Add datatypes for storing generator (meta)data in a structured and defined way

README.md

@@ -39,12 +39,18 @@ A generic event data model for future HEP collider experiments.
 | [MCRecoTrackerHitPlaneAssociation](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L644) | [MCRecoCaloParticleAssociation](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L653) | [MCRecoClusterParticleAssociation](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L662) |
 | [MCRecoTrackParticleAssociation](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L671)   | [RecoParticleVertexAssociation](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L680) |                                                                                                      |
 
-**Interfaces**
+**Generator related (meta-)data**
 
 | | | |
 |-|-|-|
-| [TrackerHit](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L787) | | |
+| [GeneratorEventParameters](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L790) | | |
+| [GeneratorPdfInfo](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L807) | | |
+
+**Interfaces**
 
+| | | |
+|-|-|-|
+| [TrackerHit](https://github.com/key4hep/EDM4hep/blob/main/edm4hep.yaml#L818) | | |
 
 The tests and examples in the `tests` directory show how to read, write, and use these types in your code.
 

doc/GeneratorInfo.md

@@ -0,0 +1,35 @@
+# Dealing with Generator (meta-)data
+
+EDM4hep provides data types and helper functions to handle the storage of generator meta data, both at run and event level.
+
+## Storing and retrieving run level information
+At run level, information about generator tools (name, version, description) are being stored. They can be written via
+
+
+```cpp
+    #include "edm4hep/GeneratorToolInfo.h"
+    // ...
+    // write some generator tool info into the run
+    auto toolInfo = edm4hep::GeneratorToolInfo();
+    auto toolInfos = std::vector<edm4hep::GeneratorToolInfo>();
+    toolInfo.name = "something";
+    toolInfo.version = "v1";
+    toolInfo.description = "some tool";
+    toolInfos.emplace_back(std::move(toolInfo));
+
+    edm4hep::utils::putGenToolInfos(run, toolInfos);
+```
+
+and read-back via:
+
+```cpp
+  #include "edm4hep/GeneratorToolInfo.h"
+  // ...
+  auto toolInfos = edm4hep::utils::getGenToolInfos(run);
+
+```
+
+### Storing and retrieving event specific generator parameters and PDF information
+
+For storing information about event level parameters of generators one can use the type `GeneratorEventParameters`.
+For storing information about PDFs, one can use the type `GeneratorPdfInfo`.

edm4hep.yaml

@@ -783,6 +783,37 @@ datatypes:
     OneToOneRelations:
       - edm4hep::Track track           // the corresponding track
 
+
+  #===== Generator related data =====
+
+  #---------- GeneratorEventParameters
+  edm4hep::GeneratorEventParameters:
+    Description: "Generator event parameters"
+    Author: "EDM4hep authors"
+    Members:
+      - double eventScale              // event scale
+      - double alphaQED                 // alpha_QED
+      - double alphaQCD                 // alpha_QCD
+      - int signalProcessId           // id of signal process
+      - double sqrts  [GeV]            // sqrt(s)
+    VectorMembers:
+      - double crossSections [pb]       // list of cross sections
+      - double crossSectionErrors [pb]  // list of cross section errors
+    OneToManyRelations:
+      - edm4hep::MCParticle signalVertex // List of initial state MCParticle that are the source of the hard interaction
+
+
+  #---------- GeneratorPdfInfo
+  edm4hep::GeneratorPdfInfo:
+    Description: "Generator pdf information"
+    Author: "EDM4hep authors"
+    Members:
+      - std::array<int, 2> partonId      // Parton PDG id
+      - std::array<int, 2> lhapdfId             // LHAPDF PDF id (see https://lhapdf.hepforge.org/pdfsets.html)
+      - std::array<double, 2> x           // Parton momentum fraction
+      - std::array<double, 2> xf          // PDF value
+      - double scale [GeV]                // Factorisation scale
+
 interfaces:
   edm4hep::TrackerHit:
     Description: "Tracker hit interface class"

include/edm4hep/Constants.h

@@ -37,6 +37,14 @@ namespace labels {
   static constexpr const char* PIDParameterNames = "ParameterNames";
   static constexpr const char* PIDAlgoName = "AlgoName";
   static constexpr const char* PIDAlgoType = "AlgoType";
+
+  // Parameter names for Generator level metadata
+  static constexpr const char* GeneratorToolVersion = "GeneratorToolVersion";
+  static constexpr const char* GeneratorToolName = "GeneratorToolName";
+  static constexpr const char* GeneratorToolDescription = "GeneratorToolDescription";
+  static constexpr const char* GeneratorEventParameters = "GeneratorEventParameters";
+  static constexpr const char* GeneratorPdfInfo = "GeneratorPdfInfo";
+  static constexpr const char* GeneratorWeightNames = "GeneratorWeightNames";
 } // namespace labels
 
