The ATLAS analysis model is a continually developing vision of how to reconcile physics analysis requirements with the ATLAS offline software and computing model constraints. In the past year this vision has influenced the evolution of the ATLAS Event Data Model, the Athena software framework, and physics analysis tools. These developments have led to a rapid refinement of the ATLAS Analysis Model in the past few months.
A Vision of the future ATLAS Analysis Model
One of the highlights of the recent Analysis Model workshop was the opening session where speakers from the D0, CDF, H1, BaBar, and CMS experiments explained their respective Analysis Models and shared their experience and recommendations. It was clear from these experiments that the speed of data access was the driving factor in the Analysis Model adopted by users, no matter what management planned or developers provided. The hind-sight of these previous experiments coupled with recent developments in Athena provide a refined vision for the ATLAS Analysis Model.
We find ourselves on the verge of having all necessary components to cope with many of the Analysis Model challenges that confronted previous experiments. The idea is to build an Analysis Model that provides a unified framework and data-format for the successive refinements which constitute the steps of algorithmic analysis.