feat: more efficient proximal total variation operator

[dpshepherd]

Implement TV operator found in: Kamilov, Ulugbek S. "A parallel proximal algorithm for anisotropic total variation minimization." IEEE Transactions on Image Processing 26.2 (2016): 539-548.

[dpshepherd]

@ptbrown1729 - I wasn't sure if you want this to live in it's own file or fold it into optimize.py. What do you prefer?

I have some questions about this approach still, but given that it is non-iterative, I think it is worth a shot to try. If it doesn't work, we can try out the other "fast" approach I mentioned via email.

[ptbrown1729]

@dpshepherd since it doesn't add any new dependencies and we don't have that many proximal operators, I am fine if it lives in optimize.py. And we can collapse shrinkage() and the optimize.py function soft_threshold(), since these should be the same function

[dpshepherd]

Perfect, I made those changes. I can test this coming week.

[dpshepherd]

That commit is from me, I hadn't changed accounts on the ODT/SIM computer yet. Should be fixed now.