To calculate electron identification and photonic electron reconstruction efficiencies, simulation dataset for both 2005 and 2006 are used. The dataset consists of Pythia events with embedded pi0, gamma, and electron tracks. To be able to calculate efficiencies, reconstructed tracks need to be properly identified as particle tracks.

The events undergo similar analysis chain as real data.

Reconstructed (rc) tracks are first associated to Monte Carlo (mc) tracks using StAssociationMaker, with the following setup:

• tracks association from Sti (IT tracks)
• using distance association (default)
• default association parameters: < 5mm distance between hits in x, y, z, 3 min. common hits required

Then the association is further narrowed down into a one-to-one association between rc and mc tracks, with the initial association maps from StAssociationMaker as the starting point. For this particular one-to-one association:

1. Initially, every rc track StAssociationMaker::mRcTrackMap has at least one mc track associated to it.
2. For every rc track, it is paired with the mc track with the highest percent of TPC hits paired between the two tracks as long as the mc track is not paired to another rc track. This means:

• Some initial rc-mc track association can be broken and the rc track left with no associated mc track. This happens if any and all possible mc track(s) is better paired with another rc track. Example: RC1 is initially associated with MC1. RC2 is initially associated with MC1, MC2. After final association, RC1 has no MC association, RC2 is associated with MC1.
• It is possible that the final rc-mc tracks pair is not the pair with the highest percent of paired TPC hits for the given rc track. This happens if the 'best' mc track for that rc track is better paired with another rc track. Example: RC1 is initially associated with MC1, MC2, and has highest percentage of common TPC hits with MC1. RC2 is initially associated with MC1, MC3 and has highest percentage of common TPC hits with MC1. RC2-MC1 common hits percentage is greater than RC1-MC1 common hits percentage. Final association yields RC1-MC2, RC2-MC1.

This method was chosen out of four possible methods of associating rc to mc tracks. More details and comparison of the methods can be found here.

One-to-one association allows definite particle identification for a given reconstructed track. It may  not be 100% accurate, but for this analysis, a track won't be identified as two (or more) different particles.

After the rc-mc track association, particle identification is done for particles of interest, namely:

• single electron (from electron embedding only): geantId==3||geantId==2
• electron from pi0 Dalitz decay (from pi0 embedding only): (geantId==2||geantId==3) && parentVertexNumberOfDaughters==3 && parentGeantId==149
• electron from gamma conversion from pi0 (from pi0 embedding only): (geantId==2||geantId==3) && parentGeantId==1 && parentVertexNumberOfDaughters==2 && parentParentGeantId==149
• electron from embedded gamma conversion (from gamma embedding only): (geantId==2||geantId==3) &&parentGeantId==1 && parentVertexNumberOfDaughters==2

GeantId = 149 is the embedded pi0 with increased Dalitz decay branching ratio.

These tracks are now treated like any real tracks and is projected to the EMC and matched to EMC points.