HcalPulseShapes.cc

Go to the documentation of this file.

#include "CalibCalorimetry/HcalAlgos/interface/HcalPulseShapes.h"
#include "FWCore/Utilities/interface/Exception.h"
#include "FWCore/Framework/interface/ESHandle.h"
#include "FWCore/Framework/interface/EventSetup.h"
#include "FWCore/MessageLogger/interface/MessageLogger.h"
#include "CLHEP/Random/RandFlat.h"
#include "CalibFormats/HcalObjects/interface/HcalDbRecord.h"

// #include "CalibCalorimetry/HcalAlgos/interface/HcalDbASCIIIO.h"
#include <cmath>
#include <iostream>
#include <fstream>
#include "TMath.h"

HcalPulseShapes::HcalPulseShapes() 
: theDbService(nullptr),
  theShapes()
{
/*

Reco  MC
-------------------------------------------------------------------------------------------
000                                              not used (reserved)
101   101   hpdShape_                            HPD (original version)
102   102   =101                                 HPD BV 30 volts in HBP iphi54
103   123   hpdShape_v2,hpdShapeMC_v2            HPD (2011. oct version)
104   124   hpdBV30Shape_v2,hpdBV30ShapeMC_v2    HPD bv30 in HBP iph54
105   125   hpdShape_v2,hpdShapeMC_v2            HPD (2011.11.12 version)
201   201   siPMShapeHO_                         SiPMs Zecotec shape   (HO)
202   202   =201,                                SiPMs Hamamatsu shape (HO)
205   203   siPMShapeData2017_,siPMShapeMC2017_  SiPMs from Data, Hamamatsu shape (HE 2017)
207   206   siPMShapeData2018_,siPMShapeMC2018_  SiPMs from Data, Hamamatsu shape (HE 2018)
301   301   hfShape_                             regular HF PMT shape
401   401                                        regular ZDC shape
-------------------------------------------------------------------------------------------

*/

 
  float ts1, ts2, ts3, thpd, tpre, wd1, wd2, wd3;

  //  HPD Shape  Version 1 (used before CMSSW5, until Oct 2011)
  ts1=8. ; ts2=10. ; ts3=29.3; thpd=4.0; tpre=9.0; wd1=2.0; wd2=0.7; wd3=1.0;  
  computeHPDShape(ts1,ts2,ts3,thpd,tpre,wd1,wd2,wd3, hpdShape_);
  theShapes[101] = &hpdShape_;
  theShapes[102] = theShapes[101];

  //  HPD Shape  Version 2 for CMSSW 5. Nov 2011  (RECO and MC separately)
  ts1=8. ; ts2=10. ; ts3=25.0; thpd=4.0; tpre=9.0; wd1=2.0; wd2=0.7; wd3=1.0;
  computeHPDShape(ts1,ts2,ts3,thpd,tpre,wd1,wd2,wd3, hpdShape_v2);
  theShapes[103] = &hpdShape_v2;

  ts1=8. ; ts2=10. ; ts3=29.3; thpd=4.0; tpre=7.0; wd1=2.0; wd2=0.7; wd3=1.0;
  computeHPDShape(ts1,ts2,ts3,thpd,tpre,wd1,wd2,wd3, hpdShapeMC_v2);
  theShapes[123] = &hpdShapeMC_v2;

  //  HPD Shape  Version 3 for CMSSW 5. Nov 2011  (RECO and MC separately)
  ts1=8. ; ts2=19. ; ts3=29.3; thpd=4.0; tpre=9.0; wd1=2.0; wd2=0.7; wd3=0.32;
  computeHPDShape(ts1,ts2,ts3,thpd,tpre,wd1,wd2,wd3, hpdShape_v3);
  theShapes[105] = &hpdShape_v3;

  ts1=8. ; ts2=10. ; ts3=22.3; thpd=4.0; tpre=7.0; wd1=2.0; wd2=0.7; wd3=1.0;
  computeHPDShape(ts1,ts2,ts3,thpd,tpre,wd1,wd2,wd3, hpdShapeMC_v3);
  theShapes[125] = &hpdShapeMC_v3;

