# Matrix groups

`matrix_group`

— Method```
matrix_group(m::Int, R::Ring, V::T...) where T<:Union{MatElem,MatrixGroupElem}
matrix_group(m::Int, R::Ring, V::AbstractVector{T}) where T<:Union{MatElem,MatrixGroupElem}
matrix_group(V::T...) where T<:Union{MatElem,MatrixGroupElem}
matrix_group(V::AbstractVector{T}) where T<:Union{MatElem,MatrixGroupElem}
```

Return the matrix group generated by matrices in `V`

. If the degree `m`

and coefficient ring `R`

are not given, then `V`

must be non-empty

`MatrixGroup`

— Type`MatrixGroup{RE<:RingElem, T<:MatElem{RE}} <: GAPGroup`

Type of groups `G`

of `n x n`

matrices over the ring `R`

, where `n = degree(G)`

and `R = base_ring(G)`

.

At the moment, only rings of type `fqPolyRepField`

are supported.

`MatrixGroupElem`

— Type`MatrixGroupElem{RE<:RingElem, T<:MatElem{RE}} <: AbstractMatrixGroupElem`

Elements of a group of type `MatrixGroup{RE<:RingElem, T<:MatElem{RE}}`

`base_ring`

— Method`base_ring(G::MatrixGroup)`

Return the base ring of the matrix group `G`

.

`degree`

— Method`degree(G::MatrixGroup)`

Return the degree of the matrix group `G`

, i.e. the number of rows of its matrices.

`centralizer`

— Method`centralizer(G::MatrixGroup{T}, x::MatrixGroupElem{T})`

Return (`C`

,`f`

), where `C`

is the centralizer of `x`

in `C`

and `f`

is the embedding of `C`

into `G`

. If `G`

= `GL(n,F)`

or `SL(n,F)`

, then `f`

= `nothing`

. In this case, to get the embedding homomorphism of `C`

into `G`

, use

`is_subgroup(C, G)[2]`

## Elements of matrix groups

`matrix`

— Method`matrix(x::MatrixGroupElem)`

Return the underlying matrix of `x`

.

`base_ring`

— Method`base_ring(x::MatrixGroupElem)`

Return the base ring of the underlying matrix of `x`

.

`nrows`

— Method`nrows(x::MatrixGroupElem)`

Return the number of rows of the underlying matrix of `x`

.

`det`

— Method`det(x::MatrixGroupElem)`

Return the determinant of the underlying matrix of `x`

.

`tr`

— Method`tr(x::MatrixGroupElem)`

Return the trace of the underlying matrix of `x`

.

`multiplicative_jordan_decomposition`

— Method`multiplicative_jordan_decomposition(M::MatrixGroupElem)`

Return `S`

and `U`

in the group `G = parent(M)`

such that `S`

is semisimple, `U`

is unipotent and `M = SU = US`

.

this is *NOT*, in general, the same output returned when `M`

has type `MatElem`

.

`is_semisimple`

— Method`is_semisimple(x::MatrixGroupElem{T}) where T <: FinFieldElem`

Return whether `x`

is semisimple, i.e. has order coprime with the characteristic of its base ring.

`is_unipotent`

— Method`is_unipotent(x::MatrixGroupElem{T}) where T <: FinFieldElem`

Return whether `x`

is unipotent, i.e. its order is a power of the characteristic of its base ring.

## Sesquilinear forms

`SesquilinearForm`

— Type`SesquilinearForm{T<:RingElem}`

Type of groups `G`

of `n x n`

matrices over the ring `R`

, where `n = degree(G)`

and `R = base_ring(G)`

. At the moment, only rings of type `fqPolyRepField`

are supported.

`is_alternating_form`

— Method`is_alternating_form(f::SesquilinearForm)`

Return whether the form `f`

is an alternating form.

`is_hermitian_form`

— Method`is_hermitian_form(f::SesquilinearForm)`

Return whether the form `f`

is a hermitian form.

`is_quadratic_form`

— Method`is_quadratic_form(f::SesquilinearForm)`

Return whether the form `f`

is a quadratic form.

`is_symmetric_form`

— Method`is_symmetric_form(f::SesquilinearForm)`

Return whether the form `f`

is a symmetric form.

`alternating_form`

— Method`alternating_form(B::MatElem{T})`

Return the alternating form with Gram matrix `B`

.

`symmetric_form`

— Method`symmetric_form(B::MatElem{T})`

Return the symmetric form with Gram matrix `B`

.

`hermitian_form`

