Quadratic space with isometry
We call quadratic space with isometry any pair $(V, f)$ consisting of a non-degenerate quadratic space $V$ together with an isometry $f\in O(V)$. We refer to the section about quadratic spaces of the documentation for new users.
Note that currently, we support only rational quadratic forms, i.e. quadratic spaces defined over the rational.
In Oscar, such a pair is encoded by the type called QuadSpaceWithIsom
:
QuadSpaceWithIsom
— TypeQuadSpaceWithIsom
A container type for pairs (V, f)
consisting on an rational quadratic space V
of type QuadSpace
and an isometry f
given as a QQMatrix
representing the action on the standard basis of V
.
We store the order of f
too, which can finite or of infinite order.
To construct an object of type QuadSpaceWithIsom
, see the set of functions called quadratic_space_with_isometry
Examples
julia> V = quadratic_space(QQ, 4);
julia> quadratic_space_with_isometry(V, neg=true)
Quadratic space of dimension 4
with isometry of finite order 2
given by
[-1 0 0 0]
[ 0 -1 0 0]
[ 0 0 -1 0]
[ 0 0 0 -1]
julia> L = root_lattice(:E, 6);
julia> V = ambient_space(L);
julia> f = matrix(QQ, 6, 6, [ 1 2 3 2 1 1;
-1 -2 -2 -2 -1 -1;
0 1 0 0 0 0;
1 0 0 0 0 0;
-1 -1 -1 0 0 -1;
0 0 1 1 0 1]);
julia> Vf = quadratic_space_with_isometry(V, f)
Quadratic space of dimension 6
with isometry of finite order 8
given by
[ 1 2 3 2 1 1]
[-1 -2 -2 -2 -1 -1]
[ 0 1 0 0 0 0]
[ 1 0 0 0 0 0]
[-1 -1 -1 0 0 -1]
[ 0 0 1 1 0 1]
and it is seen as a triple $(V, f, n)$ where $n$ is the order of $f$. We actually support isometries of finite and infinite order. In the case where $f$ is of infinite order, then n = PosInf
. If $V$ has rank 0, then any isometry $f$ of $V$ is trivial and we set by default n = -1
.
Given a quadratic space with isometry $(V, f)$, we provide the following accessors to the elements of the previously described triple:
isometry
— Methodisometry(Vf::QuadSpaceWithIsom) -> QQMatrix
Given a quadratic space with isometry $(V, f)$, return the underlying isometry f
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> isometry(Vf)
[-1 0]
[ 0 -1]
order_of_isometry
— Methodorder_of_isometry(Vf::QuadSpaceWithIsom) -> IntExt
Given a quadratic space with isometry $(V, f)$, return the order of the underlying isometry f
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> order_of_isometry(Vf) == 2
true
space
— Methodspace(Vf::QuadSpaceWithIsom) -> QuadSpace
Given a quadratic space with isometry $(V, f)$, return the underlying space V
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> space(Vf) === V
true
The main purpose of the definition of such objects is to define a contextual ambient space for quadratic lattices endowed with an isometry. Indeed, as we will see in the next section, lattices with isometry are attached to an ambient quadratic space with an isometry inducing the one on the lattice.
Constructors
For simplicity, we have gathered the main constructors for objects of type QuadSpaceWithIsom
under the same name quadratic_space_with_isometry
. The user has then the choice on the parameters depending on what they intend to do:
quadratic_space_with_isometry
— Methodquadratic_space_with_isometry(V:QuadSpace, f::QQMatrix; check::Bool = false)
-> QuadSpaceWithIsom
Given a quadratic space V
and a matrix f
, if f
defines an isometry of V
of order n
(possibly infinite), return the corresponding quadratic space with isometry pair $(V, f)$.
Examples
julia> V = quadratic_space(QQ, QQ[ 2 -1;
-1 2])
Quadratic space of dimension 2
over rational field
with gram matrix
[ 2 -1]
[-1 2]
julia> f = matrix(QQ, 2, 2, [1 1;
0 -1])
[1 1]
[0 -1]
julia> Vf = quadratic_space_with_isometry(V, f)
Quadratic space of dimension 2
with isometry of finite order 2
given by
[1 1]
[0 -1]
quadratic_space_with_isometry
— Methodquadratic_space_with_isometry(V::QuadSpace; neg::Bool = false) -> QuadSpaceWithIsom
Given a quadratic space V
, return the quadratic space with isometry pair $(V, f)$ where f
is represented by the identity matrix.
