Classification of Canons
Home Page
Harald Fripertinger
Title Page
Karl-Franzens-Universität Graz
Contents
Page 1 of 35
Go Back
Full Screen
Close
Quit
Musical Enumeration Theory Applied to the
Classification of Canons
Home Page
Harald Fripertinger
Title Page
Karl-Franzens-Universität Graz
Contents
Page 1 of 35
Full Screen
Close
Quit
Musical Enumeration Theory Applied to the
Classification of Canons
Home Page
Harald Fripertinger
Title Page
Karl-Franzens-Universität Graz
Contents
Page 1 of 35
Close
Quit
Musical Enumeration Theory Applied to the
Classification of Canons
Home Page
Harald Fripertinger
Title Page
Karl-Franzens-Universität Graz
Contents
Page 1 of 35
Quit
Musical Enumeration Theory Applied to the
Classification of Canons
Home Page
Harald Fripertinger
Title Page
Karl-Franzens-Universität Graz
Contents
Page 1 of 35
Home Page
Harald Fripertinger
Title Page
Karl-Franzens-Universität Graz
Contents
Page 1 of 35
Title Page
Contents
JJ II
J I
Page 2 of 35
Go Back
Full Screen
Close
Quit
Discrete Structures
Contents
JJ II
J I
Page 2 of 35
Go Back
Full Screen
Close
Quit
Discrete Structures
JJ II
J I
Page 2 of 35
Go Back
Full Screen
Close
Quit
Discrete Structures
JJ II
J I
Page 2 of 35
Go Back
Full Screen
Close
Quit
Discrete Structures
J I
Page 2 of 35
Go Back
Full Screen
Close
Quit
Discrete Structures
Page 2 of 35
Go Back
Full Screen
Close
Quit
Discrete Structures
Close
Quit
Classification of Discrete Structures I
JJ II
J I
Page 3 of 35
Go Back
Full Screen
Close
Quit
Classification of Discrete Structures I
J I
Page 3 of 35
Go Back
Full Screen
Close
Quit
Classification of Discrete Structures I
Page 3 of 35
Go Back
Full Screen
Close
Quit
Classification of Discrete Structures I
Go Back
Full Screen
Close
Quit
Classification of Discrete Structures I
Go Back
step 4: Generate the objects of a discrete structure uniformly at random.
Full Screen
Close
Quit
Labelled Graphs
A simple labelled graph is described by a finite set of labelled vertices V
and by a set of edges connecting two vertices. An edge is usually
described as a 2-set of vertices, indicating which two vertices are
Home Page
Contents
JJ II
J I
Page 4 of 35
Go Back
Full Screen
Close
Quit
Labelled Graphs
A simple labelled graph is described by a finite set of labelled vertices V
and by a set of edges connecting two vertices. An edge is usually
described as a 2-set of vertices, indicating which two vertices are
Home Page
3u 4
u
JJ II
the graph u u
on 4 vertices.
2 1
J I
Page 4 of 35
Go Back
Full Screen
Close
Quit
Labelled Graphs
A simple labelled graph is described by a finite set of labelled vertices V
and by a set of edges connecting two vertices. An edge is usually
described as a 2-set of vertices, indicating which two vertices are
Home Page
3u 4
u
JJ II
the graph u u
on 4 vertices. Usually the labelling of the vertices is not
2 1
J I
important, whence this graph is considered to be essentially the same as
Page 4 of 35 2u 3
u
u u
.
Go Back
1 4
Full Screen
Close
Quit
Labelled Graphs
A simple labelled graph is described by a finite set of labelled vertices V
and by a set of edges connecting two vertices. An edge is usually
described as a 2-set of vertices, indicating which two vertices are
Home Page
3u 4
u
JJ II
the graph u u
on 4 vertices. Usually the labelling of the vertices is not
2 1
J I
important, whence this graph is considered to be essentially the same as
Page 4 of 35 2u 3
u
u u
. All graphs obtained by relabelling of one graph are collected to an
Go Back
1 4
Full Screen
unlabelled graph, an equivalence class of labelled graphs. In our example
u u
Close
we get u u
.
Quit
Unlabelled Graphs on 4 Vertices
J I
Page 5 of 35
Go Back
Full Screen
Close
Quit
Unlabelled Graphs on 4 Vertices
Page 5 of 35
Go Back
Full Screen
Close
Quit
Unlabelled Graphs on 4 Vertices
Go Back
Full Screen
Close
Quit
Unlabelled Graphs on 4 Vertices
Close
Quit
Unlabelled Graphs on 4 Vertices
Title Page
Contents
JJ II
J I
Page 6 of 35
Go Back
Full Screen
Close
Quit
Classification of Discrete Structures II
Contents
JJ II
J I
Page 6 of 35
Go Back
Full Screen
Close
Quit
Classification of Discrete Structures II
Contents
3. If the objects of a discrete structure are equivalence classes, introduce
a suitable group action on X to describe the equivalence classes as orbits.
