Page 194 - Linear Algebra Done Right
P. 194

Jordan Form
                      Jordan Form
                         We know that if V is a complex vector space, then for every T ∈L(V)               183
                      there is a basis of V with respect to which T has a nice upper-triangular
                      matrix (see 8.28). In this section we will see that we can do even better—
                      there is a basis of V with respect to which the matrix of T contains zeros
                      everywhere except possibly on the diagonal and the line directly above
                      the diagonal.
                         We begin by describing the nilpotent operators. Consider, for ex-
                                                          n
                      ample, the nilpotent operator N ∈L(F ) defined by
                                        N(z 1 ,...,z n ) = (0,z 1 ,...,z n−1 ).

                                                                                  n
                      If v = (1, 0,..., 0), then clearly (v,Nv,...,N n−1 v) is a basis of F and
                      (N n−1 v) is a basis of null N, which has dimension 1.
                                                                                  5
                         As another example, consider the nilpotent operator N ∈L(F ) de-
                      fined by

                      8.39            N(z 1 ,z 2 ,z 3 ,z 4 ,z 5 ) = (0,z 1 ,z 2 , 0,z 4 ).

                      Unlike the nilpotent operator discussed in the previous paragraph, for
                                                                             5
                      this nilpotent operator there does not exist a vector v ∈ F such that
                                    3
                               2
                                                          5
                                         4
                      (v, Nv, N v, N v, N v) is a basis of F . However, if v 1 = (1, 0, 0, 0, 0)
                                                            2
                      and v 2 = (0, 0, 0, 1, 0), then (v 1 ,Nv 1 ,N v 1 ,v 2 ,Nv 2 ) is a basis of F 5
                             2
                      and (N v 1 ,Nv 2 ) is a basis of null N, which has dimension 2.
                         Suppose N ∈L(V) is nilpotent. For each nonzero vector v ∈ V, let
                      m(v) denote the largest nonnegative integer such that N m(v) v  = 0. For  Obviously m(v)
                                         5
                      example, if N ∈L(F ) is defined by 8.39, then m(1, 0, 0, 0, 0) = 2.  depends on N as well
                         The lemma below shows that every nilpotent operator N ∈L(V)      as on v, but the choice
                      behaves similarly to the example defined by 8.39, in the sense that there  of N will be clear from
                      is a finite collection of vectors v 1 ,...,v k ∈ V such that the nonzero  the context.
                                           j
                      vectors of the form N v r form a basis of V; here r varies from 1 to k
                      and j varies from 0 to m(v r ).
                      8.40   Lemma:    If N ∈L(V) is nilpotent, then there exist vectors
                      v 1 ,...,v k ∈ V such that

                      (a)   (v 1 ,Nv 1 ,...,N m(v 1 ) v 1 ,...,v k ,Nv k ,...,N m(v k ) v k ) is a basis of V;
                      (b)   (N m(v 1 ) v 1 ,...,N m(v k ) v k ) is a basis of null N.
   189   190   191   192   193   194   195   196   197   198   199