Möbius number systems based on interval covers


 Elinor Underwood
 2 years ago
 Views:
Transcription
1 Möbius number systems based on interval covers Petr Kůrka and Alexandr Kazda Center for Theoretical Study, Academy of Sciences and Charles University in Prague, Jilská, CZ Praha, Czechia Faculty of Mathematics and Physics of the Charles University in Prague, Sokolovská 83, CZ8675 Praha 8, Czechia Abstract. Given a finite alphabet A, a system of real orientationpreserving Möbius transformations (F a : R R) a A, a subshift Σ A N, and an interval cover W = {W a : a A} of R, we consider the expansion subshift Σ W Σ of all expansions of real numbers with respect to W. If the expansion quotient Q(Σ, W) is greater than then there exists a continuous and surjective symbolic mapping Φ : Σ W R and we say that (F,Σ W ) is a Möbius number system. We apply our theory to the system of binary continued fractions which is a combination of the binary signed system with the continued fractions, and to the binary square system whose transformations have stable fixed points,, and. AMS classification scheme numbers: 37B, 37F99. Introduction Möbius number systems introduced in Kůrka [5] and [8] are based on iterative systems of Möbius transformations (F a : R R) a A indexed by a finite alphabet A, acting on the extended real line R = R { }. The convergence space X F consists of all infinite words u A N such that F u F un (z) converge to a real number Φ(u) R whenever z is a complex number with positive imaginary part. Since the symbolic representation Φ : X F R is usually not continuous, we search for a subshift Σ X F such that Φ : Σ R is continuous and surjective. In this case we say that (F,Σ) is a Möbius number system. In Kůrka [8] we have developed a theory of Möbius number systems with sofic subshifts. In the present paper we consider expansion subshifts which are obtained when we expand real numbers x R into symbolic sequences u A N with Φ(u) = x. The expansion procedure uses an interval cover W = {W a : a A} of R and tests conditions x W u, Fu (x) W u, Fu u (x) W u2 etc. We limit the expansions to a predefined subshift Σ A N which may be the full shift or the subshift which forbids identities u with F u F un = Id. The expansion subshift Σ W consists of all expansions which belong to Σ and respect W. We define the expansion quotient Q(Σ,W) and show that if Q(Σ,W) >, then (F,Σ W ) is a Möbius number system (Corollary 2). There is a converse (Theorem 3) which uses an additional condition on images of cylinders of finite words. While the expansion subshifts Σ W are in general not sofic, the arithmetical algorithms which work with them are simpler than in the sofic case. If all transformations F a have rational entries, and if their expansion intervals W a overlap and have rational endpoints, then there exist arithmetical algorithms which generalize
2 the arithmetical algorithms for exact real computation described in Gosper [], Vuillemin [3], or Potts et al. []. We present algorithms for expansions of rational or algebraic numbers and for computation of rational functions and fractional bilinear functions such as the sum and product. Using the continued fractions of Gauss (see Wall [4]), algorithms for many transcendental functions can be obtained similarly as in Potts []. We apply our theory to several examples. We show that both the binary signed system and the system of continued fractions can be conceived as Möbius number systems. The system of binary continued fractions considered in Kůrka [8] is a combination of these two classical systems and corresponds to the continued logarithms of Gosper []. Then we treat polygonal number systems considered in Kůrka [5], in particular the binary square system consisting of transformations with stable fixed points,, and. Some of the present results have appeared in Kůrka [7] and Kazda [3]. 2. Möbius transformations The extended real line R = R { } can be regarded as a projective space, i.e., the space of onedimensional subspaces of the twodimensional vector space. On R we have homogenous coordinates x = (x,x ) R 2 \{(,)} with equality x = y iff x y = x y. We regard x R as a column vector, and write it usually as x = x /x, for example = /. The open and closed intervals with endpoints a,b R are defined by (a,b) = {x R : (a x a x )(x b x b )(b a b a ) > }, [a,b] = {x R : (a x a x )(x b x b )(b a b a ) }. For distinct a,b R, we have (a,b) = {x R : a < x or x < b} { } if a > b and (a,b) = {x R : a < x < b} if a < b. Moreover, (a,a) = and [a,a] = R. A real orientationpreserving Möbius transformation (MT) is a selfmap of R of the form M (a,b,c,d) (x) = (ax+b)/(cx+d) = (ax +bx )/(cx +dx ), where a,b,c,d R and ad bc >. An MTacts alsoon the complex spherec = C { } and on the upper halfplane U = {z C : I(z) > }: if z U then M(z) U. The map d(z) = (iz +)/(z +i) maps U conformally to the unit disc D = {z C : z < } and R to the unit circle D = {z C : z = }. Define the circle distance on R by (x,y) = 2arcsin x y x y (x 2 +x 2 )(y2 +y2 ) which is the length of the shortest arc joining d(x) and d(y) in D. The length of a closed interval B r (a) = {x R : (x,a) r} is B r (a) = min{2r,2π}. The length J of a set J R is the length of the shortest interval which contains J. For x R and < ε < 2π denote by x ε,x ε R the unique points for which [x,x ε] = ε and [x ε,x] = ε. On the closed disc D := D D we get disc Möbius transformations M : D D defined by M (a,b,c,d) (z) = d M (a,b,c,d) d (z) = αz +β βz +α, where α = (a + d) + (b c)i, β = (b + c) + (a d)i. Define the norm of a Möbius transformation M = M (a,b,c,d) by M := (a 2 + b 2 + c 2 + d 2 )/(ad bc). We have 2
3 M 2, and M = 2 iff M is a rotation, i.e., if M(z) = R α (z) = e iα z for some α R. Definition The circle derivation M : R (, ) and the expansion interval V(M) of M = M (a,b,c,d) are defined by M (x) := M (d(x)) = V(M) := {x R : (M ) (x) > }. (ad bc)(x 2 +x 2 ) (ax +bx ) 2 +(cx +dx ) 2, We have M (x) = lim y x (M(y),M(x))/ (x,y), so the circle derivation is the derivation with respect to, and (MN) (x) = M (N(x)) N (x). If M is a rotation, then V(M) =, otherwise V(M) is a nonempty open interval. The expansion interval V(M)isrelatedtothevalue M() DofthecorrespondingdiscMT.IfV(M) = (l,r), then M() is the middle of the line joining d(l) and d(r). This follows from Lemma 7 of Kazda [2]. Further usefull relations between M, M() and M (x) have been given in Kůrka [8]: M() 2 = ( M 2)/( M +2), 3. Möbius number systems min{m (x) : x R} = 2 ( M M 2 4), max{m (x) : x R} = 2 ( M + M 2 4). For a finite alphabet A denote by A := m Am the set of finite words and by A + := A \ {λ} the set of finite nonempty words. The length of a word u = u...u m A m is u := m. We denote by A N the Cantor space of infinite words u = u u... equipped with metric d(u,v) := 2 k, where k = min{i : u i v i }. We denote by u [i,j) = u i...u j and u [i,j] = u i...u j. We say that v A is a subword of u A A N and write v u, if v = u [i,j) for some i j u, where we put u = for u A N. Given u A n, v A m with n, m >, denote by u.v A N the periodic word with preperiod u and period v defined by (u.v) i = u i for i < n and (u.v) n+km+i = v i for i < m and for all k. Denote by [u] the cylinder {v A N : v [, u ) = u} of u A. The shift map σ : A N A N is defined by σ(u) i = u i+. A subshift is a nonempty setσ A N whichisclosedandσinvariant,i.e.,σ(σ) Σ. ForasubshiftΣthereexists asetd A + offorbidden wordssuchthatσ = S(D) := {x A N : u x,u D}. We denote by S = A N the full shift with the alphabet A. A subshift is uniquely determined by its language L(Σ) := {u A : x Σ,u x}. We denote by L n (Σ) = L(Σ) A n. A subshift is of finite type (SFT), if the set D of forbidden words is finite. If D A n, then we say that the order of Σ is at most n. A subshift is sofic, if its language is regular (Lind and Marcus [9]). An iterative system is a continuous map F : A X X, or a family of continuous maps (F u : X X) u A satisfyingf uv = F u F v, andf λ = Id. Itisdeterminedbygenerators(F a : X X) a A. Definition 2 We say that F : A R R, is a Möbius iterative system, if all F a : R R are orientationpreserving Möbius transformations. The convergence space X F A N and the symbolic representation Φ : X F R are defined by X F := {u A N : lim F u n [,n) (i) R}, Φ(u) := lim F u n [,n) (i), 3
