Completely positive factorizations associated with Euclidean distance matrices corresponding to an arithmetic progression (Q2174096)

An \(n \times n\) real symmetric matrix \(A\) is completely positive if \(A = BB^T\) for some \(n \times k\) entry-wise non-negative matrix \(B\). The minimum number of columns \(k\) in such a non-negative factor \(B\) is called the \(cp\)-\(rank\) of \(A\). For a set of distinct real numbers \(\mathcal{S} = \{a_1, a_2,\dots, a_n\}\), the authors define the Euclidean distance matrix to be the matrix \(\mathrm{EDM}(\mathcal{S}) = (a_{ij})^n_{i,j=1}\) with \(a_{ij} = (a_j - a_i)^2\) . The matrix \(\mathrm{EDM}(\mathcal{S})\) is entry-wise non-negative, and has rank 3. The authors consider a family of Euclidean distance matrices that correspond to an arithmetic progression, i.e., \(\mathcal{S} = \{a, a + d, \ldots, a + (n - 1)d\}\) for some positive real numbers \(a\) and \(d\), and in particular for \(\mathcal{S} = \{1, 2, ..., n\}\). Let \(A_n :=\mathrm{EDM}(\{1,2,...,n\})\). Then \((A_n)_{ij} = (j - i)^2\). The authors study completely positive factorizations of translations of Euclidean distance matrices corresponding to an arithmetic progression. They also study factorizations under various other conditions.