An introduction to the Equiangular algorithm

Abstract

In this paper a generalization of the Gram-Schmidt Algorithm is presented. Actually we provide an algorithm to construct a set of equiangular vectors with a given angle $\theta\in(0,\arccos(\frac{-1}{n-1}))$ using a set of input independent vectors in $\mathbb{R}^n$. Therefore a usual type of matrix decomposition is derived. Then we discuss some properties of matrices derived from the new algorithm. The inverse and eigenvalue problems of these matrices if there exist are studied. Also, we derive some canonical forms based on the algorithm.