JJ II
J I
Page 6 of 35
Go Back
Full Screen
Close
Quit
Classification of Discrete Structures II
Contents
3. If the objects of a discrete structure are equivalence classes, introduce
a suitable group action on X to describe the equivalence classes as orbits.
JJ II
J I
4. Proceed with the four steps for classification (enumeration,
construction, random generation) as described above.
Page 6 of 35
Go Back
Full Screen
Close
Quit
Group Actions
∗: G × X → X ∗ (g, x) 7→ g ∗ x
Contents
JJ II
J I
Page 7 of 35
Go Back
Full Screen
Close
Quit
Group Actions
∗: G × X → X ∗ (g, x) 7→ g ∗ x
Contents
JJ II
such that
J I (g1g2) ∗ x = g1 ∗ (g2 ∗ x) g1, g2 ∈ G, x ∈ X
Page 7 of 35
Go Back
Full Screen
Close
Quit
Group Actions
∗: G × X → X ∗ (g, x) 7→ g ∗ x
Contents
JJ II
such that
J I (g1g2) ∗ x = g1 ∗ (g2 ∗ x) g1, g2 ∈ G, x ∈ X
Page 7 of 35
and
1∗x = x x ∈ X.
Go Back
Full Screen
Close
Quit
Group Actions
∗: G × X → X ∗ (g, x) 7→ g ∗ x
Contents
JJ II
such that
J I (g1g2) ∗ x = g1 ∗ (g2 ∗ x) g1, g2 ∈ G, x ∈ X
Page 7 of 35
and
1∗x = x x ∈ X.
Go Back
Close
Quit
Group Actions
∗: G × X → X ∗ (g, x) 7→ g ∗ x
Contents
JJ II
such that
J I (g1g2) ∗ x = g1 ∗ (g2 ∗ x) g1, g2 ∈ G, x ∈ X
Page 7 of 35
and
1∗x = x x ∈ X.
Go Back
Quit
Group Actions
∗: G × X → X ∗ (g, x) 7→ g ∗ x
Contents
JJ II
such that
J I (g1g2) ∗ x = g1 ∗ (g2 ∗ x) g1, g2 ∈ G, x ∈ X
Page 7 of 35
and
1∗x = x x ∈ X.
Go Back
Quit
Orbits under Group Actions
Contents
JJ II
J I
Page 8 of 35
Go Back
Full Screen
Close
Quit
Orbits under Group Actions
JJ II
G(x) = {gx | g ∈ G} .
J I
Page 8 of 35
Go Back
Full Screen
Close
Quit
Orbits under Group Actions
JJ II
G(x) = {gx | g ∈ G} .
J I
Go Back
G\\X := {G(x) | x ∈ X} .
Full Screen
Close
Quit
Orbits under Group Actions
JJ II
G(x) = {gx | g ∈ G} .
J I
Go Back
G\\X := {G(x) | x ∈ X} .
Full Screen
Close
Theorem. The equivalence classes of any equivalence relation can be
Quit
represented as orbits under a suitable group action.
Stabilizers and Fixed Points
Gx := {g ∈ G | gx = x} .
Contents
JJ II It is a subgroup of G.
J I
Page 9 of 35
Go Back
Full Screen
Close
Quit
Stabilizers and Fixed Points
Gx := {g ∈ G | gx = x} .
Contents
JJ II It is a subgroup of G.
J I
Lagrange Theorem. If GX is a finite group action then the size of the
Page 9 of 35
orbit of x ∈ X equals
|G|
Go Back |G(x)| = .
|Gx|
Full Screen
Close
Quit
Stabilizers and Fixed Points
Gx := {g ∈ G | gx = x} .
Contents
JJ II It is a subgroup of G.
J I
Lagrange Theorem. If GX is a finite group action then the size of the
Page 9 of 35
orbit of x ∈ X equals
|G|
Go Back |G(x)| = .
|Gx|
Full Screen
Finally, the set of all fixed points of g ∈ G is denoted by
Close
Quit
Xg := {x ∈ X | gx = x} .
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
JJ II
J I
Page 10 of 35
Go Back
Full Screen
Close
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
J I
Page 10 of 35
Go Back
Full Screen
Close
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
J I
|G\\X| =
Page 10 of 35
Go Back
Full Screen
Close
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
Go Back
Full Screen
Close
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
Proof.
∑ |Xg| =
Go Back
Close
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
Proof.
∑ |Xg| = ∑ ∑
Go Back
1=
Full Screen g∈G g∈G x:gx=x
Close
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
Proof.
∑ |Xg| = ∑ ∑ ∑ ∑
Go Back
1= 1=
Full Screen g∈G g∈G x:gx=x x∈X g:gx=x
Close
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
Proof.
