preserve precision of input values when generating results

[pkienzle]

A density entered as 2.2e-4 is reported as 0.00. Instead, values should be treated as if there is an uncertainty of 1 in the last digit and reported with that uncertainty.

This is primarily an issue with the online calculator from the scattering/activation repo, but the periodictable package may need to be updated to support uncertain values in its calculations to ensure that the errors are propagated correctly.

Integer values (e.g., density = 2 g/cm^3) should perhaps be treated as having unlimited precision, and so the number of digits is driven by the uncertainties in cross sections and half-lives. That does mean that "2" and "2." will show different results, which could be confusing.