# Integer Lattices

An integer lattice $L$ is a finitely generated $\mathbb{Z}$-submodule of a quadratic vector space $V = \mathbb{Q}^n$ over the rational numbers. Integer lattices are also known as quadratic forms over the integers. We will refer to them as $\mathbb{Z}$-lattices.

A $\mathbb{Z}$-lattice $L$ has the type ZZLat. It is given in terms of its ambient quadratic space $V$ together with a basis matrix $B$ whose rows span $L$, i.e. $L = \mathbb{Z}^r B$ where $r$ is the ($\mathbb{Z}$-module) rank of $L$.

To access $V$ and $B$ see ambient_space(L::ZZLat) and basis_matrix(L::ZZLat).

## Creation of integer lattices

### From a gram matrix

integer_latticeMethod
integer_lattice([B::MatElem]; gram) -> ZZLat

Return the Z-lattice with basis matrix $B$ inside the quadratic space with Gram matrix gram.

If the keyword gram is not specified, the Gram matrix is the identity matrix. If $B$ is not specified, the basis matrix is the identity matrix.

Examples

julia> L = integer_lattice(matrix(QQ, 2, 2, [1//2, 0, 0, 2]));

julia> gram_matrix(L) == matrix(QQ, 2, 2, [1//4, 0, 0, 4])
true

julia> L = integer_lattice(gram = matrix(ZZ, [2 -1; -1 2]));

julia> gram_matrix(L) == matrix(ZZ, [2 -1; -1 2])
true

latticeMethod
lattice(V::AbstractSpace, basis::MatElem ; check::Bool = true) -> AbstractLat

Given an ambient space V and a matrix basis, return the lattice spanned by the rows of basis inside V. If V is hermitian (resp. quadratic) then the output is a hermitian (resp. quadratic) lattice.

By default, basis is checked to be of full rank. This test can be disabled by setting check to false.

### Special lattices

root_latticeMethod
root_lattice(R::Symbol, n::Int) -> ZZLat

Return the root lattice of type R given by :A, :D or :E with parameter n.

hyperbolic_plane_latticeMethod
hyperbolic_plane_lattice(n::RationalUnion = 1) -> ZZLat

Return the hyperbolic plane with intersection form of scale n, that is, the unique (up to isometry) even unimodular hyperbolic $\mathbb Z$-lattice of rank 2, rescaled by n.

Examples

julia> L = hyperbolic_plane_lattice(6);

julia> gram_matrix(L)
[0   6]
[6   0]

julia> L = hyperbolic_plane_lattice(ZZ(-13));

julia> gram_matrix(L)
[  0   -13]
[-13     0]
integer_latticeMethod
integer_lattice(S::Symbol, n::RationalUnion = 1) -> ZZlat

Given S = :H or S = :U, return a $\mathbb Z$-lattice admitting $n*J_2$ as Gram matrix in some basis, where $J_2$ is the 2-by-2 matrix with 0's on the main diagonal and 1's elsewhere.

leech_latticeFunction
leech_lattice() -> ZZLat

Return the Leech lattice.

leech_lattice(niemeier_lattice::ZZLat) -> ZZLat, QQMatrix, Int

Return a triple L, v, h where L is the Leech lattice.

L is an h-neighbor of the Niemeier lattice N with respect to v. This means that L / L ∩ N ≅ ℤ / h ℤ. Here h is the Coxeter number of the Niemeier lattice.

This implements the 23 holy constructions of the Leech lattice in J. H. Conway, N. J. A. Sloane (1999).

Examples

julia> R = integer_lattice(gram=2 * identity_matrix(ZZ, 24));