∑ |Xg| = ∑ ∑ ∑ ∑ ∑ |Gx| =
Go Back
1= 1=
Full Screen g∈G g∈G x:gx=x x∈X g:gx=x x∈X
Close
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
Proof.
∑ |Xg| = ∑ ∑ ∑ ∑ ∑ |Gx| =
Go Back
1= 1=
Full Screen g∈G g∈G x:gx=x x∈X g:gx=x x∈X
1
Close
= |G| ∑ =
Quit
x∈X |G(x)|
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
Proof.
∑ |Xg| = ∑ ∑ ∑ ∑ ∑ |Gx| =
Go Back
1= 1=
Full Screen g∈G g∈G x:gx=x x∈X g:gx=x x∈X
1 1
Close
= |G| ∑ = |G| ∑ ∑ =
x∈X |G(x)| ω∈G\\X x∈ω
|ω|
Quit
Enumeration under Group Actions
Let GX be finite group action. The main tool for determining the number of
Home Page
different orbits is the
Title Page
Proof.
∑ |Xg| = ∑ ∑ ∑ ∑ ∑ |Gx| =
Go Back
1= 1=
Full Screen g∈G g∈G x:gx=x x∈X g:gx=x x∈X
1 1
Close
= |G| ∑ = |G| ∑ ∑ = |G||G\\X|.
x∈X |G(x)| ω∈G\\X x∈ω
|ω|
Quit
Symmetry types of mappings
Contents
JJ II
J I
Page 11 of 35
Go Back
Full Screen
Close
Quit
Symmetry types of mappings
— Then G acts on Y X by
Contents
JJ II
G ×Y X → Y X , (g, f ) 7→ f ◦ ḡ−1.
J I
Page 11 of 35
Go Back
Full Screen
Close
Quit
Symmetry types of mappings
— Then G acts on Y X by
Contents
JJ II
G ×Y X → Y X , (g, f ) 7→ f ◦ ḡ−1.
J I
Close
Quit
Symmetry types of mappings
— Then G acts on Y X by
Contents
JJ II
G ×Y X → Y X , (g, f ) 7→ f ◦ ḡ−1.
J I
Quit
(H × G) ×Y X → Y X , ((h, g), f ) 7→ h̄ ◦ f ◦ ḡ−1.
Discrete structures in Music Theory
JJ II
J I
Page 12 of 35
Go Back
Full Screen
Close
Quit
Discrete structures in Music Theory
Full Screen
Close
Quit
Discrete structures in Music Theory
Close
Quit
Discrete structures in Music Theory
Quit
i ∈ Zn there is exactly one P ∈ π with i ∈ P.
Discrete structures in Music Theory (cont.)
where f (i) is the pitch class of the tone which occurs in i-th position in the
Contents
tone-row. Usually two tone-rows f1, f2 are considered to be equivalent if
JJ II f1 can be constructed by transposing, inversion and retrograde inversion
R of f2. Thus the similarity classes of tone-rows are the Dn × hRi orbits
J I
on the set of these bijections.
Page 13 of 35
Go Back
Full Screen
Close
Quit
Discrete structures in Music Theory (cont.)
where f (i) is the pitch class of the tone which occurs in i-th position in the
Contents
tone-row. Usually two tone-rows f1, f2 are considered to be equivalent if
JJ II f1 can be constructed by transposing, inversion and retrograde inversion
R of f2. Thus the similarity classes of tone-rows are the Dn × hRi orbits
J I
on the set of these bijections.
Page 13 of 35
In the most simple model of a motive, each note is described by its onset
Go Back
and by its pitch. In other words, we put a coordinate system over a line of
Full Screen
notes. If there are exactly m possible onsets and n pitch classes, a
Close k-motive is considered to be a k-subset of Zm × Zn. In the case m = n,
Mazzola showed that the group of all affine mappings on Zn2 is a musically
Quit
motivated group.
Rhythms
Contents
JJ II
J I
Page 14 of 35
Go Back
Full Screen
Close
Quit
Rhythms
Contents
We also assume that we found a subdivision of the rhythm, i.e. a regular
JJ II pulsation, into equidistant beats such that all rhythmical events coincide
with some of these beats. If the rhythm is covered by a pulsation of n
J I
beats, then it can be described as a 0, 1-vector of length n, whence as a
Page 14 of 35
function f : {0, . . . , n − 1} → {0, 1}.
Go Back
Contents
We also assume that we found a subdivision of the rhythm, i.e. a regular
JJ II pulsation, into equidistant beats such that all rhythmical events coincide
with some of these beats. If the rhythm is covered by a pulsation of n
J I
beats, then it can be described as a 0, 1-vector of length n, whence as a
Page 14 of 35
function f : {0, . . . , n − 1} → {0, 1}.
Go Back
Contents
JJ II
If we write f as a vector f = ( f (0), . . . , f (n − 1)), then
f ◦ πnj = ( f (n + j), . . . , f (n − 1), f (0), . . . , f (n + j − 1)). Hence, the orbit
J I
Cn( f ) consists of all cyclic shifts of f .
