ITS/GetGainModuleLevel.C

   1 #include "TMath.h"\r
   2 #include "TList.h"\r
   3 #include "TH2F.h"\r
   4 #include "TH1F.h"\r
   5 #include "TF1.h"\r
   6 #include "TH1D.h"\r
   7 #include "TProfile.h"\r
   8 #include "TFile.h"\r
   9 #include "TCanvas.h"\r
  10 #include <Riostream.h>\r
  11 #include <TROOT.h>\r
  12 #include <TString.h>\r
  13 #include "TStyle.h"\r
  14 #include "TGrid.h"\r
  15  \r
  16  \r
  17 class TList;\r
  18 class TH2F;\r
  19 class TH1F;\r
  20 class TH1D;\r
  21 class TProfile;\r
  22 class TFile;\r
  23 class TCanvas;\r
  24 class TStyle;\r
  25 class TF1;\r
  26 class TGrid;\r
  27 \r
  28 \r
  29 \r
  30 \r
  31 \r
  32 Double_t convolution(Double_t* x,Double_t *par)\r
  33 {\r
  34           //Fit parameters:\r
  35    //par[0]=Width (scale) parameter of Landau density\r
  36    //par[1]=Most Probable (MP, location) parameter of Landau density\r
  37    //par[2]=Total area (integral -inf to inf, normalization constant)\r
  38    //par[3]=Width (sigma) of convoluted Gaussian function\r
  39    //\r
  40    //In the Landau distribution (represented by the CERNLIB approximation),\r
  41    //the maximum is located at x=-0.22278298 with the location parameter=0.\r
  42    //This shift is corrected within this function, so that the actual\r
  43    //maximum is identical to the MP parameter.\r
  44 \r
  45       // Numeric constants\r
  46       Double_t invsq2pi = 0.3989422804014;   // (2 pi)^(-1/2)\r
  47       Double_t mpshift  = -0.22278298;       // Landau maximum location\r
  48 \r
  49       // Control constants\r
  50       Double_t np = 200.0;      // number of convolution steps\r
  51       Double_t sc =   5.0;      // convolution extends to +-sc Gaussian sigmas\r
  52 \r
  53       // Variables\r
  54       Double_t xx;\r
  55       Double_t mpc;\r
  56       Double_t fland;\r
  57       Double_t sum = 0.0;\r
  58       Double_t xlow,xupp;\r
  59       Double_t step;\r
  60       Double_t i;\r
  61 \r
  62 \r
  63       // MP shift correction\r
  64       mpc = par[1] - mpshift * par[0];\r
  65 \r
  66       // Range of convolution integral\r
  67       xlow = x[0] - sc * par[3];\r
  68       xupp = x[0] + sc * par[3];\r
  69 \r
  70       step = (xupp-xlow) / np;\r
  71 \r
  72       // Convolution integral of Landau and Gaussian by sum\r
  73       for(i=1.0; i<=np/2; i++) \r
  74       {\r
  75                  xx = xlow + (i-.5) * step;\r
  76                 fland = TMath::Landau(xx,mpc,par[0]) / par[0];\r
  77                 sum += fland * TMath::Gaus(x[0],xx,par[3]);\r
  78 \r
  79                 xx = xupp - (i-.5) * step;\r
  80                 fland = TMath::Landau(xx,mpc,par[0]) / par[0];\r
  81                 sum += fland * TMath::Gaus(x[0],xx,par[3]);\r
  82       }\r
  83 \r
  84       return (par[2] * step * sum * invsq2pi / par[3]);\r
  85 }\r
  86 \r
  87 \r
  88 \r
  89 \r
  90 void GetGainModuleLevel(TString filename,Bool_t normal=1,Int_t ntofit=500, Bool_t grid=0)\r
  91 {\r
  92         gROOT->SetStyle("Plain");\r
  93         gStyle->SetPalette(1,0);\r
  94         gStyle->SetOptStat(111111);\r
  95         gStyle->SetOptFit(1);   \r
  96         if(grid)\r
  97                 TGrid::Connect("alien://");\r
  98         \r
  99         TFile* file_data=TFile::Open(filename);\r
 100         if(!file_data)\r
 101                 return;\r
 102         TList *listin=0x0;\r
 103         listin=(TList*)file_data->Get("output");\r