  // HPD with Bias Voltage 30 volts, wider pulse.  (HBPlus iphi54)

  ts1=8. ; ts2=12. ; ts3=31.7; thpd=9.0; tpre=9.0; wd1=2.0; wd2=0.7; wd3=1.0;
  computeHPDShape(ts1,ts2,ts3,thpd,tpre,wd1,wd2,wd3, hpdBV30Shape_v2);
  theShapes[104] = &hpdBV30Shape_v2;

  ts1=8. ; ts2=12. ; ts3=31.7; thpd=9.0; tpre=9.0; wd1=2.0; wd2=0.7; wd3=1.0;
  computeHPDShape(ts1,ts2,ts3,thpd,tpre,wd1,wd2,wd3, hpdBV30ShapeMC_v2);
  theShapes[124] = &hpdBV30ShapeMC_v2;

  // HF and SiPM

  computeHFShape();
  computeSiPMShapeHO();
  computeSiPMShapeHE203();
  computeSiPMShapeHE206();
  computeSiPMShapeData2017();
  computeSiPMShapeData2018();

  theShapes[201] = &siPMShapeHO_;
  theShapes[202] = theShapes[201];
  theShapes[203] = &siPMShapeMC2017_;
  theShapes[205] = &siPMShapeData2017_;
  theShapes[206] = &siPMShapeMC2018_;
  theShapes[207] = &siPMShapeData2018_;
  theShapes[301] = &hfShape_;
  //theShapes[401] = new CaloCachedShapeIntegrator(&theZDCShape);

}


HcalPulseShapes::~HcalPulseShapes() {
}


void HcalPulseShapes::beginRun(edm::EventSetup const & es)
{
  edm::ESHandle<HcalDbService> conditions;
  es.get<HcalDbRecord>().get(conditions);
  theDbService = conditions.product();
}

void HcalPulseShapes::beginRun(const HcalDbService* conditions)
{
  theDbService = conditions;
}

//void HcalPulseShapes::computeHPDShape()
void HcalPulseShapes::computeHPDShape(float ts1, float ts2, float ts3, float thpd, float tpre,
                                float wd1, float wd2, float wd3, Shape &tmphpdShape_)
{

// pulse shape time constants in ns
/*
  const float ts1  = 8.;          // scintillation time constants : 1,2,3
  const float ts2  = 10.;           
  const float ts3  = 29.3;         
  const float thpd = 4.;          // HPD current collection drift time
  const float tpre = 9.;          // preamp time constant (refit on TB04 data)
  
  const float wd1 = 2.;           // relative weights of decay exponents 
  const float wd2 = 0.7;
  const float wd3 = 1.;
*/  
  // pulse shape components over a range of time 0 ns to 255 ns in 1 ns steps
  unsigned int nbin = 256;
  tmphpdShape_.setNBin(nbin);
  std::vector<float> ntmp(nbin,0.0);  // zeroing output pulse shape
  std::vector<float> nth(nbin,0.0);   // zeroing HPD drift shape
  std::vector<float> ntp(nbin,0.0);   // zeroing Binkley preamp shape
  std::vector<float> ntd(nbin,0.0);   // zeroing Scintillator decay shape

  unsigned int i,j,k;
  float norm;

  // HPD starts at I and rises to 2I in thpd of time
  norm=0.0;
  for(j=0;j<thpd && j<nbin;j++){
    nth[j] = 1.0 + ((float)j)/thpd;
    norm += nth[j];
  }
  // normalize integrated current to 1.0
  for(j=0;j<thpd && j<nbin;j++){
    nth[j] /= norm;
  }
  
  // Binkley shape over 6 time constants
  norm=0.0;
  for(j=0;j<6*tpre && j<nbin;j++){
    ntp[j] = ((float)j)*exp(-((float)(j*j))/(tpre*tpre));
    norm += ntp[j];
  }
  // normalize pulse area to 1.0
  for(j=0;j<6*tpre && j<nbin;j++){
    ntp[j] /= norm;
  }