Page 15 of 35
Go Back
Full Screen
Close
Quit
0,1-Vectors
Contents
JJ II
If we write f as a vector f = ( f (0), . . . , f (n − 1)), then
f ◦ πnj = ( f (n + j), . . . , f (n − 1), f (0), . . . , f (n + j − 1)). Hence, the orbit
J I
Cn( f ) consists of all cyclic shifts of f .
n
Page 15 of 35
Using the natural order 0 < 1, the set {0, 1} is totally ordered by the
n
Go Back
lexicographical order. For f , g ∈ {0, 1} we say
Close
representative.
0,1-Vectors (cont.)
For example the orbit of f = (10011010) under C8 contains the vectors
(10011010), (00110101), (01101010), (11010100), (10101001),
(01010011), (10100110), and (01001101).
Home Page
Title Page
Contents
JJ II
J I
Page 16 of 35
Go Back
Full Screen
Close
Quit
0,1-Vectors (cont.)
For example the orbit of f = (10011010) under C8 contains the vectors
(10011010), (00110101), (01101010), (11010100), (10101001),
(01010011), (10100110), and (01001101).
Home Page
n
The stabilizer of f ∈ {0, 1} is a subgroup of Cn, whence again a cyclic
group. We call f acyclic if its stabilizer consists of the identity only. If f is
Title Page
Go Back
Full Screen
Close
Quit
0,1-Vectors (cont.)
For example the orbit of f = (10011010) under C8 contains the vectors
(10011010), (00110101), (01101010), (11010100), (10101001),
(01010011), (10100110), and (01001101).
Home Page
n
The stabilizer of f ∈ {0, 1} is a subgroup of Cn, whence again a cyclic
group. We call f acyclic if its stabilizer consists of the identity only. If f is
Title Page
Go Back The vector (01100110) has a nontrivial stabilizer, since π48 6= id is also
Full Screen
contained in it.
Close
Quit
0,1-Vectors (cont.)
For example the orbit of f = (10011010) under C8 contains the vectors
(10011010), (00110101), (01101010), (11010100), (10101001),
(01010011), (10100110), and (01001101).
Home Page
n
The stabilizer of f ∈ {0, 1} is a subgroup of Cn, whence again a cyclic
group. We call f acyclic if its stabilizer consists of the identity only. If f is
Title Page
Go Back The vector (01100110) has a nontrivial stabilizer, since π48 6= id is also
Full Screen
contained in it.
Close
We identify a 0,1-vector f with the set f −1({1}) of pre-images of 1. Then
f is the characteristic function of f −1({1}). In this case we also assume
Quit
J I hK − Ki := hk − l | k, l ∈ Ki = Zn.
Page 17 of 35
Go Back
Full Screen
Close
Quit
Canons
J I hK − Ki := hk − l | k, l ∈ Ki = Zn.
Page 17 of 35
Go Back
Two canons {V1, . . . ,Vt } and {W1, . . . ,Ws} are called isomorphic if s = t
and if there exists some σ ∈ Cn and a permutation τ in the symmetric
group St such that σ(Vi) = Wτ(i) for 1 ≤ i ≤ t .
Full Screen
Close
Quit
Inner and Outer Rhythm of a Canon
The canon {V1, . . . ,Vt } can be described as a pair (V1, f ), where V1 is the
Home Page
inner and f the outer rhythm of the canon. The inner rhythm describes
Title Page
the rhythm of any voice. The outer rhythm determines how the different
voices are distributed over the n beats of a canon.
Contents
JJ II
J I
Page 18 of 35
Go Back
Full Screen
Close
Quit
Inner and Outer Rhythm of a Canon
The canon {V1, . . . ,Vt } can be described as a pair (V1, f ), where V1 is the
Home Page
inner and f the outer rhythm of the canon. The inner rhythm describes
Title Page
the rhythm of any voice. The outer rhythm determines how the different
voices are distributed over the n beats of a canon.
Contents
Page 18 of 35 10011010
Go Back
01010011
11010100
Full Screen
Close
Hence the outer rhythm of this canon is f = (10010100).
Quit
Inner and Outer Rhythm of a Canon (cont.)
JJ II
J I
Page 19 of 35
Go Back
Full Screen
Close
Quit
Inner and Outer Rhythm of a Canon (cont.)
J I
Page 19 of 35
Go Back
Full Screen
Close
Quit
Inner and Outer Rhythm of a Canon (cont.)
Go Back
length n. The pair (L, f ) does not describe a canon if and only if there
Full Screen
exists an integer d > 1 such that d | n, d | k − l for all k, l ∈ L, and d | k − l
for all k, l ∈ f .
