Trigonalization over C
About points...
We associate a certain number of points with each exercise.
When you click an exercise into a collection, this number will be taken as points for the exercise, kind of "by default".
But once the exercise is on the collection, you can edit the number of points for the exercise in the collection independently, without any effect on "points by default" as represented by the number here.
That being said... How many "default points" should you associate with an exercise upon creation?
As with difficulty, there is no straight forward and generally accepted way.
But as a guideline, we tend to give as many points by default as there are mathematical steps to do in the exercise.
Again, very vague... But the number should kind of represent the "work" required.
When you click an exercise into a collection, this number will be taken as points for the exercise, kind of "by default".
But once the exercise is on the collection, you can edit the number of points for the exercise in the collection independently, without any effect on "points by default" as represented by the number here.
That being said... How many "default points" should you associate with an exercise upon creation?
As with difficulty, there is no straight forward and generally accepted way.
But as a guideline, we tend to give as many points by default as there are mathematical steps to do in the exercise.
Again, very vague... But the number should kind of represent the "work" required.
About difficulty...
We associate a certain difficulty with each exercise.
When you click an exercise into a collection, this number will be taken as difficulty for the exercise, kind of "by default".
But once the exercise is on the collection, you can edit its difficulty in the collection independently, without any effect on the "difficulty by default" here.
Why we use chess pieces? Well... we like chess, we like playing around with \(\LaTeX\)-fonts, we wanted symbols that need less space than six stars in a table-column... But in your layouts, you are of course free to indicate the difficulty of the exercise the way you want.
That being said... How "difficult" is an exercise? It depends on many factors, like what was being taught etc.
In physics exercises, we try to follow this pattern:
Level 1 - One formula (one you would find in a reference book) is enough to solve the exercise. Example exercise
Level 2 - Two formulas are needed, it's possible to compute an "in-between" solution, i.e. no algebraic equation needed. Example exercise
Level 3 - "Chain-computations" like on level 2, but 3+ calculations. Still, no equations, i.e. you are not forced to solve it in an algebraic manner. Example exercise
Level 4 - Exercise needs to be solved by algebraic equations, not possible to calculate numerical "in-between" results. Example exercise
Level 5 -
Level 6 -
When you click an exercise into a collection, this number will be taken as difficulty for the exercise, kind of "by default".
But once the exercise is on the collection, you can edit its difficulty in the collection independently, without any effect on the "difficulty by default" here.
Why we use chess pieces? Well... we like chess, we like playing around with \(\LaTeX\)-fonts, we wanted symbols that need less space than six stars in a table-column... But in your layouts, you are of course free to indicate the difficulty of the exercise the way you want.
That being said... How "difficult" is an exercise? It depends on many factors, like what was being taught etc.
In physics exercises, we try to follow this pattern:
Level 1 - One formula (one you would find in a reference book) is enough to solve the exercise. Example exercise
Level 2 - Two formulas are needed, it's possible to compute an "in-between" solution, i.e. no algebraic equation needed. Example exercise
Level 3 - "Chain-computations" like on level 2, but 3+ calculations. Still, no equations, i.e. you are not forced to solve it in an algebraic manner. Example exercise
Level 4 - Exercise needs to be solved by algebraic equations, not possible to calculate numerical "in-between" results. Example exercise
Level 5 -
Level 6 -
Question
Solution
Short
Video
\(\LaTeX\)
No explanation / solution video to this exercise has yet been created.
Visit our YouTube-Channel to see solutions to other exercises.
Don't forget to subscribe to our channel, like the videos and leave comments!
Visit our YouTube-Channel to see solutions to other exercises.
Don't forget to subscribe to our channel, like the videos and leave comments!
Exercise:
Let V be a finite dimensional vector space over mathbbC. Then every Tin textEndV is trigonalizable.