If neg
is set to true, then the isometry f
is negative the identity on V
.
Examples
julia> V = quadratic_space(QQ, QQ[ 2 -1;
-1 2])
Quadratic space of dimension 2
over rational field
with gram matrix
[ 2 -1]
[-1 2]
julia> Vf = quadratic_space_with_isometry(V)
Quadratic space of dimension 2
with isometry of finite order 1
given by
[1 0]
[0 1]
By default, the first constructor always checks whether the matrix defines an isometry of the quadratic space. We recommend not to disable this parameter to avoid any complications. Note however that in the rank 0 case, the checks are avoided since all isometries are necessarily trivial.
Attributes and first operations
Given a quadratic space with isometry $Vf := (V, f)$, one has access to most of the attributes of $V$ and $f$ by calling the similar functions on the pair $(V, f)$ itself. For instance, in order to know the rank of $V$, one can simply call rank(Vf)
. Here is a list of what are the current accessible attributes:
characteristic_polynomial
— Methodcharacteristic_polynomial(Vf::QuadSpaceWithIsom) -> QQPolyRingElem
Given a quadratic space with isometry $(V, f)$, return the characteristic polynomial of the underlying isometry f
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> characteristic_polynomial(Vf)
x^2 + 2*x + 1
det
— Methoddet(Vf::QuadSpaceWithIsom) -> QQFieldElem
Given a quadratic space with isometry $(V, f)$, return the determinant of the underlying space V
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> is_one(det(Vf))
true
diagonal
— Methoddiagonal(Vf::QuadSpaceWithIsom) -> Vector{QQFieldElem}
Given a quadratic space with isometry $(V, f)$, return the diagonal of the underlying space V
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> diagonal(Vf)
2-element Vector{QQFieldElem}:
1
1
dim
— Methoddim(Vf::QuadSpaceWithIsom) -> Integer
Given a quadratic space with isometry $(V, f)$, return the dimension of the underlying space of V
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> dim(Vf) == 2
true
discriminant
— Methoddiscriminant(Vf::QuadSpaceWithIsom) -> QQFieldElem
Given a quadratic space with isometry $(V, f)$, return the discriminant of the underlying space V
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> discriminant(Vf)
-1
gram_matrix
— Methodgram_matrix(Vf::QuadSpaceWithIsom) -> QQMatrix
Given a quadratic space with isometry $(V, f)$, return the Gram matrix of the underlying space V
with respect to its standard basis.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> is_one(gram_matrix(Vf))
true
is_definite
— Methodis_definite(Vf::QuadSpaceWithIsom) -> Bool
Given a quadratic space with isometry $(V, f)$, return whether the underlying space V
is definite.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> is_definite(Vf)
true
is_positive_definite
— Methodis_positive_definite(Vf::QuadSpaceWithIsom) -> Bool
Given a quadratic space with isometry $(V, f)$, return whether the underlying space V
is positive definite.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> is_positive_definite(Vf)
true
is_negative_definite
— Methodis_negative_definite(Vf::QuadSpaceWithIsom) -> Bool
Given a quadratic space with isometry $(V, f)$, return whether the underlying space V
is negative definite.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> is_negative_definite(Vf)
false
minimal_polynomial
— Methodminimal_polynomial(Vf::QuadSpaceWithIsom) -> QQPolyRingElem
Given a quadratic space with isometry $(V, f)$, return the minimal polynomial of the underlying isometry f
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> minimal_polynomial(Vf)
x + 1
rank
— Methodrank(Vf::QuadSpaceWithIsom) -> Integer
Given a quadratic space with isometry $(V, f)$, return the rank of the underlying space V
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> rank(Vf) == 2
true
signature_tuple
— Methodsignature_tuple(Vf::QuadSpaceWithIsom) -> Tuple{Int, Int, Int}
Given a quadratic space with isometry $(V, f)$, return the signature tuple of the underlying space V
.
Examples
julia> V = quadratic_space(QQ, 2);
julia> Vf = quadratic_space_with_isometry(V; neg = true);
julia> signature_tuple(Vf)
(2, 0, 0)
Similarly, some basic operations on quadratic spaces are available for quadratic spaces with isometry.