// ignore stochastic variation of photoelectron emission
// <...>

// effective tile plus wave-length shifter decay time over 4 time constants
  unsigned int tmax = 6 * (int)ts3;
 
  norm=0.0;
  for(j=0;j<tmax && j<nbin;j++){
    ntd[j] = wd1 * exp(-((float)j)/ts1) + 
      wd2 * exp(-((float)j)/ts2) + 
      wd3 * exp(-((float)j)/ts3) ; 
    norm += ntd[j];
  }
  // normalize pulse area to 1.0
  for(j=0;j<tmax && j<nbin;j++){
    ntd[j] /= norm;
  }
  
  unsigned int t1,t2,t3,t4;
  for(i=0;i<tmax && i<nbin;i++){
    t1 = i;
    //    t2 = t1 + top*rand;
    // ignoring jitter from optical path length
    t2 = t1;
    for(j=0;j<thpd && j<nbin;j++){
      t3 = t2 + j;
      for(k=0;k<4*tpre && k<nbin;k++){       // here "4" is set deliberately,
    t4 = t3 + k;                         // as in test fortran toy MC ...
    if(t4<nbin){                         
      unsigned int ntb=t4;                        
      ntmp[ntb] += ntd[i]*nth[j]*ntp[k];
    }
      }
    }
  }
  
  // normalize for 1 GeV pulse height
  norm = 0.;
  for(i=0;i<nbin;i++){
    norm += ntmp[i];
  }

  for(i=0; i<nbin; i++){
    ntmp[i] /= norm;
  }

  for(i=0; i<nbin; i++){
    tmphpdShape_.setShapeBin(i,ntmp[i]);
  }
}

void HcalPulseShapes::computeHFShape() {
  // first create pulse shape over a range of time 0 ns to 255 ns in 1 ns steps
  unsigned int nbin = 256;
  hfShape_.setNBin(nbin);
  std::vector<float> ntmp(nbin,0.0);  // 

  const float k0=0.7956; // shape parameters
  const float p2=1.355;
  const float p4=2.327;
  const float p1=4.3;    // position parameter

  float norm = 0.0;

  for(unsigned int j = 0; j < 25 && j < nbin; ++j){

    float r0 = j-p1;
    float sigma0 = (r0<0) ? p2 : p2*p4;
    r0 /= sigma0;
    if(r0 < k0) ntmp[j] = exp(-0.5*r0*r0);
    else ntmp[j] = exp(0.5*k0*k0-k0*r0);
    norm += ntmp[j];
  }
  // normalize pulse area to 1.0
  for(unsigned int j = 0; j < 25 && j < nbin; ++j){
    ntmp[j] /= norm;
    hfShape_.setShapeBin(j,ntmp[j]);
  }
}


void HcalPulseShapes::computeSiPMShapeData2018()
{

   //Combination of all phase scan data (May,Jul,Oct2017)
   //runs:  294736-294740, 294929-294950, 298594-298598 and 305744-305758

  unsigned int nbin = 250;

