In this thesis, we show that several natural questions about automatic sequences can be expressed as logical predicates and then decided mechanically. We extend known results in this area to broader classes of sequences (e.g., paperfolding words), introduce new operations that extend the space of possible queries, and show how to process the results. We begin with the fundamental concepts and problems related to automatic sequences, and the corresponding numeration systems. Building on that foundation, we discuss the general logical framework that formalizes the questions we can mechanically answer. We start with a first-order logical theory, and then extend it with additional predicates and operations. Then we explain a slightly different technique that works on a monadic second- order theory, but show that it is ultimately subsumed by an extension of the first-order theory. Next, we give two applications: critical exponent and paperfolding words. In the critical exponent example, we mechanically construct an automaton that describes a set of rational numbers related to a given automatic sequence. Then we give a polynomial-time algorithm to compute the supremum of this rational set, allowing us to compute the critical exponent and many similar quantities. In the paperfolding example, we extend our mechanical procedure to the paperfolding words, an uncountably infinite collection of infinite words. In the following chapter, we address abelian and additive problems on automatic sequences. We give an example of a natural predicate which is provably inexpressible in our first-order theory, and discuss alternate methods for solving abelian and additive problems on automatic sequences. We close with a chapter of open problems, drawn from the earlier chapters.

The Erdős discrepancy problem asks, "Does there exist a sequence t = {t_i}_{1≤i<∞} with each t_i ∈ {-1,1} and a constant c such that |∑_{1≤i≤n} t_{id}| ≤ c for all n,c ∈ ℕ = {1,2,3,...}?" The discrepancy of t equals sup_{n≥1} |∑_{1≤i≤n} t_{id}|. Erdős conjectured in 1957 that no such sequence exists. We examine versions of this problem with fixed values for c and where the values of d are restricted to particular subsets of ℕ. By examining a wide variety of different subsets, we hope to learn more about the original problem. When the values of d are restricted to the set {1,2,4,8,...}, we show that there are exactly two infinite {-1,1} sequences with discrepancy bounded by 1 and an uncountable number of in nite {-1,1} sequences with discrepancy bounded by 2. We also show that the number of {-1,1} sequences of length n with discrepancy bounded by 1 is 2^{s2(n)} where s2(n) is the number of 1s in the binary representation of n. When the values of d are restricted to the set {1,b,b^2,b^3,...} for b > 2, we show there are an uncountable number of infinite sequences with discrepancy bounded by 1. We also give a recurrence for the number of sequences of length n with discrepancy bounded by 1. When the values of d are restricted to the set {1,3,5,7,..} we conjecture that there are exactly 4 in finite sequences with discrepancy bounded by 1 and give some experimental evidence for this conjecture. We give descriptions of the lexicographically least sequences with D-discrepancy c for certain values of D and c as fixed points of morphisms followed by codings. These descriptions demonstrate that these automatic sequences. We introduce the notion of discrepancy-1 maximality and prove that {1,2,4,8,...} and {1,3,5,7,...} are discrepancy-1 maximal while {1,b,b^2,...} is not for b > 2. We conclude with some open questions and directions for future work.

In this thesis, we show that several natural questions about automatic sequences can be expressed as logical predicates and then decided mechanically. We extend known results in this area to broader classes of sequences (e.g., paperfolding words), introduce new operations that extend the space of possible queries, and show how to process the results. We begin with the fundamental concepts and problems related to automatic sequences, and the corresponding numeration systems. Building on that foundation, we discuss the general logical framework that formalizes the questions we can mechanically answer. We start with a first-order logical theory, and then extend it with additional predicates and operations. Then we explain a slightly different technique that works on a monadic second- order theory, but show that it is ultimately subsumed by an extension of the first-order theory. Next, we give two applications: critical exponent and paperfolding words. In the critical exponent example, we mechanically construct an automaton that describes a set of rational numbers related to a given automatic sequence. Then we give a polynomial-time algorithm to compute the supremum of this rational set, allowing us to compute the critical exponent and many similar quantities. In the paperfolding example, we extend our mechanical procedure to the paperfolding words, an uncountably infinite collection of infinite words. In the following chapter, we address abelian and additive problems on automatic sequences. We give an example of a natural predicate which is provably inexpressible in our first-order theory, and discuss alternate methods for solving abelian and additive problems on automatic sequences. We close with a chapter of open problems, drawn from the earlier chapters.