^
— Method^(Vf::QuadSpaceWithIsom, n::Int) -> QuadSpaceWithIsom
Given a quadratic space with isometry $(V, f)$ and an integer $n$, return the pair $(V, f^n)$.
Examples
julia> V = quadratic_space(QQ, QQ[ 2 -1;
-1 2])
Quadratic space of dimension 2
over rational field
with gram matrix
[ 2 -1]
[-1 2]
julia> f = matrix(QQ, 2, 2, [1 1;
0 -1])
[1 1]
[0 -1]
julia> Vf = quadratic_space_with_isometry(V, f)
Quadratic space of dimension 2
with isometry of finite order 2
given by
[1 1]
[0 -1]
julia> Vf^2
Quadratic space of dimension 2
with isometry of finite order 1
given by
[1 0]
[0 1]
biproduct
— Methodbiproduct(x::Vector{QuadSpaceWithIsom}) -> QuadSpaceWithIsom, Vector{AbstractSpaceMor}, Vector{AbstractSpaceMor}
biproduct(x::Vararg{QuadSpaceWithIsom}) -> QuadSpaceWithIsom, Vector{AbstractSpaceMor}, Vector{AbstractSpaceMor}
Given a collection of quadratic spaces with isometries $(V_1, f_1), \ldots, (V_n, f_n)$, return the quadratic space with isometry $(V, f)$ together with the injections $V_i \to V$ and the projections $V \to V_i$, where V
is the biproduct $V := V_1 \oplus \ldots \oplus V_n$ and f
is the isometry of V
induced by the diagonal actions of the $f_i$'s.
For objects of type QuadSpaceWithIsom
, finite direct sums and finite direct products agree and they are therefore called biproducts. If one wants to obtain $(V, f)$ as a direct sum with the injections $V_i \to V$, one should call direct_sum(x)
. If one wants to obtain $(V, f)$ as a direct product with the projections $V \to V_i$, one should call direct_product(x)
.
Examples
julia> V1 = quadratic_space(QQ, QQ[2 5;
5 6])
Quadratic space of dimension 2
over rational field
with gram matrix
[2 5]
[5 6]
julia> Vf1 = quadratic_space_with_isometry(V1, neg=true)
Quadratic space of dimension 2
with isometry of finite order 2
given by
[-1 0]
[ 0 -1]
julia> V2 = quadratic_space(QQ, QQ[ 2 -1;
-1 2])
Quadratic space of dimension 2
over rational field
with gram matrix
[ 2 -1]
[-1 2]
julia> f = matrix(QQ, 2, 2, [1 1;
0 -1])
[1 1]
[0 -1]
julia> Vf2 = quadratic_space_with_isometry(V2, f)
Quadratic space of dimension 2
with isometry of finite order 2
given by
[1 1]
[0 -1]
julia> Vf3, inj, proj = biproduct(Vf1, Vf2)
(Quadratic space with isometry of finite order 2, AbstractSpaceMor[Map with following data
Domain:
=======
Quadratic space of dimension 2
Codomain:
=========
Quadratic space of dimension 4, Map with following data
Domain:
=======
Quadratic space of dimension 2
Codomain:
=========
Quadratic space of dimension 4], AbstractSpaceMor[Map with following data
Domain:
=======
Quadratic space of dimension 4
Codomain:
=========
Quadratic space of dimension 2, Map with following data
Domain:
=======
Quadratic space of dimension 4
Codomain:
=========
Quadratic space of dimension 2])
julia> Vf3
Quadratic space of dimension 4
with isometry of finite order 2
given by
[-1 0 0 0]
[ 0 -1 0 0]
[ 0 0 1 1]
[ 0 0 0 -1]
julia> space(Vf3)
Quadratic space of dimension 4
over rational field
with gram matrix
[2 5 0 0]
[5 6 0 0]
[0 0 2 -1]
[0 0 -1 2]
julia> matrix(compose(inj[1], proj[1]))
[1 0]
[0 1]
julia> matrix(compose(inj[1], proj[2]))
[0 0]
[0 0]
direct_product
— Methoddirect_product(F::FreeMod{T}...; task::Symbol = :prod) where T
Given free modules $F_1\dots F_n$, say, return the direct product $\prod_{i=1}^n F_i$.