4 where i U is the imaginary unit. If Σ X F is a subshift such that Φ : Σ R is continuous and surjective, then we say that (F, Σ) is a Möbius number system. We say that a Möbius number system is redundant, if for every continuous map g : R R there exists a continuous map f : Σ Σ such that Φf = gφ. If u X F then Φ(u) = lim n F u[,n) (z) for every z U (see Kazda [2]). For v A +, u A N we have vu X F iff u X F, and then Φ(vu) = F v (Φ(u)). Since Φ is usually not continuous on X F, we search for a subshift Σ X F such that Φ : Σ R is continuous and surjective, so that we get a symbolic representation of R by a Cantor space Σ. If the system is redundant, then continuous functions g : R R can be lifted to their continuous symbolic extensions f : Σ Σ. The continuity of a function f : Σ Σ is necessary for its computability (see e.g. Weihrauch [5]). We show that the system is redundant if the images of its cylinders overlap (Theorem ). We give some sufficient and some necessary conditions for Φ(u) = x, which will be needed in the proofs. Lemma 3 (Kůrka [8]) Let u A N and x R. Then Φ(u) = x iff there exists c > and a sequence of intervals I m x such that liminf n Fu [,n) (I m ) > c for each m, and lim m I m =. Lemma 4 Let u A N and x R. () If lim n (F u [,n) ) (x) =, then Φ(u) = x. (2) If Φ(u) = x, then for every r < there exists n such that (F u [,n) ) (x) r n. Proof: () Denote z = d(x), M n (z) = F u[,n) (z) = (α n z + β n )/(β n z + α n ), where α n 2 β n 2 =, so(m n ) (x) = ( M n ) (z) = α n β n z 2. Iflim n (M n ) (z) = then lim n α n β n z =. Since α n > for n sufficiently large, there exists < r < such that β n > r for all sufficiently large n. It follows α n lim = z = n β n z = lim β n = lim M n () Φ(u) = x. n α n n (2) Assume by contradiction that there exists r < such that (Fu [,n) ) (x) < r n for all n and choose s < with r < s. Since ln(fa ) (x) is uniformly continuous on R, there exists δ > such that if (x,y) < δ, then (Fa ) (y) < (Fa ) (x) s/r for each a A. Denote by I = (x δ,x δ). For y I denote by y n = Fu [,n) (y), x n = Fu [,n) (x). We prove by induction that (Fu [,n) ) (y) < s n and (y n,x n ) < δ. If the condition holds for all k n, then (F u [,n+) ) (y) = (F u ) (y ) (F u n ) (y n ) < (F u ) (x ) (F u n ) (x n ) (s/r) n+ = (F u [,n+) ) (x) (s/r) n+ s n+. Since F u [,n+) is a contraction on I, we get (x n+,y n+ ) < δ. Since each F u [,n) is a contraction on I, we cannot have Φ(u) = x by Lemma 3. 4
5 4. Interval systems We define the expansion subshift of all expansions of real numbers which are chosen from a predefined subshift Σ and respect an interval cover or almostcover {W a : a A} of R. A natural choice for Σ is either the full shift S = A N or the subshift with forbidden identities D = {u A + : F u = Id}. Definition 5 Let F : A R R be a Möbius iterative system and Σ A N a subshift. () We say that W = {W u : u L(Σ)} is an interval system for F and Σ, if each W u R is a finite union of open intervals, W u = R iff u = λ, and W uv = W u F u (W v ) for each uv L(Σ). (2) We say that W is almostcompatible with F and Σ if W u = {W ua : a A,ua L(Σ)} for each u L(Σ). (3) We say that W is compatible with F and Σ if W u = {W ua : a A,ua L(Σ)} for each u L(Σ). (4) Denote by Σ W := {u Σ : n,w u[,n) } the expansion subshift of Σ and W. (5) Denote by W n (Σ) := {W u : u L n (Σ W )}. (6) Denote by l(w) := sup{l > : ( I R)( I < l a A,I W a )} the Lebesgue number of W. WehaveW u = W u F u (W u ) F u[,2) (W u2 ) F u[,n) (W un )foreachu L n+ (Σ), so W is determined by W = {W a : a A}. In the examples, W a are usually open intervals, but their intersections need not be always intervals. A compatible interval system expresses the process of expansion of real numbers: Given x R, we construct nondeterministically its expansion u Σ W as follows: Choose u with x W u, choose u with Fu (x) W u and u u L(Σ), choose u 2 with Fu u (x) W u2 and u u u 2 L(Σ), etc. Then x W u[,n) for each n. In Theorem 9 we show that if W a = V(F a ) are the expansion intervals of F a (Definition ) and if u is the expansion of x, then Φ(u) = x. If W is almostcompatible with F and Σ then W (Σ) is an almostcover of R, i.e., a W a = R. If W is compatible with F and Σ then W (Σ) is a cover of W λ = R, i.e., a W a = R. If Σ = S = A N is the full shift, then we have equivalencies: W is almostcompatible iff W (S) is an almostcover, and W is compatible iff W (S) is a cover. This is a special case of Proposition 6. Proposition 6 Let Σ be a SFT of order at most n + and let W be an interval system. () If W u = {W ua : a A,ua L(Σ)} holds for all u L(Σ) with u n, then W is almostcompatible with F and Σ. (2) If W u = {W ua : a A,ua L(Σ)} holds for all u L(Σ) with u n, then W is compatible with F and Σ. Proof: () We use the fact that if U,V are finite unions of open intervals, then U V U V. Let v,u A + and u = n. If x W vu, then for some y x either 5
6 (x,y) W vu or (y,x) W vu. In the former case we have ( ) (x,y) W v F v (W u ) W v F v {Wua : a A,ua L(Σ)} = {W v F v (W ua ) : a A,vua L(Σ)} {W v F v (W ua ) : a A,vua L(Σ)} = {W vua : a A,vua L(Σ)}, so x {W vua : a A,vua L(Σ)}. The proof of (2) is analogous. Proposition 7 Let W be almostcompatible with F and Σ. Then each W n (Σ) is an almostcover of R. If W is compatible, then each W n (Σ) is a cover of R. Proof: Given x R there exists u A, y R such that (x,y ) W u. There exist u A, y R with u [,] L(Σ) and (x,y ) W u[,], and we continue in this way till we get (x,y n ) W u[,n), so x W u[,n). Proposition 8 Let W be an interval system for S = A N and assume that for each u A m, a A we have W a F a (W u ) F a (W u ) W a. Then S W is a SFT of order at most m+. Proof: Recall that a subshift Σ has order at most p, if u A N belongs to Σ whenever all subwords of u of length p belong to L(Σ). Assume that u A n+, and for all v u with v = m+ we have W v. Then and W u[n m,n], so W u. W u = W u F u (W u[,m] ) F u[,m] (W u[m+,n] ) = F u (W u[,m] ) F u[,m] (W u[m+,n] ) = F u (W u[,n] ) = = F u[,n m) (W u[n m,n] ) Theorem 9 Let F : A R R be a Möbius iterative system and assume that V = {V a = V(F a ) : a A} (see Definition ) is almostcompatible with S = A N, i.e., a A V a = R. Then (F,S V ) is a Möbius number system and for each u L(S V ) we have Φ([u] S V ) = V u. Proof: Denote by l a,r a the endpoints of V a = (l a,r a ) and by Va ε := (l a ε,r a ε) provided 2ε < V a. Since F a are contractions on Fa (V a ), there exists an increasing continuous function ψ : [,2π) [,2π) such that ψ() =, < ψ(t) < t for t >, and F a (I) ψ( I ) for each a A and I Fa (V a ). Given u S V and m n we have Fu [,m] (V u[,n] ) Fu [,m] F u[,m) (V um ) = Fu m (V um ), so V u[,n] = F u F u (V u[,n] ) ψ( F u (V u[,n] ) ) = ψ( F u F u [,] (V u[,n] ) ) ψ 2 ( F u [,] (V u[,n] ) ) ψ n ( F u [,n] (V u[,n] ) ) ψ n (2π). Since ψ(t) < t and the only fixed point of ψ is zero, we get lim n V u[,n) =, so there exists a unique point x n V u [,n). We show that Φ(u) = x. There exists 6