 104         if(!listin)\r
 105                 listin=(TList*)file_data->Get("PWGPPdEdxSSDQA/output");\r
 106         if(!listin)     \r
 107                 listin=(TList*)file_data->Get("PWGPPdEdxSSDQA/SSDdEdxQA");\r
 108         if(!listin)     \r
 109                 listin=(TList*)file_data->Get("SSDdEdxQA");\r
 110         if(!listin)     \r
 111                 return;\r
 112         TH2F* fHistQ=0x0;\r
 113         if (normal)             \r
 114                 fHistQ=(TH2F*)listin ->FindObject("QACharge");\r
 115         else\r
 116                 fHistQ=(TH2F*)listin ->FindObject("QAChargeCorrected");\r
 117         if(!fHistQ)\r
 118                 return;\r
 119         TH2F* fHistCR=(TH2F*)listin ->FindObject("QAChargeRatio");\r
 120         if(!fHistCR)\r
 121                 return;\r
 122         TList *listout1=new TList();\r
 123         \r
 124         TList *listout2=new TList();\r
 125         \r
 126         TH1F* fHistMPVs=new TH1F("HistMPVS","HistMPVs;MPV;N",75,70,95);\r
 127         fHistMPVs->SetDirectory(0);\r
 128         listout2->Add(fHistMPVs);\r
 129         \r
 130         TH1F* fHistSL=new TH1F("HistSL","#sigma_{Landau};#sigma_{Landau};N",40,0,16);\r
 131         fHistSL->SetDirectory(0);\r
 132         listout2->Add(fHistSL);\r
 133         \r
 134         TH1F* fHistSG=new TH1F("HistSG","#sigma_{Gaus};#sigma_{Gaus};N",40,0,16);\r
 135         fHistSG->SetDirectory(0);\r
 136         listout2->Add(fHistSG);\r
 137         \r
 138         TH1F* fHistCRmean=new TH1F("HistCRmean","HistCRmean;CRmean;N",200,-1,1);\r
 139         fHistCRmean->SetDirectory(0);\r
 140         listout2->Add(fHistCRmean);\r
 141         \r
 142         TH1F* fHistCRRMS=new TH1F("HistCRRMS","HistCRRMS;CRRMS;N",100,0,1);\r
 143         fHistCRRMS->SetDirectory(0);\r
 144         listout2->Add(fHistCRRMS);\r
 145         \r
 146         TH1F* fHistGainP=new TH1F("HistGainP","HistGainP;CRGainPcorr;N",120,0.5,2.0);\r
 147         fHistGainP->SetDirectory(0);\r
 148         listout2->Add(fHistGainP);\r
 149         \r
 150         TH1F* fHistGainN=new TH1F("HistGainN","HistGainN;CRGainNcorr;N",120,0.5,2.0);\r
 151         fHistGainN->SetDirectory(0);\r
 152         listout2->Add(fHistGainN);\r
 153         \r
 154         TH1F *fMPVGraph = new TH1F("MPVgraph","MPVgraph;Module number;MPV",1698,-0.5,1697.5);\r
 155         fMPVGraph->SetMarkerColor(kRed);\r
 156         fMPVGraph->SetMarkerStyle(22);\r
 157         listout2->Add(fMPVGraph);\r
 158         \r
 159         TH1F *fCRmeanGraph = new TH1F("CRmeangraph","CRmeangraph;Module number;MPV",1698,-0.5,1697.5);\r
 160         fCRmeanGraph->SetMarkerColor(kBlue);\r
 161         fCRmeanGraph->SetMarkerStyle(23);\r
 162         listout2->Add(fCRmeanGraph);\r
 163         \r
 164         Float_t gainP[1698];\r
 165         Float_t gainN[1698];\r
 166         Float_t mpv[1698];\r
 167         Int_t flag[1698];       \r
 168         \r
 169         ofstream outfiletxt;\r
 170         outfiletxt.open("gain.txt");\r
 171          outfiletxt.width(10) ;\r
 172         outfiletxt.setf(outfiletxt.left);\r
 173         outfiletxt<<"MODULE"<<"\t";\r
 174         outfiletxt.width(10);\r
 175         outfiletxt.setf(outfiletxt.left);\r
 176         outfiletxt<<"FLAG"<<"\t";\r
 177         outfiletxt.width(10) ;\r
 178         outfiletxt.setf(outfiletxt.left);\r
 179         outfiletxt<<"GainPcorr"<<"\t";\r
