Floating Point Expression Inspector

A welcome endeavor. Too many programmers are used to treat floats as cursed entities.

More awareness is needed. If implementation details such as rounding mode, CPU features, etc.. are pinned down there is little reason not to write tests that are meant to be accurate to the last bit. Instead of the awkward "near enough" approach.

There's also a floating point expression generator that tries to make your calculations more accurate, but it doesn't seem to give an analysis like this. These two tools would seem to go well together.

https://herbie.uwplse.org/doc/latest/tutorial.html