visiavg.cc 12.7 KB
Newer Older
OP's avatar
OP committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139
// Utilisation de SOPHYA pour faciliter les tests ...
#include "sopnamsp.h"
#include "machdefs.h"

/* ---------------------------------------------------------- 
   Projet BAORadio/PAON4 - (C) LAL/IRFU  2017

   visiavg: programme de lecture des fichiers matrices de 
   visibilites de PAON4, calcul de visibilities moyennes 
    en bin de temps et de frequence
   O. Perdereau, R.Ansari   -  LAL
   ---------------------------------------------------------- */

// include standard c/c++
#include <math.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

#include <iostream>
#include <string>

#include "pexceptions.h"
#include "tvector.h"
#include "fioarr.h"
// #include "tarrinit.h"
#include "ntuple.h" 
#include "datatable.h" 
#include "histinit.h" 
#include "matharr.h" 
#include "timestamp.h"
#include <utilarr.h>

// include sophya mesure ressource CPU/memoire ...
#include "resusage.h"
#include "ctimer.h"
#include "timing.h"

// include lecteur de fichiers visibilites 
#include "p4autils.h"
#include "visip4reader.h"
#include "p4gnugain.h"

int Usage(void);
int Usage(void)
{
  cout << " --- visiavg.cc : Read PPF files produced by mfacq time-frequency\n" << endl;
  cout << " Usage: visiavg [-arguments] \n" << endl;
  P4AnaParams::UsageOptions();
  cout<< endl;
  return 1;
}

//----------------------------------------------------
int main(int narg, const char* arg[])
{
  // --- Decoding parameters 
  if( (narg<2) || ((narg>1)&&(strcmp(arg[1],"-h")==0) ) )  return Usage();
  P4AnaParams params;
  params.DecodeArgs(narg, arg);
  string outfile = params.outfile_;
  if (outfile.length()<1)  outfile = "visavg.ppf";
  int deltaIavg = params.TFMtimebin_;
  sa_size_t TFMfbin = params.TFMfreqbin_;
  int Imin = params.Imin_, Imax = params.Imax_, Istep = params.Istep_; 
  int prtlev = params.prtlev_;
  bool FgTFMAC = true;
  bool FgTFMCX = true;
  string desctfmap;
  bool FgTFM = params.fgTFM_;   // true -> create time-frequency maps

  params.Print(cout);
  cout <<"visiavg/Info: Path BAO5:"<<params.inpath5_<<" BAO6:"<<params.inpath6_<<"\n"
       <<"fgreorderfreq="<<params.fgreorderfreq_<<"\n"
       <<"Imin,max,step="<<Imin<<","<<Imax<<","<<Istep<<" DeltaIAvg="<<deltaIavg<<"\n"
       <<"outfile="<<outfile<<" PrtLev="<<prtlev<<endl;

  if (!FgTFM) {
    cout<<" visiavg/parameter error : specify Time-Frequency map parameter with -tfm "<<endl;
    return 5;
  }

  P4AVisiNumEncoder  visiencod;
  vector<sa_size_t> KVAC = visiencod.getAllAutoCor();
  vector<sa_size_t> KVCXHH = visiencod.getAllHCrossCor();
  cout << " List of AutoCorrelation rows:"<<endl;
  for(size_t k=0; k<KVAC.size(); k++) {
    cout << "KVAC["<<k<<"]="<<KVAC[k]<<"  ->"<<visiencod.Convert2VisiName(KVAC[k])<<endl;
  }
  cout << " List of HH X-cor rows:"<<endl;
  for(size_t k=0; k<KVCXHH.size(); k++) {
    cout << "KVCXHH["<<k<<"]="<<KVCXHH[k]<<"  ->"<<visiencod.Convert2VisiName(KVCXHH[k])<<endl;
  }
  // --- Open file to store visibility matrices if requested

  // ---
  HiStatsInitiator _inia;
  int rc = 0;
  try {
    ResourceUsage resu;

    // Gain correction class
    P4gnuGain p4g(params.gain_gnu_file_);

    // setting up input visi reader
    vector<string> paths; 
    paths.push_back(params.inpath5_); 
    paths.push_back(params.inpath6_);

