Source

fcla / src / section-CB.xml

Full commit
   1
   2
   3
   4
   5
   6
   7
   8
   9
  10
  11
  12
  13
  14
  15
  16
  17
  18
  19
  20
  21
  22
  23
  24
  25
  26
  27
  28
  29
  30
  31
  32
  33
  34
  35
  36
  37
  38
  39
  40
  41
  42
  43
  44
  45
  46
  47
  48
  49
  50
  51
  52
  53
  54
  55
  56
  57
  58
  59
  60
  61
  62
  63
  64
  65
  66
  67
  68
  69
  70
  71
  72
  73
  74
  75
  76
  77
  78
  79
  80
  81
  82
  83
  84
  85
  86
  87
  88
  89
  90
  91
  92
  93
  94
  95
  96
  97
  98
  99
 100
 101
 102
 103
 104
 105
 106
 107
 108
 109
 110
 111
 112
 113
 114
 115
 116
 117
 118
 119
 120
 121
 122
 123
 124
 125
 126
 127
 128
 129
 130
 131
 132
 133
 134
 135
 136
 137
 138
 139
 140
 141
 142
 143
 144
 145
 146
 147
 148
 149
 150
 151
 152
 153
 154
 155
 156
 157
 158
 159
 160
 161
 162
 163
 164
 165
 166
 167
 168
 169
 170
 171
 172
 173
 174
 175
 176
 177
 178
 179
 180
 181
 182
 183
 184
 185
 186
 187
 188
 189
 190
 191
 192
 193
 194
 195
 196
 197
 198
 199
 200
 201
 202
 203
 204
 205
 206
 207
 208
 209
 210
 211
 212
 213
 214
 215
 216
 217
 218
 219
 220
 221
 222
 223
 224
 225
 226
 227
 228
 229
 230
 231
 232
 233
 234
 235
 236
 237
 238
 239
 240
 241
 242
 243
 244
 245
 246
 247
 248
 249
 250
 251
 252
 253
 254
 255
 256
 257
 258
 259
 260
 261
 262
 263
 264
 265
 266
 267
 268
 269
 270
 271
 272
 273
 274
 275
 276
 277
 278
 279
 280
 281
 282
 283
 284
 285
 286
 287
 288
 289
 290
 291
 292
 293
 294
 295
 296
 297
 298
 299
 300
 301
 302
 303
 304
 305
 306
 307
 308
 309
 310
 311
 312
 313
 314
 315
 316
 317
 318
 319
 320
 321
 322
 323
 324
 325
 326
 327
 328
 329
 330
 331
 332
 333
 334
 335
 336
 337
 338
 339
 340
 341
 342
 343
 344
 345
 346
 347
 348
 349
 350
 351
 352
 353
 354
 355
 356
 357
 358
 359
 360
 361
 362
 363
 364
 365
 366
 367
 368
 369
 370
 371
 372
 373
 374
 375
 376
 377
 378
 379
 380
 381
 382
 383
 384
 385
 386
 387
 388
 389
 390
 391
 392
 393
 394
 395
 396
 397
 398
 399
 400
 401
 402
 403
 404
 405
 406
 407
 408
 409
 410
 411
 412
 413
 414
 415
 416
 417
 418
 419
 420
 421
 422
 423
 424
 425
 426
 427
 428
 429
 430
 431
 432
 433
 434
 435
 436
 437
 438
 439
 440
 441
 442
 443
 444
 445
 446
 447
 448
 449
 450
 451
 452
 453
 454
 455
 456
 457
 458
 459
 460
 461
 462
 463
 464
 465
 466
 467
 468
 469
 470
 471
 472
 473
 474
 475
 476
 477
 478
 479
 480
 481
 482
 483
 484
 485
 486
 487
 488
 489
 490
 491
 492
 493
 494
 495
 496
 497
 498
 499
 500
 501
 502
 503
 504
 505
 506
 507
 508
 509
 510
 511
 512
 513
 514
 515
 516
 517
 518
 519
 520
 521
 522
 523
 524
 525
 526
 527
 528
 529
 530
 531
 532
 533
 534
 535
 536
 537
 538
 539
 540
 541
 542
 543
 544
 545
 546
 547
 548
 549
 550
 551
 552
 553
 554
 555
 556
 557
 558
 559
 560
 561
 562
 563
 564
 565
 566
 567
 568
 569
 570
 571
 572
 573
 574
 575
 576
 577
 578
 579
 580
 581
 582
 583
 584
 585
 586
 587
 588
 589
 590
 591
 592
 593
 594
 595
 596
 597
 598
 599
 600
 601
 602
 603
 604
 605
 606
 607
 608
 609
 610
 611
 612
 613
 614
 615
 616
 617
 618
 619
 620
 621
 622
 623
 624
 625
 626
 627
 628
 629
 630
 631
 632
 633
 634
 635
 636
 637
 638
 639
 640
 641
 642
 643
 644
 645
 646
 647
 648
 649
 650
 651
 652
 653
 654
 655
 656
 657
 658
 659
 660
 661
 662
 663
 664
 665
 666
 667
 668
 669
 670
 671
 672
 673
 674
 675
 676
 677
 678
 679
 680
 681
 682
 683
 684
 685
 686
 687
 688
 689
 690
 691
 692
 693
 694
 695
 696
 697
 698
 699
 700
 701
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
<?xml version="1.0" encoding="UTF-8" ?>
<section acro="CB">
<title>Change of Basis</title>

<!-- %%%%%%%%%% -->
<!-- % -->
<!-- %  Section CB -->
<!-- %  Change of Basis -->
<!-- % -->
<!-- %%%%%%%%%% -->
<introduction>
<p>We have seen in <acroref type="section" acro="MR" /> that a linear transformation can be represented by a matrix, once we pick bases for the domain and codomain.  How does the matrix representation change if we choose different bases?  Which bases lead to especially nice representations?  From the infinite possibilities, what is the best possible representation?  This section will begin to answer these questions.  But first we need to define eigenvalues for linear transformations and the change-of-basis matrix.</p>

</introduction>

<subsection acro="EELT">
<title>Eigenvalues and Eigenvectors of Linear Transformations</title>

<p>We now define the notion of an eigenvalue and eigenvector of a linear transformation.  It should not be too surprising, especially if you remind yourself of the close relationship between matrices and linear transformations.</p>

<definition acro="EELT" index="eigenvalue!linear transformation">
<title>Eigenvalue and Eigenvector of a Linear Transformation</title>
<indexlocation index="eigenvector!linear transformation" />
<p>Suppose that $\ltdefn{T}{V}{V}$ is a linear transformation.  Then a nonzero vector $\vect{v}\in V$ is an <define>eigenvector</define> of $T$ for the <define>eigenvalue</define> $\lambda$ if $\lt{T}{\vect{v}}=\lambda\vect{v}$.</p>

</definition>

<p>We will see shortly the best method for computing the eigenvalues and eigenvectors of a linear transformation, but for now, here are some examples to verify that such things really do exist.</p>

<example acro="ELTBM" index="eigenvectors!linear transformation">
<title>Eigenvectors of linear transformation between matrices</title>

<p>Consider the linear transformation $\ltdefn{T}{M_{22}}{M_{22}}$ defined by
<equation>
<![CDATA[\lt{T}{\begin{bmatrix}a&b\\c&d\end{bmatrix}}]]>
=
\begin{bmatrix}
-17a+11b+8c-11d
<![CDATA[&]]>
-57a+35b+24c-33d
\\
-14a+10b+6c-10d
<![CDATA[&]]>
-41a+25b+16c-23d
\end{bmatrix}
</equation>
and the vectors
<alignmath>
\vect{x}_1
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[ 0 & 1 \\ 0 & 1]]>
\end{bmatrix}
<![CDATA[&]]>
\vect{x}_2
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[ 1 & 1 \\ 1 & 0]]>
\end{bmatrix}
<![CDATA[&]]>
\vect{x}_3
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[ 1 & 3 \\ 2 & 3]]>
\end{bmatrix}
<![CDATA[&]]>
\vect{x}_4
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[ 2 & 6 \\ 1 & 4]]>
\end{bmatrix}
</alignmath>
</p>

<p>Then compute
<alignmath>
\lt{T}{\vect{x}_1}
<![CDATA[&=]]>
<![CDATA[\lt{T}{\begin{bmatrix} 0 & 1 \\ 0 & 1\end{bmatrix}}]]>
=
\begin{bmatrix}
<![CDATA[ 0 & 2 \\ 0 & 2]]>
\end{bmatrix}
=
2\vect{x}_1\\
\lt{T}{\vect{x}_2}
<![CDATA[&=]]>
<![CDATA[\lt{T}{\begin{bmatrix} 1 & 1 \\ 1 & 0\end{bmatrix}}]]>
=
\begin{bmatrix}
<![CDATA[ 2 & 2 \\ 2 & 0]]>
\end{bmatrix}
=
2\vect{x}_2\\
\lt{T}{\vect{x}_3}
<![CDATA[&=]]>
<![CDATA[\lt{T}{\begin{bmatrix} 1 & 3 \\ 2 & 3\end{bmatrix}}]]>
=
\begin{bmatrix}
<![CDATA[ -1 & -3 \\ -2 & -3]]>
\end{bmatrix}
=
(-1)\vect{x}_3\\
\lt{T}{\vect{x}_4}
<![CDATA[&=]]>
<![CDATA[\lt{T}{\begin{bmatrix} 2 & 6 \\ 1 & 4\end{bmatrix}}]]>
=
\begin{bmatrix}
<![CDATA[ -4 & -12 \\ -2 & -8]]>
\end{bmatrix}
=
(-2)\vect{x}_4\\
</alignmath>
</p>

<p>So $\vect{x}_1$, $\vect{x}_2$, $\vect{x}_3$, $\vect{x}_4$ are eigenvectors of $T$ with eigenvalues (respectively) $\lambda_1=2$, $\lambda_2=2$, $\lambda_3=-1$, $\lambda_4=-2$.</p>

</example>

<p>Here's another.</p>

<example acro="ELTBP" index="eigenvectors!linear transformation">
<title>Eigenvectors of linear transformation between polynomials</title>

<p>Consider the linear transformation $\ltdefn{R}{P_2}{P_2}$ defined by
<equation>
\lt{R}{a+bx+cx^2}=
(15a+8b-4c)+(-12a-6b+3c)x+(24a+14b-7c)x^2
</equation>
and the vectors
<alignmath>
\vect{w}_1
<![CDATA[&=1-x+x^2]]>
<![CDATA[&]]>
\vect{w}_2
<![CDATA[&=x+2x^2]]>
<![CDATA[&]]>
\vect{w}_3
<![CDATA[&=1+4x^2]]>
<![CDATA[&]]>
</alignmath>
</p>

<p>Then compute
<alignmath>
\lt{R}{\vect{w}_1}
<![CDATA[&=]]>
\lt{R}{1-x+x^2}
=
3-3x+3x^2
=3\vect{w}_1\\
\lt{R}{\vect{w}_2}
<![CDATA[&=]]>
\lt{R}{x+2x^2}
=
0+0x+0x^2
=0\vect{w}_2\\
\lt{R}{\vect{w}_3}
<![CDATA[&=]]>
\lt{R}{1+4x^2}
=
-1-4x^2
=(-1)\vect{w}_3\\
</alignmath>
</p>

<p>So $\vect{w}_1$, $\vect{w}_2$, $\vect{w}_3$ are eigenvectors of $R$ with eigenvalues (respectively) $\lambda_1=3$, $\lambda_2=0$, $\lambda_3=-1$.  Notice how the eigenvalue $\lambda_2=0$ indicates that the eigenvector $\vect{w}_2$ is a non-trivial element of the kernel of $R$, and therefore $R$ is not injective (<acroref type="exercise" acro="CB.T15" />).</p>

</example>

<p>Of course, these examples are meant only to illustrate the definition of eigenvectors and eigenvalues for linear transformations, and therefore beg the question, <q>How would I <em>find</em> eigenvectors?</q>  We'll have an answer before we finish this section.  We need one more construction first.</p>

<sageadvice acro="ENDO" index="endomorphisms">
<title>Endomorphisms</title>
An <define>endomorphism</define> is an <q>operation-preserving</q> function (a <q>morphism</q>) whose domain and codomain are equal.  Sage takes this definition one step further for linear transformations and requires that the domain and codomain have the same bases (either a default echelonized basis or the same user basis).  When a linear transformation meets this extra requirement, several natural methods become available.<br /><br />
Principally, we can compute the eigenvalues provided by <acroref type="definition" acro="EELT" />.  We also get a natural notion of a characteristic polynomial.
<sage>
<input>x1, x2, x3, x4 = var('x1, x2, x3, x4')
outputs = [ 4*x1 + 2*x2 -   x3 + 8*x4,
            3*x1 - 5*x2 - 9*x3       ,
                   6*x2 + 7*x3 + 6*x4,
           -3*x1 + 2*x2 + 5*x3 - 3*x4]
T_symbolic(x1, x2, x3, x4) = outputs
T = linear_transformation(QQ^4, QQ^4, T_symbolic)
T.eigenvalues()
</input>
<output>[3, -2, 1, 1]
</output>
</sage>

<sage>
<input>cp = T.characteristic_polynomial()
cp
</input>
<output>x^4 - 3*x^3 - 3*x^2 + 11*x - 6
</output>
</sage>

<sage>
<input>cp.factor()
</input>
<output>(x - 3) * (x + 2) * (x - 1)^2
</output>
</sage>

Now the question of eigenvalues being elements of the set of scalars used for the vector space becomes even more obvious.  If we define an endomorphism on a vector space whose scalars are the rational numbers, should we <q>allow</q> irrational or complex eigenvalues?  You will now recognize our use of the complex numbers in the text for the gross convenience that it is.


</sageadvice>
</subsection>

<subsection acro="CBM">
<title>Change-of-Basis Matrix</title>

<p>Given a vector space, we know we can usually find many different bases for the vector space, some nice, some nasty.  If we choose a single vector from this vector space, we can build many different representations of the vector by constructing the representations relative to different bases.  How are these different representations related to each other?  A change-of-basis matrix answers this question.</p>

<definition acro="CBM" index="change-of-basis matrix">
<title>Change-of-Basis Matrix</title>
<p>Suppose that $V$ is a vector space, and $\ltdefn{I_V}{V}{V}$ is the identity linear transformation on $V$.  Let $B=\set{\vectorlist{v}{n}}$ and $C$ be two bases of $V$.  Then the <define>change-of-basis matrix</define> from $B$ to $C$ is the matrix representation of $I_V$ relative to $B$ and $C$,
<alignmath>
<![CDATA[\cbm{B}{C}&=\matrixrep{I_V}{B}{C}\\]]>
<![CDATA[&=\matrixrepcolumns{I_V}{C}{v}{n}\\]]>
<![CDATA[&=\left\lbrack]]>
\left.\vectrep{C}{\vect{v}_1}\right|
\left.\vectrep{C}{\vect{v}_2}\right|
\left.\vectrep{C}{\vect{v}_3}\right|
\ldots
\left|\vectrep{C}{\vect{v}_n}\right.
\right\rbrack
</alignmath>
</p>

</definition>

<p>Notice that this definition is primarily about a single vector space ($V$) and two bases of $V$ ($B$, $C$).  The linear transformation ($I_V$) is necessary but not critical.  As you might expect, this matrix has something to do with changing bases.  Here is the theorem that gives the matrix its name (not the other way around).</p>

<theorem acro="CB" index="change-of-basis">
<title>Change-of-Basis</title>
<statement>
<p>Suppose that $\vect{v}$ is a vector in the vector space $V$ and $B$ and $C$ are bases of $V$.  Then
<equation>
\vectrep{C}{\vect{v}}=\cbm{B}{C}\vectrep{B}{\vect{v}}
</equation>
</p>

</statement>

<proof>
<p>
<alignmath>
\vectrep{C}{\vect{v}}
<![CDATA[&=\vectrep{C}{\lt{I_V}{\vect{v}}}]]>
<![CDATA[&&]]>\text{<acroref type="definition" acro="IDLT" />}\\
<![CDATA[&=\matrixrep{I_V}{B}{C}\vectrep{B}{\vect{v}}]]>
<![CDATA[&&]]>\text{<acroref type="theorem" acro="FTMR" />}\\
<![CDATA[&=\cbm{B}{C}\vectrep{B}{\vect{v}}]]>
<![CDATA[&&]]>\text{<acroref type="definition" acro="CBM" />}
</alignmath>
</p>

</proof>
</theorem>

<p>So the change-of-basis matrix can be used with matrix multiplication to convert a vector representation of a vector ($\vect{v}$) relative to one basis ($\vectrep{B}{\vect{v}}$) to a representation of the same vector relative to a second basis ($\vectrep{C}{\vect{v}}$).</p>

<theorem acro="ICBM" index="change-of-basis matrix!inverse">
<title>Inverse of Change-of-Basis Matrix</title>
<statement>
<p>Suppose that $V$ is a vector space, and $B$ and $C$ are bases of $V$.
Then the change-of-basis matrix $\cbm{B}{C}$ is nonsingular and
<equation>
\inverse{\cbm{B}{C}}=\cbm{C}{B}
</equation>
</p>

</statement>

<proof>
<p>The linear transformation $\ltdefn{I_V}{V}{V}$ is invertible, and its inverse is itself, $I_V$ (check this!). So by <acroref type="theorem" acro="IMR" />, the matrix $\matrixrep{I_V}{B}{C}=\cbm{B}{C}$ is invertible.  <acroref type="theorem" acro="NI" /> says an invertible matrix is nonsingular.</p>

<p>Then
<alignmath>
\inverse{\cbm{B}{C}}
<![CDATA[&=\inverse{\left(\matrixrep{I_V}{B}{C}\right)}&&]]>\text{<acroref type="definition" acro="CBM" />}\\
<![CDATA[&=\matrixrep{\ltinverse{I_V}}{C}{B}&&]]>\text{<acroref type="theorem" acro="IMR" />}\\
<![CDATA[&=\matrixrep{I_V}{C}{B}&&]]>\text{<acroref type="definition" acro="IDLT" />}\\
<![CDATA[&=\cbm{C}{B}&&]]>\text{<acroref type="definition" acro="CBM" />}\\
</alignmath>
</p>

</proof>
</theorem>

<example acro="CBP" index="change of basis!between polynomials">
<title>Change of basis with polynomials</title>

<p>The vector space $P_4$ (<acroref type="example" acro="VSP" />) has two nice bases (<acroref type="example" acro="BP" />),
<alignmath>
<![CDATA[B&=\set{1,x,x^2,x^3,x^4}\\]]>
<![CDATA[C&=\set{1,1+x,1+x+x^2,1+x+x^2+x^3,1+x+x^2+x^3+x^4}]]>
</alignmath>
</p>

<p>To build the change-of-basis matrix between $B$ and $C$, we must first build a vector representation of each vector in $B$ relative to $C$,
<alignmath>
\vectrep{C}{1}
<![CDATA[&=\vectrep{C}{(1)\left(1\right)}]]>
=\colvector{1\\0\\0\\0\\0}\\
\vectrep{C}{x}
<![CDATA[&=\vectrep{C}{(-1)\left(1\right)+(1)\left(1+x\right)}]]>
=\colvector{-1\\1\\0\\0\\0}\\
\vectrep{C}{x^2}
<![CDATA[&=\vectrep{C}{(-1)\left(1+x\right)+(1)\left(1+x+x^2\right)}]]>
=\colvector{0\\-1\\1\\0\\0}\\
\vectrep{C}{x^3}
<![CDATA[&=\vectrep{C}{(-1)\left(1+x+x^2\right)+(1)\left(1+x+x^2+x^3\right)}]]>
=\colvector{0\\0\\-1\\1\\0}\\
\vectrep{C}{x^4}
<![CDATA[&=\vectrep{C}{(-1)\left(1+x+x^2+x^3\right)+(1)\left(1+x+x^2+x^3+x^4\right)}]]>
=\colvector{0\\0\\0\\-1\\1}
</alignmath>
</p>

<p>Then we package up these vectors as the columns of a matrix,
<equation>
\cbm{B}{C}=
\begin{bmatrix}
<![CDATA[1 &-1 & 0 & 0 & 0\\]]>
<![CDATA[0 & 1 &-1 & 0 & 0\\]]>
<![CDATA[0 & 0 & 1 &-1 & 0\\]]>
<![CDATA[0 & 0 & 0 & 1 &-1\\]]>
<![CDATA[0 & 0 & 0 & 0 & 1\\]]>
\end{bmatrix}
</equation>
</p>

<p>Now, to illustrate <acroref type="theorem" acro="CB" />, consider the vector $\vect{u}=5-3x+2x^2+8x^3-3x^4$.  We can build the representation of $\vect{u}$ relative to $B$ easily,
<equation>
\vectrep{B}{\vect{u}}=
\vectrep{B}{5-3x+2x^2+8x^3-3x^4}=
\colvector{5\\-3\\2\\8\\-3}
</equation>
</p>

<p>Applying <acroref type="theorem" acro="CB" />, we obtain a second representation of $\vect{u}$, but now relative to $C$,
<alignmath>
\vectrep{C}{\vect{u}}
<![CDATA[&=\cbm{B}{C}\vectrep{B}{\vect{u}}&&]]>\text{<acroref type="theorem" acro="CB" />}\\
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[1 &-1 & 0 & 0 & 0\\]]>
<![CDATA[0 & 1 &-1 & 0 & 0\\]]>
<![CDATA[0 & 0 & 1 &-1 & 0\\]]>
<![CDATA[0 & 0 & 0 & 1 &-1\\]]>
<![CDATA[0 & 0 & 0 & 0 & 1\\]]>
\end{bmatrix}
\colvector{5\\-3\\2\\8\\-3}\\
<![CDATA[&=\colvector{8\\-5\\-6\\11\\-3}&&]]>\text{<acroref type="definition" acro="MVP" />}
</alignmath>
</p>

<p>We can check our work by unraveling this second representation,
<alignmath>
\vect{u}
<![CDATA[&=\vectrepinv{C}{\vectrep{C}{\vect{u}}}]]>
<![CDATA[&&]]>\text{<acroref type="definition" acro="IVLT" />}\\
<![CDATA[&=\vectrepinv{C}{\colvector{8\\-5\\-6\\11\\-3}}\\]]>
<![CDATA[&=8(1)+(-5)(1+x)+(-6)(1+x+x^2)\\]]>
<![CDATA[&\quad\quad+(11)(1+x+x^2+x^3)+(-3)(1+x+x^2+x^3+x^4)]]>
<![CDATA[&&]]>\text{<acroref type="definition" acro="VR" />}\\
<![CDATA[&=5-3x+2x^2+8x^3-3x^4]]>
</alignmath>
</p>

<p>The change-of-basis matrix from $C$ to $B$ is actually easier to build.  Grab each vector in the basis $C$ and form its representation relative to $B$
<alignmath>
\vectrep{B}{1}
<![CDATA[&=\vectrep{B}{(1)1}]]>
=\colvector{1\\0\\0\\0\\0}\\
\vectrep{B}{1+x}
<![CDATA[&=\vectrep{B}{(1)1+(1)x}]]>
=\colvector{1\\1\\0\\0\\0}\\
\vectrep{B}{1+x+x^2}
<![CDATA[&=\vectrep{B}{(1)1+(1)x+(1)x^2}]]>
=\colvector{1\\1\\1\\0\\0}\\
\vectrep{B}{1+x+x^2+x^3}
<![CDATA[&=\vectrep{B}{(1)1+(1)x+(1)x^2+(1)x^3}]]>
=\colvector{1\\1\\1\\1\\0}\\
\vectrep{B}{1+x+x^2+x^3+x^4}
<![CDATA[&=\vectrep{B}{(1)1+(1)x+(1)x^2+(1)x^3+(1)x^4}]]>
=\colvector{1\\1\\1\\1\\1}\\
</alignmath>
</p>

<p>Then we package up these vectors as the columns of a matrix,
<equation>
\cbm{C}{B}=
\begin{bmatrix}
<![CDATA[1 & 1 & 1 & 1 & 1\\]]>
<![CDATA[0 & 1 & 1 & 1 & 1\\]]>
<![CDATA[0 & 0 & 1 & 1 & 1\\]]>
<![CDATA[0 & 0 & 0 & 1 & 1\\]]>
<![CDATA[0 & 0 & 0 & 0 & 1\\]]>
\end{bmatrix}
</equation>
</p>

<p>We formed two representations of the vector $\vect{u}$ above, so we can again provide a check on our computations by converting from the representation of $\vect{u}$ relative to $C$ to the representation of $\vect{u}$ relative to $B$,
<alignmath>
\vectrep{B}{\vect{u}}
<![CDATA[&=\cbm{C}{B}\vectrep{C}{\vect{u}}&&]]>\text{<acroref type="theorem" acro="CB" />}\\
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[1 & 1 & 1 & 1 & 1\\]]>
<![CDATA[0 & 1 & 1 & 1 & 1\\]]>
<![CDATA[0 & 0 & 1 & 1 & 1\\]]>
<![CDATA[0 & 0 & 0 & 1 & 1\\]]>
<![CDATA[0 & 0 & 0 & 0 & 1\\]]>
\end{bmatrix}
\colvector{8\\-5\\-6\\11\\-3}\\
<![CDATA[&=\colvector{5\\-3\\2\\8\\-3}&&]]>\text{<acroref type="definition" acro="MVP" />}\\
</alignmath>
</p>

<p>One more computation that is either a check on our work, or an illustration of a theorem.  The two change-of-basis matrices, $\cbm{B}{C}$ and $\cbm{C}{B}$, should be inverses of each other, according to <acroref type="theorem" acro="ICBM" />.  Here we go,
<equation>
\cbm{B}{C}\cbm{C}{B}=
\begin{bmatrix}
<![CDATA[1 &-1 & 0 & 0 & 0\\]]>
<![CDATA[0 & 1 &-1 & 0 & 0\\]]>
<![CDATA[0 & 0 & 1 &-1 & 0\\]]>
<![CDATA[0 & 0 & 0 & 1 &-1\\]]>
<![CDATA[0 & 0 & 0 & 0 & 1\\]]>
\end{bmatrix}
\begin{bmatrix}
<![CDATA[1 & 1 & 1 & 1 & 1\\]]>
<![CDATA[0 & 1 & 1 & 1 & 1\\]]>
<![CDATA[0 & 0 & 1 & 1 & 1\\]]>
<![CDATA[0 & 0 & 0 & 1 & 1\\]]>
<![CDATA[0 & 0 & 0 & 0 & 1\\]]>
\end{bmatrix}
=
\begin{bmatrix}
<![CDATA[1 & 0 & 0 & 0 & 0\\]]>
<![CDATA[0 & 1 & 0 & 0 & 0\\]]>
<![CDATA[0 & 0 & 1 & 0 & 0\\]]>
<![CDATA[0 & 0 & 0 & 1 & 0\\]]>
<![CDATA[0 & 0 & 0 & 0 & 1\\]]>
\end{bmatrix}
</equation>
</p>

</example>

<p>The computations of the previous example are not meant to present any labor-saving devices, but instead are meant to illustrate the <em>utility</em> of the change-of-basis matrix.  However, you might have noticed that $\cbm{C}{B}$ was easier to compute than $\cbm{B}{C}$.  If you needed $\cbm{B}{C}$, then you could first compute $\cbm{C}{B}$ and then compute its inverse, which by <acroref type="theorem" acro="ICBM" />, would equal $\cbm{B}{C}$.</p>

<p>Here's another illustrative example.  We have been concentrating on working with abstract vector spaces, but all of our theorems and techniques apply just as well to $\complex{m}$, the vector space of column vectors.  We only need to use more complicated bases than the standard unit vectors (<acroref type="theorem" acro="SUVB" />) to make things interesting.</p>

<example acro="CBCV" index="change-of-basis!between column vectors">
<title>Change of basis with column vectors</title>

<p>For the vector space $\complex{4}$ we have the two bases,
<alignmath>
<![CDATA[B&=\set{]]>
\colvector{1 \\ -2 \\ 1 \\ -2},\,
\colvector{-1 \\ 3 \\ 1 \\ 1},\,
\colvector{2 \\ -3 \\ 3 \\ -4},\,
\colvector{-1 \\ 3 \\ 3 \\ 0}
}
<![CDATA[&]]>
<![CDATA[C&=\set{]]>
\colvector{1 \\ -6 \\ -4 \\ -1},\,
\colvector{-4 \\ 8 \\ -5 \\ 8},\,
\colvector{-5 \\ 13 \\ -2 \\ 9},\,
\colvector{3 \\ -7 \\ 3 \\ -6}
}
</alignmath>
</p>

<p>The change-of-basis matrix from $B$ to $C$ requires writing each vector of $B$ as a linear combination the vectors in $C$,
<alignmath>
\vectrep{C}{\colvector{1 \\ -2 \\ 1 \\ -2}}
<![CDATA[&=\vectrep{C}{]]>
(1)\colvector{1 \\ -6 \\ -4 \\ -1}+
(-2)\colvector{-4 \\ 8 \\ -5 \\ 8}+
(1)\colvector{-5 \\ 13 \\ -2 \\ 9}+
(-1)\colvector{3 \\ -7 \\ 3 \\ -6}
}
=\colvector{1\\-2\\1\\-1}\\
\vectrep{C}{\colvector{-1 \\ 3 \\ 1 \\ 1}}
<![CDATA[&=\vectrep{C}{]]>
(2)\colvector{1 \\ -6 \\ -4 \\ -1}+
(-3)\colvector{-4 \\ 8 \\ -5 \\ 8}+
(3)\colvector{-5 \\ 13 \\ -2 \\ 9}+
(0)\colvector{3 \\ -7 \\ 3 \\ -6}
}
=\colvector{2\\-3\\3\\0}\\
\vectrep{C}{\colvector{2 \\ -3 \\ 3 \\ -4}}
<![CDATA[&=\vectrep{C}{]]>
(1)\colvector{1 \\ -6 \\ -4 \\ -1}+
(-3)\colvector{-4 \\ 8 \\ -5 \\ 8}+
(1)\colvector{-5 \\ 13 \\ -2 \\ 9}+
(-2)\colvector{3 \\ -7 \\ 3 \\ -6}
}
=\colvector{1\\-3\\1\\-2}\\
\vectrep{C}{\colvector{-1 \\ 3 \\ 3 \\ 0}}
<![CDATA[&=\vectrep{C}{]]>
(2)\colvector{1 \\ -6 \\ -4 \\ -1}+
(-2)\colvector{-4 \\ 8 \\ -5 \\ 8}+
(4)\colvector{-5 \\ 13 \\ -2 \\ 9}+
(3)\colvector{3 \\ -7 \\ 3 \\ -6}
}
=\colvector{2\\-2\\4\\3}\\
</alignmath>
</p>

<p>Then we package these vectors up as the change-of-basis matrix,
<equation>
\cbm{B}{C}=
\begin{bmatrix}
<![CDATA[ 1 & 2 & 1 & 2 \\]]>
<![CDATA[ -2 & -3 & -3 & -2 \\]]>
<![CDATA[ 1 & 3 & 1 & 4 \\]]>
<![CDATA[ -1 & 0 & -2 & 3]]>
\end{bmatrix}
</equation>
</p>

<p>Now consider a single (arbitrary) vector $\vect{y}=\colvector{2\\6\\-3\\4}$.  First, build the vector representation of $\vect{y}$ relative to $B$.  This will require writing $\vect{y}$ as a linear combination of the vectors in $B$,
<alignmath>
\vectrep{B}{\vect{y}}
<![CDATA[&=\vectrep{B}{\colvector{2\\6\\-3\\4}}\\]]>
<![CDATA[&=\vectrep{B}{]]>
(-21)\colvector{1 \\ -2 \\ 1 \\ -2}+
(6)\colvector{-1 \\ 3 \\ 1 \\ 1}+
(11)\colvector{2 \\ -3 \\ 3 \\ -4}+
(-7)\colvector{-1 \\ 3 \\ 3 \\ 0}
}
<![CDATA[&=\colvector{-21\\6\\11\\-7}]]>
</alignmath>
</p>

<p>Now, applying <acroref type="theorem" acro="CB" /> we can convert the representation of $\vect{y}$ relative to $B$ into a representation relative to $C$,
<alignmath>
\vectrep{C}{\vect{y}}
<![CDATA[&=\cbm{B}{C}\vectrep{B}{\vect{y}}&&]]>\text{<acroref type="theorem" acro="CB" />}\\
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[1 & 2 & 1 & 2 \\]]>
<![CDATA[-2 & -3 & -3 & -2 \\]]>
<![CDATA[1 & 3 & 1 & 4 \\]]>
<![CDATA[-1 & 0 & -2 & 3]]>
\end{bmatrix}
\colvector{-21\\6\\11\\-7}\\
<![CDATA[&=\colvector{-12\\5\\-20\\-22}&&]]>\text{<acroref type="definition" acro="MVP" />}
</alignmath>
</p>