7 ε > such that whenever Fa ([l a,l a ε]) (V b \Vb ε ) then F a (l a ) {l b,r b } and similarly, if Fa ([r a ε,r a ]) (V b \Vb ε ) then F a (r a ) {l b,r b }. By choosing ε small enough, we can also guarantee that Fa ([l a,l a ε]) V b and Fa ([r a ε,r a ]) V b are intervals (as opposed to unions of two nonempty disjoint intervals) for all a, b. Denote by x n := Fu [,n) (x). Since V u[,n] F u[,n) (V un ) F u[,n) (V un ), we get x V u[,n] F u[,n) (V un ), so x n V un. We have (F u [,n) ) (x) = (F u ) (x ) (F u ) (x ) (F u n ) (x n ), and each factor in this product is at least. If x n Vu ε n for an infinite number of n, then lim n (Fu [,n) ) (x) = and Φ(u) = x by Lemma 4. Assume therefore that there exists n such that x n V un \ Vu ε n for each n n. Assume that x n [l un,l un ε] and observe that x n+ = Fu n (x n ). By the definition of ε, we then have that Fu n (l un ) {l un+,r un+ }. However, if Fu n (l un ) = r un+, then x n V un F un (V un+ ) and therefore x V u[,n+] which is a contradiction. Thus Fu n+ (l un ) = l un+ and x n [l un,l un ε] foralln > n. IfI V a then Fa (I) ψ ( I ). GivenanyintervalI δ = [x,x δ], where δ >, there exists n > n such that F u [,n) (I δ ) max{ V a : a A} ε, so Φ(u) = x by Lemma 3. If x n [r un ε,r un ] we proceed similarly. Therefore in all cases Φ(u) = x and S V X F. We show Φ([v] S V ) = V v for each v L(S V ). If u S V [v], then Φ(u) V v by the preceding proof, so Φ([v] S V ) V v. If x V v, then x n V vn for each n < v. If there are j < k < v with x j = l uj, x k = r uk, then x j V uj F u[j,k) (V uk ), which is a contradiction. Thus either x n r vn for each n < v or x n l vn for each n < v. We define u [v] S V by u n = v n for n < v and for n v by induction. If x k r vk for all k < n then there exists u n such that x n = F u [,n) (x) [l un,r un ). Then u S V [v] and Φ(u) = x so V v Φ([v] S V ). This works also for V λ = R, so Φ : S V R is surjective. Finally, since lim n Φ([u [,n) ] S V ) =, Φ : S V R is continuous. 5. Expansion quotients The condition of Theorem 9 is not necessary for the inclusion of Σ W in X F and much larger interval covers may yield Möbius number systems as well. We obtain a better condition with the concept of the expansion quotient Q(Σ,W) of Σ and W. Definition If W is almostcompatible with F and Σ, then we define q(u) := inf{(fu ) (x) : x W u }, u L(Σ W ) Q n (Σ,W) := min{q(u) : u L n (Σ W )}, n Q(Σ,W) := lim n Qn (Σ,W). For x W uv we have (F uv ) (x) = (F u ) (x) (F v ) (F u (x)) q(u) q(v), and therefore q(uv) q(u) q(v). It follows Q n+m (Σ,W) Q n (Σ,W) Q m (Σ,W), so the limit Q(Σ,W) exists and Q(Σ,W) n Q n (Σ,W) for each n. Theorem Let W be almostcompatible with Σ, and assume that for some n > we have Q n (Σ,W) and no F u with u L n (Σ) is a rotation. Then (F,Σ W ) is a Möbius number system and Φ([u] Σ W ) = W u for each u L(Σ W ). If W is compatible with F and Σ then (F,Σ W ) is redundant. 7
8 Proof: For each u L n (Σ W ) we have W u V(F u ). Consider the alphabet B = L n (Σ W ) and the Möbius number system G : B R R given by G u = F u. Then V = {V(F u ) : u B} is an almostcover, so (G,(B N ) V ) is a Möbius number system by Theorem 9. Given u A N, define ũ B N by ũ k = u [kn,(k+)n). If u Σ W, then ũ (B N ) V, so lim k F u[,kn) (z) = Φ G (ũ) for any z U. In particular the condition is satisfied for each z = F v (i), where v A +, v < n. It follows lim k F u[,k) (i) = Φ G (ũ), so Φ F (u) = Φ G (ũ). Thus Σ W X F and Φ : Σ W R is continuous, since lim n W u[,n) =. Since each W n (Σ) is an almostcover, Φ : Σ W R is surjective. By Theorem 9 we get Φ([u] Σ W ) = W u for each u L(Σ W ). We show that if W is compatible with F and Σ, then (F,Σ W ) is redundant. If g : R R is continuous, then gφ : Σ W R is uniformly continuous. Given u Σ W, we construct v = f(u) Σ W by induction so that for each n there exists k n such that gφ([u [,kn)]) W v[,n). If the condition holds for n, then there exists k n+ > k n such that gφ([u [,kn+)]) is a subset of W v[,n+) for some v n with v [,n+) L(Σ W ). Thus f : Σ W Σ W is continuous and Φf = gφ. Corollary 2 If W is almostcompatible with F and Σ, and if Q(Σ,W) >, then (F,Σ W ) is a Möbius number system and Φ([u] Σ W ) = W u for each u L(Σ W ). Proof: We get Q n (Σ,W) > for some n and it follows that no F u with u L n (Σ W ) is a rotation. Theorem 3 Assume that W is almostcompatible with F and Σ, (F,Σ W ) is a Möbius number system, and Φ([u] Σ W ) = W u for each u L(Σ W ). Then Q(Σ,W). Proof: Assume the contrary and choose r with Q(Σ,W) < r <. Then Q n (Σ,W) < r n for n larger than some n. For v L(Σ W ) denote by α(v) = q(v)/r v, and a n = min{α(v) : v L(Σ W ), v n}, so a n Q n (Σ,W)/r n (Q(Σ,W)/r) n. Denote by v (n) L(Σ W ) the word of minimal length such that α(v (n) ) = a n and by x n R the corresponding number for which the minimum is attained. We show that for each prefix u of v (n) we have (Fu ) (x n ) r u. Assume the contrary with v (n) = uw. Then α(uw) = (F uw) (x n ) = (F w ) (Fu (x n )) r uw r w > (F w ) (Fu (x n )) α(w), r w (F u ) (x n ) r u which is a contradiction since w is shorter than uw and therefore α(uw) α(w). Since Σ W is compact, there exist v Σ W such that (F v [,n) ) (x m ) r n for each n m, and x m W v[,m). By compactness of R there exists x R such that (F v [,n) ) (x) r n for each n m, so Φ(v) x by Lemma 4. On the other hand x n W v [,n) = Φ([v [,n) ] Σ W ), so x = Φ(v) by the continuity of Φ, and this is a contradiction. If (F,Σ) is a Möbius number system and W a = R is the trivial interval system then Σ W = Σ so (F,Σ W ) is a Möbius number system but Q(Σ,W) <. While 8
9 the trivial interval cover is excluded in Definition 5, the example shows that some condition like Φ([u] Σ W ) = W u is necessary in Theorem 3. There are some constraints on expansion subshifts of Möbius number systems. For example, no full shift is an expansion subshift of any Möbius number systems (see Kazda [3]). 6. Arithmetical algorithms In arithmetical algorithms we work with the extended rational numbers Q = Q { } with homogenous integer coordinates x = x /x Z 2 \ {/}. Denote by I the set of closed intervals I = [I,I ] with endpoints I I in Q. Denote by M the set of MT M = M (a,b,c,d) whose coefficients a,b,c,d Z are integers with ad bc >. We assume that Σ is a sofic subshift, whose language is accepted by a deterministic finite automaton (DFA). This is a triple A = (Y, δ, i), where Y is a finite set of states, δ : Y A Y is a partial transition function, and i Y is the initial state. The function δ is extended to a partial function δ : Y A Y by δ(p,λ) = p, δ(p,ua) = δ(δ(p,u),a), where the lefthandside is defined iff the righthandside is defined. The language L(A) of A consists of words u A which are accepted, i.e., for which δ(i, u) is defined. We assume that F : A R R is a Möbius iterative system, Σ A N is a sofic subshift whose language is accepted by a DFA A = (Y,δ,i), and W = {W u : u L(Σ)} is an interval system compatible with F and Σ, such that F a M and W a I for each a A. We also assume that for some n >, Q n (Σ,W) and no F u with u L n (Σ W ) is a rotation. Finally we assume that l(w) >, so (F,Σ W ) is a redundant Möbius number system. We generalize arithmetical algorithms of Gosper [] (see also Vuillemin [3], Kornerup and Matula [4] or Potts et al. []) using infinite labelled graphs. This is a structure G = (V,E,s,t,l) where V is the set of vertices, E is the set of edges, s,t : E V is the source and target map and l : E A {λ} is the labelling function. An infinite path with source s(u ) is any sequence of edges u E N such that t(u i ) = s(u i+ ). Its label is l(u )l(u )... A N A. Definition 4 The vertices of the rational expansion graph are pairs (x,q) a Q Y. Its labelled edges are (x,q) (Fa (x),δ(q,a)) such that δ(q,a) and x W a. Proposition 5 For each x Q there exists an infinite path with source (x,i). If u A N is its label, then u Σ W and Φ(u) = x. Proof: Set x = x, q = i. At step n choose u n such that x n W un and δ(q n,u n ). Set x n+ = F u n (x n ), q n+ = δ(q n,u n ) and continue. Then we have x W u[,n), and u [,n) L(Σ W ), so u Σ W and Φ(u) = x. The nondeterministic expansion algorithm based on the rational expansion graph ofdefinition4canberealizedbyamachinewhoseinnerstatesarepairs(x,q) Q Y. The machine is attached to an output tape where the expansion is written. The algorithm can be made deterministic by further specifications, like in the arithmetic expansion algorithm of Definition 2. Definition 6 The vertices of the linear graph are triples (M,q,u) where M M, q Y and u Σ W. Its edges are a (M,q,u) (Fa M,δ(q,a),u) if δ(q,a) & M(W u ) W a, (M,q,u) λ (MF u,q,σ(u)). 9
