Definition: A set of mn numbers (scalars) arranged in the form of a rectangular array denoted by or Ay a a 2 2 \ : Fm Am? Sinn where m is the number of rows and 7 is the number of columns is called a matrix of order m x n. The numbers 2,1. 0,3, +g, aFe Called the elements of the matrix. The marrix 4 can be represented by A = (a,],,.,, where i= 1, 2... mand j=, 2,...., n. In the matrix the Horizontal lines are called rows or row yectors and vertical lines are DETERMINANT OF A MATRIX Let M be the set of all square matrices of order 7. Then det : M— R be a map defined by det (4) =| 4], Let 4 be a square matrix of order 2 given by _ ie e 4) Ay Then |A|=|% ay, ayy | ~ 2% ~ 420, and let A be a square matrix of order 3 given by 1 2 ay A 42) dy day 4 Gy ayy Then | A | =a, 142943, ~ daydy,) ~ Arata tt) + 05, aay) columns or column vectors. The (i,j) gj, can is denoted by a,, that is the first suffi, amt to the number of rows andthe Second sufi bl the numberof columns. For example 2,5 he gg of intersection of 3" row and 4' column Of the mate Here, there are few examples of the matrix, 123 A145 GL, Similarly, we can define the determinant square matrix of order n, Example 1: Find the determinant of the matrix 123 456 769) Solution: The determinant of the matrix A is 124 l4|=|4 5 6 789 = 1(AS ~ 48) ~ 2(36 - 42) + 332 - 35) : 3+12-9=0 Minors and Cofactors of a Determinant: Let4 determinant of order n. If we eliminate the fro column of the determinant 4, then we obtain & of ary be any and j® 82Mare determinant of order (1 ~ 1) called the minor of the (i,j) element of the determinant 4 denoted by M,, ‘And the cofactors of (i, /)"* element ie. a, element is defined by a c= CDM, Example 2: Find the minors and cofactors of the determinant 123 Al=|4 56 789) Solution: The minors are given by and the cofactors are given by Cy = DEM, = 3, C= 6, Cy = CDP My, = 6, Cop 1 = CI My, = -3, C, Properties of determinant (i) The value of a determinant does not change when rows and columns are interchanged. (ii) If any two rows (or columns) of a determinant are interchanged. Then the value of determinant is multiplied by -1. (ii) If any row (or column) of a matrix 4 is completely zero, then | 4 |= 0. Also if any two rows (or columns) of a matrix are identical, then | 4 |=0. (iv) If all the elements of the one row (or column) ofa determinant are multiplied by same number ALGEBRA OF MATRICES (a) Addition of Matrices: Let A = [a,] and B= {b,] be two matrices of same order m * n. Then their sum (4+ B) is defined to the matrix C of the same order m xn obtained by adding the corresponding elements of given matrices A and B that is A+B where and = 1, 2, coy f= 1,2, 05 In other words if 83 the ‘then the value of determinant is times th value of given determinant. Thus. if ay a a3 ay aq | = 141 lay dp dbs [Aan ain ais Then| kay) a as fay ay ays] (0) If A be mrowed square matrix and & be any sealar, then [ka |= R14 | (vi) In a determinant the sum of the products of all elements of any row (or column) with the cofactors of corresponding elements of the same row (or column) is equal to the determinant value. Thus, if =KHAL My 2 Hy jAj=|an aa as na Then JA]= a Cy tay aC Coy + AyCy9 + Ay3Cyy Fay Cy + a5,Cyq + 033C35 Fay Cy | ay Ca, | a54C3, AC) + An Cy* aC, ie 3Cia + OxCoy + O30 yy (vif) Ina determinant the sum of products of the all elements of any row (or column) with the cofactors of corresponding elements of other row (or column) is zero. (viii) If the elements of a row (or column) of a determinant are added k times the corresponding elements of another row (or column), the value of determinant obtained is equal to the value of original determinant. (i) [a"|= (4 1" " in My, Ay 4, elt Amt Ama Foun Jeux by ba bn by by bay andB=| >: s byt Ona Pan84 ™ thy aa thy Ay + On ay 2 ay thy On thn ayy tbr» thy at 4+R-C> 4, +b, diy tgs nn Pm in * Pron Sen \dition of Matrices Properties for Ad m {i Commutative Property: If-4 and B are two matrices of same order m1, Then, A+B=B+A Thus, matrix addition is Commutative, (ii) Associative Property: If 4, B and C are three matrices of same order, then (A+ B)+C=A+(B+C) Thus, matrix addition is associative, (iii) Additive Identity: For each matrix 4 of order m xn, there exist a null matrix O of same order m x n such that A+O=A=O+A The null O matrix is called Additive Identity. (iv) Existence of Additive inverse: Let A be a matrix of order m * n, then the matrix B is said to be Additive inverse of 4 if : Let A, B and C be three comparable matrices of same order m m, then A+B=A+C implies B= Cis known as lefi Cancellation law. And B+ A= C +A implies B= C is known as right Cancellation law. (6) Scalar Multipli jon of matrices: Let A = {a,] be an m x n matrix and k be a scalar, then the scalar multiplication of matrix A by kis defined by the matrix B of order m * n whose each clement is multiplied by k. be two matrices of same order m % m and k, ky, ky be scalars. Then (i) (A + B) = kA + kB Gi) (ke, + kA = kd + hy Gi) k (kA) = (yk) (iv) CRA) = CDA = kA) Multipl 7 s is possible only when the number of first matrix is equal to the number of rows gp Kng, matrix, where the first matrix is said to be preg, second matrix is said to be post factor and these matrices are called Conformable for multipticayigg Po Let A = [ay] be matrix of order m x » ing [by] be matix of order” p. Then the NBs is a matrix C of order m * p thay; ™ AB=C=[c,] A where ey= Ls 4aby Properties of Multiplication of Matrices (i) Commutative Property: Let 4 and 2 yg matrices of appropriate order, then in AB #BA ‘Thus matrix multiplication is not Commu, Infact if the product AB exist, then it jg» necessary thatthe product BA will als ge! For example, 0 Beneray 12 3 lg teta=|4 0 3],e-|2 3 78-1 3.4 ‘Then the product AB is exist because thy number of columns of 4 is equal to the numbey of rows of B. And 12 3] 5] [is ay AB=|4 0 3//2 3) =|13 3 7 8-113 4] [20 55 But the product BA does not exist because the number of columns of B is not equal to the number of rows of A. (ii) Associative property: Let A, B and C be three matrices of appropriate order, then A(BC) = (AB)C Thus the matrix multiplication is associative. (iii) Distributive law: Let A, B and C be three matrices of appropriate order, then A(B + C)=AB+ AC called left distribution law with respect to addition and (B+ OA=BA+CA called right distribution law.with respect addition, Thus, the matrix multiplication is distributive with respect to addition.85 ion law: Let A, B and C be three matrices of appropriate order, then AB = AC implies B= C provided A is non-singular called Cancellation law of matrices. (v) Existence of Multiplicative identity: Let 4 be any square matrix of order n, then there exist identity matrix / of order n such that RELATED MATRICES (a) Submatrix: A matrix obtained from a given matrix by deleting some row or column or both is called a 123 submatrix. For example: IfA=|4 5 6 231 12 23 wens“ [; | andA,=|5 6| are submatrices of A. af (6) Transpose of a matrix: If in a matrix 4, we interchange all rows by columns and all columns by rows, then the new matrix so obtained is called transpose of the matrix 4 and is denoted by AT ot 4’, 123 For example: 1fA=|6 4 5] then transpose 231 162 of A=AT=|2 4 3 351 Properties of Transpose of a Matrix: Let A’ and BT be the transpose of matrices A and B respectively then (AY =A (A + BY = AT BT (iii) (A) = 2A where 2 being any scalar. (iv) (4BY = BAT (©) Adjoint of a matrix: Let 4 = (ay},,, be a square matrix of order n x m and C,, be the cofactor of a, Then the transpose of the matrix of cofactor is called adjoint of the matrix A and is denoted by adj (4). Thus, If ay a2 ay A= lay ay ayy ay ay ay Js,5 Al=A=IA 00... 