Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 ›...
Transcript of Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 ›...
![Page 1: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/1.jpg)
Extensions of the Pascal Triangle to Words,and Related Counting Problems
Manon Stipulanti
LiègeApril 2, 2019
![Page 2: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/2.jpg)
Classical Pascal triangle
P: (m, k) ∈ N×N 7→(mk
)∈ N
(mk
)k
0 1 2 3 4 5 6 7 · · ·0 1 0 0 0 0 0 0 01 1 1 0 0 0 0 0 02 1 2 1 0 0 0 0 0
m 3 1 3 3 1 0 0 0 04 1 4 6 4 1 0 0 05 1 5 10 10 5 1 0 06 1 6 15 20 15 6 1 07 1 7 21 35 35 21 7 1...
. . .
Binomial coefficients of integers:(m
k
)=
m!
(m− k)! k!
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 2
![Page 3: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/3.jpg)
A specific construction
• Grid: first 2n rows andcolumns of the Pascal triangle((
m
k
)mod 2
)
0≤m,k<2n
• Color each square in• white if
(mk
)≡ 0 mod 2
• black if(mk
)≡ 1 mod 2
• Normalize by a homothetyof ratio 1/2n
(bring into [0, 1]2)
sequence of compact setsbelonging to [0, 1]2
N2 ∩ [0, 2n]2
0
1
...
2n − 1
2n
0 1 · · · 2n − 1 2n
(00
) (01
) · · ·(
02n−1
)
(10
) (11
) · · ·(
12n−1
)
......
. . ....
(2n−10
) (2n−11
) · · ·(2n−12n−1
)
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 3
![Page 4: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/4.jpg)
A specific construction
• Grid: first 2n rows andcolumns of the Pascal triangle((
m
k
)mod 2
)
0≤m,k<2n
• Color each square in• white if
(mk
)≡ 0 mod 2
• black if(mk
)≡ 1 mod 2
• Normalize by a homothetyof ratio 1/2n
(bring into [0, 1]2)
sequence of compact setsbelonging to [0, 1]2
N2 ∩ [0, 2n]2
0
1
...
2n − 1
2n
0 1 · · · 2n − 1 2n
(00
) (01
) · · ·(
02n−1
)
(10
) (11
) · · ·(
12n−1
)
......
. . ....
(2n−10
) (2n−11
) · · ·(2n−12n−1
)
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 3
![Page 5: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/5.jpg)
A specific construction
• Grid: first 2n rows andcolumns of the Pascal triangle((
m
k
)mod 2
)
0≤m,k<2n
• Color each square in• white if
(mk
)≡ 0 mod 2
• black if(mk
)≡ 1 mod 2
• Normalize by a homothetyof ratio 1/2n
(bring into [0, 1]2)
sequence of compact setsbelonging to [0, 1]2
N2 ∩ [0, 2n]2
0
1
...
2n − 1
2n
0 1 · · · 2n − 1 2n
(00
) (01
) · · ·(
02n−1
)
(10
) (11
) · · ·(
12n−1
)
......
. . ....
(2n−10
) (2n−11
) · · ·(2n−12n−1
)
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 3
![Page 6: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/6.jpg)
A specific construction
• Grid: first 2n rows andcolumns of the Pascal triangle((
m
k
)mod 2
)
0≤m,k<2n
• Color each square in• white if
(mk
)≡ 0 mod 2
• black if(mk
)≡ 1 mod 2
• Normalize by a homothetyof ratio 1/2n
(bring into [0, 1]2)
sequence of compact setsbelonging to [0, 1]2
N2 ∩ [0, 2n]2
0
1
...
2n − 1
2n
0 1 · · · 2n − 1 2n
(00
) (01
) · · ·(
02n−1
)
(10
) (11
) · · ·(
12n−1
)
......
. . ....
(2n−10
) (2n−11
) · · ·(2n−12n−1
)
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 3
![Page 7: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/7.jpg)
The first six elements of the sequence
0 1
1
0 2
2
0 22
22
1 0 0 0
1 1 0 0
1 2 1 0
1 3 3 1
0 23
23
0 24
24
0 25
25
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 4
![Page 8: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/8.jpg)
Folklore factThe latter sequence of compact setsconverges to the Sierpiński gasket(w.r.t. the Hausdorff distance).
Definitions:• ε-fattening of a subset S ⊂ R2
[S]ε =⋃
x∈SB(x, ε)
• (H(R2), dh) complete space of the non-empty compact subsets ofR2 equipped with the Hausdorff distance dh
dh(S, S′) = inf{ε ∈ R>0 | S ⊂ [S′]ε and S′ ⊂ [S]ε}Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 5
![Page 9: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/9.jpg)
Theorem (von Haeseler, Peitgen, and Skordev, 1992)Let p be a prime and s > 0.The sequence of compact sets corresponding to
((m
k
)mod ps
)
0≤m,k<pn
converges when n tends to infinity (w.r.t. the Hausdorff distance).
p = 2, s = 1 p = 2, s = 2 p = 2, s = 3
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 6
![Page 10: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/10.jpg)
Part I
![Page 11: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/11.jpg)
Binomial coefficient of finite words
Binomial coefficient of wordsLet u, v be two finite words over the alphabet A.The binomial coefficient
(uv
)of u and v is the number of times v occurs
as a subsequence of u (meaning as a “scattered” subword).
Example: u = 101001 v = 101
101001,101001,101001,101001, 101001, 101001
}⇒(
101001
101
)= 6
Natural generalization:
testtesttesttest
(am
ak
)=
(m times︷ ︸︸ ︷a · · · aa · · · a︸ ︷︷ ︸k times
)=
(m
k
)∀m, k ∈ N
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 8
![Page 12: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/12.jpg)
Binomial coefficient of finite words
Binomial coefficient of wordsLet u, v be two finite words over the alphabet A.The binomial coefficient
(uv
)of u and v is the number of times v occurs
as a subsequence of u (meaning as a “scattered” subword).
Example: u = 101001 v = 101
101001,101001,101001,101001, 101001, 101001
}⇒(
101001
101
)= 6
Natural generalization:
testtesttesttest
(am
ak
)=
(m times︷ ︸︸ ︷a · · · aa · · · a︸ ︷︷ ︸k times
)=
(m
k
)∀m, k ∈ N
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 8
![Page 13: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/13.jpg)
Binomial coefficient of finite words
Binomial coefficient of wordsLet u, v be two finite words over the alphabet A.The binomial coefficient
(uv
)of u and v is the number of times v occurs
as a subsequence of u (meaning as a “scattered” subword).
Example: u = 101001 v = 101
101001,101001,101001,101001, 101001, 101001
}⇒(
101001
101
)= 6
Natural generalization:
testtesttesttest
(am
ak
)=
(m times︷ ︸︸ ︷a · · · aa · · · a︸ ︷︷ ︸k times
)=
(m
k
)∀m, k ∈ N
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 8
![Page 14: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/14.jpg)
Generalized Pascal triangles
Let (A,<) be a totally ordered alphabet.
Let L ⊂ A∗ be an infinite language over A.The words in L are genealogically ordered
w0 <gen w1 <gen w2 <gen · · · .
The generalized Pascal triangle PL associated with L is defined by
PL : (m, k) ∈ N×N 7→(wmwk
)∈ N .
Questions:• With a similar construction, can we expect the convergence to an
analogue of the Sierpiński gasket?• In particular, where should we cut to normalize a given generalized
Pascal triangle?• Could we describe this limit object?
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 9
![Page 15: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/15.jpg)
Generalized Pascal triangles
Let (A,<) be a totally ordered alphabet.Let L ⊂ A∗ be an infinite language over A.
The words in L are genealogically ordered
w0 <gen w1 <gen w2 <gen · · · .
The generalized Pascal triangle PL associated with L is defined by
PL : (m, k) ∈ N×N 7→(wmwk
)∈ N .
Questions:• With a similar construction, can we expect the convergence to an
analogue of the Sierpiński gasket?• In particular, where should we cut to normalize a given generalized
Pascal triangle?• Could we describe this limit object?
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 9
![Page 16: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/16.jpg)
Generalized Pascal triangles
Let (A,<) be a totally ordered alphabet.Let L ⊂ A∗ be an infinite language over A.The words in L are genealogically ordered
w0 <gen w1 <gen w2 <gen · · · .
The generalized Pascal triangle PL associated with L is defined by
PL : (m, k) ∈ N×N 7→(wmwk
)∈ N .
Questions:• With a similar construction, can we expect the convergence to an
analogue of the Sierpiński gasket?• In particular, where should we cut to normalize a given generalized
Pascal triangle?• Could we describe this limit object?
