| 1 | #include "sopnamsp.h" | 
|---|
| 2 | #include "machdefs.h" | 
|---|
| 3 | #include <iostream> | 
|---|
| 4 | #include <math.h> | 
|---|
| 5 | #include <complex> | 
|---|
| 6 | #include "sphericaltransformserver.h" | 
|---|
| 7 | #include "tvector.h" | 
|---|
| 8 | #include "nbrandom.h" | 
|---|
| 9 | #include "nbmath.h" | 
|---|
| 10 | #include "timing.h" | 
|---|
| 11 | //#include "spherehealpix.h" | 
|---|
| 12 |  | 
|---|
| 13 |  | 
|---|
| 14 | /*! \class SOPHYA::SphericalTransformServer | 
|---|
| 15 |  | 
|---|
| 16 | Class for performing analysis and synthesis of sky maps using spin-0 or spin-2 spherical harmonics. | 
|---|
| 17 |  | 
|---|
| 18 | Maps must be SOPHYA SphericalMaps (SphereGorski or SphereThetaPhi). | 
|---|
| 19 |  | 
|---|
| 20 | Temperature and polarization (Stokes parameters) can be developped on spherical harmonics : | 
|---|
| 21 | \f[ | 
|---|
| 22 | \frac{\Delta T}{T}(\hat{n})=\sum_{lm}a_{lm}^TY_l^m(\hat{n}) | 
|---|
| 23 | \f] | 
|---|
| 24 | \f[ | 
|---|
| 25 | Q(\hat{n})=\frac{1}{\sqrt{2}}\sum_{lm}N_l\left(a_{lm}^EW_{lm}(\hat{n})+a_{lm}^BX_{lm}(\hat{n})\right) | 
|---|
| 26 | \f] | 
|---|
| 27 | \f[ | 
|---|
| 28 | U(\hat{n})=-\frac{1}{\sqrt{2}}\sum_{lm}N_l\left(a_{lm}^EX_{lm}(\hat{n})-a_{lm}^BW_{lm}(\hat{n})\right) | 
|---|
| 29 | \f] | 
|---|
| 30 | \f[ | 
|---|
| 31 | \left(Q \pm iU\right)(\hat{n})=\sum_{lm}a_{\pm 2lm}\, _{\pm 2}Y_l^m(\hat{n}) | 
|---|
| 32 | \f] | 
|---|
| 33 |  | 
|---|
| 34 | \f[ | 
|---|
| 35 | Y_l^m(\hat{n})=\lambda_l^m(\theta)e^{im\phi} | 
|---|
| 36 | \f] | 
|---|
| 37 | \f[ | 
|---|
| 38 | _{\pm}Y_l^m(\hat{n})=_{\pm}\lambda_l^m(\theta)e^{im\phi} | 
|---|
| 39 | \f] | 
|---|
| 40 | \f[ | 
|---|
| 41 | W_{lm}(\hat{n})=\frac{1}{N_l}\,_{w}\lambda_l^m(\theta)e^{im\phi} | 
|---|
| 42 | \f] | 
|---|
| 43 | \f[ | 
|---|
| 44 | X_{lm}(\hat{n})=\frac{-i}{N_l}\,_{x}\lambda_l^m(\theta)e^{im\phi} | 
|---|
| 45 | \f] | 
|---|
| 46 |  | 
|---|
| 47 | (see LambdaLMBuilder, LambdaPMBuilder, LambdaWXBuilder classes) | 
|---|
| 48 |  | 
|---|
| 49 | power spectra : | 
|---|
| 50 |  | 
|---|
| 51 | \f[ | 
|---|
| 52 | C_l^T=\frac{1}{2l+1}\sum_{m=0}^{+ \infty }\left|a_{lm}^T\right|^2=\langle\left|a_{lm}^T\right|^2\rangle | 
|---|
| 53 | \f] | 
|---|
| 54 | \f[ | 
|---|
| 55 | C_l^E=\frac{1}{2l+1}\sum_{m=0}^{+\infty}\left|a_{lm}^E\right|^2=\langle\left|a_{lm}^E\right|^2\rangle | 
|---|
| 56 | \f] | 
|---|
| 57 | \f[ | 
|---|
| 58 | C_l^B=\frac{1}{2l+1}\sum_{m=0}^{+\infty}\left|a_{lm}^B\right|^2=\langle\left|a_{lm}^B\right|^2\rangle | 
|---|
| 59 | \f] | 
|---|
| 60 |  | 
|---|
| 61 | \arg | 
|---|
| 62 | \b Synthesis : Get temperature and polarization maps  from \f$a_{lm}\f$ coefficients or from power spectra, (methods GenerateFrom...). | 
|---|
| 63 |  | 
|---|
| 64 | \b Temperature: | 
|---|
| 65 | \f[ | 
|---|
| 66 | \frac{\Delta T}{T}(\hat{n})=\sum_{lm}a_{lm}^TY_l^m(\hat{n}) = \sum_{-\infty}^{+\infty}b_m(\theta)e^{im\phi} | 
|---|
| 67 | \f] | 
|---|
| 68 |  | 
|---|
| 69 | with | 
|---|
| 70 | \f[ | 
|---|
| 71 | b_m(\theta)=\sum_{l=\left|m\right|}^{+\infty}a_{lm}^T\lambda_l^m(\theta) | 
|---|
| 72 | \f] | 
|---|
| 73 |  | 
|---|
| 74 | \b Polarisation | 
|---|
| 75 | \f[ | 
|---|
| 76 | Q \pm iU = \sum_{-\infty}^{+\infty}b_m^{\pm}(\theta)e^{im\phi} | 
|---|
| 77 | \f] | 
|---|
| 78 |  | 
|---|
| 79 | where : | 
|---|
| 80 | \f[ | 
|---|
| 81 | b_m^{\pm}(\theta) = \sum_{l=\left|m\right|}^{+\infty}a_{\pm 2lm}\,_{\pm}\lambda_l^m(\theta) | 
|---|
| 82 | \f] | 
|---|
| 83 |  | 
|---|
| 84 | or : | 
|---|
| 85 | \f[ | 
|---|
| 86 | Q  = \sum_{-\infty}^{+\infty}b_m^{Q}(\theta)e^{im\phi} | 
|---|
| 87 | \f] | 
|---|
| 88 | \f[ | 
|---|
| 89 | U  = \sum_{-\infty}^{+\infty}b_m^{U}(\theta)e^{im\phi} | 
|---|
| 90 | \f] | 
|---|
| 91 |  | 
|---|
| 92 | where: | 
|---|
| 93 | \f[ | 
|---|
| 94 | b_m^{Q}(\theta) = \frac{1}{\sqrt{2}}\sum_{l=\left|m\right|}^{+\infty}\left(a_{lm}^E\,_{w}\lambda_l^m(\theta)-ia_{lm}^B\,_{x}\lambda_l^m(\theta)\right) | 
|---|
| 95 | \f] | 
|---|
| 96 | \f[ | 
|---|
| 97 | b_m^{U}(\theta) = \frac{1}{\sqrt{2}}\sum_{l=\left|m\right|}^{+\infty}\left(ia_{lm}^E\,_{x}\lambda_l^m(\theta)+a_{lm}^B\,_{w}\lambda_l^m(\theta)\right) | 
|---|
| 98 | \f] | 
|---|
| 99 |  | 
|---|
| 100 | Since the pixelization provides "slices" with constant \f$\theta\f$ and \f$\phi\f$ equally distributed  on \f$2\pi\f$  \f$\frac{\Delta T}{T}\f$, \f$Q\f$,\f$U\f$  can be computed by FFT. | 
|---|
| 101 |  | 
|---|
| 102 |  | 
|---|
| 103 | \arg | 
|---|
| 104 | \b Analysis :  Get \f$a_{lm}\f$ coefficients or  power spectra from temperature and polarization maps   (methods DecomposeTo...). | 
|---|
| 105 |  | 
|---|
| 106 | \b Temperature: | 
|---|
| 107 | \f[ | 
|---|
| 108 | a_{lm}^T=\int\frac{\Delta T}{T}(\hat{n})Y_l^{m*}(\hat{n})d\hat{n} | 
|---|
| 109 | \f] | 
|---|
| 110 |  | 
|---|
| 111 | approximated as : | 
|---|
| 112 | \f[ | 
|---|
| 113 | a_{lm}^T=\sum_{\theta_k}\omega_kC_m(\theta_k)\lambda_l^m(\theta_k) | 
|---|
| 114 | \f] | 
|---|
| 115 | where : | 
|---|
| 116 | \f[ | 
|---|
| 117 | C_m (\theta _k)=\sum_{\phi _{k\prime}}\frac{\Delta T}{T}(\theta _k,\phi_{k\prime})e^{-im\phi _{k\prime}} | 
|---|
| 118 | \f] | 
|---|
| 119 | Since the pixelization provides "slices" with constant \f$\theta\f$ and \f$\phi\f$ equally distributed  on \f$2\pi\f$ (\f$\omega_k\f$ is the solid angle of each pixel of the slice \f$\theta_k\f$) \f$C_m\f$ can be computed by FFT. | 
|---|
| 120 |  | 
|---|
| 121 | \b polarisation: | 
|---|
| 122 |  | 
|---|
| 123 | \f[ | 
|---|
| 124 | a_{\pm 2lm}=\sum_{\theta_k}\omega_kC_m^{\pm}(\theta_k)\,_{\pm}\lambda_l^m(\theta_k) | 
|---|
| 125 | \f] | 
|---|
| 126 | where : | 
|---|
| 127 | \f[ | 
|---|
| 128 | C_m^{\pm} (\theta _k)=\sum_{\phi _{k\prime}}\left(Q \pm iU\right)(\theta _k,\phi_{k\prime})e^{-im\phi _{k\prime}} | 
|---|
| 129 | \f] | 
|---|
| 130 | or : | 
|---|
| 131 |  | 
|---|
| 132 | \f[ | 
|---|
| 133 | a_{lm}^E=\frac{1}{\sqrt{2}}\sum_{\theta_k}\omega_k\left(C_m^{Q}(\theta_k)\,_{w}\lambda_l^m(\theta_k)-iC_m^{U}(\theta_k)\,_{x}\lambda_l^m(\theta_k)\right) | 
|---|
| 134 | \f] | 
|---|
| 135 | \f[ | 
|---|
| 136 | a_{lm}^B=\frac{1}{\sqrt{2}}\sum_{\theta_k}\omega_k\left(iC_m^{Q}(\theta_k)\,_{x}\lambda_l^m(\theta_k)+C_m^{U}(\theta_k)\,_{w}\lambda_l^m(\theta_k)\right) | 