<p>We could continue further with this example, perhaps by computing the representation of $\vect{y}$ relative to the basis $C$ directly as a check on our work (<acroref type="exercise" acro="CB.C20" />).  Or we could choose another vector to play the role of $\vect{y}$ and compute two different representations of this vector relative to the two bases $B$ and $C$.</p>

</example>

<sageadvice acro="CBM" index="change-of-basis matrix">
<title>Change-of-Basis Matrix</title>
To create a change-of-basis matrix, it is enough to construct an identity linear transformation relative to a domain and codomain with the specified user bases, which is simply a straight application of <acroref type="definition" acro="CBM" />.  Here we go with two arbitrary bases.
<sage>
<input>b0 = vector(QQ, [-5, 8,  0, 4])
b1 = vector(QQ, [-3, 9, -2, 4])
b2 = vector(QQ, [-1, 4, -1, 2])
b3 = vector(QQ, [-1, 2,  0, 1])
B = [b0, b1, b2, b3]
U = (QQ^4).subspace_with_basis(B)
c0 = vector(QQ, [ 0,  2, -7,  5])
c1 = vector(QQ, [-1,  2, -1,  4])
c2 = vector(QQ, [ 1, -3,  5, -7])
c3 = vector(QQ, [ 1,  1, -8,  3])
C = [c0, c1, c2, c3]
V = (QQ^4).subspace_with_basis(C)
x1, x2, x3, x4 = var('x1, x2, x3, x4')
id_symbolic(x1, x2, x3, x4) = [x1, x2, x3, x4]
S = linear_transformation(U, V, id_symbolic)
CB = S.matrix(side='right')
CB
</input>
<output>[ 36  25   8   7]
[ 27  34  15   7]
[ 35  35  14   8]
[-13  -4   0  -2]
</output>
</sage>

<sage>
<input>S.is_invertible()
</input>
<output>True
</output>
</sage>

We can demonstrate that <code>CB</code> is indeed the change-of-basis matrix from <code>B</code> to <code>C</code>, converting vector representations relative to <code>B</code> into vector representations relative to <code>C</code>.  We choose an arbitrary vector, <code>x</code>, to experiment with (you could experiment with other possibilities).  We use the Sage conveniences to create vector representations relative to the two bases, and then verify <acroref type="theorem" acro="CBM" />.  Recognize that <code>x</code>, <code>u</code> and <code>v</code> are all the same vector.
<sage>
<input>x = vector(QQ, [-45, 62, 171, 85])
u = U.coordinate_vector(x)
u
</input>
<output>(-103, -108, 45, 839)
</output>
</sage>

<sage>
<input>v = V.coordinate_vector(x)
v
</input>
<output>(-175, 95, -43, 93)
</output>
</sage>

<sage>
<input>v == CB*u
</input>
<output>True
</output>
</sage>

We can also verify the construction above by building the change-of-basis matrix directly (i.e., without constructing a linear transformation).
<sage>
<input>cols = [V.coordinate_vector(u) for u in U.basis()]
M = column_matrix(cols)
M
</input>
<output>[ 36  25   8   7]
[ 27  34  15   7]
[ 35  35  14   8]
[-13  -4   0  -2]
</output>
</sage>



</sageadvice>
</subsection>

<subsection acro="MRS">
<title>Matrix Representations and Similarity</title>

<p>Here is the main theorem of this section.  It looks a bit involved at first glance, but the proof should make you realize it is not all that complicated.  In any event, we are more interested in a special case.</p>

<theorem acro="MRCB" index="change-of-basis!matrix representation">
<title>Matrix Representation and Change of Basis</title>
<statement>
<p>Suppose that $\ltdefn{T}{U}{V}$ is a linear transformation, $B$ and $C$ are bases for $U$, and $D$ and $E$ are bases for $V$.  Then
<equation>
\matrixrep{T}{B}{D}=\cbm{E}{D}\matrixrep{T}{C}{E}\cbm{B}{C}
</equation>
</p>

</statement>

<proof>
<p>
<alignmath>
\cbm{E}{D}\matrixrep{T}{C}{E}\cbm{B}{C}
<![CDATA[&=\matrixrep{I_V}{E}{D}\matrixrep{T}{C}{E}\matrixrep{I_U}{B}{C}&&]]>\text{<acroref type="definition" acro="CBM" />}\\
<![CDATA[&=\matrixrep{I_V}{E}{D}\matrixrep{\compose{T}{I_U}}{B}{E}&&]]>\text{<acroref type="theorem" acro="MRCLT" />}\\
<![CDATA[&=\matrixrep{I_V}{E}{D}\matrixrep{T}{B}{E}&&]]>\text{<acroref type="definition" acro="IDLT" />}\\
<![CDATA[&=\matrixrep{\compose{I_V}{T}}{B}{D}&&]]>\text{<acroref type="theorem" acro="MRCLT" />}\\
<![CDATA[&=\matrixrep{T}{B}{D}&&]]>\text{<acroref type="definition" acro="IDLT" />}
</alignmath>
</p>

</proof>
</theorem>

<p>We will be most interested in a special case of this theorem (<acroref type="theorem" acro="SCB" />), but here's an example that illustrates the full generality of <acroref type="theorem" acro="MRCB" />.</p>

<example acro="MRCM" index="matrix representations!converting with change-of-basis">
<title>Matrix representations and change-of-basis matrices</title>

<p>Begin with two vector spaces, $S_2$, the subspace of $M_{22}$ containing all $2\times 2$ symmetric matrices, and $P_3$ (<acroref type="example" acro="VSP" />), the vector space of all polynomials of degree 3 or less.  Then define the linear transformation $\ltdefn{Q}{S_2}{P_3}$ by
<equation>
<![CDATA[\lt{Q}{\begin{bmatrix}a&b\\b&c\end{bmatrix}}]]>
=
(5a-2b+6c)+(3a-b+2c)x+(a+3b-c)x^2+(-4a+2b+c)x^3
</equation>
</p>

<p>Here are two bases for each vector space, one nice, one nasty.  First for $S_2$,
<alignmath>
<![CDATA[B&=]]>
\set{
<![CDATA[\begin{bmatrix}5&-3\\-3&-2\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}2&-3\\-3&0\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}1&2\\2&4\end{bmatrix}]]>
}
<![CDATA[&]]>
<![CDATA[C&=]]>
\set{
<![CDATA[\begin{bmatrix}1&0\\0&0\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}0&1\\1&0\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
</alignmath>
and then for $P_3$,
<alignmath>
<![CDATA[D&=\set{]]>
2+x-2x^2+3x^3,\,
-1-2x^2+3x^3,\,
-3-x+x^3,\,
-x^2+x^3
}\\
<![CDATA[E&=\set{1,\,x,\,x^2,\,x^3}]]>
</alignmath>
</p>

<p>We'll begin with a matrix representation of $Q$ relative to $C$ and $E$.  We first find vector representations of the elements of $C$ relative to $E$,
<alignmath>
<![CDATA[\vectrep{E}{\lt{Q}{\begin{bmatrix}1&0\\0&0\end{bmatrix}}}]]>
<![CDATA[&=\vectrep{E}{5+3x+x^2-4x^3}=\colvector{5\\3\\1\\-4}\\]]>
<![CDATA[\vectrep{E}{\lt{Q}{\begin{bmatrix}0&1\\1&0\end{bmatrix}}}]]>
<![CDATA[&=\vectrep{E}{-2-x+3x^2+2x^3}=\colvector{-2\\-1\\3\\2}\\]]>
<![CDATA[\vectrep{E}{\lt{Q}{\begin{bmatrix}0&0\\0&1\end{bmatrix}}}]]>
<![CDATA[&=\vectrep{E}{6+2x-x^2+x^3}=\colvector{6\\2\\-1\\1}\\]]>
</alignmath>
</p>

<p>So
<alignmath>
\matrixrep{Q}{C}{E}
=
\begin{bmatrix}
<![CDATA[5 & -2 & 6\\]]>
<![CDATA[3 & -1 & 2\\]]>
<![CDATA[1 & 3 & -1\\]]>
<![CDATA[-4 & 2 & 1]]>
\end{bmatrix}
</alignmath>
</p>

<p>Now we construct two change-of-basis matrices.  First, $\cbm{B}{C}$ requires vector representations of the elements of $B$, relative to $C$.  Since $C$ is a nice basis, this is straightforward,
<alignmath>
<![CDATA[\vectrep{C}{\begin{bmatrix}5&-3\\-3&-2\end{bmatrix}}]]>
<![CDATA[&=\vectrep{C}{]]>
<![CDATA[(5)\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[(-3)\begin{bmatrix}0&1\\1&0\end{bmatrix}+]]>
<![CDATA[(-2)\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{5\\-3\\-2}\\
<![CDATA[\vectrep{C}{\begin{bmatrix}2&-3\\-3&0\end{bmatrix}}]]>
<![CDATA[&=\vectrep{C}{]]>
<![CDATA[(2)\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[(-3)\begin{bmatrix}0&1\\1&0\end{bmatrix}+]]>
<![CDATA[(0)\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{2\\-3\\0}\\
<![CDATA[\vectrep{C}{\begin{bmatrix}1&2\\2&4\end{bmatrix}}]]>
<![CDATA[&=\vectrep{C}{]]>
<![CDATA[(1)\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[(2)\begin{bmatrix}0&1\\1&0\end{bmatrix}+]]>
<![CDATA[(4)\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{1\\2\\4}
</alignmath>
</p>

<p>So
<alignmath>
<![CDATA[\cbm{B}{C}&=]]>
\begin{bmatrix}
<![CDATA[5 & 2 & 1\\]]>
<![CDATA[-3 & -3 & 2\\]]>
<![CDATA[-2 & 0 & 4]]>
\end{bmatrix}
</alignmath>
</p>

<p>The other change-of-basis matrix we'll compute is $\cbm{E}{D}$.  However, since
$E$ is a nice basis (and $D$ is not) we'll turn it around and instead compute $\cbm{D}{E}$ and apply <acroref type="theorem" acro="ICBM" /> to use an inverse to compute $\cbm{E}{D}$.
<alignmath>
\vectrep{E}{2+x-2x^2+3x^3}
<![CDATA[&=\vectrep{E}{(2)1+(1)x+(-2)x^2+(3)x^3}]]>
=\colvector{2\\1\\-2\\3}\\
\vectrep{E}{-1-2x^2+3x^3}
<![CDATA[&=\vectrep{E}{(-1)1+(0)x+(-2)x^2+(3)x^3}]]>
=\colvector{-1\\0\\-2\\3}\\
\vectrep{E}{-3-x+x^3}
<![CDATA[&=\vectrep{E}{(-3)1+(-1)x+(0)x^2+(1)x^3}]]>
=\colvector{-3\\-1\\0\\1}\\
\vectrep{E}{-x^2+x^3}
<![CDATA[&=\vectrep{E}{(0)1+(0)x+(-1)x^2+(1)x^3}]]>
=\colvector{0\\0\\-1\\1}
</alignmath>
</p>

<p>So, we can package these column vectors up as a matrix to obtain $\cbm{D}{E}$ and then,
<alignmath>
\cbm{E}{D}
<![CDATA[&=\inverse{\left(\cbm{D}{E}\right)}]]>
<![CDATA[&&]]>\text{<acroref type="theorem" acro="ICBM" />}\\
<![CDATA[&=\inverse{]]>
\begin{bmatrix}
<![CDATA[ 2 & -1 & -3 & 0 \\]]>
<![CDATA[ 1 & 0 & -1 & 0 \\]]>
<![CDATA[ -2 & -2 & 0 & -1 \\]]>
<![CDATA[ 3 & 3 & 1 & 1]]>
\end{bmatrix}
}\\
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[ 1 & -2 & 1 & 1 \\]]>
<![CDATA[ -2 & 5 & -1 & -1 \\]]>
<![CDATA[ 1 & -3 & 1 & 1 \\]]>
<![CDATA[ 2 & -6 & -1 & 0]]>
\end{bmatrix}
</alignmath>
</p>

<p>We are now in a position to apply <acroref type="theorem" acro="MRCB" />.  The matrix representation of $Q$ relative to $B$ and $D$ can be obtained as follows,
<alignmath>
\matrixrep{Q}{B}{D}
<![CDATA[&=\cbm{E}{D}\matrixrep{Q}{C}{E}\cbm{B}{C}]]>
<![CDATA[&&]]>\text{<acroref type="theorem" acro="MRCB" />}\\
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[ 1 & -2 & 1 & 1 \\]]>
<![CDATA[ -2 & 5 & -1 & -1 \\]]>
<![CDATA[ 1 & -3 & 1 & 1 \\]]>
<![CDATA[ 2 & -6 & -1 & 0]]>
\end{bmatrix}
\begin{bmatrix}
<![CDATA[5 & -2 & 6\\]]>
<![CDATA[3 & -1 & 2\\]]>
<![CDATA[1 & 3 & -1\\]]>
<![CDATA[-4 & 2 & 1]]>
\end{bmatrix}
\begin{bmatrix}
<![CDATA[5 & 2 & 1\\]]>
<![CDATA[-3 & -3 & 2\\]]>
<![CDATA[-2 & 0 & 4]]>
\end{bmatrix}\\
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[ 1 & -2 & 1 & 1 \\]]>
<![CDATA[ -2 & 5 & -1 & -1 \\]]>
<![CDATA[ 1 & -3 & 1 & 1 \\]]>
<![CDATA[ 2 & -6 & -1 & 0]]>
\end{bmatrix}
\begin{bmatrix}
<![CDATA[ 19 & 16 & 25 \\]]>
<![CDATA[ 14 & 9 & 9 \\]]>
<![CDATA[ -2 & -7 & 3 \\]]>
<![CDATA[ -28 & -14 & 4]]>
\end{bmatrix}\\
<![CDATA[&=]]>
\begin{bmatrix}
<![CDATA[ -39 & -23 & 14 \\]]>
<![CDATA[ 62 & 34 & -12 \\]]>
<![CDATA[ -53 & -32 & 5 \\]]>
<![CDATA[ -44 & -15 & -7]]>
\end{bmatrix}
</alignmath>
</p>

<p>Now check our work by computing $\matrixrep{Q}{B}{D}$ directly (<acroref type="exercise" acro="CB.C21" />).</p>

</example>

<p>Here is a special case of the previous theorem, where we choose $U$ and $V$ to be the same vector space, so the matrix representations and the change-of-basis matrices are all square of the same size.</p>

<theorem acro="SCB" index="change-of-basis!similarity">
<title>Similarity and Change of Basis</title>
<statement>
<p>Suppose that $\ltdefn{T}{V}{V}$ is a linear transformation and $B$ and $C$ are bases of $V$.  Then
<equation>
\matrixrep{T}{B}{B}=\inverse{\cbm{B}{C}}\matrixrep{T}{C}{C}\cbm{B}{C}
</equation>
</p>

</statement>

<proof>
<p>In the conclusion of <acroref type="theorem" acro="MRCB" />, replace $D$ by $B$, and replace $E$ by $C$,
<alignmath>
\matrixrep{T}{B}{B}
<![CDATA[&=\cbm{C}{B}\matrixrep{T}{C}{C}\cbm{B}{C}&&]]>\text{<acroref type="theorem" acro="MRCB" />}\\
<![CDATA[&=\inverse{\cbm{B}{C}}\matrixrep{T}{C}{C}\cbm{B}{C}&&]]>\text{<acroref type="theorem" acro="ICBM" />}
</alignmath>
</p>

</proof>
</theorem>

<p>This is the third surprise of this chapter.  <acroref type="theorem" acro="SCB" /> considers the special case where a linear transformation has the same vector space for the domain and codomain ($V$).  We build a matrix representation of $T$ using the basis $B$ simultaneously for both the domain and codomain ($\matrixrep{T}{B}{B}$), and then we build a second matrix representation of $T$, now using the basis $C$ for both the domain and codomain ($\matrixrep{T}{C}{C}$).  Then these two representations are related via a similarity transformation (<acroref type="definition" acro="SIM" />) using a change-of-basis matrix ($\cbm{B}{C}$)!</p>

<example acro="MRBE" index="matrix representation!basis of eigenvectors">
<title>Matrix representation with basis of eigenvectors</title>

<p>We return to the linear transformation $\ltdefn{T}{M_{22}}{M_{22}}$ of <acroref type="example" acro="ELTBM" /> defined by
<equation>
<![CDATA[\lt{T}{\begin{bmatrix}a&b\\c&d\end{bmatrix}}]]>
=
\begin{bmatrix}
-17a+11b+8c-11d
<![CDATA[&]]>
-57a+35b+24c-33d
\\
-14a+10b+6c-10d
<![CDATA[&]]>
-41a+25b+16c-23d
\end{bmatrix}
</equation>
</p>