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 9
![Page 17: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/17.jpg)
Generalized Pascal triangles
Let (A,<) be a totally ordered alphabet.Let L ⊂ A∗ be an infinite language over A.The words in L are genealogically ordered
w0 <gen w1 <gen w2 <gen · · · .
The generalized Pascal triangle PL associated with L is defined by
PL : (m, k) ∈ N×N 7→(wmwk
)∈ N .
Questions:• With a similar construction, can we expect the convergence to an
analogue of the Sierpiński gasket?• In particular, where should we cut to normalize a given generalized
Pascal triangle?• Could we describe this limit object?
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 9
![Page 18: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/18.jpg)
Generalized Pascal triangles
Let (A,<) be a totally ordered alphabet.Let L ⊂ A∗ be an infinite language over A.The words in L are genealogically ordered
w0 <gen w1 <gen w2 <gen · · · .
The generalized Pascal triangle PL associated with L is defined by
PL : (m, k) ∈ N×N 7→(wmwk
)∈ N .
Questions:• With a similar construction, can we expect the convergence to an
analogue of the Sierpiński gasket?• In particular, where should we cut to normalize a given generalized
Pascal triangle?• Could we describe this limit object?
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 9
![Page 19: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/19.jpg)
Linear numeration systemDefinitionA numeration system is a sequence U = (U(n))n≥0 of integers s.t.• U increasing• U(0) = 1
• supn≥0U(n+1)U(n) bounded by a constant finite alphabet.
A numeration system U is linear if ∃k ≥ 1,∃a0, . . . , ak−1 ∈ Z s.t.
U(n+ k) = ak−1 U(n+ k − 1) + · · ·+ a0 U(n) ∀n ≥ 0.
Greedy representation in (U(n))n≥0:
n =∑̀
i=0
ciU(i) withj−1∑
i=0
ciU(i) < U(j)
repU (n) = c` · · · c0 ∈ LU = repU (N)︸ ︷︷ ︸numeration language
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 10
![Page 20: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/20.jpg)
Linear numeration systemDefinitionA numeration system is a sequence U = (U(n))n≥0 of integers s.t.• U increasing• U(0) = 1
• supn≥0U(n+1)U(n) bounded by a constant finite alphabet.
A numeration system U is linear if ∃k ≥ 1,∃a0, . . . , ak−1 ∈ Z s.t.
U(n+ k) = ak−1 U(n+ k − 1) + · · ·+ a0 U(n) ∀n ≥ 0.
Greedy representation in (U(n))n≥0:
n =∑̀
i=0
ciU(i) withj−1∑
i=0
ciU(i) < U(j)
repU (n) = c` · · · c0 ∈ LU = repU (N)︸ ︷︷ ︸numeration language
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 10
![Page 21: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/21.jpg)
Linear numeration systemDefinitionA numeration system is a sequence U = (U(n))n≥0 of integers s.t.• U increasing• U(0) = 1
• supn≥0U(n+1)U(n) bounded by a constant finite alphabet.
A numeration system U is linear if ∃k ≥ 1,∃a0, . . . , ak−1 ∈ Z s.t.
U(n+ k) = ak−1 U(n+ k − 1) + · · ·+ a0 U(n) ∀n ≥ 0.
Greedy representation in (U(n))n≥0:
n =∑̀
i=0
ciU(i) withj−1∑
i=0
ciU(i) < U(j)
repU (n) = c` · · · c0 ∈ LU = repU (N)︸ ︷︷ ︸numeration language
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 10
![Page 22: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/22.jpg)
Parry numbers
β ∈ R>1 Aβ = {0, 1, . . . , dβe − 1}
x ∈ [0, 1] x =
+∞∑
j=1
cjβ−j , cj ∈ Aβ
Greedy way: cjβ−j + cj+1β−j−1 + · · · < β−(j−1)
β-expansion of x: dβ(x) = c1c2c3 · · ·
Definitionβ ∈ R>1 is a Parry number if dβ(1) is ultimately periodic.
Example: b ∈ N>1: db(1) = (b− 1)ω
Example: Golden ratio ϕ: dϕ(1) = 110ω
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 11
![Page 23: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/23.jpg)
Parry numbers
β ∈ R>1 Aβ = {0, 1, . . . , dβe − 1}
x ∈ [0, 1] x =
+∞∑
j=1
cjβ−j , cj ∈ Aβ
Greedy way: cjβ−j + cj+1β−j−1 + · · · < β−(j−1)
β-expansion of x: dβ(x) = c1c2c3 · · ·
Definitionβ ∈ R>1 is a Parry number if dβ(1) is ultimately periodic.
Example: b ∈ N>1: db(1) = (b− 1)ω
Example: Golden ratio ϕ: dϕ(1) = 110ω
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 11
![Page 24: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/24.jpg)
Parry numbers
β ∈ R>1 Aβ = {0, 1, . . . , dβe − 1}
x ∈ [0, 1] x =
+∞∑
j=1
cjβ−j , cj ∈ Aβ
Greedy way: cjβ−j + cj+1β−j−1 + · · · < β−(j−1)
β-expansion of x: dβ(x) = c1c2c3 · · ·
Definitionβ ∈ R>1 is a Parry number if dβ(1) is ultimately periodic.
Example: b ∈ N>1: db(1) = (b− 1)ω
Example: Golden ratio ϕ: dϕ(1) = 110ω
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 11
![Page 25: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/25.jpg)
Parry numeration system
Parry number β ∈ R>1 linear numeration system (Uβ(n))n≥0
• dβ(1) = t1 · · · tm0ω
Uβ(0) = 1Uβ(i) = t1Uβ(i− 1) + · · ·+ tiUβ(0) + 1 ∀ 1 ≤ i ≤ m− 1Uβ(n) = t1Uβ(n− 1) + · · ·+ tmUβ(n−m) ∀n ≥ m
• dβ(1) = t1 · · · tm(tm+1 · · · tm+k)ω
Uβ(0) = 1Uβ(i) = t1Uβ(i− 1) + · · ·+ tiUβ(0) + 1 ∀ 1 ≤ i ≤ m+ k − 1Uβ(n) = t1Uβ(n− 1) + · · ·+ tm+kUβ(n−m− k) ∀n ≥ m+ k
+Uβ(n− k)−t1Uβ(n− k − 1)− · · · − tmUβ(n−m− k)
Examples:b ∈ N>1 (bn)n≥0 base bGolden ratio ϕ (F (n))n≥0 Fibonacci numeration system
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 12
![Page 26: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/26.jpg)
Special case of generalized Pascal trianglesParry number β ∈ R>1 linear numeration system Uβ
Pβ : (m, k) ∈ N×N 7→(
repUβ (m)
repUβ (k)
)∈ N
Examples:Base-2 numeration system
ε 1 10 11 100 101 110 111ε 1 0 0 0 0 0 0 01 1 1 0 0 0 0 0 010 1 1 1 0 0 0 0 011 1 2 0 1 0 0 0 0100 1 1 2 0 1 0 0 0101 1 2 1 1 0 1 0 0110 1 2 2 1 0 0 1 0111 1 3 0 3 0 0 0 1
L2 = 1{0, 1}∗ ∪ {ε}
Fibonacci numeration systemε 1 10 100 101 1000 1001 1010
ε 1 0 0 0 0 0 0 01 1 1 0 0 0 0 0 010 1 1 1 0 0 0 0 0100 1 1 2 1 0 0 0 0101 1 2 1 0 1 0 0 01000 1 1 3 3 0 1 0 01001 1 2 2 1 2 0 1 01010 1 2 3 1 1 0 0 1
Lϕ = 1{01, 0}∗ ∪ {ε}
Remark: Copies of the usual Pascal triangle
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 13
![Page 27: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/27.jpg)
Special case of generalized Pascal trianglesParry number β ∈ R>1 linear numeration system Uβ
Pβ : (m, k) ∈ N×N 7→(
repUβ (m)
repUβ (k)
)∈ N
Examples:Base-2 numeration system
ε 1 10 11 100 101 110 111ε 1 0 0 0 0 0 0 01 1 1 0 0 0 0 0 010 1 1 1 0 0 0 0 011 1 2 0 1 0 0 0 0100 1 1 2 0 1 0 0 0101 1 2 1 1 0 1 0 0110 1 2 2 1 0 0 1 0111 1 3 0 3 0 0 0 1
L2 = 1{0, 1}∗ ∪ {ε}
Fibonacci numeration systemε 1 10 100 101 1000 1001 1010