— Method`hermitian_form(B::MatElem{T})`

Return the hermitian form with Gram matrix `B`

.

`quadratic_form`

— Method`quadratic_form(B::MatElem{T})`

Return the quadratic form with Gram matrix `B`

.

`quadratic_form`

— Method`quadratic_form(f::MPolyRingElem{T}; check=true)`

Return the quadratic form described by the polynomial `f`

. Here, `f`

must be a homogeneous polynomial of degree 2. If `check`

is set as `false`

, it does not check whether the polynomial is homogeneous of degree 2. To define quadratic forms of dimension 1, `f`

can also have type `PolyRingElem{T}`

.

`corresponding_bilinear_form`

— Method`corresponding_bilinear_form(Q::SesquilinearForm)`

Given a quadratic form `Q`

, return the bilinear form `B`

defined by `B(u,v) = Q(u+v)-Q(u)-Q(v)`

.

`corresponding_quadratic_form`

— Method`corresponding_quadratic_form(Q::SesquilinearForm)`

Given a symmetric form `f`

, returns the quadratic form `Q`

defined by `Q(v) = f(v,v)/2`

. It is defined only in odd characteristic.

`gram_matrix`

— Method`gram_matrix(B::SesquilinearForm)`

Return the Gram matrix of a sesquilinear or quadratic form `B`

.

`defining_polynomial`

— Method`defining_polynomial(f::SesquilinearForm)`

Return the polynomial that defines the quadratic form `f`

.

`radical`

— Method`radical(f::SesquilinearForm{T})`

Return the radical of the sesquilinear form `f`

, i.e. the subspace of all `v`

such that `f(u,v)=0`

for all `u`

. The radical of a quadratic form `Q`

is the set of vectors `v`

such that `Q(v)=0`

and `v`

lies in the radical of the corresponding bilinear form.

` radical(A::AbsAlgAss) -> AbsAlgAssIdl`

Returns the Jacobson-Radical of $A$.

`witt_index`

— Method`witt_index(f::SesquilinearForm{T})`

Return the Witt index of the form induced by `f`

on `V/Rad(f)`

. The Witt Index is the dimension of a maximal totally isotropic (singular for quadratic forms) subspace.

`is_degenerate`

— Method`is_degenerate(f::SesquilinearForm{T})`

Return whether `f`

is degenerate, i.e. `f`

has nonzero radical. A quadratic form is degenerate if the corresponding bilinear form is.

`is_singular`

— Method`is_singular(Q::SesquilinearForm{T})`

For a quadratic form `Q`

, return whether `Q`

is singular, i.e. `Q`

has nonzero radical.

`is_congruent`

— Method`is_congruent(f::SesquilinearForm{T}, g::SesquilinearForm{T}) where T <: RingElem`

If `f`

and `g`

are sesquilinear forms, return (`true`

, `C`

) if there exists a matrix `C`

such that `f^C = g`

, or equivalently, `CBC* = A`

, where `A`

and `B`

are the Gram matrices of `f`

and `g`

respectively, and `C*`

is the transpose-conjugate matrix of `C`

. If such `C`

does not exist, then return (`false`

, `nothing`

).

If `f`

and `g`

are quadratic forms, return (`true`

, `C`

) if there exists a matrix `C`

such that `f^A = ag`

for some scalar `a`

. If such `C`

does not exist, then return (`false`

, `nothing`

).

## Invariant forms

`invariant_bilinear_forms`

— Method`invariant_bilinear_forms(G::MatrixGroup)`

Return a generating set for the vector spaces of bilinear forms preserved by the group `G`

.

At the moment, elements of the generating set are returned of type `mat_elem_type(G)`

.

`invariant_sesquilinear_forms`

— Method`invariant_sesquilinear_forms(G::MatrixGroup)`

Return a generating set for the vector spaces of sesquilinear non-bilinear forms preserved by the group `G`

. An exception is thrown if `base_ring(G)`

is not a finite field with even degree over its prime subfield.

At the moment, elements of the generating set are returned of type `mat_elem_type(G)`

.

`invariant_quadratic_forms`

— Method`invariant_quadratic_forms(G::MatrixGroup)`

Return a generating set for the vector spaces of quadratic forms preserved by the group `G`

