Determinant of sparse $4\times 4$ matrix.

93 Views Asked by At

Let $x$ be a real number such that $|x|<1$. Let the matrix to be checked be

$$\rho = \frac{1}{2}\begin{bmatrix} 1 & 0 & 0 & (1-x) \\ 0& 0& 0& 0 \\ 0& 0& 0& 0 \\ (1-x)& 0 & 0 & 1 \end{bmatrix}$$

I want to check that this matrix $\rho$ is positive semi - definite.

To do so I compute the eigenvalues and check to see if they are non - negative.

However, this is a sparse matrix and I am running into some trouble.

Here is my attempt:

$\rho v = \lambda v \\ \rightarrow det(\rho - \lambda I) = 0 \\ \rightarrow \frac{1}{2}(1-\lambda)(-1)^{2} \begin{vmatrix} -\lambda & 0& 0 \\ 0& -\lambda & 0&\\0& 0& (1-\lambda)& \end{vmatrix} -\frac{1}{2}\begin{vmatrix} 0& -\lambda& 0& \\ 0& 0& -\lambda& \\ (1-x) &0 &0 \end{vmatrix} =0 $

As can be observed, the 3x3 determinant can be computed again but the zeros at the top rows for each determinant are causing issues.

Any help is appreciated.

1

There are 1 best solutions below

1
On

This symmetric matrix has a $2$-dimensional kernel. To check for positive semi-definiteness, it suffices to restrict to the orthogonal complement of the kernel, which is also $2$-dimensional. Finding the eigenvalues of the restriction is relatively straight forward.