julia> N = maximal_even_lattice(R) # Some Niemeier lattice
Integer lattice of rank 24 and degree 24
with gram matrix
[2   1   1   1   0   0   0   0   0   0   0   0   0   0   0   0   1   0   1   1   0   0   0   0]
[1   2   1   1   0   0   0   0   0   0   0   0   0   0   0   0   1   1   0   1   0   0   0   0]
[1   1   2   1   0   0   0   0   0   0   0   0   0   0   0   0   1   1   1   0   0   0   0   0]
[1   1   1   2   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0]
[0   0   0   0   2   1   1   1   0   0   0   0   1   0   1   1   0   0   0   0   0   0   0   0]
[0   0   0   0   1   2   1   1   0   0   0   0   1   1   0   1   0   0   0   0   0   0   0   0]
[0   0   0   0   1   1   2   1   0   0   0   0   1   1   1   0   0   0   0   0   0   0   0   0]
[0   0   0   0   1   1   1   2   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0]
[0   0   0   0   0   0   0   0   2   1   1   1   0   0   0   0   0   0   0   0   1   1   1   0]
[0   0   0   0   0   0   0   0   1   2   1   1   0   0   0   0   0   0   0   0   1   0   1   1]
[0   0   0   0   0   0   0   0   1   1   2   1   0   0   0   0   0   0   0   0   1   1   0   1]
[0   0   0   0   0   0   0   0   1   1   1   2   0   0   0   0   0   0   0   0   0   0   0   0]
[0   0   0   0   1   1   1   0   0   0   0   0   2   1   1   1   0   0   0   0   0   0   0   0]
[0   0   0   0   0   1   1   0   0   0   0   0   1   2   0   0   0   0   0   0   0   0   0   0]
[0   0   0   0   1   0   1   0   0   0   0   0   1   0   2   0   0   0   0   0   0   0   0   0]
[0   0   0   0   1   1   0   0   0   0   0   0   1   0   0   2   0   0   0   0   0   0   0   0]
[1   1   1   0   0   0   0   0   0   0   0   0   0   0   0   0   2   1   1   1   0   0   0   0]
[0   1   1   0   0   0   0   0   0   0   0   0   0   0   0   0   1   2   0   0   0   0   0   0]
[1   0   1   0   0   0   0   0   0   0   0   0   0   0   0   0   1   0   2   0   0   0   0   0]
[1   1   0   0   0   0   0   0   0   0   0   0   0   0   0   0   1   0   0   2   0   0   0   0]
[0   0   0   0   0   0   0   0   1   1   1   0   0   0   0   0   0   0   0   0   2   1   1   1]
[0   0   0   0   0   0   0   0   1   0   1   0   0   0   0   0   0   0   0   0   1   2   0   0]
[0   0   0   0   0   0   0   0   1   1   0   0   0   0   0   0   0   0   0   0   1   0   2   0]
[0   0   0   0   0   0   0   0   0   1   1   0   0   0   0   0   0   0   0   0   1   0   0   2]

julia> minimum(N)
2

julia> det(N)
1

julia> L, v, h = leech_lattice(N);

julia> minimum(L)
4

julia> det(L)
1

julia> h == index(L, intersect(L, N))
true


We illustrate how the Leech lattice is constructed from N, h and v.

julia> Zmodh = residue_ring(ZZ, h);

julia> V = ambient_space(N);

julia> vG = map_entries(x->Zmodh(ZZ(x)), inner_product(V, v, basis_matrix(N)));

julia> LN = transpose(lift(kernel(vG)))*basis_matrix(N); # vectors whose inner product with v is divisible by h.

julia> lattice(V, LN) == intersect(L, N)
true

julia> gensL = vcat(LN, 1//h * v);

julia> lattice(V, gensL, isbasis=false) == L
true


### From a genus

Integer lattices can be created as representatives of a genus. See (representative(L::ZZGenus))

rescaleMethod
rescale(L::ZZLat, r::RationalUnion) -> ZZLat

Return the lattice L in the quadratic space with form r \Phi.

Examples

This can be useful to apply methods intended for positive definite lattices.

julia> L = integer_lattice(gram=ZZ[-1 0; 0 -1])
Integer lattice of rank 2 and degree 2
with gram matrix
[-1    0]
[ 0   -1]

julia> shortest_vectors(rescale(L, -1))
2-element Vector{Vector{ZZRingElem}}:
[0, 1]
[1, 0]

## Attributes

ambient_spaceMethod
ambient_space(L::AbstractLat) -> AbstractSpace

Return the ambient space of the lattice L. If the ambient space is not known, an error is raised.

basis_matrixMethod
basis_matrix(L::ZZLat) -> QQMatrix

Return the basis matrix $B$ of the integer lattice $L$.

The lattice is given by the row span of $B$ seen inside of the ambient quadratic space of $L$.

gram_matrixMethod
gram_matrix(L::ZZLat) -> QQMatrix

Return the gram matrix of $L$.

Examples

julia> L = integer_lattice(matrix(ZZ, [2 0; -1 2]));

julia> gram_matrix(L)
[ 4   -2]
[-2    5]
rational_spanMethod
rational_span(L::ZZLat) -> QuadSpace

Return the rational span of $L$, which is the quadratic space with Gram matrix equal to gram_matrix(L).