    
    VisiP4ReaderBase * reader = VisiP4ReaderBase::getReader(paths);
    VisiP4ReaderBase & vreader = (*reader);
    vreader.setFreqReordering(params.fgreorderfreq_);
    if (!params.fgserall_ && !params.fgtmsel_) {
      cout << " vreader.SelectSerialNum(Imin="<<Imin<<" ,Imax="<<Imax<<" ,Istep="<<Istep<<")"<<endl;
      vreader.SelectSerialNum(Imin,Imax,Istep);
    }
    else if (params.fgserall_) {
      cout << " vreader.SelectAll() ... " << endl;
      vreader.SelectAll();
    }
    else {
      TimeStamp tustart = params.tmsel_tu_;  tustart.ShiftSeconds(-params.tmsel_duration_*30.);
      TimeStamp tuend = params.tmsel_tu_;  tuend.ShiftSeconds(params.tmsel_duration_*30.);
      cout << " vreader.SelectTimeFrame(TUStart="<<tustart.ToString()<<" ,TUEnd="<<tuend.ToString()
	   <<" ,Istep="<<Istep<<")"<<endl;
      vreader.SelectTimeFrame(tustart, tuend, Istep);
    }
    vreader.setPrintLevel(prtlev);

    Imin = vreader.getSerialFirst(); Imax =  vreader.getSerialLast();   Istep =  vreader.getSerialStep();
    cout << "visiavg/Info: processing visibility matrix serial/sequence number range "
	 <<Imin<<" <= seq <= " << Imax << " with step="<<Istep<<endl;

    bool fgok=true;
    // un vecteur avec les temps 
    TVector< double > timevec((Imax-Imin)/Istep/deltaIavg); 
    TMatrix< complex<r_4> > vismtx;
    TMatrix< complex<r_4> > acsum;
OP's avatar
OP committed
140
    TMatrix< r_4 > acsum_sq; // wil sum only the real part ^2
OP's avatar
OP committed
141
    TMatrix< complex<r_4> > cxsum;
OP's avatar
OP committed
142 143 144
    // for sums of real and imag parts 
    TMatrix< r_4 > cxsum_sq_rp;
    TMatrix< r_4 > cxsum_sq_ip;
OP's avatar
OP committed
145 146 147 148 149 150

    TimeStamp dateobs, cfdate,datestart;
    TimeStamp dateorg(2015,1,1,12,0,0.);  // Date origine 1 jan 2015
    double mttag;
    int cnt=0, cntnt=0, pcntnt=0;
    int I=0; 
OP's avatar
OP committed
151
    // for
OP's avatar
OP committed
152 153
    //----- 6 H-H cross-cor TimeFrequency maps 
    vector< TArray< complex<r_4> > > vtfm;
OP's avatar
OP committed
154 155 156 157
    //----- 6 H-H cross-cor TimeFrequency maps for the variances of real and imag parts 
    vector< TArray< r_4 > > vtfm_rp_sq;
    vector< TArray< r_4 > > vtfm_ip_sq;

OP's avatar
OP committed
158 159
    //----- 8 auto-corr TimeFrequency maps 
    vector< TArray< r_4 > > vtfmac;
OP's avatar
OP committed
160 161
    //----- 8 auto-corr TimeFrequency variance maps 
    vector< TArray< r_4 > > vtfmac_sq;
OP's avatar
OP committed
162 163 164 165 166 167
    
    //---- for the time-freqency map filling    
    sa_size_t TFMtmidx=0;
    sa_size_t tfmSX, tfmSY;

    while (fgok) {
OP's avatar
OP committed
168
      //reads next visimtx
OP's avatar
OP committed
169
      fgok=vreader.ReadNext(vismtx, cfdate, mttag);
OP's avatar
OP committed
170

OP's avatar
OP committed
171 172 173
      if (!fgok)  break;
      // Apply gain g(nu)
      p4g.applyGain(vismtx);
OP's avatar
OP committed
174

OP's avatar
OP committed
175 176
      if (cnt==0)  {    //resizing matrices for sum of auto-correlations and sum of 6 cross-correlations 
	acsum.SetSize(8, vismtx.NCols());
OP's avatar
OP committed
177
	acsum_sq.SetSize(8, vismtx.NCols());
OP's avatar
OP committed
178
	cxsum.SetSize(6, vismtx.NCols());
OP's avatar
OP committed
179 180 181
	cxsum_sq_rp.SetSize(6, vismtx.NCols());
	cxsum_sq_ip.SetSize(6, vismtx.NCols());