  std::array<float, 250> nt {{
    5.22174e-12, 7.04852e-10, 3.49584e-08, 7.78029e-07, 9.11847e-06, 6.39666e-05, 0.000297587, 0.000996661, 0.00256618, 0.00535396,
    0.00944073, 0.0145521, 0.020145, 0.0255936, 0.0303632, 0.0341078, 0.0366849, 0.0381183, 0.0385392, 0.0381327,
    0.0370956, 0.0356113, 0.0338366, 0.0318978, 0.029891, 0.0278866, 0.0259336, 0.0240643, 0.0222981, 0.0206453,
    0.0191097, 0.0176902, 0.0163832, 0.0151829, 0.0140826, 0.0130752, 0.0121533, 0.01131, 0.0105382, 0.00983178,
    0.00918467, 0.00859143, 0.00804709, 0.0075471, 0.00708733, 0.00666406, 0.00627393, 0.00591389, 0.00558122, 0.00527344,
    0.00498834, 0.00472392, 0.00447837, 0.00425007, 0.00403754, 0.00383947, 0.00365465, 0.00348199, 0.00332052, 0.00316934,
    0.00302764, 0.0028947, 0.00276983, 0.00265242, 0.00254193, 0.00243785, 0.00233971, 0.00224709, 0.0021596, 0.00207687,
    0.0019986, 0.00192447, 0.00185421, 0.00178756, 0.0017243, 0.00166419, 0.00160705, 0.00155268, 0.00150093, 0.00145162,
    0.00140461, 0.00135976, 0.00131696, 0.00127607, 0.00123699, 0.00119962, 0.00116386, 0.00112963, 0.00109683, 0.0010654,
    0.00103526, 0.00100634, 0.000978578, 0.000951917, 0.000926299, 0.000901672, 0.000877987, 0.000855198, 0.00083326, 0.000812133,
    0.000791778, 0.000772159, 0.000753242, 0.000734994, 0.000717384, 0.000700385, 0.000683967, 0.000668107, 0.000652779, 0.00063796,
    0.000623629, 0.000609764, 0.000596346, 0.000583356, 0.000570777, 0.000558592, 0.000546785, 0.00053534, 0.000524243, 0.000513481,
    0.00050304, 0.000492907, 0.000483072, 0.000473523, 0.000464248, 0.000455238, 0.000446483, 0.000437974, 0.0004297, 0.000421655,
    0.00041383, 0.000406216, 0.000398807, 0.000391595, 0.000384574, 0.000377736, 0.000371076, 0.000364588, 0.000358266, 0.000352104,
    0.000346097, 0.00034024, 0.000334528, 0.000328956, 0.00032352, 0.000318216, 0.000313039, 0.000307986, 0.000303052, 0.000298234,
    0.000293528, 0.000288931, 0.000284439, 0.00028005, 0.000275761, 0.000271567, 0.000267468, 0.000263459, 0.000259538, 0.000255703,
    0.000251951, 0.00024828, 0.000244688, 0.000241172, 0.00023773, 0.000234361, 0.000231061, 0.00022783, 0.000224666, 0.000221566,
    0.000218528, 0.000215553, 0.000212636, 0.000209778, 0.000206977, 0.00020423, 0.000201537, 0.000198896, 0.000196307, 0.000193767,
    0.000191275, 0.000188831, 0.000186432, 0.000184079, 0.000181769, 0.000179502, 0.000177277, 0.000175092, 0.000172947, 0.000170841,
    0.000168772, 0.000166741, 0.000164745, 0.000162785, 0.000160859, 0.000158967, 0.000157108, 0.00015528, 0.000153484, 0.000151719,
    0.000149984, 0.000148278, 0.000146601, 0.000144951, 0.000143329, 0.000141734, 0.000140165, 0.000138622, 0.000137104, 0.00013561,
    0.000134141, 0.000132695, 0.000131272, 0.000129871, 0.000128493, 0.000127136, 0.000125801, 0.000124486, 0.000123191, 0.000121917,
    0.000120662, 0.000119426, 0.000118209, 0.00011701, 0.000115829, 0.000114665, 0.000113519, 0.00011239, 0.000111278, 0.000110182,
    0.000109102, 0.000108037, 0.000106988, 0.000105954, 0.000104935, 0.00010393, 0.000102939, 0.000101963, 0.000101, 0.000100051,
    9.91146e-05, 9.81915e-05, 9.7281e-05, 9.63831e-05, 9.54975e-05, 9.46239e-05, 9.37621e-05, 9.2912e-05, 9.20733e-05, 9.12458e-05
  }};
 
  siPMShapeData2018_.setNBin(nbin);

  double norm = 0.;
  for (unsigned int j = 0; j < nbin; ++j) {
    norm += (nt[j]>0) ? nt[j] : 0.;
  }

  for (unsigned int j = 0; j < nbin; ++j) {
    nt[j] /= norm;
    siPMShapeData2018_.setShapeBin(j,nt[j]);
  }
}
 


void HcalPulseShapes::computeSiPMShapeData2017()
{
  //From Jay Lawhorn: derived from data Edward Laird phase scan may2017
  //https://indico.cern.ch/event/641978/contributions/2604491/attachments/1468666/2271582/17-05-31-hcal-hep17-pulse-shape.pdf
  //Run numbers are 294736-294740 and 294929-294950

  unsigned int nbin = 250;