0 0 0 0 where J=| 0; ; 000.1 is called multiplicative identity of matrix 4 Then,adj A = Transpose of cofactor matrix Cu On Cy Co Ca Ce Cs Cx Gs Properties of Adjoint of Matrix: Let A be the square matrix of order mand J be the identity of same order nm. Then, (i) A (adj A) =| A [I= (adj A) A (ii)| adj A|=| A |?" provided | 4 | # 0 in general [adjagj...adj A] = | A" (ili) adj (adj A) =A (iv) lf] | #0, then 1 | aaal= (ar) (») If kis any scalar number, then adj (kA) = (adj A) ket (vi) adj AT = (adj A)" (@ Inverse of a matrix: Let A be a square invertible matrix, then there exist another matrix B such that AB = BA = J where is a unit matrix, then the matrix B is called inverse of the matrix A and is denoted by B = A and is defined by adja we Tay Poised | 41 #0 Properties of Inverse matrix (i) Inverse of a square matrix 4 exist if and only if [4 [#0 that is 4 is non-singular. (ii) Inverse of a square matrix ifit exist is unique. (iii) Let A and B are non-singular square matrices of same order then A and B are inverse of each other if and only if AB = = BAMan : 86 a i ay Tyra te a non-singular square matrix then (oi Let A be a non-singlar square may aed land Let 4 and B be square matrices of same order Then 4B is invertible if and only if both A and are invertible and (48) '= BA! (uy Let 4 Be a non-singular square matrix, then ated TYPES OF MATRICES W.R.T. SIZE w (9 Row matrix: A matrix of onder | * 1 is called Row matrix, That is A matrix in which the number of columns is more than 1 but the umber of rows is 1, For example, [yz], 5 is @ row matrix of order equal to 1% 3 (ii) Column matrix: A matrix of order m * 1 is called Column matrix that is A matrix in which the number of rows is more than | or equal to cone but the number of column is one ) Horizontal matrix: A matrix of order m * n where 7 > m is called Horizontal matrix that is 2 matrix in which the number of columns is more than the number of rows is called Horizontal matrix. abed Forexample:|¢ f gh ie matrix of order 3 * 4. is aHorizontal bea A 'y th (viii) Only for a 2 * 2 matrix A d -t | ad—bel-c 4. (iv) Vertical matrix: A matrix of order m x, n> niscalled vertical matrix that isa pga which the number of FOWS is more hy number of columns he | le: isacolumn mat © Sox 3x1 (1) Rectangular matrix: A matrix of order where m #7 called rectangular mati thy” a matrix in which the number of rows: different from the number of column, 12 Forexample:}3. 4 5 6 of order 3 * 2 is arectangular may Ded ‘Note: Horizontal matrix and vertical matrix are special kind of Rectangular matrix. (vi) Square matrix: A matrix of order m *-n where ‘m= nis called square matrix of order that is a matrix in which the number of rows is equal to the number of column. ay Oy For example: isa ay a Om Oa Ryleen square matrix of order n. The element a), a, sy are called diagonal element and the ine going from left hand top comer to right hans bottom comer having elements a), 335.0 is called principal diagonal of the square matrix 4. TYPES OF MATRICES WITH RESPECT TO VIRTUE (i) Null matrix: The matrix whose all elements are ero is called null matrix or 2ero matrix in other word the matrix A = [dj}y,, i8 null matrix if 4, = 0 forall i,j Itis usually denoted by 0. Note: Zero matrix or null matrix may not be square matrix, isa zero cco eco cos ba 0 For example: O = |0 0 4. matrix of order 3Marr Gi) Dia ix: Bigonit shatrixs A square matrix whose all other than the leading diagonal or main diagonal elements are zero is called diagonal en is called diagonal in’ other word the matrix nen is diagonal matrix if a, =0 for all 7, i Note: 1. Diagonal matri trix must be @ square matt 2. Zero matrix isa diagonal mati. Properties of Diagonal matrix (diag fa, b, c] + diag [x y vers diag [a xb (ii) diag [a b c] x di alma ii {diag [a b ly" a fy ra a [ax by cz] (0) {diag [@ 6 c}}-" = diag fa"! 6"! e*}) Note: Every scalar matrix is a diagonal matrix. (iv) Unit or identity matrix: Diagonal matrix = [4] .» i8 called unit or identity matrix ifits diagonal elements are unit that is one, usually it is denoted by 1, a i [ il ‘or example: [, = . 7 [0 | I2x2 87 > J and j> i usually itis denoted by D- No 0 For example: |Q\2\0} isa diagonal matrix aS of order 3. (iii) Scalar matrix: Diagonal matrix D = [a,] is called scalar matrix if al its diagonal element are equal. 0 0 For example: |O\XQ\0| is ascalar matrix. 0 ON Gs. 100 1000 =|0 1 0) Vt, one eet o10 pv ooo! lees are identity matrices. Note: Every identity matrix is a scalar matrix and hence diagonal matrix. Properties of identity matrix: (i For any square matrix A AI=IA=A (iI is the identity element for matrix multiplication so it is called multiplicative identity. (iii) Det (1D) (@yret or = (v) Triangular matrix: There are two types of triangular matrix as follows: (a) Upper Triangular matrix: A square matrix [a,) is said o be upper triangular matrix if, =0 forall > j nother words a square matrix is said to be upper triangular matrix ifall its elements below the leading diagonal are zero 105 For example: [D2 4] is an upper 3 triangular matrix. (b) Lower triangular matrix: A square matrix A= [aq] is said to be lower triangular matt if {aj} = 0 for all 1 3 are orthogonal matrices. Example 3: Let A cosa 0 sina o 1 0 msina 0 cosa Then show that A is orthogonal. Solution: We are given that cosa 0 sina o 1 0 A ~sinae 0 cost Therefore, cosa 0 sina A=! 