
3.1 SOLUTIONS
Notes:Someexercisesinthissectionprovidepracticeincomputingdeterminants,whileothersallowthe studenttodiscoverthepropertiesofdeterminantswhichwillbestudiedinthenextsection.Determinants aredevelopedthroughthecofactorexpansion,whichisgiveninTheorem1.Exercises33–36inthis sectionprovidethefirststepintheinductiveproofofTheorem3inthenextsection.
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.167

Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
k,andthedeterminantismultipliedby k.
21 34 18202, 56 =−=− 34 53643(64)(53)42 kk kk =+−+=− ++
Therowoperationreplacesrow2withktimesrow1plusrow2,andthedeterminantisunchanged.
22., ab adbc cd =− ()() akcbkd akcdcbkdadkcdbckcdadbc cd ++ =+−+=+−−=−
Therowoperationreplacesrow1withktimesrow2plusrow1,andthedeterminantisunchanged.
23.
111 3841(4)1(2)1(7)5, 232 −−=−+−=− 384(4)(2)(7)5 232
kkk kkkk −−=−+−=−
Therowoperationscalesrow1by k,andthedeterminantismultipliedby k.
abc abcabc =−+=−+
24.322(2)(6)(3)263, 656
3(65)2(66)2(56)263
Therowoperationswapsrows1and2ofthematrix,andthesignofthedeterminantisreversed.
25.Sincethematrixistriangular,byTheorem2thedeterminantistheproductofthediagonalentries: 100 010(1)(1)(1)1 01 k ==
26.Sincethematrixistriangular,byTheorem2thedeterminantistheproductofthediagonalentries: 100 010(1)(1)(1)1 01 k ==
27.Sincethematrixistriangular,byTheorem2thedeterminantistheproductofthediagonalentries:
28.Sincethematrixistriangular,byTheorem2thedeterminantistheproductofthediagonalentries:
Ineachofthesecases,thematrixistriangularanditsdeterminantistheproductofitsdiagonal entries,whichis1.Thusthedeterminantofa3 × 3elementaryrowreplacementmatrixis1.
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
32.A3 × 3elementaryscalingmatrixwith k onthediagonallookslikeoneofthethreematrices
00100100
⎡⎤⎡⎤⎡⎤
010,00,010
00100100
k k k
Ineachofthesecases,thematrixistriangularanditsdeterminantistheproductofitsdiagonal entries,whichisk.Thusthedeterminantofa3 × 3elementaryscalingmatrixwith k onthediagonal is k.
33. 01 10, E ⎡⎤ = ⎢⎥ ⎣⎦ , ab A cd ⎡⎤ = ⎢⎥ ⎣⎦ cd EA ab ⎡⎤ = ⎢⎥ ⎣⎦
det E =–1,det A = ad – bc, det EA = cb – da =–1(ad – bc)=(det E)(det A)
34. 10 0, E k ⎡⎤ = ⎢⎥ ⎣⎦ , ab A cd ⎡⎤ = ⎢⎥ ⎣⎦ ab EA kckd ⎡⎤ = ⎢⎥ ⎣⎦
det E = k,det A = ad – bc, det EA = a(kd)–(kc)b = k(ad – bc)=(det E)(det A)
35. 1 01, k E ⎡⎤ = ⎢⎥ ⎣⎦ , ab A cd ⎡⎤ = ⎢⎥ ⎣⎦ akcbkd EA cd ++ ⎡⎤ = ⎢⎥ ⎣⎦
det E =1,det A = ad – bc, det EA =(a + kc)d – c(b + kd)= ad + kcd – bc – kcd =1(ad – bc)=(det E)(det A)
36 10 1, E k ⎡⎤ = ⎢⎥ ⎣⎦ , ab A cd ⎡⎤ = ⎢⎥ ⎣⎦ ab EA kackbd ⎡⎤ = ⎢⎥ ++ ⎣⎦
det E =1,det A = ad – bc, det EA = a(kb + d)–(ka + c)b = kab + ad – kab – bc =1(ad – bc)=(det E)(det A)
37. 31 42, A ⎡⎤
155 5, 2010 A
det A =2,det5A =50 ≠ 5det A
, kakb kA kckd ⎡⎤ = ⎢⎥ ⎣⎦ det A = ad – bc, 22 det()()()()()det kAkakdkbkckadbckA =−=−=
39a
.True.Seetheparagraphprecedingthedefinitionofthedeterminant.
b.False.Seethedefinitionofcofactor,whichprecedesTheorem1.