  std::array<float, 250> nt {{
        3.97958e-29, 1.11634e-22, 9.96106e-18, 6.25334e-14, 5.08863e-11, 8.59141e-09, 4.32285e-07, 8.56617e-06, 8.28549e-05, 0.000461447,
        0.00168052, 0.00441395, 0.00901637, 0.0151806, 0.0220314, 0.028528, 0.0338471, 0.0375578, 0.0395985, 0.0401567,
        0.0395398, 0.0380776, 0.0360669, 0.0337474, 0.0312984, 0.0288457, 0.0264721, 0.0242276, 0.0221393, 0.0202181,
        0.0184647, 0.0168731, 0.0154335, 0.0141346, 0.0129639, 0.0119094, 0.0109594, 0.0101031, 0.0093305, 0.00863267,
        0.0080015, 0.00742977, 0.00691107, 0.00643969, 0.00601059, 0.00561931, 0.00526188, 0.00493483, 0.00463505, 0.00435981,
        0.00410667, 0.00387348, 0.00365832, 0.00345949, 0.00327547, 0.0031049, 0.00294656, 0.00279938, 0.00266237, 0.00253467,
        0.00241548, 0.0023041, 0.00219989, 0.00210227, 0.00201072, 0.00192476, 0.00184397, 0.00176795, 0.00169634, 0.00162884,
        0.00156512, 0.00150494, 0.00144803, 0.00139418, 0.00134317, 0.00129481, 0.00124894, 0.00120537, 0.00116398, 0.00112461,
        0.00108715, 0.00105147, 0.00101747, 0.000985042, 0.000954096, 0.000924545, 0.000896308, 0.000869311, 0.000843482, 0.000818758,
        0.000795077, 0.000772383, 0.000750623, 0.000729747, 0.00070971, 0.000690466, 0.000671977, 0.000654204, 0.00063711, 0.000620663,
        0.000604831, 0.000589584, 0.000574894, 0.000560735, 0.000547081, 0.00053391, 0.0005212, 0.000508929, 0.000497078, 0.000485628,
        0.000474561, 0.000463862, 0.000453514, 0.000443501, 0.000433811, 0.000424429, 0.000415343, 0.00040654, 0.00039801, 0.000389741,
        0.000381722, 0.000373944, 0.000366398, 0.000359074, 0.000351964, 0.00034506, 0.000338353, 0.000331838, 0.000325505, 0.00031935,
        0.000313365, 0.000307544, 0.000301881, 0.000296371, 0.000291009, 0.000285788, 0.000280705, 0.000275755, 0.000270932, 0.000266233,
        0.000261653, 0.00025719, 0.000252837, 0.000248593, 0.000244454, 0.000240416, 0.000236475, 0.00023263, 0.000228876, 0.000225212,
        0.000221633, 0.000218138, 0.000214724, 0.000211389, 0.00020813, 0.000204945, 0.000201831, 0.000198787, 0.000195811, 0.0001929,
        0.000190053, 0.000187268, 0.000184543, 0.000181876, 0.000179266, 0.000176711, 0.00017421, 0.000171761, 0.000169363, 0.000167014,
        0.000164713, 0.000162459, 0.00016025, 0.000158086, 0.000155964, 0.000153885, 0.000151847, 0.000149848, 0.000147888, 0.000145966,
        0.000144081, 0.000142232, 0.000140418, 0.000138638, 0.000136891, 0.000135177, 0.000133494, 0.000131843, 0.000130221, 0.00012863,
        0.000127066, 0.000125531, 0.000124023, 0.000122543, 0.000121088, 0.000119658, 0.000118254, 0.000116874, 0.000115518, 0.000114185,
        0.000112875, 0.000111587, 0.000110321, 0.000109076, 0.000107851, 0.000106648, 0.000105464, 0.000104299, 0.000103154, 0.000102027,
        0.000100918, 9.98271e-05, 9.87537e-05, 9.76974e-05, 9.66578e-05, 9.56346e-05, 9.46274e-05, 9.3636e-05, 9.26599e-05, 9.16989e-05,
        9.07526e-05, 8.98208e-05, 8.89032e-05, 8.79995e-05, 8.71093e-05, 8.62325e-05, 8.53688e-05, 8.45179e-05, 8.36796e-05, 8.28536e-05,
        8.20397e-05, 8.12376e-05, 8.04471e-05, 7.96681e-05, 7.89002e-05, 7.81433e-05, 7.73972e-05, 7.66616e-05, 7.59364e-05, 7.52213e-05,
        7.45163e-05, 7.3821e-05, 7.31354e-05, 7.24592e-05, 7.17923e-05, 7.11345e-05, 7.04856e-05, 6.98455e-05, 6.9214e-05, 6.8591e-05
  }};