0 1 0 sina. 0 cosa. Now, cosa 0 sina ]{cosa 0 -sina A=] 0 1 0 |} O 1 0 sina 0 cosa][sina 0 cosa cos? a+ sin® a 0 ~cos asin a+ .cos a sina: 0 1 0 -sin cos asin cosa 0 sin?a-+cos? a 100 ={0 1 Ol=/ oor Hence, A is an orthogonal matrix. Properties of orthogonal matrix. (i) If 4 isan orthogonal matrix, then det (4) =+ 1 For, since A is orthogonal matrix, therefore, AAT =I Taking the determinant on both sides, we get det (AAT) = det (/) or equivalently det(A) + det (47) = 1 [det (ADF Si det (A= #4 orequivalently / tin Ie isan orthogonal AUN, then gi. singular and °” aed For, sinee Ais orthogonal matrix, therefog yy Multiplying by A on both sides, we pep" (AAA = TA orequivalently 4 (ATA) = IA. Since |A| = #1 #0. Therefore, Dividing both sides by A, We get ATA Hence, AAT=1= ATA. Thus, A is non singuay ay Atedl. (ip Transpose of an onthogonal matrixsorthogy For, let A is an orthogonal matrix, then 1 AAT =1= ATA taking the transpose on both ay sides, we get or equiralently (AAT = I= align (jv) Product of two orthogonal matrices ig gy ‘orthogonal matrix. For, Let 4 and B be two orthogonal matric, we need to show that AB and BA are orthogong, matrices. (AB) (AB)" = (AB) (BTAT) = A(BBT ygr Using BBT = J, we get (AB) (AB)! = AAT Using AA = I, we get (AB) (AB)" = I. and (BA) (BA)" = (BA) (ATBY) = BCAA") Br Using AAT =I, we get (BA) (BAY! = BB Using BBT = I, we get (BA) (BAY! = 1. (v) Inverse of an orthogonal matrix is orthogonal matrix. For, Let 4 is an orthogonal matrix. Then 4"= 1= ATA Taking the inverse on both sides, we get (AAT)! = Ph = (tay orequivalently (47)"(4-!) == (4-4) (47) or equivalently (4 “")"(4-') = 1 = (4-4) (4-7Marr. ed: Show thatthe marie ~ 4] 4 4 7] . 1-2) is orthogonal of Solutton: For orthogonal Solution: For oogony of «mate, we have to 1a Hence, i 2 1 Now, 4’ wie t ‘ x be Hence, the given matrix ‘4’ is orthogonal matrix. (iv) Conjugate of a Matrix Let z =x + iy be a complex number then Z called the conjugate of complex number z. The matrix obtained from any given matrix A on replacing its elements by the corresponding complex conjugate number is called conjugate of the matrix 4 represented by A.. Thus if 4 = [uj], ,is any matrix, then, | I} Properties of Conjugate of a Matrix iis a, conjugate of A is A = [GijImxn- Iti For example: \f a=[i z_[i-i Then A fe Let A and B be the conjugate matrices of matrices A and B respectively, then (i) (A) =A (i) AEB)=ALB (iti) ia) = A. where k is any scalar number (iv) If A and B are conformable for multiplication then (AB) = (BA) 0) a=" (vi) A=A iff is purely real matrix (vil) A=—A iff is purely imaginary matrix. 91 Tranjugate or Transpose of Conjugate ofa Matrix The transpose of conjugate of any given mat by interchanging the rows and columns of of the matrix denoted by A° called Tranjugate of the ™ trix A isobtained fF the conjugate trix A.Thus A= (AyT =(A") For example, Let lei Ii a(n | ier ie @>|4n - lei “itl ist -i- oe et = itl 4i+D. Properties of Tranjugate of a Matrix Let A® and B? be the tranjugate of the matrices A and B respectively then (AY =4 (i) (A= BP =A EB (ii) (ka)? = F A®, where is any scalar number. then thus as (iv) If A and R are conformable for multiplication, then (4B)° = B°A® (4m? = 7. (v) Unitary Matrix ‘A square matrix 4 of order m is called unitary matrix if AAS = I= APA Lisi For example, Let A= =[ . Blo Then and ei and A°A i teeter B Thus, 4 is unitary matrix.MATH MG, 2 : Properties of Unitary Matrix () Let isan unitary matrix, then A" is also unitary matrix. For, since 4 is an unitary matrix, therefore, AA = T= AMA Taking the inverse on both sides, we get (Aa8y1 = Fh = (aay! (yl4at ANA orequivalently (A841 = 7= AAO (i) Let 4 is an unitary matrix, then A is also unitary matrix, for. since A is an unitany matrix. Therefore, or equivalently. AA? =1= AA Taking the transpose on both sides, we get (4ayr aeayr orequivalently (4°74 ATA orequivalently — (A7)°47 =] = AT(AT YP (iii) Let 4 and B are two unitary matrices of same order, then AB and BA are also unitary matrices of same order. For. (AB) (ABY® = (AB) (B°4°) = A(BBS)A® Using BB® = J, we get (AB) (AB)? = Aa® Using A4® = I, we get (AB) (AB)? = 1. and (BA) (BA)? = (BA) (4°B°) BiAA®) BP Using 44? = J, we get (BA) (BA)® = BB® Using BB® = I, we get (BA) (BAY? = 1. (vi) Hermitian Matrix ‘A square matrix A is said to be Hermitian matrix if =A For example, Iti 1-i Let 224i 2-1 3 l-i 1+i Then 2 2-1 241 3 and Hence, 4 is Hermi matrix for any real scal Properties of Hermitian Matrix = is Hermit (i) A square matrix A= [4j]y Feritan ma, if and only if a= Hi, a 4, forall ij. iti ix is necessaril; (ii) Hermitian matrix is YB 8 Su matrix. : (iti) Ina Hermitian matrix, the elements on py diagonal are real numbers. For, let A =[ag]y 0384 Hermitian mati, yy a, for all i,j Neipay we gel Fi = 4, for alt; Hence, a, are real numbers for all i. Tae as “ ‘ (iv) IfA is Hermitian matrix, then isalso Hemi ar number k. ‘Substituting, For, (AYP =k Ae since kis real and A® = 4, we get (kA = kA (9) If A and Bare Hermitian matrices of same orde, then k,A # k,B are also Hermitian matrix of same order for any real scalars k, and k,, a Ay? + (kB)? =k Ao tks Using A= A, B=B and i = kh =k, we get (k, A# kB) =k ASB. (vi) If A and Bare Hermitian matrices of same order, then AB is Hermitian matrix if and only if AB = BA. For, (AB)® = B8A® A® = A, B® = B, we get (AB)® = BA AB = BA, we get (AB)? = AB (vii) If is any square matrix, then 44° and 4°4 ar using, using (449)? = (4)°4® (4) = 4, we get (449)? = 44° and (42a)? = 4498 using (4°) = A, we get (494)® = A°A. (viii) IfA and B are Hermitian matrices of same ord, then AB + BA is also Hermitian matrix. For, (AB + BA)? = (4B)° + (BA)® = B°4° + ABMaran using 4° =4, BP = B, we get (AB + BA)” ; = BA+AB= AB + BA (ix) 1C-4 is any Hermitian mattis, then all positive integral powers of 4 are Hermitian matrix. For, (ary? = (ay using A®= A, we get (4") (2) IFA is a Hermitian matrix. Then A is also Hermition matrix. For, (A) = (a®) using AP =4, we get (A)? (ai) If A is any square matrix, then 4 + A® is a Hermitian matrn. For, (A+ A) = 48 + (49)? using (A°)=A, we get (A + A%®= 4? +A4=AtA® (xii) Any square matrix can be expressed as 4 + iB, where 4 and B are Hermitian matrix. (vii) Skew Hermitian Matrix ‘A square matrix A is said to be skew Hermitian matrix if AP= HA. For example, Let Then Hence, A is Skew Hermitian matrix. Properties of Skew Hermitian Matrix (i) A square matrix A= (a), ., is Skew-Hermitian matrix if and only if 4G, =-a, forall i,j. (ii) Skew-Hermitian matrix is neccessarly square matrix (iti) In a skew-Hermitian matrix, the elements on the principal diagonal must be purely imaginary umber or zero. 93 For, Let A= [agh-n be a Skew-Hermitian matrix, then 4, for all i. a,¥i a= Substituting i= /, we get dy = orequivalently a, +a, =O ‘or equivalently Real part ofa, (iv) IA is Skew-Hermitian matrix, then () is also Skew-Hermitian matrix for any real scalar For, (ety? (since k is real) and a ck (v) IfA and Bare! ‘Skew-Hermitian Matrices ofsame order, then f, 4 + ,B are also Skew-Hermitian inatrices of same order for any scalars &, and ky For, (Atk BY