This repository contains two different ways of implementing a code that solves a system of linear equations: Ax=b, where A is an NXN matrix, b is an N-vector and x is the unknown N-vector that we want to solve for.

• One implementation restricts to the case in which A is an 8x8 dimensional matrix with the last column given by A17=A27=A37=A47=A57=A67=0 (but A77!=0). The implementation simply consists of hard-coding the previously-computed expression for the solution. It takes longer to compile due to the length of the expressions.
• The other implementation employs the so-called LU factorisation algorithm, employed, for example, in LAPACK dgesv function (see here). This implementation was used to remove the need for LAPACK in the GRFolres code. It turns out that this algorithm is faster than LAPACK's one.