OP's avatar
OP committed
182 183 184 185 186
	tfmSX=(Imax-Imin)/Istep/deltaIavg;
	tfmSY=vismtx.NCols()/TFMfbin;
	//allocating 8 Auto-Corr time-frequency maps 
	cout<<"visiavg/Info: allocating 8 AutoCor Time-Frequency maps : Time->NX="<<tfmSX<<" x Freq->NY="<<tfmSY<<endl;
	for(int k=0; k<8; k++) vtfmac.push_back( TArray< r_4 >(tfmSX, tfmSY) ); 
OP's avatar
OP committed
187 188 189
	cout <<" and 8 for the the variance maps "<<endl;
	for(int k=0; k<8; k++) vtfmac_sq.push_back( TArray< r_4 >(tfmSX, tfmSY) ); 

OP's avatar
OP committed
190 191 192
	//allocating 6 Cross-Corr H-H time-frequency maps 
	cout<<"visiavg/Info: allocating H-H cross-cor Time-Frequency maps : Time->NX="<<tfmSX<<" x Freq->NY="<<tfmSY<<endl;
	for(int k=0; k<6; k++) vtfm.push_back( TArray< complex<r_4> >(tfmSX, tfmSY) );
OP's avatar
OP committed
193 194 195
	cout << "and the 2x6 for squares of rela & imag parts " << endl;
	for(int k=0; k<6; k++) vtfm_rp_sq.push_back( TArray< r_4 >(tfmSX, tfmSY) );
	for(int k=0; k<6; k++) vtfm_ip_sq.push_back( TArray< r_4 >(tfmSX, tfmSY) );
OP's avatar
OP committed
196 197 198 199 200 201 202 203 204 205

	// recupere le jour de depart @ 0h
	datestart = TimeStamp(cfdate.DaysPart(),0.);
 
      }
      if (I==0) {   // start filling a new time bin 
	dateobs=cfdate;
	if (prtlev>0) 
	  cout<<"visiavg/Info:  dateobs="<<dateobs<<" SecondsPart()="<<dateobs.SecondsPart()<<endl;
	acsum = complex<r_4>(0.,0.);
OP's avatar
OP committed
206
	acsum_sq = 0.;
OP's avatar
OP committed
207
	cxsum = complex<r_4>(0.,0.);
OP's avatar
OP committed
208 209
	cxsum_sq_rp = 0.;
	cxsum_sq_ip = 0.;
OP's avatar
OP committed
210 211 212 213
      }
      
      //   sum (integration) along the time axis 
      for(size_t k=0; k<KVAC.size(); k++)      acsum.Row(k) += vismtx.Row(KVAC[k]);     // Les auto-correlations 
OP's avatar
OP committed
214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230
      for(size_t k=0; k<KVAC.size(); k++)      {
	TVector<r_4> tmp = real(vismtx.Row(KVAC[k]));
	acsum_sq.Row(k) += tmp.MulElt(tmp,tmp) ;
      }

      for(size_t k=0; k<KVCXHH.size(); k++){
	cxsum.Row(k) += vismtx.Row(KVCXHH[k]);   // les cross-correlations 

	TVector<r_4> tmp = real(vismtx.Row(KVCXHH[k]));
	cxsum_sq_rp.Row(k) +=  tmp.MulElt(tmp,tmp) ;

	tmp = imag(vismtx.Row(KVCXHH[k]));
	cxsum_sq_ip.Row(k) +=  tmp.MulElt(tmp,tmp) ;

