1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

A proof needed, problem in Perron-frobenius theory

  1. Apr 18, 2010 #1
    1. The problem statement, all variables and given/known data
    Let A be an operator from V to V(finite). If the matrix of A is positive, i.e. all entries are strictly positive, and A^k=A for some integer k>1 what is the rank of A?

    2. Relevant equations



    3. The attempt at a solution

    I think that rank(A)=1 and i believe that i can prove in the case where k=2.

    Proof: if A^2=A then the eigenvalues of A are 0 and 1, which means that the perron root of A is 1. Perron-frobenius theory then states the the eigenspace of 1 has geometric mult.=algebraic mult. = 1. Furthermore since A(Ax)=(Ax) for all vectors x in V, the range is a subspace of the eigenspace corresponding to the eigenvalue 1. And since Ax=x is a subspace of the range, the range of A = the eigenspace corresponding to the eigenvalue 1, which has geom. mult = dim(range space) = rank(A)=1

    Im have trouble finding a general proof. Any help is greatly appreciated. I know I'm new here but I promise to look around and see if I can lend a hand to anyone :)
     
  2. jcsd
  3. Apr 20, 2010 #2
    Allright so I solved this myself. Here is the solution if anyone is interested:

    Let y be an eigenvalue to A. Then y^n=y which implies y^n-y=0 so y(y^(n-1)-1)=0.
    This implies either that y=0 or that y=e^(2pi(k-1)i/(n-1) ) for k=1,...,n-2. This means that all y which are not 0 has abs(y)=1 and that 1 is an eigenvalue. But this means that y=1 is the perron root of A and so by the perron theorem all other eigenvalues must be smaller in modulus, which implies that all other eigenvalues are zero.

    But y=0 has the algebraic multiplicity 1 and by the perron theorem so does y=1. Since V can be written as the direct sum of the generalized eigenspaces this means that range(A) = the eigenspace corresponding to y=1 which has dimension 1 by the perron theorem. So rank(A)=1 QED.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: A proof needed, problem in Perron-frobenius theory
Loading...