  siPMShapeData2017_.setNBin(nbin);

  double norm = 0.;
  for (unsigned int j = 0; j < nbin; ++j) {
    norm += (nt[j]>0) ? nt[j] : 0.;
  }

  for (unsigned int j = 0; j < nbin; ++j) {
    nt[j] /= norm;
    siPMShapeData2017_.setShapeBin(j,nt[j]);
  }
}

void HcalPulseShapes::computeSiPMShapeHO()
{

  unsigned int nbin = 128; 

//From Jake Anderson: toy MC convolution of SiPM pulse + WLS fiber shape + SiPM nonlinear response
  std::array<float, 128> nt {{
        2.782980485851731e-6, 4.518134885954626e-5, 2.7689305197392056e-4, 9.18328418900969e-4, .002110072599166349, .003867856860331454, .006120046224897771, .008754774090536956,
        0.0116469503358586, .01467007449455966, .01770489955229477, .02064621450689512, .02340678093764222, .02591874610854916, .02813325527435303, 0.0300189241965647,
        .03155968107671164, .03275234052577155, .03360415306318798, .03413048377960748, .03435270899678218, .03429637464659661, .03398962975487166, .03346192884394954,
        .03274298516247742, .03186195009136525, .03084679116113031, 0.0297238406141036, .02851748748929785, .02724998816332392, .02594137274487424, .02460942736731527,
        .02326973510736116, .02193576080366117, 0.0206189674254987, .01932895378564653, 0.0180736052958666, .01685925112650875, 0.0156908225633535, .01457200857138456,
        .01350540559602467, .01249265947824805, .01153459805300423, .01063135355597282, .009782474412011936, .008987026319784546, 0.00824368281357106, .007550805679909604,
        .006906515742762193, .006308754629755056, .005755338185695127, .005244002229973356, .004772441359900532, .004338341490928299, .003939406800854143, 0.00357338171220501,
        0.0032380685079891, .002931341133259233, .002651155690306086, .002395558090237333, .002162689279320922, .001950788415487319, .001758194329648101, .001583345567913682,
        .001424779275191974, .001281129147671334, 0.00115112265163774, .001033577678808199, 9.273987838127585e-4, 8.315731274976846e-4, 7.451662302008696e-4, 6.673176219006913e-4,
        5.972364609644049e-4, 5.341971801529036e-4, 4.775352065178378e-4, 4.266427928961177e-4, 3.8096498904225923e-4, 3.3999577417327287e-4, 3.032743659102713e-4, 2.703817158798329e-4,
        2.4093719775272793e-4, 2.145954900503894e-4, 1.9104365317752797e-4, 1.6999839784346724e-4, 1.5120354022478893e-4, 1.3442763782650755e-4, 1.1946179895521507e-4, 1.0611765796993575e-4,
        9.422550797617687e-5, 8.363258233342666e-5, 7.420147621931836e-5, 6.580869950304933e-5, 5.834335229919868e-5, 5.17059147771959e-5, 4.5807143072062634e-5, 4.0567063461299446e-5,
        3.591405732740723e-5, 3.178402980354131e-5, 2.811965539165646e-5, 2.4869694240316126e-5, 2.1988373166730962e-5, 1.9434825899529382e-5, 1.717258740121378e-5, 1.5169137499243157e-5,
        1.339548941011129e-5, 1.1825819079078403e-5, 1.0437131581057595e-5, 9.208961130078894e-6, 8.12310153137994e-6, 7.163364176588591e-6, 6.315360932244386e-6, 5.566309502463164e-6,
        4.904859063429651e-6, 4.320934164082596e-6, 3.8055950719111903e-6, 3.350912911083174e-6, 2.9498580949517117e-6, 2.596200697612328e-6, 2.2844215378879293e-6, 2.0096328693141094e-6,
        1.7675076766686654e-6, 1.5542166787225756e-6, 1.366372225473431e-6, 1.200978365778838e-6, 1.0553864128982371e-6, 9.272554464808518e-7, 8.145171945902259e-7, 7.153448381918271e-7
  }};

