Introduction
This page is a summary of tasks that can be performed based on information contained in the various data formats. It is assumed that there is a strict hierarchy, i.e. any task that can be done on AOD, can also be done on ESD and RAW, and any task that can be done on ESD, can also be done on RAW (if necessary by running a reconstruction step before carrying out the task).
Tasks that need RAW data
By definition all reconstruction tasks can be carried out starting from RAW.
Items listed here are those that can
only be done from the RAW and
not from the ESD or AOD.
Trigger
- Rerun of the trigger algorithms for trigger studies: As the trigger objects used in calculating the trigger decisions are not all stored in ESD, a rerun of the trigger needs to start from RAW data.
Tracking
- Apply Pixel Noise Map
- Apply Strip Noise Map
- Apply mask to TRT straws
- Pixel Clustering: the Pixel clustering can only be done from the raw as only the raw contain the hit information. Reclustering may be needed e.g. if there was an improvement in the clustering code, or a bug fix, or if e.g. for the Pixels a new ToT calibration becomes available.
- (Caveat
- this is only true in the long term. In the commissioing ESDs the full hit information is currently kept)
- SCT Clustering: the SCT clustering can only be done from the raw as only the raw contain the hit information. Reclustering may be needed e.g. if there was an improvement in the clustering code, or a bug fix.
- (Caveat
- this is only true in the long term. In the commissioing ESDs the full hit information is currently kept)
Calorimeters
- Corrections and Calibrations at the cell level of any kind
Muon System
Tasks that need ESD
Tracking
- Pattern Recognition: the pattern recognition is based on clusters and drift circles. Since all clusters and drift circles are contained in the ESD's a full new pattern recognition can be run.
- Refit all tracks: instead of doing the pattern recognition from scratch one can also do track refits to e.g. account for a new alignment, a new TRT calibration or a new masked channel list.
- Determine Hit Efficiencies and Residuals: since all clusters are available hit efficiencies and residuals can be determined
- TRT Drift Circles: the calibration of the TRT drift circles can be redone from the ESD, e.g. using a new R-t relation or a new t0. In the current commissioning ESD in fact all the RAW data for the Inner Detector are contained so that in fact anything that can be done from the RAW can also be done from these.
Muon system
- Muon alignment: Alignment studies will be carried out using ESD
Tasks that can be done on AOD
Muon system
- Track refit: A track refit on AOD data may be done if not only the hits on the track, but also outlier hits are retained in the AOD objects.
- Tag and probe efficiency studies: Can be carried out on AOD if the detector is well understood, otherwise, fall back on (d)ESD
- ID-MS relative alignment: Up to now the AODs were used to perform the ID-MS alignment. Track parameters, covariance matrix and some hits informations are used to construct the Chi2.
Tracking
AOD keeps clusters only on identified muons and electrons (?), therefore can
- determine general track properties: including the error matrix and hit counts on tracks via the TrackSummaryTool
- refit the muon or electron tracks: since the clusters are being kept for those
- determine tracking efficiencies: using an external probe (e.g. muon or electron)
- do studies of material and resolutions: etc. using the track parameters e.g. conversions, Kshorts,...
- redo vertexing of primary or secondary vertices
- determine hit efficiencies on identified muons or electrons
Note one cannot (ESD are needed):
- refit low pt tracks
- determine hit efficiencies on generic tracks
Tasks suitable for AODfix
Here tasks are listed that may be applied to data based on AOD objects, and are therefore suitable for AODfix,
which at the time of application will only have AOD information available (plus the corresponding meta- and
conditions- data). The corrections may well be derived from studies involving RAW and ESD data, however the
application has to be based on AOD.
Calorimeter
Possible are:
- Overall EM scale (at least for EM clusters).
- Dead material map (for e/gamma corrections at least)
- Jet-level corrections
- MET corrections to reflect AODFix changes in constituents
- Perhaps some alignment changes (e.g. calorimeter moves with respect to ID, leading to update of EM cluster)
- Perhaps some beam spot changes.
- Perhaps updated EM calibration of the TileCal, if it's done in some parametrized form to apply, layer-by-layer, in topo clusters.