Closed
Description
From rust-ndarray/ndarray#649 (comment)
Given that a typical use case in scientific calculations is to repeatedly perform QR factorization on a matrix of the same dimensions, it would be great to have a struct QR that contains all pre-allocated data structures and comes equipped with a method compute taking a matrix A and storing the result inside it. The factors Q and R can then be accessed by reference and cloned, if needed.
We need appropriate benchmark problem before going to implement this.