Additionally, return
- a vector containing the canonical projections $\prod_{i=1}^n F_i\to F_i$ if
task = :prod
(default), - a vector containing the canonical injections $F_i\to\prod_{i=1}^n F_i$ if
task = :sum
, - two vectors containing the canonical projections and injections, respectively, if
task = :both
, - none of the above maps if
task = :none
.
direct_product(M::ModuleFP{T}...; task::Symbol = :prod) where T
Given modules $M_1\dots M_n$, say, return the direct product $\prod_{i=1}^n M_i$.
Additionally, return
- a vector containing the canonical projections $\prod_{i=1}^n M_i\to M_i$ if
task = :prod
(default), - a vector containing the canonical injections $M_i\to\prod_{i=1}^n M_i$ if
task = :sum
, - two vectors containing the canonical projections and injections, respectively, if
task = :both
, - none of the above maps if
task = :none
.
direct_product(x::Vector{QuadSpaceWithIsom}) -> QuadSpaceWithIsom, Vector{AbstractSpaceMor}
direct_product(x::Vararg{QuadSpaceWithIsom}) -> QuadSpaceWithIsom, Vector{AbstractSpaceMor}
Given a collection of quadratic spaces with isometries $(V_1, f_1), \ldots, (V_n, f_n)$, return the quadratic space with isometry $(V, f)$ together with the projections $V \to V_i$, where V
is the direct product $V := V_1 \times \ldots \times V_n$ and f
is the isometry of V
induced by the diagonal actions of the $f_i$'s.
For objects of type QuadSpaceWithIsom
, finite direct sums and finite direct products agree and they are therefore called biproducts. If one wants to obtain $(V, f)$ as a direct sum with the injections $V_i \to V$, one should call direct_sum(x)
. If one wants to obtain $(V, f)$ as a biproduct with the injections $V_i \to V$ and the projections $V \to V_i$, one should call biproduct(x)
.
Examples
julia> V1 = quadratic_space(QQ, QQ[2 5;
5 6])
Quadratic space of dimension 2
over rational field
with gram matrix
[2 5]
[5 6]
julia> Vf1 = quadratic_space_with_isometry(V1, neg=true)
Quadratic space of dimension 2
with isometry of finite order 2
given by
[-1 0]
[ 0 -1]
julia> V2 = quadratic_space(QQ, QQ[ 2 -1;
-1 2])
Quadratic space of dimension 2
over rational field
with gram matrix
[ 2 -1]
[-1 2]
julia> f = matrix(QQ, 2, 2, [1 1;
0 -1])
[1 1]
[0 -1]
julia> Vf2 = quadratic_space_with_isometry(V2, f)
Quadratic space of dimension 2
with isometry of finite order 2
given by
[1 1]
[0 -1]
julia> Vf3, proj = direct_product(Vf1, Vf2)
(Quadratic space with isometry of finite order 2, AbstractSpaceMor[Map with following data
Domain:
=======
Quadratic space of dimension 4
Codomain:
=========
Quadratic space of dimension 2, Map with following data
Domain:
=======
Quadratic space of dimension 4
Codomain:
=========
Quadratic space of dimension 2])
julia> Vf3
Quadratic space of dimension 4
with isometry of finite order 2
given by
[-1 0 0 0]
[ 0 -1 0 0]
[ 0 0 1 1]
[ 0 0 0 -1]
julia> space(Vf3)
Quadratic space of dimension 4
over rational field
with gram matrix
[2 5 0 0]
[5 6 0 0]
[0 0 2 -1]
[0 0 -1 2]
direct_product(algebras::AlgAss...; task::Symbol = :sum)
-> AlgAss, Vector{AbsAlgAssMor}, Vector{AbsAlgAssMor}
direct_product(algebras::Vector{AlgAss}; task::Symbol = :sum)
-> AlgAss, Vector{AbsAlgAssMor}, Vector{AbsAlgAssMor}
Returns the algebra $A = A_1 \times \cdots \times A_k$. task
can be ":sum", ":prod", ":both" or ":none" and determines which canonical maps are computed as well: ":sum" for the injections, ":prod" for the projections.
direct_sum
— Methoddirect_sum(M::ModuleFP{T}...; task::Symbol = :sum) where T
Given modules $M_1\dots M_n$, say, return the direct sum $\bigoplus_{i=1}^n M_i$.
