README TwoDValidations

Git repository with the code/stuff to handle the validations for two-d deconvolution data processing

Useful tools

• TITUS evd for local usage is in here, instructions are a WIP
• Log of the test between 1D/2D is in here

Setting up TITUS (evd.py) for local usage

MacOS setup

As a staring point the needed software is brew (installation can be found on the site brew.sh), required to install a couple of tools.

• XQuartz, downloaded with brew install --cask xquartz. This provides the

Running docker-TITUS

1. Open XQuartz with open -a Xquartz, then infer the DISPLAY port with ps aux | grep Xquartz, it should be something like :0, :1 or any :N.
2. Go to the /path/where/is/your/data/ and start the docker with the command

docker run -it --rm -e "DISPLAY=host.docker.internal:N" -v /path/where/is/your/data/:/data sfbaylaser/titus /bin/bash

3. At this point setup titus and it should work...

Warning (updated Nov. 26th) The current docker image is not working correctly so for now this is NON-functioning code.

Initial tests (logbook)

In v10_06_00_04p04 at least there is something strange with the reco path

For the 2D deconvolution stage the combineHitsCryo module it used, as shown in

reco: [
      "gaushit2dTPCEW",
      "gaushit2dTPCEE",
      "combineHitsCryoE",
      "gaushitPT2dTPCEW",
      "gaushitPT2dTPCEE",
      "cluster3DCryoE",
      "pandoraGausCryoE",
      "pandoraTrackGausCryoE",
      "pandoraKalmanTrackGausCryoE",
      "SBNShowerGausCryoE",
      "fmatchCryoE",
      "fmatchopCryoE",
      "tpcpmtbarycentermatchCryoE",
      "crttrack",
      "CRTT0Tagging",
      "caloskimCalorimetryCryoE",
      "mcassociationsGausCryoE",
      "mcreco"
   ]

Both Pandora modules (E/W) are using the HitFinderModuleLabel: "combineHitsCryoE" label...

Opposite to this, the same path for the 1D deconvolution chain is only using the cluster3D module as shown in

reco: [
      "gaushit1dTPCEW",
      "gaushit1dTPCEE",
      "cluster3DCryoE",
      "pandoraGausCryoE",
      "pandoraTrackGausCryoE",
      "pandoraKalmanTrackGausCryoE",
      "SBNShowerGausCryoE",
      "fmatchCryoE",
      "fmatchopCryoE",
      "tpcpmtbarycentermatchCryoE",
      "crttrack",
      "CRTT0Tagging",
      "caloskimCalorimetryCryoE",
      "mcassociationsGausCryoE",
      "mcreco"
   ]

1D deconvolution chain

Aside from all the junk seen before, the 1D chain requires the addition of

# Tentative fix for mcreco
physics.producers.mcreco.SimChannelLabel: "shifted"

# Tentative fix for caloskim
physics.analyzers.caloskimW.SimChannelproducer: "daq:simpleSC"
physics.analyzers.caloskimE.SimChannelproducer: "daq:simpleSC"

in the stage1 reconstruction. mcreco is used for SPINE, whereas the caloskim analyzer is used for what?

Identical setting to be added to the 2D+DNNROI chain

# Tentative fix for mcreco
physics.producers.mcreco.SimChannelLabel: "shifted"

# Tentative fix for caloskim
physics.analyzers.caloskimW.SimChannelproducer: "daq:simpleSC"
physics.analyzers.caloskimE.SimChannelproducer: "daq:simpleSC"

These steps are done to remove the YZ-sim labels (which are now on by default)