 DEPRECATED_LABEL(CellIDEncoding, CellIDEncoding);

include/edm4hep/GeneratorToolInfo.h

@@ -0,0 +1,88 @@
+#ifndef EDM4HEP_GENERATORTOOLINFO_H
+#define EDM4HEP_GENERATORTOOLINFO_H
+
+#include "edm4hep/Constants.h"
+#include "podio/Frame.h"
+#include <string>
+#include <vector>
+
+namespace edm4hep {
+
+/// Meta information class to group information about the used generator (tools)
+/// to create a file
+///
+/// @note Since this is all rather loosely coupled and stored via podio Frame
+/// parameters, use of the @ref getGenToolInfos and @ref putGenToolInfos utility
+/// functions for retrieval, resp. storage of this information is crucial to
+/// ensure consistent information.
+struct GeneratorToolInfo {
+  std::string name{};        ///< The name of the tool
+  std::string version{};     ///< The version of the tool
+  std::string description{}; ///< A brief description of the tool
+
+  /// Construct a generator tool info object with all empty fields
+  GeneratorToolInfo() = default;
+
+  /// Construct a complete tool info object from all ingredients
+  ///
+  /// @param name        The name of the tool
+  /// @param version     The version of the tool
+  /// @param description The brief description of the tool
+  GeneratorToolInfo(const std::string& name, const std::string& version, const std::string& description) :
+      name(name), version(version), description(description){};
+};
+
+namespace utils {
+
+  /// Get all the generator tool infos that are available from the passed
+  /// (metadata) frame.
+  ///
+  /// Tries to retrieve all meta information that are available and that have
+  /// been stored via the @ref putGenToolInfos function.
+  ///
+  /// @param frame The (metadata) frame that should be queried for the
+  ///              information
+  ///
+  /// @returns The GeneratorToolInfo that were found in the Frame. If none ar
+  ///          found an empty vector will be returned.
+  const inline std::vector<GeneratorToolInfo> getGenToolInfos(const podio::Frame& frame) {
+    using namespace edm4hep::labels;
+    auto toolInfos = std::vector<GeneratorToolInfo>();
+    const auto names =
+        frame.getParameter<std::vector<std::string>>(GeneratorToolName).value_or(std::vector<std::string>{});
+    const auto versions =
+        frame.getParameter<std::vector<std::string>>(GeneratorToolVersion).value_or(std::vector<std::string>{});
+    const auto descriptions =
+        frame.getParameter<std::vector<std::string>>(GeneratorToolDescription).value_or(std::vector<std::string>{});
+    for (unsigned int i = 0; i < names.size(); ++i) {
+      toolInfos.emplace_back(names[i], versions[i], descriptions[i]);
+    }
+    return toolInfos;
+  };
+
+  /// Put the generator tool meta information into a (metadata) frame
+  ///
+  /// In order to guarantee consistent storage and retrieval of these metadata
+  /// it is necessary to use this utility function.
+  ///
+  /// @param frame     The (metadata) Frame into which the generator tool info should go
+  /// @param toolInfos The generator tool infos that should be stored
+  void inline putGenToolInfos(podio::Frame& frame, const std::vector<GeneratorToolInfo>& toolInfos) {
+    auto names = std::vector<std::string>();
+    auto versions = std::vector<std::string>();
+    auto descriptions = std::vector<std::string>();
+    for (auto& toolInfo : toolInfos) {
+      names.push_back(toolInfo.name);
+      versions.push_back(toolInfo.version);
+      descriptions.push_back(toolInfo.description);
+    }
+
+    using namespace edm4hep::labels;
+    frame.putParameter(GeneratorToolName, std::move(names));
+    frame.putParameter(GeneratorToolVersion, std::move(versions));
+    frame.putParameter(GeneratorToolDescription, std::move(descriptions));
+  };
+} // namespace utils
+} // namespace edm4hep
+
+#endif // EDM4HEP_GENERATORTOOLINFO_H