.

At the moment, elements of the generating set are returned of type `mat_elem_type(G)`

.

`invariant_symmetric_forms`

— Method`invariant_symmetric_forms(G::MatrixGroup)`

Return a generating set for the vector spaces of symmetric forms preserved by the group `G`

.

At the moment, elements of the generating set are returned of type `mat_elem_type(G)`

.

Work properly only in odd characteristic. In even characteristic, only alternating forms are found.

`invariant_alternating_forms`

— Method`invariant_alternating_forms(G::MatrixGroup)`

Return a generating set for the vector spaces of alternating forms preserved by the group `G`

.

At the moment, elements of the generating set are returned of type `mat_elem_type(G)`

.

`invariant_hermitian_forms`

— Method`invariant_hermitian_forms(G::MatrixGroup)`

Return a generating set for the vector spaces of hermitian forms preserved by the group `G`

. An exception is thrown if `base_ring(G)`

is not a finite field with even degree over its prime subfield.

At the moment, elements of the generating set are returned of type `mat_elem_type(G)`

.

`invariant_bilinear_form`

— Method`invariant_bilinear_form(G::MatrixGroup)`

Return an invariant bilinear form for the group `G`

. An exception is thrown if the module induced by the action of `G`

is not absolutely irreducible.

At the moment, the output is returned of type `mat_elem_type(G)`

.

`invariant_sesquilinear_form`

— Method`invariant_sesquilinear_form(G::MatrixGroup)`

Return an invariant sesquilinear (non bilinear) form for the group `G`

. An exception is thrown if the module induced by the action of `G`

is not absolutely irreducible or if the group is defined over a finite field of odd degree over the prime field.

At the moment, the output is returned of type `mat_elem_type(G)`

.

`invariant_quadratic_form`

— Method`invariant_quadratic_form(G::MatrixGroup)`

Return an invariant quadratic form for the group `G`

. An exception is thrown if the module induced by the action of `G`

is not absolutely irreducible.

At the moment, the output is returned of type `mat_elem_type(G)`

.

`preserved_quadratic_forms`

— Method`preserved_quadratic_forms(G::MatrixGroup)`

Uses random methods to find all of the quadratic forms preserved by `G`

up to a scalar (i.e. such that `G`

is a group of similarities for the forms). Since the procedure relies on a pseudo-random generator, the user may need to execute the operation more than once to find all invariant quadratic forms.

`preserved_sesquilinear_forms`

— Method`preserved_sesquilinear_forms(G::MatrixGroup)`

Uses random methods to find all of the sesquilinear forms preserved by `G`

up to a scalar (i.e. such that `G`

is a group of similarities for the forms). Since the procedure relies on a pseudo-random generator, the user may need to execute the operation more than once to find all invariant sesquilinear forms.

`isometry_group`

— Method`isometry_group(f::SesquilinearForm{T})`

Return the group of isometries for the sesquilinear form `f`

.

`orthogonal_sign`

— Method`orthogonal_sign(G::MatrixGroup)`

For absolutely irreducible `G`

of degree `n`

and such that `base_ring(G)`

is a finite field, return

`nothing`

if`G`

does not preserve a nonzero quadratic form,`0`

if`n`

is odd and`G`

preserves a nonzero quadratic form,`1`

if`n`

is even and`G`

preserves a nonzero quadratic form of`+`

type,`-1`

if`n`

is even and`G`

preserves a nonzero quadratic form of`-`

type.

## Utilities for matrices (replace by available functions, or document elsewhere?)

`pol_elementary_divisors`

— Method```
pol_elementary_divisors(x::MatElem)
pol_elementary_divisors(x::MatrixGroupElem)
```

