Dear all,
I want to compute eigenvector centrality (a network measure). I have obtained help on the Mata forum concerning the matrix computation (see http://www.statalist.org/forums/foru...tor-centrality). However, I am a little stuck on solving a linear equation (see attached, I tried and failed with paste from word, or upload an image)
i,j are subscripts in the formula
In this formula, lambda is constant (already computed), I,j are industry pairs, Aij are elements in a matrix A for row i and column j. In this equation, we sum over all the j’s that are not equal to i. Hence, to compute Ci, we have to solve multiple equations recursively.
If you could suggest some routines or algorithm for computing it, that will be greatly appreciated. If you need me to further clarify the problem, please let me know.
P.S. I posted on general forum assuming solving this equation does not depend on mata .
Thanks so much,
Rochelle
I want to compute eigenvector centrality (a network measure). I have obtained help on the Mata forum concerning the matrix computation (see http://www.statalist.org/forums/foru...tor-centrality). However, I am a little stuck on solving a linear equation (see attached, I tried and failed with paste from word, or upload an image)
i,j are subscripts in the formula
In this formula, lambda is constant (already computed), I,j are industry pairs, Aij are elements in a matrix A for row i and column j. In this equation, we sum over all the j’s that are not equal to i. Hence, to compute Ci, we have to solve multiple equations recursively.
If you could suggest some routines or algorithm for computing it, that will be greatly appreciated. If you need me to further clarify the problem, please let me know.
P.S. I posted on general forum assuming solving this equation does not depend on mata .
Thanks so much,
Rochelle
