distribution properties of digital functions over
play

Distribution properties of digital functions over Gaussian integers - PowerPoint PPT Presentation

Distribution properties of digital functions over Gaussian integers Peter Grabner (joint work with M. Drmota (TU Vienna) and P. Liardet (Marseille)) Institut f ur Mathematik A, Graz University of Technology Journ ees Num eration,


  1. Distribution properties of digital functions over Gaussian integers Peter Grabner (joint work with M. Drmota (TU Vienna) and P. Liardet (Marseille)) Institut f¨ ur Mathematik A, Graz University of Technology Journ´ ees Num´ eration, Prague 26/05/2008

  2. Digital expansions A digital representation of the elements of A (= N , Z , Z [ i ], Z K . . . ) is a bijection rep : A → L ⊂ D ∗ , where L is a language over the finite alphabet of “digits” D .

  3. Digital expansions A digital representation of the elements of A (= N , Z , Z [ i ], Z K . . . ) is a bijection rep : A → L ⊂ D ∗ , where L is a language over the finite alphabet of “digits” D . Usually, this bijection is given by a base sequence ( G n ) n ∈ N 0 with G 0 = 1 and its inverse map val : L → A m � ( ε 0 , . . . , ε m ) �→ ε ℓ G ℓ . ℓ =0

  4. Digital expansions A digital representation of the elements of A (= N , Z , Z [ i ], Z K . . . ) is a bijection rep : A → L ⊂ D ∗ , where L is a language over the finite alphabet of “digits” D . Usually, this bijection is given by a base sequence ( G n ) n ∈ N 0 with G 0 = 1 and its inverse map val : L → A m � ( ε 0 , . . . , ε m ) �→ ε ℓ G ℓ . ℓ =0 A different approach was given by M. Rigo for A = N . In this case an ordering on D is used to impose an order on the regular language L . Then n ∈ N is represented by the n -th element of L ,

  5. Digital functions Digital functions are functions which depend in a simple way on a given digital representation of A , for instance additively or multiplicatively on single digits or blocks of digits m � f (val( ε 0 , . . . , ε m )) = g ( ε ℓ ) ℓ =0 m � f (val( ε 0 , . . . , ε m )) = g ( ε ℓ ) ℓ =0 m + L � f (val( ε 0 , . . . , ε m )) = g ( ε ℓ , ε ℓ +1 , . . . , ε ℓ + L − 1 ) , ℓ = − L where we set ε − k = 0 for k ∈ N .

  6. Distribution properties In the classical case of digital representations of the positive integers, several different types of distribution results are known: ◮ uniform distribution of the values of f in residue classes modulo m for integer valued f

  7. Distribution properties In the classical case of digital representations of the positive integers, several different types of distribution results are known: ◮ uniform distribution of the values of f in residue classes modulo m for integer valued f ◮ uniform distribution of the values of f modulo 1, if f attains one irrational value

  8. Distribution properties In the classical case of digital representations of the positive integers, several different types of distribution results are known: ◮ uniform distribution of the values of f in residue classes modulo m for integer valued f ◮ uniform distribution of the values of f modulo 1, if f attains one irrational value ◮ uniform distribution of the values of f along Følner sequences

  9. Distribution properties In the classical case of digital representations of the positive integers, several different types of distribution results are known: ◮ uniform distribution of the values of f in residue classes modulo m for integer valued f ◮ uniform distribution of the values of f modulo 1, if f attains one irrational value ◮ uniform distribution of the values of f along Følner sequences ◮ asymptotic normality of the values of f

  10. Distribution properties In the classical case of digital representations of the positive integers, several different types of distribution results are known: ◮ uniform distribution of the values of f in residue classes modulo m for integer valued f ◮ uniform distribution of the values of f modulo 1, if f attains one irrational value ◮ uniform distribution of the values of f along Følner sequences ◮ asymptotic normality of the values of f ◮ local limit theorems for integer valued f

  11. Gaussian integers Every z ∈ Z [ i ] can be represented in the form m ε ℓ b ℓ , ε ℓ ∈ { 0 , . . . , | b | 2 − 1 } =: D , � z = ℓ =0 if and only if b = − a ± i , a ∈ N .

  12. Gaussian integers Every z ∈ Z [ i ] can be represented in the form m ε ℓ b ℓ , ε ℓ ∈ { 0 , . . . , | b | 2 − 1 } =: D , � z = ℓ =0 if and only if b = − a ± i , a ∈ N . The corresponding sum-of-digits function is given by m � s b ( z ) = ε ℓ . ℓ =0

  13. Gaussian integers Every z ∈ Z [ i ] can be represented in the form m ε ℓ b ℓ , ε ℓ ∈ { 0 , . . . , | b | 2 − 1 } =: D , � z = ℓ =0 if and only if b = − a ± i , a ∈ N . The corresponding sum-of-digits function is given by m � s b ( z ) = ε ℓ . ℓ =0 Similarly, for every F : D L → R with F (0 , 0 , . . . , 0) = 0 m + L � s F ( z ) = F ( ε ℓ , ε ℓ +1 , . . . , ε ℓ + L − 1 ) ℓ = − L defines a block additive function.