Close
Quit
Enumeration of Canons
Kn := ∑ µ(d)λ(n/d)α(n/d),
Home Page
Title Page
d|n
JJ II
1
J I λ(r) = ∑ µ(s)2r/s for r > 1,
r s|r
Page 20 of 35
1
Go Back
α(r) = ∑ ϕ(s)2r/s − 1 for r ≥ 1,
r s|r
Full Screen
Quit
Enumeration of Canons
Kn := ∑ µ(d)λ(n/d)α(n/d),
Home Page
Title Page
d|n
JJ II
1
J I λ(r) = ∑ µ(s)2r/s for r > 1,
r s|r
Page 20 of 35
1
Go Back
α(r) = ∑ ϕ(s)2r/s − 1 for r ≥ 1,
r s|r
Full Screen
The ideas of this proof and of the previous Lemma can be used in order
Quit
10011010
Home Page
01010011
11010100
Title Page
Contents
we saw that there are beats where no, one, two, or three voices are
playing at the same time.
JJ II
J I
Page 21 of 35
Go Back
Full Screen
Close
Quit
Rhythmic Tiling Canons
In our last example of a canon, ((00110101), (01001010))
10011010
Home Page
01010011
11010100
Title Page
Contents
we saw that there are beats where no, one, two, or three voices are
playing at the same time.
JJ II
Close
Quit
Rhythmic Tiling Canons
In our last example of a canon, ((00110101), (01001010))
10011010
Home Page
01010011
11010100
Title Page
Contents
we saw that there are beats where no, one, two, or three voices are
playing at the same time.
JJ II
00001010
Enumeration of Tiling Canons
5 1 41
Go Back
6 3 110
Full Screen
7 1 341
8 6 1035
Close
9 4 3298
Quit
Enumeration of Tiling Canons
5 1 41 13 1 397529
Go Back
6 3 110 14 13 1.370798
Full Screen
7 1 341 15 25 4.780715
8 6 1035 16 49 16788150
Close
9 4 3298 17 1 59451809
Quit
Enumeration of Tiling Canons
5 1 41 13 1 397529
Go Back
6 3 110 14 13 1.370798
Full Screen
7 1 341 15 25 4.780715
8 6 1035 16 49 16788150
Close
9 4 3298 17 1 59451809
Quit
Contents
JJ II
J I
Page 23 of 35
Go Back
Full Screen
Close
Quit
Regular Complementary Canons of Maximal Category
J I
Page 23 of 35
Go Back
Full Screen
Close
Quit
Regular Complementary Canons of Maximal Category
Go Back
Full Screen
Close
Quit
Regular Complementary Canons of Maximal Category
Go Back Zn is not a Hajós group if and only if n can be expressed in the form
Full Screen
p1 p2n1n2n3 with p1, p2 primes, ni ≥ 2 for 1 ≤ i ≤ 3, and
gcd(n1 p1, n2 p2) = 1.
Close
How does such a canon sound?
Quit
Contents
JJ II
J I
Page 24 of 35
Go Back
Full Screen
Close
Quit
Vuza’s Algorithm
Contents Hence, both L or f can serve as the inner rhythm and the other one as
JJ II
the outer rhythm of an RCMC-canon.
J I
Page 24 of 35
Go Back
Full Screen
Close
Quit
Vuza’s Algorithm
Contents Hence, both L or f can serve as the inner rhythm and the other one as
JJ II
the outer rhythm of an RCMC-canon.
Full Screen
Close
Quit
Vuza’s Algorithm
Contents Hence, both L or f can serve as the inner rhythm and the other one as
JJ II
the outer rhythm of an RCMC-canon.
Full Screen
He also proves that when (L, f ) describes an RCMC-canon, then also
(kL, f ), (kL, k f ) have this property for all k ∈ Zn∗. (Here
Close
kL = {k` | ` ∈ L}.)
Quit
Contents
JJ II
J I
Page 25 of 35
Go Back
Full Screen
Close
Quit
Enumeration of Vuza Constructible Canons
J I
Page 25 of 35
Go Back
Full Screen
Close
Quit
n p1 p2 n1 n2 n3 #L # f #(L, f )
72 2 3 2 3 2 3 6 18
108 2 3 2 3 3 3 180 540
120 2 3 2 5 2 16 20 320
Home Page
120 2 5 2 3 2 8 6 48
Title Page
144 2 3 4 3 2 6 36 216
144 2 3 2 3 4 3 2808 8424
Contents 168 2 3 2 7 2 104 42 4368
168 2 7 2 3 2 16 6 96
JJ II
180 2 3 2 3 5 3 45360 136080
J I
180 2 3 2 5 3 16 1000 16000
180 2 5 2 3 3 8 252 2016
Page 26 of 35 180 2 3 5 3 2 9 60 540
180 3 5 3 2 2 6 12 72
Go Back
200 2 5 2 5 2 125 20 2500
Full Screen
240 2 3 4 5 2 32 120 3840
240 2 5 4 3 2 16 36 576
Close 240 2 3 2 5 4 16 26000 416000
240 2 5 2 3 4 8 6264 50112
Quit
Do there exist RCMC-canons which are not Vuza
constructible?