Solution:
Proof of Longrightarrow. If exists a basis mathcalB s.t. T_mathcalB^mathcalB pmatrix lambda_ & & & * & lambda_ & & & & ddots & & & &lambda_n pmatrix. then P_TxtextdetleftT_mathcalB^mathcalB-x Irighttextdet pmatrix lambda_-x & & & * & lambda_-x & & & & ddots & & & &lambda_n-x pmatrix lambda_-x...lambda_n-x. Consider T_A:K^nlongrightarrow K^n T_AuA u. By the induction hypothesis exists a basis mathcalC of K^n s.t. T_A_mathcalC^mathcalC pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. Denote by epsilon the standard basis of K^n. T_A_mathcalC^mathcalCid_mathcalC^epsilon T_A_epsilon^epsilon id_epsilon^mathcalC Q^-AQ. So Q^-AQ pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. Define P: pmatrix & & hdots & & & & vdots & & Q & & & & & pmatrixin M_n+timesn+K. claim. abcliste abc P is invertible and P^- pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix. abc P^-MP pmatrix lambda_ & *' & hdots & *' & lambda_ & & * vdots & & ddots & & & & lambda_n+ pmatrix. abcliste Proof of the claim. Direct calculation. abcliste abc P P^- pmatrix & & hdots & & & & vdots & & Q Q^- & & & & & pmatrix pmatrix & & hdots & & & & vdots & & I_n & & & & & pmatrix and similarly for P^- P. abc P^-MP pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & & pmatrix pmatrix & & hdots & & & & vdots & & Q & & & & & pmatrix pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & & & vdots & & AQ & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & & & vdots & & Q^-AQ & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & lambda_ & & * vdots & & ddots & & & & lambda_n+ pmatrix. This proves the claim. abcliste We continue with the proof of the theorem. Recall we have a basis mathcalB' for V s.t. T_mathcalB'^mathcalB' pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & pmatrixM . And we found an invertible matrix P s.t. P^-MP pmatrix lambda_ & & & * & lambda_ & & & & ddots & & & & lambda_n+ pmatrix Claim. exists basis mathcalB for V s.t. id_mathcalB'^mathcalBP. Proof of the claim. Write P pmatrix P_ & P_ & hdots & P_k & hdots & P_m vdots & vdots & hdots & vdots & hdots & vdots P_m & P_m & hdots & P_mk & hdots & P_mm pmatrix mtextdimVn+. Write mathcalB'u_u_...u_m. Define v_&:P_u_+P_u_+...+P_mu_m v_&:P_u_+P_u_+...+P_mu_m &vdots v_k&:P_ku_+P_ku_+...+P_mku_m &vdots v_m&:P_mu_+P_mu_+...+P_mmu_m exr. show that mathcalB:v_...v_m is a basis for V. H. Since textdimVm it's enough to show v_...v_m are linearly indepent. Now suppose c_v_+...