|---|
| 137 | \f] | 
|---|
| 138 |  | 
|---|
| 139 | where : | 
|---|
| 140 | \f[ | 
|---|
| 141 | C_m^{Q} (\theta _k)=\sum_{\phi _{k\prime}}Q(\theta _k,\phi_{k\prime})e^{-im\phi _{k\prime}} | 
|---|
| 142 | \f] | 
|---|
| 143 | \f[ | 
|---|
| 144 | C_m^{U} (\theta _k)=\sum_{\phi _{k\prime}}U(\theta _k,\phi_{k\prime})e^{-im\phi _{k\prime}} | 
|---|
| 145 | \f] | 
|---|
| 146 |  | 
|---|
| 147 | */ | 
|---|
| 148 |  | 
|---|
| 149 | /*! \fn void SOPHYA::SphericalTransformServer::GenerateFromAlm( SphericalMap<T>& map, int_4 pixelSizeIndex, const Alm<T>& alm) const | 
|---|
| 150 |  | 
|---|
| 151 | synthesis of a temperature  map from  Alm coefficients | 
|---|
| 152 | */ | 
|---|
| 153 | template<class T> | 
|---|
| 154 | void SphericalTransformServer<T>::GenerateFromAlm( SphericalMap<T>& map, int_4 pixelSizeIndex, const Alm<T>& alm) const | 
|---|
| 155 | { | 
|---|
| 156 | /*======================================================================= | 
|---|
| 157 | computes a map from its alm for the HEALPIX pixelisation | 
|---|
| 158 | map(theta,phi) = sum_l_m a_lm Y_lm(theta,phi) | 
|---|
| 159 | = sum_m {e^(i*m*phi) sum_l a_lm*lambda_lm(theta)} | 
|---|
| 160 |  | 
|---|
| 161 | where Y_lm(theta,phi) = lambda(theta) * e^(i*m*phi) | 
|---|
| 162 |  | 
|---|
| 163 | * the recurrence of Ylm is the standard one (cf Num Rec) | 
|---|
| 164 | * the sum over m is done by FFT | 
|---|
| 165 |  | 
|---|
| 166 | =======================================================================*/ | 
|---|
| 167 | int_4 nlmax=alm.Lmax(); | 
|---|
| 168 | int_4 nmmax=nlmax; | 
|---|
| 169 | int_4 nsmax=0; | 
|---|
| 170 | // le Resize est suppose mettre a zero | 
|---|
| 171 | map.Resize(pixelSizeIndex); | 
|---|
| 172 | string sphere_type=map.TypeOfMap(); | 
|---|
| 173 | int premiereTranche = 0; | 
|---|
| 174 | int derniereTranche = map.NbThetaSlices()-1; | 
|---|
| 175 | if (sphere_type.substr(0,4) == "RING") | 
|---|
| 176 | { | 
|---|
| 177 | nsmax=map.SizeIndex(); | 
|---|
| 178 | } | 
|---|
| 179 | else | 
|---|
| 180 | { | 
|---|
| 181 | // pour une sphere Gorski le nombre de pixels est 12*nsmax**2 | 
|---|
| 182 | // on calcule une quantite equivalente a nsmax pour la sphere-theta-phi | 
|---|
| 183 | // en vue de l'application du critere Healpix : nlmax<=3*nsmax-1 | 
|---|
| 184 | // c'est approximatif ; a raffiner. | 
|---|
| 185 | if (sphere_type.substr(0,6) == "TETAFI") | 
|---|
| 186 | { | 
|---|
| 187 | nsmax=(int_4)sqrt(map.NbPixels()/12.); | 
|---|
| 188 | premiereTranche++; | 
|---|
| 189 | derniereTranche--; | 
|---|
| 190 | } | 
|---|
| 191 | else | 
|---|
| 192 | { | 
|---|
| 193 | cout << " unknown type of sphere : " << sphere_type << endl; | 
|---|
| 194 | throw IOExc(" unknown type of sphere: " + (string)sphere_type ); | 
|---|
| 195 | } | 
|---|
| 196 | //  cout << "GenerateFromAlm: the sphere is of type : " << sphere_type << endl; | 
|---|
| 197 | //  cout << "GenerateFromAlm: size index (nside) of the sphere= " << nsmax << endl; | 
|---|
| 198 | //  cout << "GenerateFromAlm: nlmax (from Alm) = " << nlmax << endl; | 
|---|
| 199 | //  if (nlmax>3*nsmax-1) | 
|---|
| 200 | //  { | 
|---|
| 201 | //     cout << "GenerateFromAlm: nlmax should be <= 3*nside-1" << endl; | 
|---|
| 202 | //   if (strncmp(sphere_type,"TETAFI",6) == 0) | 
|---|
| 203 | //    { | 
|---|
| 204 | //      cout << "GenerateFromAlm: nlmax should be <= 3*nside-1" << endl; | 
|---|
| 205 | //      cout << " (for this criterium, nsmax is computed as sqrt(nbPixels/12))" << endl; | 
|---|
| 206 | //    } | 
|---|
| 207 | //} | 
|---|
| 208 | } | 
|---|
| 209 | Bm<complex<T> > b_m_theta(nmmax); | 
|---|
| 210 |  | 
|---|
| 211 | //  map.Resize(nsmax); | 
|---|
| 212 |  | 
|---|
| 213 |  | 
|---|
| 214 | // pour chaque tranche en theta | 
|---|
| 215 | for (int_4 ith = premiereTranche; ith <= derniereTranche;ith++) | 
|---|
| 216 | { | 
|---|
| 217 | int_4 nph; | 
|---|
| 218 | r_8 phi0; | 
|---|
| 219 | r_8 theta; | 
|---|
| 220 | TVector<int_4> pixNumber; | 
|---|
| 221 | TVector<T> datan; | 
|---|
| 222 |  | 
|---|
| 223 | map.GetThetaSlice(ith,theta,phi0, pixNumber,datan); | 
|---|
| 224 | nph = pixNumber.NElts(); | 
|---|
| 225 |  | 
|---|
| 226 | //       ----------------------------------------------------- | 
|---|
| 227 | //              for each theta, and each m, computes | 
|---|
| 228 | //              b(m,theta) = sum_over_l>m (lambda_l_m(theta) * a_l_m) | 
|---|
| 229 | //              ------------------------------------------------------ | 
|---|
| 230 | LambdaLMBuilder lb(theta,nlmax,nmmax); | 
|---|
| 231 | //  somme sur m de 0 a l'infini | 
|---|
| 232 | int m; | 
|---|
| 233 | for (m = 0; m <= nmmax; m++) | 
|---|
| 234 | { | 
|---|
| 235 | b_m_theta(m) = (T)( lb.lamlm(m,m) ) * alm(m,m); | 
|---|
| 236 | for (int l = m+1; l<= nlmax; l++) | 
|---|
| 237 | { | 
|---|
| 238 | b_m_theta(m) += (T)( lb.lamlm(l,m) ) * alm(l,m); | 
|---|
| 239 | } | 
|---|
| 240 | } | 
|---|
| 241 | //        obtains the negative m of b(m,theta) (= complex conjugate) | 
|---|
| 242 |  | 
|---|
| 243 | for (m=1;m<=nmmax;m++) | 
|---|
| 244 | { | 
|---|
| 245 | b_m_theta(-m) = conj(b_m_theta(m)); | 
|---|
| 246 | } | 
|---|
| 247 | // --------------------------------------------------------------- | 
|---|
| 248 | //    sum_m  b(m,theta)*exp(i*m*phi)   -> f(phi,theta) | 
|---|
| 249 | // ---------------------------------------------------------------*/ | 
|---|
| 250 |  | 
|---|
| 251 |  | 
|---|
| 252 | if (sphere_type.substr(0,4) == "RING") | 
|---|
| 253 | { | 
|---|
| 254 | TVector<T> Temp = RfourierSynthesisFromB(b_m_theta,nph,phi0); | 
|---|
| 255 | for (int i=0;i< nph;i++) map(pixNumber(i))=Temp(i); | 
|---|
| 256 | } | 
|---|
| 257 | else | 
|---|
| 258 | // pour des pixelisations quelconques (autres que HEALPix | 
|---|
| 259 | //  nph n'est pas toujours pair | 
|---|
| 260 | // ca fait des problemes pour les transformees de Fourier | 
|---|
| 261 | // car le server de TF ajuste la longueur du vecteur reel | 
|---|
| 262 | // en sortie de TF, bref, la securite veut qu'on prenne une | 
|---|
| 263 | // TF complexe | 
|---|
| 264 | { | 
|---|