Examples

julia> L = integer_lattice(matrix(ZZ, [2 0; -1 2]));

julia> rational_span(L)
over rational field
with gram matrix
[ 4   -2]
[-2    5]

## Invariants

rankMethod
rank(L::AbstractLat) -> Int

Return the rank of the underlying module of the lattice L.

detMethod
det(L::ZZLat) -> QQFieldElem

Return the determinant of the gram matrix of L.

scaleMethod
scale(L::ZZLat) -> QQFieldElem

Return the scale of L.

The scale of L is defined as the positive generator of the $\mathbb Z$-ideal generated by $\{\Phi(x, y) : x, y \in L\}$.

normMethod
norm(L::ZZLat) -> QQFieldElem

Return the norm of L.

The norm of L is defined as the positive generator of the $\mathbb Z$- ideal generated by $\{\Phi(x,x) : x \in L\}$.

isevenMethod
iseven(L::ZZLat) -> Bool

Return whether L is even.

An integer lattice L in the rational quadratic space $(V,\Phi)$ is called even if $\Phi(x,x) \in 2\mathbb{Z}$ for all $x in L$.

is_integralMethod
is_integral(L::AbstractLat) -> Bool

Return whether the lattice L is integral.

is_primary_with_primeMethod
is_primary_with_prime(L::ZZLat) -> Bool, ZZRingElem

Given a $\mathbb Z$-lattice L, return whether L is primary, that is whether L is integral and its discriminant group (see discriminant_group) is a p-group for some prime number p. In case it is, p is also returned as second output.

Note that for unimodular lattices, this function returns (true, 1). If the lattice is not primary, the second return value is -1 by default.

is_primaryMethod
is_primary(L::ZZLat, p::Union{Integer, ZZRingElem}) -> Bool

Given an integral $\mathbb Z$-lattice L and a prime number p, return whether L is p-primary, that is whether its discriminant group (see discriminant_group) is a p-group.

is_elementary_with_primeMethod
is_elementary_with_prime(L::ZZLat) -> Bool, ZZRingElem

Given a $\mathbb Z$-lattice L, return whether L is elementary, that is whether L is integral and its discriminant group (see discriminant_group) is an elemenentary p-group for some prime number p. In case it is, p is also returned as second output.

Note that for unimodular lattices, this function returns (true, 1). If the lattice is not elementary, the second return value is -1 by default.

is_elementaryMethod
is_elementary(L::ZZLat, p::Union{Integer, ZZRingElem}) -> Bool

Given an integral $\mathbb Z$-lattice L and a prime number p, return whether L is p-elementary, that is whether its discriminant group (see discriminant_group) is an elementary p-group.

### The Genus

For an integral lattice The genus of an integer lattice collects its local invariants. genus(::ZZLat)

massMethod
mass(L::ZZLat) -> QQFieldElem

Return the mass of the genus of L.

genus_representativesMethod
genus_representatives(L::ZZLat) -> Vector{ZZLat}

Return representatives for the isometry classes in the genus of L.

### Real invariants

signature_tupleMethod
signature_tuple(L::ZZLat) -> Tuple{Int,Int,Int}

Return the number of (positive, zero, negative) inertia of L.

is_positive_definiteMethod
is_positive_definite(L::AbstractLat) -> Bool

Return whether the rational span of the lattice L is positive definite.

is_negative_definiteMethod
is_negative_definite(L::AbstractLat) -> Bool

Return whether the rational span of the lattice L is negative definite.

is_definiteMethod
is_definite(L::AbstractLat) -> Bool

Return whether the rational span of the lattice L is definite.

## Isometries

automorphism_group_generatorsMethod
automorphism_group_generators(E::EllCrv) -> Vector{EllCrvIso}

Return generators of the automorphism group of $E$.

automorphism_group_generators(L::AbstractLat; ambient_representation::Bool = true)
-> Vector{MatElem}

Given a definite lattice L, return generators for the automorphism group of L. If ambient_representation == true (the default), the transformations are represented with respect to the ambient space of L. Otherwise, the transformations are represented with respect to the (pseudo-)basis of L.

automorphism_group_orderMethod
automorphism_group_order(L::AbstractLat) -> Int

Given a definite lattice L, return the order of the automorphism group of L.

is_isometricMethod
is_isometric(L::AbstractLat, M::AbstractLat) -> Bool

Return whether the lattices L and M are isometric.

is_locally_isometricMethod
is_locally_isometric(L::ZZLat, M::ZZLat, p::Int) -> Bool

Return whether L and M are isometric over the p-adic integers.

i.e. whether $L \otimes \mathbb{Z}_p \cong M\otimes \mathbb{Z}_p$.

# Root lattices

root_lattice_recognitionMethod
root_lattice_recognition(L::ZZLat)

Return the ADE type of the root sublattice of L.