using "4 ~B we get (ky A BB YP =k) + CB) =4k,A + kB) (vi) IA and B are Hermitian matrix of same order, then AB BA is Skew-Hermitian matrix. For, (4B BA)? = (4B)? - (BA? = BAY — AP using A? = A, B°= B, we get (AB - BA)’ = Bd ~ AB = (AB ~ BA). (vii) If A is any square matrix, then 4 ~ 4° is Skew-Hermitian matrix. For, (4a) = 48 — (499 using A = (4°)®, we get ° (4-49) = AoA = {4 - A°) (viii) If A is Hermitian matrix, then id is Skew- Hermitian matrix. For (iA)? = -i49 using A° = A, we get (ia)? = -id. (i) IF A is Skew-Hermitian matrix, then (iA) is Hermitian matrix, For, (ia)® = -i4® using A° (iA)! = -i(CA) = id. (x) If is a Skew-Hermitian matrix, then A isalso Skew-Hermitian matri. A, we get(ay = (AM using AT= A, we gel (a= A= > every square matrix canbe uniquely expressed aye sum of a Hermitian matrix and a Skew- 0 Hmitan mati. The Henmitan pats 7) and Skew-Hermitian partis a 6 AtAe ANA + 2 2 Vemian shew Heian Thus, 4 Px Pan For example, Let +i 1 243i 4 i : 3 3 1-1 =i 14i 342) i 5 i 44 2 S+2i 3421 0 A 2 Ii] ft 2438 4-7 3-2i|-i-i i 0 vi} [lei gear i oa 0 1 Therefore, [2-3 4ti Idempotent Matrix ‘A square matrix d is said to be ee mattix it fe Forexample, identity ee isan idempotent may Example 5: Show that the matrix 22 4 A=|-1 3 4 1-2 -3 is idempotent matrix. Solution: We need to show that the matrix 224 1034 12 ~ is idempotent matrix. Now, a= Therefore, the matrix 4 is an idempotent matrix. Properties of Idempotent Matrix (i) An Idempotent matrix is necessarily be squat matrix. (ii) IFA and B are idempotent matrices of same order, then A + B is idempotent matrix iff AB +BA = 0, where 0 is zero matrix. For, (A+ BY =A? + AB + BA + BPMarnie B=B +B. (ii) IPA and B onder, then AB i A For, using. (A+ and AB = BA=0, we gat are idempotent I matrices of same idempotent matrix iff . BA, (ABY = (4B) (AB) = A = ABA using AB = BA, we get 2 (AB) =A(AB)B = (4A) (BB) = 2B? using 4 and B? = B, we get _ BY = 4B, (iv) 1-4 is an Idempotent matrix and A += /, then B is idempotent and 4B = i the zero matrix, eos B= A, we get | AB=0 and BA= (I-A) A=A-a? A= A, we get using using BA=0. (v) If AB = A and BA = B, then A and B are idempotent matrices. For, AB=A implies BAB = BA using BA = B, we get BAB=B or equivalently AB=I which implies A"! = B. Therefore, AB= A implies44! = A or A=I Thus A= B= / and hence idempotent (vi) diag (1,1, 1,..-1] is an idempotent matrix. (vil) If Ay, Ag, Ay are direction cosines, then the matrix AE May ade diy Mo Iady Dyds Aydy is an idempotent matrix. (ix) Involuntary Mai ‘A square matrix A is said to be an involuntary matrix if #=l. For example, identity matrix isan involuntary matrix 95 Example 6: Show that the matrix: 433 A=|-1 0 -I 44-3 is an hwvolumary matrix. Solution: We need to show that the matrix 4 3) A=|-1 0-1 is an Involuntary matrix. Now 43 374 3 3 At=|-1 0 -1i}-1 0 -1 4-4 3||4 4 3 Therefore, 4 is an involuntary matrix. Properties of Involuntary Matrix (i diag (1, 1, 1, ..- 1] is an Involuntary matrix. For fdiag (1,1, 1y...5 1 = diag(l Is eo = 2 (ii) A square matraix A is Involuntary if and only if A-DA+D=0. For, A? = implies A? - or equivalently (4—J) (4+) = 0. (iii) A isnxn Involuntary matrix. Then, Trace (A) € (1, (11), +10) 1, 25 oy de =0 (x) Nilpotent Matrix ‘A square matrix Ais said to be nilpotent of index k if At = 0, where k is positive integer. Example 7: Show that the matrix 7 A=}5 2,6 op is nilpotent matrix of index 3. Solution: We need to show that the matrix is a nilpotent matrix of index 3. 11 3yfa Now, #=| 5 2 6/5 -2 -1 -3)[2 B33 0 3ha, 96 7 ani Pad oo offi a ooo 3 1 offs 2 of-]o 0 0 1-1 -3}f-2 000 Therefore, the matrix 4 is nilpotent of index 3. RANK OF A MATRIX The most useful application o: finding the rank of a matrix is the computation of the number of solutions of a system of linear equations A positive number r is said to be the rank of a given matrix 4 of order m x nif it satisfies the follo conditions. (:) The determinant of order higher than r should be zero ii) There is atleast one square submatrix of order r whose determinant is not equal to zero. The rank of matrix A is denoted by r(4) or p(4)- From the definition of the rank of matrix 4 of order m * n we conclude that (i) If the matrix 4 does not possess any minor of order (r + 1), then p(4) r. For example. The rank of the matrix is 2 because det(A) = =Oand i 2 14 1-1 Properties of Rank (i) If every (r + 1)" order minor of the matrix is zero, then any higher order minor will also be zero and rank of the martix is less than or equal tor+l. Gi) The rank of null martrix is defined as 0. iii) The rank of unit matrix J, of order n is n. (iv) The rank of any m * n non-singular matrix is n (») The rank of any n * m singular matrix is less than n. (vi) The rank of a matrix is same as the number of Jinearly independent row vectors in the matrix or number of linearly independent columns vectors in the matrix, det 1=-240. i riodic Matrix es rig said to be periodic matrix if : ala Where n is positive integer. Ive take n= then periodic marx jg, ow, idempotent mat Nay (vil) The rank of a matrix A does not change pre multiplication or post multiplication «°Y he non singular matrix P, that is pcg) = phan p(AP) where the multiplication 4p ayy is . conformable. ate (vii IFA and B are matrices of same order, p(d + B) S p(A) + p(B) . i and B are matrices such ” ltl ication AB is conformable, thet the p(AB) < min{p(4), p(B)}, Thus p(4B) < p(A) and (4a) < on (&) Let A be any matrix, then (4) = p(47) ang i pda") = pl) (xi) Let A be any matrix, Then p(4°) = p (A) ang p(Aa®) = p(A) (xii) If A is a matrix of order m * n, then p(A) < min{m, n}. and hence p(4) < m and p(A) p(A) + p(B) —n. (xiv) The rank of a Skew-symmetric matrix can not be one. Elementary Operations Anelementary operation applied on row is called elem row operation and applied on column is called elementary column aperation. There are three types of elementary operations: (i The interchange of any two rows (columns) i th and jth of the matrix denoted by R, © R, (o C,3C). (ii) The multiplication of ith row (column) of the matrix by non-zero scalar k denoted by R, > AR, (or C, > kC). (iii) The addition of constant & multiple of the elements of ith row (column) to the corresponding elements of the jth row (column) denoted by R, > R, + kR, (or C, => C, + KC)MATROC Elementary Matrices A matrix obtained from a unit m; atrix by applying any of There are three ty ‘Pes of the clemer comesponing thre elem epeations the clementaty matrices obtained from unit eas &y applying elementary operations fro Band C, & Care same and denoted by y (if) The elementary matrices obtained from unit Pere! by applying elementary operations ih fe Rand C, > G are same and denoted by ;(#), where k is any constant. (iti) The elementary matrices obtained from unit motrin by applying elementary operations ade AR, and G > C+ KC, are same noted by E,(k), where k any scala. Properties of Elementary Matrices (i Elementary operations applied on a matrix does not change the rank of the matrix. (Gi) The matrix obtained by applying a row operation on agiven matrix is same as the matrix obtained by premultiplication of the given elementary mat For example, LeA= |> 4 Pees |80 be a matrix , then by operating Ry, we get 8 9 A= [i Al Now applying R,, on unit matrix. 