Matrix inequalities and partial isometries that arise in x-ray crystallography (Q677918)

The paper considers matrices of the form \(P=(X^TX)/M\), where \(X\in\mathbb{R}^{M\times n}\) has rows \(X_i\), with \(1\leq i\leq M\) such that \(|X_jv|\leq c\) and \(|X_i|=0\) for some given \(c\in(0,1)\) and \(v\in\mathbb{R}^n\). The authors extend the results of \textit{B. Ortner} and \textit{A. R. Kräuter} [Linear Algebra Appl. 236, 147-180 (1996; Zbl 0846.15004)] pertaining to the lower bounds for \(\det(X^TX)^{-1}\) and \(\text{tr}(X^TX)^{-1}\), to a set of inequalities involving eigenvalues of \((X^TX)\). The investigation also includes matrices \(X\) that minimise \(\det(X^TX)^{-1}\) and \(\text{tr}(X^TX)^{-1}\) and the constructions of optimal matrices.