 180         outfiletxt.width(10) ;\r
 181         outfiletxt.setf(outfiletxt.left);\r
 182         outfiletxt<<"GainNcorr"<<"\t";\r
 183         outfiletxt.width(10) ;\r
 184         outfiletxt.setf(outfiletxt.left);\r
 185         outfiletxt<<"MPV"<<endl;\r
 186         \r
 187         \r
 188         ofstream outfiletxtbad;\r
 189         outfiletxtbad.open("badModules.txt");\r
 190         \r
 191         \r
 192         \r
 193         \r
 194         for (int i =0;i<1698;i++)\r
 195         {\r
 196                 cout<<i<<endl;\r
 197                 TString tmpQ("Q");\r
 198                 tmpQ+=i;\r
 199                 TString tmpCR("CR");\r
 200                 tmpCR+=i;\r
 201                 TH1D* fHist1DCR= fHistCR->ProjectionY(tmpCR,i+1,i+1);\r
 202                 Double_t mean=fHist1DCR->GetMean();\r
 203                 if(!(TMath::Abs(mean)<1.0)||fHist1DCR->GetEntries()<10)\r
 204                 {               \r
 205                         flag[i]=-2;\r
 206                         gainN[i]=1.0;\r
 207                         gainP[i]=1.0;\r
 208                         mpv[i]=1.0;\r
 209                         continue;\r
 210                 }\r
 211                 fHistCRmean->Fill(mean);\r
 212                 fHistCRRMS->Fill(fHist1DCR->GetRMS());\r
 213                 gainN[i]=1.0/(1.0+mean);\r
 214                 gainP[i]=1.0/(1.0-mean);\r
 215                 fHistGainP->Fill(gainP[i]);\r
 216                 fHistGainN->Fill(gainN[i]);\r
 217                 fCRmeanGraph->SetBinContent(i+1,mean);\r
 218                 fCRmeanGraph->SetBinError(i+1,fHist1DCR->GetRMS());\r
 219                 \r
 220                 \r
 221                 TH1D* fHist1DQ=fHistQ->ProjectionY(tmpQ,i+1,i+1);\r
 222                  fHist1DQ->SetDirectory(0);\r
 223                  listout1->Add(fHist1DQ);\r
 224                 if(fHist1DQ->GetEntries()<ntofit)\r
 225                 {\r
 226                         flag[i]=-1;\r
 227                         mpv[i]=1.0;\r
 228                          outfiletxtbad<<"Low statistic \t module= "<<i<<" netries="<<fHist1DQ->GetEntries()<<endl;\r
 229                         continue;\r
 230                 }\r
 231                 else\r
 232                 {\r
 233                         tmpQ+="fit";\r
 234                         Float_t range=fHist1DQ->GetBinCenter(fHist1DQ->GetMaximumBin());\r
 235                         TF1 *f1 = new TF1(tmpQ,convolution,range*0.45,range*3.0,4);\r
 236                         f1->SetParameters(7.0,range,1.0,5.5);\r
 237                         Float_t normalization=fHist1DQ->GetEntries()*fHist1DQ->GetXaxis()->GetBinWidth(2)/f1->Integral(range*0.45,range*3.0);\r
 238                         f1->SetParameters(7.0,range,normalization,5.5);\r
 239                         //f1->SetParameters(7.0,range,fHist1DQ->GetMaximum(),5.5);\r
 240                         f1->SetParNames("sigma Landau","MPV","N","sigma Gaus");\r
 241                         f1->SetParLimits(0,2.0,100.0);\r
 242                         f1->SetParLimits(3,0.0,100.0);\r
 243                         if(fHist1DQ->Fit(tmpQ,"BRQ")==0)\r
 244                         {\r
 245                                 mpv[i]=f1->GetParameter(1);\r
 246                                 fHistMPVs->Fill(mpv[i]);\r
 247                                 fHistSL->Fill(f1->GetParameter(0));\r