0 1 | we get elementary matrix 01 [td Therefore,E;» al? alle al-B ‘| which is same as the matrix A. (iii) The matrix obtained by applying a column operation on a given matrix is same as the matrix. obtained by post multiplication of the given matrix by the corresponding elementary mattis (iv) The effect of row operation on the product of two matrices is equivalent to the effect of the same row operation applied only to the pre- factor. 7 eration product pperation on the ivatent to the effect of ied only to the (0) The effect of acolumn 0 of two matrices is equ the same column operation appli post factor. Equivalent Matrices Two matrices 4 and B are said to be equivalent if one can obtained from the other by a sequence of elementary operations denoted by 4 ~ B and read as is equivalent 0 3B. Equivalent matrices have the same order and rank Echelon Form or Triangular Form “The matrix is said to bein echelon form if (i) All non-zer0 rows of the matrix preceeding the zero rows of the matrix. (ii) The first non-zero entry in every rows is unity (iii) The number ‘of zeros preceeding the first non- zero entry in a row is less than that the number ‘of such zeros in the next row. (iv) The number of non-zero rows of the reduced echelon from isthe rank of the matrix. Normal Form or Canonical From To reduce the matrix 4 to the normal form, apply row and column operations on 4 so as to make the diagonal tlements possible after which the diagonal elements if existare all zeo and non-ciagonal elements are made zeros, (On performing elementary row and column ‘operations. Any matrix of order m % rr can be reduced t0 one of the following form I, « [ i @) [% Ostet normal form. Ol, (i), 0} Example 8: Find the rank of the matrix: 4213 6 a4 7 gro 4213 Solution: We have, A=|6 347 2101 Performing R > Ry +4 112 V4 34 as 3. 4 T 2 eae Performing R, -> Ry ~ 6R,, Ry > Ry ~ya a4 s2 52 2-12 Hence, number of non-zero rows = 2 So, rank of given matrix = 2, Example 9: Find the normal form of the matrix 2. 2-1 27 4 123 6 2-65 21 23 65 Applying row operations R, > R, -2R,, Ry = R, 3, on A, we get CHECK POINT EXERCISE - 3.1 10-4 304 5 0 6 7 aH 2 “i=l 2 is nilpotent and find its index Show that the matrix is invertible and find 4~! Show that the matrix 12 90 90 5 Applying C, © Cs, We Bet 100 4 A~|o 13 q 025 Applying R; > Rs — 2), we get 10 0 og A~10 1 3 6 4 00-1 og Applying C, -> C3-3Cy Cy C,+6C;, yg 10 000 A~|O 1 000 00-100 Applying & > (1) Ry, we get 100:0¢ A~]0 10:00 00 1:09 =U; O} Which is normal form of A, where O is the ny) matrix of order 3 x 2. Thus, Rank of 4 = 3 -[l@ 4 3. A lo {]: Then 0 ° eccos] 4, Find the rank of the matrix. A= ecccce coon raid Find the rank of the matrix = aa |99 Marra n-10 0 , owe eth 1 mati f 7 1 rank of the matrix 00 pS 66 1 then find the value ofp. ; youl 1. Find the rank of the matrix => 92 1) where the wohl yy) ate collinear Doints (vy. Je (ye Ng) and (ty 8. Find the rank of matrix, whose every element is one 9, Find the rank of the matrix of order 7, whose every element is 2020. 10, Show that every odd order skew symmetric matrix is singular. SYSTEM OF SIMULTANEOUS LINEAR EQUATIONS Letus consider the following linear system of m equations inn unknowns: yy FX TAyky ay, + ayy + us + 5,8, = by Gy¥ where x, in dnt ’m Ix ‘The matrix is called the coefficient matrix of the system. Solution of System of Equations ‘A set of values of the variables x, "3, ...3, Which satisfy all the linear equations simultaneously is called a solution of the system of equations, For example: x=~3,y=2 is solution of the system of inear equations x+2y=1 and 2x+3y=0 +2Q)=1 and 2x (-3)+3*2 Consistent and Inconsistent System The system of linear equations having at least one solution is called consistent otherwise it is said to be inconsistent. For example: The system of equations 2x+3y=5 and 4r+6y=10 y= U3 are its is consistent, since solutions. But the system of linear equations Qe+3y=5 and 4x+6y=9 is inconsistent, since it has no solutions. Homogeneous and Non-homogeneous System of Equations A system of equations 4x = b is called, homogeneous if b = 0, othenvise itis called non-homogeneous system of equations. For example: The system of equations extmy=0 mtey=0 isa homogeneous system of equations, While the system Wty=3 and +5 is a non-homogeneous system of equations. Solutions of Non-homogeneous System of Linear Equations In this article we shall discuss the method of solving non- homogenous system of equation. There are two category of non-homogencous system of equations (i) The non-homogeneous system of equation which the number of equations and the number of variables are same. (ii) The non-homogeneous system of equations in which the number of equations and the number of variables are different. ‘Now, there are two method for solving the system of non-homogeneous equation having, number of equations and variable same. (a) Cramer's rule (determinate method), (b) Matrix Method (a) Cramer's Rule: The solution of the system of linear equations100 : de is piven by Ms ye ye Band 2 = A A A a, a: where, A=/a:, a: 4)/. 4,7 Ja, hay] fas) by ay| and d= [ay by as Provided a0. Frample 10: Solve the following system of lin equations using Cramer rule Sx—7yte-I 6x - By ax + 2y-62=7 Solution: We are given that Se-Ty+ year 6x-8) 3x + 2y- 62-7 sal Therefore, 6 -8 -l 32 4 = $(48 + 2) + 1(-36 +3) + 1(12 +24) = 55 #0 n7o1 A,=|15 -8 -1 724 = 11(48 + 2) + 11-90 + 7) + 1G0 + 56) = 55 sul 6 15 =I 37-6 =5(-90 + 7) — 11(-36 + 3) + 1(42 - 45) = -55 5-7 and A,=|6 -8 15 3027 =5(-S6 ~ 30) + 1(42 ~ 45) + 11(12 + 24) = -55 Therefor, by Cramer's rule ruse = bar =ab and == 1 isthe soy, given system ‘of equations. J pe (p) Matrix Method: Suppo thesssem oy sions in n unknowns is given by ~ equal sisi ‘There are two cases: When matrix A is non-singula, Case I: Se , exists therefore, by Pre multiplication of 4-1,°" "en 4, AMAN) = ANB implies imp! (41A)X = 4°'b (Associative law) or equivalently 'b (since 4-14 ~ y ‘Therefore, the system of linear equations ats, ion, given b} * has solution, given by yoy Next, let X; and X, be tWo solutions ofthe sy, of linear equations AX=b Then AX,=b and AX,=b Which implies AX, = 4%, orequivalently 