10 Proposition 7 If u Σ W and M M then there exists an infinite path with source (M,i,u) whose label w belongs to A N. For any such w we have w Σ W and Φ(w) = M(Φ(u)). Proof: Weshowbyinductionthatifw A isthelabelofapathwithsource(m,i,u), then its target is (Fw MF u[,k),δ(i,w),u [k, ) ) for some k, and M(W u[,k] ) W w. The statement holds for w = λ and k =. If the statement holds for w and some k then we have two possible continuations: (Fw λ MF u[,k),δ(i,w),u [k, ) ) (Fw MF u[,k],δ(i,w),u [k+, ) ), (Fw a MF u[,k),δ(i,w),u [k, ) ) (FwaMF u[,k),δ(i,wa),u [k, ) ). In the former case we get M(W u[,k+] ) M(W u[,k] ) W w. In the latter case we use the assumption Fw MF u[,k) (W uk ) W a, to get M(W u[,k] ) M(F u[,k) (W uk )) F w (W a ), and M(W u[,k] ) W w F w (W a ) = W wa. Since u Σ W, we have lim k W u[,k] =, so for each j there exists k and w j such that M(W u[,k) ) W w[,j). Thus there exists a path with label w Σ W and M(Φ(u)) = Φ(w). The algorithm based on the linear graph of Definition 6 can be realized by a machine whose inner states are (M,q) M Y. The machine is attached to an input and output tapes and can see the first letter of the input tape. 7. Fractional bilinear functions To obtain algorithms for functions of two variables like the sum or product, we consider orientationreversing MT M (a,b,c,d) with ad bc <, singular MT with ad bc = and a + b + c + d >, and the zero MT M (,,,) =. Each MT defines a closed relation M = {(x,y) R 2 : (ax +bx )y = (cx +dx )y }. We define the stable point s M R of a singular MT by s M { a c, b d } R. Note that a c or b d can be R, but not both. Similarly the unstable point u M R of a singular MT is defined by u M { b a, d c } R. A singular MT yields the relation M = (R {s M }) ({u M } R), and the zero MT yields the full relation M = R 2. The value M(I) = {y R : x I,(x,y) M} of M on a closed interval I = [I,I ] is [M(I ),M(I )] if ad bc >, [M(I ),M(I )] if ad bc <, M(I) = {s M } if ad bc = & M & u M I, R if ad bc = & M & u M I, R if M =. Denote by M (,) the set of fractional bilinear functions with integer coefficients. These are functions P : R R R { } of two variables x,y R of the form which represent closed relations P(x,y) = ax y +bx y +cx y +dx y ex y +fx y +gx y +hx y P = {(x,y,z) R 3 : (ax y +bx y +cx y +dx y )z = (ex y +fx y +gx y +hx y )z }.
11 For each x,y R, P(x, ), P(,y) are MT with matrices [ ] [ ax +cx P(x, ) = bx +dx ay +by, P(,y) = cy +dy ex +gx fx +hx ey +fy gy +hy ], so if x,y R and I,J are closed intervals, then P(x,J), P(I,y) are well defined intervals. Since P(x, ) and P(,y) are monotone, the image P(I,J) of closed intervals I = [I,I ], J = [J,J ] coincides with the image of the boundary of I J: P(I,J) := {z R : x I, y J,(x,y,z) P} = P(I,J) P(I,J ) P(I,J) P(I,J ). Given a (2 2)matrix M = M (a,b,c,d), define the (4 4)matrices M x and M y by M x = a b a b c d c d, My = a b c d a b c d Then both P(Mx,y) = PM x (x,y) and P(x,My) = PM y (x,y), as well as MP are fractional bilinear functions. In these formulas, P is regarded as a (2 4)matrix. Definition 8 The bilinear graph has vertices (P,q,u,v), where P M (,), q Y and u,v Σ W. The edges are a (P,q,u,v) (Fa P,δ(q,a),u,v) if δ(q,a) & P(W u,w v ) W a, λ (P,q,u,v) (PFu x,q,σ(u),v), (P,q,u,v) λ (PF y v,q,u,σ(v)). Proposition 9 If u,v Σ W and w A N is a label of a path with the source (P,i,u,v), then w Σ W and Φ(w) = P(Φ(u),Φ(v)). The proof is similar as in Proposition 7. However, for u,v Σ W there need not exist any path with source (P,i,u,v) whose label belongs to A N. The algorithm may give only a finite output, for example when we add words representing Rational functions Denote by M n the set of rational functions R(x) = a x n +a x x n + +a n x n b x n +b x x n + +b n x n of degree n with integer coefficients. If M M, then both composed functions RM and MR belong to M n. Denote by t(x) := argd(x) = 2arctanx the isomorphism of R with ( π,π). The circle derivation R : R R of R is defined by R (x) := (trt ) (t(x)) = R (x)(+x 2 ) +R 2. (x) We have R (x) = lim y x (R(y),R(x))/ (y,x). A monotone element is a pair (R,I) such that I is a closed interval, R M n is monotone on I, i.e., R (x) does.
12 not change sign on I, and R has no poles in I, i.e., R(x) for x I. We say that a monotone element (R,I) is signchanging, if either R(I ) < < R(I ) and R (x) > for x I, or R(I ) > > R(I ) and R (x) < for x I. A signchanging monotone element (R,I) has a unique root x I with R(x) =. Proposition 2 () There exists an algorithm which for R M n gives a list of signchanging elements (R,I ),...,(R k,i k ), such that each root of R is a root of some (R j,i j ). (2) It is decidable, whether (R, I) is a monotone element. (3) Given closed intervals I,J, it is decidable whether R(I) J. These algorithms can be obtained from the Sturm theorem (see e.g., van der Waerden [2]), which counts the number of roots of a polynomial P in an interval using the Euclidean algorithm applied to P and its derivative P. The condition (3) can be decided without evaluating the interval R(I), which may have irrational endpoints: We use the the Sturm theorem to test whether R(x) J and R(x) J have roots in I. Definition 2 The algebraic expansion graph is a graph whose vertices are (R,q,I), where (R,I) is a signchanging element, and q Y. Its edges are (R,q,I) a (RF a,δ(q,a),f a (I W a )), such that δ(q,a) and (R,I W a ) is a signchanging element. Proposition 22 For each signchanging element (R, I) there exists an infinite path with source (R,i,I). If w is its label, then w Σ W, Φ(w) I, and RΦ(w) =. Proof: We show by induction that if u is a label of a finite path with source (R,i,I), then its target is (RF u,δ(i,u),j), where F u (J) = I W u, (RF u,j) is a signchanging element, and there exists a path with source (R,i,I) and label ua for some a A. If this holds for u A, then u L(Σ W ), and (R,F u (J)) = (R,I W u ) is a signchanging element, so there exists a A such that ua L(Σ W ), and (R,I W ua ) = (R,I W u F u (W a )) = (R,F u (J W a )) is a signchanging element. Thus (RF u,j W a ) is a signchanging element, so there exists a path with label ua and target (RF ua,δ(i,ua),j ), where J = Fa (J W a ). It follows that (RF ua,j ) is a signchanging element and F ua (J ) = F u (J) F u (W a ) = I W ua. The graph for the computation of a rational function R M n at Φ(u) has the same structure as the linear graph from Definition 6. To test the condition R(W u ) W a, we use the Sturm theorem rather than the simple comparison of the endpoints of intervals. However, if R is monotone in a neighbourhood I of Φ(u), the test simplifies. Finally we have a conversion algorithm between two redundant Möbius number systems: Definition 23 The (G,Θ V ) to (F,Σ W ) conversion graph has vertices (M,q,u), where M M, q Y is the state of a DFA for Σ W and u Θ V. The labelled edges are a (M,q,u) (Fa M,δ(q,a),u) if δ(q,a) & V u W a, λ (M,q,u) (MG u,q,σ(u)). 2
13 4 / Proposition 24 Assume that (G,Θ V ) and (F,Σ W ) are Möbius number systems and (F,Σ W ) is redundant. If u Θ V then there exists an infinite path with source (Id,i,u) whose label v belongs to Σ W, and Φ F (v) = Φ G (u). Using the continued fractions of Gauss (see Wall [4]), algorithms for many transcendental functions can be obtained similarly as in Potts [] or Potts et al. []. / 3/ /2 3/ /2 3/4 3/4 2/3 2/3 /2 /2 /2 /2 /3 /4 /4 /4 /4 /3 Figure. The binary signed system (left) and the regular continued fractions (right) 9. The binary signed system and the system of continued fractions We give some examples of Möbius number systems based on positional systems and continued fractions. We use the alphabet A = {,,, } whose letters stand for numbers,,,. Example The Möbius binary signed system (BSS, Figure left) consists of the alphabet A = {,,,}, transformations F (x) = +x, F (x) = x/2, F (x) = +x, F (x) = 2x, and the interval system W = (( 2, ),(,),(,2),(2, 2)). The subshift S W is sofic with forbidden words D = {,,,,,,, } and each u L(S W ) can be written as u = n v, where v {,} {,}. Then F u (x) = 2 n (s +2 s + +2 k s k +2 k x) for some s i which all belong either to {,} or to {,}. The system is not redundant and Q(S,W) >. A redundant system can be obtained as S W where W = (( 3, 2 ),(,),( 2,3),(2, 2)), or as Σ W2, where Σ = S({,,,}) forbids some identities, and W 2 = ((,),(,),(, ),(, )). We have Q n (Σ,W 2 ) = for each n 5. The means F u () of words u L(S W ) can be seen in Figure left. The curves between these means are constructed as follows. For each MT M there exists a family of MT (M t ) t R such that M = Id, M = M, and M t+s = M t M s. In Figures 3