+c_mv_m ** substitute * ** use the fact that u_...u_m are linearly indepent and obtain that P pmatrix c_ vdots c_m pmatrix pmatrix vdots pmatrix. But P is invertible Longrightarrow pmatrix c_ vdots c_m pmatrix pmatrix vdots pmatrix. . We are done with the claim. We now have T_mathcalB^mathcalBid_mathcalB^mathcalB'T_mathcalB'^mathcalB'id_mathcalB'^mathcalB pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. bf An alternative and more elegant proof of Longleftarrow of the theorem Proof of Longleftarrow. We'll use induction on ntextdimV and also quotient spaces. Recall: If Usubseteq V is a linear subspace we have the quotient space V/U whose elements are equivalence classes v of vin V where vv' if v-v'in U. Recall also that if V is finite dimensional then so is V/U and textdimV/UtextdimV-textdimU. Now to the proof. n: In this case V is one dimensional and T_mathcalB^mathcalBlambda so T is even diagonalizable. Let ngeq . Ase that Longleftarrow of the theorem holds for all vector spaces of textdimVleq n and every omorphism of such a space. Let V ve an n+-dimensional vector space and Tin textEndV be s.t. P_Tx splits as a product of linear factors. Let lambdain K be a zero of P_Tx Longrightarrow exists an eigenvector v_in textKerT-lambda id. Ext v_ to a basis mathcalB_v_w_...w_n+ of V. Then T_mathcalB_^mathcalB_ pmatrix lambda_ & & & & & & vdots & & * & & & & pmatrix pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & pmatrix We have P_Txlambda_-x P_Ax. Define U:textSpv_ and consider V/U. Since TUsubseteq U we obtain a new omorphism overlineT:V/Ulongrightarrow V/U defined by overlineTv:Tv. The vectors w_...w_n+ form a basis for V/U and in this basis overlineT is represented by A. So P_Txlambda_-x P_overlineTx. Since P_Tx splits o a product of linear factors so does also P_overlineTx. By induction hypothesis exists a basis mathcalCz_...z_n+ for V/U s.t. overlineT_mathcalC^mathcalC pmatrix lambda_ & & * & ddots & & & lambdan+ pmatrix. Now each element z_i in mathcalC is represented by some w_iin V i.e. z_iv_i. Such a v_i is not unique but we pick one for every i. Define mathcalBv_...v_n+. We claim that mathcalB is a basis for V. Indeed v_...v_n are linearly indepent because if alpha_v_+alpha_v_+...+alpha_n+v_n+ Longrightarrow alpha_v_+...+alpha_n+v_n+ -alpha_v_in U Longrightarrow alpha_v_+...+alpha_n+v_n+ Longrightarrow alpha_z_+...+alpha_n+z_n+ Longrightarrow alpha_...alpha_n+. This shows that v_...v_n+ are linearly indepent. Since textdimVn+ mathcalBv_...v_n+ is a basis for V. We claim that: T_mathcalB^mathcalB pmatrix lambda_ & * & vdots & vdots & * & lambda_ & * & vdots & * & & & ddots & vdots & & & &lambda_n+ pmatrix. Indeed Tv_lambda_v_ and for leq ileq n+ we have Tv_ioverlineTv_ioverlineTz_ialpha_iz_+alpha_iz_+...+alpha_i-iz_i-+lambda_iz_i for some alpha_i...alpha_i-iin K. Longrightarrow Tv_ialpha_iv_+...+alpha_i-iv_i-+lambda_iv_i Longrightarrow Tv_ialpha_iv_+...+alpha_i-iv_i-+lambda_iv_i+u for some uin U. Write ubeta_i v_. Longrightarrow Tv_ibeta_iv_+alpha_iv_+...+alpha_i-iv_i-+lambda_iv_i. This shows T is trigonalizable. The induction is now complete.
Let V be a finite dimensional vector space over mathbbC. Then every Tin textEndV is trigonalizable.
