The commutant of a 2 × 2 matrix #
A 2 × 2 matrix over a field is regular as soon as it is not scalar: it is then cyclic
(nonderogatory), and the matrices commuting with it are exactly the polynomials in it. In size 2
those polynomials are the affine combinations a • 1 + b • M, so the commutant of a non-scalar
M : Matrix (Fin 2) (Fin 2) F is the two-dimensional algebra F[M] — in particular it is
commutative.
Mathlib's Matrix.mem_range_scalar_iff_commute_transvectionStruct describes the matrices commuting
with everything; what is proved here is the commutant of a single matrix, in size two, read off
the three independent entry equations of M * N = N * M by splitting on which of the two
off-diagonal entries or the diagonal difference is nonzero.
Being scalar is recorded over a semiring; only the commutant computation itself, which divides by a matrix entry, needs a field.
Main results #
TauCeti.mem_range_scalar_fin_two_iff: a2 × 2matrix is scalar exactly when its off-diagonal entries vanish and its two diagonal entries agree.TauCeti.commute_fin_two_iff: a matrix commutes with a non-scalar2 × 2matrixMexactly when it is of the forma • 1 + b • M.TauCeti.commute_of_commute_fin_two: two matrices commuting with a common non-scalar2 × 2matrix commute with each other.
References #
The commutant is used to compute centralizers in GL₂ in
TauCeti.LinearAlgebra.Matrix.GeneralLinearGroup.Centralizer, for the
character theory roadmap,
Layer 9, "The conjugacy classes (a build target)".
The commutant of a non-scalar 2 × 2 matrix. A matrix commutes with a non-scalar
M : Matrix (Fin 2) (Fin 2) F exactly when it lies in the two-dimensional subalgebra spanned by
1 and M; equivalently, a non-scalar 2 × 2 matrix is cyclic, so its commutant is F[M].
The hypothesis is necessary: everything commutes with a scalar matrix.
Two matrices commuting with a common non-scalar 2 × 2 matrix commute with each other: both
lie in the commutative algebra F[M].