Implement FAASTA

[zaccharieramzi]

The FAASTA algorithm allows us to perform a FISTA algorithm in the analysis formulation. In our case we could see if the FAASTA still performs well with the new restarting options.

Since the FISTA with an orthogonal transform is way faster than the Condat-Vu, we are going to see now if this holds true even when we have to compute an extra inner iteration.

[sfarrens]

@zaccharieramzi are you planning a PR for this issue?

[zaccharieramzi]

Yes I am. I am currently working on the best way to package it but it's basically done. Do you want to have a WIP PR to see if the direction is statisfying?

[zaccharieramzi]

@sfarrens even with the trimming (#50 ), the FAASTA is still way slower than Condat for reconstruction problems with undecimated wavelets. Should we still have it in Modopt?

[sfarrens]

@zaccharieramzi I think having a diverse set of reconstruction algorithms would be useful regardless. So if you have already implemented it then feel free to open a PR.

[zaccharieramzi]

Ok I will package it in a PR.

[sfarrens]

@zaccharieramzi are you still planning a PR for this algorithm?

[zaccharieramzi]

Not at the moment. Do you still want to keep this open? I will probably do it at some point but I don't think I can find the time right now.

[sfarrens]

We can keep it open, I just wanted to check on the status. :)