search

optimas-org / optimas

Optimization at scale, powered by libEnsemble
https://optimas.readthedocs.io
Other
22 stars 14 forks source link

SciPy Generators (e.g., Nelder-Mead) #189

Open ax3l opened 3 months ago

ax3l commented 3 months ago

Currently, optimas provides BO/MF-BO, random and grid sampling generators: https://optimas.readthedocs.io/en/latest/api/generators.html

For many optimization jobs, it would also be helpful to have a fast converging, conventional algorithm available. For this purpose, optimas could wrap a few SciPy Optimize methods, e.g., Nelder-Mead and L-BFGS-B.

https://docs.scipy.org/doc/scipy/reference/optimize.html

AngelFP commented 3 months ago

That's a good suggestion, but last time I had a look at this it wasn't possible because the scipy optimizers don't have an ask-tell interface that can separate the generation of trials from their execution. Maybe there are other libraries that could allow this.

ax3l commented 2 months ago

Oh, good point! I searched a little and found so far this: