I have a sparse matrix A, and need to solve the system (AtA)X = B. Right now I'm multiplying At by A and using AtA to solve the system. Is there a faster alternative (like passing A to UMFPACK and have it find the factorization quicker)?
A is rectangular by the way.