<p>In <acroref type="example" acro="ELTBM" /> we showcased four eigenvectors of $T$.  We will now put these four vectors in a set,
<equation>
B=\set{\vect{x}_1,\,\vect{x}_2,\,\vect{x}_3,\,\vect{x}_4}
=\set{
\begin{bmatrix}
<![CDATA[ 0 & 1 \\ 0 & 1]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 1 & 1 \\ 1 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 1 & 3 \\ 2 & 3]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 2 & 6 \\ 1 & 4]]>
\end{bmatrix}
}
</equation>
</p>

<p>Check that $B$ is a basis of $M_{22}$ by first establishing the linear independence of $B$ and then employing <acroref type="theorem" acro="G" /> to get the spanning property easily.  Here is a second set of $2\times 2$ matrices, which also forms a basis of $M_{22}$ (<acroref type="example" acro="BM" />),
<equation>
C=\set{\vect{y}_1,\,\vect{y}_2,\,\vect{y}_3,\,\vect{y}_4}
=\set{
\begin{bmatrix}
<![CDATA[ 1 & 0 \\ 0 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 0 & 1 \\ 0 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 0 & 0 \\ 1 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 0 & 0 \\ 0 & 1]]>
\end{bmatrix}
}
</equation>
</p>

<p>We can build two matrix representations of $T$, one relative to $B$ and one relative to $C$.  Each is easy, but for wildly different reasons.  In our computation of the matrix representation relative to $B$ we borrow some of our work in <acroref type="example" acro="ELTBM" />.  Here are the representations, then the explanation.
<alignmath>
\vectrep{B}{\lt{T}{\vect{x}_1}}
<![CDATA[&=]]>
\vectrep{B}{2\vect{x}_1}
=\vectrep{B}{2\vect{x}_1+0\vect{x}_2+0\vect{x}_3+0\vect{x}_4}
=\colvector{2\\0\\0\\0}\\
\vectrep{B}{\lt{T}{\vect{x}_2}}
<![CDATA[&=]]>
\vectrep{B}{2\vect{x}_2}
=\vectrep{B}{0\vect{x}_1+2\vect{x}_2+0\vect{x}_3+0\vect{x}_4}
=\colvector{0\\2\\0\\0}\\
\vectrep{B}{\lt{T}{\vect{x}_3}}
<![CDATA[&=]]>
\vectrep{B}{(-1)\vect{x}_3}
=\vectrep{B}{0\vect{x}_1+0\vect{x}_2+(-1)\vect{x}_3+0\vect{x}_4}
=\colvector{0\\0\\-1\\0}\\
\vectrep{B}{\lt{T}{\vect{x}_4}}
<![CDATA[&=]]>
\vectrep{B}{(-2)\vect{x}_4}
=\vectrep{B}{0\vect{x}_1+0\vect{x}_2+0\vect{x}_3+(-2)\vect{x}_4}
=\colvector{0\\0\\0\\-2}
</alignmath>
</p>

<p>So the resulting representation is
<alignmath>
\matrixrep{T}{B}{B}
=
\begin{bmatrix}
<![CDATA[2 & 0 & 0 & 0\\]]>
<![CDATA[0 & 2 & 0 & 0\\]]>
<![CDATA[0 & 0 & -1 & 0\\]]>
<![CDATA[0 & 0 & 0 & -2\\]]>
\end{bmatrix}
</alignmath></p>

<p>Very pretty.</p>

<p>Now for the matrix representation relative to $C$ first compute,
<alignmath>
<![CDATA[&\vectrep{C}{\lt{T}{\vect{y}_1}}]]>
<![CDATA[=\vectrep{C}{\begin{bmatrix}-17&-57\\-14&-41\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{C}{]]>
<![CDATA[(-17)\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[(-57)\begin{bmatrix}0&1\\0&0\end{bmatrix}+]]>
<![CDATA[(-14)\begin{bmatrix}0&0\\1&0\end{bmatrix}+]]>
<![CDATA[(-41)\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{-17\\-57\\-14\\-41}\\
<![CDATA[&\vectrep{C}{\lt{T}{\vect{y}_2}}]]>
<![CDATA[=\vectrep{C}{\begin{bmatrix}11&35\\10&25\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{C}{]]>
<![CDATA[11\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[35\begin{bmatrix}0&1\\0&0\end{bmatrix}+]]>
<![CDATA[10\begin{bmatrix}0&0\\1&0\end{bmatrix}+]]>
<![CDATA[25\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{11\\35\\10\\25}\\
<![CDATA[&\vectrep{C}{\lt{T}{\vect{y}_3}}]]>
<![CDATA[=\vectrep{C}{\begin{bmatrix}8&24\\6&16\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{C}{]]>
<![CDATA[8\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[24\begin{bmatrix}0&1\\0&0\end{bmatrix}+]]>
<![CDATA[6\begin{bmatrix}0&0\\1&0\end{bmatrix}+]]>
<![CDATA[16\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{8\\24\\6\\16}\\
<![CDATA[&\vectrep{C}{\lt{T}{\vect{y}_4}}]]>
<![CDATA[=\vectrep{C}{\begin{bmatrix}-11&-33\\-10&-23\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{C}{]]>
<![CDATA[(-11)\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[(-33)\begin{bmatrix}0&1\\0&0\end{bmatrix}+]]>
<![CDATA[(-10)\begin{bmatrix}0&0\\1&0\end{bmatrix}+]]>
<![CDATA[(-23)\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{-11\\-33\\-10\\-23}\\
</alignmath>
</p>

<p>So the resulting representation is
<alignmath>
\matrixrep{T}{C}{C}
=
\begin{bmatrix}
<![CDATA[ -17 & 11 & 8 & -11 \\]]>
<![CDATA[ -57 & 35 & 24 & -33 \\]]>
<![CDATA[ -14 & 10 & 6 & -10 \\]]>
<![CDATA[ -41 & 25 & 16 & -23]]>
\end{bmatrix}
</alignmath>
</p>

<p>Not quite as pretty.</p>

<p>The purpose of this example is to illustrate <acroref type="theorem" acro="SCB" />.  This theorem says that the two matrix representations, $\matrixrep{T}{B}{B}$ and $\matrixrep{T}{C}{C}$, of the one linear transformation, $T$, are related by a similarity transformation using the change-of-basis matrix $\cbm{B}{C}$.  Lets compute this change-of-basis matrix.  Notice that since $C$ is such a nice basis, this is fairly straightforward,
<alignmath>
\vectrep{C}{\vect{x}_1}
<![CDATA[&=\vectrep{C}{\begin{bmatrix}0 & 1 \\ 0 & 1\end{bmatrix}}]]>
=\vectrep{C}{
<![CDATA[0\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[1\begin{bmatrix}0&1\\0&0\end{bmatrix}+]]>
<![CDATA[0\begin{bmatrix}0&0\\1&0\end{bmatrix}+]]>
<![CDATA[1\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{0\\1\\0\\1}\\
\vectrep{C}{\vect{x}_2}
<![CDATA[&=\vectrep{C}{\begin{bmatrix}1 & 1 \\ 1 & 0\end{bmatrix}}]]>
=\vectrep{C}{
<![CDATA[1\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[1\begin{bmatrix}0&1\\0&0\end{bmatrix}+]]>
<![CDATA[1\begin{bmatrix}0&0\\1&0\end{bmatrix}+]]>
<![CDATA[0\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{1\\1\\1\\0}\\
\vectrep{C}{\vect{x}_3}
<![CDATA[&=\vectrep{C}{\begin{bmatrix}1 & 3 \\ 2 & 3\end{bmatrix}}]]>
=\vectrep{C}{
<![CDATA[1\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[3\begin{bmatrix}0&1\\0&0\end{bmatrix}+]]>
<![CDATA[2\begin{bmatrix}0&0\\1&0\end{bmatrix}+]]>
<![CDATA[3\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{1\\3\\2\\3}\\
\vectrep{C}{\vect{x}_4}
<![CDATA[&=\vectrep{C}{\begin{bmatrix}2 & 6 \\ 1 & 4\end{bmatrix}}]]>
=\vectrep{C}{
<![CDATA[2\begin{bmatrix}1&0\\0&0\end{bmatrix}+]]>
<![CDATA[6\begin{bmatrix}0&1\\0&0\end{bmatrix}+]]>
<![CDATA[1\begin{bmatrix}0&0\\1&0\end{bmatrix}+]]>
<![CDATA[4\begin{bmatrix}0&0\\0&1\end{bmatrix}]]>
}
=\colvector{2\\6\\1\\4}
</alignmath>
</p>

<p>So we have,
<equation>
\cbm{B}{C}
=
\begin{bmatrix}
<![CDATA[ 0 & 1 & 1 & 2 \\]]>
<![CDATA[ 1 & 1 & 3 & 6 \\]]>
<![CDATA[ 0 & 1 & 2 & 1 \\]]>
<![CDATA[ 1 & 0 & 3 & 4]]>
\end{bmatrix}
</equation>
</p>

<p>Now, according to <acroref type="theorem" acro="SCB" /> we can write,
<alignmath>
<![CDATA[\matrixrep{T}{B}{B}&=\inverse{\cbm{B}{C}}\matrixrep{T}{C}{C}\cbm{B}{C}\\]]>
\begin{bmatrix}
<![CDATA[2 & 0 & 0 & 0\\]]>
<![CDATA[0 & 2 & 0 & 0\\]]>
<![CDATA[0 & 0 & -1 & 0\\]]>
<![CDATA[0 & 0 & 0 & -2\\]]>
\end{bmatrix}
<![CDATA[&=]]>
\inverse{
\begin{bmatrix}
<![CDATA[ 0 & 1 & 1 & 2 \\]]>
<![CDATA[ 1 & 1 & 3 & 6 \\]]>
<![CDATA[ 0 & 1 & 2 & 1 \\]]>
<![CDATA[ 1 & 0 & 3 & 4]]>
\end{bmatrix}
}
\begin{bmatrix}
<![CDATA[ -17 & 11 & 8 & -11 \\]]>
<![CDATA[ -57 & 35 & 24 & -33 \\]]>
<![CDATA[ -14 & 10 & 6 & -10 \\]]>
<![CDATA[ -41 & 25 & 16 & -23]]>
\end{bmatrix}
\begin{bmatrix}
<![CDATA[ 0 & 1 & 1 & 2 \\]]>
<![CDATA[ 1 & 1 & 3 & 6 \\]]>
<![CDATA[ 0 & 1 & 2 & 1 \\]]>
<![CDATA[ 1 & 0 & 3 & 4]]>
\end{bmatrix}
</alignmath>
</p>

<p>This should look and feel exactly like the process for diagonalizing a matrix, as was described in <acroref type="section" acro="SD" />.  And it is.</p>

</example>

<sageadvice acro="MRCB" index="matrix representation!change-of-basis">
<title>Matrix Representation and Change-of-Basis</title>
In <acroref type="sage" acro="MR" /> we built two matrix representations of one linear transformation, relative to two different pairs of bases.  We now understand how these two matrix representations are related <mdash /> <acroref type="theorem" acro="MRCB" /> gives the precise relationship with change-of-basis matrices, one converting vector representations on the domain, the other converting vector representations on the codomain.  Here is the demonstration.  We use <code>MT</code> as the prefix of names for matrix representations, <code>CB</code> as the prefix for change-of-basis matrices, and numerals to distinguish the two domain-codomain pairs.
<sage>
<input>x1, x2, x3, x4 = var('x1, x2, x3, x4')
outputs = [3*x1 + 7*x2 + x3 - 4*x4,
           2*x1 + 5*x2 + x3 - 3*x4,
            -x1 - 2*x2      +   x4]
T_symbolic(x1, x2, x3, x4) = outputs
U = QQ^4
V = QQ^3
b0 = vector(QQ, [ 1, 1, -1, 0])
b1 = vector(QQ, [-1, 0, -2, 7])
b2 = vector(QQ, [ 0, 1, -2, 4])
b3 = vector(QQ, [-2, 0, -1, 6])
B = [b0, b1, b2, b3]
c0 = vector(QQ, [ 1,  6, -6])
c1 = vector(QQ, [ 0,  1, -1])
c2 = vector(QQ, [-2, -3,  4])
C = [c0, c1, c2]
d0 = vector(QQ, [ 1, -3,  2, -1])
d1 = vector(QQ, [ 0,  1,  0,  1])
d2 = vector(QQ, [-1,  2, -1, -1])
d3 = vector(QQ, [ 2, -8,  4, -3])
D = [d0, d1, d2, d3]
e0 = vector(QQ, [ 0,  1, -3])
e1 = vector(QQ, [-1,  2, -1])
e2 = vector(QQ, [ 2, -4,  3])
E = [e0, e1, e2]
U1 = U.subspace_with_basis(B)
V1 = V.subspace_with_basis(C)
T1 = linear_transformation(U1, V1, T_symbolic)
MTBC =T1.matrix(side='right')
MTBC
</input>
<output>[ 15 -67 -25 -61]
[-75 326 120 298]
[  3 -17  -7 -15]
</output>
</sage>

<sage>
<input>U2 = U.subspace_with_basis(D)
V2 = V.subspace_with_basis(E)
T2 = linear_transformation(U2, V2, T_symbolic)
MTDE = T2.matrix(side='right')
MTDE
</input>
<output>[ -32    8   38  -91]
[-148   37  178 -422]
[ -80   20   96 -228]
</output>
</sage>

This is as far as we could go back in <acroref type="section" acro="MR" />.  These two matrices represent the same linear transformation (namely <code>T_symbolic</code>), but the question now is <q>how are these representations related?</q>  We need two change-of-basis matrices.  Notice that with different dimensions for the domain and codomain, we get square matrices of different sizes.
<sage>
<input>identity4(x1, x2, x3, x4) = [x1, x2, x3, x4]
CU = linear_transformation(U2, U1, identity4)
CBDB = CU.matrix(side='right')
CBDB
</input>
<output>[ 6  7 -8  1]
[ 5  1 -5  9]
[-9 -6 10 -9]
[ 0  3 -1 -5]
</output>
</sage>

<sage>
<input>identity3(x1, x2, x3) = [x1, x2, x3]
CV = linear_transformation(V1, V2, identity3)
CBCE = CV.matrix(side='right')
CBCE
</input>
<output>[  8   1  -7]
[ 33   4 -28]
[ 17   2 -15]
</output>
</sage>

Finally, here is <acroref type="theorem" acro="MRCB" />, relating the the two matrix representations via the change-of-basis matrices.
<sage>
<input>MTDE == CBCE * MTBC * CBDB
</input>
<output>True
</output>
</sage>

We can walk through this theorem just a bit more carefully, step-by-step.  We will compute three matrix-vector products, using three vector representations, to demonstrate the equality above.  To prepare, we choose the vector <code>x</code> arbitrarily, and we compute its value when evaluted by <code>T_symbolic</code>, and then verify the vector and matrix representations relative to <code>D</code> and <code>E</code>.
<sage>
<input>T_symbolic(34, -61, 55, 18)
</input>
<output>(-342, -236, 106)
</output>
</sage>

<sage>
<input>x = vector(QQ, [34, -61, 55, 18])
u_D = U2.coordinate_vector(x)
u_D
</input>
<output>(25, 24, -13, -2)
</output>
</sage>

<sage>
<input>v_E = V2.coordinate_vector(vector(QQ, [-342, -236, 106]))
v_E
</input>
<output>(-920, -4282, -2312)
</output>
</sage>

<sage>
<input>v_E == MTDE*u_D
</input>
<output>True
</output>
</sage>

So far this is not really new, we have just verified the representation <code>MTDE</code> in the case of one input vector (<code>x</code>), but now we will use the alternate version of this matrix representation, <code>CBCE * MTBC * CBDB</code>, in steps.<br /><br />
First, convert the input vector from a representation relative to <code>D</code> to a representation relative to <code>B</code>.
<sage>
<input>u_B = CBDB*u_D
u_B
</input>
<output>(420, 196, -481, 95)
</output>
</sage>

Now apply the matrix representation, which expects <q>input</q> coordinatized relative to <code>B</code> and produces <q>output</q> coordinatized relative to <code>C</code>.
<sage>
<input>v_C = MTBC*u_B
v_C
</input>
<output>(-602, 2986, -130)
</output>
</sage>

Now convert the output vector from a representation relative to <code>C</code> to a representation relative to <code>E</code>.
<sage>
<input>v_E = CBCE*v_C
v_E
</input>
<output>(-920, -4282, -2312)
</output>
</sage>