4" (AX,) = 4-(AX,) Using Associative law, we get (414) X, = (414) X, Using A"!A = I, we getX, = xX Hence, the given system of linear equations 4Y=5 has unique solutions. For example; the system of equations me tey=m extmy=e has unique solution, because| 4 |= Case I: Consider the system of linear equations Qety=3 4x + 2y=5 21 lAl= | a3 and it has no solution, Thus if 4 is singular matrix, then the non-homogeneous system of equations given by AX= b may be consistent with infinitely many solutions oF i may be inconsist also. Then 4-4=0Marna a et the following creterion of consistency. (IF |A | # 0, then the system ae system is consistent has aunque solution given by X=A"b, (ii) If|A |= 0 and (adj 4)b = 0, consistent and has infinitely many volun a (iii) If] A |= O and (adj) b a iA) b #0, then the system is Rank Method This is one of the method for solvin i solving a non-homogene system of equations havi auations syaem of uations having different number of equations Let us consider m linear equations in n unknowns Ay Faye to tay, = by May + apy to + Oak = By Opty + Ugh + 2+ Oy, which can be written asAX=b where CT a m1 nn Soon dae a 4 % x sai 3-|"? Fn dnxt Pn It ‘The matrix A is called the coefficient matrix and the matrix ay «by bq 2, iscalled the augmented matrix ofthe given system of linear equations. For Example, The system of equations dxty-z=2 xt2y-22= 1 x-yt3z=3 has augmented matrix 101 aint? (aaj=]t 2-2! i | 3:3 “Theorem 1: The system of simultaneous finear equations and only ifthe rank ofthe augmented AX=bis consistent if tmatrix [As b] is equal tothe rank of the coefficient mat. A. ‘Theorem 2: Let AX b bea system of m linear equa inn variables. Case I: When m > n. Then (iy IF Rank(A) = Rank [4 :b] = n then system of linear equations has a unique solution. (ii) If Rank (4) = Rank (A : 5] Ry~3R, and R, -> Ry 2Ry, 1-3 -8 -10 ~}0 10 20 30 oN 2 33.wing R, — F. Applying Ry > Fs, 1-3 -8 0 0 10 20 w 000: 0 Hence, Rank (4) = Rankld: 6] =2 which is less than the number of variables. Therefore, the given system of equations is consistent with infinitely many solutions Now from the last matrix of the echelon form, we get following system of equations: -10 30 J, where & is any real number, we get y=3-2k and x=2k-1. Hence x = 2k 1, y= 3 ~ 2k and 2 = k, where kis any real number. isthe required solution. Example 12: For Whar values of aand B. the following simuliancous equations have an infinite number of Putting : solution Solution: Putting the system of simultaneous equations in the form N= B tie Operating Ry + R,~ Ry and Ry 4 Ry - Ry {t | $s we get 02 2 4 : [0 1 a-1 i Bs Operating R, >, +2 i 5 or 2 Ob alt ps Operating Ry >; on a. Man, 1 Se 0 | X J 1 5 9005 2] Now, for infinite solution the las roy A | Therefore. Wh 1 ~2=0 implies o ~ " and B-7=0 implies p- 7 Solution of A Homogeneous s,., tm Linear Equations aoe . Here we shal seuss tHE MAX Method ang method ; hod, Le us consider a homogeneoy Matrix Method. Let us cons sf near equations in variables as vq, AX=0 suppose |A | #0 that s mat is on then dens. Therefore, by Premuipticajy we get ANAK) = 410 using Associative property, we get wana (Since 4-95 ATA=1, we get \ X=0 Thus, if the matrix A is non-singular, ty homogeneous system of linear ‘equations has a ige solution given by X= 0 that x, 70M solution is known as trivial solution. Thus, in ode iy the system AX'= 0 has non-trivial solution itis neesay that | |= 0. : ‘Theorem: 3 A homogeneous system AX'= Ooty linear equations in n variables has non-trivial soluioait the coefficient matrix 4 is singular that is | 4 |= 0. Example 13: For what value of the following sytem equations has non-trivial solution? ty +22=0 xtyt3z=0 det yt ba=0 Solution: Since the system of homogeneous equation is non-trivial solution, Hence, implies implies implies 26-9) - Wb - 3 (3-4)=0 = b-8=0-6=8Marne 103 Example 1: Show shat heh ta the homogeneous system of i 'm of linear ‘equations KB 42=O,ety 3x t hy- Sz =0 ‘ : hhas non-trivial solution. Also find the soluai solution, Solution: We are given that ¥-+z2=Oxtyoz= - nytt which can be written as AX=0 12 where A=i1 1 an 6 Now 1A +2(-5 +3) + (6-3)=0 Thus the given homo n geneous system of linear equations has a non-trivial solution. Now putting z= k in the first two equations, we get x-2y= which can be written as bet tot Now, |A[=1+2=340 Hence, A" exists and kxty=k or equivalently where, A= Now x esl IE] [x | implies 3 and y = 24/3. Hence, x=h3,y= 23 and 2=k Hence x = /3, y= 2/3 and z = k where k is any real number satisfy the given system of equations. Rank Method: We know that in a homogeneous system of linear equations. The rank of the augmented matrix is always same as that of the coefficient matrix. Therefore, a homogeneous linear system of equations is always consistent. y x Which implies [ | x Tp =n (the number of Case 1: If Rank (A) has an infinite variables), then the system of equations ‘number of solutions. Case I: If Rank (A) =r =n (the number of variables), then the system of equations has unique solutions X= 0 that is x, = x)=... =, = 0 Example 15: Using the rank method, solve the following system of equations x+dy-2=0 De-y+4e=0 x-Hy+ Iz Solution: We are given that xt3y-2 2e-ytdz xo y+ 142 which can be written asAX = Now, Applying Ry > R; -2Ry, we get 132 ~|o -7 8 0 0 0 Hence, Rank (4)=2, which is less than the number of variables. Therefore, the given system of equation has infinite number of solutions. Now number of variables ~ Rank (4) = 3-2 = 1. Variable can be assigned arbitrary values. From, the least matrix of echelon form, the given system of linear equations can be written as x+3y-22=0 “Ty + 82=0 Putting z = &, where k is any real number, we get 8k =10k ode fe 7 au y= Sand e=k, where k is any real number is a general solution of the given homogeneous system of linear equations. Hence, xoo *—TiGEN VALUES AND EINGE| te a square matrix of order 1 over the orc), Then Ae Fis called an cigen non zero vector VE vad Ie field P= (OOF : fue of the matrix 4 if there exist a > such that AY= AY The vector V'is called eigen vector corresponding to eigen value A Now, AX = N40. orequivalently (4 -MX=0, The matrix, 4-N ay ay Mn ay yh An Gq) M2 is known as characteristic matrix and f (R) = | 4 ~ Mis Known as characteristic polynomial ofthe matrix 4 in This is monic polynomial. Now (A-N)X=0,X#0. Therefore, for non-zero vector X, A-M=0 orequivalently | 4-2|=0 The equation f(A) =|4-A|=0 is known as characteristic equation of the matrix 4. Thus the roots of the characteristic equation say A,, Ay are eigen values of matrix A of order n. Eigen values are also called eigen roots or latent roots or characteristic roots or characteristic value of the matrix A Spectrum of matrix A: The