test/read_events.h

@@ -3,6 +3,9 @@
 
 // test data model
 #include "edm4hep/CaloHitContributionCollection.h"
+#include "edm4hep/GeneratorEventParametersCollection.h"
+#include "edm4hep/GeneratorPdfInfoCollection.h"
+#include "edm4hep/GeneratorToolInfo.h"
 #include "edm4hep/MCParticleCollection.h"
 #include "edm4hep/RawTimeSeriesCollection.h"
 #include "edm4hep/SimCalorimeterHitCollection.h"
@@ -16,6 +19,27 @@
 // STL
 #include <iostream>
 
+void processRun(const podio::Frame& run) {
+  //===============================================================================
+  // get generator tool info from the run
+  auto toolInfos = edm4hep::utils::getGenToolInfos(run);
+  auto toolinfo = toolInfos[0];
+  if (toolinfo.name != "something")
+    throw std::runtime_error("toolinfo.name != 'something'");
+  if (toolinfo.version != "v1")
+    throw std::runtime_error("toolinfo.version != 'v1'");
+  if (toolinfo.description != "some tool")
+    throw std::runtime_error("toolinfo.description != 'some tool'");
+
+  //===============================================================================
+  // get generator weight names
+  auto weightNames = run.getParameter<std::vector<std::string>>(edm4hep::labels::GeneratorWeightNames).value();
+  if (weightNames[0] != "oneWeight")
+    throw std::runtime_error("weightNames[0] != 'oneWeight'");
+  if (weightNames[1] != "anotherWeight")
+    throw std::runtime_error("weightNames[1] != 'anotherWeight'");
+}
+
 void processEvent(const podio::Frame& event) {
   auto& mcps = event.get<edm4hep::MCParticleCollection>("MCParticles");
   auto& sths = event.get<edm4hep::SimTrackerHitCollection>("SimTrackerHits");
@@ -232,6 +256,17 @@ void processEvent(const podio::Frame& event) {
     throw std::runtime_error("Collection 'TrackerHitPlanes' should be present");
   }
 
+  //===============================================================================
+  // check the generator meta data
+  auto& genParametersCollection =
+      event.get<edm4hep::GeneratorEventParametersCollection>(edm4hep::labels::GeneratorEventParameters);
+  auto genParam = genParametersCollection[0];
+  if (genParam.getEventScale() != 23)
+    throw std::runtime_error("Event_scale != 23");
+
+  auto& generatorPdfInfoCollection = event.get<edm4hep::GeneratorPdfInfoCollection>(edm4hep::labels::GeneratorPdfInfo);
+  auto genPdfInfo = generatorPdfInfoCollection[0];
+
   // //===============================================================================
   //  if( sccons.isValid() ){
   //  } else {
@@ -251,6 +286,9 @@ void read_events(const std::string& filename) {
   ReaderT reader;
   reader.openFile(filename);
 
+  const auto run = podio::Frame(reader.readNextEntry("runs"));
+  processRun(run);
+
   unsigned nEvents = reader.getEntries("events");
   for (unsigned i = 0; i < nEvents; ++i) {
     std::cout << "reading event " << i << std::endl;

test/write_events.h

@@ -2,7 +2,11 @@
 #define EDM4HEP_TEST_WRITE_EVENTS_H
 
 // Data model
+
 #include "edm4hep/CaloHitContributionCollection.h"
+#include "edm4hep/GeneratorEventParametersCollection.h"
+#include "edm4hep/GeneratorPdfInfoCollection.h"
+#include "edm4hep/GeneratorToolInfo.h"
 #include "edm4hep/MCParticleCollection.h"
 #include "edm4hep/RawTimeSeriesCollection.h"
 #include "edm4hep/SimCalorimeterHitCollection.h"
@@ -28,6 +32,7 @@ void write(std::string outfilename) {
   for (unsigned i = 0; i < nevents; ++i) {
     std::cout << " --- processing event " << i << std::endl;
     auto event = podio::Frame();
+    auto run = podio::Frame();
 