40. a.False.SeeTheorem1.
b.False.SeeTheorem2.
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
41.Theareaoftheparallelogramdeterminedby
0 is6,sincethebaseof theparallelogramhaslength3andtheheightoftheparallelogramis2.Bythesamereasoning,the areaoftheparallelogramdeterminedby
Alsonotethat [] 31 detdet6, 02
matrixwhosecolumnsarethosevectorswhichdefinethesidesoftheparallelogramadjacentto 0 is equaltotheareaoftheparallelogram
42.Theareaoftheparallelogramdeterminedby
theparallelogramhaslengthcandtheheightoftheparallelogramisb.
thematrixwhosecolumnsarethosevectorswhichdefinethesidesoftheparallelogramadjacentto 0 eitherisequaltotheareaoftheparallelogramorisequaltothenegativeoftheareaofthe parallelogram. 43[M] Answerswillvary.Theconclusionshouldbethatdet(A + B) ≠ det A +det B 44[M] Answerswillvary.Theconclusionshouldbethatdet(AB)=(det A)(det B).
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
45. [M] Answerswillvary.For4 × 4matrices,theconclusionsshouldbethatdetdet, T AA = det(–A)=det A,det(2A)=16det A,and4det(10)10detAA = .For5 × 5matrices,theconclusions shouldbethatdetdet, T AA = det(–A)=–det A,det(2A)=32det A,and5det(10)10det. AA = For6 × 6matrices,theconclusionsshouldbethatdetdet T AA = ,det(–A)=det A,det(2A)=64det A,and 6 det(10)10det. AA =
46[M] Answerswillvary.Theconclusionshouldbethat1det1/det. AA =
3.2 SOLUTIONS
Notes:Thissectionpresentsthemainpropertiesofthedeterminant,includingtheeffectsofrow operationsonthedeterminantofamatrix.ThesepropertiesarefirststudiedbyexamplesinExercises1–20.Thepropertiesaretreatedinamoretheoreticalmannerinlaterexercises.Anefficientmethodfor computingthedeterminantusingrowreductionandselectivecofactorexpansionispresentedinthis sectionandusedinExercises11–14.Theorems4and6areusedextensivelyinChapter5.Thelinearity propertyofthedeterminantstudiedinthetextisoptional,butisusedinmoreadvancedcourses.
1.Rows1and2areinterchanged,sothedeterminantchangessign(Theorem3b.).
2.Theconstant2maybefactoredoutoftheRow1(Theorem3c.).
3.Therowreplacementoperationdoesnotchangethedeterminant(Theorem3a.).
4.Therowreplacementoperationdoesnotchangethedeterminant(Theorem3a.).
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
11.Firstusearowreplacementtocreatezerosinthesecondcolumn,andthenexpanddownthesecond column:
Nowusearowreplacementtocreatezerosinthefirstcolumn,andthenexpanddownthefirst
12.Firstusearowreplacementtocreatezerosinthefourthcolumn,andthenexpanddownthefourth column:
Nowusearowreplacementtocreatezerosinthefirstcolumn,andthenexpanddownthefirst
13.Firstusearowreplacementtocreatezerosinthefourthcolumn,andthenexpanddownthefourth column:
27a.True.SeeTheorem3.
b.True.SeetheparagraphfollowingExample2.
c.True.SeetheparagraphfollowingTheorem4.
d.False.SeethewarningfollowingExample5.
