Vectors are small letters. Matrices are capital letters.
Matrix notation
Dimensions
\(A = [a_{ij}]\) is mn matrix: \(\in C^{m \times n}\). \(q = \min \set{m,n}\). Square A: \(m \times m\). \(\hat{A}\) is rectangular.
Matrices related to A
\((v_{1}, v_{2} \dots)\) is a column vector v.
\(a_{i}\): ith element of vector a or ith column of A. \(a_{i}^{*}\): by context: ith row of \(A\) or transpose of \(a_{i}\). \(A_{k+1:m,k:f}\): a submatrix; other unambiguous matlab notations \(A_{i,:}\) for ith row etc.. \(a_{i,j}\) an element of A. \(A_{i,j}\): by context: an element of \(A\) or a submatrix of A.
Conjugate matrix \(\conj{A}\). Adjoint (Hermitian conjugate) of A: \(A^{*}=\conj{A^{T}}\). \(\tilde{A}\): \(A\) as stored on computer; or as calculated by alg.
Dilation of matrix A: add rows and cols to A.
Special matrices
Permutation matrix, P. Lower triangular matrix L. Upper triangular matrix U. Diagonal matrix \(D = diag(d_{i}) = diag(d)\). Orthogonal (or Orthonormal) matrix Q, \(\hat{Q}\). Identity I.
Special vectors
ith col of I: \(e_{i}\) (Canonical unit vector). e or 1: col vector of 1’s. \(P_{\perp q}\): projector to space \(\perp q\).
Abused notation
\(y = O(\eps) \implies \exists c = f(m,n), \forall x \lim_{\eps_{M} \to 0} y \leq c\eps_{M}\). Extra Defn: If \(y = \frac{a(x)}{b(x)}\), at \(b(x) = 0\), \(y = O(\eps)\) means \(a(x) = O(\eps b(x))\).