      }


OP's avatar
OP committed
231 232 233 234 235 236
      I++;    // incrementing DeltaTime counter 
      
      if (I==deltaIavg) {
	//---- On s'occupe d'abord des autocorrelations P1 ... P8 
	for(size_t k=0; k<KVAC.size(); k++) {  // Loop over the 8 auto-correlations 
	  TVector<r_4> vac = real(acsum.Row(k));
OP's avatar
OP committed
237
	  TVector<r_4> vacsq = acsum_sq.Row(k);
OP's avatar
OP committed
238 239
	  if (TFMtmidx<tfmSX) {  // we check that our time index did not go beyond the allocated array size (might not be necessary)
	    TArray< r_4 > & tfmap = vtfmac[k];
OP's avatar
OP committed
240
	    TArray< r_4 > & tfmap_sq = vtfmac_sq[k];
OP's avatar
OP committed
241 242
	    for(sa_size_t jy=0; jy<tfmSY; jy++) {  // frequency binning 
	      tfmap(TFMtmidx, jy) = vac( Range(jy*TFMfbin, (jy+1)*TFMfbin-1) ).Sum();
OP's avatar
OP committed
243 244
	      tfmap_sq(TFMtmidx, jy) = vacsq( Range(jy*TFMfbin, (jy+1)*TFMfbin-1) ).Sum();

OP's avatar
OP committed
245
	    } 
OP's avatar
OP committed
246 247
	  }
  
OP's avatar
OP committed
248
	}  //----- end of loop over the 8 AutoCor
OP's avatar
OP committed
249
	//---- On s'occupe des 6 cross-correlations  1H-2H ... 3H-4H 
OP's avatar
OP committed
250 251
 	for(size_t k=0; k<KVCXHH.size(); k++)   {   // loop over the 6 Xcor 	  
	  TVector< complex<r_4> > vcx = cxsum.Row(k);
OP's avatar
OP committed
252 253
	  TVector<r_4>  vcxprsq = cxsum_sq_rp.Row(k);
	  TVector<r_4>  vcxpisq = cxsum_sq_ip.Row(k);
OP's avatar
OP committed
254 255
	  if (TFMtmidx<tfmSX) {  // we check that our time index did not go beyond the allocated array size (might not be necessary) 
	    TArray< complex<r_4> > & tfmap = vtfm[k];
OP's avatar
OP committed
256 257
	    TArray< r_4 > & tfmapsqpr = vtfm_rp_sq[k];
	    TArray< r_4 > & tfmapsqpi = vtfm_ip_sq[k];
OP's avatar
OP committed
258 259
	    for(sa_size_t jy=0; jy<tfmSY; jy++) {
	      tfmap(TFMtmidx, jy) = vcx( Range(jy*TFMfbin, (jy+1)*TFMfbin-1) ).Sum();
OP's avatar
OP committed
260 261
	      tfmapsqpr(TFMtmidx, jy) = vcxprsq( Range(jy*TFMfbin, (jy+1)*TFMfbin-1) ).Sum();
	      tfmapsqpi(TFMtmidx, jy) = vcxpisq( Range(jy*TFMfbin, (jy+1)*TFMfbin-1) ).Sum();
OP's avatar
OP committed
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282
	    } 
	  }  
	}  //----- end of loop over the 6 Xcor 
	timevec(TFMtmidx) = cfdate.TimeDifferenceSeconds(cfdate,datestart);
	TFMtmidx++;
	//  ... done 
	I=0;  cntnt++;
      }
      cnt++;
      if ((cnt>0)&&(cntnt%10==0)&&(cntnt>pcntnt)) {
	cout<<"visiavg/Info: TFM-Map fill cnt="<<cntnt<<" VisMtxCount="<<cnt<<" /Max="<<Imax<<" DateObs="<<dateobs<<endl;
	pcntnt=cntnt;
      }
    }
    cout<<"visiavg/Info: count="<<cnt<<" visimtx read "<<endl;

    // --- Sauvegarde cartes temps-frequence 
    POutPersist potfm(outfile);
    // --- renormalizing and saving AutoCorr time-frequency maps 
    cout<<"  visiavg/Info: Saving 8 AutoCorr time-frequency maps to PPF file "<<outfile<<endl;
    const char* tfm_names[8]={"TFM_1H", "TFM_2H", "TFM_3H", "TFM_4H", "TFM_1V", "TFM_2V", "TFM_3V", "TFM_4V"};
OP's avatar
OP committed
283
    const char* tfmsq_names[8]={"VARTFM_1H", "VARTFM_2H", "VARTFM_3H", "VARTFM_4H", "VARTFM_1V", "VARTFM_2V", "VARTFM_3V", "VARTFM_4V"};
OP's avatar
OP committed
284 285 286
    for(int k=0; k<8; k++)  {   // loop over the 8 AutoCorr 
      TArray< r_4 > & tfmap = vtfmac[k];
      tfmap *= (r_4)(1./((double)deltaIavg*(double)TFMfbin));
OP's avatar
OP committed
287 288 289
      TArray< r_4 > & tfmapsq = vtfmac_sq[k];
      tfmapsq *= (r_4)(1./((double)deltaIavg*(double)TFMfbin));
      tfmapsq = tfmapsq - tfmap.MulElt(tfmap,tfmap) ; 
OP's avatar
OP committed
290
      potfm << PPFNameTag(tfm_names[k]) << tfmap;
OP's avatar
OP committed
291 292
      potfm << PPFNameTag(tfmsq_names[k]) << tfmapsq;

