HYPercomplex Automatic Differentiation (HYPAD) is a methodology to infuse existing codes with automatic differentiation capabilities. The method consists of “augmenting” variables with one or more imaginary units to obtain partial derivatives that are step-size independent and machine precision accurate. It is an attractive method because the existing numerical method is enhanced in a straightforward manner. The purpose of this web site is to provide educational material (see Lecture Notes), example Python notebooks (see Software Downloads), and supporting libraries (see Software Downloads) to facilitate the implementation of HYPAD across scientific and engineering software.
