This is an old revision of the document!
When analyzing a large sample of events, it may be useful to write a subsample of events (which may have passed) in a new DST file, which then can be used later on and is looped through faster than the original one.
In your Python options-file, add:
InputCopyStream().Output = "DATAFILE='PFN:yourdatafile.dst' TYP='POOL_ROOTTREE' OPT='REC'" YourSequence.Members.append(InputCopyStream())