Home Page
Title Page
Contents
JJ II
J I
Page 27 of 35
Go Back
Full Screen
Close
Quit
Do there exist RCMC-canons which are not Vuza
constructible?
The answer is:
Home Page
Title Page
Contents
JJ II
J I
Page 27 of 35
Go Back
Full Screen
Close
Quit
Do there exist RCMC-canons which are not Vuza
constructible?
The answer is: YES!
Home Page
Title Page
Contents
JJ II
J I
Page 27 of 35
Go Back
Full Screen
Close
Quit
Do there exist RCMC-canons which are not Vuza
constructible?
The answer is: YES! For a long time it was an open question. I gave this
answer at the Third International Seminar on Mathematical Music Theory
Home Page
Contents
JJ II
J I
Page 27 of 35
Go Back
Full Screen
Close
Quit
Do there exist RCMC-canons which are not Vuza
constructible?
The answer is: YES! For a long time it was an open question. I gave this
answer at the Third International Seminar on Mathematical Music Theory
Home Page
Idea: If you have an RCMC-canon, then subdivide each beat into 2 beats.
Contents
Replace each quarter note or rest by 2 eighth notes or rests in the inner
JJ II
rhythm and enlarge the outer rhythm in a suitable way.
J I
Page 27 of 35
Go Back
Full Screen
Close
Quit
Do there exist RCMC-canons which are not Vuza
constructible?
The answer is: YES! For a long time it was an open question. I gave this
answer at the Third International Seminar on Mathematical Music Theory
Home Page
Idea: If you have an RCMC-canon, then subdivide each beat into 2 beats.
Contents
Replace each quarter note or rest by 2 eighth notes or rests in the inner
JJ II
rhythm and enlarge the outer rhythm in a suitable way.
J I
Construction: Let (L, f ) be an RCMC-canon. Construct L0 by replacing in
Page 27 of 35
L each occurrence of 1 by 11 and 0 by 00. And construct f 0 by replacing
Go Back each 1 in f by 01 and 0 in f by 00. This way we construct from the
Full Screen
RCMC-canon (L, f ) of length n an RCMC-canon (L0, f 0) of length 2n.
Close
Quit
Do there exist RCMC-canons which are not Vuza
constructible?
The answer is: YES! For a long time it was an open question. I gave this
answer at the Third International Seminar on Mathematical Music Theory
Home Page
Idea: If you have an RCMC-canon, then subdivide each beat into 2 beats.
Contents
Replace each quarter note or rest by 2 eighth notes or rests in the inner
JJ II
rhythm and enlarge the outer rhythm in a suitable way.
J I
Construction: Let (L, f ) be an RCMC-canon. Construct L0 by replacing in
Page 27 of 35
L each occurrence of 1 by 11 and 0 by 00. And construct f 0 by replacing
Go Back each 1 in f by 01 and 0 in f by 00. This way we construct from the
Full Screen
RCMC-canon (L, f ) of length n an RCMC-canon (L0, f 0) of length 2n.
Close Among the 216 RCMC-canons of length 2 · 72 = 144 with |L| = 12 we did
not find a canon which was constructed in this way from the 18 canons of
Quit
length 72.
From L equal to
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1]
Home Page
Title Page
Contents
JJ II
J I
Page 28 of 35
Go Back
Full Screen
Close
Quit
From L equal to
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1]
and f equal to
Home Page
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,1,1,0,0,0,0,1,0,0,0,0,0,1,0,
0,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,1,0,0,1,1]
Title Page
Contents
JJ II
J I
Page 28 of 35
Go Back
Full Screen
Close
Quit
From L equal to
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1]
and f equal to
Home Page
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,1,1,0,0,0,0,1,0,0,0,0,0,1,0,
0,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,1,0,0,1,1]
we get the canon with L0 equal to
Title Page
Contents
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
JJ II 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,
J I
0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1]
Page 28 of 35
Go Back
Full Screen
Close
Quit
From L equal to
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1]
and f equal to
Home Page
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,1,1,0,0,0,0,1,0,0,0,0,0,1,0,
0,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,1,0,0,1,1]
we get the canon with L0 equal to
Title Page
Contents
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
JJ II 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,
J I
0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1]
Page 28 of 35
and f 0 equal to
Go Back [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
Full Screen
0,0,0,1,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,1,0,0,
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,
Close
0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,1]
Quit
We can do much better
n m
Define a composition of two 0,1-vectors f ∈ {0, 1} , g ∈ {0, 1} . Then
f (g) is the vector in {0, 1}nm where each 1 in f is replaced by g and each
0 in f is replaced by 0m = 0 . . . 0. Each i ∈ nm can uniquely be written as
Home Page
i = qm + r with q ∈ n and r ∈ m. Then
Title Page
f (g)(i) = f (g)(qm + r) = f (q)g(r).