ε 1 0 0 0 0 0 0 01 1 1 0 0 0 0 0 010 1 1 1 0 0 0 0 0100 1 1 2 1 0 0 0 0101 1 2 1 0 1 0 0 01000 1 1 3 3 0 1 0 01001 1 2 2 1 2 0 1 01010 1 2 3 1 1 0 0 1
Lϕ = 1{01, 0}∗ ∪ {ε}
Remark: Copies of the usual Pascal triangle
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 13
![Page 28: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/28.jpg)
Special case of generalized Pascal trianglesParry number β ∈ R>1 linear numeration system Uβ
Pβ : (m, k) ∈ N×N 7→(
repUβ (m)
repUβ (k)
)∈ N
Examples:Base-2 numeration system
ε 1 10 11 100 101 110 111ε 1 0 0 0 0 0 0 01 1 1 0 0 0 0 0 010 1 1 1 0 0 0 0 011 1 2 0 1 0 0 0 0100 1 1 2 0 1 0 0 0101 1 2 1 1 0 1 0 0110 1 2 2 1 0 0 1 0111 1 3 0 3 0 0 0 1
L2 = 1{0, 1}∗ ∪ {ε}
Fibonacci numeration systemε 1 10 100 101 1000 1001 1010
ε 1 0 0 0 0 0 0 01 1 1 0 0 0 0 0 010 1 1 1 0 0 0 0 0100 1 1 2 1 0 0 0 0101 1 2 1 0 1 0 0 01000 1 1 3 3 0 1 0 01001 1 2 2 1 2 0 1 01010 1 2 3 1 1 0 0 1
Lϕ = 1{01, 0}∗ ∪ {ε}
Remark: Copies of the usual Pascal triangleGeneralized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 13
![Page 29: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/29.jpg)
Sequence of compact sets (first Uβ(n) rows and columns of Pβ) in [0, 1]2:
Uβn =1
Uβ(n)
⋃
u,v∈LUβ ,|u|,|v|≤n(uv)≡1 mod 2
valUβ (v, u) + [0, 1]2
Examples: Base-2 numeration systemε 1
1
ε 10
10
ε 102
102
1 0 0 0
1 1 0 0
1 1 1 0
1 2 0 1
ε 103
103
0 104
104
0 105
105
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 14
![Page 30: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/30.jpg)
Sequence of compact sets (first Uβ(n) rows and columns of Pβ) in [0, 1]2:
Uβn =1
Uβ(n)
⋃
u,v∈LUβ ,|u|,|v|≤n(uv)≡1 mod 2
valUβ (v, u) + [0, 1]2
Examples: Fibonacci numeration systemε 1
1
ε 10
10
ε 102
102
ε 103
103
ε 104
104
ε 105
105
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 14
![Page 31: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/31.jpg)
Base-2 numeration system
Lines of slopes: 2n, n ≥ 0
Fibonacci numeration system
Lines of slopes: ϕn, n ≥ 0
General case: Lines of slopes: βn, n ≥ 0
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 15
![Page 32: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/32.jpg)
(?) condition
p(u, v) ∈ N s.t. u0p(u,v)w, v0p(u,v)w ∈ LUβ for all w ∈ 0∗LUβ
(?)
(u, v) satisfies (?) iff u = v = ε or
u, v 6= ε(u0p(u,v)
v0p(u,v)
)≡ 1 (mod 2)(
u0p(u,v)
v0p(u,v)a
)= 0 ∀ a ∈ AUβ .
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 16
![Page 33: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/33.jpg)
• The (?) condition describes lines of slope 1 in [0, 1]2.0 v 1
u
1
Su,v
(u, v) ∈ LUβ × LUβ satisfying (?) closed segment Su,v• slope 1
• length√
2 · β−|u|−p(u,v)
• origin Au,v = (0.0|u|−|v|v, 0.u)
• New compact set containing those lines:
Aβ0 =⋃
(u,v)satisfying(?)
Su,v ⊂ [0, 1]2
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
0.8
1.0
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 17
![Page 34: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/34.jpg)
• The (?) condition describes lines of slope 1 in [0, 1]2.0 v 1
u
1
Su,v
(u, v) ∈ LUβ × LUβ satisfying (?) closed segment Su,v• slope 1
• length√
2 · β−|u|−p(u,v)
• origin Au,v = (0.0|u|−|v|v, 0.u)
• New compact set containing those lines:
Aβ0 =⋃
(u,v)satisfying(?)
Su,v ⊂ [0, 1]2
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
0.8
1.0
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 17
![Page 35: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/35.jpg)
• Two maps c : (x, y) 7→ (xβ ,yβ ) and h : (x, y) 7→ (x, βy)
Example: 01β 1
1β
1
• New compact set containing lines of slopes 1, β, . . . , βn:
Aβn =⋃
0≤i≤n, 0≤j≤ihj(ci(Aβ0 )) ⊂ [0, 1]2.
• The compact sets (Aβn)n≥0 are increasingly nested and their unionis bounded.(Aβn)n≥0 converges to Lβ =
⋃n≥0A
βn w.r.t. the Hausdorff distance.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 18
![Page 36: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/36.jpg)
• Two maps c : (x, y) 7→ (xβ ,yβ ) and h : (x, y) 7→ (x, βy)
Example: 01β
1β2 1
1β2
1β
1
• New compact set containing lines of slopes 1, β, . . . , βn:
Aβn =⋃
0≤i≤n, 0≤j≤ihj(ci(Aβ0 )) ⊂ [0, 1]2.
• The compact sets (Aβn)n≥0 are increasingly nested and their unionis bounded.(Aβn)n≥0 converges to Lβ =
⋃n≥0A
βn w.r.t. the Hausdorff distance.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 18
![Page 37: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/37.jpg)
• Two maps c : (x, y) 7→ (xβ ,yβ ) and h : (x, y) 7→ (x, βy)
Example: 01β
1β2 1
1β2
1β
1
• New compact set containing lines of slopes 1, β, . . . , βn:
Aβn =⋃
0≤i≤n, 0≤j≤ihj(ci(Aβ0 )) ⊂ [0, 1]2.
• The compact sets (Aβn)n≥0 are increasingly nested and their unionis bounded.(Aβn)n≥0 converges to Lβ =
⋃n≥0A
βn w.r.t. the Hausdorff distance.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 18
![Page 38: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/38.jpg)
• Two maps c : (x, y) 7→ (xβ ,yβ ) and h : (x, y) 7→ (x, βy)
Example: 01β
1β2
1β3 1
1β3
1β2
1β
1
• New compact set containing lines of slopes 1, β, . . . , βn:
Aβn =⋃
0≤i≤n, 0≤j≤ihj(ci(Aβ0 )) ⊂ [0, 1]2.
• The compact sets (Aβn)n≥0 are increasingly nested and their unionis bounded.(Aβn)n≥0 converges to Lβ =
⋃n≥0A
βn w.r.t. the Hausdorff distance.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 18
![Page 39: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/39.jpg)
• Two maps c : (x, y) 7→ (xβ ,yβ ) and h : (x, y) 7→ (x, βy)
Example: 01β
1β2
1β3 1
1β3
1β2
1β
1
• New compact set containing lines of slopes 1, β, . . . , βn:
Aβn =⋃
0≤i≤n, 0≤j≤ihj(ci(Aβ0 )) ⊂ [0, 1]2.
• The compact sets (Aβn)n≥0 are increasingly nested and their unionis bounded.(Aβn)n≥0 converges to Lβ =
⋃n≥0A
βn w.r.t. the Hausdorff distance.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 18
![Page 40: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/40.jpg)
• Two maps c : (x, y) 7→ (xβ ,yβ ) and h : (x, y) 7→ (x, βy)
Example: 01β
1β2
1β3 1
1β3
1β2
1β
1
• New compact set containing lines of slopes 1, β, . . . , βn:
Aβn =⋃
0≤i≤n, 0≤j≤ihj(ci(Aβ0 )) ⊂ [0, 1]2.
• The compact sets (Aβn)n≥0 are increasingly nested and their unionis bounded.(Aβn)n≥0 converges to Lβ =
⋃n≥0A
βn w.r.t. the Hausdorff distance.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 18
![Page 41: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/41.jpg)
• Two maps c : (x, y) 7→ (xβ ,yβ ) and h : (x, y) 7→ (x, βy)
Example: 01β
1β2
1β3 1
1β3
1β2
1β
1
• New compact set containing lines of slopes 1, β, . . . , βn:
Aβn =⋃
0≤i≤n, 0≤j≤ihj(ci(Aβ0 )) ⊂ [0, 1]2.