Input:

L – a definite and integral $\mathbb{Z}$-lattice.

Output:

Two lists, the first one containing the ADE types and the second one the irreducible root sublattices.

For more recognizable gram matrices use root_lattice_recognition_fundamental.

Examples

julia> L = integer_lattice(gram=ZZ[4  0 0  0 3  0 3  0;
0 16 8 12 2 12 6 10;
0  8 8  6 2  8 4  5;
0 12 6 10 2  9 5  8;
3  2 2  2 4  2 4  2;
0 12 8  9 2 12 6  9;
3  6 4  5 4  6 6  5;
0 10 5  8 2  9 5  8])
Integer lattice of rank 8 and degree 8
with gram matrix
[4    0   0    0   3    0   3    0]
[0   16   8   12   2   12   6   10]
[0    8   8    6   2    8   4    5]
[0   12   6   10   2    9   5    8]
[3    2   2    2   4    2   4    2]
[0   12   8    9   2   12   6    9]
[3    6   4    5   4    6   6    5]
[0   10   5    8   2    9   5    8]

julia> R = root_lattice_recognition(L)
([(:A, 1), (:D, 6)], ZZLat[Integer lattice of rank 1 and degree 8, Integer lattice of rank 6 and degree 8])
root_lattice_recognition_fundamentalMethod
root_lattice_recognition_fundamental(L::ZZLat)

Return the ADE type of the root sublattice of L as well as the corresponding irreducible root sublattices with basis given by a fundamental root system.

Input:

L – a definite and integral $\mathbb Z$-lattice.

Output:

• the root sublattice, with basis given by a fundamental root system
• a Vector consisting of the irreducible root sublattices.

Examples

julia> L = integer_lattice(gram=ZZ[4  0 0  0 3  0 3  0;
0 16 8 12 2 12 6 10;
0  8 8  6 2  8 4  5;
0 12 6 10 2  9 5  8;
3  2 2  2 4  2 4  2;
0 12 8  9 2 12 6  9;
3  6 4  5 4  6 6  5;
0 10 5  8 2  9 5  8])
Integer lattice of rank 8 and degree 8
with gram matrix
[4    0   0    0   3    0   3    0]
[0   16   8   12   2   12   6   10]
[0    8   8    6   2    8   4    5]
[0   12   6   10   2    9   5    8]
[3    2   2    2   4    2   4    2]
[0   12   8    9   2   12   6    9]
[3    6   4    5   4    6   6    5]
[0   10   5    8   2    9   5    8]

julia> R = root_lattice_recognition_fundamental(L);

julia> gram_matrix(R)
[2    0    0    0    0    0    0]
[0    2    0   -1    0    0    0]
[0    0    2   -1    0    0    0]
[0   -1   -1    2   -1    0    0]
[0    0    0   -1    2   -1    0]
[0    0    0    0   -1    2   -1]
[0    0    0    0    0   -1    2]

ADE_typeMethod
ADE_type(G::MatrixElem) -> Tuple{Symbol,Int64}

Return the type of the irreducible root lattice with gram matrix G.

See also root_lattice_recognition.

Examples

julia> Hecke.ADE_type(gram_matrix(root_lattice(:A,3)))
(:A, 3)
coxeter_numberMethod
coxeter_number(ADE::Symbol, n) -> Int

Return the Coxeter number of the corresponding ADE root lattice.

If $L$ is a root lattice and $R$ its set of roots, then the Coxeter number $h$ is $|R|/n$ where n is the rank of $L$.

Examples

julia> coxeter_number(:D, 4)
6

highest_rootMethod
highest_root(ADE::Symbol, n) -> ZZMatrix

Return coordinates of the highest root of root_lattice(ADE, n).

Examples

julia> highest_root(:E, 6)
[1   2   3   2   1   2]

## Module operations

Most module operations assume that the lattices live in the same ambient space. For instance only lattices in the same ambient space compare.

==Method

Return true if both lattices have the same ambient quadratic space and the same underlying module.

is_sublatticeMethod
is_sublattice(L::AbstractLat, M::AbstractLat) -> Bool

Return whether M is a sublattice of the lattice L.

is_sublattice_with_relationsMethod
is_sublattice_with_relations(M::ZZLat, N::ZZLat) -> Bool, QQMatrix

Returns whether $N$ is a sublattice of $M$. In this case, the second return value is a matrix $B$ such that $B B_M = B_N$, where $B_M$ and $B_N$ are the basis matrices of $M$ and $N$ respectively.

