Hioven Enonernna Manenatcs is equivalent to Ay + Bhy + Bg =0, BA, ~ BAy—2y =O, 4h, + hy + Bhy = 0, 2hy + Wy + By = 0 As these are satisfied by the values A, = 1, Ay = 1, Ay =—2 which are not zero, the given vectors are linearly dependent. Also we have the relation, x, +x,-2x,=0 by means of which any of the given vectors can be expressed as a linear combination of the others. ‘Obs. Applying elementary row operations to the vectors x. Xp. Xy, we see that the matrices 1 x; A=|xg| and B= Xs Xs X1 + Xp —2xy have the same rank. The rank of B being 2, the rank of A is also 2. Moreover x,, x, are linearly independent and x, can be expressed as @ linear combination of x, and x, [+ xy =}(x, +x,)]. Similar results will hold for column operations and for any matrix. In general, we have the following results : Ifa given matrix has r linearly independent vectors (rows or columns) and the remaining vectors are linear combina- tions of these r vectors, then rank of the matrix is r. Conversely, if a matrix is of rank r, it contains r linearly independent tuectors are remaining vectors (if any) can be expressed as a linear combination of these vectors. ees 1. Represent each of the transformations y= 8y, + Bandy =H, + ey and aye - 7,4 Mn Vy = 3, by the use of matrices and find the composite transformation which express x,, xin terms of, 2, WE =xcos ay sin, n= 1 sin a +y cos a, write the matrix A of transformation and prove that A the inverse transformation. ‘B. Atransformation from the variables: ,;,%, 10,)42»¥5i8 given by Y =X, and another transformation fFom Ys y Yq t0 2.2.25 is given by Z = BY, where (Hence write 21 @ did a-[ o4 4]. 2 4] tin eration in sy 12 74 13 5 4, Find the inverse transformation of y, =%, + 2¢q°+ 8x 531g = 2x + 4g + Uy jy = —xp-* Bap 5, Verify that the following matrix is orthogonal = v3) 2a 28 ‘e080 0 sin (9/2/3113 ~ 2/3) (Hissar, 20058; PU, 2003) i) | 0 1 0 (Kurukshetra. 2005) 23-213 WB sin® 0 c058 i 0 | 6 Find the values ofa,6,cifA=|a 6 ~c| is orthogonal ? (Mumbai, 2005 S) a-b « Im on 0 7. Prove that’ ° 9 8) % is orthogonal when I = 2/7, m = 8/7,n = 6/7! -mn =1 0 8, If and B are orthogonal matrices, prove that AB is also orthogonal, (Anno, 2005) 9, Aro the following vectors linearly dependent. If so, find the relation between them : G@)(2, 1, , (2, 0,= 0, (4,2). (Mumbai, 2009) Gi) A, 1, 1,9),0,2, 8, 4), (2,3, 4,9) +2, 4), = (2,= 1,3), xq = ©, 1, 2. x= 8,7, 2). (G.P-T.U., 2003 ; Nogpur, 2001) (1) EIGEN VALUES IfA is any square matrix of order n, we can form the matrix A — J, where / is the nth order unit matrix. ‘The determinant of this matrix equated to zero,is called the characteristic equation of A. On expanding the determinant, the characteristic equation takes the form CAA + NE + AE HA, = 5 where #’s are expressible in terms of the elements a,,. The roots of this equation are called the eigenvalues or latent roots or characteristic roots of the matrix A. (2) Kigen vectors * My yy om My x= |*?) anda =|“ % ~ |, then the linear transformation ¥ = AX Ai) Sq ig yg om a carries the column vector X' into the column vector Y by means of the square-matrix A. In practice, it is often required to find such vectors which transform into themselves or to a scalar multiple of themselves. ‘Let X be such a vector which transforms into AX by means of the transformation (i).. ‘Then IX=AX or AX-NX=0 or [A-NIX=0 elii) ‘This matrix equation represents n homogeneous linear equations (ayy = Mix, + Gye +--+ Ain %, =O Gy1% + (zg — NXg +. + BagXy =O GgiXy + OggXy + on. (Gyn — AMy =O which will have a non-trivial solution only if the coefficient matrix is singular, i.e., if | A- AI | = 0. This is called the characteristic equation of the transformation and is same as the characteristic equation of the matrix A. It has n roots and corresponding to each root, the equation (ii) [or (iii)] will have a non-zero solution, X= [xy)Xp) 0 I’, Which is known as the eigen vector or latent vector. i Cor a 2. = ie, | Fi a’ [=o or #=Th4+6=0 or Q-6A-1)=0 2. 2261. ‘Thus the eigen values are 6 and 1. Ix, be the components of an eigen veetor corresponding to the eigen value, then woanee fh) Corresponding to = 6,wehave |-} _4]/*] <0 ponding to=6,wehave |" _4 |] 5] = which gives only one independent equation —x + 4y = 0 iving the eigen vector (4, 1)Corresponding to A = 1, we have [i ‘] (F =0 which gives only one independent equation x + y = 0. tS giving the eigen vector (1, - 1). 1--A 1 3 Solution. The characteristic equation is |A-—M|=| 1 5-A 1 |ieatmessese 3 1 1-A Since 4 = - 2 satifies it, we can write this equation as (A+ 2)(Q?-924 18)=0 or (A+2)(A-3)(A-6)=0. ‘Thus the eigen values of A are A =~ 2, 3, 6. Ifx, y z be the components of an eigen vector corresponding to the eigen value A, we have W-A 1 8] fx {A-a]X=| 1 5-4 1 |lyl=0 ol) 3 1 1-al lz Putting A =~ 2, we have Sx + y + 82 =0,x+ Ty +2=0, Br +y +82 =0. The first and third equations being the same, we have from the first two Hence the eigen vector is (— 1, 0, 1). Also every non-zero multiple of this vector is an eigen vector corresponding to i= —2. Similarly, the eigen vectors corresponding to A = 3 and A = 6 are the arbitrary non-zero multiples of the vectors (1,—1, 1) and (1, 2, 1) which are obtained from (i). Hence the three eigen vectors may be taken as (— 1, 0, 1), (1,~1, 1), (1, 2, D. Solution. The characteristic equation is 3-2 (A-¢MI=0, ie, 0 0 or (8-2) (2-9) (6-2)=0 Thus the eigen values of A are 2, 3, 5. Ifx, y,z be the components of an eigen vector corresponding to the eigen value A, we have 8-2 1 4 Wx a-wxX=| 0 2-2 6 =0 o 0 5-all2 Putting A = 2, we havex +y +42 =0,62=0,32=0,ie,x+y=0 and 2=0. Hence the eigen vector corresponding to i= 2 is &, (1,~1, 0). Putting 2=3,wehavey+4z=0,-y+62=0,2=0, ie,y=0, 2=0.Hence the eigen vector corresponding to A = 3 is k, (1, 0, 0). Similarly, the eigen vector corresponding to A = 5 is k, (3, 2, 1). Ba PROPERTIES OF EIGEN VALUES 1. Any square matrix A and its transpose A’ have the same eigen values. We have (A-MYy =A’ =A | A-My | =| Aa] JA-Ar[=|A'-N | “1B 1=1BI | AAI | = Oifand only if | A’—27 | =0 i.e., 2 is an eigen value of A if and only if it is an eigen value of A’. IL. The eigen values of a triangular matrix are just the diagonal elements of the matrix. hy he Qn yg Aan Let 2 . | be a triangular matrix of order n. 0 0 . ay Phen | A=AL | = (ayy 2) bgg = Wragg = 20. a, Roots of | A=2L | =O are A= ayy, 2y99-05 Gye Hence the eigen values of A are the diagonal elements of A, ie., 545 go» + Gye Cor. The eigen values of a diagonal matrix are just the diagonal elements of the matrix. UL. The eigen values of an idempotent matrix are either zero or unity. Let A be an idempotent matrix so that A® = A. If i be an eigen value of A, then there exists a non-zero vector X such that AX =2X A(AX)=AQX), ie, A®X = MAX) ie. AX = 200) [» A?=A and AX =2X a AX = 2X AB) From (1) and (2), we get 4X = AX or (A? 4) X= 0 or 22 —2.