• The compact sets (Aβn)n≥0 are increasingly nested and their unionis bounded.(Aβn)n≥0 converges to Lβ =
⋃n≥0A
βn w.r.t. the Hausdorff distance.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 18
![Page 42: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/42.jpg)
Results
Theorem
(Uβn )n≥0 converges to Lβ w.r.t. the Hausdorff distance.
3 L3 ϕ Lϕ ϕ2 Lϕ2
β1 ≈ 2.47098 Lβ1 β2 ≈ 1.38028 Lβ2 β3 ≈ 2.80399 Lβ3
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 19
![Page 43: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/43.jpg)
Results (continued)
Previous result: even and odd binomial coefficients
TheoremLet p be a prime and r ∈ {1, . . . , p − 1}. When considering binomialcoefficients congruent to r (mod p), the sequence (Uβn,p,r)n≥0 convergesto a well-defined compact set Lβp,r w.r.t. the Hausdorff distance.
Example: L23,1 ∪ L2
3,2
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 20
![Page 44: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/44.jpg)
Part II
![Page 45: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/45.jpg)
Counting scattered subwords
Example: Pϕ (Fibonacci numeration system)
ε 1 10 100 101 1000 1001 1010 n Sϕ(n)ε 1 0 0 0 0 0 0 0 0 11 1 1 0 0 0 0 0 0 1 2
10 1 1 1 0 0 0 0 0 2 3100 1 1 2 1 0 0 0 0 3 4101 1 2 1 0 1 0 0 0 4 4
1000 1 1 3 3 0 1 0 0 5 51001 1 2 2 1 2 0 1 0 6 61010 1 2 3 1 1 0 0 1 7 6
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 22
![Page 46: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/46.jpg)
Classical Pascal triangle: S(n) = n+ 1 ∀n ≥ 0
Base-2 numeration system:S2 = 1, 2, 3, 3, 4, 5, 5, 4, 5, 7, 8, 7, 7, 8, 7, 5, 6, 9, 11, 10, 11, 13, 12, . . .
50 100 150 200 250
10
20
30
40
50
OEIS tag: A007306
Fibonacci numeration system:Sϕ = 1, 2, 3, 4, 4, 5, 6, 6, 6, 8, 9, 8, 8, 7, 10, 12, 12, 12, 10, 12, 12, 8, 12, . . .
50 100 150 200
20
40
60
80
OEIS tag: A282717
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 23
![Page 47: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/47.jpg)
Classical Pascal triangle: S(n) = n+ 1 ∀n ≥ 0
Base-2 numeration system:S2 = 1, 2, 3, 3, 4, 5, 5, 4, 5, 7, 8, 7, 7, 8, 7, 5, 6, 9, 11, 10, 11, 13, 12, . . .
50 100 150 200 250
10
20
30
40
50
OEIS tag: A007306
Fibonacci numeration system:Sϕ = 1, 2, 3, 4, 4, 5, 6, 6, 6, 8, 9, 8, 8, 7, 10, 12, 12, 12, 10, 12, 12, 8, 12, . . .
50 100 150 200
20
40
60
80
OEIS tag: A282717
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 23
![Page 48: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/48.jpg)
Regularity
Example: Sum-of-digits function Sum2 in base 2Example: Sum2(n) = # of 1’s in rep2(n)
n 0 1 2 3 4 5 6 7 8 9 10
rep2(n) ε 1 10 11 100 101 110 111 1000 1001 1010
Sum2(n) 0 1 1 2 1 2 2 3 1 2 2
2-kernel K2(Sum2):
Suffix ε (Sum2(n))n≥0
Suffix 0 (Sum2(2n))n≥0 Suffix 1 (Sum2(2n+ 1))n≥0
Suffix 00 (Sum2(4n))n≥0 Suffix 01 (Sum2(4n+ 1))n≥0
Suffix 10 (Sum2(4n+ 2))n≥0 Suffix 11 (Sum2(4n+ 3))n≥0
Sum2(2n) = Sum2(n) Sum2(2n+ 1) = Sum2(n) + 1
Sum2 2-regular:sequences in K2(Sum2) are Z-linear combinations of Sum2, 1
e.g. Sum2(4n+ 1) = Sum2(2(2n) + 1) = Sum2(2n) + 1 = Sum2(n) + 1Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 24
![Page 49: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/49.jpg)
Regularity
Example: Sum-of-digits function Sum2 in base 2Example: Sum2(n) = # of 1’s in rep2(n)
n 0 1 2 3 4 5 6 7 8 9 10
rep2(n) ε 1 10 11 100 101 110 111 1000 1001 1010
Sum2(n) 0 1 1 2 1 2 2 3 1 2 2
2-kernel K2(Sum2):
Suffix ε (Sum2(n))n≥0
Suffix 0 (Sum2(2n))n≥0 Suffix 1 (Sum2(2n+ 1))n≥0
Suffix 00 (Sum2(4n))n≥0 Suffix 01 (Sum2(4n+ 1))n≥0
Suffix 10 (Sum2(4n+ 2))n≥0 Suffix 11 (Sum2(4n+ 3))n≥0
Sum2(2n) = Sum2(n) Sum2(2n+ 1) = Sum2(n) + 1
Sum2 2-regular:sequences in K2(Sum2) are Z-linear combinations of Sum2, 1
e.g. Sum2(4n+ 1) = Sum2(2(2n) + 1) = Sum2(2n) + 1 = Sum2(n) + 1Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 24
![Page 50: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/50.jpg)
Regularity
Example: Sum-of-digits function Sum2 in base 2Example: Sum2(n) = # of 1’s in rep2(n)
n 0 1 2 3 4 5 6 7 8 9 10
rep2(n) ε 1 10 11 100 101 110 111 1000 1001 1010
Sum2(n) 0 1 1 2 1 2 2 3 1 2 2
2-kernel K2(Sum2):
Suffix ε (Sum2(n))n≥0
Suffix 0 (Sum2(2n))n≥0 Suffix 1 (Sum2(2n+ 1))n≥0
Suffix 00 (Sum2(4n))n≥0 Suffix 01 (Sum2(4n+ 1))n≥0
Suffix 10 (Sum2(4n+ 2))n≥0 Suffix 11 (Sum2(4n+ 3))n≥0
Sum2(2n) = Sum2(n) Sum2(2n+ 1) = Sum2(n) + 1
Sum2 2-regular:sequences in K2(Sum2) are Z-linear combinations of Sum2, 1
e.g. Sum2(4n+ 1) = Sum2(2(2n) + 1) = Sum2(2n) + 1 = Sum2(n) + 1
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 24
![Page 51: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/51.jpg)
Regularity
Example: Sum-of-digits function Sum2 in base 2Example: Sum2(n) = # of 1’s in rep2(n)
n 0 1 2 3 4 5 6 7 8 9 10
rep2(n) ε 1 10 11 100 101 110 111 1000 1001 1010
Sum2(n) 0 1 1 2 1 2 2 3 1 2 2
2-kernel K2(Sum2):
Suffix ε (Sum2(n))n≥0
Suffix 0 (Sum2(2n))n≥0 Suffix 1 (Sum2(2n+ 1))n≥0
Suffix 00 (Sum2(4n))n≥0 Suffix 01 (Sum2(4n+ 1))n≥0
Suffix 10 (Sum2(4n+ 2))n≥0 Suffix 11 (Sum2(4n+ 3))n≥0
Sum2(2n) = Sum2(n) Sum2(2n+ 1) = Sum2(n) + 1
Sum2 2-regular:sequences in K2(Sum2) are Z-linear combinations of Sum2, 1
e.g. Sum2(4n+ 1) = Sum2(2(2n) + 1) = Sum2(2n) + 1 = Sum2(n) + 1
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 24
![Page 52: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/52.jpg)
Regularity
Example: Sum-of-digits function Sum2 in base 2Example: Sum2(n) = # of 1’s in rep2(n)
n 0 1 2 3 4 5 6 7 8 9 10
rep2(n) ε 1 10 11 100 101 110 111 1000 1001 1010
Sum2(n) 0 1 1 2 1 2 2 3 1 2 2
2-kernel K2(Sum2):
Suffix ε (Sum2(n))n≥0
Suffix 0 (Sum2(2n))n≥0 Suffix 1 (Sum2(2n+ 1))n≥0
Suffix 00 (Sum2(4n))n≥0 Suffix 01 (Sum2(4n+ 1))n≥0
Suffix 10 (Sum2(4n+ 2))n≥0 Suffix 11 (Sum2(4n+ 3))n≥0
Sum2(2n) = Sum2(n) Sum2(2n+ 1) = Sum2(n) + 1
Sum2 2-regular:sequences in K2(Sum2) are Z-linear combinations of Sum2, 1
e.g. Sum2(4n+ 1) = Sum2(2(2n) + 1) = Sum2(2n) + 1 = Sum2(n) + 1
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 24
![Page 53: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/53.jpg)
Regularity
Example: Sum-of-digits function Sum2 in base 2Example: Sum2(n) = # of 1’s in rep2(n)
n 0 1 2 3 4 5 6 7 8 9 10
rep2(n) ε 1 10 11 100 101 110 111 1000 1001 1010
Sum2(n) 0 1 1 2 1 2 2 3 1 2 2
2-kernel K2(Sum2):
Suffix ε (Sum2(n))n≥0
Suffix 0 (Sum2(2n))n≥0 Suffix 1 (Sum2(2n+ 1))n≥0
Suffix 00 (Sum2(4n))n≥0 Suffix 01 (Sum2(4n+ 1))n≥0
Suffix 10 (Sum2(4n+ 2))n≥0 Suffix 11 (Sum2(4n+ 3))n≥0
