A basis of a symmetric tensor power #
A basis b : Basis κ R M induces a basis of every symmetric tensor power Sym[R]^n M, indexed by
the unordered n-tuples Sym κ n of basis indices: the basis vector at s is the pure
symmetric tensor of the basis vectors listed by s. This is the symmetric counterpart of
Basis.piTensorProduct, whose index type is the ordered tuples Fin n → κ, and of
Module.Basis.exteriorPower, whose index type is the n-element subsets.
The coordinate map is built from the tensor-power basis: reading a pure tensor of basis vectors as
the unordered tuple of its indices is invariant under permuting the factors, so it descends to the
symmetric power by SymmetricPower.lift. Its inverse sends an unordered tuple to the pure
symmetric tensor of the corresponding basis vectors, which is well defined because any two
orderings of an unordered tuple differ by a permutation.
Main definitions #
SymmetricPower.tprodOfSym: the pure symmetric tensor whose factors are listed, without order, by a point ofSym κ n.Module.Basis.symmetricPower: the induced basis ofSym[R]^n M.
Main results #
Module.Basis.map_symmetricPower_of_apply: an endomorphism diagonal in a basis is diagonal in the induced basis of the symmetric power, with eigenvalue the product of the eigenvalues listed by the index.SymmetricPower.finrank_eq: the rank ofSym[R]^n MisNat.multichoose (finrank R M) n.Module.Basis.trace_map_symmetricPower_of_apply: the trace onSym[R]^n Mof an endomorphism that is diagonal in a basis is the sum, over the unorderedn-tuples of basis indices, of the product of the corresponding eigenvalues. This is the complete homogeneous symmetric polynomial in the eigenvalues, and is the symmetric counterpart ofModule.Basis.trace_map_exteriorPower_of_apply.
Pure symmetric tensors indexed by unordered tuples #
The pure symmetric tensor whose factors are the members of a family v : κ → M listed, with
multiplicity but without order, by a point of Sym κ n.
Equations
- SymmetricPower.tprodOfSym R v s = ⨂ₛ[R] (i : Fin n), v (SymmetricPower.orderOfSym✝ s i)
Instances For
Reading off an ordered tuple of factors gives the pure symmetric tensor of those factors: the
choice of ordering hidden in the definition of tprodOfSym does not matter.
The coordinate map in a basis #
A basis of the symmetric tensor power: a basis of M induces a basis of Sym[R]^n M
indexed by the unordered n-tuples of basis indices.
Equations
- Module.Basis.symmetricPower n b = { repr := SymmetricPower.linearEquivFinsupp✝ n b }
Instances For
The basis vector of Sym[R]^n M indexed by an unordered tuple s : Sym κ n of basis indices
is the pure symmetric tensor SymmetricPower.tprodOfSym R b s of the corresponding basis
vectors.
An endomorphism diagonal in a basis is diagonal in the induced basis of the symmetric
power: the basis vector indexed by s is an eigenvector, with eigenvalue the product of the
eigenvalues listed by s. Summing these eigenvalues gives the trace,
Module.Basis.trace_map_symmetricPower_of_apply.
Freeness and rank #
A symmetric power of a free module is free, on the unordered tuples of basis indices.
The rank of a symmetric power of a finite free module: the number of unordered n-tuples of
basis indices, Nat.multichoose (finrank R M) n.
Traces #
If an endomorphism of a semimodule over a commutative semiring is diagonal in a finite basis,
then its trace on the nth symmetric power is the sum, over the unordered n-tuples of basis
indices, of the product of the corresponding eigenvalues.