The Erdős discrepancy problem asks, "Does there exist a sequence t = {t_i}_{1≤i<∞} with each t_i ∈ {-1,1} and a constant c such that |∑_{1≤i≤n} t_{id}| ≤ c for all n,c ∈ ℕ = {1,2,3,...}?" The discrepancy of t equals sup_{n≥1} |∑_{1≤i≤n} t_{id}|. Erdős conjectured in 1957 that no such sequence exists. We examine versions of this problem with fixed values for c and where the values of d are restricted to particular subsets of ℕ. By examining a wide variety of different subsets, we hope to learn more about the original problem. When the values of d are restricted to the set {1,2,4,8,...}, we show that there are exactly two infinite {-1,1} sequences with discrepancy bounded by 1 and an uncountable number of in nite {-1,1} sequences with discrepancy bounded by 2. We also show that the number of {-1,1} sequences of length n with discrepancy bounded by 1 is 2^{s2(n)} where s2(n) is the number of 1s in the binary representation of n. When the values of d are restricted to the set {1,b,b^2,b^3,...} for b > 2, we show there are an uncountable number of infinite sequences with discrepancy bounded by 1. We also give a recurrence for the number of sequences of length n with discrepancy bounded by 1. When the values of d are restricted to the set {1,3,5,7,..} we conjecture that there are exactly 4 in finite sequences with discrepancy bounded by 1 and give some experimental evidence for this conjecture. We give descriptions of the lexicographically least sequences with D-discrepancy c for certain values of D and c as fixed points of morphisms followed by codings. These descriptions demonstrate that these automatic sequences. We introduce the notion of discrepancy-1 maximality and prove that {1,2,4,8,...} and {1,3,5,7,...} are discrepancy-1 maximal while {1,b,b^2,...} is not for b > 2. We conclude with some open questions and directions for future work.

In 1912 Axel Thue sparked the study of combinatorics on words when he showed that the Thue-Morse sequence contains no overlaps, that is, factors of the form ayaya. Since then many interesting properties of sequences began to be discovered and studied. In this thesis, we consider a class of infinite sequences generated by automata, called the k-automatic sequences. In particular, we present a logical theory in which many properties of k-automatic sequences can be expressed as predicates and we show that such predicates are decidable. Our main contribution is the implementation of a theorem prover capable of practically characterizing many commonly sought-after properties of k-automatic sequences. We showcase a panoply of results achieved using our method. We give new explicit descriptions of the recurrence and appearance functions of a list of well-known k-automatic sequences. We define a related function, called the condensation function, and give explicit descriptions for it as well. We re-affirm known results on the critical exponent of some sequences and determine it for others where it was previously unknown. On the more theoretical side, we show that the subword complexity p(n) of k-automatic sequences is k-synchronized, i.e., the language of pairs (n, p(n)) (expressed in base k) is accepted by an automaton. Furthermore, we prove that the Lyndon factorization of k-automatic sequences is also k-automatic and explicitly compute the factorization for several sequences. Finally, we show that while the number of unbordered factors of length n is not k-synchronized, it is k-regular.

In 1912 Axel Thue sparked the study of combinatorics on words when he showed that the Thue-Morse sequence contains no overlaps, that is, factors of the form ayaya. Since then many interesting properties of sequences began to be discovered and studied. In this thesis, we consider a class of infinite sequences generated by automata, called the k-automatic sequences. In particular, we present a logical theory in which many properties of k-automatic sequences can be expressed as predicates and we show that such predicates are decidable. Our main contribution is the implementation of a theorem prover capable of practically characterizing many commonly sought-after properties of k-automatic sequences. We showcase a panoply of results achieved using our method. We give new explicit descriptions of the recurrence and appearance functions of a list of well-known k-automatic sequences. We define a related function, called the condensation function, and give explicit descriptions for it as well. We re-affirm known results on the critical exponent of some sequences and determine it for others where it was previously unknown. On the more theoretical side, we show that the subword complexity p(n) of k-automatic sequences is k-synchronized, i.e., the language of pairs (n, p(n)) (expressed in base k) is accepted by an automaton. Furthermore, we prove that the Lyndon factorization of k-automatic sequences is also k-automatic and explicitly compute the factorization for several sequences. Finally, we show that while the number of unbordered factors of length n is not k-synchronized, it is k-regular.

