Ever have a situation in Golang where you want the precision of big.Float, but most operations don't need it? This library may be for you!

This library's main object is maybebig.Float, which wraps a float64 which is always evaluated, and a big.Float which may or may not be evaluated. The maybefloat.Float retains pointers to it's dependencies in such a way that if you ever want to compute the precice value, you can do so by calling myMaybeFloat.GetBig(), which will compute and return the full precision result.

The library automatically does the big.Float computations in a couple of circumstances:

• when the accumulated number of floating point operations reaches a certain size (configurable)
• when a comparison yields a difference that is within some tolerance (configurable)
• when a user asks for it via maybebig.Float.ForceComputation()

This library is a WIP, used in a few of my projects, but only supports the functions I've built thus far. If you need some functions that aren't supported, just ask, I'm happy to build them.

Please feel free to file a bug if you find one, I'm motivated to keep this library well supported.