How surprising that they release a product that contains so many bugs. Very impressive that all of them come with a small an nice reproducing example.

0Thread HN

Optimization Without Derivatives: Prima Fortran Version and Inclusion in SciPy 3 years ago

“Renowned for their robustness and efficiency, these solvers are used in a wide spectrum of applications, for instance, aeronautical engineering [25], astronomy [39], computer vision [33], robotics [40], and statistics [5].” These algorithms seem truly used in science and engineering.

[5] D. M. Bates, M. Mächler, B. Bolker, and S. Walker. Fitting linear mixed-effects models using lme4. J. Stat. Softw., 67:1–48, 2015.

[25] F. Gallard et al. GEMS: a Python library for automation of multidisciplinary design optimization process generation. In 2018 AIAA/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, Kissimmee, FL, USA, 2018. AIAA.

[33] H. Izadinia, Q. Shan, and S. M. Seitz. IM2CAD. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 5134–5143, San Juan, PR, USA, 2017. IEEE.

[40] K. Mombaur, A. Truong, and J. P. Laumond. From human to humanoid locomotion—an inverse optimal control approach. Auton. Robot., 28:369–383, 2010.

[39] G. A. Mamon, A. Biviano, and G. Boué. MAMPOSSt: modelling anisotropy and mass profiles of observed spherical systems I. Gaussian 3D velocities. Mon. Not. R. Astron. Soc., 429:3079–3098, 2013.

0Thread HN

Optimization Without Derivatives: Prima Fortran Version and Inclusion in SciPy 3 years ago

LFortran sounds to be a genenius project. The interactivity is a critical block that is missing from FORTRAN ecosystem. Hope things will get better with LFortran!

0Thread HN

Optimization Without Derivatives: Prima Fortran Version and Inclusion in SciPy 3 years ago

`they are also much simpler to implement ...'

According to the papers metnioned by others, the algos under discussion do not seem to support this point. The author of the project mentiones that the FORTRAN 77 implementation is `nontrivial to understand', which motivates this work.

`Caveat: derivative free optimizers are terrible for constrained optimization however (especially if your constraints are anything but simple bounds). You can use barrier methods but still they’re not great.'

Lincoa and cobyla in the package can handle problems with nontrivial constriants according to the documentation of the git repository, though I do not know how well they work.

0Thread HN