Refactor

[nikhar-abbas]

Refactor

Some large updates to re-factor a number of dependencies that the ROSCO toolbox has had thus far. The main motivation for this is to ease the use of ROSCO as a stand-alone tool.

Major changes

• remove dependency on WIES's Aeroelasticse. For now, the Openfast I/O files and casegen-related scripts have been copied into a new folder in the ROSCO toolbox called ofTools. Over time, I think we can do some re-org to clean up and ease the workflow in these tools for controls and ROSCO specific purposes.

Minor Changes

• None, yet

To-Do list

Here's a list of items we need before this PR is ready to be merged. They can be added to the major and minor changes above as they get checked off.

• [x] Move plotting from ROSCO_toolbox/utilities.py to ofTools
• [x] Test tuning scripts
• [x] Add to ofTools README to provide better descriptions of the scripts
• [x] Update primary README to accommodate all changes
• [x] Push BAR model
• [x] Check examples
• [x] Check ROSCO_walkthrough.ipynb
• [ ] Get rid of pCrunch dependency
• [x] Update setup.py to compile ROSCO better

This refactor will also necessitate changes to pCrunch and weis.

Some notes on possibilities:

• The DISCON read/write function is pretty non-congruous with the aeroelasticse FAST_reader and FAST_writer functions. Good practice would be to consolidate all of these into a single function that works consistently, but this would involve some fairly significant restructuring to how the controller and turbine classes are handled or something like that (maybe a good thing anyways). This would end up trickling down to changes in how the read_DISCON and write_DISCON functions work, and call for updates to WEIS itself. Not a big deal, just should be kept in mind.

[nikhar-abbas]

I removed the ROSCO submodule in b6295b3 and added a section to check for the conda-forge distribution of ROSCO in the setup.py file in 5689dbe.

[dzalkind]

There are some issues in the ROSCO_testing scripts. They want to use pCrunch and pCrunch wants to use ROSCO_toolbox utilities. Do we want to remove dependency on pCrunch while that is being revamped?

[nikhar-abbas]

There are some issues in the ROSCO_testing scripts. They want to use pCrunch and pCrunch wants to use ROSCO_toolbox utilities. Do we want to remove dependency on pCrunch while that is being revamped?

I think it might be smart to get a parallel pCrunch branch going to call the new ROSCO toolbox utilities, and eventually coordinate their merges into their respective develop/master branches.

[dzalkind]

This looks good. I'd like to merge it into develop. There is still one place with a pCrunch dependency, but I don't think it is used often: generate_rotperf_fast() in turbine.py.