28a.True.SeeTheorem3.
b.False.SeetheparagraphsfollowingExample2.
c.False.SeeExample3.
d.False.SeeTheorem5.
29.ByTheorem6,555det(det)(2)32 BB==−=−
30.Supposethetworowsofasquarematrix A areequal.Byswappingthesetworows,thematrix A is notchangedsoitsdeterminantshouldnotchange.Butsinceswappingrowschangesthesignofthe determinant,det A =–det A.Thisisonlypossibleifdet A =0.Thesamemaybeproventruefor columnsbyapplyingtheaboveresultto TA andusingTheorem5.
31.ByTheorem6,1(det)(det)det1 AAI== ,so 1 det1/det. AA =
32.Byfactoringan r outofeachofthe n rows,det()det. n rArA =
33.ByTheorem6,det AB =(det A)(det B)=(det B)(det A)=det BA.
34.ByTheorem6andExercise31, 111 det()(det)(det)(det)(det)(det)(det) PAPPAPPPA == 1 (det)(det)1det det PAA P
det A =
35.ByTheorem6andTheorem5,2 det(det)(det)(det). TT UUUUU == Since, T UUI = detdet1 T UUI== ,so 2 (det)1. U = Thusdet U = ±1.
36.ByTheorem644 det(det) AA = .Since4det0 A = ,then4(det)0 A = .Thusdet A =0,and A isnot invertiblebyTheorem4.
37.OnemaycomputeusingTheorem2thatdet A =3anddet B =8,while 60 174 AB ⎡⎤ = ⎢⎥ ⎣⎦ .Thus det AB =24=3 × 8=(det A)(det B).
38.Onemaycomputethatdet A =0anddet B =–2,while 60 20 AB ⎡ ⎤ = ⎢ ⎥ ⎣ ⎦ .Thus det AB =0=0 × –2=(det A)(det B).
39. a.ByTheorem6,det AB =(det A)(det B)=4 × –3=–12.
b.ByExercise32,3det55det1254500 AA==×= .
c.ByTheorem5,detdet3 T BB==− .
d.ByExercise31,1det1/det1/4 AA== .
e.ByTheorem6,333det(det)464 AA===
40. a.ByTheorem6,det AB =(det A)(det B)=–1 × 2=–2.
b.ByTheorem6,555det(det)232 BB===
c.ByExercise32,4det22det16116 AA==×−=− .
d.ByTheorems5and6,det(det)(det)(det)(det)111 TT AAAAAA === × =
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
e.ByTheorem6andExercise31, 11
det(det)(det)(det)(1/det)(det)(det)det1 BABBABBABA === =
41.det A =(a + e)d – c(b + f)= ad + ed – bc – cf =(ad – bc)+(ed – cf)=det B +det C
42. 1 det()(1)(1)1detdet 1 ab ABadcbadadcbAadB cd + +==++−=+++−=+++ + ,so det(A + B)=det A +det B ifandonlyif a + d =0.
43.Computedet A byusingacofactorexpansiondownthethirdcolumn:
111322233333 det()det()det()det AuvAuvAuvA =+−+++
113223333113223333 detdetdetdetdetdet uAuAuAvAvAvA =−++−+ detdetBC=+
44.ByTheorem5,detdet(). AEAET = Since()TTT AEEA = ,detdet(). TTAEEA = Now T E isitself anelementarymatrix,sobytheproofofTheorem3,det()(det)(det). TTTT EAEA = Thusitistrue thatdet(det)(det), TT AEEA = andbyapplyingTheorem5,det AE =(det E)(det A).
45. [M] Answerswillvary,butwillshowthatdet T AA alwaysequals0whiledet TAA shouldseldom bezero.Toseewhy T AA -shouldnotbeinvertible(andthusdet0 T AA = ),let A beamatrixwith morecolumnsthanrows.Thenthecolumnsof A mustbelinearlydependent,sotheequation Ax = 0 musthaveanon-trivialsolution x.Thus()(), TTT AAAAA=== xx00 andtheequation() T AA = x0 hasa non-trivialsolution.Since T AA isasquarematrix,theInvertibleMatrixTheoremnowsaysthat T AA isnotinvertible.Noticethatthesameargumentwillnotworkingeneralfor, TAA since TA hasmorerowsthancolumns,soitscolumnsarenotautomaticallylinearlydependent.