It is no surprise that this version of <code>v_E</code> equals the previous one, since we have checked the equality of the matrices earlier.  But it may be instructive to see the input converted by change-of-basis matrices before and after being hit by the linear transformation (as a matrix representation).
Now we will perform another example, but this time using Sage endomorphisms, linear transformations with equal bases for the domain and codomain.  This will allow us to illustrate <acroref type="theorem" acro="SCB" />.  Just for fun, we will do something large. Notice the labor-saving device for manufacturing many symbolic variables at once.
<sage>
<input>[var('x{0}'.format(i)) for i in range(1, 12)]
</input>
<output>[x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11]
</output>
</sage>

<sage>
<input>x = vector(SR, [x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11])
A = matrix(QQ, 11, [
  [ 146, -225, -10,  212,  419, -123, -73,  3,  219, -100, -57],
  [ -24,   32,   1,  -33,  -66,   13,  16,  1,  -33,   18,   3],
  [  79, -131, -15,  124,  235,  -74, -33, -3,  128,  -57, -29],
  [  -1,   13, -16,   -1,  -27,    3,  -5, -4,   -9,    6,   2],
  [-104,  170,  20, -162, -307,   95,  45,  3, -167,   75,  37],
  [ -16,   59, -19,  -34, -103,   27, -10, -1,  -51,   27,   8],
  [  36,  -41,  -7,   46,   80,  -25, -26,  2,   42,  -18, -16],
  [  -5,    0,   1,   -4,   -3,    2,   6, -1,    0,   -2,   3],
  [ 105, -176, -28,  168,  310, -103, -41, -4,  172,  -73, -40],
  [   1,    7,   0,   -3,   -9,    5,  -6, -2,   -7,    3,   2],
  [  74, -141,   4,  118,  255,  -72, -23, -1,  133,  -63, -26]
                   ])
out = (A*x).list()
T_symbolic(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11) = out
V1 = QQ^11
C = V1.basis()
T1 = linear_transformation(V1, V1, T_symbolic)
MC = T1.matrix(side='right')
MC
</input>
<output>[ 146 -225  -10  212  419 -123  -73    3  219 -100  -57]
[ -24   32    1  -33  -66   13   16    1  -33   18    3]
[  79 -131  -15  124  235  -74  -33   -3  128  -57  -29]
[  -1   13  -16   -1  -27    3   -5   -4   -9    6    2]
[-104  170   20 -162 -307   95   45    3 -167   75   37]
[ -16   59  -19  -34 -103   27  -10   -1  -51   27    8]
[  36  -41   -7   46   80  -25  -26    2   42  -18  -16]
[  -5    0    1   -4   -3    2    6   -1    0   -2    3]
[ 105 -176  -28  168  310 -103  -41   -4  172  -73  -40]
[   1    7    0   -3   -9    5   -6   -2   -7    3    2]
[  74 -141    4  118  255  -72  -23   -1  133  -63  -26]
</output>
</sage>

Not very interesting, and perhaps even transparent, with a definiton from a matrix and with the standard basis attached to <code>V1 == QQ^11</code>.  Let's use a different basis to obtain a more interesting representation.  We will input the basis compactly as the columns of a nonsingular matrix.
<sage>
<input>D = matrix(QQ, 11,
    [[ 1,  2, -1, -2,  4,  2,  2, -2,  4,  4,  8],
     [ 0,  1,  0,  2, -2,  1,  1, -1, -7,  5,  3],
     [ 1,  0,  0, -2,  3,  0, -1, -1,  6, -1, -1],
     [ 0, -1,  1, -1,  3, -2, -3,  0,  5, -8,  2],
     [-1,  0,  0,  3, -4,  0,  1,  1, -8,  1,  2],
     [-1, -1,  1,  0,  3, -3, -4, -1,  0, -7,  3],
     [ 0,  1,  0,  0,  2,  0,  0, -1,  0, -1,  8],
     [ 0,  0,  0, -1,  0,  0,  0,  1,  5, -4,  1],
     [ 1,  0,  0, -2,  3,  0, -2, -3,  3,  3, -4],
     [ 0, -1,  0,  0,  1, -1, -2, -1,  2, -4,  0],
     [ 1,  0, -1, -2,  0,  2,  2,  0,  5,  3, -1]])
E = D.columns()
V2 = (QQ^11).subspace_with_basis(E)
T2 = linear_transformation(V2, V2, T_symbolic)
MB = T2.matrix(side='right')
MB
</input>
<output>[ 2  1  0  0  0  0  0  0  0  0  0]
[ 0  2  1  0  0  0  0  0  0  0  0]
[ 0  0  2  0  0  0  0  0  0  0  0]
[ 0  0  0  2  1  0  0  0  0  0  0]
[ 0  0  0  0  2  0  0  0  0  0  0]
[ 0  0  0  0  0 -1  1  0  0  0  0]
[ 0  0  0  0  0  0 -1  1  0  0  0]
[ 0  0  0  0  0  0  0 -1  1  0  0]
[ 0  0  0  0  0  0  0  0 -1  1  0]
[ 0  0  0  0  0  0  0  0  0 -1  1]
[ 0  0  0  0  0  0  0  0  0  0 -1]
</output>
</sage>

Well, now <em>that</em> is interesting!  What a nice representation.  Of course, it is all due to the choice of the basis (which we have not explained).  To explain the relationship between the two matrix representations, we need a change-of-basis-matrix, and its inverse.  <acroref type="theorem" acro="SCB" /> says we need the matrix that converts vector representations relative to <code>B</code> into vector representations relative to <code>C</code>.
<sage>
<input>out = [x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11]
id11(x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11) = out
L = linear_transformation(V2, V1, id11)
CB = L.matrix(side='right')
CB
</input>
<output>[ 1  2 -1 -2  4  2  2 -2  4  4  8]
[ 0  1  0  2 -2  1  1 -1 -7  5  3]
[ 1  0  0 -2  3  0 -1 -1  6 -1 -1]
[ 0 -1  1 -1  3 -2 -3  0  5 -8  2]
[-1  0  0  3 -4  0  1  1 -8  1  2]
[-1 -1  1  0  3 -3 -4 -1  0 -7  3]
[ 0  1  0  0  2  0  0 -1  0 -1  8]
[ 0  0  0 -1  0  0  0  1  5 -4  1]
[ 1  0  0 -2  3  0 -2 -3  3  3 -4]
[ 0 -1  0  0  1 -1 -2 -1  2 -4  0]
[ 1  0 -1 -2  0  2  2  0  5  3 -1]
</output>
</sage>

OK, all set.
<sage>
<input>CB^-1*MC*CB
</input>
<output>[ 2  1  0  0  0  0  0  0  0  0  0]
[ 0  2  1  0  0  0  0  0  0  0  0]
[ 0  0  2  0  0  0  0  0  0  0  0]
[ 0  0  0  2  1  0  0  0  0  0  0]
[ 0  0  0  0  2  0  0  0  0  0  0]
[ 0  0  0  0  0 -1  1  0  0  0  0]
[ 0  0  0  0  0  0 -1  1  0  0  0]
[ 0  0  0  0  0  0  0 -1  1  0  0]
[ 0  0  0  0  0  0  0  0 -1  1  0]
[ 0  0  0  0  0  0  0  0  0 -1  1]
[ 0  0  0  0  0  0  0  0  0  0 -1]
</output>
</sage>

Which is <code>MB</code>.  So the conversion from a <q>messy</q> matrix representation relative to a standard basis to a <q>clean</q> representation relative to some other basis is just a similarity transformation by a change-of-basis matrix.  Oh, I almost forgot.  Where did that basis come from?  Hint: work your way up to <acroref type="section" acro="JCF" />.


</sageadvice>
<p>We can now return to the question of computing an eigenvalue or eigenvector of a linear transformation.  For a linear transformation of the form $\ltdefn{T}{V}{V}$, we know that representations relative to different bases are similar matrices.  We also know that similar matrices have equal characteristic polynomials by <acroref type="theorem" acro="SMEE" />.   We will now show that eigenvalues of a linear transformation $T$ are precisely the eigenvalues of <em>any</em> matrix representation of $T$.  Since the choice of a different matrix representation leads to a similar matrix, there will be no <q>new</q> eigenvalues obtained from this second representation.  Similarly, the change-of-basis matrix can be used to show that eigenvectors obtained from one matrix representation will be precisely those obtained from any other representation.  So we can determine the eigenvalues and eigenvectors of a linear transformation by forming one matrix representation, using <em>any</em> basis we please, and analyzing the matrix in the manner of <acroref type="chapter" acro="E" />.</p>

<theorem acro="EER" index="eigenvalues, eigenvectors!vector, matrix representations">
<title>Eigenvalues, Eigenvectors, Representations</title>
<statement>
<p>Suppose that $\ltdefn{T}{V}{V}$ is a linear transformation and $B$ is a basis of $V$.  Then $\vect{v}\in V$ is an eigenvector of $T$ for the eigenvalue $\lambda$ if and only if $\vectrep{B}{\vect{v}}$ is an eigenvector of $\matrixrep{T}{B}{B}$ for the eigenvalue $\lambda$.</p>

</statement>

<proof>
<p><implyforward /> Assume that $\vect{v}\in V$ is an eigenvector of $T$ for the eigenvalue $\lambda$.  Then
<alignmath>
\matrixrep{T}{B}{B}\vectrep{B}{\vect{v}}
<![CDATA[&=\vectrep{B}{\lt{T}{\vect{v}}}&&]]>\text{<acroref type="theorem" acro="FTMR" />}\\
<![CDATA[&=\vectrep{B}{\lambda\vect{v}}&&]]>\text{<acroref type="definition" acro="EELT" />}\\
<![CDATA[&=\lambda\vectrep{B}{\vect{v}}&&]]>\text{<acroref type="theorem" acro="VRLT" />}
</alignmath>
which by <acroref type="definition" acro="EEM" /> says that $\vectrep{B}{\vect{v}}$ is an eigenvector of the matrix $\matrixrep{T}{B}{B}$ for the eigenvalue $\lambda$.</p>

<p><implyreverse />  Assume that $\vectrep{B}{\vect{v}}$ is an eigenvector of $\matrixrep{T}{B}{B}$ for the eigenvalue $\lambda$.  Then
<alignmath>
\lt{T}{\vect{v}}
<![CDATA[&=\vectrepinv{B}{\vectrep{B}{\lt{T}{\vect{v}}}}&&]]>\text{<acroref type="definition" acro="IVLT" />}\\
<![CDATA[&=\vectrepinv{B}{\matrixrep{T}{B}{B}\vectrep{B}{\vect{v}}}&&]]>\text{<acroref type="theorem" acro="FTMR" />}\\
<![CDATA[&=\vectrepinv{B}{\lambda\vectrep{B}{\vect{v}}}&&]]>\text{<acroref type="definition" acro="EEM" />}\\
<![CDATA[&=\lambda\vectrepinv{B}{\vectrep{B}{\vect{v}}}&&]]>\text{<acroref type="theorem" acro="ILTLT" />}\\
<![CDATA[&=\lambda\vect{v}&&]]>\text{<acroref type="definition" acro="IVLT" />}
</alignmath>
which by <acroref type="definition" acro="EELT" /> says $\vect{v}$ is an eigenvector of $T$ for the eigenvalue $\lambda$.</p>

</proof>
</theorem>

</subsection>

<subsection acro="CELT">
<title>Computing Eigenvectors of Linear Transformations</title>

<p><acroref type="theorem" acro="EER" /> tells us that the eigenvalues of a linear transformation are the eigenvalues of <em>any</em> representation, no matter what the choice of the basis $B$ might be.  So we could now unambiguously define items such as the characteristic polynomial of a linear transformation, which we would define as the characteristic polynomial of any matrix representation.  We'll say that again <mdash /> eigenvalues, eigenvectors, and characteristic polynomials are intrinsic properties of a linear transformation, independent of the choice of a basis used to construct a matrix representation.</p>

<p>As a practical matter, how does one compute the eigenvalues and eigenvectors of a linear transformation of the form $\ltdefn{T}{V}{V}$?  Choose a nice basis $B$ for $V$, one where the vector representations of the values of the linear transformations necessary for the matrix representation are easy to compute.  Construct the matrix representation relative to this basis, and find the eigenvalues and eigenvectors of this matrix using the techniques of <acroref type="chapter" acro="E" />.  The resulting eigenvalues of the matrix are precisely the eigenvalues of the linear transformation.  The eigenvectors of the matrix are column vectors that need to be converted to vectors in $V$ through application of $\ltinverse{\vectrepname{B}}$ (this is part of the content of <acroref type="theorem" acro="EER" />).</p>

<p>Now consider the case where the matrix representation of a linear transformation is diagonalizable.  The $n$ linearly independent eigenvectors that must exist for the matrix (<acroref type="theorem" acro="DC" />) can be converted (via $\ltinverse{\vectrepname{B}}$) into eigenvectors of the linear transformation.  A matrix representation of the linear transformation relative to a basis of eigenvectors will be a diagonal matrix <mdash /> an especially nice representation!  Though we did not know it at the time, the diagonalizations of <acroref type="section" acro="SD" /> were really about finding especially pleasing matrix representations of linear transformations.</p>

<p>Here are some examples.</p>

<example acro="ELTT" index="eigenvectors!of a linear transformation">
<title>Eigenvectors of a linear transformation, twice</title>

<p>Consider the linear transformation $\ltdefn{S}{M_{22}}{M_{22}}$ defined by
<equation>
<![CDATA[\lt{S}{\begin{bmatrix}a&b\\c&d\end{bmatrix}}=]]>
\begin{bmatrix}
<![CDATA[-b - c - 3d & -14a - 15b - 13c + d\\]]>
<![CDATA[18a + 21b + 19c + 3d &  -6a - 7b - 7c - 3d]]>
\end{bmatrix}
</equation>
</p>

<p>To find the eigenvalues and eigenvectors of $S$ we will build a matrix representation and analyze the matrix.  Since <acroref type="theorem" acro="EER" /> places no restriction on the choice of the basis $B$, we may as well use a basis that is easy to work with.  So set
<equation>
B=\set{\vect{x}_1,\,\vect{x}_2,\,\vect{x}_3,\,\vect{x}_4}
=\set{
\begin{bmatrix}
<![CDATA[ 1 & 0 \\ 0 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 0 & 1 \\ 0 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 0 & 0 \\ 1 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 0 & 0 \\ 0 & 1]]>
\end{bmatrix}
}
</equation>
</p>

<p>Then to build the matrix representation of $S$ relative to $B$ compute,
<alignmath>
<![CDATA[\vectrep{B}{\lt{S}{\vect{x}_1}}&=]]>
<![CDATA[\vectrep{B}{\begin{bmatrix}0 & -14 \\ 18 & -6\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{B}{0\vect{x}_1+(-14)\vect{x}_2+18\vect{x}_3+(-6)\vect{x}_4}=]]>
\colvector{0\\-14\\18\\-6}\\
<![CDATA[\vectrep{B}{\lt{S}{\vect{x}_2}}&=]]>
<![CDATA[\vectrep{B}{\begin{bmatrix}-1 & -15\\21 & -7\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{B}{(-1)\vect{x}_1+(-15)\vect{x}_2+21\vect{x}_3+(-7)\vect{x}_4}=]]>
\colvector{-1\\-15\\21\\-7}\\
<![CDATA[\vectrep{B}{\lt{S}{\vect{x}_3}}&=]]>
<![CDATA[\vectrep{B}{\begin{bmatrix}-1 & -13\\19 & -7\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{B}{(-1)\vect{x}_1+(-13)\vect{x}_2+19\vect{x}_3+(-7)\vect{x}_4}=]]>
\colvector{-1\\-13\\19\\-7}\\
<![CDATA[\vectrep{B}{\lt{S}{\vect{x}_4}}&=]]>
<![CDATA[\vectrep{B}{\begin{bmatrix}-3 & 1\\3 & -3\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{B}{(-3)\vect{x}_1+1\vect{x}_2+3\vect{x}_3+(-3)\vect{x}_4}=]]>
\colvector{-3\\1\\3\\-3}
</alignmath>
</p>

<p>So by <acroref type="definition" acro="MR" /> we have
<equation>
M=\matrixrep{S}{B}{B}=
\begin{bmatrix}
<![CDATA[ 0 & -1 & -1 & -3 \\]]>
<![CDATA[ -14 & -15 & -13 & 1 \\]]>
<![CDATA[ 18 & 21 & 19 & 3 \\]]>
<![CDATA[ -6 & -7 & -7 & -3]]>
\end{bmatrix}
</equation>
</p>

<p>Now compute eigenvalues and eigenvectors of the matrix representation of $M$ with the techniques of <acroref type="section" acro="EE" />.  First the characteristic polynomial,
<equation>
\charpoly{M}{x}=\detname{M-xI_4}=x^4-x^3-10 x^2+4 x+24=(x-3) (x-2) (x+2)^2
</equation>
</p>