= 0 whence 2. = 0 or 1. Hence the result. IV. The sum of the eigen values of a matrix is the sum of the elements of the principal diagonal. This property will be proved for a matrix of order 3, but the method will be capable of easy extension to matrices of any order] Consider the square matrix a) a %2 hs Azlan a2 aa ma) My gy Os ay 2 ayy Ms so that [A-aT} =| Gq ak aay (On expanding) ay Ag Ogg == AP + AMG yy + ayy + gg) ~ Adi) If Ay» hy Ay be the eigen values of A, then | A—AI | =(- 1) (-2,) (A—A,) (Ay) BDH D2 Oy + Ay + Ig) — Equating the right hand sides of (ii) and (iii) and comparing coefficients of 22, we get Dy + hy +g = yy + gy + gy, Hence the result. V. The product of the eigen values ofa matrix A is equal to its determinant. Putting 2=0 in (iii), we get the result. VI. If his an eigen value of a matrix A, then 1/.is the eigen value of A“. IfX be the eigen vector corresponding to 2, then AX = AXPremultiplying both sides by A~!, we get A! AX = AX ie, IX=)A4X or X=MA'X), ie, AIX=()X ‘This being of the same form as (i), shows that V/A is an eigen value of the inverse matrix A“ VII. If his an eigen value of an orthogonal matrix, then 1/2. is also its eigen value. We know that if is an eigen value of a matrix A, then V/A is an eigen value of A~. (Property V1. Since A is an orthogonal matrix, A is same as its transpose A’. + Vikis an eigen value of A’. But the matrices A and A’ have the same eigen values, since the determinants | A-M | and | A’~M | are the same. Hence 1/, is also an eigen value of A VII. If hey hg soy hy are the eigen values of a matrix A, then A” has the eigen values hy, a! voy hy! (mn being a positive integer). (Mumbai, 2006) Let A, be the eigen value of A and X, the corresponding eigen vector. Then @ We have Similarly, Hence }/”' is an eigen value of A”. ‘The corresponding eigen vector is the same X,. CAYLEY-HAMILTON THEOREM* Every square matrix satisfies its own characteristic equation ; i. order square matrix A is , ifthe characteristic equation for the nth [A-aT | =Car eka li) then Can ans kan Let the adjoint of the matrix A 2 be P. Clearly, the elements of P will be polynomials of the (n ~ 1)th degree in 2, for the cofactors of the elements in | A — AJ | will be such polynomials. + Pan be split up into a number of matrices, containing terms with the same powers of A, such that P=PA-'4 PAM 24 +P, ht P, where P,, P,, ... P,, are all the square matrices of order n whose elements are functions of the elements of A. Since the product of a matrix by its adjoint = determinant of the matrix x unit matrix. [A-MIP=|A-M| xT by @ and (ii), 1A - AN IP, 2°-1 +P, 2-2 +... P,_y AEP, [IP at ky ot thy Re RE Equating the coefficients of various powers of A, we get -P,=CT AP,-P,=k,1, AP, -Py=hyl, AP. =keT. Now pre-multiplying the equations by A", A"-1, .... A, respectively and adding, we get CAVA" + RANT tk, AtK THO, for the terms on the left cancel in pairs. This proves the theorem. Cor. Another method of finding the inverse. Multiplying (it) by A+, we get Caan) kya 1 +h, gl +k,At=0 whence a [ayant kA 24 ek, Ml “See footnote on p.17. William Rowan Hamilton (1805-1865) an Irish mathematician who is known for his work in dynamics." Linean Avcc@Ra : Detenwnaers, Marnices EI ‘This result gives the inverse of A in terms of n—1 powers of A and is considered as @ practical method for the computation of the inverse of the large matrices. As a by-product of the computation, the characteristic equa- tion and the determinant of the matrix are also obtained. Example 2.48. Verify Cayley-Hamilton.theorem for the matrix A~ (3 4] and find its inverse. Also express A® 444 —7A9 + 11A® ~A ~ 10] a8 a linear polynomial in A. (Bhopal, 2009) Solution. The characteristic equation of A is ieee aed or 22-44-5=0 li) By Cayley-Hamilton theorem, A must satisfy its characteristic equation (i), so that ~4A-51=0 ii) 2_4a-sr= [1 4][1 4]_4f2 4]_gf1 0 ‘on woamoeE ab lb 3)-sf = {2 1)_[4 1]_/s 9). g - ~ [8 17)7|8 12. 