FixedProfiles actor and optimization

[daveweisberg]

I've published my development branch to the FUSE (https://github.com/ProjectTorreyPines/FUSE.jl/tree/weisberg_dev) and IMASDD (https://github.com/ProjectTorreyPines/IMASDD.jl/tree/weisberg_dev) repos. One of the main additions is the FixedProfiles actor, which is a transport actor that allows for the "feed-forward" profile generation I presented last Tuesday. I've also added "opt_workflow.ipynb" (https://github.com/ProjectTorreyPines/FUSE.jl/blob/44bee0dd29850562566e9accc02557f0d68286e3/playground/weisberg_FPP_opt/opt_workflow.ipynb), which is the primary optimization workflow I've been using.

For this issue, I have two requests:

1. Please look over my developments and comment on how to improve them re: run-time and computational efficiency.
2. Please run opt_workflow.ipynb for yourselves and look into the error between betaN_MHD and betaN_kinetic. I've tried with all three equilibrium models, and I still get significant errors of up to 10% for certain cases. What am I doing wrong that is resulting in this discrepancy?

[orso82]

@daveweisberg when asking for people to look over your work it's better open a pull request, as we did for https://github.com/ProjectTorreyPines/FUSE.jl/pull/374

I am going to close this and copy your comments over to #374