By default, the calculation on the active subsystem is performed using all basis functions employed by the calculation on the full system. The computational cost of a high-level WF calculation on even a small active subsystem can thus be prohibitively high if the full system is sufficiently large. This cost can be greatly reduced by performing the calculation on the active subsystem using only the basis functions that are most important for representation of the MOs associated with the active subsystem. Two methods are implemented to facilitate truncation of the active subsystem basis set, and these methods are described in the subsections below.

molpro@molpro.net 2019-03-22