3.4. Visualization

Visualization is of paramount importance in order to easily understand the behavior of the detector and simulation. In this section, several tools of visualizing the detector geometry and the particle tracks are provided.

3.4.1. Key4hep tool: geoDisplay

The tool geoDisplay is available as part of Key4hep project. To use it, first we need to source it from LCG

source /cvmfs/sft.cern.ch/lcg/views/dev4/latest/x86_64-centos7-gcc11-opt/setup.sh 

To use it, just pass as first argument the xml file with the detector description, for example

geoDisplay $DD4hepINSTALL/DDDetectors/compact/SiD.xml

3.4.2. Geant4

The Geant 4 visualization capabilities can be accessed as

source /cvmfs/sft.cern.ch/lcg/views/dev4/latest/x86_64-centos7-gcc11-opt/setup.sh 
ddsim --compactFile $DD4hepINSTALL/DDDetectors/compact/SiD.xml --runType vis

This version of Geant 4 is built with Qt interface, and can be explicitly called as

ddsim --compactFile $DD4hepINSTALL/DDDetectors/compact/SiD.xml --runType qt

As in the previous step, the LCG version of Key4hep was sourced.

3.4.3. Phoenix@FCC

Phoenix is a web based event display for High Energy Physics. To visualize FCC events one needs to provide detector geometry and generated events — event data.

In this tutorial we will be working with files/programs stored on two computers. First computer will be the one which can source Key4hep stack, e.g. lxplus and the second one will be yours with the recent web browser. We will call the first one the remote machine and the second one the local machine.

3.4.3.1. Event Data from CLD Reconstructed Events

Let’s start with the visualization of the event data in the established detector design – CLD, which started its life as the detector designed for the linear collider concept CLIC. The detectors in Key4hep are described using DD4hep compact files. The compact files are written in XML and expose configuration options for the detector.

Example CLD compact file can be examined after sourcing of the Key4hep stack on the remote machine

source /cvmfs/sw-nightlies.hsf.org/key4hep/setup.sh

like so

less ${K4GEO}/FCCee/CLD/compact/CLD_o2_v07/CLD_o2_v07.xml

This file is part part of the k4geo repository, which stores among other detector concepts integrated in Key4hep also FCC-ee detector designs.

The Phoenix web application already showcases the CLD detector in different options here and we will simulate few events in order to visualize them there.

On the remote machine we will need to clone the CLDConfig repository

git clone https://github.com/key4hep/CLDConfig.git
cd CLDConfig/CLDConfig

run the simulation

ddsim --compactFile ${K4GEO}/FCCee/CLD/compact/CLD_o2_v07/CLD_o2_v07.xml \
      --steeringFile cld_steer.py \
      --inputFiles ../test/yyxyev_000.stdhep \
      --outputFile tops_cld_SIM_edm4hep.root \
      --numberOfEvents 5

and after that, run the reconstruction with the help of the CLDReconstruction.py steering file

k4run CLDReconstruction.py --inputFiles tops_cld_SIM_edm4hep.root \
                           --outputBasename tops_cld \
                           --num-events -1

There should now be a file called tops_cld_REC_edm4hep.root inside our working directory. More information about running of the CLD FullSim you can visit FCC-ee Detector Full Sim page.

In order to visualize the events inside tops_cld_REC_edm4hep.root file we need to convert the EDM4hep ROOT file into intermediate JSON representation with the command:

edm4hep2json -e 0 tops_cld_REC_edm4hep.root

EDM4hep in JSON

Be careful when converting EDM4hep ROOT files into JSON files as JSON is very ineffective in storing information leading to very large file size.

Always check size of your file to be around 200MB at maximum. One can check the file size using du command

du -h tops_cld_REC.edm4hep.json

The edm4hep2json offers possibility to limit either number of events exported or which collections. See edm4hep2json -h for more details.

Now we download the obtained file from the remote machine into our local one. Most easily done with scp

scp lxplus.cern.ch:CLDConfig/CLDConfig/tops_cld_REC.edm4hep.json .

To upload the EDM4hep JSON file into the Phoenix use the upload button in the lover right corner of the web page

to bring up modal with our desired upload option

The detector and event data might look similar to this screenshot

Example reconstructed event in CLD detector

EDM4hep JSON File

The obtained EDM4hep JSON file should look similar to this one [Right click to download].

3.4.3.2. Event Data from Delphes Fast Simulation

Let’s start by generating sample files using Generate and Simulate Events with DelphesEDM4Hep tutorial section.

Produced EDM4hep ROOT file(s) needs to be converted into intermediate JSON format with the edm4hep2json converter. Let’s look at the file more closely, so we can decide what to keep/convert.

Primary tool for inspecting EDM4hep files is podio-dump. This command can show you list of all collections and also metadata stored in the file. To list the collections in the default frame use

podio-dump p8_ee_ZH_ecm240_edm4hep.root

Alternatively, one can always use ROOT to inspect the file. Start by running

root p8_ee_ZH_ecm240_edm4hep.root

and execute the Show() command on the events TTree

root [1] events->Show(2);

this will show you all the branches which are by the PODIO reader to recreate EDM4hep objects and their relations in memory.

