Santiago Soler

Also, it would be nice to make `BaseRegularization` and `WeightedLeastSquares` to raise an errors if the passed `weights` is not a dictionary. And also add a few tests that check...

cc @dccowan

The confusion raised when working in a 2.5D model. When we want to introduce a infinite cylinder in the mesh, currently we would have to use `get_indices_sphere` to do so....

I noticed by running some tests on the drafted code that some regularization classes like `JointTotalVariation`, `CrossGradient` and `LinearCorrespondence` don't implement the `f_m` and/or the `f_m_deriv` methods. They are all...

After discussing with @lheagy, we decided not to touch the `BaseRegularization` class in this PR. That branch of the hierarchy tree deserves some more thoughts and trying to tackle it...

Love this cleanup! As I suggested in today's meeting, I think we should take this opportunity to make all these `.py` files private by adding a heading underscore to their...

I just ran an inversion on synthetic gravity data with SimPEG `v0.20.0` and with the latest changes in this branch (`54da8d025b00bba9b7641f1349f5cee1397fc891`) and got comparable results. They pass a `np.allclose()` after...

Thanks @lheagy for the changes and the reply. I just edited my review comment, striking out the suggestions that were already solved: - all the changes in the PGI were...

Thanks @winnerer123 for willing to collaborate. @domfournier would it be fine if we let @winnerer123 tackle this issue? She's eager to collaborate more in SimPEG and I think this issue...

Nice catch @winnerer123! Yes, those lines should also be changed since `WeightedLeastSquares`'s constructor (the `__init__()` method) will broadcast the passed `weights` to all regularization objects within it.