make an MC recon file for 2019, 2021 spaced readout #1066
Conversation
There was a problem hiding this comment.
Thanks @bloodyyugo ! Add a few comments, mostly just to confirm changes. I think the biggest comment is on the KFOutputDriver ... if we keep this in, you will need to merge that driver in too.
| <driver name="ReconClusterer" type="org.hps.recon.ecal.cluster.ReconClusterDriver"> | ||
| <logLevel>WARNING</logLevel> | ||
| <outputClusterCollectionName>EcalClusters</outputClusterCollectionName> | ||
| <hitEnergyThreshold>0.030</hitEnergyThreshold> |
There was a problem hiding this comment.
Are these MC specific? We don't configure an energy threshold in our data config.
| <beamPositionY> 0.0 </beamPositionY> | ||
| <beamSigmaY> 0.02 </beamSigmaY> | ||
| <beamPositionZ> 0.0 </beamPositionZ> | ||
| <trackClusterTimeOffset>25</trackClusterTimeOffset> |
There was a problem hiding this comment.
Just to confirm that this and the line below should be different between data and MC. Our data is
<maxVertexClusterDt>40.0</maxVertexClusterDt>
<maxMatchDt>40</maxMatchDt>
| <debug>false</debug> | ||
| </driver> | ||
|
|
||
| <driver name="ReconParticleDriver_Kalman" type="org.hps.recon.particle.HpsReconParticleDriver" > |
There was a problem hiding this comment.
Do we need to enable makeCollerCols in MC? I guess it may not make sense for the majority of MC, but thought to check.
| <driver name="CleanupDriver" type="org.lcsim.recon.tracking.digitization.sisim.config.ReadoutCleanupDriver"/> | ||
|
|
||
| <driver name="KFOutputDriver" type="org.hps.recon.tracking.kalman.KFOutputDriver"> | ||
| <outputPlotsFilename>${outputFile}.root</outputPlotsFilename> |
There was a problem hiding this comment.
Are we planning to save this output? I wonder whether we should do the same for data. Might be nice to have this.
By the way, did you ever merge the KFOutputDriver into hps-java? I think it's private code you have locally only?
I am adding the tracking changes to this MC steering file
|
I have added the optimized parameters into the KalmanPatRecDriver. Should be the same as used in data as it was validated in both. |
|
I believe now all the reco changes I have optimized for have been incorporated. |
Added a steering file for 2019/2021 MC readout. This has Rory's fitting/clustering fixes. The SVT hit time corrections should put the track time to around ~0 and the cluster-track time offset in the recon-particle driver should be correct.
I ran this on some MC I had and it runs and gives decent tracks/clusters.