14 and 2, each mean F u () is joined to F ua () by the curve ( F u Ft a ()) t. The labels u A + at F u () are written in the direction of the tangent vectors F u(). Example 2 The Möbius system of regular continued fraction (RCF, Figure right) consists of the alphabet A = {,, }, transformations F (x) = +x, F (x) = /x, F (x) = +x, and the interval system W = ((, ),(,),(, )). By Proposition 7, the subshift S W is of finite type with forbidden words D = {,,,,} and Q n (S,W) = for each n. For each u L(S W ), the transformation F u can be written as F u (x) = F a F F a F F an (x) where a i Z, a i a i+ and a i for i >. Thus we obtain a continued fraction whose partial quotients ( ) i a i are either all positive or all negative. There is no reasonable way to get redundant continued fractions. We obtain semiregular CF with the interval system W = ((, 2 ),(,),( 2, )), but W and W cannot be extended beyond. / / 3 3 3/2 2 3/2 5/3 2 5/ /3 /2 /2 2/3 3/5 /2 /2 3/5 /3 /4 /4 /3 /3 /4 /4 /3 Figure 2. Binary continued fractions (left) and the binary square system (right). Binary continued fractions Example 3 The Möbius system of binary continued fractions (BCF, Figure 2 left) consists of the alphabet A = {,,,}, transformations F (x) = +x, F (x) = /x, F (x) = +x, F (x) = 2x, and the interval system W = ((, 2 ),(,),( 2, ),( 2,2)). We have l(w) > and Q n (W) = for each n, so (F,S W ) is redundant and S W is sofic. In BCF, each rational number has an eventually periodic expansion with period length of the form u.. The proof of this fact for a variant of the BCF system in Kůrka [8] can be easily adapted to the present situation. To obtain shorter expansions, we can test the parities of the numerator and denominator: 4
15 8/ 5/ /4 6/ /3 3/2 8/ /4 5/ 3/2 6/ /3 4/ 3/ /2 4/ /2 3/ 2/ 2/ / / / / /. /4. /3. 2/5. /2. 3/5. 2/3. 3/4. /. 4/3. 3/2. 5/3. 2/. 5/2. 3/. 4/. Figure 3. Arithmetical expansions of rational numbers in the system of binary continued fractions. Definition 25 The arithmetical expansion graph (Figure 3 top) for the BCF system has vertices x = (x,x ) Q, with x and gcd(x,x ) =, and labelled edges x (x /2,x ) if x 2 x & 2 x, x (x,2x ) if x 2 x & 2 x, x (x +x,x ) if x x (2x < x & 2 x ), x (x x,x ) if x x (2x > x & 2 x ), x ( x sgn(x ), x ) otherwise. In the expansion procedure of Definition 25, the first applicable rule is used, so each vertex has outdegree and we get a deterministic expansion function E : Q Σ W, such that E(x) is the label of the unique infinite path with source x and Φ(E(x)) = x (see Figure 3 top). Each rational number has expansion of the form u. and integers have the same expansions as in the binary system. An integer can be written as x = x +2x + +2 k x k, where x i {,,} are either all nonnegative, or all nonpositive. Then E(x) = s s...s k s k., where s i is empty if x i =, s i = if x i =, and s i = if x i = (see Figure 3 bottom).. Hyperbolic polygonal systems An nary hyperbolic polygonal system with quotient q > considered in Kůrka [5] has alphabet A = {,,...,n } and transformations F a = R 2πa/n C q R 2πa/n, where C q (x) = x/q and R α is the rotation by α. In particular for n = 4 and q = 2 we get Definition 26 The Möbius binary square system (Figure 2 right) with quotient q = 2 consists of the alphabet A = {,,,}, transformations F (x) = 3x x+3, F (x) = x 2, F (x) = 3x+ x+3, F (x) = 2x, and the interval system W = (( 3, 3 ),( 2, 2 ),( 3,3),(2, 2)). 5
16 Note that all W a have the same length. Moreover we have F (W ) = (,) and F (W ) = (, ). It follows that S W is the sofic subshift S W = {,} N {,} N {,} N {,} N which is a union of four full shifts, one for each quadrant: Φ({,} N ) = [,], Φ({,} N ) = [,], Φ({,} N ) = [, ] and Φ({,} N ) = [, ]. A disadvantage of this system is, however, that not all rational numbers have periodic expansions. This is shown in Kůrka [6], where some conditions for the periodicity of expansions of rational numbers are given. 2. Conclusions Arithmetical algorithms simplify if Σ = S is the full shift, since no DFA is necessary in this case. However, if we forbid some identities, we can have larger expansion intervals W a and therefore a more redundant system with larger Lebesgue number, whose arithmetical algorithms may be faster. If Σ is a SFT, its compatibility can be verified by Proposition 6. In particular cases, the arithmetical algorithms can be further simplified, which is the case of the Binary signed system. While Corollary 2 and Theorem 3 formulate a tight condition on Σ and W to form a Möbius number system, the case Q(Σ,W) = has not yet been completely analysed. Acknowledgments The research was supported by the Research Program CTS MSM and by the Czech Science Foundation research project GAČR 2/9/854. A part of this paper has been written during the stay of the first author in the MaxPlanckInstitut für Mathematik in Bonn. References [] R. W. Gosper. Continued fractions arithmetic. unpublished manuscript, [2] A. Kazda. Convergence in Möbius number systems. Integers, 9:26 279, 29. [3] A. Kazda. Symbolic representations of compact spaces. Master s thesis, Faculty of Mathematics and Physics of the Charles University in Prague, Prague, 29. [4] P. Kornerup and D. W. Matula. An algorithm for redundant binary bitpipelined rational arithmetic. IEEE Transactions on Computers, 39(8):6 5, August 99. [5] P. Kůrka. A symbolic representation of the real Möbius group. Nonlinearity, 2:63 623, 28. [6] P. Kůrka. Expansion of rational numbers in Möbius number systems. 29. submitted. [7] P. Kůrka. Geometry of Möbius number systems. Technical Report MPIM 2959, MaxPlanck Institut für Mathematik, Bonn, 29. [8] P. Kůrka. Möbius number systems with sofic subshifts. Nonlinearity, 22: , 29. [9] D. Lind and B. Marcus. An Introduction to Symbolic Dynamics and Coding. Cambridge University Press, Cambridge, 995. [] P. J. Potts. Exact real arithmetic using Möbius transformations. PhD thesis, University of London, Imperial College, London, 998. [] P. J. Potts, A. Edalat, and M. H. Escardó. Semantics of exact real computation. In Proceedings of the twelfth annual IEEE symposium in computer science, pages , Warsaw, 997. [2] B. L. van der Waerden. Algebra I. SpringerVerlag, Berlin, 97. [3] J. E. Vuillemin. Exact real computer arithmetic with continued fractions. IEEE Transactions on Computers, 39(8):87 5, August 99. [4] H. S. Wall. Analytic theory of continued fractions. van Nostrand, New York, 948. [5] K. Weihrauch. Computable analysis. An introduction. EATCS Monographs on Theoretical Computer Science. SpringerVerlag, Berlin, 2. 6
Course 221: Analysis Academic year , First Semester
Course 221: Analysis Academic year 200708, First Semester David R. Wilkins Copyright c David R. Wilkins 1989 2007 Contents 1 Basic Theorems of Real Analysis 1 1.1 The Least Upper Bound Principle................
More informationThe Exact Real Arithmetical Algorithm in Binary Continued Fractions
The Exact Real Arithmetical Algorithm in Binary Continued Fractions Petr Kůrka Center for Theoretical Study Academy of Sciences and Charles University in Prague Jilská, CZ000 Praha, Czechia Abstract The
More informationMathematics Course 111: Algebra I Part IV: Vector Spaces
Mathematics Course 111: Algebra I Part IV: Vector Spaces D. R. Wilkins Academic Year 19967 9 Vector Spaces A vector space over some field K is an algebraic structure consisting of a set V on which are
More informationPUTNAM TRAINING POLYNOMIALS. Exercises 1. Find a polynomial with integral coefficients whose zeros include 2 + 5.
