search

PyLops / pyproximal

PyProximal – Proximal Operators and Algorithms in Python
https://pyproximal.readthedocs.io
GNU Lesser General Public License v3.0
52 stars 12 forks source link

Let `epsg` in `AcceleratedProximalGradient` be a vector of values #41

Closed eurunuela closed 2 years ago

eurunuela commented 2 years ago

Now that the pylops FISTA works on matrices of data, it makes sense that the regularization parameter can be a vector of values (one per vector of data in the data matrix).

Implementing this feature will require that the proximals work at the matrix level rather than the vector level.

Edit: I actually think Python would already take care of making proximals works with matrices.

mrava87 commented 2 years ago

Good point :) I’ll check if the current implementation allows it, probably yes

mrava87 commented 2 years ago

Sorry for taking so long! Seems like it was as easy as allowing tau to be a vector in the check for tau>0. Of course this will not work with any proximal but for the case of L1 it can work. I am not planning to change also FISTA in PyLops as by now AcceleratedProximalGradient is more powerful in many ways and I recommend using it over FISTA in PyLops

eurunuela commented 2 years ago

Thank you @mrava87! I believe I can now use pylops for my analyses.