Return a list of pairs `(f_i,m_i)`

, for irreducible polynomials `f_i`

and positive integers `m_i`

, where the `f_i^m_i`

are the elementary divisors of `x`

.

`generalized_jordan_block`

— Method`generalized_jordan_block(f::T, n::Int) where T<:PolyRingElem`

Return the Jordan block of dimension `n`

corresponding to the polynomial `f`

.

`generalized_jordan_form`

— Method`generalized_jordan_form(A::MatElem{T}; with_pol::Bool=false) where T`

Return (`J`

,`Z`

), where `Z^-1*J*Z = A`

and `J`

is a diagonal join of Jordan blocks (corresponding to irreducible polynomials).

`matrix`

— Method`matrix(A::Vector{AbstractAlgebra.Generic.FreeModuleElem})`

Return the matrix whose rows are the vectors in `A`

. All vectors in `A`

must have the same length and the same base ring.

`upper_triangular_matrix`

— Method`upper_triangular_matrix(L)`

Return the upper triangular matrix whose entries on and above the diagonal are the elements of `L`

.

An exception is thrown whenever the length of `L`

is not equal to $n(n+1)/2$, for some integer $n$.

`lower_triangular_matrix`

— Method`lower_triangular_matrix(L)`

Return the upper triangular matrix whose entries on and below the diagonal are the elements of `L`

.

An exception is thrown whenever the length of `L`

is not equal to $n(n+1)/2$, for some integer $n$.

`conjugate_transpose`

— Method`conjugate_transpose(x::MatElem{T}) where T <: FinFieldElem`

If the base ring of `x`

is `GF(q^2)`

, return the matrix `transpose( map ( y -> y^q, x) )`

. An exception is thrown if the base ring does not have even degree.

`complement`

— Method`complement(V::AbstractAlgebra.Generic.FreeModule{T}, W::AbstractAlgebra.Generic.Submodule{T}) where T <: FieldElem`

Return a complement for `W`

in `V`

, i.e. a subspace `U`

of `V`

such that `V`

is direct sum of `U`

and `W`

.

`permutation_matrix`

— Method```
permutation_matrix(F::Ring, Q::AbstractVector{T}) where T <: Int
permutation_matrix(F::Ring, p::PermGroupElem)
```

Return the permutation matrix over the ring `R`

corresponding to the sequence `Q`

or to the permutation `p`

. If `Q`

is a sequence, then `Q`

must contain exactly once every integer from 1 to some `n`

.

**Examples**

```
julia> s = perm([3,1,2])
(1,3,2)
julia> permutation_matrix(QQ,s)
[0 0 1]
[1 0 0]
[0 1 0]
```

`is_skewsymmetric_matrix`

— Method`is_skewsymmetric_matrix(B::MatElem{T}) where T <: Ring`

Return whether the matrix `B`

is skew-symmetric, i.e. `B = -transpose(B)`

and `B`

has zeros on the diagonal. Return `false`

if `B`

is not a square matrix.

`is_hermitian_matrix`

— Method`is_hermitian_matrix(B::MatElem{T}) where T <: FinFieldElem`

Return whether the matrix `B`

is hermitian, i.e. `B = conjugate_transpose(B)`

. Return `false`

if `B`

is not a square matrix, or the field has not even degree.

## Classical groups

`general_linear_group`

— Method```
general_linear_group(n::Int, q::Int)
general_linear_group(n::Int, R::Ring)
GL = general_linear_group
```

Return the general linear group of dimension `n`

over the ring `R`

respectively the field `GF(q)`

.

Currently, this function only supports rings of type `fqPolyRepField`

.

**Examples**

```
julia> F = GF(7,1)
Finite field of degree 1 over GF(7)
julia> H = general_linear_group(2,F)
GL(2,7)
julia> gens(H)
2-element Vector{MatrixGroupElem{fqPolyRepFieldElem, fqPolyRepMatrix}}:
[3 0; 0 1]
[6 1; 6 0]
```

`special_linear_group`

— Method```
special_linear_group(n::Int, q::Int)
special_linear_group(n::Int, R::Ring)
SL = special_linear_group
```

Return the special linear group of dimension `n`

over the ring `R`

respectively the field `GF(q)`

.

Currently, this function only supports rings of type `fqPolyRepField`

.

**Examples**

```
julia> F = GF(7,1)
Finite field of degree 1 over GF(7)
julia> H = special_linear_group(2,F)
SL(2,7)
julia> gens(H)
2-element Vector{MatrixGroupElem{fqPolyRepFieldElem, fqPolyRepMatrix}}:
[3 0; 0 5]
[6 1; 6 0]
```

`symplectic_group`

— Method```
symplectic_group(n::Int, q::Int)
symplectic_group(n::Int, R::Ring)
Sp = symplectic_group
```

Return the symplectic group of dimension `n`

over the ring `R`

respectively the field `GF(q)`

. The dimension `n`

must be even.