46. [M] Onemaycomputeforthismatrixthatdet A =–4008andcond A ≈ 16.3.Notethatthisisthe 2 conditionnumber,whichisusedinSection2.3.Sincedet A ≠ 0,itisinvertibleand
837181207297
175057430654
4008171195871095
2118781639
Thedeterminantisverysensitivetoscaling,as4det1010det40,080,000 AA==− and det0.1 A = 4 (0.1)det0.4008. A =−
Theconditionnumberisnotchangedatallbyscaling: cond(10A)=cond(0.1A)=cond A ≈ 16.3.When4 AI = ,det A=1andcond A =1.Asbeforethe determinantissensitivetoscaling:4det1010det10,000 AA== and4det0.1(0.1)det0.0001. AA==
Yettheconditionnumberisnotchangedbyscaling:cond(10A)=cond(0.1A)=cond A =1.
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
3.3 SOLUTIONS
Notes:Thissectionfeaturesseveralindependenttopicsfromwhichtochoose.Thegeometric interpretationofthedeterminant(Theorem10)providesthekeytochangesofvariablesinmultiple integrals.StudentsofeconomicsandengineeringarelikelytoneedCramer’sRuleinlatercourses. Exercises1–10concernCramer’sRule,exercises11–18dealwiththeadjugate,andexercises19–32 coverthegeometricinterpretationofthedeterminant.Inparticular,Exercise25examinesstudents’ understandingoflinearindependenceandrequiresacarefulexplanation,whichisdiscussedinthe Study Guide.The StudyGuide alsocontainsaheuristicproofofTheorem9for2 × 2matrices.
1212 3753 (),(),det6,det()5,det()1, 1421 AAAAA
6146 (),(),det3,det()5,det()2, 7257 AAAAA
12 det()det()52 ,. det3det3 AA xx AA ====−
3.Thesystemisequivalentto Ax = b,where 32
1212 7237 (),(),det8,det()32,det()20, 5655 AAAAA
=====
12 12 det()det()32205 4, det8det82 AA xx AA ====== bb
4.Thesystemisequivalentto Ax = b,where 53 31 A ⎡ ⎤ = ⎢ ⎥ ⎣ ⎦ and 9 5 ⎡ ⎤ = ⎢ ⎥ ⎣ ⎦ b .Wecompute
=== −==
1212 9359 (),(),det4,det()6,det()2, 5135 AAAAA
12 12 det()det()6321 ,. det42det42 AA xx AA ===−=== bb
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
123
710270217 ()801,()381,()308, 312032013 AAA
123 det4,det()6,det()16,det()14, AAAA ====− bbb
123 123 det()det()det()6316147 ,4,. det42det4det42 AAA xxx
AAA =========− bbb
411241214 ()202,()122,()102, 213323312 AAA
det4,det()16,det()52,det()4, AAAA ==−==−
123 det()det()det()16524 4,13,1.
bbbb
AA
=
35
bbbb
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
9.Thesystemisequivalentto Ax = b,where
1212 121 (),(),det()2,det()43. 4634 ss AAAsAs s ⎡⎤⎡⎤ ==== + ⎢⎥⎢⎥
Since2det666(1)0 Assss=+=+= for s =0,–1,thesystemwillhaveauniquesolutionwhen s ≠ 0, –1.Forsuchasystem,thesolutionwillbe
12 12 det()det()2143
,. det6(1)3(1)det6(1) AAss xx AsssAss + ===== ++ + bb
10.Thesystemisequivalentto Ax = b,where 21 36
1212 1121 (),(),det()62,det(). 2632 s
Since2det1233(41)0 Assss =−=−= for s =0,1/4,thesystemwillhaveauniquesolutionwhen s ≠ 0,1/4.Forsuchasystem,thesolutionwillbe
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
18.Eachcofactorof A isanintegersinceitisasumofproductsofentriesin A.Henceallentriesinadj A willbeintegers.Sincedet A =1,theinverseformulainTheorem8showsthatalltheentriesin1 A willbeintegers.