Sum2(2n) = Sum2(n) Sum2(2n+ 1) = Sum2(n) + 1
Sum2 2-regular:sequences in K2(Sum2) are Z-linear combinations of Sum2, 1
e.g. Sum2(4n+ 1) = Sum2(2(2n) + 1) = Sum2(2n) + 1 = Sum2(n) + 1
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 24
![Page 54: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/54.jpg)
Regularity
Example: Sum-of-digits function Sum2 in base 2Example: Sum2(n) = # of 1’s in rep2(n)
n 0 1 2 3 4 5 6 7 8 9 10
rep2(n) ε 1 10 11 100 101 110 111 1000 1001 1010
Sum2(n) 0 1 1 2 1 2 2 3 1 2 2
2-kernel K2(Sum2):
Suffix ε (Sum2(n))n≥0
Suffix 0 (Sum2(2n))n≥0 Suffix 1 (Sum2(2n+ 1))n≥0
Suffix 00 (Sum2(4n))n≥0 Suffix 01 (Sum2(4n+ 1))n≥0
Suffix 10 (Sum2(4n+ 2))n≥0 Suffix 11 (Sum2(4n+ 3))n≥0
Sum2(2n) = Sum2(n) Sum2(2n+ 1) = Sum2(n) + 1
Sum2 2-regular:sequences in K2(Sum2) are Z-linear combinations of Sum2, 1
e.g. Sum2(4n+1) = Sum2(2(2n)+1) = Sum2(2n)+1 = 1·Sum2(n)+1·1Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 24
![Page 55: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/55.jpg)
Regularity: general caseLet U = (U(n))n≥0 be a numeration system.Let s = (s(n))n≥0 be a sequence.
• The sequence (s(iw(n)))n≥0 is called the subsequence of s with leastsignificant digits equal to w w.r.t. representations in the numera-tion system U .• The U -kernel of s is the set
KU (s) = {(s(iw(n)))n≥0 | for all suffixes w}.
• A sequence s of integers is U -regular if there exists a finite num-ber of sequences t1 = (t1(n))n≥0, . . . , t` = (t`(n))n≥0 s.t. everysequence in the U -kernel KU (s) is a Z-linear combination of thesequences t1, . . . , t`.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 25
![Page 56: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/56.jpg)
Method
Define, study, use a new tree structure called tries of scattered subwords. easily count/enumerate scattered subwords
Example: in base 2
•
•
•
•
•
•
•
•
•
•
•
•
1
0
1
1
0
Scattered subwords of 10110:
ε, 1, 10, 11, 100, 101, 110, 111,
1010, 1011, 1110, 10110
S2(val2(10110)) = S2(22) = # nodes = 12
Internal structure: subtree of root 11 u subtree of root 101
Consequence: Study trees to deduce properties of Sβ
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 26
![Page 57: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/57.jpg)
Method
Define, study, use a new tree structure called tries of scattered subwords. easily count/enumerate scattered subwords
Example: in base 2
•
•
•
•
•
•
•
•
•
•
•
•
1
0
1
1
0
Scattered subwords of 10110:
ε, 1, 10, 11, 100, 101, 110, 111,
1010, 1011, 1110, 10110
S2(val2(10110)) = S2(22) = # nodes = 12
Internal structure: subtree of root 11 u subtree of root 101
Consequence: Study trees to deduce properties of Sβ
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 26
![Page 58: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/58.jpg)
Method
Define, study, use a new tree structure called tries of scattered subwords. easily count/enumerate scattered subwords
Example: in base 2
•
•
•
•
•
•
•
•
•
•
•
•
1
0
1
1
0
Scattered subwords of 10110:
ε, 1, 10, 11, 100, 101, 110, 111,
1010, 1011, 1110, 10110
S2(val2(10110)) = S2(22) = # nodes = 12
Internal structure: subtree of root 11 u subtree of root 101
Consequence: Study trees to deduce properties of Sβ
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 26
![Page 59: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/59.jpg)
Method
Define, study, use a new tree structure called tries of scattered subwords. easily count/enumerate scattered subwords
Example: in base 2
•
•
•
•
•
•
•
•
•
•
•
•
1
0
1
1
0
Scattered subwords of 10110:
ε, 1, 10, 11, 100, 101, 110, 111,
1010, 1011, 1110, 10110
S2(val2(10110)) = S2(22) = # nodes = 12
Internal structure: subtree of root 11 u subtree of root 101
Consequence: Study trees to deduce properties of Sβ
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 26
![Page 60: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/60.jpg)
Method
Define, study, use a new tree structure called tries of scattered subwords. easily count/enumerate scattered subwords
Example: in base 2
•
•
•
•
•
•
•
•
•
•
•
•
1
0
1
1
0
Scattered subwords of 10110:
ε, 1, 10, 11, 100, 101, 110, 111,
1010, 1011, 1110, 10110
S2(val2(10110)) = S2(22) = # nodes = 12
Internal structure: subtree of root 11 u subtree of root 101
Consequence: Study trees to deduce properties of Sβ
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 26
![Page 61: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/61.jpg)
Method
Define, study, use a new tree structure called tries of scattered subwords. easily count/enumerate scattered subwords
Example: in base 2
•
•
•
•
•
•
•
•
•
•
•
•
1
0
1
1
0
Scattered subwords of 10110:
ε, 1, 10, 11, 100, 101, 110, 111,
1010, 1011, 1110, 10110
S2(val2(10110)) = S2(22) = # nodes = 12
Internal structure: subtree of root 11 u subtree of root 101
Consequence: Study trees to deduce properties of Sβ
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 26
![Page 62: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/62.jpg)
Method
Define, study, use a new tree structure called tries of scattered subwords. easily count/enumerate scattered subwords
Example: in base 2
•
•
•
•
•
•
•
•
•
•
•
•
1
0
1
1
0
Scattered subwords of 10110:
ε, 1, 10, 11, 100, 101, 110, 111,
1010, 1011, 1110, 10110
S2(val2(10110)) = S2(22) = # nodes = 12
Internal structure: subtree of root 11 u subtree of root 101
Consequence: Study trees to deduce properties of Sβ
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 26
![Page 63: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/63.jpg)
Results
Integer base case:• The sequence (Sb(n))n≥0 satisfies recurrence relations.
• The sequence (Sb(n))n≥0 is palindromic over intervals of the form[(b− 1)b`, b`+1].• The sequence (Sb(n))n≥0 is b-regular.• We obtain a matrix representation for the sequence (Sb(n))n≥0.
Fibonacci case:• The sequence (Sϕ(n))n≥0 satisfies recurrence relations.• The sequence (Sϕ(n))n≥0 is Fibonacci-regular.• We obtain a matrix representation for the sequence (Sϕ(n))n≥0.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 27
![Page 64: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/64.jpg)
Results
Integer base case:• The sequence (Sb(n))n≥0 satisfies recurrence relations.• The sequence (Sb(n))n≥0 is palindromic over intervals of the form
[(b− 1)b`, b`+1].
• The sequence (Sb(n))n≥0 is b-regular.• We obtain a matrix representation for the sequence (Sb(n))n≥0.
