Proof of Dejean’s conjecture for alphabets
نویسندگان
چکیده
Axel Thue proved that overlapping factors could be avoided in arbitrarily long words on a two-letter alphabet while, on the same alphabet, square factors always occur in words longer than 3. Françoise Dejean stated an analogous result for three-letter alphabets : every long enough word has a factor which is a fractional power with an exponent at least 7/4 and there exist arbitrary long words in which no factor is a fractional power with an exponent strictly greater than 7/4. The number 7/4 is called the repetition threshold of the three-letter alphabets. Thereafter, she proposed the following conjecture : the repetition threshold of the k-letter alphabets is equal to k/(k − 1) except in the particular cases k = 3, where this threshold is 7/4, and k = 4, where it is 7/5. For k = 4, this conjecture was proved by Jean Jacques Pansiot. In this paper, we give a computer-aided proof of Dejean’s conjecture for several other values : 5, 6, 7, 8, 9, 10, 11. Résumé. Axel Thue a montré que l’on peut éviter les chevauchements dans des mots arbitrairement grands sur un alphabet à deux lettres alors que l’on ne peut éviter les carrés. On peut ainsi dire que l’exposant 2 est le seuil de répétition des alphabets à deux lettres. Françoise Dejean a démontré que le seuil de répétition des alphabets à trois lettres est 7/4 ; elle a alors proposé la conjecture suivante : le seuil sk de répétition des alphabets à k lettres est égal à k/(k − 1) sauf pour les cas particuliers 3 et 4 (s3 = 7/4, s4 = 7/5). Jean Jacques Pansiot a établi le résultat pour k = 4. Nous donnons ici une démonstration assistée par ordinateur de cette conjecture pour k = 5, 6, 7, 8, 9, 10 et 11.
منابع مشابه
A proof of Dejean's conjecture
We prove Dejean’s conjecture. Specifically, we show that Dejean’s conjecture holds for the last remaining open values of n, namely 15 ≤ n ≤ 26.
متن کاملA short proof of the maximum conjecture in CR dimension one
In this paper and by means of the extant results in the Tanaka theory, we present a very short proof in the specific case of CR dimension one for Beloshapka's maximum conjecture. Accordingly, we prove that each totally nondegenerate model of CR dimension one and length >= 3 has rigidity. As a result, we observe that the group of CR automorphisms associated with each of such models contains onl...
متن کاملOn the D0L Repetition Threshold
The repetition threshold is a measure of the extent to which there need to be consecutive (partial) repetitions of finite words within infinite words over alphabets of various sizes. Dejean’s Conjecture, which has been recently proven, provides this threshold for all alphabet sizes. Motivated by a question of Krieger, we deal here with the analogous threshold when the infinite word is restricte...
متن کاملPartial proof of Graham Higman's conjecture related to coset diagrams
Graham Higman has defined coset diagrams for PSL(2,ℤ). These diagrams are composed of fragments, and the fragments are further composed of two or more circuits. Q. Mushtaq has proved in 1983 that existence of a certain fragment γ of a coset diagram in a coset diagram is a polynomial f in ℤ[z]. Higman has conjectured that, the polynomials related to the fragments are monic and for a fixed degree...
متن کاملOn the Closed-Form Solution of a Nonlinear Difference Equation and Another Proof to Sroysang’s Conjecture
The purpose of this paper is twofold. First we derive theoretically, using appropriate transformation on x(n), the closed-form solution of the nonlinear difference equation x(n+1) = 1/(±1 + x(n)), n ∈ N_0. The form of solution of this equation, however, was first obtained in [10] but through induction principle. Then, with the solution of the above equation at hand, we prove a case ...
متن کامل