PUTNAM TRAINING POLYNOMIALS (Last updated: November 17, 2015) Remark. This is a list of exercises on polynomials. Miguel A. Lerma Exercises 1. Find a polynomial with integral coefficients whose zeros include
More informationCourse 421: Algebraic Topology Section 1: Topological Spaces
Course 421: Algebraic Topology Section 1: Topological Spaces David R. Wilkins Copyright c David R. Wilkins 1988 2008 Contents 1 Topological Spaces 1 1.1 Continuity and Topological Spaces...............
More informationChapter 4, Arithmetic in F [x] Polynomial arithmetic and the division algorithm.
Chapter 4, Arithmetic in F [x] Polynomial arithmetic and the division algorithm. We begin by defining the ring of polynomials with coefficients in a ring R. After some preliminary results, we specialize
More informationContinued Fractions and the Euclidean Algorithm
Continued Fractions and the Euclidean Algorithm Lecture notes prepared for MATH 326, Spring 997 Department of Mathematics and Statistics University at Albany William F Hammond Table of Contents Introduction
More informationIntroduction to Algebraic Geometry. Bézout s Theorem and Inflection Points
Introduction to Algebraic Geometry Bézout s Theorem and Inflection Points 1. The resultant. Let K be a field. Then the polynomial ring K[x] is a unique factorisation domain (UFD). Another example of a
More informationMA651 Topology. Lecture 6. Separation Axioms.
MA651 Topology. Lecture 6. Separation Axioms. This text is based on the following books: Fundamental concepts of topology by Peter O Neil Elements of Mathematics: General Topology by Nicolas Bourbaki Counterexamples
More information1 if 1 x 0 1 if 0 x 1
Chapter 3 Continuity In this chapter we begin by defining the fundamental notion of continuity for real valued functions of a single real variable. When trying to decide whether a given function is or
More informationModule MA1S11 (Calculus) Michaelmas Term 2016 Section 3: Functions
Module MA1S11 (Calculus) Michaelmas Term 2016 Section 3: Functions D. R. Wilkins Copyright c David R. Wilkins 2016 Contents 3 Functions 43 3.1 Functions between Sets...................... 43 3.2 Injective
More informationBasic Concepts of Point Set Topology Notes for OU course Math 4853 Spring 2011
Basic Concepts of Point Set Topology Notes for OU course Math 4853 Spring 2011 A. Miller 1. Introduction. The definitions of metric space and topological space were developed in the early 1900 s, largely
More informationAlgebra Unpacked Content For the new Common Core standards that will be effective in all North Carolina schools in the 201213 school year.
This document is designed to help North Carolina educators teach the Common Core (Standard Course of Study). NCDPI staff are continually updating and improving these tools to better serve teachers. Algebra
More information9 More on differentiation
Tel Aviv University, 2013 Measure and category 75 9 More on differentiation 9a Finite Taylor expansion............... 75 9b Continuous and nowhere differentiable..... 78 9c Differentiable and nowhere monotone......
More informationMetric Spaces. Chapter 1
Chapter 1 Metric Spaces Many of the arguments you have seen in several variable calculus are almost identical to the corresponding arguments in one variable calculus, especially arguments concerning convergence
More informationTHE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS
THE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS KEITH CONRAD 1. Introduction The Fundamental Theorem of Algebra says every nonconstant polynomial with complex coefficients can be factored into linear
More informationAdvanced Higher Mathematics Course Assessment Specification (C747 77)
Advanced Higher Mathematics Course Assessment Specification (C747 77) Valid from August 2015 This edition: April 2016, version 2.4 This specification may be reproduced in whole or in part for educational
More informationPURE MATHEMATICS AM 27
AM Syllabus (015): Pure Mathematics AM SYLLABUS (015) PURE MATHEMATICS AM 7 SYLLABUS 1 AM Syllabus (015): Pure Mathematics Pure Mathematics AM 7 Syllabus (Available in September) Paper I(3hrs)+Paper II(3hrs)
More informationPURE MATHEMATICS AM 27
AM SYLLABUS (013) PURE MATHEMATICS AM 7 SYLLABUS 1 Pure Mathematics AM 7 Syllabus (Available in September) Paper I(3hrs)+Paper II(3hrs) 1. AIMS To prepare students for further studies in Mathematics and
More informationMathematical Methods of Engineering Analysis
Mathematical Methods of Engineering Analysis Erhan Çinlar Robert J. Vanderbei February 2, 2000 Contents Sets and Functions 1 1 Sets................................... 1 Subsets.............................
More information4. Expanding dynamical systems
4.1. Metric definition. 4. Expanding dynamical systems Definition 4.1. Let X be a compact metric space. A map f : X X is said to be expanding if there exist ɛ > 0 and L > 1 such that d(f(x), f(y)) Ld(x,
More informationQuotient Rings and Field Extensions
Chapter 5 Quotient Rings and Field Extensions In this chapter we describe a method for producing field extension of a given field. If F is a field, then a field extension is a field K that contains F.
More informationit is easy to see that α = a
21. Polynomial rings Let us now turn out attention to determining the prime elements of a polynomial ring, where the coefficient ring is a field. We already know that such a polynomial ring is a UF. Therefore
More informationMathematics for Computer Science/Software Engineering. Notes for the course MSM1F3 Dr. R. A. Wilson
Mathematics for Computer Science/Software Engineering Notes for the course MSM1F3 Dr. R. A. Wilson October 1996 Chapter 1 Logic Lecture no. 1. We introduce the concept of a proposition, which is a statement
More informationIntroduction to Flocking {Stochastic Matrices}
Supelec EECI Graduate School in Control Introduction to Flocking {Stochastic Matrices} A. S. Morse Yale University Gif sur  Yvette May 21, 2012 CRAIG REYNOLDS  1987 BOIDS The Lion King CRAIG REYNOLDS
More informationMultiplicity. Chapter 6
Chapter 6 Multiplicity The fundamental theorem of algebra says that any polynomial of degree n 0 has exactly n roots in the complex numbers if we count with multiplicity. The zeros of a polynomial are
More informationReal Roots of Univariate Polynomials with Real Coefficients
Real Roots of Univariate Polynomials with Real Coefficients mostly written by Christina Hewitt March 22, 2012 1 Introduction Polynomial equations are used throughout mathematics. When solving polynomials
More informationINTRODUCTORY SET THEORY
M.Sc. program in mathematics INTRODUCTORY SET THEORY Katalin Károlyi Department of Applied Analysis, Eötvös Loránd University H1088 Budapest, Múzeum krt. 68. CONTENTS 1. SETS Set, equal sets, subset,
More informationThe Division Algorithm for Polynomials Handout Monday March 5, 2012
The Division Algorithm for Polynomials Handout Monday March 5, 0 Let F be a field (such as R, Q, C, or F p for some prime p. This will allow us to divide by any nonzero scalar. (For some of the following,
More informationProblem Set. Problem Set #2. Math 5322, Fall December 3, 2001 ANSWERS
Problem Set Problem Set #2 Math 5322, Fall 2001 December 3, 2001 ANSWERS i Problem 1. [Problem 18, page 32] Let A P(X) be an algebra, A σ the collection of countable unions of sets in A, and A σδ the collection
More information2 Complex Functions and the CauchyRiemann Equations
2 Complex Functions and the CauchyRiemann Equations 2.1 Complex functions In onevariable calculus, we study functions f(x) of a real variable x. Likewise, in complex analysis, we study functions f(z)
More informationDiscrete Mathematics, Chapter 5: Induction and Recursion
Discrete Mathematics, Chapter 5: Induction and Recursion Richard Mayr University of Edinburgh, UK Richard Mayr (University of Edinburgh, UK) Discrete Mathematics. Chapter 5 1 / 20 Outline 1 Wellfounded
More informationFunctions and Equations
Centre for Education in Mathematics and Computing Euclid eworkshop # Functions and Equations c 014 UNIVERSITY OF WATERLOO Euclid eworkshop # TOOLKIT Parabolas The quadratic f(x) = ax + bx + c (with a,b,c
More informationFactoring of Prime Ideals in Extensions
Chapter 4 Factoring of Prime Ideals in Extensions 4. Lifting of Prime Ideals Recall the basic AKLB setup: A is a Dedekind domain with fraction field K, L is a finite, separable extension of K of degree
More informationMetric Spaces. Chapter 7. 7.1. Metrics
Chapter 7 Metric Spaces A metric space is a set X that has a notion of the distance d(x, y) between every pair of points x, y X. The purpose of this chapter is to introduce metric spaces and give some
More informationWeek 910: Recurrence Relations and Generating Functions
Week 910: Recurrence Relations and Generating Functions March 31, 2015 1 Some number sequences An infinite sequence (or just a sequence for short is an ordered array a 0, a 1, a 2,..., a n,... of countably
More informationSOLUTIONS TO EXERCISES FOR. MATHEMATICS 205A Part 3. Spaces with special properties
SOLUTIONS TO EXERCISES FOR MATHEMATICS 205A Part 3 Fall 2008 III. Spaces with special properties III.1 : Compact spaces I Problems from Munkres, 26, pp. 170 172 3. Show that a finite union of compact subspaces
More informationRESULTANT AND DISCRIMINANT OF POLYNOMIALS
RESULTANT AND DISCRIMINANT OF POLYNOMIALS SVANTE JANSON Abstract. This is a collection of classical results about resultants and discriminants for polynomials, compiled mainly for my own use. All results
More informationMATHEMATICS (CLASSES XI XII)
MATHEMATICS (CLASSES XI XII) General Guidelines (i) All concepts/identities must be illustrated by situational examples. (ii) The language of word problems must be clear, simple and unambiguous. (iii)
More informationUnique Factorization
Unique Factorization Waffle Mathcamp 2010 Throughout these notes, all rings will be assumed to be commutative. 1 Factorization in domains: definitions and examples In this class, we will study the phenomenon
More informationCHAPTER SIX IRREDUCIBILITY AND FACTORIZATION 1. BASIC DIVISIBILITY THEORY
January 10, 2010 CHAPTER SIX IRREDUCIBILITY AND FACTORIZATION 1. BASIC DIVISIBILITY THEORY The set of polynomials over a field F is a ring, whose structure shares with the ring of integers many characteristics.
