In Matlab, how could I convert, if possible, this solving of generalized eigen values problem :
[V,D,W] = EIG(A,B) also produces a full matrix W whose columns are the
corresponding left eigenvectors so that `W'*A = D*W'*B eq(1)`.
Into the following classical problem, that is, to find a passing matrix W
such that :
A*W = D1*W
and B*W = D2*W
where D1
and D2
coming directly from diagonlisation of A
and B
??
I would have thought that multiplying (eq 1
) on the right by B^-1
would be enough but I wonder if I can conclude on a common basis for both matrices A
and B
.
If someone could give an advise, I would be grateful
question from:
https://stackoverflow.com/questions/65844244/from-generalized-eigen-values-problem-to-classical-eigen-values-problem 与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…