+Method
+(L::AbstractLat, M::AbstractLat) -> AbstractLat

Return the sum of the lattices L and M.

The lattices L and M must have the same ambient space.

*Method
*(a::RationalUnion, L::ZZLat) -> ZZLat

Return the lattice $aM$ inside the ambient space of $M$.

intersectMethod
intersect(L::AbstractLat, M::AbstractLat) -> AbstractLat

Return the intersection of the lattices L and M.

The lattices L and M must have the same ambient space.

inMethod
Base.in(v::Vector, L::ZZLat) -> Bool

Return whether the vector v lies in the lattice L.

inMethod
Base.in(v::QQMatrix, L::ZZLat) -> Bool

Return whether the row span of v lies in the lattice L.

primitive_closureMethod
primitive_closure(M::ZZLat, N::ZZLat) -> ZZLat

Given two $\mathbb Z$-lattices M and N with $N \subseteq \mathbb{Q} M$, return the primitive closure $M \cap \mathbb{Q} N$ of N in M.

Examples

julia> M = root_lattice(:D, 6);

julia> N = lattice_in_same_ambient_space(M, 3*basis_matrix(M)[1,:]);

julia> basis_matrix(N)
[3   0   0   0   0   0]

julia> N2 = primitive_closure(M, N)
Integer lattice of rank 1 and degree 6
with gram matrix


julia> basis_matrix(N2)
[1   0   0   0   0   0]

julia> M2 = primitive_closure(dual(M), M);

julia> is_integral(M2)
false

is_primitiveMethod
is_primitive(M::ZZLat, N::ZZLat) -> Bool

Given two $\mathbb Z$-lattices $N \subseteq M$, return whether N is a primitive sublattice of M.

Examples

julia> U = hyperbolic_plane_lattice(3);

julia> bU = basis_matrix(U);

julia> e1, e2 = bU[1,:], bU[2,:]
([1 0], [0 1])

julia> N = lattice_in_same_ambient_space(U, e1 + e2)
Integer lattice of rank 1 and degree 2
with gram matrix


julia> is_primitive(U, N)
true

julia> M = root_lattice(:A, 3);

julia> f = matrix(QQ, 3, 3, [0 1 1; -1 -1 -1; 1 1 0]);

julia> N = kernel_lattice(M, f+1)
Integer lattice of rank 1 and degree 3
with gram matrix


julia> is_primitive(M, N)
true
is_primitiveMethod
is_primitive(L::ZZLat, v::Union{Vector, QQMatrix}) -> Bool

Return whether the vector v is primitive in L.

A vector v in a $\mathbb Z$-lattice L is called primitive if for all w in L such that $v = dw$ for some integer d, then $d = \pm 1$.

divisibilityMethod
divisibility(L::ZZLat, v::Union{Vector, QQMatrix}) -> QQFieldElem

Return the divisibility of v with respect to L.

For a vector v in the ambient quadratic space $(V, \Phi)$ of L, we call the divisibility of v with the respect to L the non-negative generator of the fractional $\mathbb Z$-ideal $\Phi(v, L)$.

## Embeddings

### Categorical constructions

direct_sumMethod
direct_sum(x::Vararg{ZZLat}) -> ZZLat, Vector{AbstractSpaceMor}
direct_sum(x::Vector{ZZLat}) -> ZZLat, Vector{AbstractSpaceMor}

Given a collection of $\mathbb Z$-lattices $L_1, \ldots, L_n$, return their direct sum $L := L_1 \oplus \ldots \oplus L_n$, together with the injections $L_i \to L$. (seen as maps between the corresponding ambient spaces).

For objects of type ZZLat, finite direct sums and finite direct products agree and they are therefore called biproducts. If one wants to obtain L as a direct product with the projections $L \to L_i$, one should call direct_product(x). If one wants to obtain L as a biproduct with the injections $L_i \to L$ and the projections $L \to L_i$, one should call biproduct(x).

direct_productMethod
direct_product(x::Vararg{ZZLat}) -> ZZLat, Vector{AbstractSpaceMor}
direct_product(x::Vector{ZZLat}) -> ZZLat, Vector{AbstractSpaceMor}

Given a collection of $\mathbb Z$-lattices $L_1, \ldots, L_n$, return their direct product $L := L_1 \times \ldots \times L_n$, together with the projections $L \to L_i$. (seen as maps between the corresponding ambient spaces).

