Optimization of the calculation of structure factors for large molecules

A method is described for the rapid calculation of structure factors by computing all the contributions of a given atom to all reflections at a time, instead of the contributions from all atoms to each reflection. This method requires no multiplication in the calculation of the arguments of the sine and cosine terms. These arguments, accumulated by integer addition, are used as subscripts in referencing a trigonometric function table. The method can be used in structure refinement, and is particularly suited to a Gauss-Seidel least-squares approach.