More informationMath Review. for the Quantitative Reasoning Measure of the GRE revised General Test
Math Review for the Quantitative Reasoning Measure of the GRE revised General Test www.ets.org Overview This Math Review will familiarize you with the mathematical skills and concepts that are important
More information1. R In this and the next section we are going to study the properties of sequences of real numbers.
+a 1. R In this and the next section we are going to study the properties of sequences of real numbers. Definition 1.1. (Sequence) A sequence is a function with domain N. Example 1.2. A sequence of real
More information1 Sets and Set Notation.
LINEAR ALGEBRA MATH 27.6 SPRING 23 (COHEN) LECTURE NOTES Sets and Set Notation. Definition (Naive Definition of a Set). A set is any collection of objects, called the elements of that set. We will most
More informationa 1 x + a 0 =0. (3) ax 2 + bx + c =0. (4)
ROOTS OF POLYNOMIAL EQUATIONS In this unit we discuss polynomial equations. A polynomial in x of degree n, where n 0 is an integer, is an expression of the form P n (x) =a n x n + a n 1 x n 1 + + a 1 x
More information2.3 Convex Constrained Optimization Problems
42 CHAPTER 2. FUNDAMENTAL CONCEPTS IN CONVEX OPTIMIZATION Theorem 15 Let f : R n R and h : R R. Consider g(x) = h(f(x)) for all x R n. The function g is convex if either of the following two conditions
More information3. INNER PRODUCT SPACES
. INNER PRODUCT SPACES.. Definition So far we have studied abstract vector spaces. These are a generalisation of the geometric spaces R and R. But these have more structure than just that of a vector space.
More informationMathematics Review for MS Finance Students
Mathematics Review for MS Finance Students Anthony M. Marino Department of Finance and Business Economics Marshall School of Business Lecture 1: Introductory Material Sets The Real Number System Functions,
More informationAutomata and Languages
Automata and Languages Computational Models: An idealized mathematical model of a computer. A computational model may be accurate in some ways but not in others. We shall be defining increasingly powerful
More informationUnderstanding Basic Calculus
Understanding Basic Calculus S.K. Chung Dedicated to all the people who have helped me in my life. i Preface This book is a revised and expanded version of the lecture notes for Basic Calculus and other
More informationPROBLEM SET 6: POLYNOMIALS
PROBLEM SET 6: POLYNOMIALS 1. introduction In this problem set we will consider polynomials with coefficients in K, where K is the real numbers R, the complex numbers C, the rational numbers Q or any other
More informationThe Dirichlet Unit Theorem
Chapter 6 The Dirichlet Unit Theorem As usual, we will be working in the ring B of algebraic integers of a number field L. Two factorizations of an element of B are regarded as essentially the same if
More informationDecember 4, 2013 MATH 171 BASIC LINEAR ALGEBRA B. KITCHENS
December 4, 2013 MATH 171 BASIC LINEAR ALGEBRA B KITCHENS The equation 1 Lines in twodimensional space (1) 2x y = 3 describes a line in twodimensional space The coefficients of x and y in the equation
More informationThe Goldberg Rao Algorithm for the Maximum Flow Problem
The Goldberg Rao Algorithm for the Maximum Flow Problem COS 528 class notes October 18, 2006 Scribe: Dávid Papp Main idea: use of the blocking flow paradigm to achieve essentially O(min{m 2/3, n 1/2 }
More informationRieszFredhölm Theory
RieszFredhölm Theory T. Muthukumar tmk@iitk.ac.in Contents 1 Introduction 1 2 Integral Operators 1 3 Compact Operators 7 4 Fredhölm Alternative 14 Appendices 18 A AscoliArzelá Result 18 B Normed Spaces
More informationCollinear Points in Permutations
Collinear Points in Permutations Joshua N. Cooper Courant Institute of Mathematics New York University, New York, NY József Solymosi Department of Mathematics University of British Columbia, Vancouver,
More information7  Linear Transformations
7  Linear Transformations Mathematics has as its objects of study sets with various structures. These sets include sets of numbers (such as the integers, rationals, reals, and complexes) whose structure
More informationMATH 304 Linear Algebra Lecture 9: Subspaces of vector spaces (continued). Span. Spanning set.
MATH 304 Linear Algebra Lecture 9: Subspaces of vector spaces (continued). Span. Spanning set. Vector space A vector space is a set V equipped with two operations, addition V V (x,y) x + y V and scalar
More informationThis chapter describes set theory, a mathematical theory that underlies all of modern mathematics.
Appendix A Set Theory This chapter describes set theory, a mathematical theory that underlies all of modern mathematics. A.1 Basic Definitions Definition A.1.1. A set is an unordered collection of elements.
More informationNOTES ON MEASURE THEORY. M. Papadimitrakis Department of Mathematics University of Crete. Autumn of 2004
NOTES ON MEASURE THEORY M. Papadimitrakis Department of Mathematics University of Crete Autumn of 2004 2 Contents 1 σalgebras 7 1.1 σalgebras............................... 7 1.2 Generated σalgebras.........................
More informationMAT2400 Analysis I. A brief introduction to proofs, sets, and functions
MAT2400 Analysis I A brief introduction to proofs, sets, and functions In Analysis I there is a lot of manipulations with sets and functions. It is probably also the first course where you have to take
More informationApproximating the entropy of a 2dimensional shift of finite type
Approximating the entropy of a dimensional shift of finite type Tirasan Khandhawit c 4 July 006 Abstract. In this paper, we extend the method used to compute entropy of dimensional subshift and the technique
More informationFIRST YEAR CALCULUS. Chapter 7 CONTINUITY. It is a parabola, and we can draw this parabola without lifting our pencil from the paper.
FIRST YEAR CALCULUS WWLCHENW L c WWWL W L Chen, 1982, 2008. 2006. This chapter originates from material used by the author at Imperial College, University of London, between 1981 and 1990. It It is is
More informationCopy in your notebook: Add an example of each term with the symbols used in algebra 2 if there are any.
Algebra 2  Chapter Prerequisites Vocabulary Copy in your notebook: Add an example of each term with the symbols used in algebra 2 if there are any. P1 p. 1 1. counting(natural) numbers  {1,2,3,4,...}
More informationMetric Spaces Joseph Muscat 2003 (Last revised May 2009)
1 Distance J Muscat 1 Metric Spaces Joseph Muscat 2003 (Last revised May 2009) (A revised and expanded version of these notes are now published by Springer.) 1 Distance A metric space can be thought of
More informationComplex Numbers and the Complex Exponential
Complex Numbers and the Complex Exponential Frank R. Kschischang The Edward S. Rogers Sr. Department of Electrical and Computer Engineering University of Toronto September 5, 2005 Numbers and Equations
More informationClassification of Cartan matrices
Chapter 7 Classification of Cartan matrices In this chapter we describe a classification of generalised Cartan matrices This classification can be compared as the rough classification of varieties in terms
More informationInfinite product spaces
Chapter 7 Infinite product spaces So far we have talked a lot about infinite sequences of independent random variables but we have never shown how to construct such sequences within the framework of probability/measure
More information1. Prove that the empty set is a subset of every set.