Solution:
Proof of Longrightarrow. If exists a basis mathcalB s.t. T_mathcalB^mathcalB pmatrix lambda_ & & & * & lambda_ & & & & ddots & & & &lambda_n pmatrix. then P_TxtextdetleftT_mathcalB^mathcalB-x Irighttextdet pmatrix lambda_-x & & & * & lambda_-x & & & & ddots & & & &lambda_n-x pmatrix lambda_-x...lambda_n-x. Consider T_A:K^nlongrightarrow K^n T_AuA u. By the induction hypothesis exists a basis mathcalC of K^n s.t. T_A_mathcalC^mathcalC pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. Denote by epsilon the standard basis of K^n. T_A_mathcalC^mathcalCid_mathcalC^epsilon T_A_epsilon^epsilon id_epsilon^mathcalC Q^-AQ. So Q^-AQ pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. Define P: pmatrix & & hdots & & & & vdots & & Q & & & & & pmatrixin M_n+timesn+K. claim. abcliste abc P is invertible and P^- pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix. abc P^-MP pmatrix lambda_ & *' & hdots & *' & lambda_ & & * vdots & & ddots & & & & lambda_n+ pmatrix. abcliste Proof of the claim. Direct calculation. abcliste abc P P^- pmatrix & & hdots & & & & vdots & & Q Q^- & & & & & pmatrix pmatrix & & hdots & & & & vdots & & I_n & & & & & pmatrix and similarly for P^- P. abc P^-MP pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & & pmatrix pmatrix & & hdots & & & & vdots & & Q & & & & & pmatrix pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & & & vdots & & AQ & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & & & vdots & & Q^-AQ & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & lambda_ & & * vdots & & ddots & & & & lambda_n+ pmatrix. This proves the claim. abcliste We continue with the proof of the theorem. Recall we have a basis mathcalB' for V s.t. T_mathcalB'^mathcalB' pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & pmatrixM . And we found an invertible matrix P s.t. P^-MP pmatrix lambda_ & & & * & lambda_ & & & & ddots & & & & lambda_n+ pmatrix Claim. exists basis mathcalB for V s.t. id_mathcalB'^mathcalBP. Proof of the claim. Write P pmatrix P_ & P_ & hdots & P_k & hdots & P_m vdots & vdots & hdots & vdots & hdots & vdots P_m & P_m & hdots & P_mk & hdots & P_mm pmatrix mtextdimVn+. Write mathcalB'u_u_...u_m. Define v_&:P_u_+P_u_+...+P_mu_m v_&:P_u_+P_u_+...+P_mu_m &vdots v_k&:P_ku_+P_ku_+...+P_mku_m &vdots v_m&:P_mu_+P_mu_+...+P_mmu_m exr. show that mathcalB:v_...v_m is a basis for V. H. Since textdimVm it's enough to show v_...v_m are linearly indepent. Now suppose c_v_+...+c_mv_m ** substitute * ** use the fact that u_...u_m are linearly indepent and obtain that P pmatrix c_ vdots c_m pmatrix pmatrix vdots pmatrix. But P is invertible Longrightarrow pmatrix c_ vdots c_m pmatrix pmatrix vdots pmatrix. . We are done with the claim. We now have T_mathcalB^mathcalBid_mathcalB^mathcalB'T_mathcalB'^mathcalB'id_mathcalB'^mathcalB pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. bf An alternative and more elegant proof of Longleftarrow of the theorem Proof of Longleftarrow. We'll use induction on ntextdimV and also quotient spaces. Recall: If Usubseteq V is a linear subspace we have the quotient space V/U whose elements are equivalence classes v of vin V where vv' if v-v'in U. Recall also that if V is finite dimensional then so is V/U and textdimV/UtextdimV-textdimU. Now to the proof. n: In this case V is one dimensional and T_mathcalB^mathcalBlambda so T is even diagonalizable. Let ngeq . Ase that Longleftarrow of the theorem holds for all vector spaces of textdimVleq n and every omorphism of such a space. Let V ve an n+-dimensional vector space and Tin textEndV be s.t. P_Tx splits as a product of linear factors. Let lambdain K be a zero of P_Tx Longrightarrow exists an eigenvector v_in textKerT-lambda id. Ext v_ to a basis mathcalB_v_w_...w_n+ of V. Then T_mathcalB_^mathcalB_ pmatrix lambda_ & & & & & & vdots & & * & & & & pmatrix pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & pmatrix We have P_Txlambda_-x P_Ax. Define U:textSpv_ and consider V/U. Since TUsubseteq U we obtain a new omorphism overlineT:V/Ulongrightarrow V/U defined by overlineTv:Tv. The vectors w_...w_n+ form a basis for V/U and in this basis overlineT is represented by A. So P_Txlambda_-x P_overlineTx. Since P_Tx splits o a product of linear factors so does also P_overlineTx. By induction hypothesis exists a basis mathcalCz_...z_n+ for V/U s.t. overlineT_mathcalC^mathcalC pmatrix lambda_ & & * & ddots & & & lambdan+ pmatrix. Now each element z_i in mathcalC is represented by some w_iin V i.e. z_iv_i. Such a v_i is not unique but we pick one for every i. Define mathcalBv_...v_n+. We claim that mathcalB is a basis for V. Indeed v_...v_n are linearly indepent because if alpha_v_+alpha_v_+...+alpha_n+v_n+ Longrightarrow alpha_v_+...+alpha_n+v_n+ -alpha_v_in U Longrightarrow alpha_v_+...+alpha_n+v_n+ Longrightarrow alpha_z_+...+alpha_n+z_n+ Longrightarrow alpha_...alpha_n+. This shows that v_...v_n+ are linearly indepent. Since textdimVn+ mathcalBv_...v_n+ is a basis for V. We claim that: T_mathcalB^mathcalB pmatrix lambda_ & * & vdots & vdots & * & lambda_ & * & vdots & * & & & ddots & vdots & & & &lambda_n+ pmatrix. Indeed Tv_lambda_v_ and for leq ileq n+ we have Tv_ioverlineTv_ioverlineTz_ialpha_iz_+alpha_iz_+...+alpha_i-iz_i-+lambda_iz_i for some alpha_i...alpha_i-iin K. Longrightarrow Tv_ialpha_iv_+...+alpha_i-iv_i-+lambda_iv_i Longrightarrow Tv_ialpha_iv_+...+alpha_i-iv_i-+lambda_iv_i+u for some uin U. Write ubeta_i v_. Longrightarrow Tv_ibeta_iv_+alpha_iv_+...+alpha_i-iv_i-+lambda_iv_i. This shows T is trigonalizable. The induction is now complete.
Meta Information
Exercise:
Let V be a finite dimensional vector space over mathbbC. Then every Tin textEndV is trigonalizable.
Solution:
Proof of Longrightarrow. If exists a basis mathcalB s.t. T_mathcalB^mathcalB pmatrix lambda_ & & & * & lambda_ & & & & ddots & & & &lambda_n pmatrix. then P_TxtextdetleftT_mathcalB^mathcalB-x Irighttextdet pmatrix lambda_-x & & & * & lambda_-x & & & & ddots & & & &lambda_n-x pmatrix lambda_-x...lambda_n-x. Consider T_A:K^nlongrightarrow K^n T_AuA u. By the induction hypothesis exists a basis mathcalC of K^n s.t. T_A_mathcalC^mathcalC pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. Denote by epsilon the standard basis of K^n. T_A_mathcalC^mathcalCid_mathcalC^epsilon T_A_epsilon^epsilon id_epsilon^mathcalC Q^-AQ. So Q^-AQ pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. Define P: pmatrix & & hdots & & & & vdots & & Q & & & & & pmatrixin M_n+timesn+K. claim. abcliste abc P is invertible and P^- pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix. abc P^-MP pmatrix lambda_ & *' & hdots & *' & lambda_ & & * vdots & & ddots & & & & lambda_n+ pmatrix. abcliste Proof of the claim. Direct calculation. abcliste abc P P^- pmatrix & & hdots & & & & vdots & & Q Q^- & & & & & pmatrix pmatrix & & hdots & & & & vdots & & I_n & & & & & pmatrix and similarly for P^- P. abc P^-MP pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & & pmatrix pmatrix & & hdots & & & & vdots & & Q & & & & & pmatrix pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & & & vdots & & AQ & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & & & vdots & & Q^-AQ & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & lambda_ & & * vdots & & ddots & & & & lambda_n+ pmatrix. This proves the claim. abcliste We continue with the proof of the theorem. Recall we have a basis mathcalB' for V s.t. T_mathcalB'^mathcalB' pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & pmatrixM . And we found an invertible matrix P s.t. P^-MP pmatrix lambda_ & & & * & lambda_ & & & & ddots & & & & lambda_n+ pmatrix Claim. exists basis mathcalB for V s.t. id_mathcalB'^mathcalBP. Proof of the claim. Write P pmatrix P_ & P_ & hdots & P_k & hdots & P_m vdots & vdots & hdots & vdots & hdots & vdots P_m & P_m & hdots & P_mk & hdots & P_mm pmatrix mtextdimVn+. Write mathcalB'u_u_...u_m. Define v_&:P_u_+P_u_+...