Fibonacci case:• The sequence (Sϕ(n))n≥0 satisfies recurrence relations.• The sequence (Sϕ(n))n≥0 is Fibonacci-regular.• We obtain a matrix representation for the sequence (Sϕ(n))n≥0.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 27
![Page 65: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/65.jpg)
Results
Integer base case:• The sequence (Sb(n))n≥0 satisfies recurrence relations.• The sequence (Sb(n))n≥0 is palindromic over intervals of the form
[(b− 1)b`, b`+1].• The sequence (Sb(n))n≥0 is b-regular.
• We obtain a matrix representation for the sequence (Sb(n))n≥0.
Fibonacci case:• The sequence (Sϕ(n))n≥0 satisfies recurrence relations.• The sequence (Sϕ(n))n≥0 is Fibonacci-regular.• We obtain a matrix representation for the sequence (Sϕ(n))n≥0.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 27
![Page 66: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/66.jpg)
Results
Integer base case:• The sequence (Sb(n))n≥0 satisfies recurrence relations.• The sequence (Sb(n))n≥0 is palindromic over intervals of the form
[(b− 1)b`, b`+1].• The sequence (Sb(n))n≥0 is b-regular.• We obtain a matrix representation for the sequence (Sb(n))n≥0.
Fibonacci case:• The sequence (Sϕ(n))n≥0 satisfies recurrence relations.• The sequence (Sϕ(n))n≥0 is Fibonacci-regular.• We obtain a matrix representation for the sequence (Sϕ(n))n≥0.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 27
![Page 67: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/67.jpg)
Results
Integer base case:• The sequence (Sb(n))n≥0 satisfies recurrence relations.• The sequence (Sb(n))n≥0 is palindromic over intervals of the form
[(b− 1)b`, b`+1].• The sequence (Sb(n))n≥0 is b-regular.• We obtain a matrix representation for the sequence (Sb(n))n≥0.
Fibonacci case:• The sequence (Sϕ(n))n≥0 satisfies recurrence relations.
• The sequence (Sϕ(n))n≥0 is Fibonacci-regular.• We obtain a matrix representation for the sequence (Sϕ(n))n≥0.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 27
![Page 68: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/68.jpg)
Results
Integer base case:• The sequence (Sb(n))n≥0 satisfies recurrence relations.• The sequence (Sb(n))n≥0 is palindromic over intervals of the form
[(b− 1)b`, b`+1].• The sequence (Sb(n))n≥0 is b-regular.• We obtain a matrix representation for the sequence (Sb(n))n≥0.
Fibonacci case:• The sequence (Sϕ(n))n≥0 satisfies recurrence relations.• The sequence (Sϕ(n))n≥0 is Fibonacci-regular.
• We obtain a matrix representation for the sequence (Sϕ(n))n≥0.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 27
![Page 69: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/69.jpg)
Results
Integer base case:• The sequence (Sb(n))n≥0 satisfies recurrence relations.• The sequence (Sb(n))n≥0 is palindromic over intervals of the form
[(b− 1)b`, b`+1].• The sequence (Sb(n))n≥0 is b-regular.• We obtain a matrix representation for the sequence (Sb(n))n≥0.
Fibonacci case:• The sequence (Sϕ(n))n≥0 satisfies recurrence relations.• The sequence (Sϕ(n))n≥0 is Fibonacci-regular.• We obtain a matrix representation for the sequence (Sϕ(n))n≥0.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 27
![Page 70: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/70.jpg)
More results in the base-2 case
• Connection with the Farey tree (every reduced positive rationalless than 1 exactly once).
Let w ∈ 1{0, 1}∗with val2(w) = 2k + r.
01
S2(r)S2(2k+r)
w
01
11
12
13
14
15
27
25
38
37
23
35
47
58
34
57
45
1
0 1
0 1 0 1
0 1 0 1 0 1 0 1
• The Stern–Brocot sequence (SB(n))n≥0 contains the numeratorsand denominators in the Farey tree.Then S2(n) = SB(2n+ 1) for all n ≥ 0.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 28
![Page 71: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/71.jpg)
Part III
![Page 72: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/72.jpg)
Behavior of summatory functions
Example: Sum-of-digits function Sum2 in base 2Example: Sum2(n) = # of 1’s in rep2(n)Example: Sum2 is 2-regular
Summatory function A of Sum2
A(n) =
n−1∑
j=0
Sum2(j) ∀n ≥ 0
Theorem (Delange, 1975)There exists a continuous nowhere differentiable periodic function G ofperiod 1 s.t.
A(n)
n=
1
2log2 n+ G(log2 n).
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 30
![Page 73: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/73.jpg)
Summatory functions of b-regular sequences algebraic or analytic methods
New method: to tackle the behavior of the summatory functionNew method: As = (As(n))n≥0 of a regular sequence s = (s(n))n≥0
• Find r = (r(n))n≥0 and t = (t(n))n≥0
• each satisfying a linear recurrence relation,• verifying As ◦ r = t.
• Recurrence relation for s recurrence relation for As in which t is involved.• Relevant representations of As in some exotic numeration system
associated with t.“Exotic” possibly unbounded coefficients• The behavior of As depends on
• the dominant root of the characteristic polynomial of the linearrecurrence relation defining t,
• a periodic fluctuation.
Definition: Aβ = (Aβ(n))n≥0 is the summatory function of Sβ .
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 31
![Page 74: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/74.jpg)
Summatory functions of b-regular sequences algebraic or analytic methods
New method: to tackle the behavior of the summatory functionNew method: As = (As(n))n≥0 of a regular sequence s = (s(n))n≥0
• Find r = (r(n))n≥0 and t = (t(n))n≥0
• each satisfying a linear recurrence relation,• verifying As ◦ r = t.
• Recurrence relation for s recurrence relation for As in which t is involved.• Relevant representations of As in some exotic numeration system
associated with t.“Exotic” possibly unbounded coefficients• The behavior of As depends on
• the dominant root of the characteristic polynomial of the linearrecurrence relation defining t,
• a periodic fluctuation.
Definition: Aβ = (Aβ(n))n≥0 is the summatory function of Sβ .
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 31
![Page 75: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/75.jpg)
Summatory functions of b-regular sequences algebraic or analytic methods
New method: to tackle the behavior of the summatory functionNew method: As = (As(n))n≥0 of a regular sequence s = (s(n))n≥0
• Find r = (r(n))n≥0 and t = (t(n))n≥0
• each satisfying a linear recurrence relation,• verifying As ◦ r = t.
• Recurrence relation for s recurrence relation for As in which t is involved.• Relevant representations of As in some exotic numeration system
associated with t.“Exotic” possibly unbounded coefficients• The behavior of As depends on
• the dominant root of the characteristic polynomial of the linearrecurrence relation defining t,
• a periodic fluctuation.
Definition: Aβ = (Aβ(n))n≥0 is the summatory function of Sβ .
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 31
![Page 76: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/76.jpg)
Summatory functions of b-regular sequences algebraic or analytic methods
New method: to tackle the behavior of the summatory functionNew method: As = (As(n))n≥0 of a regular sequence s = (s(n))n≥0
• Find r = (r(n))n≥0 and t = (t(n))n≥0
• each satisfying a linear recurrence relation,• verifying As ◦ r = t.
• Recurrence relation for s recurrence relation for As in which t is involved.
• Relevant representations of As in some exotic numeration systemassociated with t.“Exotic” possibly unbounded coefficients• The behavior of As depends on
• the dominant root of the characteristic polynomial of the linearrecurrence relation defining t,
• a periodic fluctuation.
Definition: Aβ = (Aβ(n))n≥0 is the summatory function of Sβ .
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 31
![Page 77: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/77.jpg)
Summatory functions of b-regular sequences algebraic or analytic methods
New method: to tackle the behavior of the summatory functionNew method: As = (As(n))n≥0 of a regular sequence s = (s(n))n≥0
• Find r = (r(n))n≥0 and t = (t(n))n≥0
• each satisfying a linear recurrence relation,• verifying As ◦ r = t.
• Recurrence relation for s recurrence relation for As in which t is involved.• Relevant representations of As in some exotic numeration system
associated with t.“Exotic” possibly unbounded coefficients
• The behavior of As depends on• the dominant root of the characteristic polynomial of the linear
recurrence relation defining t,• a periodic fluctuation.
Definition: Aβ = (Aβ(n))n≥0 is the summatory function of Sβ .
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 31
![Page 78: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/78.jpg)
Summatory functions of b-regular sequences algebraic or analytic methods
New method: to tackle the behavior of the summatory functionNew method: As = (As(n))n≥0 of a regular sequence s = (s(n))n≥0
• Find r = (r(n))n≥0 and t = (t(n))n≥0
• each satisfying a linear recurrence relation,• verifying As ◦ r = t.