0 ~ ‘This verifies the theorem. ‘Multiplying (ii) by A+, we ea -41-5A*=0 * wefan 2h Sell Shalt 4] ‘Now dividing the polynomial ie 4d — 723 + 1102 -2.— 101 by the polynomial 4? - 44. —5, we obtain ~ 4M — TAS A - 101 = (A? — 4. — 5) (A820 43) 4245, A485 [By (i) Hence A® ~ 4A‘ —7A% + 11A2— A ~ 107 =A + 5, which is a linear polynomial in A. Fie Si tif | Example 2.46, Find the characteristie equation of the matrix A =| 1 3 ~8| andhence find its | a2 4-4} “inverse. 1-a 1 3 Solution. The characteristic equationis} 1 3-2 — ), ie, M-200+8=0, 4-4-2 By Cayley-Hamilton theorem, A° ~ 20A + 81 = 0, whence A! = 57 far, 5 100 1 -4 -8 -12 3 1 3/2 =3]0 1 0)/-2)10 22 6|-|-5/4 -14 ~a/4 lef. Ex. 2.21 ool 2 2 22 -W4 -4 -/4 J F \ 247 Example 247. Find the characteristic equation of thé matrix, A=|0 1 0) and henee compute A> 112 f (W.2.U. 2010) Also find the matrix represented by A8_GA7 + 7AS- 3A5 +A‘ 5A" + 8A? 2A 41. (Anna, 2009; Raethan, 20083 U-P.T-U., 2003) Solution. The characteristic equation of the matrix A is 2-- I 2 o 1-2 0 11 2=2 0 or (8-54? +7A-3=0)Ea ‘Hianer Enoneenine Mariewanes. According to Cayley-Hamilton theorem, we have A® — 5A? + 7A — 31 =0 od) ‘Multiplying () by A“, we get $UAt— 5A +70) ii) 211)[211] [44041 24141 24042] [5 44 But A®=/0 1 0]/0 1 0 |=| 0+0+0 0+1+0 0+0+0 |=| 0 1 0 11 2][11 2] [2+0+2 14142 14044] | 4 4 5 544 211 100) [2-1-1 At-GA+7=|0 1 0|-5)0 1 0]/+7/01 0|=| 0 3 0 445 Hig oo1]|-1-1 2 1 2-1 -1 Hence from (ii),A*=2/ 9 3 0 3|-1 -1 2 Now AS— 5A? + 7AS—3A5 + At— 5A + 8A?— 24 +1 * (A? — 5A? + TA - 31) + A(AS— 5A? + 7-3 +A? +A4T =A? + Ae] [- A? -5A?+7A-31=0) 544 211 100 855 =|o 1 o|+/o 1 o]+/0 1 0 |=|0 3 0]. 445 112 ool 558 g14 1. Find the sum and product of the eigen values of | 0 2 6 (Madras, 2006) o05 2, Find the eigen values and cigen vectors of the matrices : @ i H| (WBTU., 2005) (6) it | (Bhopal, 2002 8) 8. Find the latent roots and the latent vectors of the matrices : ! 8-6 2 7 wi|-t 7 +] (Bhopal, 2008 ; Nagarjuna, 2008 ; 8.V.T.U., 2008 ; J.N.1.U,, 2006) 2-4 8 : 207 10 20 U.N.T.U,, 2008 ; Kurukshetra, 2008) 102 2 easy |.2. 1-6 (Mumbai, 2006 »B P-TU., 2006 ; U.P-T.U., 2006) 418 0) 6-2 2 24168 (d)|-2 3-1 (Kurukshetra, 2006) fe)|/ i 1-2 (Madras, 2006) 241° 3 1-2) 1 4, ICA bean eigen value of a non-singular matrix A, show that | A |/Lis aneigen value of the matrix adj A. 7; (UP-TU, 2001) 234 5, Find the eigen values of adj A and of A?- 24+ I, where A= |0 4 2: (Mumbai, 2006) — Oo 0 3 2 24 ‘i 6. Two eigen values of the matrixA =/1 3 1) are = 1 each. Find the eigen values of A+. 122 % Show that if, 2g)» Ay are the latent roots of a matrix A, then A® has the latent foots af, 2%, .. 