient to have it smaller than 1/2 we change to beta = 1-tau = (3-sqrt(5))/2.
This changes the morphism 0->101, 1->10101 to its mirror image psi given by 0->01010, 1->010.
Let psi_1 and psi_2 be the elementary Sturmian morphisms given by
psi_1(0)=01 , psi_1(1)=1, psi_2(0)=10, psi_2(1)=0.
Then psi = psi_2^2 psi_1.
This already shows that psi generates a Sturmian sequence with certain parameters alpha and rho: s(alpha,rho) = ([(n+1)*alpha+rho]-[n*alpha+rho]).
Since psi is the composition psi_2^2psi_1, the parameters of s(alpha,rho) are given by the composition T:=T_2^2T_1 of the fractional linear maps
T_1(x,y) = ((1-x)/(2-x),(1-y)/(2-x)),
T_2(x,y) = ((1-x)/(2-x), (2-x-y)/(2-x)).
Since one can verify that T(beta,1/2)=(beta,1/2), it follows that
alpha = beta, and rho = 1/2.
(End)
REFERENCES
N. J. A. Sloane and Simon Plouffe, The Encyclopedia of Integer Sequences, Academic Press, 1995 (includes this sequence).
LINKS
Michel Dekking, Substitution invariant Sturmian words and binary trees, arXiv:1705.08607 [math.CO], (2017).
Michel Dekking, Substitution invariant Sturmian words and binary trees, Integers, Electronic Journal of Combinatorial Number Theory 18A (2018), #A17.
D. R. Hofstadter, Eta-Lore [Cached copy, with permission]
D. R. Hofstadter, Pi-Mu Sequences [Cached copy, with permission]
D. R. Hofstadter and N. J. A. Sloane, Correspondence, 1977 and 1991
M. Lothaire, Algebraic combinatorics on words, Cambridge University Press. Online publication date: April 2013; Print publication year: 2002.
MATHEMATICA
Differences[ Table[ Round[ GoldenRatio*n], {n, 0, 93}]] (* Jean-François Alcover, Aug 13 2012 *)
PROG
(PARI) rt(n) = my(tau=(1 + sqrt(5))/2); round(tau*n)
a(n) = rt(n+1)-rt(n) \\ Felix Fröhlich, Aug 26 2018
CROSSREFS
Differs from A014675 in many places. Cf. A245216.
Sequence in context: A172155 A080573 A186440 * A270641 A076371 A175044
KEYWORD
nonn,easy,nice
AUTHOR
D. R. Hofstadter, Jul 15 1977
EXTENSIONS
Extended by N. J. A. Sloane, Nov 07 2001
STATUS
approved