Cette thèse se situe à l'interface de trois grands domaines : la combinatoire des mots, la théorie des automates et la théorie des nombres. Plus précisément, nous montrons comment des outils provenant de la combinatoire des mots et de la théorie des automates interviennent dans l'étude de problèmes arithmétiques concernant les séries formelles à coefficients dans un corps fini.Le point de départ de cette thèse est un célèbre théorème de Christol qui caractérise les séries de Laurent algébriques sur le corps F_q(T), l'entier q désignant une puissance d'un nombre premier p, en termes d'automates finis et dont l'énoncé est : « Une série de Laurent à coefficients dans le corps fini F_q est algébrique si et seulement si la suite de ses coefficients est engendrée par un p-automate fini ». Ce résultat, qui révèle dans un certain sens la simplicité de ces séries de Laurent, a donné naissance à des travaux importants parmi lesquels de nombreuses applications et généralisations.L'objet principal de cette thèse est, dans un premier temps, d'exploiter la simplicité de séries de Laurent algébriques à coefficients dans un corps fini afin d'obtenir des résultats diophantiens, puis d'essayer d'étendre cette étude à des fonctions transcendantes arithmétiquement intéressantes. Nous nous concentrons tout d'abord sur une classe de séries de Laurent algébriques particulières qui généralisent la fameuse cubique de Baum et Sweet. Le résultat principal obtenu pour ces dernières est une description explicite de leur développement en fraction continue, généralisant ainsi certains travaux de Mills et Robbins. Rappelons que le développement en fraction continue permet généralement d'obtenir des informations très précises sur l'approximation rationnelle ; les meilleures approximations étant obtenues directement à partir de la suite des quotients partiels. Malheureusement, il est souvent très difficile d'obtenir le développement en fraction continue d'une série de Laurent algébrique, que celle-ci soit donné par une équation algébrique ou par son développement en série de Laurent. La deuxième étude que nous présentons dans cette thèse fournit une information diophantienne à priori moins précise que la description du développement en fraction continue, mais qui a le mérite de concerner toutes les séries de Laurent algébriques (à coefficients dans un corps fini). L'idée principale est d'utiliser l'automaticité de la suite des coefficients de ces séries de Laurent afin d'obtenir une borne générale pour leur exposant d'irrationalité. Malgré la généralité de ce résultat, la borne obtenue n'est pas toujours satisfaisante. Dans certains cas, elle peut s'avérer plus mauvaise que celle provenant de l'inégalité de Mahler. Cependant, dans de nombreuses situations, il est possible d'utiliser notre approche pour fournir, au mieux, la valeur exacte de l'exposant d'irrationalité, sinon des encadrements très précis de ce dernier.Dans un dernier travail nous nous plaçons dans un cadre plus général que celui des séries de Laurent algébriques, à savoir celui des séries de Laurent dont la suite des coefficients a une « basse complexité ». Nous montrons que cet ensemble englobe quelques fonctions remarquables, comme les séries algébriques et l'inverse de l'analogue du nombre \pi dans le module de Carlitz. Il possède, par ailleurs, des propriétés de stabilité intéressantes : entre autres, il s'agit d'un espace vectoriel sur le corps des fractions rationnelles à coefficients dans un corps fini (ce qui, d'un point de vue arithmétique, fournit un critère d'indépendance linéaire), il est de plus laissé invariant par diverses opérations classiques comme le produit de Hadamard / This thesis looks at the interplay of three important domains: combinatorics on words, theory of finite-state automata and number theory. More precisely, we show how tools coming from combinatorics on words and theory of finite-state automata intervene in the study of arithmetical problems concerning the Laurent series with coefficients in a finite field.The starting point of this thesis is a famous theorem of Christol which characterizes algebraic Laurent series over the field F_q(T), q being a power of the prime number p, in terms of finite-state automata and whose statement is the following : “A Laurent series with coefficients in a finite field F_q is algebraic over F_q(T) if and only if the sequence of its coefficients is p-automatic”.This result, which reveals, somehow, the simplicity of these Laurent series, has given rise to important works including numerous applications and generalizations. The theory of finite-state automata and the combinatorics on words naturally occur in number theory and, sometimes, prove themselves to be indispensable in establishing certain important results in this domain.The main purpose of this thesis is, foremost, to exploit the simplicity of the algebraic Laurent series with coefficients in a finite field in order to obtain some Diophantine results, then to try to extend this study to some interesting transcendental functions. First, we focus on a particular set of algebraic Laurent series that generalize the famous cubic introduced by Baum and Sweet. The main result we obtain concerning these Laurent series gives the explicit description of its continued fraction expansion, generalizing therefore some articles of Mills and Robbins.Unfortunately, it is often very difficult to find the continued fraction representation of a Laurent series, whether it is given by an algebraic equation or by its Laurent series expansion. The second study that we present in this thesis provides a Diophantine information which, although a priori less complete than the continued fraction expansion, has the advantage to characterize any algebraic Laurent series. The main idea is to use some the automaticity of the sequence of coefficients of these Laurent series in order to obtain a general bound for their irrationality exponent. In the last part of this thesis we focus on a more general class of Laurent series, namely the one of Laurent series of “low” complexity. We prove that this set includes some interesting functions, as for example the algebraic series or the inverse of the analogue of the real number \pi. We also show that this set satisfy some nice closure properties : in particular, it is a vector space over the field over F_q(T).