 248                                 fHistSG->Fill(f1->GetParameter(3));\r
 249                                 flag[i]=1;              \r
 250                                 fMPVGraph->SetBinContent(i+1,f1->GetParameter(1));\r
 251                                 fMPVGraph->SetBinError(i+1,f1->GetParError(1));\r
 252                                 if(mpv[i]<75.0)\r
 253                                 {\r
 254                                         outfiletxtbad<<"MPV lower than 75 \t module="<<i<<endl;\r
 255                                         flag[i]=0;\r
 256                                 }       \r
 257                                 if(mpv[i]>100.0)\r
 258                                 {\r
 259                                         outfiletxtbad<<"MPV higher than 100 \t module="<<i<<endl;\r
 260                                         flag[i]=0;\r
 261                                         \r
 262                                 }\r
 263                                 if(f1->GetParError(1)>1.0)\r
 264                                 {\r
 265                                         outfiletxtbad<<"MPV high error on MPV  \t module="<<i<<endl;                            \r
 266                                         flag[i]=0;\r
 267                                 }\r
 268                         }\r
 269                         else\r
 270                         {\r
 271                                 mpv[i]=1;\r
 272                                 flag[i]=0;\r
 273                                  outfiletxtbad<<"BAD FIT \t module="<<i<<endl;\r
 274                                 //continue;\r
 275                         }       \r
 276                 }       \r
 277         }       \r
 278         \r
 279         for (int i=0;i<1698;i++)\r
 280         {       \r
 281                 outfiletxt.setf(outfiletxt.scientific);\r
 282                 outfiletxt.precision(2);        \r
 283                  outfiletxt.width(10) ;\r
 284                 outfiletxt.setf(outfiletxt.left);\r
 285                 outfiletxt<<i<<"\t";\r
 286                  outfiletxt.width(10) ;\r
 287                 outfiletxt.setf(outfiletxt.left);\r
 288                 outfiletxt<<flag[i]<<"\t";\r
 289                  outfiletxt.width(10) ;\r
 290                 outfiletxt.setf(outfiletxt.left);\r
 291                 outfiletxt<<gainP[i]<<"\t";\r
 292                  outfiletxt.width(10) ;\r
 293                 outfiletxt.setf(outfiletxt.left);\r
 294                 outfiletxt<<gainN[i]<<"\t";\r
 295                  outfiletxt.width(10) ;\r
 296                 outfiletxt.setf(outfiletxt.left);\r
 297                 outfiletxt<<mpv[i]<<endl;\r
 298         }\r
 299                  \r
 300         TCanvas *c1 = new TCanvas("1","1",1200,800);\r
 301         c1->Divide(2,1);\r
 302         c1->cd(1);\r
 303         fHistQ->Draw("colz");\r
 304         c1->cd(2);\r
 305         fHistCR->Draw("colz");\r
 306         \r
 307         \r
 308         \r
 309         TFile* fout1=TFile::Open("gain_all_fits.root","recreate");\r
 310         listout1->Write("output",TObject::kSingleKey);  \r
 311         fout1->Close();\r
 312         \r
 313         TFile* fout2=TFile::Open("gain_control_plots.root","recreate");\r
 314         listout2->Write("output",TObject::kSingleKey);  \r
 315         fout2->Close();\r
 316         \r
 317         \r
 318         \r
 319 }\r