Currently, this function only supports rings of type `fqPolyRepField`

.

**Examples**

```
julia> F = GF(7,1)
Finite field of degree 1 over GF(7)
julia> H = symplectic_group(2,F)
Sp(2,7)
julia> gens(H)
2-element Vector{MatrixGroupElem{fqPolyRepFieldElem, fqPolyRepMatrix}}:
[3 0; 0 5]
[6 1; 6 0]
```

`orthogonal_group`

— Method```
orthogonal_group(e::Int, n::Int, R::Ring)
orthogonal_group(e::Int, n::Int, q::Int)
GO = orthogonal_group
```

Return the orthogonal group of dimension `n`

over the ring `R`

respectively the field `GF(q)`

, and of type `e`

, where `e`

in {`+1`

,`-1`

} for `n`

even and `e`

=`0`

for `n`

odd. If `n`

is odd, `e`

can be omitted.

Currently, this function only supports rings of type `fqPolyRepField`

.

**Examples**

```
julia> F = GF(7,1)
Finite field of degree 1 over GF(7)
julia> H = symplectic_group(2,F)
Sp(2,7)
julia> gens(H)
2-element Vector{MatrixGroupElem{fqPolyRepFieldElem, fqPolyRepMatrix}}:
[3 0; 0 5]
[6 1; 6 0]
```

`special_orthogonal_group`

— Method```
special_orthogonal_group(e::Int, n::Int, R::Ring)
special_orthogonal_group(e::Int, n::Int, q::Int)
SO = special_orthogonal_group
```

Return the special orthogonal group of dimension `n`

over the ring `R`

respectively the field `GF(q)`

, and of type `e`

, where `e`

in {`+1`

,`-1`

} for `n`

even and `e`

=`0`

for `n`

odd. If `n`

is odd, `e`

can be omitted.

Currently, this function only supports rings of type `fqPolyRepField`

.

**Examples**

```
julia> F = GF(7,1)
Finite field of degree 1 over GF(7)
julia> H = special_orthogonal_group(1,2,F)
SO+(2,7)
julia> gens(H)
3-element Vector{MatrixGroupElem{fqPolyRepFieldElem, fqPolyRepMatrix}}:
[3 0; 0 5]
[5 0; 0 3]
[1 0; 0 1]
```

`omega_group`

— Method```
omega_group(e::Int, n::Int, R::Ring)
omega_group(e::Int, n::Int, q::Int)
```

Return the Omega group of dimension `n`

over the field `GF(q)`

of type `e`

, where `e`

in {`+1`

,`-1`

} for `n`

even and `e`

=`0`

for `n`

odd. If `n`

is odd, `e`

can be omitted.

Currently, this function only supports rings of type `fqPolyRepField`

.

**Examples**

```
julia> F = GF(7,1)
Finite field of degree 1 over GF(7)
julia> H = omega_group(1,2,F)
Omega+(2,7)
julia> gens(H)
1-element Vector{MatrixGroupElem{fqPolyRepFieldElem, fqPolyRepMatrix}}:
[2 0; 0 4]
```

`unitary_group`

— Method```
unitary_group(n::Int, q::Int)
GU = unitary_group
```

Return the unitary group of dimension `n`

over the field `GF(q^2)`

.

**Examples**

```
julia> H = unitary_group(2,3)
GU(2,3)
julia> gens(H)
2-element Vector{MatrixGroupElem{fqPolyRepFieldElem, fqPolyRepMatrix}}:
[o 0; 0 2*o]
[2 2*o+2; 2*o+2 0]
```

`special_unitary_group`

— Method```
special_unitary_group(n::Int, q::Int)
SU = special_unitary_group
```

Return the special unitary group of dimension `n`

over the field with `q^2`

elements.

**Examples**

```
julia> H = special_unitary_group(2,3)
SU(2,3)
julia> gens(H)
2-element Vector{MatrixGroupElem{fqPolyRepFieldElem, fqPolyRepMatrix}}:
[1 2*o+2; 0 1]
[0 2*o+2; 2*o+2 0]
```