Cette thèse est composée de deux parties : l’une traite des propriétés combinatoires de mots infinis et l’autre des problèmes de colorations des graphes.La première partie du manuscrit concerne les structures régulières dans les mots apériodiques infinis, à savoir les sous-séquences arithmétiques et les premiers retours complets.Nous étudions la fonction qui donne la longueur maximale d’une sous-séquence arithmétique monochromatique (une progression arithmétique) en fonction de la différence commune d pour une famille de mots morphiques uniformes, qui inclut le mot de Thue-Morse. Nous obtenons la limite supérieure explicite du taux de croissance de la fonction et des emplacements des progressions arithmétiques de longueurs maximales et de différences d. Pour étudier des sous-séquences arithmétiques périodiques dans des mots infinis, nous définissons la notion d'indice arithmétique et obtenons des bornes supérieures et inférieures sur le taux de croissance de la fonction donnant l’indice arithmétique dans la même famille de mots.Dans la même veine, une autre question concerne l’étude de deux nouvelles fonctions de complexité de mots infinis basées sur les notions de mots ouverts et fermés. Nous dérivons des formules explicites pour les fonctions de complexité ouverte et fermée pour un mot d'Arnoux-Rauzy sur un alphabet de cardinalité finie.La seconde partie de la thèse traite des colorations parfaites (des partitions équitables) de graphes infinis de degré borné. Nous étudions les graphes de Caley de groupes additifs infinis avec un ensemble de générateurs fixé. Nous considérons le cas où l'ensemble des générateurs est composé d'entiers de l'intervalle [-n, n], et le cas où les générateurs sont des entiers impairs de [-2n-1, 2n+1], où n est un entier positif. Pour les deux familles de graphes, nous obtenons une caractérisation complète des colorations parfaites à deux couleurs / The content of the thesis is comprised of two parts: one deals with combinatorial properties of infinite words and the other with graph coloring problems.The first main part of the manuscript concerns regular structures in infinite aperiodic words, such as arithmetic subsequences and complete first returns.We study the function that outputs the maximal length of a monochromatic arithmetic subsequence (an arithmetic progression) as a function of the common difference d for a family of uniform morphic words, which includes the Thue-Morse word. We obtain the explicit upper bound on the rate of growth of the function and locations of arithmetic progressions of maximal lengths and difference d. To study periodic arithmetic subsequences in infinite words we define the notion of an arithmetic index and obtain upper and lower bounds on the rate of growth of the function of arithmetic index in the same family of words.Another topic in this direction involves the study of two new complexity functions of infinite words based on the notions of open and closed words. We derive explicit formulae for the open and closed complexity functions for an Arnoux-Rauzy word over an alphabet of finite cardinality.The second main part of the thesis deals with perfect colorings (a.k.a. equitable partitions) of infinite graphs of bounded degree. We study Caley graphs of infinite additive groups with a prescribed set of generators. We consider the case when the set of generators is composed of integers from the interval [-n,n], and the case when the generators are odd integers from [-2n-1,2n+1], where n is a positive integer. For both families of graphs, we obtain a complete characterization of perfect 2-colorings