For objects of type ZZLat, finite direct sums and finite direct products agree and they are therefore called biproducts. If one wants to obtain L as a direct sum with the injections $L_i \to L$, one should call direct_sum(x). If one wants to obtain L as a biproduct with the injections $L_i \to L$ and the projections $L \to L_i$, one should call biproduct(x).

biproductMethod
biproduct(x::Vararg{ZZLat}) -> ZZLat, Vector{AbstractSpaceMor}, Vector{AbstractSpaceMor}
biproduct(x::Vector{ZZLat}) -> ZZLat, Vector{AbstractSpaceMor}, Vector{AbstractSpaceMor}

Given a collection of $\mathbb Z$-lattices $L_1, \ldots, L_n$, return their biproduct $L := L_1 \oplus \ldots \oplus L_n$, together with the injections $L_i \to L$ and the projections $L \to L_i$. (seen as maps between the corresponding ambient spaces).

For objects of type ZZLat, finite direct sums and finite direct products agree and they are therefore called biproducts. If one wants to obtain L as a direct sum with the injections $L_i \to L$, one should call direct_sum(x). If one wants to obtain L as a direct product with the projections $L \to L_i$, one should call direct_product(x).

### Orthogonal sublattices

orthogonal_submoduleMethod
orthogonal_submodule(L::ZZLat, S::ZZLat) -> ZZLat

Return the largest submodule of L orthogonal to S.

irreducible_componentsMethod
irreducible_components(L::ZZLat) -> Vector{ZZLat}

Return the irreducible components $L_i$ of the positive definite lattice $L$.

This yields a maximal orthogonal splitting of L as

$$$L = \bigoplus_i L_i.$$$

### Dual lattice

dualMethod
dual(L::AbstractLat) -> AbstractLat

Return the dual lattice of the lattice L.

### Overlattices

glue_mapMethod
glue_map(L::ZZLat, S::ZZLat, R::ZZLat; check=true)
-> Tuple{TorQuadModuleMor, TorQuadModuleMor, TorQuadModuleMor}

Given three integral $\mathbb Z$-lattices L, S and R, with S and R primitive sublattices of L and such that the sum of the ranks of S and R is equal to the rank of L, return the glue map $\gamma$ of the primitive extension $S+R \subseteq L$, as well as the inclusion maps of the domain and codomain of $\gamma$ into the respective discriminant groups of S and R.

Example

julia> M = root_lattice(:E,8);

julia> f = matrix(QQ, 8, 8, [-1 -1  0  0  0  0  0  0;
1  0  0  0  0  0  0  0;
0  1  1  0  0  0  0  0;
0  0  0  1  0  0  0  0;
0  0  0  0  1  0  0  0;
0  0  0  0  0  1  1  0;
-2 -4 -6 -5 -4 -3 -2 -3;
0  0  0  0  0  0  0  1]);

julia> S = kernel_lattice(M ,f-1)
Integer lattice of rank 4 and degree 8
with gram matrix
[12   -3    0   -3]
[-3    2   -1    0]
[ 0   -1    2    0]
[-3    0    0    2]

julia> R = kernel_lattice(M , f^2+f+1)
Integer lattice of rank 4 and degree 8
with gram matrix
[ 2   -1    0    0]
[-1    2   -6    0]
[ 0   -6   30   -3]
[ 0    0   -3    2]

julia> glue, iS, iR = glue_map(M, S, R)
(Map with following data
Domain:
=======
Finite quadratic module: (Z/3)^2 -> Q/2Z
Codomain:
=========
Finite quadratic module: (Z/3)^2 -> Q/2Z, Map with following data
Domain:
=======
Finite quadratic module: (Z/3)^2 -> Q/2Z
Codomain:
=========
Finite quadratic module: (Z/3)^2 -> Q/2Z, Map with following data
Domain:
=======
Finite quadratic module: (Z/3)^2 -> Q/2Z
Codomain:
=========
Finite quadratic module: (Z/3)^2 -> Q/2Z)

julia> is_bijective(glue)
true
overlatticeMethod
overlattice(glue_map::TorQuadModuleMor) -> ZZLat

Given the glue map of a primitive extension of $\mathbb Z$-lattices $S+R \subseteq L$, return L.

