This is a non-linear optimization and regression library written for Swift, using the Accelerate vectorization framework. It's intendended to make implementing simple optimization and regression methods in iOS apps easy and efficient using Swift subclassing and protocols. Currently used to perform non-linear regression of thermal models for the (EcoChef) iOS app.
While the framework is currently being used in an existing iOS app this project is just started and there's a lot of work left to do. Collaboration partners would be welcome! The architecture is largely done, and so I think porting in or wrapping existing algorithms should be fairly straighforward. One possibility is wrapping the GNU Scientific Library optimization and regression functions, which are written in C.