Gram Matrices of Monomial Matrices
Abstract
Exact diagonals obtained from monomial matrices and their transposes.
Theorem 1.1 (Diagonal conjugation by a monomial matrix).
Proof. Machine-checked in Lean as D5/S3/Quantum/Matrix/MonomialGram.monomial_mul_diagonal_mul_transpose (✓ std3). ∎
Source. Repository-derived.
Commentary.
Conjugating a diagonal matrix by a monomial matrix gives a diagonal matrix whose entry at i is the weight read at the permuted index sigma(i), times the square of the row scale c(i).
No mixing occurs because a monomial matrix has at most one nonzero entry per row: a surviving term needs the same column index for two rows, which forces the rows equal because sigma is injective.
The frozen isDiag_monomial_mul_diagonal_mul_transpose asserts only that this conjugate is diagonal, a property with no entry values; this module sharpens that to the exact diagonal, while the frozen statement is neither restated nor amended.
Theorem 1.2 (Gram matrix of a monomial matrix).
Proof. Machine-checked in Lean as D5/S3/Quantum/Matrix/MonomialGram.monomial_mul_transpose (✓ std3). ∎
Source. Repository-derived.
Commentary.
Taking d = 1 gives the Gram matrix of the monomial matrix itself: its diagonal entries are the squares of the row scales.
References
- Truth anchor:
D5/S3/Quantum/Matrix/MonomialGram.monomial_mul_diagonal_mul_transpose - Truth anchor:
D5/S3/Quantum/Matrix/MonomialGram.monomial_mul_transpose - Dependency: D5/S3/Quantum/Matrix/MonomialDiagonalPreserving