| Analysis design [message #616] |
Tue, 22 June 2004 10:45 |
Hans Essel
Messages: 6
Registered: February 2004
Location: GSI - EE
|
occasional visitor |
From: depc35.gsi.de
|
|
In the current Go4 (v2.7) the analysis is organized in
sequential steps. Each step has a processor implementing the
event function, an event source filling an input event,
an output event filled by the event function, and an optional
event store. The input events of each step can be either filled
from a file (root tree) or from the previous step processor.
For Go4 v3 we open a discussion for an advanced step model.
The main goals are to provide a more flexible configuration
of analysis tasks and IO.
The new analysis design shall allow to configure meshes of
analysis jobs with arbitrary IO or event object access. A
Go4 analysis job could be a combination of GO4 steps and
ROOT tasks.
Another feature is an event stack allowing to store events on
a stack before processing. This is needed for time ordered
events from independent MBS branches like RISING, or data from
decay experiments.
Ideas or requirements are welcome.
Hans Essel, 2491
|
|
|
|
GSI Forum
Search
Help
Members
Login
Home