     // place the following generator event to the MCParticle collection
     //
@@ -108,6 +113,47 @@ void write(std::string outfilename) {
       }
     }
 
+    //===============================================================================
+    // write some generator event data
+    auto genParametersCollection = edm4hep::GeneratorEventParametersCollection();
+    auto genParam = genParametersCollection.create();
+    genParam.setEventScale(23);
+    genParam.setAlphaQED(1 / 127);
+    genParam.setAlphaQCD(0.1);
+    genParam.setSignalProcessId(42);
+    genParam.setSqrts(90);
+    genParam.addToCrossSections(10);
+    genParam.addToCrossSectionErrors(3);
+    genParam.addToSignalVertex(mcp1);
+    genParam.addToSignalVertex(mcp2);
+    event.put(std::move(genParametersCollection), edm4hep::labels::GeneratorEventParameters);
+
+    auto genPdfInfoCollection = edm4hep::GeneratorPdfInfoCollection();
+    auto genPdfInfo = genPdfInfoCollection.create();
+    genPdfInfo.setPartonId(1, 2);
+    genPdfInfo.setLhapdfId({20, 20});
+    genPdfInfo.setX({0.5, 0.5});
+    genPdfInfo.setXf({0.5, 0.5});
+    genPdfInfo.setScale(23);
+    event.put(std::move(genPdfInfoCollection), edm4hep::labels::GeneratorPdfInfo);
+
+    //===============================================================================
+    // write some generator tool info into the run
+    auto toolInfos = std::vector<edm4hep::GeneratorToolInfo>();
+    auto toolInfo = edm4hep::GeneratorToolInfo();
+    toolInfo.name = "something";
+    toolInfo.version = "v1";
+    toolInfo.description = "some tool";
+    toolInfos.emplace_back(std::move(toolInfo));
+    edm4hep::utils::putGenToolInfos(run, toolInfos);
+
+    //===============================================================================
+    // write some generator weightname info into the run
+    auto weightNames = std::vector<std::string>();
+    weightNames.emplace_back("oneWeight");
+    weightNames.emplace_back("anotherWeight");
+    run.putParameter(edm4hep::labels::GeneratorWeightNames, std::move(weightNames));
+
     // fixme: should this become a utility function ?
     //-------------------------------------------------------------
 
@@ -237,6 +283,7 @@ void write(std::string outfilename) {
     event.putParameter("EventType", "test");
 
     writer.writeFrame(event, "events");
+    writer.writeFrame(run, "runs");
   }
 
   writer.finish();

tools/include/edm4hep2json.hxx

@@ -6,6 +6,8 @@
 #include "edm4hep/CalorimeterHitCollection.h"
 #include "edm4hep/ClusterCollection.h"
 #include "edm4hep/EventHeaderCollection.h"
+#include "edm4hep/GeneratorEventParametersCollection.h"
+#include "edm4hep/GeneratorPdfInfoCollection.h"
 #include "edm4hep/MCParticleCollection.h"
 #include "edm4hep/ParticleIDCollection.h"
 #include "edm4hep/RawCalorimeterHitCollection.h"
@@ -154,6 +156,12 @@ nlohmann::json processEvent(const podio::Frame& frame, std::vector<std::string>&
       insertIntoJson<podio::UserDataCollection<uint32_t>>(jsonDict, coll, collList[i]);
     } else if (coll->getTypeName() == "podio::UserDataCollection<uint64_t>") {
       insertIntoJson<podio::UserDataCollection<uint64_t>>(jsonDict, coll, collList[i]);
+    }
+    // Generator (meta-)data
+    else if (coll->getTypeName() == "podio::GeneratorParametersCollection") {
+      insertIntoJson<edm4hep::GeneratorEventParametersCollection>(jsonDict, coll, collList[i]);
+    } else if (coll->getTypeName() == "podio::GeneratorPdfInfoCollection") {
+      insertIntoJson<edm4hep::GeneratorPdfInfoCollection>(jsonDict, coll, collList[i]);
     } else {
       std::cout << "WARNING: Collection type not recognized!\n"
                 << "         " << coll->getTypeName() << "\n";