  siPMShapeHO_.setNBin(nbin);

  double norm = 0.;
  for (unsigned int j = 0; j < nbin; ++j) {
    norm += (nt[j]>0) ? nt[j] : 0.;
  }

  for (unsigned int j = 0; j < nbin; ++j) {
    nt[j] /= norm;
    siPMShapeHO_.setShapeBin(j,nt[j]);
  }
}

void HcalPulseShapes::computeSiPMShapeHE203()
{
  //numerical convolution of SiPM pulse + WLS fiber shape
  std::vector<double> nt = convolve(nBinsSiPM_,analyticPulseShapeSiPMHE,Y11203);

  siPMShapeMC2017_.setNBin(nBinsSiPM_);

  //skip first bin, always 0
  double norm = 0.;
  for (unsigned int j = 1; j <= nBinsSiPM_; ++j) {
    norm += (nt[j]>0) ? nt[j] : 0.;
  }

  for (unsigned int j = 1; j <= nBinsSiPM_; ++j) {
    nt[j] /= norm;
    siPMShapeMC2017_.setShapeBin(j,nt[j]);
  }
}

void HcalPulseShapes::computeSiPMShapeHE206()
{
  //numerical convolution of SiPM pulse + WLS fiber shape
  std::vector<double> nt = convolve(nBinsSiPM_,analyticPulseShapeSiPMHE,Y11206);

  siPMShapeMC2018_.setNBin(nBinsSiPM_);

  //Aligning 206 phase closer to 205 in order to have good reco agreement
  int shift = -2;

  //skip first bin, always 0
  double norm = 0.;
  for (unsigned int j = std::max(1,-1*shift); j<=nBinsSiPM_; j++) {
    norm += std::max(0., nt[j-shift]);
  }
  double normInv=1./norm;
  for ( int j = 1; j<=nBinsSiPM_; j++) {
    if ( j-shift>=0 ) {
      nt[j-shift]*=normInv;
      siPMShapeMC2018_.setShapeBin(j,nt[j-shift]);
    }
    else{
      siPMShapeMC2018_.setShapeBin(j,0);
    }
  }
}

const HcalPulseShapes::Shape &
HcalPulseShapes::getShape(int shapeType) const
{
  ShapeMap::const_iterator shapeMapItr = theShapes.find(shapeType);
  if(shapeMapItr == theShapes.end()) {
   throw cms::Exception("HcalPulseShapes") << "unknown shapeType";
   return  hpdShape_;   // should not return this, but...
  } else {
    return *(shapeMapItr->second);
  }
}


const HcalPulseShapes::Shape &
HcalPulseShapes::shape(const HcalDetId & detId) const
{
  if(!theDbService) {
    return defaultShape(detId);
  }
  int shapeType = theDbService->getHcalMCParam(detId)->signalShape();

  ShapeMap::const_iterator shapeMapItr = theShapes.find(shapeType);
  if(shapeMapItr == theShapes.end()) {
    return defaultShape(detId);
  } else {
    return *(shapeMapItr->second);
  }
}

const HcalPulseShapes::Shape &
HcalPulseShapes::shapeForReco(const HcalDetId & detId) const
{
  if(!theDbService) {
    return defaultShape(detId);
  }
  int shapeType = theDbService->getHcalRecoParam(detId.rawId())->pulseShapeID();

  ShapeMap::const_iterator shapeMapItr = theShapes.find(shapeType);
  if(shapeMapItr == theShapes.end()) {
    return defaultShape(detId);
  } else {
    return *(shapeMapItr->second);
  }
}


const HcalPulseShapes::Shape &
HcalPulseShapes::defaultShape(const HcalDetId & detId) const
{
  edm::LogWarning("HcalPulseShapes") << "Cannot find HCAL MC Params ";
  HcalSubdetector subdet = detId.subdet();
  switch(subdet) {
  case HcalBarrel:
    return hbShape();
  case HcalEndcap:
    return heShape();
  case HcalForward:
    return hfShape();
  case HcalOuter:
    //FIXME doesn't look for SiPMs
    return hoShape(false);
  default:
    throw cms::Exception("HcalPulseShapes") << "unknown detId";
    break;
  }
}