19.Theparallelogramisdeterminedbythecolumnsof 56 24 A
20.Theparallelogramisdeterminedbythecolumnsof 14 35 A
21.Firsttranslateonevertextotheorigin.Forexample,subtract(–1,0)fromeachvertextogetanew parallelogramwithvertices(0,0),(1,5),(2,–4),and(3,1).Thisparallelogramhasthesameareaas theoriginal,andisdeterminedbythecolumnsof 12
22.Firsttranslateonevertextotheorigin.Forexample,subtract(0,–2)fromeachvertextogetanew parallelogramwithvertices(0,0),(6,1),(–3,3),and(3,4).Thisparallelogramhasthesameareaas theoriginal,andisdeterminedbythecolumnsof
23.Theparallelepipedisdeterminedbythecolumnsof
24.Theparallelepipedisdeterminedbythecolumnsof
parallelepipedis|det A|=|–15|=15.
25.TheInvertibleMatrixTheoremsaysthata3 × 3matrix A isnotinvertibleifandonlyifitscolumns arelinearlydependent.Thiswillhappenifandonlyifoneofthecolumnsisalinearcombinationof theothers;thatis,ifoneofthevectorsisintheplanespannedbytheothertwovectors.Thisis equivalenttotheconditionthattheparallelepipeddeterminedbythethreevectorshaszerovolume, whichisinturnequivalenttotheconditionthatdet A =0.
26.Bydefinition, p + S isthesetofallvectorsoftheform p + v,where v isin S.Applying T toatypical vectorin p + S,wehave T(p + v)= T(p)+ T(v).Thisvectorisinthesetdenotedby T(p)+ T(S).This provesthat T mapstheset p + S intotheset T(p)+ T(S).Conversely,anyvectorin T(p)+ T(S)has theform T(p)+ T(v)forsome v in S.Thisvectormaybewrittenas T(p + v).Thisshowsthatevery vectorin T(p)+ T(S)istheimageunder T ofsomepoint p + v in p + S.
27.Sincetheparallelogram S
S}=6 ⋅ 4=24.Alternatively,onemaycomputethevectorsthatdeterminethe image,namely,thecolumnsof
Thedeterminantofthismatrixis–24,sotheareaoftheimageis24.
A|{areaof S}=5 ⋅ 4=20.Alternatively,onemaycomputethevectorsthatdeterminetheimage, namely,thecolumnsof
7240142 117131
Thedeterminantofthismatrixis20,sotheareaoftheimageis20.
29.Theareaofthetrianglewillbeonehalfoftheareaoftheparallelogramdeterminedby1 v and2 v ByTheorem9,theareaofthetrianglewillbe(1/2)|det A|,where [] 12. A = vv
30.Translate R toanewtriangleofequalareabysubtracting33(,) xy fromeachvertex.Thenewtriangle hasvertices(0,0),1313 (,) xxyy ,and2323 (,). xxyy ByExercise29,theareaofthetriangle willbe
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
uxa = ,22/uxb = ,and33/uxc = ,and u liesinside S (or
1231uuu++≤ )ifandonlyif x liesinside T(S)(or 222 123 2221xxx abc ++≤ ).
b.BythegeneralizationofTheorem10, {volumeofellipsoid}{volumeof()} TS =
32a.Alineartransformation T thatmaps S onto S ′ willmap1 e to1, v 2 e to2, v and3 e to3; v thatis, 11() T = ev ,22 () T = ev ,and33 (). T = ev Thestandardmatrixforthistransformationwillbe [] [] 123123 ()()().ATTT ==eeevvv
b.Theareaofthebaseof S is(1/2)(1)(1)=1/2,sothevolumeof S is(1/3)(1/2)(1)=1/6.Byparta.
