Re: [eigen] Bug(s) report: SparseQR on tall-thin matrices

[ Thread Index | Date Index | More lists.tuxfamily.org/eigen Archives ]



Take a look at https://bitbucket.org/sergarrido/eigen_pr/branch/block-lma-awf-expts which includes some fixes related to those you’ve already pointed to, and a revamp of LevenbergMarquardt using much faster specialist QRs:

Looking at your tmp2 in the SparseQR_QProduct, if you swap the j and k loops you can save the time of repeatedly making the tmp2 here for multi-column results. However, I haven't run into needing a tmp on GCC, clang or MSVC14. Which compiler are you using?

Julian



Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/