22 (PTW, 2005)Ina ALceona : Devens, Marrices L st | 8. For a symmetrical square matrix, show that the eigen vectors corresponding to twa unequal eigen values are orthogonal, 9. Using Cayley-Hamilton theorem, find the inverse of e 40 3 @ (5 a @f2 4-1 (Osmania, 2000 8) b-1 4 a) fie alt a f Gii)]1 ~3 -3| (Bhopal, 20088) 0 -2 0 (UP.7U,, 2006) 2-4 <4 0 08 137 10. Find the characteristic equation of the matrix A =|4 2 3]. Show that the equation is satisfied: by A and hence iad obtain the inverse of the given matrix. (Bhopal, 2008; Anna, 2005 ; Kerala, 2005) 11. Verify Cayley-Hamilton theorem for the matrix A and find its inverse, 2-14 ol 2 4 (Anna, 2009 ; S.V-1.U., 2008 ; Madras, 2006) Leb 2 Hes ry (i) |-6 -1 2| (Coimbatore, 2001) 2 (PTL, 2006) 6 2-1 3 12, Using Cayley-Hamilton theorem, find A% if, [3 ri (Anna, 2003) 2-1 2 13, A=|-1 2 -1|, na at, (Madras, 2006) 1-1 2 12.0 14. Using Cayley-Hamilton theorem, find A, where A=| 2-1 0 (Bhopal, 2008) 0 0-1 46 6 A=] 1 3 2 |, evaluate A4,Atand A. =hi4 +8 100 16, fA =| 1 0.1 |, show that A" = A"? + A®— 1. Hence find AY, © Ctumbai, 2006) o10 (1) REDUCTION TO DIAGONAL FORM Ifa square matrix A of order n has n linearly independent eigen vectors, then a matrix P can be found such that P'! AP isa diagonal matrix. {This result will be proved for a square matrix of order 3 but the method will be capable of easy extension to matrices of any order] Let A be a square matrix of order 3. Let A,, 4), A, be its eigen values and % 2 ct 4 2 ‘yg | be the corresponding eigen vectors. 2 HH ty Denoting the square matrix [X,X,X,] =|, yp 94 | by P, we have Ah AP = AIX, X,X,] = (AX, AX,, AX] = JX, Xp, A Xsiti Rote Aste] [mr 22 22] [ar 0 0 {i Aod2 Aas |=|91 Je Ya |X| O Aq O | =PD, where D is the diagonal matrix. yar ate Meta] La ze 4] [0 0 dy P*! AP = P| PD = D, which proves the theorem. (2) Similarity of matrices. A square matrix A of order nis called similar to a square matix A of order n if A = PAP for some non-singular n x n matrix P. ‘This transformation of a matrix A by a non-singular matrix P to A is called a similarity transforma- g 3 If'x is an eigen vector of A, then y = P*! xis an eigen vector of A corresponding to the same eigen value. (8) Powers of a matrix. Diagonalisation of a matrix is quite useful for obtaining powers of « matrix. Let A be the square matrix. Then a non-singular matrix P can be found such that D=P1AP (D? = (PAP) (PIAP) = P-1A®P fe PPt=1 ilarly, D® = P+! A®P and in general, D® = P! A"P oD) ‘To obtain A", premultiply (i) by P and post-multiply by P-". Then PD"P- = PP’ A" PP = A" whieh gives A". uO 0 ‘Thus, A"=PD'Pt where, D"=|0 25 0 00% Working procedure : 1. Find the eigen values of the square matrix A. 2, Find the corresponding eigen vectors and write the modal matrix P. 3, Find the diagonal matrix D from D = P-1AP 4, Obtain A* from A” = PD"P*, [oo 2-2) 12-2 1/=0 or i8-22-544+5=0. <1 -1 -2 Solving, we get 4, = 1,2,= V5 ,%,=— V6 as the eigen values of A. When 4 = 1, the corresponding eigen vector is given by ~2r + 2y-2e=Oxtyt2= » 0 Solving the first two equations, we get ~y siving the eigen vector (1, 0,~ 1) When A= V5 , the corresponding eigen vector is given by (C1 - V5 )x + 2y -22 = 0,2 + (2- V5 )y +2=0,-x-y- Vz =0.Solving 2nd and 3rd equations, we get — a 6-25 -1+V5 giving the eigen vector ( /5 — 1, 1,- 1). Similarly the eigen vector corresponding to A= — V5, is (J5 +1,-1, D. Writing the three eigen vectors as the three columns, we get the transformation (modal) matrix as |: VB-1 a =|o 1a al. 1-1 1 Hence the diagonal matrix is Solution. The eigen values of A (found in Ex. 2.43) are — 2,3, 6 and the « eigen vectors are (-1, 0, 1), G,-1,),(1,2, D. Writing these eigen vectors as the three columns, the required transformation matrix (modal matrix) is 1 oad =| 0-12 11d -1 11f [qh q ‘To find P-!, [P|=| © -1 2l=|a, b c]} (say) 1 11] la, & & 3,B,=2,C,=1,A,=0,B,=—2,C,=2,A,= 8, By=2,C,=1 Also | WA, +b, By +0, Cy = ‘Thus if-t 1 1)[16 0 0 ][-3 0 3] Hence At=PD'P+=1) 9 -1 2|/0 81 0 || 2-2 2 a 1ijlo o 16], 1 21 “1 8] [251 485 236 =| 0 4 2 oof 27 |=|485 1051 485 1 216 512 216| (295 485 251Solution. The characteristic equation of A is a/2-v 2 . V2 3/2-a| 70 be, @R-AP— 4 =0. 