T(S)= S ′ ,sothegeneralizationofTheorem10givesthatthevolumeof S ′ is|det A|{volumeof S}=(1/6)|det A|.
33. [M] Answerswillvary.InMATLAB,entriesin B –inv(A)areapproximately1510orsmaller.
34[M] Answerswillvary,aswillthecommandswhichproducethesecondentryof x.Forexample,the MATLABcommandis x2 = det([A(:,1) b A(:,3:4)])/det(A) whiletheMathematica commandis x2 = Det[{Transpose[A][[1]],b,Transpose[A][[3]], Transpose[A][[4]]}]/Det[A].
35. [M] MATLABStudentVersion4.0uses57,771flopsforinv A and14,269,045flopsfortheinverse formula.The inv(A) commandrequiresonlyabout0.4%oftheoperationsfortheinverseformula.
Chapter 3 SUPPLEMENTARY EXERCISES
1. a.True.Thecolumnsof A arelinearlydependent.
b.True.SeeExercise30inSection3.2.
c.False.SeeTheorem3(c);inthiscase3det55detAA = .
d.False.Consider
e.False.ByTheorem6,33 det2 A = .
f.False.SeeTheorem3(b).
g.True.SeeTheorem3(c).
h.True.SeeTheorem3(a).
i.False.SeeTheorem5.
j.False.SeeTheorem3(c);thisstatementisfalsefor n × n invertiblematriceswith n aneven integer.
k.True.SeeTheorems6and5;2 det(det) T AAA = .
l.False.Thecoefficientmatrixmustbeinvertible.
m.False.Theareaofthe triangle is5.
n.True.SeeTheorem6;33 det(det) AA = .
o.False.SeeExercise31inSection3.2.
p.True.SeeTheorem6.
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
Thisisanequationoftheform ax + by + c =0,andsincethepoints11(,) xy and22(,) xy aredistinct, atleastoneof a and b isnotzero.Thustheequationistheequationofaline.Thepoints11(,) xy and 22 (,) xy areontheline,becausewhenthecoordinatesofoneofthepointsaresubstitutedfor x and y, tworowsofthematrixareequalandsothedeterminantiszero.
10.Expandingalongthefirstrowwillshowthat23 ()det0123. ftVcctctct ==+++ ByExercise9,
since1 x ,2 x ,and3 x aredistinct.Thus f (t)isacubicpolynomial.Thepoints1(,0) x ,2(,0) x ,and 3 (,0) x areonthegraphof f,sincewhenanyof1 x ,2 x or3 x aresubstitutedfor t,thematrixhastwo equalrowsandthusitsdeterminant(whichis f (t))iszero.Thus()0 ifx = for i =1,2,3.
11.Totellifaquadrilateraldeterminedbyfourpointsisaparallelogram,firsttranslateoneofthe verticestotheorigin.Ifwelabeltheverticesofthisnewquadrilateralas 0,1 v ,2 v ,and3 v ,then theywillbetheverticesofaparallelogramifoneof1 v ,2 v ,or3 v isthesumoftheothertwo.In thisexample,subtract(1,4)fromeachvertextogetanewparallelogramwithvertices 0 =(0,0), 1(2,1) =− v ,2(2,5) = v ,and3(4,4) = v .Since231 =+ vvv ,thequadrilateralisaparallelogramas stated.Thetranslatedparallelogramhasthesameareaastheoriginal,andisdeterminedbythe columnsof [] 13 24 14 A ⎡⎤ == ⎢⎥
vv
12.A2 × 2matrix A isinvertibleifandonlyiftheparallelogramdeterminedbythecolumnsof A has nonzeroarea.
13.ByTheorem8,1 1 (adj)detAAAAI A ⋅== .BytheInvertibleMatrixTheorem,adj A isinvertible and11 (adj)detAA A = .
