All the event-analysis routines above are defined on an event-by-event basis. Once found, the quantities are about equally often used to define inclusive distributions as to select specific classes of events for continued study. For instance, the thrust routine might be used either to find the inclusive distribution or to select events with . Other measures, although still defined for the individual event, only make sense to discuss in terms of averages over many events. A small set of such measures is found in PYTABU. This routine has to be called once after each event to accumulate statistics, and once in the end to print the final tables. Of course, among the wealth of possibilities imaginable, the ones collected here are only a small sample, selected because the authors at some point have found a use for them.