
#1
Feb2306, 10:21 AM

P: 66

I have matrix A
[tex] \left(\begin{array}{ccc}6&2&2\\2&2&2\\2&2&2\end{array} \right) [/tex] Its characteristic polynomial is [tex] p(\lambda)=\lambda^3  10\lambda^2 + 32\lambda 32 [/tex] Finding minimal polynomial i get: [tex] (I\lambdaA)^\vee=\left(\begin{array}{ccc}\lambda6&2&2\\2&\lambda2&2\\2&2&\lambda2\end{array}\right)^\vee [/tex] I cant understand why this last result equals with [tex] \left(\begin{array}{ccc}\lambda^24\lambda&2\lambda+8&2\lambda8\\2\lambda8&\lambda^28\lambda+16&2\lambda8\\2\lambda+8&2\lambda8&\lambda^28\lambda+16\end{array} \right) [/tex] can someone explain? 



#2
Feb2306, 10:26 AM

Sci Advisor
HW Helper
P: 9,398

would you mind retypsetting that so it is more legible?




#3
Feb2306, 10:30 AM

Sci Advisor
HW Helper
P: 9,398

Oh, and if you check that the char poly has 3 distinct roots it will be its minimal poly. other than that, it is of the form (xa)^2(xb) a different from b or (xa)^3, check if those happen and if they do the minimal polys are divisors of these and you can check which, A satisfies.




#4
Feb2306, 10:57 AM

Sci Advisor
HW Helper
P: 2,589

Minimal polynomial
Also, you can observe right away that (0 1 0)^{T} is an eigenvector with eigenvalue 2 (look at the middle column of A). So 2 is a root of your characteristic polynomial, which will make it easier for you to factor.




#5
Feb2406, 04:04 AM

P: 66

I guess im on my own again.




#6
Feb2406, 06:25 AM

Math
Emeritus
Sci Advisor
Thanks
PF Gold
P: 38,881

Perhaps it would help if you explain what the "v" in
[tex] (I\lambdaA)^\vee[/tex] means! In any case, it was explained to you that the minimal polynomial the polynomial of smallest degree that A satisfies is just the characteristic polynomial with duplicate factors removed. And it was pointed out that [itex]\lambda 2[/itex] is a factor so it should be simple to determine all the factors. What more do you want? 



#7
Feb2406, 01:26 PM

Sci Advisor
HW Helper
P: 2,589

If you have a new question, make a new post or a new thread, don't edit an old post which every one thinks has already been answered. Moreover, it appears you only gave it less than 5 hours before complaining that "you're on your own again." Be more patient. 



#8
Feb2406, 01:42 PM

Sci Advisor
HW Helper
P: 9,398

That in my mind is a little misleading. It is the characteristic poly with 'unnecessary' duplicate factors removed in some sense. 



#9
Feb2506, 07:13 AM

P: 66

I have matrix A
[tex] \left(\begin{array}{ccc}6&2&2\\2&2&2\\2&2&2\end{array} \right) [/tex] Its characteristic polynomial is [tex] p(\lambda)=\lambda^3  10\lambda^2 + 32\lambda 32[/tex] when i factor it i get [tex]p(\lambda)=(2\lambda)(\lambda4)^2 [/tex] So removing the duplicate factor [tex](\lambda4)[/tex] i get that minimal polynomial is [tex](\psi)=(2\lambda)(\lambda 4)[/tex] and A really satisfies it. But lets assume I have matrix B [tex]\left(\begin{array}{ccc}6&2&2\\2&2&0\\0&0&2\end{array} \right)[/tex] Its characteristic polynomial is the same as for A and following the same method as HallsOfIvy pointed out i get the same [tex]p(\psi)[/tex] for B as for A but when u put B into the equation you dont get 0 so this method doesnt work. 



#10
Feb2506, 08:05 AM

Sci Advisor
HW Helper
P: 9,398

Erm, no, as was pointed out you cannot *just* remove factors and assume that the minimal poly is the product of the distinct linear factors. All this tells you if indeed your manipulations are correct is that the minimal poly and the characterstic poly are the same. The method is correct, and has given you the minimal poly, it is just that you don't recognize this.
This was what I wished to clarify from HallsofIvy's post. The minimal poly is not the product of the distinct linear factors but the product of the distinct linear factors with the smallest multiplicity of each possible so that the matrix satisfies the polynomial. 



#11
Feb2506, 09:36 AM

Sci Advisor
HW Helper
P: 9,421

to illustrate matt's point about repeated factors in mminimal polynomials, consider the derivative operator D, acting on functions. on the space spanned by the functions e^x and xe^x, the minimal polynomial is (D1)^2. while on the span of e^x, it is just D1.
expressed in matrices, i.e. stripped of all meaning, for the ordered basis e^x, xe^x, we get a matrix of D with columns [ 1 0]^t and [1 1]^t, where ^t means transpose. the point is that the matrix is not disgonalizable, but is a sum of a diagonal matrix (namely the identity matrix), and a "nilpotent matrix", here the matrix which is all zeroes except for a 1 in the upper right corner. It is the rpoesence of this nilpotent summand which prevents the minimal polynomial from having distinct irreducible factors. or think about the map D, on the space of polynomials of degree at most n1. The eigenvalues are all zero, and the characteristic polynomial is D^n, which is also the minimal polynomial. 


Register to reply 
Related Discussions  
Minimal Polynomial A nxn Matrix  Calculus & Beyond Homework  5  
a question on minimal polynomial (LA)  Calculus & Beyond Homework  5  
Irreducible polynomial on polynomial ring  Calculus & Beyond Homework  1  
Matrix Minimal Polynomial  Linear & Abstract Algebra  2  
Minimal and characteristic polynomial  Linear & Abstract Algebra  4 