We have changed the way that we organize the ACD reconstruction data by adding an extra layer in which we group all the data about each track. Before all the data from all the track were stored together. This improvement has two main advantages:
- It makes it easier to sort and access the information, since we are often just working with the best track in the event, or we only care about the best associations for a given track.
- It makes it easier to interface to the event display, since for each track we only want to display the best associations for that track
New ACD recon data structures
Event Based
Every event now has a AcdEventTopology structure in AcdRecon to keep track of all the event based data such as AcdTotalEnergy, AcdTileCount, etc. Before these were all stored individually. The wrapper is mainly just there to clean up the interfaces a bit.
Track Based
Each track (and the vertex) has an AcdTkrAssoc object which stores:
- Some infomation about the track
- Data about where the track enters and leave the idealized Acd volume (AcdTkrPoint objects)
- Data about all the associations between the track and various ACD tiles and ribbons (AcdTkrHitPoca objects)
- Data about all the associations between the track and various gaps in the ACD (AcdTkrGapPoca objects)
Comparison between old and new
There a three main differences between the old and new versions:
- In the old version all the associations are stored in one collection for all the tracks, in the new version the associations are sorted by track
- In the old version the associations are sorted by active distance, in the new version they are sorted by veto probablity
- In the old version all the event level data are stored directly in the AcdRecon object, in the new version they are grouped into a sub object