| 265 | TVector<complex<T> > Temp = fourierSynthesisFromB(b_m_theta,nph,phi0); | 
|---|
| 266 | for (int i=0;i< nph;i++) map(pixNumber(i))=Temp(i).real(); | 
|---|
| 267 | } | 
|---|
| 268 | } | 
|---|
| 269 | } | 
|---|
| 270 |  | 
|---|
| 271 |  | 
|---|
| 272 |  | 
|---|
| 273 | /*! \fn TVector< complex<T> >  SOPHYA::SphericalTransformServer::fourierSynthesisFromB(const Bm<complex<T> >& b_m,  int_4 nph, r_8 phi0) const | 
|---|
| 274 |  | 
|---|
| 275 | \return a vector with nph elements  which are  sums :\f$\sum_{m=-mmax}^{mmax}b_m(\theta)e^{im\varphi}\f$ for nph values of \f$\varphi\f$ regularly distributed in \f$[0,\pi]\f$ ( calculated by FFT) | 
|---|
| 276 |  | 
|---|
| 277 | The object b_m (\f$b_m\f$) of the class Bm is a special vector which index goes from -mmax to mmax. | 
|---|
| 278 | */ | 
|---|
| 279 | template<class T> | 
|---|
| 280 | TVector< complex<T> >  SphericalTransformServer<T>::fourierSynthesisFromB(const Bm<complex<T> >& b_m,  int_4 nph, r_8 phi0) const | 
|---|
| 281 | { | 
|---|
| 282 | /*======================================================================= | 
|---|
| 283 | dataout(j) = sum_m datain(m) * exp(i*m*phi(j)) | 
|---|
| 284 | with phi(j) = j*2pi/nph + kphi0*pi/nph and kphi0 =0 or 1 | 
|---|
| 285 |  | 
|---|
| 286 | as the set of frequencies {m} is larger than nph, | 
|---|
| 287 | we wrap frequencies within {0..nph-1} | 
|---|
| 288 | ie  m = k*nph + m' with m' in {0..nph-1} | 
|---|
| 289 | then | 
|---|
| 290 | noting bw(m') = exp(i*m'*phi0) | 
|---|
| 291 | * sum_k (datain(k*nph+m') exp(i*k*pi*kphi0)) | 
|---|
| 292 | with bw(nph-m') = CONJ(bw(m')) (if datain(-m) = CONJ(datain(m))) | 
|---|
| 293 | dataout(j) = sum_m' [ bw(m') exp (i*j*m'*2pi/nph) ] | 
|---|
| 294 | = Fourier Transform of bw | 
|---|
| 295 | is real | 
|---|
| 296 |  | 
|---|
| 297 | NB nph is not necessarily a power of 2 | 
|---|
| 298 |  | 
|---|
| 299 | =======================================================================*/ | 
|---|
| 300 | //********************************************************************** | 
|---|
| 301 | // pour une valeur de phi (indexee par j) la temperature est la transformee | 
|---|
| 302 | // de Fourier de bm (somme sur m de -nmax a +nmmax de bm*exp(i*m*phi)). | 
|---|
| 303 | // on demande nph (nombre de pixels sur la tranche) valeurs de transformees, pour nph valeurs de phi, regulierement reparties sur 2*pi. On a: | 
|---|
| 304 | //      DT/T(j) = sum_m b(m) * exp(i*m*phi(j)) | 
|---|
| 305 | // sommation de -infini a +infini, en fait limitee a -nmamx, +nmmax | 
|---|
| 306 | // On pose m=k*nph + m', avec m' compris entre 0 et nph-1. Alors : | 
|---|
| 307 | // DT/T(j) = somme_k somme_m'  b(k*nph + m')*exp(i*(k*nph + m')*phi(j)) | 
|---|
| 308 | // somme_k : de -infini a +infini | 
|---|
| 309 | // somme_m' : de 0 a nph-1 | 
|---|
| 310 | // On echange les sommations : | 
|---|
| 311 | // DT/T(j) = somme_m' (exp(i*m'*phi(j)) somme_k b(k*nph + m')*exp(i*(k*nph*phi(j)) | 
|---|
| 312 | // mais phi(j) est un multiple entier de 2*pi/nph, la seconde exponentielle | 
|---|
| 313 | // vaut 1. | 
|---|
| 314 | // Il reste a calculer les transformees de Fourier de somme_m' b(k*nph + m') | 
|---|
| 315 | // si phi0 n'est pas nul, il y a juste un decalage a faire. | 
|---|
| 316 | //********************************************************************** | 
|---|
| 317 |  | 
|---|
| 318 | TVector< complex<T> > bw(nph); | 
|---|
| 319 | TVector< complex<T> > dataout(nph); | 
|---|
| 320 | TVector< complex<T> > data(nph); | 
|---|
| 321 |  | 
|---|
| 322 |  | 
|---|
| 323 | for (int kk=0; kk<bw.NElts(); kk++) bw(kk)=(T)0.; | 
|---|
| 324 | int m; | 
|---|
| 325 | for (m=-b_m.Mmax();m<=-1;m++) | 
|---|
| 326 | { | 
|---|
| 327 | int maux=m; | 
|---|
| 328 | while (maux<0) maux+=nph; | 
|---|
| 329 | int iw=maux%nph; | 
|---|
| 330 | double aux=(m-iw)*phi0; | 
|---|
| 331 | bw(iw) += b_m(m) * complex<T>( (T)cos(aux),(T)sin(aux) )  ; | 
|---|
| 332 | } | 
|---|
| 333 | for (m=0;m<=b_m.Mmax();m++) | 
|---|
| 334 | { | 
|---|
| 335 | //      int iw=((m % nph) +nph) % nph; //between 0 and nph = m' | 
|---|
| 336 | int iw=m%nph; | 
|---|
| 337 | double aux=(m-iw)*phi0; | 
|---|
| 338 | bw(iw)+=b_m(m) * complex<T>( (T)cos(aux),(T)sin(aux) ); | 
|---|
| 339 | } | 
|---|
| 340 |  | 
|---|
| 341 | //     applies the shift in position <-> phase factor in Fourier space | 
|---|
| 342 | for (int mprime=0; mprime < nph; mprime++) | 
|---|
| 343 | { | 
|---|
| 344 | complex<double> aux(cos(mprime*phi0),sin(mprime*phi0)); | 
|---|
| 345 | data(mprime)=bw(mprime)* | 
|---|
| 346 | (complex<T>)(complex<double>(cos(mprime*phi0),sin(mprime*phi0))); | 
|---|
| 347 | } | 
|---|
| 348 |  | 
|---|
| 349 | //sortie.ReSize(nph); | 
|---|
| 350 | TVector< complex<T> > sortie(nph); | 
|---|
| 351 |  | 
|---|
| 352 | fftIntfPtr_-> FFTBackward(data, sortie); | 
|---|
| 353 |  | 
|---|
| 354 | return sortie; | 
|---|
| 355 | } | 
|---|
| 356 |  | 
|---|
| 357 | //******************************************** | 
|---|
| 358 | /*! \fn TVector<T>  SOPHYA::SphericalTransformServer::RfourierSynthesisFromB(const Bm<complex<T> >& b_m,  int_4 nph, r_8 phi0) const | 
|---|
| 359 |  | 
|---|
| 360 | same as fourierSynthesisFromB, but return a real vector, taking into account the fact that b(-m) is conjugate of b(m) */ | 
|---|
| 361 | template<class T> | 
|---|
| 362 | TVector<T>  SphericalTransformServer<T>::RfourierSynthesisFromB(const Bm<complex<T> >& b_m,  int_4 nph, r_8 phi0) const | 
|---|
| 363 | { | 
|---|
| 364 | /*======================================================================= | 
|---|
| 365 | dataout(j) = sum_m datain(m) * exp(i*m*phi(j)) | 
|---|
| 366 | with phi(j) = j*2pi/nph + kphi0*pi/nph and kphi0 =0 or 1 | 
|---|
| 367 |  | 
|---|
| 368 | as the set of frequencies {m} is larger than nph, | 
|---|
| 369 | we wrap frequencies within {0..nph-1} | 
|---|
| 370 | ie  m = k*nph + m' with m' in {0..nph-1} | 
|---|
| 371 | then | 
|---|
| 372 | noting bw(m') = exp(i*m'*phi0) | 
|---|
| 373 | * sum_k (datain(k*nph+m') exp(i*k*pi*kphi0)) | 
|---|
| 374 | with bw(nph-m') = CONJ(bw(m')) (if datain(-m) = CONJ(datain(m))) | 
|---|
| 375 | dataout(j) = sum_m' [ bw(m') exp (i*j*m'*2pi/nph) ] | 
