Re: [eigen] cppduals - dual number implementation with Eigen specializations |
[ Thread Index | Date Index | More lists.tuxfamily.org/eigen Archives ]
Yes, of course. It's mentioned in the paper. The dual approach is faster (one multiplication fewer per op) and exact (doesn't rely on +h). The only place where the complex-step might be better is when a) your function is real, and b) you dont need much accuracy, and c) you have BLAS's complex-complex matrix operations that are faster than Eigen's optimization of the dual-valued matrix ops.There is/was some performance problem with Eigen's optimization of mat-mult of complex-valued matrices (and by extension, my dual-valued Eiegn matrices). The real-valued operations are/were very close to OpenBLAS on my machines, but the complex valued ops are quite a bit slower. I mentioned it on the Eigen chat room once, but I didn't have time to track it down and come up with a proper bug-report. (sorry, it will happen some day.. in case it hasn't already been fixed.)On Tue, Dec 3, 2019 at 4:04 PM Ian Bell <ian.h.bell@xxxxxxxxx> wrote:Have you compared against naive complex step derivatives : https://sinews.siam.org/Details-Page/differentiation-without-a-difference ? For first derivatives, CSD are trivial to apply, and doesn't require any additional machinations. I think that would serve as a nice benchmark.On Tue, Dec 3, 2019, 8:36 AM Michael Tesch <tesch1@xxxxxxxxx> wrote:Hello,I've written (yet another!) Dual Number implementation for automatic differentiation. It is meant to be used as the value-type in Eigen matrices, and has templates for vectorization (shockingly) similar to (and based on) Eigen's complex-type vectorizations. It is quite fast for first-order forward diff, and imho pretty easy to use. There are also SSE/SSE3/AVX vectorizations for std::complex<dual< float | double >> types.The library is here: https://gitlab.com/tesch1/cppduals , and there's a small paper in JOSS too: https://doi.org/10.21105/joss.01487I hope this could be useful for someone and would be glad for any feedback, improvements, etc.It would be interesting to compare this approach to others, by hand-wavey arguments I believe it should ultimately be faster in certain cases.Cheers,Michael
Mail converted by MHonArc 2.6.19+ | http://listengine.tuxfamily.org/ |