<p>We could now make statements about the eigenvalues of $M$, but in light of <acroref type="theorem" acro="EER" /> we can refer to the eigenvalues of $S$ and mildly abuse (or extend) our notation for multiplicities to write
<alignmath>
<![CDATA[\algmult{S}{3}&=1]]>
<![CDATA[&]]>
<![CDATA[\algmult{S}{2}&=1]]>
<![CDATA[&]]>
<![CDATA[\algmult{S}{-2}&=2]]>
</alignmath>
</p>

<p>Now compute the eigenvectors of $M$,
<alignmath>
<![CDATA[\lambda&=3&M-3I_4&=]]>
\begin{bmatrix}
<![CDATA[ -3 & -1 & -1 & -3 \\]]>
<![CDATA[ -14 & -18 & -13 & 1 \\]]>
<![CDATA[ 18 & 21 & 16 & 3 \\]]>
<![CDATA[ -6 & -7 & -7 & -6]]>
\end{bmatrix}
\rref
\begin{bmatrix}
<![CDATA[ \leading{1} & 0 & 0 & 1 \\]]>
<![CDATA[ 0 & \leading{1} & 0 & -3 \\]]>
<![CDATA[ 0 & 0 & \leading{1} & 3 \\]]>
<![CDATA[ 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[&&\eigenspace{M}{3}&=\nsp{M-3I_4}]]>
=\spn{\set{\colvector{-1\\3\\-3\\1}}}
</alignmath>
<alignmath>
<![CDATA[\lambda&=2&M-2I_4&=]]>
\begin{bmatrix}
<![CDATA[ -2 & -1 & -1 & -3 \\]]>
<![CDATA[ -14 & -17 & -13 & 1 \\]]>
<![CDATA[ 18 & 21 & 17 & 3 \\]]>
<![CDATA[ -6 & -7 & -7 & -5]]>
\end{bmatrix}
\rref
\begin{bmatrix}
<![CDATA[ \leading{1} & 0 & 0 & 2 \\]]>
<![CDATA[ 0 & \leading{1} & 0 & -4 \\]]>
<![CDATA[ 0 & 0 & \leading{1} & 3 \\]]>
<![CDATA[ 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[&&\eigenspace{M}{2}&=\nsp{M-2I_4}]]>
=\spn{\set{\colvector{-2\\4\\-3\\1}}}
</alignmath>
<alignmath>
<![CDATA[\lambda&=-2&M-(-2)I_4&=]]>
\begin{bmatrix}
<![CDATA[ 2 & -1 & -1 & -3 \\]]>
<![CDATA[ -14 & -13 & -13 & 1 \\]]>
<![CDATA[ 18 & 21 & 21 & 3 \\]]>
<![CDATA[ -6 & -7 & -7 & -1]]>
\end{bmatrix}
\rref
\begin{bmatrix}
<![CDATA[ \leading{1} & 0 & 0 & -1 \\]]>
<![CDATA[ 0 & \leading{1} & 1 & 1 \\]]>
<![CDATA[ 0 & 0 & 0 & 0 \\]]>
<![CDATA[ 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[&&\eigenspace{M}{-2}&=\nsp{M-(-2)I_4}]]>
=\spn{\set{\colvector{0\\-1\\1\\0},\,\colvector{1\\-1\\0\\1}}}
</alignmath>
</p>

<p>According to <acroref type="theorem" acro="EER" /> the eigenvectors just listed as basis vectors for the eigenspaces of $M$ are vector representations (relative to $B$) of eigenvectors for $S$.  So the application if the inverse function $\vectrepinvname{B}$ will convert these column vectors into elements of the vector space $M_{22}$ ($2\times 2$ matrices) that are eigenvectors of $S$.  Since $\vectrepname{B}$ is an isomorphism (<acroref type="theorem" acro="VRILT" />), so is $\vectrepinvname{B}$.  Applying the inverse function will then preserve linear independence and spanning properties, so with a sweeping application of the <miscref type="principle" text="Coordinatization Principle" /> and some extensions of our previous notation for eigenspaces and geometric multiplicities, we can write,
<alignmath>
\vectrepinv{B}{\colvector{-1\\3\\-3\\1}}
<![CDATA[&=]]>
(-1)\vect{x}_1+3\vect{x}_2+(-3)\vect{x}_3+1\vect{x}_4=
<![CDATA[\begin{bmatrix}-1 & 3\\-3 & 1\end{bmatrix}\\]]>
\vectrepinv{B}{\colvector{-2\\4\\-3\\1}}
<![CDATA[&=]]>
(-2)\vect{x}_1+4\vect{x}_2+(-3)\vect{x}_3+1\vect{x}_4=
<![CDATA[\begin{bmatrix}-2 & 4\\-3 & 1\end{bmatrix}\\]]>
\vectrepinv{B}{\colvector{0\\-1\\1\\0}}
<![CDATA[&=]]>
0\vect{x}_1+(-1)\vect{x}_2+1\vect{x}_3+0\vect{x}_4=
<![CDATA[\begin{bmatrix}0 & -1\\1 & 0\end{bmatrix}\\]]>
\vectrepinv{B}{\colvector{1\\-1\\0\\1}}
<![CDATA[&=]]>
1\vect{x}_1+(-1)\vect{x}_2+0\vect{x}_3+1\vect{x}_4=
<![CDATA[\begin{bmatrix}1 & -1\\0 & 1\end{bmatrix}\\]]>
</alignmath>
</p>

<p>So
<alignmath>
<![CDATA[\eigenspace{S}{3}&=]]>
<![CDATA[\spn{\set{\begin{bmatrix}-1 & 3\\-3 & 1\end{bmatrix}}}\\]]>
<![CDATA[\eigenspace{S}{2}&=]]>
<![CDATA[\spn{\set{\begin{bmatrix}-2 & 4\\-3 & 1\end{bmatrix}}}\\]]>
<![CDATA[\eigenspace{S}{-2}&=]]>
<![CDATA[\spn{\set{\begin{bmatrix}0 & -1\\1 & 0\end{bmatrix},\,\begin{bmatrix}1 & -1\\0 & 1\end{bmatrix}}}]]>
</alignmath>
with geometric multiplicities given by
<alignmath>
<![CDATA[\geomult{S}{3}&=1]]>
<![CDATA[&]]>
<![CDATA[\geomult{S}{2}&=1]]>
<![CDATA[&]]>
<![CDATA[\geomult{S}{-2}&=2]]>
</alignmath>
</p>

<p>Suppose we now decided to build another matrix representation of $S$, only now relative to a linearly independent set of eigenvectors of $S$, such as
<equation>
C=
\set{
<![CDATA[\begin{bmatrix}-1 & 3\\-3 & 1\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}-2 & 4\\-3 & 1\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}0 & -1\\1 & 0\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}1 & -1\\0 & 1\end{bmatrix}]]>
}
</equation>
</p>

<p>At this point you should have computed enough matrix representations to predict that the result of representing $S$ relative to $C$ will be a diagonal matrix.  Computing this representation is an example of how <acroref type="theorem" acro="SCB" /> generalizes the diagonalizations from <acroref type="section" acro="SD" />.  For the record, here is the diagonal representation,
<equation>
\matrixrep{S}{C}{C}
=
\begin{bmatrix}
<![CDATA[ 3 & 0 & 0 & 0 \\]]>
<![CDATA[ 0 & 2 & 0 & 0 \\]]>
<![CDATA[ 0 & 0 & -2 & 0 \\]]>
<![CDATA[ 0 & 0 & 0 & -2]]>
\end{bmatrix}
</equation>
</p>

<p>Our interest in this example is not necessarily building nice representations, but instead we want to demonstrate how eigenvalues and eigenvectors are an intrinsic property of a linear transformation, independent of any particular representation.  To this end, we will repeat the foregoing, but replace $B$ by another basis.  We will make this basis different, but not extremely so,
<equation>
D=\set{\vect{y}_1,\,\vect{y}_2,\,\vect{y}_3,\,\vect{y}_4}
=\set{
\begin{bmatrix}
<![CDATA[ 1 & 0 \\ 0 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 1 & 1 \\ 0 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 1 & 1 \\ 1 & 0]]>
\end{bmatrix}
,\,
\begin{bmatrix}
<![CDATA[ 1 & 1 \\ 1 & 1]]>
\end{bmatrix}
}
</equation>
</p>

<p>Then to build the matrix representation of $S$ relative to $D$ compute,
<alignmath>
<![CDATA[\vectrep{D}{\lt{S}{\vect{y}_1}}&=]]>
<![CDATA[\vectrep{D}{\begin{bmatrix}0 & -14\\18 & -6\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{D}{14\vect{y}_1+(-32)\vect{y}_2+24\vect{y}_3+(-6)\vect{y}_4}=]]>
\colvector{14\\-32\\24\\-6}\\
<![CDATA[\vectrep{D}{\lt{S}{\vect{y}_2}}&=]]>
<![CDATA[\vectrep{D}{\begin{bmatrix}-1 & -29 \\ 39 & -13\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{D}{28\vect{y}_1+(-68)\vect{y}_2+52\vect{y}_3+(-13)\vect{y}_4}=]]>
\colvector{28\\-68\\52\\-13}\\
<![CDATA[\vectrep{D}{\lt{S}{\vect{y}_3}}&=]]>
<![CDATA[\vectrep{D}{\begin{bmatrix}-2 & -42 \\ 58 & -20\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{D}{40\vect{y}_1+(-100)\vect{y}_2+78\vect{y}_3+(-20)\vect{y}_4}=]]>
\colvector{40\\-100\\78\\-20}\\
<![CDATA[\vectrep{D}{\lt{S}{\vect{y}_4}}&=]]>
<![CDATA[\vectrep{D}{\begin{bmatrix}-5 & -41 \\ 61 & -23\end{bmatrix}}\\]]>
<![CDATA[&=\vectrep{D}{36\vect{y}_1+(-102)\vect{y}_2+84\vect{y}_3+(-23)\vect{y}_4}=]]>
\colvector{36\\-102\\84\\-23}\\
</alignmath>
</p>

<p>So by <acroref type="definition" acro="MR" /> we have
<equation>
N=\matrixrep{S}{D}{D}=
\begin{bmatrix}
<![CDATA[ 14 & 28 & 40 & 36 \\]]>
<![CDATA[ -32 & -68 & -100 & -102 \\]]>
<![CDATA[ 24 & 52 & 78 & 84 \\]]>
<![CDATA[ -6 & -13 & -20 & -23]]>
\end{bmatrix}
</equation>
</p>

<p>Now compute eigenvalues and eigenvectors of the matrix representation of $N$ with the techniques of <acroref type="section" acro="EE" />.  First the characteristic polynomial,
<equation>
\charpoly{N}{x}=\detname{N-xI_4}=x^4-x^3-10 x^2+4 x+24=(x-3) (x-2) (x+2)^2
</equation>
</p>

<p>Of course this is not news.  We now know that $M=\matrixrep{S}{B}{B}$ and $N=\matrixrep{S}{D}{D}$ are similar matrices (<acroref type="theorem" acro="SCB" />).  But <acroref type="theorem" acro="SMEE" /> told us long ago that similar matrices have identical characteristic polynomials.  Now compute eigenvectors for the matrix representation,  which will be different than what we found for $M$,
<alignmath>
<![CDATA[\lambda&=3&N-3I_4&=]]>
\begin{bmatrix}
<![CDATA[ 11 & 28 & 40 & 36 \\]]>
<![CDATA[ -32 & -71 & -100 & -102 \\]]>
<![CDATA[ 24 & 52 & 75 & 84 \\]]>
<![CDATA[ -6 & -13 & -20 & -26]]>
\end{bmatrix}
\rref
\begin{bmatrix}
<![CDATA[ 1 & 0 & 0 & 4 \\]]>
<![CDATA[ 0 & 1 & 0 & -6 \\]]>
<![CDATA[ 0 & 0 & 1 & 4 \\]]>
<![CDATA[ 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[&&\eigenspace{N}{3}&=\nsp{N-3I_4}]]>
=\spn{\set{\colvector{-4\\6\\-4\\1}}}
</alignmath>
<alignmath>
<![CDATA[\lambda&=2&N-2I_4&=]]>
\begin{bmatrix}
<![CDATA[ 12 & 28 & 40 & 36 \\]]>
<![CDATA[ -32 & -70 & -100 & -102 \\]]>
<![CDATA[ 24 & 52 & 76 & 84 \\]]>
<![CDATA[ -6 & -13 & -20 & -25]]>
\end{bmatrix}
\rref
\begin{bmatrix}
<![CDATA[ 1 & 0 & 0 & 6 \\]]>
<![CDATA[ 0 & 1 & 0 & -7 \\]]>
<![CDATA[ 0 & 0 & 1 & 4 \\]]>
<![CDATA[ 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[&&\eigenspace{N}{2}&=\nsp{N-2I_4}]]>
=\spn{\set{\colvector{-6\\7\\-4\\1}}}
</alignmath>
<alignmath>
<![CDATA[\lambda&=-2&N-(-2)I_4&=]]>
\begin{bmatrix}
<![CDATA[ 16 & 28 & 40 & 36 \\]]>
<![CDATA[ -32 & -66 & -100 & -102 \\]]>
<![CDATA[ 24 & 52 & 80 & 84 \\]]>
<![CDATA[ -6 & -13 & -20 & -21]]>
\end{bmatrix}
\rref
\begin{bmatrix}
<![CDATA[ 1 & 0 & -1 & -3 \\]]>
<![CDATA[ 0 & 1 & 2 & 3 \\]]>
<![CDATA[ 0 & 0 & 0 & 0 \\]]>
<![CDATA[ 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[&&\eigenspace{N}{-2}&=\nsp{N-(-2)I_4}]]>
=\spn{\set{\colvector{1\\-2\\1\\0},\,\colvector{3\\-3\\0\\1}}}
</alignmath>
</p>

<p>Employing <acroref type="theorem" acro="EER" /> we can apply $\vectrepinvname{D}$ to each of the basis vectors of the eigenspaces of $N$ to obtain eigenvectors for $S$ that also form bases for eigenspaces of $S$,
<alignmath>
\vectrepinv{D}{\colvector{-4\\6\\-4\\1}}
<![CDATA[&=]]>
(-4)\vect{y}_1+6\vect{y}_2+(-4)\vect{y}_3+1\vect{y}_4=
<![CDATA[\begin{bmatrix}-1 & 3\\-3 & 1\end{bmatrix}\\]]>
\vectrepinv{D}{\colvector{-6\\7\\-4\\1}}
<![CDATA[&=]]>
(-6)\vect{y}_1+7\vect{y}_2+(-4)\vect{y}_3+1\vect{y}_4=
<![CDATA[\begin{bmatrix}-2 & 4\\-3 & 1\end{bmatrix}\\]]>
\vectrepinv{D}{\colvector{1\\-2\\1\\0}}
<![CDATA[&=]]>
1\vect{y}_1+(-2)\vect{y}_2+1\vect{y}_3+0\vect{y}_4=
<![CDATA[\begin{bmatrix}0 & -1\\1 & 0\end{bmatrix}\\]]>
\vectrepinv{D}{\colvector{3\\-3\\0\\1}}
<![CDATA[&=]]>
3\vect{y}_1+(-3)\vect{y}_2+0\vect{y}_3+1\vect{y}_4=
<![CDATA[\begin{bmatrix}1 & -2\\1 & 1\end{bmatrix}\\]]>
</alignmath>
</p>

<p>The eigenspaces for the eigenvalues of algebraic multiplicity 1 are exactly as before,
<alignmath>
<![CDATA[\eigenspace{S}{3}&=]]>
<![CDATA[\spn{\set{\begin{bmatrix}-1 & 3\\-3 & 1\end{bmatrix}}}\\]]>
<![CDATA[\eigenspace{S}{2}&=]]>
<![CDATA[\spn{\set{\begin{bmatrix}-2 & 4\\-3 & 1\end{bmatrix}}}]]>
</alignmath>
</p>

<p>However, the eigenspace for $\lambda=-2$ would at first glance appear to be different.  Here are the two eigenspaces for $\lambda=-2$, first the eigenspace obtained from $M=\matrixrep{S}{B}{B}$, then followed by the eigenspace obtained from $M=\matrixrep{S}{D}{D}$.
<alignmath>
<![CDATA[\eigenspace{S}{-2}&=]]>
<![CDATA[\spn{\set{\begin{bmatrix}0 & -1\\1 & 0\end{bmatrix},\,\begin{bmatrix}1 & -1\\0 & 1\end{bmatrix}}}]]>
<![CDATA[&]]>
<![CDATA[\eigenspace{S}{-2}&=]]>
<![CDATA[\spn{\set{\begin{bmatrix}0 & -1\\1 & 0\end{bmatrix},\,\begin{bmatrix}1 & -2\\1 & 1\end{bmatrix}}}]]>
</alignmath>
</p>