Example

julia> M = root_lattice(:E,8);

julia> f = matrix(QQ, 8, 8, [ 1  0  0  0  0  0  0  0;
0  1  0  0  0  0  0  0;
1  2  4  4  3  2  1  2;
-2 -4 -6 -5 -4 -3 -2 -3;
2  4  6  4  3  2  1  3;
-1 -2 -3 -2 -1  0  0 -2;
0  0  0  0  0 -1  0  0;
-1 -2 -3 -3 -2 -1  0 -1]);

julia> S = kernel_lattice(M ,f-1)
Integer lattice of rank 4 and degree 8
with gram matrix
[ 2   -1     0     0]
[-1    2    -1     0]
[ 0   -1    12   -15]
[ 0    0   -15    20]

julia> R = kernel_lattice(M , f^4+f^3+f^2+f+1)
Integer lattice of rank 4 and degree 8
with gram matrix
[10   -4    0    1]
[-4    2   -1    0]
[ 0   -1    4   -3]
[ 1    0   -3    4]

julia> glue, iS, iR = glue_map(M, S, R);

julia> overlattice(glue) == M
true
local_modificationMethod
local_modification(M::ZZLat, L::ZZLat, p)

Return a local modification of M that matches L at p.

INPUT:

• $M$ – a \mathbb{Z}_p-maximal lattice
• $L$ – the a lattice isomorphic to M over \QQ_p
• $p$ – a prime number

OUTPUT:

an integral lattice M' in the ambient space of M such that M and M' are locally equal at all completions except at p where M' is locally isometric to the lattice L.

maximal_integral_latticeMethod
maximal_integral_lattice(L::AbstractLat) -> AbstractLat

Given a lattice L, return a lattice M in the ambient space of L which is maximal integral and which contains L.

### Sublattices defined by endomorphisms

kernel_latticeMethod
kernel_lattice(L::ZZLat, f::MatElem;
ambient_representation::Bool = true) -> ZZLat

Given a $\mathbf{Z}$-lattice $L$ and a matrix $f$ inducing an endomorphism of $L$, return $\ker(f)$ is a sublattice of $L$.

If ambient_representation is true (the default), the endomorphism is represented with respect to the ambient space of $L$. Otherwise, the endomorphism is represented with respect to the basis of $L$.

invariant_latticeMethod
invariant_lattice(L::ZZLat, G::Vector{MatElem};
ambient_representation::Bool = true) -> ZZLat
invariant_lattice(L::ZZLat, G::MatElem;
ambient_representation::Bool = true) -> ZZLat

Given a $\mathbf{Z}$-lattice $L$ and a list of matrices $G$ inducing endomorphisms of $L$ (or just one matrix $G$), return the lattice $L^G$, consisting on elements fixed by $G$.

If ambient_representation is true (the default), the endomorphism is represented with respect to the ambient space of $L$. Otherwise, the endomorphism is represented with respect to the basis of $L$.

coinvariant_latticeMethod
coinvariant_lattice(L::ZZLat, G::Vector{MatElem};
ambient_representation::Bool = true) -> ZZLat
coinvariant_lattice(L::ZZLat, G::MatElem;
ambient_representation::Bool = true) -> ZZLat

Given a $\mathbf{Z}$-lattice $L$ and a list of matrices $G$ inducing endomorphisms of $L$ (or just one matrix $G$), return the orthogonal complement $L_G$ in $L$ of the fixed lattice $L^G$ (see invariant_lattice).

If ambient_representation is true (the default), the endomorphism is represented with respect to the ambient space of $L$. Otherwise, the endomorphism is represented with respect to the basis of $L$.

### Computing embeddings

embedMethod
embed(S::ZZLat, G::Genus, primitive::Bool=true) -> Bool, embedding

Return a (primitive) embedding of the integral lattice S into some lattice in the genus of G.

julia> G = integer_genera((8,0), 1, even=true);

julia> L, S, i = embed(root_lattice(:A,5), G);

embed_in_unimodularMethod
embed_in_unimodular(S::ZZLat, pos::Int, neg::Int, primitive=true, even=true) -> Bool, L, S', iS, iR

Return a (primitive) embedding of the integral lattice S into some (even) unimodular lattice of signature (pos, neg).

For now this works only for even lattices.

julia> NS = direct_sum(integer_lattice(:U), rescale(root_lattice(:A, 16), -1));

julia> LK3, iNS, i = embed_in_unimodular(NS, 3, 19);

julia> genus(LK3)
Genus symbol for integer lattices
Signatures: (3, 0, 19)
Local symbol:
Local genus symbol at 2: 1^22

