Numerical precision and tolerances

Quantas calculations and native HDF5 values use float64 for real quantities and complex128 for complex quantities. Display precision belongs to renderers and does not modify stored data.

Work in progress

This page will explain tolerance selection, conditioning, uncertainty-aware comparisons, and the distinction between regression tolerances and physical acceptance criteria.