22—84+2=0 whence A= 1,2. When d= 1,(A-MIX=0, gives (ie wall]-L2) » GF lle )-Le (2k 2 " , [o a [2]-[0] [By R,-R,) x,+x,=0.Ifx,=—1,x,=1, ie, the eigen vector is [1,—1]'. When 2=2,{A—MIX=0, gives Fu alle )-(4] i (3 -ill2}-Le) [i] . faba enn =x,#x,=0, ie, x=% x,=1, ie., the eigen vector is [1, 1] Replacing e by 4, we got 1/20 12]_flo 6 12 20|"| 6 10]° Ba REDUCTION OF QUADRATIC FORM TO CANONICAL FORM A homogeneous expression of the second degree in any number of variables is called a quadratic form. ah x For instance, if hob ‘|. y| andX’=[x y 2], then Bf ¢. £ KVAX = ax? + by? + 2? + 2fye + Qgax + Qhxy oi) which is a quadratic form.| Livean ALosibna : Derentnawrs, MaTrices Let Ay, Ayy Ay be the eigen values of the matrix A and be its corresponding eigen vectors in the normalized form (i.e., each element is divided by square root of sum of the squares of all the three elements in the eigen vector). a 0 0 he % % ‘Then by $2.16), P*AP=|0 2, 0|whereP=|¥ 92 33 0 0 dy 4% 4% Hence the quadratie form (i) is reduced to a canonical form (or sum of squares form or Principal axes form). Tyr? + Agy? + Age? and P is the matrix of transformation which is an orthogonal matrix. Note. Congruent (or orthogonal) transformation. The diagonal matrix D and the matrix A are called congruent matrices and the above method of reduction is called congruent (or orthogonal) transformation. Remember that the matrix A corresponding to the quadratic form ax? + by? + c2* + 2fye + Qgex + hry coeff. of x” pool of yz } coeff. of zx whe is J coett of yz coeff of »? Leveft.of xy|, & i 6 ‘| 2 2 efe 1 1 2 Meoott of 2x evel of xy coef of 2 poaoicedaeys ae teeth ha dit —Byz + Bex - joel 3 Rare 20) 3-1 1 Solution, The matrix of the given quadratic form is A = [- 1 5 - | 1-1 3 B-A -1 001 Its characteristic equation is |A-M|=0, ie, | -1 5-A -1 1-1 3-2 which gives A = 2, 3, 6 as its eigen values. Hence the given quadratic form reduces to the canonical form Aart Age+ag®, ke, 2a? + By? + Ge? To find the matrix of transformation From [A - A] X = 0, we obtain the equations (=A) x-y+2=0;-x4+6-Dy- Now corresponding to A = 2, we getx—y +2 =0;x-y+(3-A)z=0. x4+3y-z=0,andx-y+z2=0, shenee wens tO 1 ‘The eigen vector is X, (1, 0, 1) and its normalised form is (1/ V2 , 0, - V2). Similarly, corresponding to 2 = 3, the eigen vector is X, (1, 1, 1) and its normalised form is (1/3 , W/ V3 , vB). Finally, corresponding to 2. = 6, the eigen vector is X, (1, — 2, 1) and its normalised form is (1/ 6 ,- 2/6 , wJ6). [ wv2 wVS WE Hence the matrix of transformation is P= 0 W/V3 -2/V6|. <2 VWV3 V6ca -—Hiaien Enanserma ManiEnarics NATURE OF A QUADRATIC FORM Let Q = X" AX be a quadratic form in n variables x), 95 --%q- Index. The number of positive terms in its canonical form is called the index of the quadratic form. Signature (S) of the quadratic form is the difference of positive and negative terms in the canonical form. Ifthe rank of the matrix A is r and the signature of the quadratic form Q is 8, then the quadratic form is said to be ( positive definite ifr =n and s =n (ii) negative definite ifr =n and s = 0 (iii) positive semidefinite ifr 0. (ii) negative definite if all the eigen values of A <0. (iit) positive semidefinite if all the eigen values of A 20 and at least one eigen value (iv) negative semidefinite if all the eigen values of A $0 and at least one eigen valu (v) indefinite if some of the eigen values of A are positive and others negative. SR Rai tt adr fon Ba 3 eer aeey discuss its nature, i Va A pe ssednde As h 0 Solution. (i) The matrix of the given quadratic form is A = | 1 1 -A 1 1 Its characteristic equation is [A—2]=0, ie, 1-a- - i =0 1-1-2 which gives 4932+ 2=0 Solving, we get A= 1, 1,-2as the eigen values. Hence the given quadratic form reduces to the canonical form Aye + hy? +hg?=0, ie, 22 +y?-2e2=0 (i) Since some of the eigen values of A are positive and others are negative, the given quadratic form is Indefinite. (iii) To find the matrix of transformation From [A — XI] X= 0, we get the equations met y+2=0,x-hytz=0,x-y—he When 2=-2,weget 2x+y+z=0,x+2y—z=0,x-y+22=0. Solving first and second equations, we get ‘The corresponding eigen vector X, = (— 1, 1, 1) and its normalised form is (~ 1/V3, 1/ V3, 1/V3) Whend=1,weget —x+y+2=0,x-y-2=0,x-y—z=0. These equations are same. Take y = 0 so that x = 2. ‘The corresponding eigen vector X, = (1, 0, 1) and its normalised form is (1/ V2 , 0, 1/ V2) To find the eigen vector X,, = (I, m,n) (say) Since X, is orthogonal to X,, —l+m+n=0 Since X, is orthogonal toX,, -. l+n=0 L ‘These equations give 1 ‘The eigen vector X, = (1, 2, ~ 1) and normalised form is (1/V6, 2/V6,—_1/V6). 2Lingan Avceena : Deverwnants, Marrices Hence the modal matrix is -W3 W2 WE W302 |. W3 WV2 -VV6 recess =1/2 - 3/2 0 2 V3/2 0 |. HA=|-J8/2" 1/2 0] and P=|-Y3/2 1/2 0}, show that P* AP is a dingonal matrix. 0 00 o oL 2. Show that the linear transformation Hee fin 6]. where e= 3 tant 24 , changes the matrix «fe 5 to the diagonal form D = HCH". 3. Reduce the matrix A = ste x 1 tothe aia form. (BP.TWU, 2005) 4 was[f §] find at and at (Mumbai, 2006) a-1. t 6, WA=|-1 5 -1), calculate A‘, (Coimbatore, 2001) 1-2 8. 6. was[} {]s then prove chat 9 tan A= A tan (Stumbai, 2006) 6-2) 2) 7. Find the eigen vectors of the matrix |—2 _ 3. —1] and hence reduce 6x? + 3)? + 3z*~2yz + 42x — day to a'sum of 2-1 8 squares’. Also write the nature of the matrix (Calicut, 2005) ‘8. Reduce the quadratic form 2xy + 2yz + 2zx into canonical form. 10. (Anna, 2009 ; Kurukshetra, 2006 ; Muribai, 2003) 8 Sanh (6) Reduce the quadratic form x,? + $x,? + 8x32 —2e,x, to.a canonical form, (Anna, 2009) Reduce the following quadratic forms into a ‘sum of squares’ by an orthogonal transformation and give the matrix of transformation. Also state the nature of each of these. (8) 9e,2 4 S432 + Sry! + Beye + 2 yx4~ Dees. 1 0 Oo (i gat i ter oe ed ot 3 i} (ii) Bx? + Ty24 Be? — Ly — Bye + dx nna, 2002 8) 11. Find the index and signature of the quadratic form xf + 2x3 - 3x3. (Madras, 2006) 12, Find the nature of the quadratic form x* + 5y? + 2” + 2xy + 2yz + Gex. (Bhopal, 2099) 13. ‘Show that the form 5x,?-+ 26x,2-+ 10x, + dx, + Mdxge, + Gxt» is a positive semi-definite and find a non-zero set of values of x,,9,3 which make the form zero. COMPLEX MATRICES (P.T.U., 2003) So far, we have considered matrices whose elements were real numbers. The elements of a matrix can, however, be complex numbers also. (1) Conjugate of a matrix. If the elements of a matrix A= a,,] are complex numbers 0, + i Br. 0, and By, being real, then the matrix