forward.search() now throws an error when no unique base set can be found (occurs when the defined
model is exactly identified, and therefore perfectly fits the data)
added a progress argument to gCD(), LD(), and GOF() to print the progress of the iterations
several orgainizational overahauls
split LD() into LD() and GOF() for better clarity
avoid masking lavaan::sem with sem::sem
better use of the parallel package for faster computing in most functions (most notably in the forward.search implementation)
numerical tests added for better package stability
more flexible passing of lavaan-type models and other functions used within the package