BrTpcDeconvoluteClusterModule

class BrTpcDeconvoluteClusterModule : public BrModule

    private:

      virtual Bool_t DeconvoluteClusters()
        virtual void FillSubClusters(Int_t npeaks, TObjArray* subClusters)
        virtual void FillWorkingArrays(BrTpcCluster* cluster)
       virtual Int_t FindHitStatus(BrTpcCluster* cluster)
       virtual Int_t FindSubClusters(BrTpcCluster* primaryCluster, TObjArray* subClusters)

    public:

      BrTpcDeconvoluteClusterModule BrTpcDeconvoluteClusterModule()
      BrTpcDeconvoluteClusterModule BrTpcDeconvoluteClusterModule(const Char_t* name, const Char_t* title)
      BrTpcDeconvoluteClusterModule BrTpcDeconvoluteClusterModule(BrTpcDeconvoluteClusterModule&)
                       virtual void ~BrTpcDeconvoluteClusterModule()
                     static TClass* Class()
                       virtual void DefineHistograms()
                       virtual void Event(BrEventNode* inputNode, BrEventNode* outputNode)
                       virtual void Init()
                    virtual TClass* IsA() const
                       virtual void Print(Option_t* option = "B") const
                       virtual void SetHighPsigCut(Float_t val = 1.6)
                       virtual void SetHighTsigCut(Float_t val = 1.6)
                       virtual void SetMinMaxADCCut(Float_t val = 10)
                       virtual void SetMinPadCut(Int_t val = 1)
                       virtual void SetMinTimeCut(Int_t val = 1)
                       virtual void ShowMembers(TMemberInspector& insp, char* parent)
                       virtual void Streamer(TBuffer& b)
                               void StreamerNVirtual(TBuffer& b)

Data Members

    protected:

        BrDataTable* fClusterTable    pointer to data (both input and output)
      BrTpcAdcTable* fAdcTable        pointer to help table
               Int_t fClusterId       For defining unique IDs.
               Int_t fCurrentRow      Used to remember row no
               Int_t fMinPadCut       Min no of pads in good cluster 
               Int_t fMinTimeCut      Min no of time bins in good cluster 
             Float_t fMinMaxADCCut    Min max ADC in good cluster 
             Float_t fHighPsigCut     Pad cut used to define multi hits
             Float_t fHighTsigCut     Time cut used to define multi hits
               TH1F* hNpads           No of pads in good clusters
               TH1F* hNtime           No of time bins in good clusters            
               TH1F* hNclusters       No of clusters after deconvolution
               TH1F* hNnoiseclusters  No of clusters that does not survive cuts
               TH1F* hNsubclusters    No of good clusters from deconvolution
               TH1F* hNpeaks          No of good clusters pr multicluster
               TH1F* hMaxADC          Max ADC values in good clusters
               TH1F* hSumADC          Integrated ADC values in good clusters
               TH1F* hPadvar          Pad variance, good clusters
               TH1F* hTimevar         Pad variance, good clusters

Class Description

   The BrTpcDeconvoluteClusterModule used to be part of
   BrTpcClusterFinder

   It takes clusters sequences that come from BrTpcClusterModule and
   examines the status. If removes noise clusters and tries to
   deconvolute multiclusters.
   The output data is (also) clusters in the input datatable

 This constructer is no good

 This is the constructor that should be used
 Name should be "T1", "T2", "TPM1", "TPM2"

 Define the histograms used to monitor performance
 This method is usually called by BrModule::Book().

 Main method
 Gets the TPCCluster data from the inputNode
 Examines the status of the clusters and tries to deconvolute
 multihits
 The output is the same table as the input so no new table is
 appended, i.e, the changes are made to the old table.

 Scan the working array and create a set of subclusters
 according to the fClusterIdTable identification.
 The appropriate set of BrTPSequence objects that make up the
 clusters are created. These are of course NOT the inputdata
 but a subset.
 This routine needs access to fCurrentRow, and the next available ClusterId
 to be used.

 Fill the adcTable with adc values in prepartion for
 subclusterfinding.

 Determine the status of each cluster
 The outcome can be either
 noisehit (throw away)
 single hit
 multi hit (deconvolution needed)

 primaryCluster is scanned with FindHitStatus to determine if its
 noise, a good single hit cluster or a multihit candidate. The
 TObjArray is empty initially. Ownership of new BrTpcClusters from
 deconvolution is transferred to this array on exit.

 Deconvolution is done as follows :
 -  Fill working array
 -  Finds peaks and work towards valleys
 -  Identify part sequences belonging to same peak
 -  Create subClusters and new BrTpcSequences for these clusters
 -

 Initialize module. This is the proper member fct for getting
 information from the data base.

 Module Information method

 Options: (see also BrModule::Print)

Inline Functions

                                 void SetHighPsigCut(Float_t val = 1.6)
                                 void SetHighTsigCut(Float_t val = 1.6)
                                 void SetMinPadCut(Int_t val = 1)
                                 void SetMinTimeCut(Int_t val = 1)
                                 void SetMinMaxADCCut(Float_t val = 10)
                              TClass* Class()
                              TClass* IsA() const
                                 void ShowMembers(TMemberInspector& insp, char* parent)
                                 void Streamer(TBuffer& b)
                                 void StreamerNVirtual(TBuffer& b)
        BrTpcDeconvoluteClusterModule BrTpcDeconvoluteClusterModule(BrTpcDeconvoluteClusterModule&)

Copyright ; 2002 BRAHMS Collaboration <brahmlib@rcf.rhic.bnl.gov>
Last Update on 2002/08/15 17:23:31 $ by videbaek $

Validate HTML
Validate CSS