Additionally, return
- a vector containing the canonical injections $M_i\to\bigoplus_{i=1}^n M_i$ if
task = :sum
(default), - a vector containing the canonical projections $\bigoplus_{i=1}^n M_i\to M_i$ if
task = :prod
, - two vectors containing the canonical injections and projections, respectively, if
task = :both
, - none of the above maps if
task = :none
.
direct_sum(x::Vector{QuadSpaceWithIsom}) -> QuadSpaceWithIsom, Vector{AbstractSpaceMor}
direct_sum(x::Vararg{QuadSpaceWithIsom}) -> QuadSpaceWithIsom, Vector{AbstractSpaceMor}
Given a collection of quadratic spaces with isometries $(V_1, f_1), \ldots, (V_n, f_n)$, return the quadratic space with isometry $(V, f)$ together with the injections $V_i \to V$, where V
is the direct sum $V := V_1 \oplus \ldots \oplus V_n$ and f
is the isometry of V
induced by the diagonal actions of the $f_i$'s.
For objects of type QuadSpaceWithIsom
, finite direct sums and finite direct products agree and they are therefore called biproducts. If one wants to obtain $(V, f)$ as a direct product with the projections $V \to V_i$, one should call direct_product(x)
. If one wants to obtain $(V, f)$ as a biproduct with the injections $V_i \to V$ and the projections $V \to V_i$, one should call biproduct(x)
.
Examples
julia> V1 = quadratic_space(QQ, QQ[2 5;
5 6])
Quadratic space of dimension 2
over rational field
with gram matrix
[2 5]
[5 6]
julia> Vf1 = quadratic_space_with_isometry(V1, neg=true)
Quadratic space of dimension 2
with isometry of finite order 2
given by
[-1 0]
[ 0 -1]
julia> V2 = quadratic_space(QQ, QQ[ 2 -1;
-1 2])
Quadratic space of dimension 2
over rational field
with gram matrix
[ 2 -1]
[-1 2]
julia> f = matrix(QQ, 2, 2, [1 1;
0 -1])
[1 1]
[0 -1]
julia> Vf2 = quadratic_space_with_isometry(V2, f)
Quadratic space of dimension 2
with isometry of finite order 2
given by
[1 1]
[0 -1]
julia> Vf3, inj = direct_sum(Vf1, Vf2)
(Quadratic space with isometry of finite order 2, AbstractSpaceMor[Map with following data
Domain:
=======
Quadratic space of dimension 2
Codomain:
=========
Quadratic space of dimension 4, Map with following data
Domain:
=======
Quadratic space of dimension 2
Codomain:
=========
Quadratic space of dimension 4])
julia> Vf3
Quadratic space of dimension 4
with isometry of finite order 2
given by
[-1 0 0 0]
[ 0 -1 0 0]
[ 0 0 1 1]
[ 0 0 0 -1]
julia> space(Vf3)
Quadratic space of dimension 4
over rational field
with gram matrix
[2 5 0 0]
[5 6 0 0]
[0 0 2 -1]
[0 0 -1 2]
direct_sum(g1::QuadSpaceCls, g2::QuadSpaceCls) -> QuadSpaceCls
Return the isometry class of the direct sum of two representatives.
rescale
— Methodrescale(Vf::QuadSpaceWithIsom, a::RationalUnion)
Given a quadratic space with isometry $(V, f)$, return the pair $(V^a, f$) where $V^a$ is the same space as V
with the associated quadratic form rescaled by a
.
Examples
julia> V = quadratic_space(QQ, QQ[ 2 -1;
-1 2])
Quadratic space of dimension 2
over rational field
with gram matrix
[ 2 -1]
[-1 2]
julia> Vf = quadratic_space_with_isometry(V)
Quadratic space of dimension 2
with isometry of finite order 1
given by
[1 0]
[0 1]
julia> Vf2 = rescale(Vf, 1//2)
Quadratic space of dimension 2
with isometry of finite order 1
given by
[1 0]
[0 1]
julia> space(Vf2)
Quadratic space of dimension 2
over rational field
with gram matrix
[ 1 -1//2]
[-1//2 1]
Equality
We choose as a convention that two pairs $(V, f)$ and $(V', f')$ of quadratic spaces with isometries are equal if and only if $V$ and $V'$ are the same space, and $f$ and $f'$ are represented by the same matrix with respect to the standard basis of $V = V'$.