Contents
JJ II
J I
Page 29 of 35
Go Back
Full Screen
Close
Quit
We can do much better
n m
Define a composition of two 0,1-vectors f ∈ {0, 1} , g ∈ {0, 1} . Then
f (g) is the vector in {0, 1}nm where each 1 in f is replaced by g and each
0 in f is replaced by 0m = 0 . . . 0. Each i ∈ nm can uniquely be written as
Home Page
i = qm + r with q ∈ n and r ∈ m. Then
Title Page
f (g)(i) = f (g)(qm + r) = f (q)g(r).
Contents
Properties:
JJ II
1. wt( f (g)) = wt( f ) wt(g).
J I
Page 29 of 35
Go Back
Full Screen
Close
Quit
We can do much better
n m
Define a composition of two 0,1-vectors f ∈ {0, 1} , g ∈ {0, 1} . Then
f (g) is the vector in {0, 1}nm where each 1 in f is replaced by g and each
0 in f is replaced by 0m = 0 . . . 0. Each i ∈ nm can uniquely be written as
Home Page
i = qm + r with q ∈ n and r ∈ m. Then
Title Page
f (g)(i) = f (g)(qm + r) = f (q)g(r).
Contents
Properties:
JJ II
1. wt( f (g)) = wt( f ) wt(g).
n nm
J I 2. If g 6= 0, then {0, 1} 3 f 7→ f (g) ∈ {0, 1} is injective.
Page 29 of 35
Go Back
Full Screen
Close
Quit
We can do much better
n m
Define a composition of two 0,1-vectors f ∈ {0, 1} , g ∈ {0, 1} . Then
f (g) is the vector in {0, 1}nm where each 1 in f is replaced by g and each
0 in f is replaced by 0m = 0 . . . 0. Each i ∈ nm can uniquely be written as
Home Page
i = qm + r with q ∈ n and r ∈ m. Then
Title Page
f (g)(i) = f (g)(qm + r) = f (q)g(r).
Contents
Properties:
JJ II
1. wt( f (g)) = wt( f ) wt(g).
n nm
J I 2. If g 6= 0, then {0, 1} 3 f 7→ f (g) ∈ {0, 1} is injective.
Page 29 of 35
3. f (g) ◦ πqm
nm = ( f ◦ πq
n)(g).
Go Back
Full Screen
Close
Quit
We can do much better
n m
Define a composition of two 0,1-vectors f ∈ {0, 1} , g ∈ {0, 1} . Then
f (g) is the vector in {0, 1}nm where each 1 in f is replaced by g and each
0 in f is replaced by 0m = 0 . . . 0. Each i ∈ nm can uniquely be written as
Home Page
i = qm + r with q ∈ n and r ∈ m. Then
Title Page
f (g)(i) = f (g)(qm + r) = f (q)g(r).
Contents
Properties:
JJ II
1. wt( f (g)) = wt( f ) wt(g).
n nm
J I 2. If g 6= 0, then {0, 1} 3 f 7→ f (g) ∈ {0, 1} is injective.
Page 29 of 35
3. f (g) ◦ πqm
nm = ( f ◦ πq
n)(g).
4. f ≤ f 0 implies f (g) ≤ f 0(g).
Go Back
Full Screen
Close
Quit
We can do much better
n m
Define a composition of two 0,1-vectors f ∈ {0, 1} , g ∈ {0, 1} . Then
f (g) is the vector in {0, 1}nm where each 1 in f is replaced by g and each
0 in f is replaced by 0m = 0 . . . 0. Each i ∈ nm can uniquely be written as
Home Page
i = qm + r with q ∈ n and r ∈ m. Then
Title Page
f (g)(i) = f (g)(qm + r) = f (q)g(r).
Contents
Properties:
JJ II
1. wt( f (g)) = wt( f ) wt(g).
n nm
J I 2. If g 6= 0, then {0, 1} 3 f 7→ f (g) ∈ {0, 1} is injective.
Page 29 of 35
3. f (g) ◦ πqm
nm = ( f ◦ πq
n)(g).
4. f ≤ f 0 implies f (g) ≤ f 0(g).
5. If g 6= 0 and f < f 0, then f (g) < f 0(g).
Go Back
Full Screen
Close
Quit
We can do much better
n m
Define a composition of two 0,1-vectors f ∈ {0, 1} , g ∈ {0, 1} . Then
f (g) is the vector in {0, 1}nm where each 1 in f is replaced by g and each
0 in f is replaced by 0m = 0 . . . 0. Each i ∈ nm can uniquely be written as
Home Page
i = qm + r with q ∈ n and r ∈ m. Then
Title Page
f (g)(i) = f (g)(qm + r) = f (q)g(r).