+P_mu_m v_&:P_u_+P_u_+...+P_mu_m &vdots v_k&:P_ku_+P_ku_+...+P_mku_m &vdots v_m&:P_mu_+P_mu_+...+P_mmu_m exr. show that mathcalB:v_...v_m is a basis for V. H. Since textdimVm it's enough to show v_...v_m are linearly indepent. Now suppose c_v_+...+c_mv_m ** substitute * ** use the fact that u_...u_m are linearly indepent and obtain that P pmatrix c_ vdots c_m pmatrix pmatrix vdots pmatrix. But P is invertible Longrightarrow pmatrix c_ vdots c_m pmatrix pmatrix vdots pmatrix. . We are done with the claim. We now have T_mathcalB^mathcalBid_mathcalB^mathcalB'T_mathcalB'^mathcalB'id_mathcalB'^mathcalB pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. bf An alternative and more elegant proof of Longleftarrow of the theorem Proof of Longleftarrow. We'll use induction on ntextdimV and also quotient spaces. Recall: If Usubseteq V is a linear subspace we have the quotient space V/U whose elements are equivalence classes v of vin V where vv' if v-v'in U. Recall also that if V is finite dimensional then so is V/U and textdimV/UtextdimV-textdimU. Now to the proof. n: In this case V is one dimensional and T_mathcalB^mathcalBlambda so T is even diagonalizable. Let ngeq . Ase that Longleftarrow of the theorem holds for all vector spaces of textdimVleq n and every omorphism of such a space. Let V ve an n+-dimensional vector space and Tin textEndV be s.t. P_Tx splits as a product of linear factors. Let lambdain K be a zero of P_Tx Longrightarrow exists an eigenvector v_in textKerT-lambda id. Ext v_ to a basis mathcalB_v_w_...w_n+ of V. Then T_mathcalB_^mathcalB_ pmatrix lambda_ & & & & & & vdots & & * & & & & pmatrix pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & pmatrix We have P_Txlambda_-x P_Ax. Define U:textSpv_ and consider V/U. Since TUsubseteq U we obtain a new omorphism overlineT:V/Ulongrightarrow V/U defined by overlineTv:Tv. The vectors w_...w_n+ form a basis for V/U and in this basis overlineT is represented by A. So P_Txlambda_-x P_overlineTx. Since P_Tx splits o a product of linear factors so does also P_overlineTx. By induction hypothesis exists a basis mathcalCz_...z_n+ for V/U s.t. overlineT_mathcalC^mathcalC pmatrix lambda_ & & * & ddots & & & lambdan+ pmatrix. Now each element z_i in mathcalC is represented by some w_iin V i.e. z_iv_i. Such a v_i is not unique but we pick one for every i. Define mathcalBv_...v_n+. We claim that mathcalB is a basis for V. Indeed v_...v_n are linearly indepent because if alpha_v_+alpha_v_+...+alpha_n+v_n+ Longrightarrow alpha_v_+...+alpha_n+v_n+ -alpha_v_in U Longrightarrow alpha_v_+...+alpha_n+v_n+ Longrightarrow alpha_z_+...+alpha_n+z_n+ Longrightarrow alpha_...alpha_n+. This shows that v_...v_n+ are linearly indepent. Since textdimVn+ mathcalBv_...v_n+ is a basis for V. We claim that: T_mathcalB^mathcalB pmatrix lambda_ & * & vdots & vdots & * & lambda_ & * & vdots & * & & & ddots & vdots & & & &lambda_n+ pmatrix. Indeed Tv_lambda_v_ and for leq ileq n+ we have Tv_ioverlineTv_ioverlineTz_ialpha_iz_+alpha_iz_+...+alpha_i-iz_i-+lambda_iz_i for some alpha_i...alpha_i-iin K. Longrightarrow Tv_ialpha_iv_+...+alpha_i-iv_i-+lambda_iv_i Longrightarrow Tv_ialpha_iv_+...+alpha_i-iv_i-+lambda_iv_i+u for some uin U. Write ubeta_i v_. Longrightarrow Tv_ibeta_iv_+alpha_iv_+...+alpha_i-iv_i-+lambda_iv_i. This shows T is trigonalizable. The induction is now complete.
