I would like to calculate the Adjugate matrix of a given matrix $A$, and its updates in the diagonal: $B=A-\lambda I$, where $I$ is the identity matrix, $\lambda$ is a scalar. To this end, I am using the algorithm explained here, using a decomposition $A=XDY$. Unfortunately, the matrix $A$ or $B$ could be singular.

Is it possible to obtain a suitable decomposition such, that for the decomposition of $B$ I could save computer time by reusing $X$, $D$, or $Y$ from the decomposition of $A$?

  • $\begingroup$ One attempt that I have made is that I perform SVD on A'=A+I in order to remove the null-space, and decompose it as A'=X'D'Y'. Then I calculate D"=abs(D'-I) as my matrices are positive semi-definite, symmetric and real. Numerical tests in python are provided spot-on results for the adjugate using X', D" (or D"-lambda*I), and Y' but I would like to see the limits of this approach. Any suggestions? $\endgroup$ Commented Nov 23, 2020 at 15:46


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.