Matrices Definition of the sum A B 2 matrices A and B can be added if they have the same shape. Let A be an m x n matrix with elements aij Let B be an m x n matrix with elements bij Then we define A B as an m x n matrix with element (aij + bij) e.g. 3 0 1 2 A B 1 7 3 4 Then A B 2 2 4 3 Definition of numerical multiple r A : Let A be an m x n matrix with elements aij. We defined r A as an m x n matrix with elements raij. e.g. 3A 9 0 3 21 By convention we write - A for (-1) A And A - B for A +(-1) B Properties of matrix addition and numerical multiplication Let A , B and C be m x n matrices and r,s 1. r A + s A = (r + s) A 2. r A + r B = r( A + B ) 3. r(s A ) = (rs) A 4. A + B = B + A 5. ( A + B ) + C = A + ( B + C ) 6. A + 0 = A 7. A + - A = 0 Matrices And Linear Form An m x n matrix describes a function transformation from Rn to Rm e.g. Transformation from R2 to R2 z1 = 3y1 – y2 R2 R2 z2 = 5y1 + 2y2 B( 3 1 ) 5 2 A( 2 3 4 ) 1 1 2 e.g. Transformation from R3 to R2 y1 = 2x1 + 3x2 + 4x3 y2 = x1 - x2 + 2x3 R3 R2 Transformation from R3 to R2 is described by 3.2 1.1 3.3 1. 1 3.4 1.2 C( ) BA 5.2 2.1 5.3 2. 1 5.4 2.2 Some properties of matrix multiplication are 1. ( A B)C AC BC 2. A( B C ) AB AC 3. (r A) B r ( AB) A(r B ) 4. A( BC ) ( AB)C 5. 0 A A0 0 6. IA AI A Proof of A( BC ) ( AB)C Let A p x q matrix B q x r matrix AB is p x r matrix C r x s matrix BC is q x s matrix q q k 1 k 1 r r r ( A( BC )) ij aik ( BC ) kj aik bki cij i 1 q (( AB)C ) ij ( AB) ik C ij aik bki cij i 1 i 1 k 1 Q.E.D Linear Functions Let A denote a linear function from Rn to Rm that satisfies: f ( x y ) f ( x) f ( y ) xy f (x) f ( x) x Theorem: A function of form Rn to Rm is linear if and only there exists an m x n matrix A such that f(x) = A x for all real values of x. Proof: Assume f is defined by f(x) = Ax where A is an m x n matrix. f(x + y) = A (x + y) = A x + A y = f(x) + f(y) x A(r x) ( A x) f ( x) Assume f is a linear function from Rn to Rm. Let j denote the jth natural basis vector in Rn. And let x be real. n x x j e j x1e1 x2 e2 ... xn en j 1 Then f ( x) f ( x1e1 x2 e2 ...xn en ) If linear f ( x) x1 f (e1 ) x2 f (e2 ) ... xn f (en ) x1 f (e j ) R m ( f (e1 ) f (e2 )... f (en )). x2 ... xn x1 f ( x) ( f (e1 ). f (en )). x2 ... xn f ( x) a11 a12 ... a1n x1 a 21 a 22 ... ... ... a 2 n x 2 . Ax ... ... ... a m1 am2 a1 j a2 j Where f(j) = ... a mj ... a mn x n Hence we have shown that if Rn Rm is a linear function then the associated matrix is A , such that f(x)= A x with the jth column been equal to f(ej). e.g. Consider the linear transformation R2 R2 that rotates anticlockwise by an angle . e2 sin , cos e1 cos , sin cos sin Theorem R sin cos Let f be a linear function so Rn transforms to Rm And let g be a linear function so Rm transforms to Rn mxn f Axy nxm Then g(f(x)) = BA x g B yz Proof g(f(x) = g( A x) = B ( A x) ( BA ) x