Let V be a finite dimensional vector space over mathbbC. Then every Tin textEndV is trigonalizable.
Solution:
Proof of Longrightarrow. If exists a basis mathcalB s.t. T_mathcalB^mathcalB pmatrix lambda_ & & & * & lambda_ & & & & ddots & & & &lambda_n pmatrix. then P_TxtextdetleftT_mathcalB^mathcalB-x Irighttextdet pmatrix lambda_-x & & & * & lambda_-x & & & & ddots & & & &lambda_n-x pmatrix lambda_-x...lambda_n-x. Consider T_A:K^nlongrightarrow K^n T_AuA u. By the induction hypothesis exists a basis mathcalC of K^n s.t. T_A_mathcalC^mathcalC pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. Denote by epsilon the standard basis of K^n. T_A_mathcalC^mathcalCid_mathcalC^epsilon T_A_epsilon^epsilon id_epsilon^mathcalC Q^-AQ. So Q^-AQ pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. Define P: pmatrix & & hdots & & & & vdots & & Q & & & & & pmatrixin M_n+timesn+K. claim. abcliste abc P is invertible and P^- pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix. abc P^-MP pmatrix lambda_ & *' & hdots & *' & lambda_ & & * vdots & & ddots & & & & lambda_n+ pmatrix. abcliste Proof of the claim. Direct calculation. abcliste abc P P^- pmatrix & & hdots & & & & vdots & & Q Q^- & & & & & pmatrix pmatrix & & hdots & & & & vdots & & I_n & & & & & pmatrix and similarly for P^- P. abc P^-MP pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & & pmatrix pmatrix & & hdots & & & & vdots & & Q & & & & & pmatrix pmatrix & & hdots & & & & vdots & & Q^- & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & & & vdots & & AQ & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & & & vdots & & Q^-AQ & & & & & pmatrix pmatrix lambda_ & *' & hdots & *' & lambda_ & & * vdots & & ddots & & & & lambda_n+ pmatrix. This proves the claim. abcliste We continue with the proof of the theorem. Recall we have a basis mathcalB' for V s.t. T_mathcalB'^mathcalB' pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & pmatrixM . And we found an invertible matrix P s.t. P^-MP pmatrix lambda_ & & & * & lambda_ & & & & ddots & & & & lambda_n+ pmatrix Claim. exists basis mathcalB for V s.t. id_mathcalB'^mathcalBP. Proof of the claim. Write P pmatrix P_ & P_ & hdots & P_k & hdots & P_m vdots & vdots & hdots & vdots & hdots & vdots P_m & P_m & hdots & P_mk & hdots & P_mm pmatrix mtextdimVn+. Write mathcalB'u_u_...u_m. Define v_&:P_u_+P_u_+...+P_mu_m v_&:P_u_+P_u_+...+P_mu_m &vdots v_k&:P_ku_+P_ku_+...+P_mku_m &vdots v_m&:P_mu_+P_mu_+...+P_mmu_m exr. show that mathcalB:v_...v_m is a basis for V. H. Since textdimVm it's enough to show v_...v_m are linearly indepent. Now suppose c_v_+...