//SiPM helpers

inline double gexp(double t, double A, double c, double t0, double s) {
  static double const root2(sqrt(2));
  return -A*0.5*exp(c*t+0.5*c*c*s*s-c*s)*(erf(-0.5*root2/s*(t-t0+c*s*s))-1);
}

inline double onePulse(double t, double A, double sigma, double theta, double m) {
  return (t<theta) ? 0 : A*TMath::LogNormal(t,sigma,theta,m);
}

double HcalPulseShapes::analyticPulseShapeSiPMHO(double t) {
  // HO SiPM pulse shape fit from Jake Anderson ca. 2013
  double A1(0.08757), c1(-0.5257), t01(2.4013), s1(0.6721);
  double A2(0.007598), c2(-0.1501), t02(6.9412), s2(0.8710);
  return gexp(t,A1,c1,t01,s1) + gexp(t,A2,c2,t02,s2);
}

double HcalPulseShapes::analyticPulseShapeSiPMHE(double t) {
  // taken from fit to laser measurement taken by Iouri M. in Spring 2016.
  double A1(5.204/6.94419), sigma1_shape(0.5387), theta1_loc(-0.3976), m1_scale(4.428);
  double A2(1.855/6.94419), sigma2_shape(0.8132), theta2_loc(7.025),   m2_scale(12.29);
  return
    onePulse(t,A1,sigma1_shape,theta1_loc,m1_scale) +
    onePulse(t,A2,sigma2_shape,theta2_loc,m2_scale);
}

double HcalPulseShapes::generatePhotonTime(CLHEP::HepRandomEngine* engine, unsigned int signalShape) {
  if(signalShape==206) return generatePhotonTime206(engine);
  else return generatePhotonTime203(engine);
}

double HcalPulseShapes::generatePhotonTime203(CLHEP::HepRandomEngine* engine) {
  double result(0.);
  while (true) {
    result = CLHEP::RandFlat::shoot(engine, HcalPulseShapes::Y11RANGE_);
    if (CLHEP::RandFlat::shoot(engine, HcalPulseShapes::Y11MAX203_) < HcalPulseShapes::Y11203(result))
      return result;
  }
}

double HcalPulseShapes::generatePhotonTime206(CLHEP::HepRandomEngine* engine) {
  double result(0.);
  while (true) {
    result = CLHEP::RandFlat::shoot(engine, HcalPulseShapes::Y11RANGE_);
    if (CLHEP::RandFlat::shoot(engine, HcalPulseShapes::Y11MAX206_) < HcalPulseShapes::Y11206(result))
      return result;
  }
}

//Original scintillator+Y11 fit from Vasken's 2001 measurement
double HcalPulseShapes::Y11203(double t) {
  return exp(-0.0635-0.1518*t + log(t)*2.528)/2485.9;
}

//New scintillator+Y11 model from Vasken's 2017 measurement plus a Landau correction term
double HcalPulseShapes::Y11206(double t) {
  //Shifting phase to have better comparison of digi shape with data
  //If necessary, further digi phase adjustment can be done here:
  //SimCalorimetry/HcalSimProducers/python/hcalSimParameters_cfi.py 
  //by changing "timePhase"
  double shift = 7.2;

  //Fit From Deconvolved Data
  double A,n,t0,fit;
  A=0.104204; n=0.44064; t0=10.0186;
  if(t>shift) fit = A*(1-exp(-(t-shift)/n))*exp(-(t-shift)/t0);
  else fit = 0.0;

  //Correction Term
  double norm,mpv,sigma,corTerm;
  norm=0.0809882; mpv=0; sigma=20;
  if(t>shift) corTerm = norm*TMath::Landau((t-shift),mpv,sigma);
  else corTerm = 0.0;

  //Overall Y11
  double frac = 0.11;
  double val  = (1-frac)*fit + frac*corTerm;

  if(val >= 0) return val;
  else return 0.0;
}