|---|
| 376 | = Fourier Transform of bw | 
|---|
| 377 | is real | 
|---|
| 378 |  | 
|---|
| 379 | NB nph is not necessarily a power of 2 | 
|---|
| 380 |  | 
|---|
| 381 | =======================================================================*/ | 
|---|
| 382 | //********************************************************************** | 
|---|
| 383 | // pour une valeur de phi (indexee par j) la temperature est la transformee | 
|---|
| 384 | // de Fourier de bm (somme sur m de -nmax a +nmmax de bm*exp(i*m*phi)). | 
|---|
| 385 | // on demande nph (nombre de pixels sur la tranche) valeurs de transformees, pour nph valeurs de phi, regulierement reparties sur 2*pi. On a: | 
|---|
| 386 | //      DT/T(j) = sum_m b(m) * exp(i*m*phi(j)) | 
|---|
| 387 | // sommation de -infini a +infini, en fait limitee a -nmamx, +nmmax | 
|---|
| 388 | // On pose m=k*nph + m', avec m' compris entre 0 et nph-1. Alors : | 
|---|
| 389 | // DT/T(j) = somme_k somme_m'  b(k*nph + m')*exp(i*(k*nph + m')*phi(j)) | 
|---|
| 390 | // somme_k : de -infini a +infini | 
|---|
| 391 | // somme_m' : de 0 a nph-1 | 
|---|
| 392 | // On echange les sommations : | 
|---|
| 393 | // DT/T(j) = somme_m' (exp(i*m'*phi(j)) somme_k b(k*nph + m')*exp(i*(k*nph*phi(j)) | 
|---|
| 394 | // mais phi(j) est un multiple entier de 2*pi/nph, la seconde exponentielle | 
|---|
| 395 | // vaut 1. | 
|---|
| 396 | // Il reste a calculer les transformees de Fourier de somme_k b(k*nph + m') | 
|---|
| 397 | // si phi0 n'est pas nul, il y a juste un decalage a faire. | 
|---|
| 398 | //********************************************************************** | 
|---|
| 399 | TVector< complex<T> > bw(nph); | 
|---|
| 400 | TVector< complex<T> > dataout(nph); | 
|---|
| 401 | TVector< complex<T> > data(nph/2+1); | 
|---|
| 402 |  | 
|---|
| 403 |  | 
|---|
| 404 | for (int kk=0; kk<bw.NElts(); kk++) bw(kk)=(T)0.; | 
|---|
| 405 | int m; | 
|---|
| 406 | for (m=-b_m.Mmax();m<=-1;m++) | 
|---|
| 407 | { | 
|---|
| 408 | int maux=m; | 
|---|
| 409 | while (maux<0) maux+=nph; | 
|---|
| 410 | int iw=maux%nph; | 
|---|
| 411 | double aux=(m-iw)*phi0; | 
|---|
| 412 | bw(iw) += b_m(m) * complex<T>( (T)cos(aux),(T)sin(aux) )  ; | 
|---|
| 413 | } | 
|---|
| 414 | for (m=0;m<=b_m.Mmax();m++) | 
|---|
| 415 | { | 
|---|
| 416 | //      int iw=((m % nph) +nph) % nph; //between 0 and nph = m' | 
|---|
| 417 | int iw=m%nph; | 
|---|
| 418 | double aux=(m-iw)*phi0; | 
|---|
| 419 | bw(iw)+=b_m(m) * complex<T>( (T)cos(aux),(T)sin(aux) ); | 
|---|
| 420 | } | 
|---|
| 421 |  | 
|---|
| 422 | //     applies the shift in position <-> phase factor in Fourier space | 
|---|
| 423 | // cout << " TF : nph= " << nph << " vec. entree " << data.Size() << endl; | 
|---|
| 424 | for (int mprime=0; mprime <= nph/2; mprime++) | 
|---|
| 425 | { | 
|---|
| 426 | complex<double> aux(cos(mprime*phi0),sin(mprime*phi0)); | 
|---|
| 427 | data(mprime)=bw(mprime)* | 
|---|
| 428 | (complex<T>)(complex<double>(cos(mprime*phi0),sin(mprime*phi0))); | 
|---|
| 429 | } | 
|---|
| 430 |  | 
|---|
| 431 | TVector<T> sortie; | 
|---|
| 432 | fftIntfPtr_-> FFTBackward(data, sortie); | 
|---|
| 433 |  | 
|---|
| 434 | return sortie; | 
|---|
| 435 | } | 
|---|
| 436 | //******************************************* | 
|---|
| 437 |  | 
|---|
| 438 | /*! \fn  Alm<T> SOPHYA::SphericalTransformServer::DecomposeToAlm(const SphericalMap<T>& map, int_4 nlmax, r_8 cos_theta_cut) const | 
|---|
| 439 |  | 
|---|
| 440 | \return the Alm coefficients from analysis of a temperature map. | 
|---|
| 441 |  | 
|---|
| 442 | \param<nlmax> : maximum value of the l index | 
|---|
| 443 |  | 
|---|
| 444 | \param<cos_theta_cut> : cosinus of the symmetric cut EULER angle theta : cos_theta_cut=0 means no cut ; cos_theta_cut=1 all the sphere is cut. | 
|---|
| 445 |  | 
|---|
| 446 | */ | 
|---|
| 447 | template<class T> | 
|---|
| 448 | void SphericalTransformServer<T>::DecomposeToAlm(const SphericalMap<T>& map, Alm<T>& alm, int_4 nlmax, r_8 cos_theta_cut) const | 
|---|
| 449 | { | 
|---|
| 450 | DecomposeToAlm(const_cast< SphericalMap<T>& >(map), alm, nlmax, cos_theta_cut, 0); | 
|---|
| 451 | } | 
|---|
| 452 | //******************************************* | 
|---|
| 453 |  | 
|---|
| 454 | /*! \fn  Alm<T> SOPHYA::SphericalTransformServer::DecomposeToAlm(const SphericalMap<T>& map, int_4 nlmax, r_8 cos_theta_cut, int iterationOrder) const | 
|---|
| 455 |  | 
|---|
| 456 | \return the Alm coefficients from analysis of a temperature map. THE MAP CAN BE MODIFIED (if iterationOrder >0) | 
|---|
| 457 |  | 
|---|
| 458 | \param<nlmax> : maximum value of the l index | 
|---|
| 459 |  | 
|---|
| 460 | \param<cos_theta_cut> : cosinus of the symmetric cut EULER angle theta : cos_theta_cut=0 means no cut ; cos_theta_cut=1 all the sphere is cut. | 
|---|
| 461 |  | 
|---|
| 462 | \param<iterationOrder> : 1,2,3,4.... order of an iterative analysis. (Default : 0 -> standard analysis). If iterationOrder is not null, the method works with SphereHEALPix but NOT WITH SphereThetaPhi maps !  */ | 
|---|
| 463 | template<class T> | 
|---|
| 464 | void SphericalTransformServer<T>::DecomposeToAlm(SphericalMap<T>& map, Alm<T>& alm, int_4 nlmax, r_8 cos_theta_cut, int iterationOrder) const | 
|---|
| 465 | { | 
|---|
| 466 | int_4  nmmax = nlmax; | 
|---|
| 467 | //  PrtTim("appel  carteVersAlm"); | 
|---|
| 468 | carteVersAlm(map, nlmax, cos_theta_cut, alm); | 
|---|
| 469 | //  PrtTim("retour  carteVersAlm"); | 
|---|
| 470 | if (iterationOrder > 0) | 
|---|
| 471 | { | 
|---|
| 472 | TVector<int_4> fact(iterationOrder+2); | 
|---|
| 473 | fact(0) = 1; | 
|---|
| 474 | int k; | 
|---|
| 475 | for (k=1; k <= iterationOrder+1; k++) | 
|---|
| 476 | { | 
|---|
| 477 | fact(k) = fact(k-1)*k; | 
|---|
| 478 | } | 
|---|
| 479 | Alm<T> alm2(alm); | 
|---|
| 480 | T Tzero = (T)0.; | 
|---|
| 481 | complex<T> complexZero = complex<T>(Tzero, Tzero); | 
|---|
| 482 | alm = complexZero; | 
|---|
| 483 | int signe = 1; | 
|---|
| 484 | int nbIteration = iterationOrder+1; | 
|---|
| 485 | for (k=1; k <= nbIteration; k++) | 
|---|
| 486 | { | 
|---|
| 487 | T facMult = (T)(0.5*signe*fact(iterationOrder)*(2*nbIteration-k)/(fact(k)*fact(nbIteration-k))); | 