+c_mv_m ** substitute * ** use the fact that u_...u_m are linearly indepent and obtain that P pmatrix c_ vdots c_m pmatrix pmatrix vdots pmatrix. But P is invertible Longrightarrow pmatrix c_ vdots c_m pmatrix pmatrix vdots pmatrix. . We are done with the claim. We now have T_mathcalB^mathcalBid_mathcalB^mathcalB'T_mathcalB'^mathcalB'id_mathcalB'^mathcalB pmatrix lambda_ & & * & ddots & & &lambda_n+ pmatrix. bf An alternative and more elegant proof of Longleftarrow of the theorem Proof of Longleftarrow. We'll use induction on ntextdimV and also quotient spaces. Recall: If Usubseteq V is a linear subspace we have the quotient space V/U whose elements are equivalence classes v of vin V where vv' if v-v'in U. Recall also that if V is finite dimensional then so is V/U and textdimV/UtextdimV-textdimU. Now to the proof. n: In this case V is one dimensional and T_mathcalB^mathcalBlambda so T is even diagonalizable. Let ngeq . Ase that Longleftarrow of the theorem holds for all vector spaces of textdimVleq n and every omorphism of such a space. Let V ve an n+-dimensional vector space and Tin textEndV be s.t. P_Tx splits as a product of linear factors. Let lambdain K be a zero of P_Tx Longrightarrow exists an eigenvector v_in textKerT-lambda id. Ext v_ to a basis mathcalB_v_w_...w_n+ of V. Then T_mathcalB_^mathcalB_ pmatrix lambda_ & & & & & & vdots & & * & & & & pmatrix pmatrix lambda_ & * & hdots & * & & & vdots & & A & & & & pmatrix We have P_Txlambda_-x P_Ax. Define U:textSpv_ and consider V/U. Since TUsubseteq U we obtain a new omorphism overlineT:V/Ulongrightarrow V/U defined by overlineTv:Tv. The vectors w_...w_n+ form a basis for V/U and in this basis overlineT is represented by A. So P_Txlambda_-x P_overlineTx. Since P_Tx splits o a product of linear factors so does also P_overlineTx. By induction hypothesis exists a basis mathcalCz_...z_n+ for V/U s.t. overlineT_mathcalC^mathcalC pmatrix lambda_ & & * & ddots & & & lambdan+ pmatrix. Now each element z_i in mathcalC is represented by some w_iin V i.e. z_iv_i. Such a v_i is not unique but we pick one for every i. Define mathcalBv_...v_n+. We claim that mathcalB is a basis for V. Indeed v_...v_n are linearly indepent because if alpha_v_+alpha_v_+...+alpha_n+v_n+ Longrightarrow alpha_v_+...+alpha_n+v_n+ -alpha_v_in U Longrightarrow alpha_v_+...+alpha_n+v_n+ Longrightarrow alpha_z_+...+alpha_n+z_n+ Longrightarrow alpha_...alpha_n+. This shows that v_...v_n+ are linearly indepent. Since textdimVn+ mathcalBv_...v_n+ is a basis for V. We claim that: T_mathcalB^mathcalB pmatrix lambda_ & * & vdots & vdots & * & lambda_ & * & vdots & * & & & ddots & vdots & & & &lambda_n+ pmatrix. Indeed Tv_lambda_v_ and for leq ileq n+ we have Tv_ioverlineTv_ioverlineTz_ialpha_iz_+alpha_iz_+...+alpha_i-iz_i-+lambda_iz_i for some alpha_i...alpha_i-iin K. Longrightarrow Tv_ialpha_iv_+...+alpha_i-iv_i-+lambda_iv_i Longrightarrow Tv_ialpha_iv_+...+alpha_i-iv_i-+lambda_iv_i+u for some uin U. Write ubeta_i v_. Longrightarrow Tv_ibeta_iv_+alpha_iv_+...+alpha_i-iv_i-+lambda_iv_i. This shows T is trigonalizable. The induction is now complete.
Contained in these collections: