Eigen with LAPACK/BLAS backend

ExaScience / smurff

Bayesian Factorization with Side Information in C++ with Python wrapper

MIT License

70 stars 14 forks source link

Eigen with LAPACK/BLAS backend #107

Closed tvandera closed 6 years ago

tvandera commented 6 years ago

This would improve performance, especially in conda.

From #41:

A cleaner solution, still using LAPACK, that would also boost the speed of other parts in SMURFF is to use -DEIGEN_USE_BLAS and -DEIGEN_USE_LAPACKE, with Eigen. See https://eigen.tuxfamily.org/dox/TopicUsingBlasLapack.html

tvandera commented 6 years ago

Does not seem to help w.r.t. performance...

tvandera commented 6 years ago

BLAS/LAPACK enabled since ce320428193dc16788d9d5ba1715c392af4697a6