|---|
| 488 | for (int m = 0; m <= nmmax; m++) | 
|---|
| 489 | { | 
|---|
| 490 | for (int l = m; l<= nlmax; l++) | 
|---|
| 491 | { | 
|---|
| 492 | alm(l,m) += facMult*alm2(l,m); | 
|---|
| 493 | } | 
|---|
| 494 | } | 
|---|
| 495 | if (k == nbIteration) break; | 
|---|
| 496 | signe = -signe; | 
|---|
| 497 | for (int k=0; k< map.NbPixels(); k++) map(k) = (T)0.; | 
|---|
| 498 | //        synthetize a map from the estimated alm | 
|---|
| 499 | //      PrtTim("appel  GenerateFromAlm"); | 
|---|
| 500 | GenerateFromAlm( map, map.SizeIndex(), alm2); | 
|---|
| 501 | //      PrtTim("retour  GenerateFromAlm"); | 
|---|
| 502 | alm2 = complexZero; | 
|---|
| 503 | //        analyse the new map | 
|---|
| 504 | //      PrtTim("appel  carteVersAlm"); | 
|---|
| 505 | carteVersAlm(map, nlmax, cos_theta_cut, alm2); | 
|---|
| 506 | //      PrtTim("retour  carteVersAlm"); | 
|---|
| 507 | } | 
|---|
| 508 | } | 
|---|
| 509 | } | 
|---|
| 510 |  | 
|---|
| 511 | template<class T> | 
|---|
| 512 | void SphericalTransformServer<T>::carteVersAlm(const SphericalMap<T>& map, int_4 nlmax, r_8 cos_theta_cut, Alm<T>& alm) const | 
|---|
| 513 | { | 
|---|
| 514 |  | 
|---|
| 515 | /*----------------------------------------------------------------------- | 
|---|
| 516 | computes the integral in phi : phas_m(theta) | 
|---|
| 517 | for each parallele from north to south pole | 
|---|
| 518 | -----------------------------------------------------------------------*/ | 
|---|
| 519 | TVector<T> data; | 
|---|
| 520 | TVector<int_4> pixNumber; | 
|---|
| 521 | int_4  nmmax = nlmax; | 
|---|
| 522 | TVector< complex<T> > phase(nmmax+1); | 
|---|
| 523 |  | 
|---|
| 524 | alm.ReSizeToLmax(nlmax); | 
|---|
| 525 | for (int_4 ith = 0; ith < map.NbThetaSlices(); ith++) | 
|---|
| 526 | { | 
|---|
| 527 | r_8 phi0; | 
|---|
| 528 | r_8 theta; | 
|---|
| 529 | //  PrtTim("debut 1ere tranche "); | 
|---|
| 530 | map.GetThetaSlice(ith,theta,phi0,pixNumber ,data); | 
|---|
| 531 | phase = complex<T>((T)0.,(T)0.); | 
|---|
| 532 | double cth = cos(theta); | 
|---|
| 533 |  | 
|---|
| 534 | //part of the sky out of the symetric cut | 
|---|
| 535 | bool keep_it = (fabs(cth) >= cos_theta_cut); | 
|---|
| 536 |  | 
|---|
| 537 | //    PrtTim("fin 1ere tranche "); | 
|---|
| 538 |  | 
|---|
| 539 | if (keep_it) | 
|---|
| 540 | { | 
|---|
| 541 | //      phase = CFromFourierAnalysis(nmmax,data,phi0); | 
|---|
| 542 | //      PrtTim("avant Fourier "); | 
|---|
| 543 | CFromFourierAnalysis(nmmax,data,phase, phi0); | 
|---|
| 544 | //      PrtTim("apres Fourier "); | 
|---|
| 545 |  | 
|---|
| 546 | } | 
|---|
| 547 |  | 
|---|
| 548 | //      --------------------------------------------------------------------- | 
|---|
| 549 | //      computes the a_lm by integrating over theta | 
|---|
| 550 | //      lambda_lm(theta) * phas_m(theta) | 
|---|
| 551 | //      for each m and l | 
|---|
| 552 | //      ----------------------------------------------------------------------- | 
|---|
| 553 | //        PrtTim("avant instanciation LM "); | 
|---|
| 554 | LambdaLMBuilder lb(theta,nlmax,nmmax); | 
|---|
| 555 | //        PrtTim("apres instanciation LM "); | 
|---|
| 556 | r_8 domega=map.PixSolAngle(map.PixIndexSph(theta,phi0)); | 
|---|
| 557 |  | 
|---|
| 558 | //   PrtTim("avant mise a jour Alm "); | 
|---|
| 559 | complex<T> fi; | 
|---|
| 560 | T facteur; | 
|---|
| 561 | int index; | 
|---|
| 562 | for (int m = 0; m <= nmmax; m++) | 
|---|
| 563 | { | 
|---|
| 564 | fi = phase(m); | 
|---|
| 565 | for (int l = m; l<= nlmax; l++) | 
|---|
| 566 | { | 
|---|
| 567 | index = alm.indexOfElement(l,m); | 
|---|
| 568 | //  facteur = (T)(lb.lamlm(l,m) * domega); | 
|---|
| 569 | facteur = (T)(lb.lamlm(index) * domega); | 
|---|
| 570 | // alm(l,m) += facteur * fi ; | 
|---|
| 571 | alm(index) += facteur * fi ; | 
|---|
| 572 | } | 
|---|
| 573 | } | 
|---|
| 574 |  | 
|---|
| 575 |  | 
|---|
| 576 |  | 
|---|
| 577 | // | 
|---|
| 578 | // | 
|---|
| 579 | //       PrtTim("apres mise a jour Alm "); | 
|---|
| 580 | } | 
|---|
| 581 | } | 
|---|
| 582 | /*! \fn TVector< complex<T> > SOPHYA::SphericalTransformServer::CFromFourierAnalysis(int_4 nmmax, const TVector<complex<T> >datain, r_8 phi0) const | 
|---|
| 583 |  | 
|---|
| 584 | \return a vector with mmax elements  which are  sums : | 
|---|
| 585 | \f$\sum_{k=0}^{nphi}datain(\theta,\varphi_k)e^{im\varphi_k}\f$ for (mmax+1) values of \f$m\f$ from 0 to mmax. | 
|---|
| 586 | */ | 
|---|
| 587 | template<class T> | 
|---|
| 588 | TVector< complex<T> > SphericalTransformServer<T>::CFromFourierAnalysis(int_4 nmmax, const TVector<complex<T> >datain, r_8 phi0) const | 
|---|
| 589 | { | 
|---|
| 590 | /*======================================================================= | 
|---|
| 591 | integrates (data * phi-dependence-of-Ylm) over phi | 
|---|
| 592 | --> function of m can be computed by FFT | 
|---|
| 593 |  | 
|---|
| 594 | datain est modifie | 
|---|
| 595 | =======================================================================*/ | 
|---|
| 596 | int_4 nph=datain.NElts(); | 
|---|
| 597 | if (nph <= 0) | 
|---|
| 598 | { | 
|---|
| 599 | throw PException("bizarre : vecteur datain de longueur nulle (CFromFourierAnalysis)"); | 
|---|
| 600 | } | 
|---|
| 601 | TVector<complex<T> > transformedData(nph); | 
|---|
| 602 | fftIntfPtr_-> FFTForward(datain, transformedData); | 
|---|
| 603 |  | 
|---|
| 604 | TVector< complex<T> > dataout(nmmax+1); | 
|---|
| 605 |  | 
|---|
| 606 | int im_max=min(nph,nmmax+1); | 
|---|
| 607 | int i; | 
|---|
| 608 | dataout = complex<T>((T)0.,(T)0.); | 
|---|
| 609 | //  for (i=0;i< dataout.NElts();i++) dataout(i)=complex<T>((T)0.,(T)0.); | 
|---|
| 610 | for (i=0;i<im_max;i++) dataout(i)=transformedData(i); | 
|---|
| 611 |  | 
|---|
| 612 |  | 
|---|
| 613 | for (int kk=nph; kk<dataout.NElts(); kk++) dataout(kk)=dataout(kk%nph); | 
|---|
| 614 | for (i = 0;i <dataout.NElts();i++){ | 
|---|
| 615 | dataout(i)*= (complex<T>)(complex<double>(cos(-i*phi0),sin(-i*phi0))); | 
|---|
| 616 | } | 
|---|
| 617 | return dataout; | 
|---|
| 618 | } | 
|---|
| 619 |  | 
|---|
| 620 | //&&&&&&&&& nouvelle version | 
