we want to find the solution that satisfies the linear system of equations . instead of giving you the matrix , you are provided with the svd of . assume is a square non-singular matrix. the svd is defined as: where , and are all matrices with the same shape as . the setup code provides the vector as the 1d-numpy array b, and the diagonal entries of the matrix as the 1d-numpy array sigma. instead of providing you with the matrices and , we provide the following two functions: def right multiply with u(y):