1. Prove that the empty set is a subset of every set. Basic Topology Written by MenGen Tsai email: b89902089@ntu.edu.tw Proof: For any element x of the empty set, x is also an element of every set since
More informationApproximation Algorithms
Approximation Algorithms or: How I Learned to Stop Worrying and Deal with NPCompleteness Ong Jit Sheng, Jonathan (A0073924B) March, 2012 Overview Key Results (I) General techniques: Greedy algorithms
More informationBOREL SETS, WELLORDERINGS OF R AND THE CONTINUUM HYPOTHESIS. Proof. We shall define inductively a decreasing sequence of infinite subsets of N
BOREL SETS, WELLORDERINGS OF R AND THE CONTINUUM HYPOTHESIS SIMON THOMAS 1. The Finite Basis Problem Definition 1.1. Let C be a class of structures. Then a basis for C is a collection B C such that for
More informationCORRELATED TO THE SOUTH CAROLINA COLLEGE AND CAREERREADY FOUNDATIONS IN ALGEBRA
We Can Early Learning Curriculum PreK Grades 8 12 INSIDE ALGEBRA, GRADES 8 12 CORRELATED TO THE SOUTH CAROLINA COLLEGE AND CAREERREADY FOUNDATIONS IN ALGEBRA April 2016 www.voyagersopris.com Mathematical
More informationFinite Sets. Theorem 5.1. Two nonempty finite sets have the same cardinality if and only if they are equivalent.
MATH 337 Cardinality Dr. Neal, WKU We now shall prove that the rational numbers are a countable set while R is uncountable. This result shows that there are two different magnitudes of infinity. But we
More informationThe Ideal Class Group
Chapter 5 The Ideal Class Group We will use Minkowski theory, which belongs to the general area of geometry of numbers, to gain insight into the ideal class group of a number field. We have already mentioned
More informationChapter 2 Limits Functions and Sequences sequence sequence Example
Chapter Limits In the net few chapters we shall investigate several concepts from calculus, all of which are based on the notion of a limit. In the normal sequence of mathematics courses that students
More informationThe BanachTarski Paradox
University of Oslo MAT2 Project The BanachTarski Paradox Author: Fredrik Meyer Supervisor: Nadia S. Larsen Abstract In its weak form, the BanachTarski paradox states that for any ball in R, it is possible
More informationSome Polynomial Theorems. John Kennedy Mathematics Department Santa Monica College 1900 Pico Blvd. Santa Monica, CA 90405 rkennedy@ix.netcom.
Some Polynomial Theorems by John Kennedy Mathematics Department Santa Monica College 1900 Pico Blvd. Santa Monica, CA 90405 rkennedy@ix.netcom.com This paper contains a collection of 31 theorems, lemmas,
More informationPrime Numbers and Irreducible Polynomials
Prime Numbers and Irreducible Polynomials M. Ram Murty The similarity between prime numbers and irreducible polynomials has been a dominant theme in the development of number theory and algebraic geometry.
More informationNotes from FE Review for mathematics, University of Kentucky
Notes from FE Review for mathematics, University of Kentucky Dr. Alan Demlow March 19, 2012 Introduction These notes are based on reviews for the mathematics portion of the Fundamentals of Engineering
More informationLabeling outerplanar graphs with maximum degree three
Labeling outerplanar graphs with maximum degree three Xiangwen Li 1 and Sanming Zhou 2 1 Department of Mathematics Huazhong Normal University, Wuhan 430079, China 2 Department of Mathematics and Statistics
More informationStructure of Measurable Sets
Structure of Measurable Sets In these notes we discuss the structure of Lebesgue measurable subsets of R from several different points of view. Along the way, we will see several alternative characterizations
More information4. Factor polynomials over complex numbers, describe geometrically, and apply to realworld situations. 5. Determine and apply relationships among syn
I The Real and Complex Number Systems 1. Identify subsets of complex numbers, and compare their structural characteristics. 2. Compare and contrast the properties of real numbers with the properties of
More informationMATH 289 PROBLEM SET 1: INDUCTION. 1. The induction Principle The following property of the natural numbers is intuitively clear:
MATH 89 PROBLEM SET : INDUCTION The induction Principle The following property of the natural numbers is intuitively clear: Axiom Every nonempty subset of the set of nonnegative integers Z 0 = {0,,, 3,
More informationVERTICES OF GIVEN DEGREE IN SERIESPARALLEL GRAPHS
VERTICES OF GIVEN DEGREE IN SERIESPARALLEL GRAPHS MICHAEL DRMOTA, OMER GIMENEZ, AND MARC NOY Abstract. We show that the number of vertices of a given degree k in several kinds of seriesparallel labelled
More information1 = (a 0 + b 0 α) 2 + + (a m 1 + b m 1 α) 2. for certain elements a 0,..., a m 1, b 0,..., b m 1 of F. Multiplying out, we obtain
Notes on realclosed fields These notes develop the algebraic background needed to understand the model theory of realclosed fields. To understand these notes, a standard graduate course in algebra is
More informationCARDINALITY, COUNTABLE AND UNCOUNTABLE SETS PART ONE
CARDINALITY, COUNTABLE AND UNCOUNTABLE SETS PART ONE With the notion of bijection at hand, it is easy to formalize the idea that two finite sets have the same number of elements: we just need to verify
More information1 Norms and Vector Spaces
008.10.07.01 1 Norms and Vector Spaces Suppose we have a complex vector space V. A norm is a function f : V R which satisfies (i) f(x) 0 for all x V (ii) f(x + y) f(x) + f(y) for all x,y V (iii) f(λx)
More information1.5. Factorisation. Introduction. Prerequisites. Learning Outcomes. Learning Style
Factorisation 1.5 Introduction In Block 4 we showed the way in which brackets were removed from algebraic expressions. Factorisation, which can be considered as the reverse of this process, is dealt with
More informationRANDOM INTERVAL HOMEOMORPHISMS. MICHA L MISIUREWICZ Indiana University Purdue University Indianapolis
RANDOM INTERVAL HOMEOMORPHISMS MICHA L MISIUREWICZ Indiana University Purdue University Indianapolis This is a joint work with Lluís Alsedà Motivation: A talk by Yulij Ilyashenko. Two interval maps, applied
More informationREAL ANALYSIS I HOMEWORK 2
REAL ANALYSIS I HOMEWORK 2 CİHAN BAHRAN The questions are from Stein and Shakarchi s text, Chapter 1. 1. Prove that the Cantor set C constructed in the text is totally disconnected and perfect. In other
More informationCHAPTER 1 BASIC TOPOLOGY
CHAPTER 1 BASIC TOPOLOGY Topology, sometimes referred to as the mathematics of continuity, or rubber sheet geometry, or the theory of abstract topological spaces, is all of these, but, above all, it is
More informationP NP for the Reals with various Analytic Functions
P NP for the Reals with various Analytic Functions Mihai Prunescu Abstract We show that nondeterministic machines in the sense of [BSS] defined over wide classes of real analytic structures are more powerful
More informationPSEUDOARCS, PSEUDOCIRCLES, LAKES OF WADA AND GENERIC MAPS ON S 2
PSEUDOARCS, PSEUDOCIRCLES, LAKES OF WADA AND GENERIC MAPS ON S 2 Abstract. We prove a BrucknerGarg type theorem for the fiber structure of a generic map from a continuum X into the unit interval I. We
More informationFinite Fields and ErrorCorrecting Codes
Lecture Notes in Mathematics Finite Fields and ErrorCorrecting Codes KarlGustav Andersson (Lund University) (version 1.01316 September 2015) Translated from Swedish by Sigmundur Gudmundsson Contents
More information1 VECTOR SPACES AND SUBSPACES
1 VECTOR SPACES AND SUBSPACES What is a vector? Many are familiar with the concept of a vector as: Something which has magnitude and direction. an ordered pair or triple. a description for quantities such
More informationON GALOIS REALIZATIONS OF THE 2COVERABLE SYMMETRIC AND ALTERNATING GROUPS
ON GALOIS REALIZATIONS OF THE 2COVERABLE SYMMETRIC AND ALTERNATING GROUPS DANIEL RABAYEV AND JACK SONN Abstract. Let f(x) be a monic polynomial in Z[x] with no rational roots but with roots in Q p for
More informationChapter 7. Induction and Recursion. Part 1. Mathematical Induction
Chapter 7. Induction and Recursion Part 1. Mathematical Induction The principle of mathematical induction is this: to establish an infinite sequence of propositions P 1, P 2, P 3,..., P n,... (or, simply
More informationMath 120 Final Exam Practice Problems, Form: A
Math 120 Final Exam Practice Problems, Form: A Name: While every attempt was made to be complete in the types of problems given below, we make no guarantees about the completeness of the problems. Specifically,
More information9. POLYNOMIALS. Example 1: The expression a(x) = x 3 4x 2 + 7x 11 is a polynomial in x. The coefficients of a(x) are the numbers 1, 4, 7, 11.
9. POLYNOMIALS 9.1. Definition of a Polynomial A polynomial is an expression of the form: a(x) = a n x n + a n1 x n1 +... + a 1 x + a 0. The symbol x is called an indeterminate and simply plays the role
More informationDIFFERENTIAL GEOMETRY HW 2
DIFFERENTIAL GEOMETRY HW 2 CLAY SHONKWILER 2 Prove that the only orientationreversing isometries of R 2 are glide reflections, that any two glidereflections through the same distance are conjugate by
More information