|---|
| 621 | /* \fn TVector< complex<T> > SOPHYA::SphericalTransformServer::CFromFourierAnalysis(int_4 nmmax, const TVector<T> datain, r_8 phi0) const | 
|---|
| 622 |  | 
|---|
| 623 | same as previous one, but with a "datain" which is real (not complex) */ | 
|---|
| 624 | template<class T> | 
|---|
| 625 | void SphericalTransformServer<T>::CFromFourierAnalysis(int_4 nmmax, const TVector<T> datain, TVector< complex<T> >& dataout, r_8 phi0) const | 
|---|
| 626 | { | 
|---|
| 627 | //======================================================================= | 
|---|
| 628 | //    integrates (data * phi-dependence-of-Ylm) over phi | 
|---|
| 629 | //    --> function of m can be computed by FFT | 
|---|
| 630 | //   !     with  0<= m <= npoints/2 (: Nyquist) | 
|---|
| 631 | //   !     because the data is real the negative m are the conjugate of the | 
|---|
| 632 | //   !     positive ones | 
|---|
| 633 |  | 
|---|
| 634 | //    datain est modifie | 
|---|
| 635 | // | 
|---|
| 636 | //    ======================================================================= | 
|---|
| 637 | int_4 nph=datain.NElts(); | 
|---|
| 638 | if (nph <= 0) | 
|---|
| 639 | { | 
|---|
| 640 | throw PException("bizarre : vecteur datain de longueur nulle (CFromFourierAnalysis)"); | 
|---|
| 641 | } | 
|---|
| 642 | // if (nph%2 != 0 ) | 
|---|
| 643 | //  { | 
|---|
| 644 | //  throw PException("SphericalTransformServer<T>::CFromFourierAnalysis : longueur de datain impair ?"); | 
|---|
| 645 | //  } | 
|---|
| 646 | TVector<complex<T> > transformedData; | 
|---|
| 647 |  | 
|---|
| 648 | // la taille du vecteur complexe retourne est nph/2+1 (si la taille | 
|---|
| 649 | // du vecteur reel entre est nph) | 
|---|
| 650 | //   cout << " longueur de datain  = " << nph << endl; | 
|---|
| 651 | fftIntfPtr_-> FFTForward(datain, transformedData); | 
|---|
| 652 | //  cout <<  " taille de la transformee " << transformedData.Size() << endl; | 
|---|
| 653 | //  TVector< complex<T> > dataout(nmmax+1); | 
|---|
| 654 | dataout.ReSize(nmmax+1); | 
|---|
| 655 |  | 
|---|
| 656 | // on transfere le resultat de la fft dans dataout. | 
|---|
| 657 |  | 
|---|
| 658 | int maxFreqAccessiblesParFFT = min(nph/2,nmmax); | 
|---|
| 659 | int i; | 
|---|
| 660 | for (i=0;i<=maxFreqAccessiblesParFFT;i++) dataout(i)=transformedData(i); | 
|---|
| 661 |  | 
|---|
| 662 |  | 
|---|
| 663 | // si dataout n'est pas plein, on complete jusqu'a  nph+1 valeurs (a moins | 
|---|
| 664 | // que dataout ne soit plein avant d'atteindre nph) | 
|---|
| 665 | if (maxFreqAccessiblesParFFT != nmmax ) | 
|---|
| 666 | { | 
|---|
| 667 | int maxMfft = min(nph,nmmax); | 
|---|
| 668 | for (i=maxFreqAccessiblesParFFT+1; i<=maxMfft; i++) | 
|---|
| 669 | { | 
|---|
| 670 | dataout(i) = conj(dataout(nph-i) ); | 
|---|
| 671 | } | 
|---|
| 672 | // on conplete, si necessaire, par periodicite | 
|---|
| 673 | if ( maxMfft != nmmax ) | 
|---|
| 674 | { | 
|---|
| 675 | for (int kk=nph+1; kk <= nmmax; kk++) | 
|---|
| 676 | { | 
|---|
| 677 | dataout(kk)=dataout(kk%nph); | 
|---|
| 678 | } | 
|---|
| 679 | } | 
|---|
| 680 | } | 
|---|
| 681 | for (i = 0;i <dataout.NElts();i++) | 
|---|
| 682 | { | 
|---|
| 683 | dataout(i)*= (complex<T>)(complex<double>(cos(-i*phi0),sin(-i*phi0))); | 
|---|
| 684 | } | 
|---|
| 685 | //  return dataout; | 
|---|
| 686 | } | 
|---|
| 687 |  | 
|---|
| 688 | /*! \fn void SOPHYA::SphericalTransformServer::GenerateFromAlm(SphericalMap<T>& mapq, | 
|---|
| 689 | SphericalMap<T>& mapu, | 
|---|
| 690 | int_4 pixelSizeIndex, | 
|---|
| 691 | const Alm<T>& alme, | 
|---|
| 692 | const Alm<T>& almb) const | 
|---|
| 693 |  | 
|---|
| 694 | synthesis of a polarization map from  Alm coefficients. The spheres mapq and mapu contain respectively the Stokes parameters. */ | 
|---|
| 695 | template<class T> | 
|---|
| 696 | void SphericalTransformServer<T>::GenerateFromAlm(SphericalMap<T>& mapq, | 
|---|
| 697 | SphericalMap<T>& mapu, | 
|---|
| 698 | int_4 pixelSizeIndex, | 
|---|
| 699 | const Alm<T>& alme, | 
|---|
| 700 | const Alm<T>& almb) const | 
|---|
| 701 | { | 
|---|
| 702 | /*======================================================================= | 
|---|
| 703 | computes a map form its alm for the HEALPIX pixelisation | 
|---|
| 704 | map(theta,phi) = sum_l_m a_lm Y_lm(theta,phi) | 
|---|
| 705 | = sum_m {e^(i*m*phi) sum_l a_lm*lambda_lm(theta)} | 
|---|
| 706 |  | 
|---|
| 707 | where Y_lm(theta,phi) = lambda(theta) * e^(i*m*phi) | 
|---|
| 708 |  | 
|---|
| 709 | * the recurrence of Ylm is the standard one (cf Num Rec) | 
|---|
| 710 | * the sum over m is done by FFT | 
|---|
| 711 |  | 
|---|
| 712 | =======================================================================*/ | 
|---|
| 713 | int_4 nlmax=alme.Lmax(); | 
|---|
| 714 | if (nlmax != almb.Lmax()) | 
|---|
| 715 | { | 
|---|
| 716 | cout << " SphericalTransformServer: les deux tableaux alm n'ont pas la meme taille" << endl; | 
|---|
| 717 | throw SzMismatchError("SphericalTransformServer: les deux tableaux alm n'ont pas la meme taille"); | 
|---|
| 718 | } | 
|---|
| 719 | int_4 nmmax=nlmax; | 
|---|
| 720 | int_4 nsmax=0; | 
|---|
| 721 | mapq.Resize(pixelSizeIndex); | 
|---|
| 722 | mapu.Resize(pixelSizeIndex); | 
|---|
| 723 | string sphere_type=mapq.TypeOfMap(); | 
|---|
| 724 | if (sphere_type != mapu.TypeOfMap()) | 
|---|
| 725 | { | 
|---|
| 726 | cout <<  " SphericalTransformServer: les deux spheres ne sont pas de meme type" << endl; | 
|---|
| 727 | cout << " type 1 " << sphere_type << endl; | 
|---|
| 728 | cout << " type 2 " << mapu.TypeOfMap() << endl; | 
|---|
| 729 | throw SzMismatchError("SphericalTransformServer: les deux spheres ne sont pas de meme type"); | 
|---|
| 730 |  | 
|---|
| 731 | } | 
|---|
| 732 | bool healpix = true; | 
|---|
| 733 | if (sphere_type.substr(0,4) == "RING") | 
|---|
| 734 | { | 
|---|
| 735 | nsmax=mapq.SizeIndex(); | 
|---|
| 736 | } | 
|---|
| 737 | else | 
|---|
| 738 | // pour une sphere Gorski le nombre de pixels est 12*nsmax**2 | 
|---|
| 739 | // on calcule une quantite equivalente a nsmax pour la sphere-theta-phi | 
|---|
| 740 | // en vue de l'application du critere Healpix : nlmax<=3*nsmax-1 | 
|---|