14. a.Considerthematrix k k
AO A OI ⎡⎤ = ⎢⎥
,where1 ≤ k ≤ n and O isanappropriatelysizedzeromatrix.
Wewillshowthatdetdet k AA = forall1 ≤ k ≤ n bymathematicalinduction.
Firstlet k =1.Expandalongthelastrowtoobtain
(1)(1) 1 detdet(1)1detdet. 1 nn AO AAA O +++ ⎡⎤ ==−⋅⋅= ⎢⎥ ⎣⎦
Nowlet1< k ≤ n andassumethat1detdet. k AA = Expandalongthelastrowof kA toobtain
AO AAAA OI +++ ⎡⎤ ==−⋅⋅== ⎢⎥ ⎣⎦ Thuswehaveproventhe result,andthedeterminantofthematrixinquestionisdet A.
()() 11 detdet(1)1detdetdet. nknk kkk k
b.Considerthematrix k k k
IO A CD ⎡⎤ = ⎢⎥ ⎣⎦ ,where1 ≤ k ≤ n, kC isan n × k matrixand O isan appropriatelysizedzeromatrix.Wewillshowthatdetdet k AD = forall1 ≤ k ≤ n by mathematicalinduction.
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
15. a.Computetherightsideoftheequation:
Setthisequaltotheleftsideoftheequation:
Since XA = C and A isinvertible,1 . XCA = Since XB + Y = D,
16. a.Doingthegivenoperationsdoesnotchangethedeterminantof A sincethegivenoperationsare allrowreplacementoperations.Theresultingmatrixis
b.Sincecolumnreplacementoperationsareequivalenttorowoperationson TA anddetdet T AA = , thegivenoperationsdonotchangethedeterminantofthematrix.Theresultingmatrixis
c.Sincetheprecedingmatrixisatriangularmatrixwiththesamedeterminantas A, 1 det()((1)). n Aabanb =−+−
17.Firstconsiderthecase n =2.Inthiscase
=+=−+−=−=−+=−+−
Nowassumethattheformulaholdsforall(k –1) × (k –1)matrices,andlet A, B,and C be k × k matrices.Byacofactorexpansionalongthefirstcolumn,
()()((2))()((2)) kk
sincethematrixintheaboveformulaisa(k –1) × (k –1)matrix.Wecanperformaseriesofrow operationson C to“zeroout”belowthefirstpivot,andproducethefollowingmatrixwhose determinantisdet C:
Sincethisisatriangularmatrix,wehavefoundthat1 det() Cbabk =− .Thus
111 detdetdet()((2))()()((1)), kkk ABCabakbbababakb =+=−+−+−=−+− whichiswhatwastobeshown.Thustheformulahasbeenprovenbymathematicalinduction.
18[M] Sincethefirstmatrixhas a =3, b =8,and n =4,itsdeterminantis
413 (38)(3(41)8)(5)(324)(125)(27)3375. −+−=−+=−=−
Sincethesecondmatrixhas a =8, b = 3,and n =5,itsdeterminantis514(83)(8(51)3)(5)(812)(625)(20)12,500. −+−=+==
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
Toshowthis,considerusingrowreplacementoperationsto“zeroout”belowthefirstpivot.The resultingmatrixis
Nowuserowreplacementoperationsto“zeroout”belowthesecondpivot,andsoon.Thefinal matrixwhichresultsfromthisprocessis
whichisanuppertriangularmatrixwithdeterminant1.
Copyright©2012PearsonEducation,Inc.PublishingasAddison-Wesley.
Toshowthis,considerusingrowreplacementoperationsto“zeroout”belowthefirstpivot.The resultingmatrixis
Nowuserowreplacementoperationsto“zeroout”belowthesecondpivot.Thematrixwhichresults fromthisprocessis
Thismatrixhasthesamedeterminantastheoriginalmatrix,andisrecognizableasablockmatrixof theform