The symmetric and exterior squares of a continuous representation #
The tensor square ContRepresentation.tprod π π of a continuous representation acts on
V ⊗[𝕜] V by π g ⊗ π g, which commutes with the flip x ⊗ y ↦ y ⊗ x. The two eigenspaces of
that flip, TauCeti.symmetricTensors and TauCeti.antisymmetricTensors, are therefore invariant
submodules, and restricting the tensor square to them gives the symmetric square and the
exterior square of π, again as continuous representations.
Realizing the two squares inside V ⊗[𝕜] V rather than as Sym[𝕜]^2 V and ⋀[𝕜]^2 V is what
makes them continuous representations at all: the carrier of a continuous representation has to
carry a topology, and a submodule of the tensor square of an inner product space does, whereas a
quotient or a subobject of a PiTensorProduct carries none. Over RCLike 𝕜, which has
characteristic zero, the two eigenspaces are the symmetric and exterior squares, which is what
the names record: the identifications are
TauCeti.symmetricTensorsEquivSymmetricPower and
TauCeti.antisymmetricTensorsEquivExteriorPower of
TauCeti/LinearAlgebra/TensorSquare.lean, and they turn the restriction of π g ⊗ π g into
SymmetricPower.map (π g) and exteriorPower.map 2 (π g).
Main definitions #
ContRepresentation.symmetricSquare: the tensor square of a continuous representation restricted to the symmetric tensors ofV ⊗[𝕜] V.ContRepresentation.exteriorSquare: its restriction to the antisymmetric tensors.
Main statements #
ContRepresentation.continuous_symmetricSquareandContRepresentation.continuous_exteriorSquare: both squares of a continuous representation are continuous.ContRepresentation.symmetricSquare_applyandContRepresentation.exteriorSquare_apply: each square acts by the restriction ofπ g ⊗ π g, which is how their characters are computed.ContRepresentation.mem_invariants_symmetricSquare_iffandContRepresentation.mem_invariants_exteriorSquare_iff: a tensor of either eigenspace is invariant for that square exactly when it is invariant for the tensor square.
Implementation notes #
Nothing here needs a group, a measure, or compactness, so the statements are made over a
topological monoid with RCLike scalars; the consumer is
TauCeti/RepresentationTheory/Compact/FrobeniusSchur/InvariantTensors.lean, where the invariants
of the two squares are what the Frobenius-Schur indicator counts.
All declarations sit in the root ContRepresentation namespace, so that
π.symmetricSquare elaborates: ContRepresentation is Mathlib's type, and
scripts/lint-dot-notation.py asks that new declarations about it not recreate its namespace
inside TauCeti. That is why the ambient TauCeti names this file consumes are brought in by
open.
The tensor square of a continuous representation acts on the tensor square by f ⊗ f, so the
symmetric tensors are one of its invariant submodules.
The antisymmetric tensors are the other invariant submodule of the tensor square.
The symmetric square of a continuous representation: its tensor square restricted to the symmetric tensors.
Equations
- π.symmetricSquare = (π.tprod π).subrepresentation (TauCeti.symmetricTensors 𝕜 V) ⋯
Instances For
The exterior square of a continuous representation: its tensor square restricted to the
antisymmetric tensors. Over RCLike 𝕜, which has characteristic zero, those are the exterior
square ⋀[𝕜]^2 V realized inside V ⊗[𝕜] V, which is what the name records; the identification
is TauCeti.antisymmetricTensorsEquivExteriorPower (see the module docstring).
Equations
- π.exteriorSquare = (π.tprod π).subrepresentation (TauCeti.antisymmetricTensors 𝕜 V) ⋯
Instances For
The symmetric square of a continuous representation is continuous.
The exterior square of a continuous representation is continuous.
The symmetric square acts by the restriction of π g ⊗ π g.
The exterior square acts by the restriction of π g ⊗ π g.
Membership in the invariants of the symmetric square, read in the tensor square.
Membership in the invariants of the exterior square, read in the tensor square.