La présente thèse a été fortement influencée par deux conjectures, l'une de Gelfond et l'autre de Sarnak.En 1968, Gelfond a prouvé que la somme des chiffres modulo m est asymtotiquement équirépartie dans des progressions arithmétiques, et il a formulé trois problèmes nouveaux.Le deuxième et le troisième problèmes traitent des sommes des chiffres pour les nombres premiers et les suites polynomiales.En ce qui concerne les nombres premiers et les carrés, Mauduit et Rivat ont résolu ces problèmes en 2010 et 2009, respectivement.Drmota, Mauduit et Rivat ont réussi généraliser le résultat concernant la suite des sommes des chiffres des carrés.Ils ont démontré que chaque bloc apparaît asymptotiquement avec la même fréquence.Selon la conjecture de Sarnak, il n'y a pas de corrélation entre la fonction de Möbius et des fonctions simples.La présente thèse traite de la répartition de suites automatiques le long de sous-suites particulières ainsi que d'autres propriétés de suites automatiques.Selon l'un des résultats principaux du présent travail, toutes les suites automatiques vérifient la conjecture de Sarnak.Moyennant une approche légèrement modifiée, nous traitons également la répartition de suites automatiques le long de la suite des nombres premiers.Dans le cadre du traitement de suites automatiques générales, nous avons mis au point une nouvelle structure destinée aux automates finisdéterministes ouvrant une vision nouvelle pour les automates et/ou les suites automatiques.Nous étendons les résultat de Drmota, Mauduit et Rivat concernant les suites digitales.Cette approche peut également être considérée comme une généralisation du troisième problème de Gelfond. / The present dissertation was inspired by two conjectures, one by Gelfond and one of Sarnak.In 1968 Gelfond proved that the sum of digits modulo m is asymptotically equally distributed along arithmetic progressions.Furthermore, he stated three problems which are nowadays called Gelfond problems.The second and third questions are concerned with the sum of digits of prime numbers and polynomial subsequences.Mauduit and Rivat were able to solve these problems for primes and squares in 2010 and 2009 respectively.Drmota, Mauduit and Rivat generalized the result concerning the sequence of the sum of digits of squares.They showed that each block appears asymptotically equally frequently.Sarnak conjectured in 2010 that the Mobius function does not correlate with deterministic functions.This dissertation deals with the distribution of automatic sequences along special subsequences and other properties of automatic sequences.A main result of this thesis is that all automatic sequences satisfy the Sarnak conjecture.Through a slightly modified approach, we also deal with the distribution of automatic sequences along the subsequence of primes.In the course of the treatment of general automatic sequences, a new structure for deterministic finite automata is developed,which allows a new view for automata or automatic sequences.We extend the result of Drmota, Mauduit and Rivat to digital sequences.This is also a generalization of the third Gelfond problem.

Cette thèse est composée d'une partie sur la conjecture des familles stables par unions et de quatre autres chapitres consacrés aux sujets liés aux suites automatiques. Dans la première partie, on donne une condition suffisante pour qu'une version affaiblie de la conjecture soit vraie. On donne aussi un majorant de la fréquence maximale minimale dans une famille de taille $n$. Dans Chapitre 3 on démontre que la formule d'extraction des coefficients des séries algébriques connue pour les corps à caractéristique $0$ est une conséquence d'un théorème de Furstenberg qui permet d'écrire certaines séries algébriques comme les diagonales des fractions rationnelles à deux variables. Comme ce théorème est valide pour tous les corps, la formule l'est aussi. Dans Chapitre 4 on donne une généralisation des résultats de J.-P. Allouche et J. Shallit concernant certains produits infinis et les fonctions qui comptent le nombre d'occurrences d'un facteur dans l'expansion en base $B$ de $n$. Dans Chapitre 5 on donne une construction explicite d'un mot infini avec complexité en facteur de $\Theta(n^t)$ avec la valuation $p$-adique. Dans Chapitre 6 on donne une nouvelle démonstration de la transcendance de la série formelle $L(1,\chi_s)/\Pi$, où $L$ est un analogue des fonctions $L$ de Dirichlet en caractéristique finie défini par D. Goss et $\Pi$ l'analogue de $\pi$ défini par L. Carlitz. / This thesis comprises one part concerning the union-closed sets conjecture and four other chapters dedicated to subjects related to automatic sequences. In the first part, we give a sufficient condition for a weaker version of the conjecture ($\varepsilon$-union closed sets conjecture) to hold. We also give an upper bound of the minimal maximal frequency for a family of size $n$. In Chapter 3 we prove that the coefficient extraction formula for algebraic series known for fields of characteristic $0$ is a consequence of a theorem of Furstenberg that says certains algebraic series can be written as the diagonals of a rational fractions in two variables. As the theorem is true for all fields, so is the formula. In Chapter 4 we give a generalization of the result of J.-P. Allouche and J. Shallit concerning certain infinite products and block-counting functions. In Chapter 5 we give an explicit construction based on $p$-adic valuation of an infinite word with subword complexity $\Theta(n^t)$. In Chapter 6 we give a new proof of the transcendence of the power series $L(1,\chi_s)/\Pi$, where $L$ is an analogue in positive characteristics of Dirichlet $L$ functions defined by D. Goss and $\Pi$ the analogue of $\pi$ defined by L. Carlitz.