set of all eigen values of the matrix 4 is known as spectrum of the matrix 4. ‘The largest eigen value of the matrix Ais known as spectral radius of the matrix A Theorem 4: (Cayley-Hamilton Theorem) Every square matrix satisfies its own characteristic equation. or For a square matrix 4 of order n, if [AM HCIOM + aye ah 2+. +a) be the characteristic polynomial, then the equation Wr ayXt- 4a," 2+... $a] =0is satisfied by y=A je Anta An +. aA ad=0 tamer N VECTORS am implies A" liek tad : Example 16: Verif? Cayley-Hamilion theorem for 2-47 A=|-l 2 -1 Py? +4, jon: The characteristic equation of Ais|A~ Nisy a-aA -l -1 2-2 1 -l Operating R, > R, + Ry we 8 I-A 1-2 0 -1 2-2 a 1 =] 2-% Operating C, — Cp — Cy» we get 1-k 0 0 -1 3-2 -1 |=0 1-2 2-4 On expanding through first row, we get (1-MIG-N@-%)-2=0 implies-? + 647-94 +4 =0 impliesA3 - 62? + 91-4=0 a) This is characteristic equation, Now, for Cayley-Hamilton theorem, matrix 4 must satisfy Eq. (7). So, on putting 4 = A we get, AS 64 +94 -41=0 Solutic 1 |= 2-2 et implies 0 eli) Now, fee ae -1 2 -1fl-1 2 -1 toa ai- 2 6+ 5 =|-5 6 -5 5-5 6105 On putting the values of Eg. (i), we get A642 + 94-4) 2 6-5 5 2-6-5 6 = 22. 5-5 6 1 loo -I|-4]0 1 0 2] [oor Hence, the Cayley-Hamilton theorem is verified, Now, we have to find 41. So, on multiplying b A! 1t0 both sides of Eq. (i) eee 1 and Jin RHS of BA ae 14944! - gis =0 = -64 + 9f-441=0 = A -64 +91 5 2-11 = -s|-6)-1 2 -1 6 1-1 2 100 +910 1 0 001 31-1 4ft=] 13 1 11 3 31-1 Hence, ates} 1301 “11 3 Example 17: Let A be a square matrix of order 2 with entries integer number and | A | = 1. If eigen values of matrix A are rational number. Then show that trace of Ais #2 Solution: Let A, and A, be eigen values of matrix 4. Then dydy = P | ‘which implies ada t or hy therefore JV teawhen Ay = 2 Trace A) =}, when y = -G Hence, Trace (A) = #2 Characteristic of Eigen Values and Eigen Vectors (i) Any square ma the same eigen values. (ii) The sum of eigen values of any square matrix is equal (o the trace of the matrix A. (iil) The product of the eigen values of the matrix A is equal to the determinant of the matrix A. iv) If dy, Agy one Ay are n-eigen values of square matrix A, then for any scalar & (a) The eigen values of A + kI are hy + k, Ay thd, ke (6) The eigen values of kal are Ky, Ky «os Kye (c) If A is non-singular, then eigen values of 4 and its transpose AT have 1 ae 7 (@) The eigen values of A" are RY, 2", misa positive integer. (») The set of eigen vectors x,, x3, corresponding to distinct eigen values cus hy of any square matrix is linearly independent. (vi) If Ais an eigen value of a square matrix ‘Then the eigen vector corresponding to eigen value 2 is unique upto constant multiplication. ‘Theorem 5: Show that 0 is a characteristic root of matrix A ifand only ifA is singular. Solution: Let 0 be the characteristic equation of matrix A, Then 2, where x, |a-A\=0 implies | |=0. Hence a is singular Conversely let is singular matrix, then. |a-MI= [4|+@ Diti=o orequivalently = (-2)- = or ae implies106 Theorem 6: 1 is a chanacteristic noo of a matrix A ifand only if then exist a non-zero vector N such that AX= AY Solntion: ct 4 he any square matrix of onder 7 and Ais a characteristic roots of matrix 4 Then 4-AT|=0 ‘or equivalently 4 — A/is singular. Thus, there exists a non- zero vector such that (4-ADV=0 ‘oF equivalently AN =2V Conversely. let there exists a non-zero vector X such that AN = AN. or equivalently there exists a non-zero solution of the equation (A-N)X=0 Thus \4-AI]=0 Hence 2. is an eigen value of the matrix A. Theorem 7: // A and B be two square invertible matrices, then AB and BA have the same characteristic root. Solution: We need to show that AB and BA have same characteristic roots, where A and B are invertrible matrices. We know that AB =IAB = B'B(AB) = B (BA) B But the matrix BA and B-'(BA)B are similar matrices and hence BA and B'' (BA) B have same eigen values. Thus, AB and BA have the same eigen values. Theorem 8: [fA and P are square matrices of same order wid P is invertible. Then show that the matrices A and P~ ! AP have the same characteristic roots. Solution: Let B = P'! AP, then we need to show that A and B have same characteristic root. Now B-2= P\AP- MW = Pap - PMP (Since p-! Ap = 1). Then | B-A|=| PAP ~ PIMP | PY(A-M)P| =|P'|A-NIIPL IA — All PPL A-2\|P'P|=|A-M| ‘Therefore, the matrices A and B have same characteristic equation and hence same eigen values. ‘Thus, A and ?”' AP have same eigen values. Theorem 9: JfA and B be n-rowed square matrices and A js invertible then show that the matrices ‘4-1B and BA! have the some characteristic roots Solution: We need to show that the matrices AB and BA! have the same characteristic roots, where A is invertible. We know that A'B=A'BA'A)= A'(BA")A Mey characteristic F00t. Therefore, the matrices BA" ang 4, characteristic roots. ay fe 18: Find the eigen vectors oF the m t Examp = 122 A=1027 +35 122 Solution: We have, A=] 0 2 1 “122 Now, the characteristic equation is | 4 _ ~Any a 2 re 0 2-a | implies? - 54? + 84-4 =0 implies(A —1) (A 2)? = 0 implies 2=1,2,2. Hence, the eigen values are 1, 2, 2. Now, for eigen vectors [4-any=0 Putting A= 1 in Eq. (i), we get implies 2x, +2x,=0 x +x,=0 implies (i) =, lil) implies So, from Eq, (ii), we get xy -2k+x=0 implies Hence, Now, putting 4 = 2 in Eq. (i), we getMarrs “12 is] yo 00 ty} =to -12 ofl} fo Operating Ry © Ry, we get -1 2 Wf -1.2 olla] 00 ly} fo Operating Ry —>R, ~ Rand we get RR, 1 1-2 25) To 0 0 ~21Hx,}=l0 oo ila} [o Operating R, > — 1 2 12 2)fa) fo 0 0 ifz}=lo O20 ai|(| 10 Now, operating R, + Ry ~ R,, we get bee impliese, -2x, 2, -0 wv) x,=0 (0) Let a =k Then from Eq.. (iv) and (v), we get x -2k=0 implies x) =2k x] fe] f2 2 Hence, X, op = iA] or X= [ »} lo} lo 0 107 Since, 2 has two same values 2, 2. Hence, X, is also equal to 1 1.4% = -1 Example 19: Let A be a square matrix of order 3 having eigen values 1, -I and 0. Then find the determinant of 1 + am Solution: Since eigen value of matrix A are 1, -1 and 0. Therefore, the eigen values of matrix A! are 1!, (1)! and 0°" that is 1, 1 and 0. Hence, the eigen values of the matrix /+ 4"? are | +1, 14 and 1+ 0 that is 2, 2, 1. ‘Thus Hence, x 22x1 Spectrum of Eigen Values of Matrices Unginary axis Real symmetric voix and Hermitian matrix ALGEBRAIC MULTIPLICITY AND GEOMETRIC MULTIPLICITY Let A = [aj], . , be a square matrix of order whose characteristic equation is chare (x) = 0 Ifx = Ais a root of ch(x) = 0 such that ch(A) = ch’(A) ch’! (A)=0 But ch’ (A) #0 Then, ris called algebraic multiplicity of A In other words, the multiplicity of an eigen value is called algebraic multiplicity of that eigen value i... if ch (x) = (AY Ox) Then, r is algebraic multiplicity of eigen value 