A New Much Faster and Simpler Algorithm for Lapack Dgels

We present new algorithms for computing the linear least squares solution to overde-termined linear systems and the minimum norm solution to underdetermined linear systems. For both problems, we consider the standard formulation min kAX ? BkF and the transposed formulation min kA T X ? BkF , i.e, four diierent problems in all. The functionality of our implementation corresponds to that of the LAPACK routine DGELS. The new implementation is signiicantly faster and simpler. It outperforms the LAPACK DGELS for all matrix sizes tested. The improvement is usually 50{ 100% and it is as high as 400%. The four diierent problems of DGELS are essentially reduced to two, by use of explicit transposition of A. By explicit transposition we avoid computing Householder transformations on vectors with large stride. The QR factorization of block columns of A is performed using a recursive level-3 algorithm. By interleaving updates of B with the factorization of A, we reduce the number of oating point operations performed for the linear least squares problem. By avoiding redundant computations in the update of B we reduce the work needed to compute the minimum norm solution. Finally, we outline fully recursive algorithms for the four problems of DGELS as well as for QR factorization.