MATH0047: ADVANCED LINEAR ALGEBRA CHEATSHEET CHAPTER 1: Matrices and Linear Equations Size of matrix Matrix Multiplication Trace Sum of entries along main diagonal Transpose Flip matrix- make rows columns and columns rows Commutative Addition Property Distributive Property Associative Property Multiplication is NOT Commutative Transpose of Transpose Transpose of Sum Transpose of Product Trace of Product Inverse Matrices: ο· A (nxn) has inverse B (nxn) iff. > ο· If A and B are invertible (nxn) matrixes, (AB) will also be invertible with inverse Gauss Jordan Elimination- reduction to reduced row echelon form https://tinyurl.com/gaussjordan1 Conditions for Reduced row echelon form: 1) Each zero row appears below all non-zero rows 2) For each non-zero row, the first (leftmost) non-zero entry is a 1 3) The leading 1 of the lower row is to the right of the leading 1 of the row above 4) Each column that contains a leading 1 has zeros everywhere else Without (4) > matrix is in row echelon form (NOT reduced) Possible operations: 1) Multiply by non-zero number 2) Add π times a row to another row 3) Permute rows i and j ο· ο· System of real linear equations can have- 0 or 1 or ∞ solutions Can generate ∞ solutions from 2 solutions by taking combinations that add up to 1 Homogenous System: ο· Can have 1 (zero solution) or ∞ solutions Elementary Matrices ο· Multiplying (on left) by an elementary matrix corresponds to a row operation ο· Conduct the row operation on the identity matrix to obtain the elementary matrix Now AB is the same as doing the row operation on B Forms of elementary matrices: Same as identity matrix I except that-1) Diagonal Matrix One non-zero entry on diagonal 2) Elementary row operation One non-diagonal entry is non-zero 3) Permute/ Swap Rows i and j are interchanged Finding the inverse of Aο· Take augmented matrix with A on LHS, I on RHS ο· After row reduction, if LHS is I, then RHS is inverse ο· To get A-1 > flip the order of operations and inverse them Properties of invertible matrix A: 1. There is a unique sol to 2. Row reduced echelon form of A is In 3. A can be expressed as a product of elementary matrices CHAPTER 2: Determinants Det=2 Det<0 Det=0 Volume of object has doubled Reflection/ Flip Object has shrunk completely (lost) ππππ(π) = (−1)π π€βπππ π ππ π‘βπ ππ’ππππ ππ πππ£πππ ππππ ο· Inversion- Permutation when a larger number precedes a smaller one ο· Ex: (3,2,1) has 3 inversions Elementary Productο· Exactly 1 entry from each row and column ο· Ex: Determinant of Sq Matrixο· Sum of all signed elementary products of A If: Minor of A: Cofactor of an entry Aij Cofactor Expansion Method Cofactor expansion along row 1: Determinant results: 1. If A has a zero row or column, detA=0 2. 3. For upper, lower and diagonal triangular matrices, Product of entries along diagonal 4. 5. π·ππ‘(π΄−1 ) = 1 det π΄ But note > det(A+B) ≠det(A) +det(B) Operations and their effects on det(A) A is invertible iff det(A)≠0 CHAPTER 3: Vector Spaces ο· ο· Multiplication of vector by constant π can be interpreted as a ‘stretch’ Minus sign flips/ reflects the vector Real Vector Spaces Dot/ Inner Product Properties of Inner Product For real vectors Length/ Norm of vector Orthogonal Vectors Angle b/w 2 vectors (dot product) Cauchy-Schwarz Inequality Triangle Inequality Complex Vector Spaces Length of Complex Number a Inner Product Of complex vectors Properties of Inner Product For Complex Vectors Polynomials (degree ≤ 2) Inner Product For Polynomials Properties Norm of Polynomial Subspace of Rn is a subset S of Rn satisfying: i) 0 vector belongs to S ii) If iii) If Span: ο· A set of vectors (v1,…,vm) spans a subspace V if every vector in V can be written in the form: ο· Check if a vector lies in the span of 2 vectors using row reduction- Linear Independence ο· If Ax=0 ONLY has a zero solution > linearly independence ο· Otherwise, they are linearly dependent ο· Independent when A is an invertible matrix [i.e., det(A) ≠ 0] ο· Row reduced form should be identity matrix for linearly independent vectors BASE: ο· Can obtain any vector in Rn through a linear combination of the vectors in the set ο· Set {v1,...,vm} is a basis for V (subspace of Rn) ifo Set {v1,...,vm} spans V o v1,...,vm are linearly independent Dimension: Dim(V)= no. of vectors in a basis for V Orthonormal Vectors: 1. Each vector has norm 1 2. Any 2 vectors are orthogonal to each other Gram-Schmidt (Orthonormalization) Process For 2 vectors u1 and u2 1. Reference vector 2. Find 3. Make norms 1: Orthogonal Complement: If V is a subspace of Rn, then π ⊥ will also be a subspace of Rn Fourier Series: CHAPTER 4: Linear Maps Linear Map ο· Linear Map T from Rn to Rm satisfying i) T (0) = 0 ii) iii) Linear Map ο matrix: Kernel Set of vectors that the map sends to 0 Image Set of vectors attained by the map Nullity Dimension of kernel Rank Dimension of image π πππ(π) + ππ’ππ(π) = π Eigenvectors: ο· Eigenvector of A is a non-zero vector v in Rn such that: ο· π is the eigenvalue Find eigenvalues by solving the characteristic equation- ο· Find eigenvector by substitute in the value of π and solve (π΄ − ππΌ)π₯ = 0 ο· ο· Eigenvector v with eigenvalue π for A is also eigenvector for π΄π with eigenvalue ππ Any non-zero vector in the kernel of a linear map is an eigenvector with eigenvalue 0 Algebraic Multiplicity: ο· No. of times the eigenvalue π appears as a root of the characteristic equation Geometric Multiplicity: ο· Largest no. of linearly independent eigenvectors corresponding to an eigenvalue π Diagonalisation: ο· A is diagonalisable if there exists an invertible matrix P such that P-1AP is a diagonal matrix ο· Matrix CANNOT be diagonalised if algebraic multiplicity ≠ geometric multiplicity Using the diagonalised matrix of A to find a formula for π΄π : Hermitian/ Self-Adjoint Matrix Symmetrix Matrix Properties of Hermitian Matrix A: i) A is diagonalisable ii) Each eigenvalue of A is a real number iii) A has an orthonormal set of n eigenvectors