Using edm4hep2json for converting the EDM4hep ROOT file into JSON we select desired collections with -l flag and limit number of converted events to 10 with -n flag

edm4hep2json -l ReconstructedParticles,Jet,EFlowTrack,TrackerHits,Particle,MCRecoAssociations,MissingET -n 10 p8_ee_ZH_ecm240_edm4hep.root

The resulting EDM4hep JSON file can now be downloaded with scp to the local machine and uploaded into Phoenix visualization of the IDEA detector.

EDM4hep JSON File

The obtained EDM4hep JSON file should look similar to this one [Right click to download].

3.4.3.3. Detector Geometry

There are several ways how to import FCC detector geometry into Phoenix. Currently the preferred method is to convert compact DD4hep file(s) to ROOT files and from ROOT files to glTF files. The first conversion (.xml -> .root) is straightforward and can be done using script like this. With the most important part being the building of the detector geometry from the compact file

import ROOT

ROOT.gSystem.Load('libDDCore')
description = ROOT.dd4hep.Detector.getInstance()
for cfile in compact_files:
    description.fromXML(cfile)

ROOT.gGeoManager.SetVisLevel(9)
ROOT.gGeoManager.SetVisOption(0)
ROOT.gGeoManager.Export(out_path)

We will try to convert FCCee Noble Liquid Calorimeter. On the remote machine with FCCSW stack already sourced download the dd4hep2root script

wget https://fccsw.web.cern.ch/fccsw/tutorials/static/python/dd4hep2root

make it executable

chmod u+x dd4hep2root

and run the conversion with

./dd4hep2root -c ${FCCDETECTORS}/Detector/DetFCCeeIDEA-LAr/compact/FCCee_DectEmptyMaster.xml \
                 ${FCCDETECTORS}/Detector/DetFCCeeECalInclined/compact/FCCee_ECalBarrel.xml
              -o fccee_lar.root

The resulting file fccee_lar.root can already be visualized with ROOT. Let’s download it to the local machine with scp and then upload it into JSROOT. This way we don’t have to have ROOT installed on our local machine.

Resulting in the following visualization

Noble Liquid Calorimeter visualized with JSROOT

Second conversion (.root -> .gltf) requires additional configuration and can be done using the root_cern-To_gltf-Exporter. We have to create menu items, suppress drawing of detector parts which are too detailed or can affect the performance.

The root_cern-To_gltf-Exporter converter runs in the browser and is configured with an HTML page. It is recommended to run the converter on the local machine, but it should be possible to run it also on the remote machine over SSH with XWindowing enabled.

Let’s start with cloning the converter repository

git clone https://github.com/HSF/root_cern-To_gltf-Exporter.git
cd root_cern-To_gltf-Exporter

then, we create new file fccee_lar_conversion.html and use, for example, the following configuration for the FCCee Noble Liquid Calorimeter:

<html>

  <head>
    <script src="https://unpkg.com/three@0.139.1/build/three.js"> </script>
    <script src="https://unpkg.com/three@0.139.1/examples/js/exporters/GLTFExporter.js"> </script>
    <script src="https://root.cern/js/latest/scripts/JSRoot.core.js"> </script>
    <script src="https://cdnjs.cloudflare.com/ajax/libs/FileSaver.js/1.3.8/FileSaver.js"></script>
  </head>

<body>
  <script type="module">

    import { convertGeometry } from './phoenixExport.js';

    var hide_children = [
      "passive_",
      "active_",
      "PCB_"
    ];

    var subparts = {
        "Cryo Front" : [["ECAL_Cryo_front_0"], .8],
        "Cryo Back" : [["ECAL_Cryo_back_1"], .8],
        "Cryo Sides" : [["ECAL_Cryo_side_2"], .8],
        "Services Front" : [["services_front_3"], .6],
        "Services Back" : [["services_back_4"], .6],
        "Bath": [["LAr_bath_5"], true]
    }

    convertGeometry("./fccee_lar.root", "fccee_lar.gltf",
                    4, subparts, hide_children, "default", 120);
  </script>
</body>

Last missing peace is the ROOT file containing the detector geometry fccee_lar.root which we should also copy into the converter directory.

To start the converter do

./run

which will start the web server and show URL to which we can connect through the web browser

Started http server to serve requests at:
http://127.0.0.1:8000

In the web browser we should see directory listing similar to this one:

clicking on the fccee_lar_conversion.html starts the conversion. At the end of it the automatic download pop-up should appear in the browser window. The glTF file can be uploaded into Playground section of the Phoenix application by selecting the correct geometry file format:

The resulting visualization will be similar to the following screenshot

Example reconstructed event in CLD detector

glTF File

The obtained glTF file should look similar to this one [Right click to download].

LAr Clusters

Try to reconstruct calorimeter clusters in LAr Calorimeter and visualize them in the Phoenix Playground alongside with the detector.