A computational technique for evaluating and propagating the uncertainty of complex-valued quantities

This paper builds on recent contributions to ARFTG on the calculation and propagation of measurement uncertainty for complex-valued quantities according to N.M. Ridler and M.J. Salter (2000) and N.M. Ridler and M.J. Salter (2001). It describes the design of a simple software framework that substantially reduces the complexity of uncertainty calculations, while adhering to international best practice. The design is novel in several respects. It implicitly obtains the necessary partial derivatives (sensitivity coefficients) and has an internal structure that is hierarchical and modular. It can provide support for dynamic modularity (plug-and-play) and could be used in complex instrumentation systems to provide full low-level support for calculations that can propagate uncertainty.