Contents
Properties:
JJ II
1. wt( f (g)) = wt( f ) wt(g).
n nm
J I 2. If g 6= 0, then {0, 1} 3 f 7→ f (g) ∈ {0, 1} is injective.
Page 29 of 35
3. f (g) ◦ πqm
nm = ( f ◦ πq
n)(g).
4. f ≤ f 0 implies f (g) ≤ f 0(g).
5. If g 6= 0 and f < f 0, then f (g) < f 0(g).
Go Back
Full Screen
6. If f and g are the canonical representatives of Cn( f ) and Cm(g), then
Close
f (g) is the canonical representative of Cnm( f (g)).
Quit
We can do much better
n m
Define a composition of two 0,1-vectors f ∈ {0, 1} , g ∈ {0, 1} . Then
f (g) is the vector in {0, 1}nm where each 1 in f is replaced by g and each
0 in f is replaced by 0m = 0 . . . 0. Each i ∈ nm can uniquely be written as
Home Page
i = qm + r with q ∈ n and r ∈ m. Then
Title Page
f (g)(i) = f (g)(qm + r) = f (q)g(r).
Contents
Properties:
JJ II
1. wt( f (g)) = wt( f ) wt(g).
n nm
J I 2. If g 6= 0, then {0, 1} 3 f 7→ f (g) ∈ {0, 1} is injective.
Page 29 of 35
3. f (g) ◦ πqm
nm = ( f ◦ πq
n)(g).
4. f ≤ f 0 implies f (g) ≤ f 0(g).
5. If g 6= 0 and f < f 0, then f (g) < f 0(g).
Go Back
Full Screen
6. If f and g are the canonical representatives of Cn( f ) and Cm(g), then
Close
f (g) is the canonical representative of Cnm( f (g)).
7. If f is a Lyndon word and g is the canonical representative of Cm(g),
Quit
Theorem.
Home Page
1. If (L1, f1) is a canon of length n1 and (L2, f2) is a canon of length n2,
Title Page
JJ II
J I
Page 30 of 35
Go Back
Full Screen
Close
Quit
New Canons from Old Canons
Theorem.
Home Page
1. If (L1, f1) is a canon of length n1 and (L2, f2) is a canon of length n2,
Title Page
2. If (L1, f1) is a tiling canon and (L2, f2) is a tiling canon, then
JJ II
(L1(L2), f1( f2)) is a tiling canon.
J I
Page 30 of 35
Go Back
Full Screen
Close
Quit
New Canons from Old Canons
Theorem.
Home Page
1. If (L1, f1) is a canon of length n1 and (L2, f2) is a canon of length n2,
Title Page
2. If (L1, f1) is a tiling canon and (L2, f2) is a tiling canon, then
JJ II
(L1(L2), f1( f2)) is a tiling canon.
J I
Full Screen
Close
Quit
Are these all RCMC-Canons?
Home Page
Title Page
Contents
JJ II
J I
Page 31 of 35
Go Back
Full Screen
Close
Quit
Are these all RCMC-Canons?
Home Page
The answer is:
Title Page
Contents
JJ II
J I
Page 31 of 35
Go Back
Full Screen
Close
Quit
Are these all RCMC-Canons?
Home Page
The answer is: NO!
Title Page
Contents
JJ II
J I
Page 31 of 35
Go Back
Full Screen
Close
Quit
Are these all RCMC-Canons?
Home Page
The answer is: NO!
Title Page
J I
Page 31 of 35
Go Back
Full Screen
Close
Quit
Are these all RCMC-Canons?
Home Page
The answer is: NO!
Title Page
Go Back
Full Screen
Close
Quit
Complete lists of RCMC-canons
Contents
JJ II
J I
Page 32 of 35
Go Back
Full Screen
Close
Quit
Complete lists of RCMC-canons
J I
Page 32 of 35
Go Back
Full Screen
Close
Quit
Complete lists of RCMC-canons
determine the number of orbits of these Lyndon words under the action of
Full Screen
the affine group. For each of these representatives L determine all acyclic
Close outer rhythms f , such that L + f = Zn.
Quit
All RCMC-canons of length 72 and 108
There are 2.169882 Lyndon words of length 72 and weight 6 over {0, 1}.
Title Page
Page 33 of 35
Go Back
Full Screen
Close
Quit
All RCMC-canons of length 72 and 108
There are 2.169882 Lyndon words of length 72 and weight 6 over {0, 1}.
Title Page
Page 33 of 35
There are 17.717859 Lyndon words of length 108 and weight 6 over
{0, 1}. They are collected into 514754 orbits. There remains only one
Go Back
Discrete Structures
Title Page
Classification of Discrete Structures I
Contents
Labelled Graphs
Unlabelled Graphs on 4 Vertices
JJ II
Classification of Discrete Structures II
J I
Group Actions
Page 34 of 35 Orbits under Group Actions
Stabilizers and Fixed Points
Go Back