| 741 | // c'est approximatif ; a raffiner. | 
|---|
| 742 | healpix = false; | 
|---|
| 743 | if (sphere_type.substr(0,6) == "TETAFI") | 
|---|
| 744 | { | 
|---|
| 745 | nsmax=(int_4)sqrt(mapq.NbPixels()/12.); | 
|---|
| 746 | } | 
|---|
| 747 | else | 
|---|
| 748 | { | 
|---|
| 749 | cout << " unknown type of sphere : " << sphere_type << endl; | 
|---|
| 750 | throw IOExc(" unknown type of sphere "); | 
|---|
| 751 | } | 
|---|
| 752 | cout << "GenerateFromAlm: the spheres are of type : " << sphere_type << endl; | 
|---|
| 753 | cout << "GenerateFromAlm: size indices (nside) of  spheres= " << nsmax << endl; | 
|---|
| 754 | cout << "GenerateFromAlm: nlmax (from Alm) = " << nlmax << endl; | 
|---|
| 755 | if (nlmax>3*nsmax-1) | 
|---|
| 756 | { | 
|---|
| 757 | cout << "GenerateFromAlm: nlmax should be <= 3*nside-1" << endl; | 
|---|
| 758 | if (sphere_type.substr(0,6) == "TETAFI") | 
|---|
| 759 | { | 
|---|
| 760 | cout << " (for this criterium, nsmax is computed as sqrt(nbPixels/12))" << endl; | 
|---|
| 761 | } | 
|---|
| 762 | } | 
|---|
| 763 | if (alme.Lmax()!=almb.Lmax()) | 
|---|
| 764 | { | 
|---|
| 765 | cout << "GenerateFromAlm: arrays Alme and Almb have not the same size ? " << endl; | 
|---|
| 766 | throw SzMismatchError("SphericalTransformServer: arrays Alme and Almb have not the same size ?  "); | 
|---|
| 767 | } | 
|---|
| 768 | mapFromWX(nlmax, nmmax, mapq, mapu, alme, almb, healpix); | 
|---|
| 769 | // mapFromPM(nlmax, nmmax, mapq, mapu, alme, almb); | 
|---|
| 770 | } | 
|---|
| 771 | /*! \fn void SOPHYA::SphericalTransformServer::DecomposeToAlm(const SphericalMap<T>& mapq, | 
|---|
| 772 | const SphericalMap<T>& mapu, | 
|---|
| 773 | Alm<T>& alme, | 
|---|
| 774 | Alm<T>& almb, | 
|---|
| 775 | int_4 nlmax, | 
|---|
| 776 | r_8 cos_theta_cut) const | 
|---|
| 777 |  | 
|---|
| 778 | analysis of a polarization map into Alm coefficients. | 
|---|
| 779 |  | 
|---|
| 780 | The spheres \c mapq and \c mapu contain respectively the Stokes parameters. | 
|---|
| 781 |  | 
|---|
| 782 | \c a2lme and \c a2lmb will receive respectively electric and magnetic Alm's | 
|---|
| 783 | nlmax : maximum value of the l index | 
|---|
| 784 |  | 
|---|
| 785 | \c cos_theta_cut : cosinus of the symmetric cut EULER angle theta : cos_theta_cut=0 means no cut ; cos_theta_cut=1 all the sphere is cut. | 
|---|
| 786 |  | 
|---|
| 787 |  | 
|---|
| 788 | */ | 
|---|
| 789 | template<class T> | 
|---|
| 790 | void SphericalTransformServer<T>::DecomposeToAlm(const SphericalMap<T>& mapq, | 
|---|
| 791 | const SphericalMap<T>& mapu, | 
|---|
| 792 | Alm<T>& alme, | 
|---|
| 793 | Alm<T>& almb, | 
|---|
| 794 | int_4 nlmax, | 
|---|
| 795 | r_8 cos_theta_cut) const | 
|---|
| 796 | { | 
|---|
| 797 | DecomposeToAlm(const_cast< SphericalMap<T>& >(mapq), const_cast< SphericalMap<T>& >(mapu), alme, almb, nlmax, cos_theta_cut); | 
|---|
| 798 | } | 
|---|
| 799 |  | 
|---|
| 800 | /*! \fn void SOPHYA::SphericalTransformServer::DecomposeToAlm(const SphericalMap<T>& mapq, | 
|---|
| 801 | const SphericalMap<T>& mapu, | 
|---|
| 802 | Alm<T>& alme, | 
|---|
| 803 | Alm<T>& almb, | 
|---|
| 804 | int_4 nlmax, | 
|---|
| 805 | r_8 cos_theta_cut, | 
|---|
| 806 | int iterationOrder) const | 
|---|
| 807 |  | 
|---|
| 808 | analysis of a polarization map into Alm coefficients. | 
|---|
| 809 |  | 
|---|
| 810 | The spheres \c mapq and \c mapu contain respectively the Stokes parameters. | 
|---|
| 811 |  | 
|---|
| 812 | \c a2lme and \c a2lmb will receive respectively electric and magnetic Alm's | 
|---|
| 813 | nlmax : maximum value of the l index | 
|---|
| 814 |  | 
|---|
| 815 | \c cos_theta_cut : cosinus of the symmetric cut EULER angle theta : cos_theta_cut=0 means no cut ; cos_theta_cut=1 all the sphere is cut. | 
|---|
| 816 |  | 
|---|
| 817 | \param<iterationOrder> : 1,2,3,4.... order of an iterative analysis. (Default : 0 -> standard analysis). If iterationOrder is not null, the method works with SphereHEALPix but NOT WITH SphereThetaPhi maps ! | 
|---|
| 818 |  | 
|---|
| 819 | THE INPUT MAPS CAN BE MODIFIED (only if iterationOrder >0) | 
|---|
| 820 |  | 
|---|
| 821 | */ | 
|---|
| 822 | template<class T> | 
|---|
| 823 | void SphericalTransformServer<T>::DecomposeToAlm(SphericalMap<T>& mapq, | 
|---|
| 824 | SphericalMap<T>& mapu, | 
|---|
| 825 | Alm<T>& alme, | 
|---|
| 826 | Alm<T>& almb, | 
|---|
| 827 | int_4 nlmax, | 
|---|
| 828 | r_8 cos_theta_cut, | 
|---|
| 829 | int iterationOrder) const | 
|---|
| 830 | { | 
|---|
| 831 | int_4  nmmax = nlmax; | 
|---|
| 832 | carteVersAlm(mapq, mapu, alme, almb, nlmax, cos_theta_cut); | 
|---|
| 833 | if (iterationOrder > 0) | 
|---|
| 834 | { | 
|---|
| 835 | TVector<int_4> fact(iterationOrder+2); | 
|---|
| 836 | fact(0) = 1; | 
|---|
| 837 | int k; | 
|---|
| 838 | for (k=1; k <= iterationOrder+1; k++) | 
|---|
| 839 | { | 
|---|
| 840 | fact(k) = fact(k-1)*k; | 
|---|
| 841 | } | 
|---|
| 842 | Alm<T> alme2(alme); | 
|---|
| 843 | Alm<T> almb2(almb); | 
|---|
| 844 | T Tzero = (T)0.; | 
|---|
| 845 | complex<T> complexZero = complex<T>(Tzero, Tzero); | 
|---|
| 846 | alme = complexZero; | 
|---|
| 847 | almb = complexZero; | 
|---|
| 848 | int signe = 1; | 
|---|
| 849 | int nbIteration = iterationOrder+1; | 
|---|
| 850 | for (k=1; k <= nbIteration; k++) | 
|---|
| 851 | { | 
|---|
| 852 | T facMult = (T)(0.5*signe*fact(iterationOrder)*(2*nbIteration-k)/(fact(k)*fact(nbIteration-k))); | 
|---|
| 853 | for (int m = 0; m <= nmmax; m++) | 
|---|
| 854 | { | 
|---|
| 855 | for (int l = m; l<= nlmax; l++) | 
|---|
| 856 | { | 
|---|
| 857 | alme(l,m) += facMult*alme2(l,m); | 
|---|
| 858 | almb(l,m) += facMult*almb2(l,m); | 
|---|
| 859 | } | 
|---|
| 860 | } | 
|---|
| 861 | if (k == nbIteration) break; | 
|---|
| 862 | signe = -signe; | 
|---|
| 863 | for (int k=0; k< mapq.NbPixels(); k++) | 
|---|
| 864 | { | 
|---|
| 865 | mapq(k) = (T)0.; | 
|---|
| 866 | mapu(k) = (T)0.; | 
|---|
| 867 | } | 
|---|
| 868 | //        synthetize a map from the estimated alm | 
|---|
| 869 | GenerateFromAlm(mapq,mapu,mapq.SizeIndex(),alme2,almb2); | 
|---|
| 870 | alme2 = complexZero; | 
|---|
| 871 | almb2 = complexZero; | 
|---|
