Detailed insight into the differences between Fo and Fc can be very helpful during the validation of a structural model and its underlying reflection data. The recently defined R tensor (S. Parkin, Expansion of scalar validation criteria to three dimensions: the R tensor. Acta Cryst. (2000), A56: 157-162.), which was developed to signal e.g. problems like twinning, shows the anisotropy of the error distribution in reciprocal space.
However, it does not reveal resolution or measuring order dependencies of the errors due to its averaging over theta and the loss of, in the first place, the link between individual, symmetry equivalent reflections and, in the second place, the link between the order in which the data were collected and the input to the least squares refinement procedure. Information on the differences between Fo and Fc as a function of measuring order, theta, Fo or position in reciprocal space can show drifts and trends that may point to systematic errors related to instrumental malfunctioning, to data reduction problems (e.g. sloppy correction for crystal decay), to wrong absorption correction or to errors in the structural model. General qualifications like 'bad data' or 'bad crystals' can be avoided by a proper analysis of the match between the structural model and the data.
The programs SYSTER and SYSTERPLOT were developed to show systematic relationships between a number of functions of Fo and Fc vs. a variety of variables that are involved in the measuring and data reduction procedures.The options