Some Conditions Equivalent to Diagonalizability of an 8 ‚ 8 Matrix E E is diagonalizable Ì Ë There is a basis of eigenvectors of E for ‘8 Ì Ë E has 8 linearly independent eigenvectors Ì Ë The sum of the dimensions of the eigenspaces of E œ 8 Suppose E is % ‚ % Can you say whether or not E is diagonalizable from the information given? 1. The characteristic polynomial of E is GÐ-Ñ œ Ð- "ÑÐ- #ÑÐ-# - "Ñ 2. E has eigenvalues - œ #ß $ß " and 4 3. E has eigenvalues - œ #ß $ß " and these have multiplicities "ß #ß " in that order. _________________________________________ (For after today's lecture) " ! 4. E is similar to ! ! # # ! ! % " & ! & $ " ' 5. What does diagonalizable mean for a " ‚ " matrix. Think it through just to see that all the pieces of the more general theory fit together correctly. X Ò Example Z Basis Æ Ò ÓU [ Æ Ò ÓV U œ Ö , " ß , # ß ,$ × ‘$ Suppose Basis V œ Ö-" ß - #ß -$ ß -% × ‘% Ò X Ð," Ñ œ -# %-$ X Ð,# Ñ œ -" $-# %-$ -% X Ð,$ Ñ œ &-" #-$ #-% Then Q œ Ò ÒX Ð," ÑÓV ÒX Ð," ÑÓV ÒX Ð,$ ÑÓV Ó œ ! " % ! " $ % " & ! # # So if @ œ #," (,# ',$ then ÒX Ð@ÑÓV œ ! " % ! " $ % " & $( # "* ! ( œ # $# ' # & so X Ð@Ñ œ $(-" "*-# $#-$ &-% # X ÀZ œ‘ Example Suppose Ä B B [ œ ‘ is given by X œ B C . C B C $ " The standard matrix for X is E œ Ò X Ð/" Ñ X Ð/# Ñ Ó œ " " ! " " Choose two new bases: " " U œ Ö," ß ,# × œ Ö ß × # " For ‘# ! ! " V œ Ö-" ß -# ß -$ × œ Ö ! ß # ß " × " " ! $ For ‘ With respect to the bases U , V what is the matrix for X À Q œ ÒX ÓUßV œ ?? Z œ ‘# X œ mult by E Ä Æ Ò ÓU ‘# [ œ ‘$ Æ Ò ÓV mult by Q œ ?? Ä ‘$ Q œ Ò ÒX Ð," ÑÓV ÒX Ð,# ÑÓV Ó (OVER) -" To find ÒX Ð," ÑÓV œ -# ß we need to find the weights -" ß -# ß -$ so that -# " ! ! " " $ X Ð ," Ñ œ X Ð Ñ œ œ -" ! -# # -$ " : # " " " ! ! so we need to solve ! " ! # " " -" " " -# œ $ Þ ! -$ " ." Similarly to find ÒX Ð,# ÑÓV œ .# we need to find the weights ." ß .# ß .$ so that .# " ! ! " " ! ! # " : X Ð,# Ñ œ X Ð Ñ œ œ . . . " # $ " # " " ! ! so we need to solve ! " ! # " " ." " " .# œ ! Þ ! .$ # We solve both systems simultaneously by row reducing the “double-augmented” matrix: ! ! " ! # " " " ! " $ " " " ! µ ÞÞÞ µ ! # ! ! " ! ! ! " $ # " $# "# " $ # -" $ ." # Therefore -# œ ÒX Ð," ÑÓV œ and .# œ ÒX Ð,# ÑÓV œ "# Þ -$ " .$ " $ So the matrix for X with respect to bases U and V is À Q œ # " $# "# Þ "