• Recurrence relation for s recurrence relation for As in which t is involved.• Relevant representations of As in some exotic numeration system
associated with t.“Exotic” possibly unbounded coefficients• The behavior of As depends on
• the dominant root of the characteristic polynomial of the linearrecurrence relation defining t,
• a periodic fluctuation.
Definition: Aβ = (Aβ(n))n≥0 is the summatory function of Sβ .
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 31
![Page 79: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/79.jpg)
Summatory functions of b-regular sequences algebraic or analytic methods
New method: to tackle the behavior of the summatory functionNew method: As = (As(n))n≥0 of a regular sequence s = (s(n))n≥0
• Find r = (r(n))n≥0 and t = (t(n))n≥0
• each satisfying a linear recurrence relation,• verifying As ◦ r = t.
• Recurrence relation for s recurrence relation for As in which t is involved.• Relevant representations of As in some exotic numeration system
associated with t.“Exotic” possibly unbounded coefficients• The behavior of As depends on
• the dominant root of the characteristic polynomial of the linearrecurrence relation defining t,
• a periodic fluctuation.
Definition: Aβ = (Aβ(n))n≥0 is the summatory function of Sβ .Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 31
![Page 80: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/80.jpg)
Results in the integer base case
• The sequence (Ab(n))n≥0 is b-regular.
• Matrix representation for the sequence (Ab(n))n≥0.• For all n ≥ 0, Ab( bn︸︷︷︸
r(n)
) = (2b− 1)n︸ ︷︷ ︸t(n)
.
• Recurrence relations for Ab involving 2b− 1.
For all a, a′ ∈ {1, . . . , b− 1} with a 6= a′, all ` ≥ 1 and all r ∈ {0, . . . , b`−1}
Ab(ab` + r) = (2b− 2) · (2a− 1) · (2b− 1)`−1 +Ab(ab
`−1 + r) +Ab(r),
Ab(ab` + ab`−1 + r) = (4ab− 2a− 2b+ 2) · (2b− 1)`−1 + 2Ab(ab
`−1 + r)−Ab(r),
Ab(ab`+a′b`−1+r) =
(4ab− 4a− 2b+ 3) · (2b− 1)`−1 +Ab(ab
`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ < a;
(4ab− 4a− 2b+ 2) · (2b− 1)`−1 +Ab(ab`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ > a.
• (2b−1)-decomposition of Ab: mixing the base b and the base 2b−1numeration systems
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 32
![Page 81: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/81.jpg)
Results in the integer base case
• The sequence (Ab(n))n≥0 is b-regular.• Matrix representation for the sequence (Ab(n))n≥0.
• For all n ≥ 0, Ab( bn︸︷︷︸r(n)
) = (2b− 1)n︸ ︷︷ ︸t(n)
.
• Recurrence relations for Ab involving 2b− 1.
For all a, a′ ∈ {1, . . . , b− 1} with a 6= a′, all ` ≥ 1 and all r ∈ {0, . . . , b`−1}
Ab(ab` + r) = (2b− 2) · (2a− 1) · (2b− 1)`−1 +Ab(ab
`−1 + r) +Ab(r),
Ab(ab` + ab`−1 + r) = (4ab− 2a− 2b+ 2) · (2b− 1)`−1 + 2Ab(ab
`−1 + r)−Ab(r),
Ab(ab`+a′b`−1+r) =
(4ab− 4a− 2b+ 3) · (2b− 1)`−1 +Ab(ab
`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ < a;
(4ab− 4a− 2b+ 2) · (2b− 1)`−1 +Ab(ab`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ > a.
• (2b−1)-decomposition of Ab: mixing the base b and the base 2b−1numeration systems
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 32
![Page 82: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/82.jpg)
Results in the integer base case
• The sequence (Ab(n))n≥0 is b-regular.• Matrix representation for the sequence (Ab(n))n≥0.• For all n ≥ 0, Ab( bn︸︷︷︸
r(n)
) = (2b− 1)n︸ ︷︷ ︸t(n)
.
• Recurrence relations for Ab involving 2b− 1.
For all a, a′ ∈ {1, . . . , b− 1} with a 6= a′, all ` ≥ 1 and all r ∈ {0, . . . , b`−1}
Ab(ab` + r) = (2b− 2) · (2a− 1) · (2b− 1)`−1 +Ab(ab
`−1 + r) +Ab(r),
Ab(ab` + ab`−1 + r) = (4ab− 2a− 2b+ 2) · (2b− 1)`−1 + 2Ab(ab
`−1 + r)−Ab(r),
Ab(ab`+a′b`−1+r) =
(4ab− 4a− 2b+ 3) · (2b− 1)`−1 +Ab(ab
`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ < a;
(4ab− 4a− 2b+ 2) · (2b− 1)`−1 +Ab(ab`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ > a.
• (2b−1)-decomposition of Ab: mixing the base b and the base 2b−1numeration systems
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 32
![Page 83: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/83.jpg)
Results in the integer base case
• The sequence (Ab(n))n≥0 is b-regular.• Matrix representation for the sequence (Ab(n))n≥0.• For all n ≥ 0, Ab( bn︸︷︷︸
r(n)
) = (2b− 1)n︸ ︷︷ ︸t(n)
.
• Recurrence relations for Ab involving 2b− 1.
For all a, a′ ∈ {1, . . . , b− 1} with a 6= a′, all ` ≥ 1 and all r ∈ {0, . . . , b`−1}
Ab(ab` + r) = (2b− 2) · (2a− 1) · (2b− 1)`−1 +Ab(ab
`−1 + r) +Ab(r),
Ab(ab` + ab`−1 + r) = (4ab− 2a− 2b+ 2) · (2b− 1)`−1 + 2Ab(ab
`−1 + r)−Ab(r),
Ab(ab`+a′b`−1+r) =
(4ab− 4a− 2b+ 3) · (2b− 1)`−1 +Ab(ab
`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ < a;
(4ab− 4a− 2b+ 2) · (2b− 1)`−1 +Ab(ab`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ > a.
• (2b−1)-decomposition of Ab: mixing the base b and the base 2b−1numeration systems
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 32
![Page 84: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/84.jpg)
Results in the integer base case
• The sequence (Ab(n))n≥0 is b-regular.• Matrix representation for the sequence (Ab(n))n≥0.• For all n ≥ 0, Ab( bn︸︷︷︸
r(n)
) = (2b− 1)n︸ ︷︷ ︸t(n)
.
• Recurrence relations for Ab involving 2b− 1.
For all a, a′ ∈ {1, . . . , b− 1} with a 6= a′, all ` ≥ 1 and all r ∈ {0, . . . , b`−1}
Ab(ab` + r) = (2b− 2) · (2a− 1) · (2b− 1)`−1 +Ab(ab
`−1 + r) +Ab(r),
Ab(ab` + ab`−1 + r) = (4ab− 2a− 2b+ 2) · (2b− 1)`−1 + 2Ab(ab
`−1 + r)−Ab(r),
Ab(ab`+a′b`−1+r) =
(4ab− 4a− 2b+ 3) · (2b− 1)`−1 +Ab(ab
`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ < a;
(4ab− 4a− 2b+ 2) · (2b− 1)`−1 +Ab(ab`−1 + r)
+2Ab(a′b`−1 + r)− 2Ab(r), if a′ > a.
• (2b−1)-decomposition of Ab: mixing the base b and the base 2b−1numeration systems
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 32
![Page 85: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/85.jpg)
Illustration in base 2
2n+1 2n+2
α
α ∈ [0, 1) with d2(α) = d1d2d3 · · · en(α) = val2(1d1 · · · dn1)
Example: α = π − 3
n en(π − 3) 3dec(A2(en(π − 3)))
1 5 2 72 9 2 2 83 19 2 6 −2 54 37 2 6 −6 6 155 73 2 6 −6 2 8 316 147 2 6 −6 2 24 −8 147 293 2 6 −6 2 24 −24 22 538 585 2 6 −6 2 24 −24 6 30 1169 1169 2 6 −6 2 24 −24 6 30 30 13110 2337 2 6 −6 2 24 −24 6 30 30 30 146
3dec(A2(en(α)))n→+∞−−−−−→ a0(α)a1(α)a2(α) · · ·
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 33
![Page 86: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/86.jpg)
Step functions: α ∈ [0, 1) 7→ φn(α) = A2(en(α))/3log2(en(α))
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
φn(α) =
1
31+{log2(en(α))}
n∑i=0
ai(en(α))
3iif 0 ≤ α < 1
2
1
3{log2(en(α))}
n+1∑i=0
ai(en(α))
3iif 1
2≤ α < 1
(φn)n≥1 uniformly converges to the function
Φ2(α) =
1
31+log2(α+1)
+∞∑i=0
ai(α)
3iif 0 ≤ α < 1
2
1
3log2(α+1)
+∞∑i=0
ai(α)
3iif 1
2≤ α < 1
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 34
![Page 87: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/87.jpg)
• Φ2 is continuous over [0, 1) s.t Φ2(0) = 1 and limα→1−
Φ2(α) = 1.