<p>Subspaces generally have many bases, and that is the situation here.  With a careful proof of set equality, you can show that these two eigenspaces are equal sets.  The key observation to make such a proof go is that
<equation>
<![CDATA[\begin{bmatrix}1 & -2\\1 & 1\end{bmatrix}]]>
=
<![CDATA[\begin{bmatrix}0 & -1\\1 & 0\end{bmatrix}+\begin{bmatrix}1 & -1\\0 & 1\end{bmatrix}]]>
</equation>
which will establish that the second set is a subset of the first.  With equal dimensions, <acroref type="theorem" acro="EDYES" /> will finish the task.</p>

<p>So the eigenvalues of a linear transformation are independent of the matrix representation employed to compute them!</p>

</example>

<p>Another example, this time a bit larger and with complex eigenvalues.</p>

<example acro="CELT" index="eigenvalues!complex, of a linear transformation">
<title>Complex eigenvectors of a linear transformation</title>

<p>Consider the linear transformation $\ltdefn{Q}{P_4}{P_4}$ defined by
<alignmath>
<![CDATA[&\lt{Q}{a+bx+cx^2+dx^3+ex^4}\\]]>
<![CDATA[&=(-46a-22b+13c+5d+e)+(117a+57b-32c-15d-4e) x+\\]]>
<![CDATA[&\quad\quad (-69a-29b+21c-7e)x^2+(159a+73b-44c-13d+2e)x^3+\\]]>
<![CDATA[&\quad\quad (-195a-87b+55c+10d-13e)x^4]]>
</alignmath>
</p>

<p>Choose a simple basis to compute with, say
<equation>
B=\set{1,\,x,\,x^2,\,x^3,\,x^4}
</equation>
</p>

<p>Then it should be apparent that the matrix representation of $Q$ relative to $B$ is
<equation>
M=\matrixrep{Q}{B}{B}=
\begin{bmatrix}
<![CDATA[ -46 & -22 & 13 & 5 & 1 \\]]>
<![CDATA[ 117 & 57 & -32 & -15 & -4 \\]]>
<![CDATA[ -69 & -29 & 21 & 0 & -7 \\]]>
<![CDATA[ 159 & 73 & -44 & -13 & 2 \\]]>
<![CDATA[ -195 & -87 & 55 & 10 & -13]]>
\end{bmatrix}
</equation>
</p>

<p>Compute the characteristic polynomial, eigenvalues and eigenvectors according to the techniques of <acroref type="section" acro="EE" />,
<alignmath>
\charpoly{Q}{x}
<![CDATA[&=-x^5+6 x^4-x^3-88 x^2+252 x-208\\]]>
<![CDATA[&=-(x-2)^2 (x+4) \left(x^2-6x+13\right)\\]]>
<![CDATA[&=-(x-2)^2 (x+4) \left(x-(3+2i)\right) \left(x-(3-2i)\right)\\]]>
</alignmath>
<alignmath>
<![CDATA[\algmult{Q}{2}&=2]]>
<![CDATA[&]]>
<![CDATA[\algmult{Q}{-4}&=1]]>
<![CDATA[&]]>
<![CDATA[\algmult{Q}{3+2i}&=1]]>
<![CDATA[&]]>
<![CDATA[\algmult{Q}{3-2i}&=1]]>
</alignmath>
<alignmath>
<![CDATA[\lambda&=2\\]]>
<![CDATA[M-(2)I_5&=]]>
\begin{bmatrix}
<![CDATA[ -48 & -22 & 13 & 5 & 1 \\]]>
<![CDATA[ 117 & 55 & -32 & -15 & -4 \\]]>
<![CDATA[ -69 & -29 & 19 & 0 & -7 \\]]>
<![CDATA[ 159 & 73 & -44 & -15 & 2 \\]]>
<![CDATA[ -195 & -87 & 55 & 10 & -15]]>
\end{bmatrix}
\rref
\begin{bmatrix}
<![CDATA[ 1 & 0 & 0 & \frac{1}{2} & -\frac{1}{2} \\]]>
<![CDATA[ 0 & 1 & 0 & -\frac{5}{2} & -\frac{5}{2} \\]]>
<![CDATA[ 0 & 0 & 1 & -2 & -6 \\]]>
<![CDATA[ 0 & 0 & 0 & 0 & 0 \\]]>
<![CDATA[ 0 & 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[\eigenspace{M}{2}&=\nsp{M-(2)I_5}]]>
=\spn{\set{
\colvector{-\frac{1}{2}\\\frac{5}{2}\\2\\1\\0},\,
\colvector{\frac{1}{2}\\\frac{5}{2}\\6\\0\\1}
}}
=\spn{\set{
\colvector{-1\\5\\4\\2\\0},\,
\colvector{1\\5\\12\\0\\2}
}}
</alignmath>
<alignmath>
<![CDATA[\lambda&=-4\\]]>
<![CDATA[M-(-4)I_5&=]]>
\begin{bmatrix}
<![CDATA[ -42 & -22 & 13 & 5 & 1 \\]]>
<![CDATA[ 117 & 61 & -32 & -15 & -4 \\]]>
<![CDATA[ -69 & -29 & 25 & 0 & -7 \\]]>
<![CDATA[ 159 & 73 & -44 & -9 & 2 \\]]>
<![CDATA[ -195 & -87 & 55 & 10 & -9]]>
\end{bmatrix}
\rref
\begin{bmatrix}
<![CDATA[ 1 & 0 & 0 & 0 & 1 \\]]>
<![CDATA[ 0 & 1 & 0 & 0 & -3 \\]]>
<![CDATA[ 0 & 0 & 1 & 0 & -1 \\]]>
<![CDATA[ 0 & 0 & 0 & 1 & -2 \\]]>
<![CDATA[ 0 & 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[\eigenspace{M}{-4}&=\nsp{M-(-4)I_5}]]>
=\spn{\set{\colvector{-1\\3\\1\\2\\1}}}
</alignmath>
<alignmath>
<![CDATA[\lambda&=3+2i\\]]>
<![CDATA[M-(3+2i)I_5&=]]>
\begin{bmatrix}
<![CDATA[ -49-2 i & -22 & 13 & 5 & 1 \\]]>
<![CDATA[ 117 & 54-2 i & -32 & -15 & -4\\]]>
<![CDATA[ -69 & -29 & 18-2 i & 0 & -7 \\]]>
<![CDATA[ 159 & 73 & -44 & -16-2 i & 2 \\]]>
<![CDATA[ -195 & -87 & 55 & 10 & -16-2 i]]>
\end{bmatrix}\\
<![CDATA[&\quad\quad\rref]]>
\begin{bmatrix}
<![CDATA[ 1 & 0 & 0 & 0 &  -\frac{3}{4}+\frac{i}{4} \\]]>
<![CDATA[ 0 & 1 & 0 & 0 &  \frac{7}{4}-\frac{i}{4} \\]]>
<![CDATA[ 0 & 0 & 1 & 0 &  -\frac{1}{2}+\frac{i}{2} \\]]>
<![CDATA[ 0 & 0 & 0 & 1 &  \frac{7}{4}-\frac{i}{4} \\]]>
<![CDATA[ 0 & 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[\eigenspace{M}{3+2i}&=\nsp{M-(3+2i)I_5}]]>
=\spn{\set{\colvector{\frac{3}{4}-\frac{i}{4} \\ -\frac{7}{4}+\frac{i}{4} \\  \frac{1}{2}-\frac{i}{2}  \\  -\frac{7}{4}+\frac{i}{4} \\ 1}}}
=\spn{\set{\colvector{3-i\\-7+i\\2-2i\\-7+i\\4}}}
</alignmath>
<alignmath>
<![CDATA[\lambda&=3-2i\\]]>
<![CDATA[M-(3-2i)I_5&=]]>
\begin{bmatrix}
<![CDATA[ -49+2 i & -22 & 13 & 5 & 1 \\]]>
<![CDATA[ 117 & 54+2 i & -32 & -15 & -4 \\]]>
<![CDATA[ -69 & -29 & 18+2 i & 0 & -7 \\]]>
<![CDATA[ 159 & 73 & -44 & -16+2 i & 2 \\]]>
<![CDATA[ -195 & -87 & 55 & 10 & -16+2 i]]>
\end{bmatrix}\\
<![CDATA[&\quad\quad\rref]]>
\begin{bmatrix}
<![CDATA[ 1 & 0 & 0 & 0 &  -\frac{3}{4}-\frac{i}{4} \\]]>
<![CDATA[ 0 & 1 & 0 & 0 &  \frac{7}{4}+\frac{i}{4} \\]]>
<![CDATA[ 0 & 0 & 1 & 0 &  -\frac{1}{2}-\frac{i}{2} \\]]>
<![CDATA[ 0 & 0 & 0 & 1 &  \frac{7}{4}+\frac{i}{4} \\]]>
<![CDATA[ 0 & 0 & 0 & 0 & 0]]>
\end{bmatrix}\\
<![CDATA[\eigenspace{M}{3-2i}&=\nsp{M-(3-2i)I_5}]]>
=\spn{\set{\colvector{\frac{3}{4}+\frac{i}{4} \\ -\frac{7}{4}-\frac{i}{4} \\  \frac{1}{2}+\frac{i}{2}  \\  -\frac{7}{4}-\frac{i}{4} \\ 1}}}
=\spn{\set{\colvector{3+i\\-7-i\\2+2i\\-7-i\\4}}}
</alignmath>
</p>

<p>It is straightforward to convert each of these basis vectors for eigenspaces of $M$ back to elements of $P_4$ by applying the isomorphism $\vectrepinvname{B}$,
<alignmath>
<![CDATA[\vectrepinv{B}{\colvector{-1\\5\\4\\2\\0}}&=-1+5x+4x^2+2x^3\\]]>
<![CDATA[\vectrepinv{B}{\colvector{1\\5\\12\\0\\2}}&=1+5x+12x^2+2x^4\\]]>
<![CDATA[\vectrepinv{B}{\colvector{-1\\3\\1\\2\\1}}&=-1+3x+x^2+2x^3+x^4\\]]>
<![CDATA[\vectrepinv{B}{\colvector{3-i\\-7+i\\2-2i\\-7+i\\4}}&=(3-i)+(-7+i)x+(2-2i)x^2+(-7+i)x^3+4x^4\\]]>
<![CDATA[\vectrepinv{B}{\colvector{3+i\\-7-i\\2+2i\\-7-i\\4}}&=(3+i)+(-7-i)x+(2+2i)x^2+(-7-i)x^3+4x^4\\]]>
</alignmath>
</p>

<p>So we apply <acroref type="theorem" acro="EER" /> and the <miscref type="principle" text="Coordinatization Principle" /> to get the eigenspaces for $Q$,
<alignmath>
<![CDATA[\eigenspace{Q}{2}&=\spn{\set{-1+5x+4x^2+2x^3,\,1+5x+12x^2+2x^4}}\\]]>
<![CDATA[\eigenspace{Q}{-4}&=\spn{\set{-1+3x+x^2+2x^3+x^4}}\\]]>
<![CDATA[\eigenspace{Q}{3+2i}&=\spn{\set{(3-i)+(-7+i)x+(2-2i)x^2+(-7+i)x^3+4x^4}}\\]]>
<![CDATA[\eigenspace{Q}{3-2i}&=\spn{\set{(3+i)+(-7-i)x+(2+2i)x^2+(-7-i)x^3+4x^4}}]]>
</alignmath>
with geometric multiplicities
<alignmath>
<![CDATA[\geomult{Q}{2}&=2]]>
<![CDATA[&]]>
<![CDATA[\geomult{Q}{-4}&=1]]>
<![CDATA[&]]>
<![CDATA[\geomult{Q}{3+2i}&=1]]>
<![CDATA[&]]>
<![CDATA[\geomult{Q}{3-2i}&=1]]>
</alignmath>
</p>

</example>

<sageadvice acro="CELT" index="matrix representation!designing">
<title>Designing Matrix Representations</title>
How do we find the eigenvectors of a linear transformation?  How do we find pleasing (or computationally simple) matrix representations of linear transformations.  <acroref type="theorem" acro="EER" /> and <acroref type="theorem" acro="SCB" /> applied in the context of <acroref type="theorem" acro="DC" /> can answer both questions.  Here is an example.
<sage>
<input>x1, x2, x3, x4, x5, x6 = var('x1, x2, x3, x4,x5,x6')
outputs = [  9*x1 - 15*x2 - 7*x3 + 15*x4 - 36*x5 - 53*x6,
            24*x1 - 20*x2 - 9*x3 + 18*x4 - 24*x5 - 78*x6,
             8*x1 -  6*x2 - 3*x3 +  6*x4 -  6*x5 - 26*x6,
           -12*x1 -  9*x2 - 3*x3 + 13*x4 - 54*x5 - 24*x6,
            -8*x1 +  6*x2 + 3*x3 -  6*x4 +  6*x5 + 26*x6,
            -4*x1 -  3*x2 -   x3 +  3*x4 - 18*x5 -  4*x6]
T_symbolic(x1, x2, x3, x4, x5, x6) = outputs
T1 = linear_transformation(QQ^6, QQ^6, T_symbolic)
M1 = T1.matrix(side='right')
M1
</input>
<output>[  9 -15  -7  15 -36 -53]
[ 24 -20  -9  18 -24 -78]
[  8  -6  -3   6  -6 -26]
[-12  -9  -3  13 -54 -24]
[ -8   6   3  -6   6  26]
[ -4  -3  -1   3 -18  -4]
</output>
</sage>

Now we compute the eigenvalues and eigenvectors of <code>M1</code>.  Since <code>M1</code> is diagonalizable, we can find a basis of eigenvectors for use as the basis for a new representation.
<sage>
<input>ev = M1.eigenvectors_right()
ev
</input>
<output>[(4, [
(1, 6/5, 2/5, 4/5, -2/5, 1/5)
], 1), (0, [
(1, 9/7, 4/7, 3/7, -3/7, 1/7)
], 1), (-2, [
(1, 7/5, 2/5, 3/5, -2/5, 1/5)
], 1), (-3, [
(1, 3, 1, -3/2, -1, -1/2)
], 1), (1, [
(1, 0, 0, 3, 0, 1),
(0, 1, 1/3, -2, -1/3, -2/3)
], 2)]
</output>
</sage>

<sage>
<input>evalues, evectors = M1.eigenmatrix_right()
B = evectors.columns()
V = (QQ^6).subspace_with_basis(B)
T2 = linear_transformation(V, V, T_symbolic)
M2 = T2.matrix('right')
M2
</input>
<output>[ 4  0  0  0  0  0]
[ 0  0  0  0  0  0]
[ 0  0 -2  0  0  0]
[ 0  0  0 -3  0  0]
[ 0  0  0  0  1  0]
[ 0  0  0  0  0  1]
</output>
</sage>

The eigenvectors that are the basis elements in B are the eigenvectors of the linear transformation, <em>relative</em> to the standard basis.  For different representations the eigenvectors tke different forms, relative to other bases.  What are the eigenvctors of the matrix representation M2?<br /><br />
Notice that the eigenvalues of the linear transformation are totally independent of the representation.  So in a sense, they are an inherent property of the linear transformation.<br /><br />
You should be able to use these techniques with linear transformations on abstract vector spaces <mdash /> just use a mental linear transformation transforming the abstract vector space back-and-forth between a vector space of column vectors of the right size.


</sageadvice>
<sageadvice acro="SUTH4" index="sage under the hood!round 4">
<title>Sage Under The Hood, Round 4</title>
We finally have enough theorems to understand how Sage creates and manages linear transformations.  With a choice of bases for the domain and codomain, a linear transformation can be represented by a matrix.  Every interesting property of the linear transformation can be computed from the matrix representation, and we can convert between representations (of vectors and linear transformations) with change-of-basis matrices, similarity and matrix multiplication.<br /><br />
So we can understand the theory of linear algebra better by experimenting with the assistance of Sage, and the theory of linear algebra helps us understand how Sage is designed and functions.  A virtuous cycle, if there ever was one.  Keep it going.