OP's avatar
OP committed
293 294 295
    }
    // --- renormalizing and saving H-H Cross-Corr time-frequency maps 
    cout<<"  visiavg/Info: Saving 6 H-H cross-corr time-frequency maps to PPF file "<<outfile<<endl;
OP's avatar
OP committed
296

OP's avatar
OP committed
297
    const char* tfmCC_names[6]={"TFM_1H2H", "TFM_1H3H", "TFM_1H4H", "TFM_2H3H", "TFM_2H4H", "TFM_3H4H"};
OP's avatar
OP committed
298 299 300
    const char* vrtfmCC_names[6]={"RVARTFM_1H2H", "RVARTFM_1H3H", "RVARTFM_1H4H", "RVARTFM_2H3H", "RVARTFM_2H4H", "RVARTFM_3H4H"};
    const char* vitfmCC_names[6]={"IVARTFM_1H2H", "IVARTFM_1H3H", "IVARTFM_1H4H", "IVARTFM_2H3H", "IVARTFM_2H4H", "IVARTFM_3H4H"};

OP's avatar
OP committed
301 302
    for(int k=0; k<6; k++)  {   // loop over the 6 Xcor 
      TArray< complex<r_4> > & tfmap = vtfm[k];
OP's avatar
OP committed
303 304 305

      TArray< r_4 > & tfmap_sqpr = vtfm_rp_sq[k];
      TArray< r_4 > & tfmap_sqpi = vtfm_ip_sq[k];
OP's avatar
OP committed
306
      tfmap *= complex<r_4>((r_4)(1./((double)deltaIavg*(double)TFMfbin)), 0.);
OP's avatar
OP committed
307 308 309 310 311 312 313 314 315
      TArray< r_4 >  tfmapr = real(tfmap);
      TArray< r_4 >  tfmapi = imag(tfmap);

      tfmap_sqpr *= ((r_4)(1./((double)deltaIavg*(double)TFMfbin)));
      tfmap_sqpi *= ((r_4)(1./((double)deltaIavg*(double)TFMfbin)));
      tfmapr = tfmapr.MulElt(tfmapr,tfmapr) ;
      tfmap_sqpr -= tfmapr ;
      tfmapi = tfmapi.MulElt(tfmapi,tfmapi) ;
      tfmap_sqpi -= tfmapi;
OP's avatar
OP committed
316
      potfm << PPFNameTag(tfmCC_names[k]) << tfmap;
OP's avatar
OP committed
317 318
      potfm << PPFNameTag(vrtfmCC_names[k]) << tfmap_sqpr;
      potfm << PPFNameTag(vitfmCC_names[k]) << tfmap_sqpi;
OP's avatar
OP committed
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344
    }
    potfm << PPFNameTag("TimeVec") << timevec ;
    // --- FIN sauvegarde cartes temps-frequence 
    //    resu.Update();
    cout << resu;   // Update est fait lors du print
  }
  catch (PException& exc) {
    cerr << " visiavg.cc catched PException " << exc.Msg() << endl;
    rc = 77;
  }  
  catch (std::exception& sex) {
    cerr << "\n visiavg.cc std::exception :" 
         << (string)typeid(sex).name() << "\n msg= " 
         << sex.what() << endl;
    rc = 78;
  }
  catch (...) {
    cerr << " visiavg.cc catched unknown (...) exception  " << endl; 
    rc = 79; 
  } 

  cout << ">>>> visiavg.cc ------- END ----------- RC=" << rc << endl;
  return rc;

}