Φ2
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
• Define H2 with the help of Φ2.Then H2 is continuous and 1-periodic s.t. for all large enough n
A2(n) = 3log2 n H2(log2 n).
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 35
![Page 88: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/88.jpg)
Results in the integer base case (continued)TheoremThere exists a continuous and 1-periodic function Hb s.t. for all largeenough n
Ab(n) =
n−1∑
j=0
Sb(j) = (2b− 1)logb n Hb(logb n).
H2
0.0 0.2 0.4 0.6 0.8 1.0
0.98
1.00
1.02
1.04
1.06
H3
0.0 0.2 0.4 0.6 0.8 1.0
1.00
1.02
1.04
1.06
1.08
H4
0.0 0.2 0.4 0.6 0.8 1.0
1.00
1.05
1.10
H5
0.0 0.2 0.4 0.6 0.8 1.0
1.00
1.05
1.10
1.15
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 36
![Page 89: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/89.jpg)
Results in the Fibonacci case
Let (B(n))n≥0 be defined by B(0) = 1, B(1) = 3, B(2) = 6, andB(n+ 3) = 2B(n+ 2) +B(n+ 1)−B(n) for all n ≥ 0.
• For all n ≥ 0, Aϕ(F (n)− 1︸ ︷︷ ︸r(n)
) = B(n)︸ ︷︷ ︸t(n)
.
• Recurrence relations for Aϕ involving B.
If 0 ≤ r < F (`− 2), then
Aϕ(F (`) + r) = B(`)−B(`− 1) +Aϕ(F (`− 1) + r) +Aϕ(r).
If F (`− 2) ≤ r < F (`− 1), then
Aϕ(F (`) + r) = 2B(`)−B(`− 1)−B(`− 2) + 2Aϕ(r).
• B-decomposition of Aϕ: mixing the Fibonacci numeration systemand the numeration system based on B
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 37
![Page 90: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/90.jpg)
Results in the Fibonacci case (continued)TheoremLet λ be the dominant root of the characteristic polynomialPB(X) = X3 − 2X2 −X + 1 of B.Let c be a constant s.t. lim
n→+∞B(n)/λn = c.
There exists a continuous and 1-periodic function G s.t., for all largeenough n,
Aϕ(n) =
n∑
j=0
Sϕ(j) = c λlogF nG(logF n) + o(λblogF nc).
0.0 0.2 0.4 0.6 0.8 1.0
1.00
1.02
1.04
1.06
Remark: There is an error term, but the method allows us to deal withgeneralized regular sequences.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 38
![Page 91: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/91.jpg)
Summary
Part I: Pascal triangle• Generalization to Parry numeration systems• Description of the limit set Lβ (segments, maps c and h)• Works for r mod p (p prime)
Part II: Sequences counting scattered subwords• b-regularity of Sb, Fibonacci-regularity of Sϕ• Method using tries of scattered subwords• Matrix representations• In base 2: link with the Farey tree
Part III: Summatory functions• Asymptotics for Ab and Aϕ• Method mixing numeration systems (exotic decompositions)
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 39
![Page 92: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/92.jpg)
Perspectives: natural open problems/questions
Part I: Generalized Pascal triangles• Extension to other numeration systems/languages (conditions)• Other colorings (generalizations of Lucas’ theorem)• Properties of Lβ : Hausdorff dimension, Minkowski dimension,
Properties of Lβ : Hölder exponent, Lebesgue measure, etc.• Iterated functions systems (IFS)• Comparison/Classification of limit sets
Part II: Regularity of sequences counting scattered subwords• Extension to other numeration systems/languages (conditions)• b-regularity of (Sϕ(n))n≥0 (Cobham-like theorem)• Extension of the relation between (S2(n))n≥0 and the Farey tree
Part III: Asymptotics of summatory functions• Extension to other numeration systems/languages (conditions)• Regularity of (Aβ(n))n≥0
• Differentiability of periodic fluctuations• Comparison/Classification of periodic fluctuations
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 40
![Page 93: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/93.jpg)
References
• J.-P. Allouche and J. Shallit, The ring of k-regular sequences, Theoret. Com-put. Sci., 98 (1992), 163–197.
• J.-P. Allouche and J. Shallit, Automatic sequences. Theory, applications, gen-eralizations, Cambridge University Press, Cambridge, 2003.
• J.-P. Allouche, K. Scheicher, and R. F. Tichy, Regular maps in generalizednumber systems, Math. Slovaca, 50 (2000), 41–58.
• V. Berthé and M. Rigo (editors), Combinatorics, automata and number theory,volume 135 of Encyclopedia of Mathematics and its Applications, CambridgeUniversity Press, Cambridge, 2010.
• H. Delange, Sur la fonction sommatoire de la fonction “somme des chiffres”,Enseignement Math. (2), 21 (1975), 31–47.
• P. Dumas, Joint spectral radius, dilation equations, and asymptotic behaviorof radix-rational sequences, Linear Algebra Appl., 438 (2013), 2107–2126.
• P. Dumas, Asymptotic expansions for linear homogeneous divide-and-conquerrecurrences: algebraic and analytic approaches collated, Theoret. Comput.Sci., 548 (2014), 25–53.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 41
![Page 94: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/94.jpg)
References
• K. J. Falconer, The geometry of fractal sets, volume 85 of Cambridge Tractsin Mathematics, Cambridge University Press, Cambridge, 1986.
• J. Leroy, M. Rigo, and M. Stipulanti, Generalized Pascal triangle for binomialcoefficients of words, Adv. in Appl. Math., 80 (2016), 24–47.
• J. Leroy, M. Rigo, and M. Stipulanti, Counting the number of non-zero co-efficients in rows of generalized Pascal triangles, Discrete Math., 340 (2017),862–881.
• J. Leroy, M. Rigo, and M. Stipulanti, Behavior of digital sequences throughexotic numeration systems, Electron. J. Combin., 24 (2017), Paper No. 1.44,36 pages.
• J. Leroy, M. Rigo, and M. Stipulanti, Counting subword occurrences in base-bexpansions, Integers, 18A (2018), Paper No. A13, 32 pages.
• M. Lothaire, Combinatorics on Words, Cambridge Mathematical Library, Cam-bridge University Press, Cambridge, 1997.
• M. Lothaire, Algebraic Combinatorics on Words, volume 90 of Encyclopediaof Mathematics and its Applications, Cambridge University Press, Cambridge,2002.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 42
![Page 95: Extensions of the Pascal Triangle to Words, and … › ... › 233348 › 4 › Beamer_defense_these.pdfSpecialcaseofgeneralizedPascaltriangles Parrynumber 2R>1 linearnumerationsystemU](https://reader030.fdocuments.net/reader030/viewer/2022041114/5f237ff0f589be0c144864c3/html5/thumbnails/95.jpg)
References
• É. Lucas, Théorie des fonctions numériques simplement périodiques, Amer. J.Math., 1 (1878), 197–240.
• F. von Haeseler, H.-O. Peitgen and G. Skordev, Pascal’s triangle, dynamicalsystems and attractors, Ergodic Theory Dynam. Systems, 12 (1992), 479–486.
• M. Rigo and A. Maes, More on generalized automatic sequences, J. Autom.Lang. Comb., 7 (2002), 351–376.
• M. Rigo, Formal languages, automata and numeration systems, Networks andTelecommunications Series, ISTE, London, John Wiley & Sons, Inc., Hoboken,NJ, 2014.
• J. Shallit, A generalization of automatic sequences, Theoret. Comput. Sci.,61 (1988), 1–16.
• M. Stipulanti, Convergence of Pascal-like Triangles in Parry–Bertrand Numer-ation Systems, Theoret. Comput. Sci., 758 (2019), 42–60.
• É. Zeckendorf, Représentation des nombres naturels par une somme de nom-bres de Fibonacci ou de nombres de Lucas, Bull. Soc. Roy. Sci. Liège, 41(1972), 179–182.
Generalized Pascal Triangles, and Related Counting Problems M. Stipulanti (ULiège) 43