</sageadvice>
</subsection>

<!--   End of  cb.tex -->
<readingquestions>
<ol>
<li>The change-of-basis matrix is a matrix representation of which linear transformation?
</li>
<li>Find the change-of-basis matrix, $\cbm{B}{C}$, for the two bases of $\complex{2}$
<alignmath>
<![CDATA[B&=\set{\colvector{2\\3},\,\colvector{-1\\2}}&]]>
<![CDATA[C&=\set{\colvector{1\\0},\,\colvector{1\\1}}]]>
</alignmath>
</li>
<li>What is the third <q>surprise,</q> and why is it surprising?
</li></ol>
</readingquestions>

<exercisesubsection>

<exercise type="C" number="20" rough="Vector representation from Example CBCV">
<problem contributor="robertbeezer">In <acroref type="example" acro="CBCV" /> we computed the vector representation of $\vect{y}$ relative to $C$, $\vectrep{C}{\vect{y}}$, as an example of <acroref type="theorem" acro="CB" />.  Compute this same representation directly.  In other words, apply <acroref type="definition" acro="VR" /> rather than <acroref type="theorem" acro="CB" />.
</problem>
</exercise>

<exercise type="C" number="21" rough="Matrix representation from Example MRCM">
<problem contributor="robertbeezer">Perform a check on <acroref type="example" acro="MRCM" /> by computing $\matrixrep{Q}{B}{D}$ directly.  In other words, apply <acroref type="definition" acro="MR" /> rather than <acroref type="theorem" acro="MRCB" />.
</problem>
<solution contributor="robertbeezer">Apply <acroref type="definition" acro="MR" />,
<alignmath>
<![CDATA[&\vectrep{D}{\lt{Q}{\begin{bmatrix}5&-3\\-3&-2\end{bmatrix}}}]]>
=\vectrep{D}{19+14x-2x^2-28x^3}\\
<![CDATA[&=\vectrep{D}{(-39)(2+x-2x^2+3x^3)+62(-1-2x^2+3x^3)+(-53)(-3-x+x^3)+(-44)(-x^2+x^3)}\\]]>
<![CDATA[&=\colvector{-39\\62\\-53\\-44}\\]]>
<![CDATA[&\vectrep{D}{\lt{Q}{\begin{bmatrix}2&-3\\-3&0\end{bmatrix}}}]]>
=\vectrep{D}{16+9x-7x^2-14x^3}\\
<![CDATA[&=\vectrep{D}{(-23)(2+x-2x^2+3x^3)+(34)(-1-2x^2+3x^3)+(-32)(-3-x+x^3)+(-15)(-x^2+x^3)}\\]]>
<![CDATA[&=\colvector{-23\\34\\-32\\-15}\\]]>
<![CDATA[&\vectrep{D}{\lt{Q}{\begin{bmatrix}1&2\\2&4\end{bmatrix}}}]]>
=\vectrep{D}{25+9x+3x^2+4x^3}\\
<![CDATA[&=\vectrep{D}{(14)(2+x-2x^2+3x^3)+(-12)(-1-2x^2+3x^3)+5(-3-x+x^3)+(-7)(-x^2+x^3)}\\]]>
<![CDATA[&=\colvector{14\\-12\\5\\-7}\\]]>
</alignmath>
These three vectors are the columns of the matrix representation,
<equation>
\matrixrep{Q}{B}{D}=
\begin{bmatrix}
<![CDATA[ -39 & -23 & 14 \\]]>
<![CDATA[ 62 & 34 & -12 \\]]>
<![CDATA[ -53 & -32 & 5 \\]]>
<![CDATA[ -44 & -15 & -7]]>
\end{bmatrix}
</equation>
which coincides with the result obtained in <acroref type="example" acro="MRCM" />.
</solution>
</exercise>

<exercise type="C" number="30" rough="Eigenvector representation">
<problem contributor="robertbeezer">Find a basis for the vector space $P_3$  composed of eigenvectors of the linear transformation $T$.  Then find a matrix representation of $T$ relative to this basis.
<equation>
\ltdefn{T}{P_3}{P_3},\quad\lt{T}{a+bx+cx^2+dx^3}=
(a+c+d)+(b+c+d)x+(a+b+c)x^2+(a+b+d)x^3
</equation>
</problem>
<solution contributor="robertbeezer">With the domain and codomain being identical, we will build a matrix representation using the same basis for both the domain and codomain.  The eigenvalues of the matrix representation will be the eigenvalues of the linear transformation, and we can obtain the eigenvectors of the linear transformation by un-coordinatizing (<acroref type="theorem" acro="EER" />).  Since the method does not depend on <em>which</em> basis we choose, we can choose a natural basis for ease of computation, say,
<equation>
B=\set{1,\,x,\,x^2,x^3}
</equation>
The matrix representation is then,
<equation>
\matrixrep{T}{B}{B}=
\begin{bmatrix}
<![CDATA[1 &  0 &  1 &  1\\]]>
<![CDATA[0 &  1 &  1 &  1\\]]>
<![CDATA[1 &  1 &  1 &  0\\]]>
<![CDATA[1 &  1 &  0 &  1]]>
\end{bmatrix}
</equation>
The eigenvalues and eigenvectors of this matrix were computed in <acroref type="example" acro="ESMS4" />.  A basis for $\complex{4}$, composed of eigenvectors of the matrix representation is,
<equation>
C=\set{
\colvector{1\\1\\1\\1},\,
\colvector{-1\\1\\0\\0},\,
\colvector{0\\0\\-1\\1},\,
\colvector{-1\\-1\\1\\1}
}
</equation>
Applying $\vectrepinvname{B}$ to each vector of this set, yields a basis of $P_3$ composed of eigenvectors of $T$,
<equation>
D=\set{1+x+x^2+x^3, -1+x,\,-x^2+x^3,\,-1-x+x^2+x^3}
</equation>
The matrix representation of $T$ relative to the basis $D$ will be a diagonal matrix with the corresponding eigenvalues along the diagonal, so in this case we get
<equation>
\matrixrep{T}{D}{D}=
\begin{bmatrix}
<![CDATA[3 & 0 & 0 & 0\\]]>
<![CDATA[0 & 1 & 0 & 0\\]]>
<![CDATA[0 & 0 & 1 & 0\\]]>
<![CDATA[0 & 0 & 0 & -1]]>
\end{bmatrix}
</equation>
</solution>
</exercise>

<exercise type="C" number="40" rough="Basis to yield diagonalization on S_22">
<problem contributor="robertbeezer">Let $S_{22}$ be the vector space of $2\times 2$ symmetric matrices. Find a basis $B$ for $S_{22}$ that yields a diagonal matrix representation of the linear transformation $R$.
<alignmath>
\ltdefn{R}{S_{22}}{S_{22}},\quad
<![CDATA[\lt{R}{\begin{bmatrix}a&b\\b&c\end{bmatrix}}=]]>
\begin{bmatrix}
<![CDATA[-5a + 2b - 3c & -12a + 5b - 6c\\]]>
<![CDATA[-12a + 5b - 6c & 6a - 2b + 4c]]>
\end{bmatrix}
</alignmath>
</problem>
<solution contributor="robertbeezer">Begin with a matrix representation of $R$, any matrix representation, but use the same basis for both instances of $S_{22}$.  We'll choose a basis that makes it easy to compute vector representations in $S_{22}$.
<equation>
B=\set{
<![CDATA[\begin{bmatrix} 1 & 0 \\ 0 & 0 \end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix} 0 & 1 \\ 1 & 0 \end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix} 0 & 0 \\ 0 & 1 \end{bmatrix}]]>
}
</equation>
Then the resulting matrix representation of $R$  (<acroref type="definition" acro="MR" />) is
<equation>
\matrixrep{R}{B}{B}=
\begin{bmatrix}
<![CDATA[ -5 & 2 & -3 \\]]>
<![CDATA[ -12 & 5 & -6 \\]]>
<![CDATA[ 6 & -2 & 4]]>
\end{bmatrix}
</equation>
Now, compute the eigenvalues and eigenvectors of this matrix, with the goal of diagonalizing the matrix (<acroref type="theorem" acro="DC" />),
<alignmath>
<![CDATA[\lambda&=2]]>
<![CDATA[&]]>
<![CDATA[\eigenspace{\matrixrep{R}{B}{B}}{2}&=\spn{\set{\colvector{-1\\-2\\1}}}\\]]>
<![CDATA[\lambda&=1]]>
<![CDATA[&]]>
<![CDATA[\eigenspace{\matrixrep{R}{B}{B}}{1}&=\spn{\set{\colvector{-1\\0\\2},\,\colvector{1\\3\\0}}}\\]]>
</alignmath>
The three vectors that occur as basis elements for these eigenspaces will together form a linearly independent set (check this!).  So these column vectors may be employed in a matrix that will diagonalize the matrix representation.  If we <q>un-coordinatize</q> these three column vectors relative to the basis $B$, we will find three linearly independent elements of $S_{22}$ that are eigenvectors of the linear transformation $R$ (<acroref type="theorem" acro="EER" />).  A matrix representation relative to this basis of eigenvectors will be diagonal, with the eigenvalues ($\lambda=2,\,1$) as the diagonal elements.  Here we go,
<alignmath>
<![CDATA[\vectrepinv{B}{\colvector{-1\\-2\\1}}&=]]>
<![CDATA[(-1)\begin{bmatrix} 1 & 0 \\ 0 & 0 \end{bmatrix}+]]>
<![CDATA[(-2)\begin{bmatrix} 0 & 1 \\ 1 & 0 \end{bmatrix}+]]>
<![CDATA[1\begin{bmatrix} 0 & 0 \\ 0 & 1 \end{bmatrix}]]>
=
\begin{bmatrix}
<![CDATA[-1 & -2 \\-2 & 1]]>
\end{bmatrix}\\
<![CDATA[\vectrepinv{B}{\colvector{-1\\0\\2}}&=]]>
<![CDATA[(-1)\begin{bmatrix} 1 & 0 \\ 0 & 0 \end{bmatrix}+]]>
<![CDATA[0\begin{bmatrix} 0 & 1 \\ 1 & 0 \end{bmatrix}+]]>
<![CDATA[2\begin{bmatrix} 0 & 0 \\ 0 & 1 \end{bmatrix}]]>
=
\begin{bmatrix}
<![CDATA[-1 & 0 \\ 0 & 2]]>
\end{bmatrix}\\
<![CDATA[\vectrepinv{B}{\colvector{1\\3\\0}}&=]]>
<![CDATA[1\begin{bmatrix} 1 & 0 \\ 0 & 0 \end{bmatrix}+]]>
<![CDATA[3\begin{bmatrix} 0 & 1 \\ 1 & 0 \end{bmatrix}+]]>
<![CDATA[0\begin{bmatrix} 0 & 0 \\ 0 & 1 \end{bmatrix}]]>
=
\begin{bmatrix}
<![CDATA[1 & 3 \\ 3 & 0]]>
\end{bmatrix}\\
</alignmath>
So the requested basis of $S_{22}$, yielding a diagonal matrix representation of $R$, is
<equation>
\set{
\begin{bmatrix}
<![CDATA[-1 & -2 \\-2 & 1]]>
\end{bmatrix}\,\
\begin{bmatrix}
<![CDATA[-1 & 0 \\ 0 & 2]]>
\end{bmatrix},\,
\begin{bmatrix}
<![CDATA[1 & 3 \\ 3 & 0]]>
\end{bmatrix}%
}
</equation>
</solution>
</exercise>

<exercise type="C" number="41" rough="Basis to yield diagonalization on S_22">
<problem contributor="robertbeezer">Let $S_{22}$ be the vector space of $2\times 2$ symmetric matrices.  Find a basis for $S_{22}$ composed of eigenvectors of the linear transformation $\ltdefn{Q}{S_{22}}{S_{22}}$.
<equation>
\lt{Q}{
\begin{bmatrix}
<![CDATA[ a  &  b\\]]>
<![CDATA[ b  &  c]]>
\end{bmatrix}
}
=
\begin{bmatrix}
<![CDATA[ 25a + 18b + 30c  &  -16a - 11b - 20c\\]]>
<![CDATA[ -16a - 11b - 20c  &  -11a - 9b - 12c]]>
\end{bmatrix}
</equation>
</problem>
<solution contributor="robertbeezer">Use a single basis for both the domain and codomain, since they are equal.
<equation>
B=\set{
<![CDATA[\begin{bmatrix}1 & 0 \\ 0 & 0\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}0 & 1 \\ 1 & 0\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}0 & 0 \\ 0 & 1\end{bmatrix}]]>
}
</equation>
The matrix representation of $Q$ relative to $B$ is
<equation>
M=
\matrixrep{Q}{B}{B}
=
\begin{bmatrix}
<![CDATA[ 25 & 18 & 30 \\]]>
<![CDATA[ -16 & -11 & -20 \\]]>
<![CDATA[ -11 & -9 & -12]]>
\end{bmatrix}
</equation>
We can analyze this matrix with the techniques of <acroref type="section" acro="EE" /> and then apply <acroref type="theorem" acro="EER" />.  The eigenvalues of this matrix are $\lambda=-2,\,1,\,3$ with eigenspaces
<alignmath>
<![CDATA[\eigenspace{M}{-2}&=\spn{\set{\colvector{-6\\4\\3}}}]]>
<![CDATA[&]]>
<![CDATA[\eigenspace{M}{1}&=\spn{\set{\colvector{-2\\1\\1}}}]]>
<![CDATA[&]]>
<![CDATA[\eigenspace{M}{3}&=\spn{\set{\colvector{-3\\2\\1}}}]]>
</alignmath>
Because the three eigenvalues are distinct, the three basis vectors from the three eigenspaces for a linearly independent set (<acroref type="theorem" acro="EDELI" />).  <acroref type="theorem" acro="EER" /> says we can uncoordinatize these eigenvectors to obtain eigenvectors of $Q$.  By <acroref type="theorem" acro="ILTLI" /> the resulting set will remain linearly independent.  Set
<equation>
C=\set{
\vectrepinv{B}{\colvector{-6\\4\\3}},\,
\vectrepinv{B}{\colvector{-2\\1\\1}},\,
\vectrepinv{B}{\colvector{-3\\2\\1}}
}
=
\set{
<![CDATA[\begin{bmatrix}-6 & 4 \\ 4 & 3\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}-2 & 1 \\ 1 & 1\end{bmatrix},\,]]>
<![CDATA[\begin{bmatrix}-3 & 2 \\ 2 & 1\end{bmatrix}]]>
}
</equation>
Then $C$  is a linearly independent set of size 3 in the vector space $S_{22}$, which has dimension 3 as well.  By <acroref type="theorem" acro="G" />, $C$ is a basis of $S_{22}$.
</solution>
</exercise>

<exercise type="T" number="10" rough="Eigenvalues of inverse linear trans">
<problem contributor="robertbeezer">Suppose that $\ltdefn{T}{V}{V}$ is an invertible linear transformation with a nonzero eigenvalue $\lambda$.  Prove that $\displaystyle\frac{1}{\lambda}$ is an eigenvalue of $\ltinverse{T}$.
</problem>
<solution contributor="robertbeezer">Let $\vect{v}$ be an eigenvector of $T$ for the eigenvalue $\lambda$.  Then,
<alignmath>
<![CDATA[\lt{\ltinverse{T}}{\vect{v}}&=]]>
<![CDATA[\frac{1}{\lambda}\lambda\lt{\ltinverse{T}}{\vect{v}}&&\text{$\lambda\neq 0$}\\]]>
<![CDATA[&=\frac{1}{\lambda}\lt{\ltinverse{T}}{\lambda\vect{v}}&&]]>\text{<acroref type="theorem" acro="ILTLT" />}\\
<![CDATA[&=\frac{1}{\lambda}\lt{\ltinverse{T}}{\lt{T}{\vect{v}}}&&\text{$\vect{v}$ eigenvector of $T$}\\]]>
<![CDATA[&=\frac{1}{\lambda}\lt{I_V}{\vect{v}}&&]]>\text{<acroref type="definition" acro="IVLT" />}\\
<![CDATA[&=\frac{1}{\lambda}\vect{v}&&]]>\text{<acroref type="definition" acro="IDLT" />}
</alignmath>
which says that $\displaystyle\frac{1}{\lambda}$ is an eigenvalue of $\ltinverse{T}$ with eigenvector $\vect{v}$.  Note that it is possible to prove that any eigenvalue of an invertible linear transformation is never zero.  So the hypothesis that $\lambda$ be nonzero is just a convenience for this problem.
</solution>
</exercise>

<exercise type="T" number="15" rough="Injective iff no zero eigenvalue">
<problem contributor="robertbeezer">Suppose that $V$ is a vector space and $\ltdefn{T}{V}{V}$ is a linear transformation.  Prove that $T$ is injective if and only if $\lambda=0$ is not an eigenvalue of $T$.
</problem>
</exercise>

</exercisesubsection>

</section>