Handle floating point imprecision

C3 should take care of floating point rounding imprecision, i.e. values like "0.30000000000000004" (= 0.1 + 0.2), in X and Y axis ticks.

We could
- have a fixed number of decimal places by default
- try to recognize imprecision patterns, like 10 zeros in a row

Someone already reported this in #506, but no action was taken there to resolve the problem.

I think that this should be taken care of in c3, because it's a common problem (anyone dealing with floating point values is exposed to it).