julia> iNS
Integer lattice of rank 18 and degree 22
with gram matrix
[0   1    0    0    0    0    0    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]
[0   0   -2    1    0    0    0    0    0    0    0    0    0    0    0    0    0    0]
[0   0    1   -2    1    0    0    0    0    0    0    0    0    0    0    0    0    0]
[0   0    0    1   -2    1    0    0    0    0    0    0    0    0    0    0    0    0]
[0   0    0    0    1   -2    1    0    0    0    0    0    0    0    0    0    0    0]
[0   0    0    0    0    1   -2    1    0    0    0    0    0    0    0    0    0    0]
[0   0    0    0    0    0    1   -2    1    0    0    0    0    0    0    0    0    0]
[0   0    0    0    0    0    0    1   -2    1    0    0    0    0    0    0    0    0]
[0   0    0    0    0    0    0    0    1   -2    1    0    0    0    0    0    0    0]
[0   0    0    0    0    0    0    0    0    1   -2    1    0    0    0    0    0    0]
[0   0    0    0    0    0    0    0    0    0    1   -2    1    0    0    0    0    0]
[0   0    0    0    0    0    0    0    0    0    0    1   -2    1    0    0    0    0]
[0   0    0    0    0    0    0    0    0    0    0    0    1   -2    1    0    0    0]
[0   0    0    0    0    0    0    0    0    0    0    0    0    1   -2    1    0    0]
[0   0    0    0    0    0    0    0    0    0    0    0    0    0    1   -2    1    0]
[0   0    0    0    0    0    0    0    0    0    0    0    0    0    0    1   -2    1]
[0   0    0    0    0    0    0    0    0    0    0    0    0    0    0    0    1   -2]

julia> is_primitive(LK3, iNS)
true

## LLL, Short and Close Vectors

### LLL and indefinite LLL

lllMethod
lll(L::ZZLat, same_ambient::Bool = true) -> ZZLat

Given an integral $\mathbb Z$-lattice L with basis matrix B, compute a basis C of L such that the gram matrix $G_C$ of L with respect to C is LLL-reduced.

By default, it creates the lattice in the same ambient space as L. This can be disabled by setting same_ambient = false. Works with both definite and indefinite lattices.

### Short Vectors

short_vectorsFunction
short_vectors(L::ZZLat, [lb = 0], ub, [elem_type = ZZRingElem]; check::Bool = true)
-> Vector{Tuple{Vector{elem_type}, QQFieldElem}}

Returns all tuples (v, n) such that n = v G v^t satisfies lb <= n <= ub, where G is the Gram matrix of L and v is non-zero.

Note that the vectors are computed up to sign (so only one of v and -v appears).

It is assumed and checked that L is positive definite.

See also short_vectors_iterator for an iterator version.

shortest_vectorsFunction
shortest_vectors(L::ZZLat, [elem_type = ZZRingElem]; check::Bool = true)
-> QQFieldElem, Vector{elem_type}, QQFieldElem}

Returns the list of shortest non-zero vectors. Note that the vectors are computed up to sign (so only one of v and -v appears).

It is assumed and checked that L is positive definite.

See also minimum.

short_vectors_iteratorFunction
short_vectors_iterator(L::ZZLat, [lb = 0], ub,
[elem_type = ZZRingElem]; check::Bool = true)
-> Tuple{Vector{elem_type}, QQFieldElem} (iterator)

Returns an iterator for all tuples (v, n) such that n = v G v^t satisfies lb <= n <= ub, where G is the Gram matrix of L and v is non-zero.

Note that the vectors are computed up to sign (so only one of v and -v appears).

It is assumed and checked that L is positive definite.

See also short_vectors.

minimumMethod
minimum(L::ZZLat) -> QQFieldElem

Return the minimum squared length among the non-zero vectors in L.

kissing_numberMethod
kissing_number(L::ZZLat) -> Int

Return the Kissing number of the sphere packing defined by L.

This is the number of non-overlapping spheres touching any other given sphere.

### Close Vectors

close_vectorsMethod
close_vectors(L:ZZLat, v:Vector, [lb,], ub; check::Bool = false)
-> Vector{Tuple{Vector{Int}}, QQFieldElem}

Return all tuples (x, d) where x is an element of L such that d = b(v - x, v - x) <= ub. If lb is provided, then also lb <= d.

If filter is not nothing, then only those x with filter(x) evaluating to true are returned.

By default, it will be checked whether L is positive definite. This can be disabled setting check = false.

Both input and output are with respect to the basis matrix of L.

Examples

julia> L = integer_lattice(matrix(QQ, 2, 2, [1, 0, 0, 2]));

julia> close_vectors(L, [1, 1], 1)
3-element Vector{Tuple{Vector{ZZRingElem}, QQFieldElem}}:
([2, 1], 1)
([0, 1], 1)
([1, 1], 0)

julia> close_vectors(L, [1, 1], 1, 1)
2-element Vector{Tuple{Vector{ZZRingElem}, QQFieldElem}}:
([2, 1], 1)
([0, 1], 1)