As explained earlier (see Section the section called "The new Data Model"), tags are a small collection of the most important physics attributes of an event plus an association with the event in question. Tags provide a natural and efficient syntax for handling event data, speeding up queries, cuts handling, etc., substantially, while at the same time offering the possibility to easily access the original full data. Usually, the event tags are chosen in such a way that a high degree of selectivity can be obtained by first cutting on attributes in the event tag, while, if needed, the association to the full event can be exploited in order to retrieve any other information not contained in the tag.
In workgroup-wide data sets, individual physicists should still be able to have their own simpler data collection, so an easy-to-use Generic Tag is defined as well. Presently, both the Event Tag and the Generic Tag have to be created inside a C++ program before using the Interactive Analysis framework.
The data types that can be stored in a tag are long and short integers, float and double real numbers, and an 8-bit char.
Examples of the use of tags are given with more details in the HEPODBMS documentation.
The Anaphe environment provides tools to convert existing HBOOK Ntuples to Objectivity/DB.