Some obvious lemmas on matrices #
This file includes some necessary and obvious results for matrices,
such as matrix.mul_vec_eq
.
We also show that the trace of a symmetric matrix is equal to the sum of its eigenvalues.
two matrices $a,b \in M_n$ are equal iff for all vectors $c \in \mathbb{k}^n$ we have $a c = b c$, essentially, treat them as linear maps on $\mathbb{k}^n$ so you can have extentionality with vectors
the outer product of two nonzero vectors is nonzero
the transpose of vecMul_vec x y
is simply vecMul_vec y x
the identity written as a sum of the standard basis
Equations
- hA.eigenvectorMatrix = (PiLp.basisFun 2 𝕜 n).toMatrix ⇑hA.eigenvectorBasis.toBasis
Instances For
the trace of a Hermitian matrix is the sum of its eigenvalues
a Hermitian matrix applied to its eigenvector basis element equals the basis element scalar-ed by its respective eigenvalue
Equations
- instInnerForall_monlib = { inner := fun (x y : n → 𝕜) => inner ((EuclideanSpace.equiv n 𝕜).symm x) ((EuclideanSpace.equiv n 𝕜).symm y) }
for any matrix $x\in M_{n_1 \times n_2}$, we have $$x=\sum_{i,j,k,l}x_{ik}^{jl}(e_{ij} \otimes_k e_{kl}) $$
$(x \otimes_k y)^* = x^* \otimes_k y^*$
$e_{ij}^*=e_{ji}$
Equations
- x.toInvertibleMatrix = { invOf := LinearMap.toMatrix' ↑x.symm, invOf_mul_self := ⋯, mul_invOf_self := ⋯ }