2. Now, construct thematrices B= 4~2, (distinct), where Ay Ray oo By are eigen values of matrix 4. Thus, we get system of homogeneous equations, BX Now, y lok solution set of B= 0 = (XK: B,X=0} W, is called eigen space corresponding to eigen value 2, Moreover, 1, is null space of B, Thus, dim B, = Nullity of B, — Rank B, =m, This m, is called the geometric multiplicity of 2,108 Which is equal to the number of linearly independent eigen Vectors of matrix 4 corresponding to A,. Thus geometric multiplicity of, = Number of linearly independent eigen vector of 4 corressponding to A, =n PB) =n ps2) Clearly, Algebraic Multiplicity (4A of A, is greater than or equal to the geometric multiplicity (GM) of 2, Example 20: For the matriv 1 2050 2015 A=|0 -1 2020 00 1 find the algebraic multiplicity and geometric multiplicity of the cigen values Solution: The characteristic equation of matrix A is ch @x) = (= 1 & +1) MINIMAL POLYNOMIAL The smallest degree monic polynomial is called minimal Polynomial for matrix 4, denoted by m,(x) if mA) =0. Properties of Minimal Polynomial (i) For every matrix, minimal polynomial is unique. (ii) For any matrix A, minimal polynomial divides the characteristic polynomial of A i.e., deg m,(x) 2 deg ch(x). (ii) The deg m, (x) = 1 iff 4 is scalar matrix. Thus, if 4 is not scalar. Then deg m, (22 (iv) Ifx— 2 divides m (x) or m,(2) = 0, then 2 is an eigen value of A i.c., every root of m,(x) is an eigen value of A which does not give algebraic multiplicity of eigen value. (v) The order of 2 as zero of m,(x) may not be algebraic multiplicity of eigen value. _ is ix of order 7, (A) If A = [ayy «9 i8.2 Squate matrix of order 1, then ch(x) divides [m ,(x)]" ie., every irreducible factor of ch (x) is also a factor of m,(x) it -braic Hence, all the eigen values of A without alge multiplicity ean be root of m4(x) = 0. Hence, if ch (x) = Fe)" Ey Then, n(x) = TI, M Therefore the eigen valves ofthe gi ~ Mey 1, 1, and -1 Mati Hence, * AM of (2 = 1) is 2 and AM of (L=-1)is | Now, Constructing 0 B,=4-1=|9 be 2015 0 6 “3 0 2 0 A+1=]0 2019 d 0 an f 0 my 2, p(B) =3 )is 3-2 ui I)is3-3=9 ‘Thus, the number of eigen vector corres i i »POndir = 1g | while comesponding £0 A=—1 is zero, "8" Example 21: Find the minimal polynomial for the may aie see Pldlad Pilla d roxio Solution: We are given that 11 1 11 ri lol H10x10 Therefore, 1010 10 10 10 ... 10 ee =10A 10 10 10 10). or equivalently? — 104 = 0 Hence, m,(4)=0 Therefore, the minimal polynomial is F-10r=0,Maren a SIMILAR MATRICES Matrices and Bare said to be similar if there singular matrix 7? such that here exists non. trace and same determinant si ices, then t= Ppp (c) fA and B are simil t or A is said (0 be similar to 2 iF there exi mgs) = mf) mts Ph Ee fetalnesle Hence, p(l) = p(@). not conversely = ample, the matrices 1= P' BP, denoted by A For example, the matrice Properties of Similar Matrices 0100 i a i: 5 a) If AS Bic we have 0000 = ( have a non-singular matrix p A-lo 9 01, MMF jo oo} such that " 0000. 0000 aaplse eine «i p(B) = 1, while Then, and B may be singular are not similar as p() (6) If and B are similar matrices, then xt = chy (8) chy) =eh g(x)” (nye = m0) a 7 ‘Trace (A) = Trace (B) A ave same eigen value, same det (1) = dot (B) igen value of B. Eigen value of A = DIAGONALIZATION OF MATRICES ‘Any square matrix is said to be diagonalizable i lap =D similarto a diagonal matrix that is there e : fl tmatrix P such that exists an invertible ‘Where D is a diagonal matrix. — similar matrices have the same eigen values, therefore, the diagonal elements of D are the eigen values of rem 10: If a square me Py Dae specie onde eraear creat st teH pe iaes vectors, then a matrix P can be found such that P-'AP is abways a diagonal matrix, where ax, +1 +, =X 7 Pe [Xp Ny Ny] and X, are eigen vectors any +byy, +E = hip CD Proof: We prove this theorem for a square matrix of order ann + Oy + 6s 3 x3 and the proof can be easily extended to matrices of and similarly for A, and Ay, we obtain ax, +h y2 +4 higher orders. ayx; +0,) 6 Gi) a BG ayy +52 + 2 Let a, by &| bea3x3matrix. ...() a, by ey aay + Ys +23 and ai + Py +e (iy) Let Ay, Ay and A, are its eigen values and x, X, and yxy +Dyby + 0 Ay the corresponding eigen vectors of matrix A defined Now, let us consider the matrix P, such that its as follows columns are the eigen vectors of A. x x ay 4 x5 yf Xe=]oo] and X=] bes ” % a 2 a 4 % @ & alfa ms Then, aP=la, by Gl] y2 9s a & olla 2 For the eigen value A, the eigen vector is (a - Ax, +5 Fer at, + (by hy), +ayy FAM TE elas tht? Xs + yy +425 yX5 + by) + 652 aX + Ys + 652, ay tat dass = [Ai Aayr ays Lhe hae Ages [Using Eqs. (ii) (iti) and (iv)] Making use of Eqs. (ii), (iti) and (iv); we have ye lf 0 0 yyy ysffO & Of =PD 4% all0 0 % Remark: The square matrix P, which diagonaliss 4, is obtained by then resulting diagonal matrix has the eigen value of A as its diagonal elem Properties of Diagonal Matrices (i) A matrix is diagonalizable iff sum of geometric multiplicity is equal to the order of 4. (ii) A matrix is diagonalizable iff for every eigen value geometric multiplicity is equal to algebraic multiplicity. (iii) If all the eigen values of a matrix are distinct, then matrix can be diagonalizable not conversely. (iv) A matrix A can be diagonalizable iffits minimal polynomial m, (x) does not contain repeated linear factor ie., minimal polynomial is product of distinct linear terms ie, m,(x) = PA) ae F (») If all the entries of A are equal, then 4 is diagonalizable. For A= [ay x 94 = OV I] ch (x) =x"! (x= Mic) = m4(e) =x") (vi) Every idempotent matrix can be diagonalizable. (vii) Every involutory matrix can be diagonalizable. matrix is diagonalizable. (viii) Every real symmetric be diagonalizable. (ix) Every Hermition matrix can (x) Nilpotent matrix ean never unless it is zero matrix. be diagonalizable where, 0 onal mati fort 3*3 implies AP=PD preemutiplying by Pe get PIAP=P'PD=ID= p 1A issimiar 0 D. Therefore, hg agen vector Pree MTS A 10s diagonal Post multiplying Eq- (0) by P, we get fn ss resutean de gored (07% ord ° sete mts Piscaled the modEI MIA ang ved the spectral mati of : is a diag which implies tha grouping the igen vectors of 4 into square mari ang ents. Matrix by the Method of Diagy, Power of 2 nalization ; sa bea square matrix of order” *7” TREN We can such that a non-singular maui P D=PAP DP=(P'AP) (PAP) implies =(P'APP-AP) D=PIAP Cs PPIs 1) D=P'AP Pap al and Pre-multply Eq, () by P and post-multiply by P', wwe have PD! P= PP A'PP*! = A”, which gives A” Example 22: Reduce the matrix A into a diagonal matri where 86 2 A=|-6 7 4]. 2-4 3 Solution: Froseding a above its eigen values ar ted to be 0, 3,15. Also the corresponding vector ae 7 2 Vand x, =| ~ 4 af?