  14. Mean values A first result on the mean value of the sum-of-digits function s b is s b ( z ) = π N | b | 2 − 1 √ � log | b | 2 N + NF b (log | b | 2 N )+ O ( N log N ) , 2 | z | 2 < N where F b is a continuous periodic function of period 1.

  15. Mean values A first result on the mean value of the sum-of-digits function s b is s b ( z ) = π N | b | 2 − 1 √ � log | b | 2 N + NF b (log | b | 2 N )+ O ( N log N ) , 2 | z | 2 < N where F b is a continuous periodic function of period 1. The mean value of the sum-of-digits function along the real line is given by s b ( n ) = N | b | 2 − 1 log | b | 2 N + O ( N log α N ) � 2 n < N for some α < 1 for b � = − 1 ± i .

  16. Exponential sums All the distribution results shown before can be derived from precise knowledge of the behaviour of the exponential sums � e ts F ( z ) z ∈ A N with t taking values either in an interval on the real line, or along the imaginary axis, or in an open complex neighbourhood of 0.

  17. Three techniques In a recent joint work with M. Drmota and P. Liardet we have described three different techniques, which allow to derive distribution results of various kinds for block-additive (and more general) digital functions on the Gaussian integers (and other number fields).

  18. Three techniques In a recent joint work with M. Drmota and P. Liardet we have described three different techniques, which allow to derive distribution results of various kinds for block-additive (and more general) digital functions on the Gaussian integers (and other number fields). ◮ a measure theoretic technique

  19. Three techniques In a recent joint work with M. Drmota and P. Liardet we have described three different techniques, which allow to derive distribution results of various kinds for block-additive (and more general) digital functions on the Gaussian integers (and other number fields). ◮ a measure theoretic technique ◮ a technique based on Dirichlet series

  20. Three techniques In a recent joint work with M. Drmota and P. Liardet we have described three different techniques, which allow to derive distribution results of various kinds for block-additive (and more general) digital functions on the Gaussian integers (and other number fields). ◮ a measure theoretic technique ◮ a technique based on Dirichlet series ◮ an ergodic technique

  21. The measure theoretic technique The main idea is to realise that the sequence of measures z ∈ b N A e ts F ( z ) � µ N , t ( A ) = � z ∈ B N e ts F ( z ) converges weakly to a limit measure µ t , where we denote � N � ε ℓ b ℓ | ε ℓ ∈ { 0 , . . . , | b | 2 − 1 } � B N = . ℓ =0

  22. The measure theoretic technique The main idea is to realise that the sequence of measures z ∈ b N A e ts F ( z ) � µ N , t ( A ) = � z ∈ B N e ts F ( z ) converges weakly to a limit measure µ t , where we denote � N � ε ℓ b ℓ | ε ℓ ∈ { 0 , . . . , | b | 2 − 1 } � B N = . ℓ =0 Then conversely e ts F ( z ) = µ t ( A ) λ N � t + o ( λ N t ) , z ∈ b N A where λ t is the dominating eigenvalue of a weighted adjacency matrix related to the function F .

  23. Error terms In order to obtain error terms for the convergence µ N , t → µ t , we use an according version of the Berry-Esseen inequality.

  24. Error terms In order to obtain error terms for the convergence µ N , t → µ t , we use an according version of the Berry-Esseen inequality. This needs estimates on the measure dimension of µ t , which have to be worked out from the definition of µ t .

  25. Error terms In order to obtain error terms for the convergence µ N , t → µ t , we use an according version of the Berry-Esseen inequality. This needs estimates on the measure dimension of µ t , which have to be worked out from the definition of µ t . The Fourier-transform of µ N , t can be computed as z ∈ b N A e ts F ( z ) e 2 π i ℑ ( xzb − N ) � µ N , t ( x ) = ˆ . z ∈ B N e ts F ( z ) � Numerator and denominator can be expressed as matrix products in terms of weighted adjacency matrices.

  26. Putting things together. . . For a µ t -continuity set A we get e ts F ( z ) = µ t ( ANb −⌊ log | b | N ⌋ ) λ ⌊ log | b | N ⌋ � + O ( N log | b | λ t − α t ) . t z ∈ NA for | t | ≤ C with C > 0 and α t > 0.

Recommend


More recommend