Skip to content

Latest commit

 

History

History
185 lines (137 loc) · 5.42 KB

edmConverters.md

File metadata and controls

185 lines (137 loc) · 5.42 KB
Raw

Event Data Model converters

The converters are implemented as Gaudi Tools: these can be attached to any Gaudi Algorithm defined in the options file.

Reading events

Read EDM4hep events:

  1. Instantiate the Event Service from k4FWCore: k4DataSvc
  2. Indicate the event file to read.
  3. Then use the algorithm PodioInput to select the collections to read by their name.
  4. Add the algorithm PodioInput to the algorithm list.
from Configurables import k4DataSvc, PodioInput

# 1
evtsvc = k4DataSvc('EventDataSvc')
# 2
evtsvc.input = 'path/to/file.root' 

# 3
inp = PodioInput('InputReader')
inp.collections = [
    'ReconstructedParticles',
    'EFlowTrack'
]
inp.OutputLevel = DEBUG

# 4
algList.append(inp)

===

Read LCIO events:

  1. Instantiate the Event Data Service.
  2. Use LcioEvent and indicate the event file to read.
  3. Add the algorithm LcioEvent to the algorithm list.
from Configurables import EventDataSvc, LcioEvent

# 1
evtsvc = EventDataSvc()

# 2
read = LcioEvent()
read.OutputLevel = DEBUG
read.Files = ["path/to/file.slcio"]

# 3
algList.append(read)

Writing events

Write EDM4hep events:

  1. Instantiate PodioOutput and indicate event file to write.
  2. Select command for PodioOutput.
  3. Add the PodioOutput to the algorithm list.
from Configurables import PodioOutput

# 1
out = PodioOutput("PodioOutput", filename = "my_output.root")
# 2
out.outputCommands = ["keep *"]

# 3
algList.append(out)

===

Write LCIO events:

  1. Instantiate a MarlinProcessorWrapper with a relevant name.
  2. Indicate the ProcessorType to be LCIOOutputProcessor.
  3. Indicate the relevant parameters for the Marlin Processor.
  4. Add the MarlinProcessorWrapper to the algorithm list.
from Configurables import MarlinProcessorWrapper

# 1
Output_DST = MarlinProcessorWrapper("Output_DST")
Output_DST.OutputLevel = WARNING
# 2
Output_DST.ProcessorType = "LCIOOutputProcessor"
# 3
Output_DST.Parameters = {
                         "DropCollectionNames": [],
                         "DropCollectionTypes": ["MCParticle", "LCRelation", "SimCalorimeterHit"],
                         ...
                         }

# 4                         
algList.append(Output_DST)

Event Data Model converters

Note and review the following when using the converters:

  • Collection names must be unique.
  • If using the PodioInput to read events: collection names must be indicated both in the PodioInput and the EDM4hep2LcioTool.
  • Collections not indicated to be converted will not be converted even if its a dependency from an indicated collection to be converted.
  • If a converted collection is used later by a Gaudi Algorithm, and this Gaudi Algorithm indicates the use of that collection in the Parameters, the converted collection name must match the name indicated in the Gaudi Algorithm Parameters.
    • For example: A collection may be converted with the following parameters: ReconstructedParticles", "ReconstructedParticleLCIO"
    • A Gaudi Algorithm may indicate in their Parameters: "PFOCollection": ["ReconstructedParticleLCIO"]

EDM4hep to LCIO converter

Collections from events that are already read, or are produced by a Gaudi Algorithm can be converted from EDM4hep to LCIO format:

  1. Instantiate the EDM4hep2LcioTool Gaudi Tool.
  2. Indicate the collections to convert in Parameters.
  • To convert all available collections write an asterisk, like follows: edmConvTool.Parameters = ["*"]
  • Arguments are read in groups of 2: name of the EDM4hep collection, name of the LCIO converted collection.
  1. Select the Gaudi Algorithm that will convert the indicated collections.
  2. Add the Tool to the Gaudi Algorithm.
from Configurables import ToolSvc, EDM4hep2LcioTool

# 1
edmConvTool = EDM4hep2LcioTool("EDM4hep2lcio")
# 2
edmConvTool.Parameters = [
    "EFlowTrack", "EFlowTrack_LCIO",
    "ReconstructedParticles", "ReconstructedParticle_LCIO"
]
edmConvTool.OutputLevel = DEBUG

# 3
InitDD4hep = MarlinProcessorWrapper("InitDD4hep")

# 4
InitDD4hep.EDM4hep2LcioTool=edmConvTool

LCIO to EDM4hep converter

Collections from events that are already read, or are produced by a gaudi Algorithm can be converted from LCIO to EDM4hep format:

  1. Instantiate the Lcio2EDM4hepTool Gaudi Tool.
  2. Indicate the collections to convert in Parameters.
  • To convert all available collections write an asterisk, like follows: lcioConvTool.Parameters = ["*"]
  • Arguments are read in groups of 2: name of the LCIO collection, name of the EDM4hep converted collection.
  1. Select the Gaudi Algorithm that will convert the indicated collections.
  2. Add the Tool to the Gaudi Algorithm.
from Configurables import ToolSvc, Lcio2EDM4hepTool

# 1
lcioConvTool = Lcio2EDM4hepTool("LCIO2EDM4hep")
# 2
lcioConvTool.Parameters = [
    "EFlowTrackConv", "EFlowTrackEDM4hep",
    "ReconstructedParticle", "ReconstructedParticlesEDM4hep",
    "BuildUpVertices", "BuildUpVerticesEDM4hep",
    "PrimaryVertices", "PrimaryVerticesEDM4hep"
]
lcioConvTool.OutputLevel = DEBUG

# 3
JetClusteringAndRefiner = MarlinProcessorWrapper("JetClusteringAndRefiner")

# 4
JetClusteringAndRefiner.Lcio2EDM4hepTool=lcioConvTool