Dario Panici

@dpanici check singular values for different cases and check condition number

Checked these for basically every optimization or free bdry example in our notebooks. The condition number of Fxh was on average no more than 1e6, and the cutoff threshold for...

I still think it would be interesting to see the condition number of the actual proximally projected jacobian (e.g. the one used in the optimizer) with and without this regularization

Also mentioned in [this paper](https://arxiv.org/pdf/2301.13062.pdf) section V. C as way to speedup compilation times 

probably would not help so closing for now

I don't remember why I said it would not help much, probably worth spending the hour to re-write the F balance equations component wise and see if there's a difference...

Might be worth looking into `jnp.einsum` where we currently use `[None,:]` to broadcast arrays for multiplication, as this may be more efficient memory-wise

unscented transform as an option?

> Perturbations seems to be working: > > X Y Z >    > * The blue line represents the position perturbation as a function of > θ...

An issue I noticed: right now this computes a single perturbation which is applied to every coil in the coilset, i.e. every coil in the coilset has the exact same...