| 872 | //        analyse the new map | 
|---|
| 873 | carteVersAlm(mapq, mapu, alme2, almb2, nlmax, cos_theta_cut); | 
|---|
| 874 | } | 
|---|
| 875 | } | 
|---|
| 876 | } | 
|---|
| 877 |  | 
|---|
| 878 | template<class T> | 
|---|
| 879 | void SphericalTransformServer<T>::carteVersAlm(const SphericalMap<T>& mapq, | 
|---|
| 880 | const SphericalMap<T>& mapu, | 
|---|
| 881 | Alm<T>& alme, | 
|---|
| 882 | Alm<T>& almb, | 
|---|
| 883 | int_4 nlmax, | 
|---|
| 884 | r_8 cos_theta_cut) const | 
|---|
| 885 | { | 
|---|
| 886 | int_4  nmmax = nlmax; | 
|---|
| 887 | // resize et remise a zero | 
|---|
| 888 | alme.ReSizeToLmax(nlmax); | 
|---|
| 889 | almb.ReSizeToLmax(nlmax); | 
|---|
| 890 |  | 
|---|
| 891 |  | 
|---|
| 892 | TVector<T> dataq; | 
|---|
| 893 | TVector<T> datau; | 
|---|
| 894 | TVector<int_4> pixNumber; | 
|---|
| 895 |  | 
|---|
| 896 | string sphere_type=mapq.TypeOfMap(); | 
|---|
| 897 | if (sphere_type != mapu.TypeOfMap()) | 
|---|
| 898 | { | 
|---|
| 899 | cout <<  " SphericalTransformServer: les deux spheres ne sont pas de meme type" << endl; | 
|---|
| 900 | cout << " type 1 " << sphere_type << endl; | 
|---|
| 901 | cout << " type 2 " << mapu.TypeOfMap() << endl; | 
|---|
| 902 | throw SzMismatchError("SphericalTransformServer: les deux spheres ne sont pas de meme type"); | 
|---|
| 903 |  | 
|---|
| 904 | } | 
|---|
| 905 | if (mapq.NbPixels()!=mapu.NbPixels()) | 
|---|
| 906 | { | 
|---|
| 907 | cout << " DecomposeToAlm: map Q and map U have not same size ?" << endl; | 
|---|
| 908 | throw SzMismatchError("SphericalTransformServer::DecomposeToAlm: map Q and map U have not same size "); | 
|---|
| 909 | } | 
|---|
| 910 | for (int_4 ith = 0; ith < mapq.NbThetaSlices(); ith++) | 
|---|
| 911 | { | 
|---|
| 912 | r_8 phi0; | 
|---|
| 913 | r_8 theta; | 
|---|
| 914 | mapq.GetThetaSlice(ith,theta,phi0, pixNumber,dataq); | 
|---|
| 915 | mapu.GetThetaSlice(ith,theta,phi0, pixNumber,datau); | 
|---|
| 916 | if (dataq.NElts() != datau.NElts() ) | 
|---|
| 917 | { | 
|---|
| 918 | throw  SzMismatchError("the spheres have not the same pixelization"); | 
|---|
| 919 | } | 
|---|
| 920 | r_8 domega=mapq.PixSolAngle(mapq.PixIndexSph(theta,phi0)); | 
|---|
| 921 | double cth = cos(theta); | 
|---|
| 922 | //part of the sky out of the symetric cut | 
|---|
| 923 | bool keep_it = (fabs(cth) >= cos_theta_cut); | 
|---|
| 924 | if (keep_it) | 
|---|
| 925 | { | 
|---|
| 926 | //  almFromPM(pixNumber.NElts(), nlmax, nmmax, phi0, domega, theta, dataq, datau, alme, almb); | 
|---|
| 927 | almFromWX(nlmax, nmmax, phi0, domega, theta, dataq, datau, alme, almb); | 
|---|
| 928 | } | 
|---|
| 929 | } | 
|---|
| 930 | } | 
|---|
| 931 |  | 
|---|
| 932 |  | 
|---|
| 933 | /*! \fn void SOPHYA::SphericalTransformServer::almFromWX(int_4 nlmax, int_4 nmmax, | 
|---|
| 934 | r_8 phi0, r_8 domega, | 
|---|
| 935 | r_8 theta, | 
|---|
| 936 | const TVector<T>& dataq, | 
|---|
| 937 | const TVector<T>& datau, | 
|---|
| 938 | Alm<T>& alme, | 
|---|
| 939 | Alm<T>& almb) const | 
|---|
| 940 |  | 
|---|
| 941 | Compute polarized Alm's as : | 
|---|
| 942 | \f[ | 
|---|
| 943 | a_{lm}^E=\frac{1}{\sqrt{2}}\sum_{slices}{\omega_{pix}\left(\,_{w}\lambda_l^m\tilde{Q}-i\,_{x}\lambda_l^m\tilde{U}\right)} | 
|---|
| 944 | \f] | 
|---|
| 945 | \f[ | 
|---|
| 946 | a_{lm}^B=\frac{1}{\sqrt{2}}\sum_{slices}{\omega_{pix}\left(i\,_{x}\lambda_l^m\tilde{Q}+\,_{w}\lambda_l^m\tilde{U}\right)} | 
|---|
| 947 | \f] | 
|---|
| 948 |  | 
|---|
| 949 | where \f$\tilde{Q}\f$ and \f$\tilde{U}\f$ are C-coefficients computed by FFT (method CFromFourierAnalysis, called by present method) from the Stokes parameters. | 
|---|
| 950 |  | 
|---|
| 951 | \f$\omega_{pix}\f$ are solid angle of each pixel. | 
|---|
| 952 |  | 
|---|
| 953 | dataq, datau : Stokes parameters. | 
|---|
| 954 |  | 
|---|
| 955 | */ | 
|---|
| 956 | template<class T> | 
|---|
| 957 | void SphericalTransformServer<T>::almFromWX(int_4 nlmax, int_4 nmmax, | 
|---|
| 958 | r_8 phi0, r_8 domega, | 
|---|
| 959 | r_8 theta, | 
|---|
| 960 | const TVector<T>& dataq, | 
|---|
| 961 | const TVector<T>& datau, | 
|---|
| 962 | Alm<T>& alme, | 
|---|
| 963 | Alm<T>& almb) const | 
|---|
| 964 | { | 
|---|
| 965 | TVector< complex<T> > phaseq(nmmax+1); | 
|---|
| 966 | TVector< complex<T> > phaseu(nmmax+1); | 
|---|
| 967 | //  TVector<complex<T> > datain(nph); | 
|---|
| 968 | for (int i=0;i< nmmax+1;i++) | 
|---|
| 969 | { | 
|---|
| 970 | phaseq(i)=0; | 
|---|
| 971 | phaseu(i)=0; | 
|---|
| 972 | } | 
|---|
| 973 | //  for(int kk=0; kk<nph; kk++) datain(kk)=complex<T>(dataq(kk),0.); | 
|---|
| 974 |  | 
|---|
| 975 | //  phaseq = CFromFourierAnalysis(nmmax,dataq,phi0); | 
|---|
| 976 | CFromFourierAnalysis(nmmax,dataq,phaseq, phi0); | 
|---|
| 977 |  | 
|---|
| 978 | //  phaseu=  CFromFourierAnalysis(nmmax,datau,phi0); | 
|---|
| 979 | CFromFourierAnalysis(nmmax,datau,phaseu, phi0); | 
|---|
| 980 |  | 
|---|
| 981 | LambdaWXBuilder lwxb(theta,nlmax,nmmax); | 
|---|
| 982 |  | 
|---|
| 983 | r_8 sqr2inv=1/Rac2; | 
|---|
| 984 | for (int m = 0; m <= nmmax; m++) | 
|---|
| 985 | { | 
|---|
| 986 | r_8 lambda_w=0.; | 
|---|
| 987 | r_8 lambda_x=0.; | 
|---|
| 988 | lwxb.lam_wx(m, m, lambda_w, lambda_x); | 
|---|
| 989 | complex<T>  zi_lam_x((T)0., (T)lambda_x); | 
|---|
| 990 | alme(m,m) +=  ( (T)(lambda_w)*phaseq(m)-zi_lam_x*phaseu(m) )*(T)(domega*sqr2inv); | 
|---|
| 991 | almb(m,m) +=  ( (T)(lambda_w)*phaseu(m)+zi_lam_x*phaseq(m) )*(T)(domega*sqr2inv); | 
|---|
| 992 |  | 
|---|
| 993 | for (int l = m+1; l<= nlmax; l++) | 
|---|
| 994 | { | 
|---|
| 995 | lwxb.lam_wx(l, m, lambda_w, lambda_x); | 
|---|
| 996 | zi_lam_x = complex<T>((T)0., (T)lambda_x); | 
|---|
| 997 | alme(l,m) +=  ( (T)(lambda_w)*phaseq(m)-zi_lam_x*phaseu(m) )*(T)(domega*sqr2inv); | 
|---|
| 998 | almb(l,m) +=  ( (T)(lambda_w)*phaseu(m)+zi_lam_x*phaseq(m) )*(T)(domega*sqr2inv); | 
|---|
| 999 | } | 
|---|
| 1000 | } | 
|---|
| 1001 | } | 
|---|
| 1002 |  | 
|---|
| 1003 |  | 
|---|
| 1004 | /*! \fn void SOPHYA::SphericalTransformServer::almFromPM(int_4 nph, int_4 nlmax, | 
|---|
| 1005 | int_4 nmmax, | 
|---|
| 1006 | r_8 phi0, r_8 domega, | 
|---|
| 1007 | r_8 theta, | 
|---|
| 1008 | const TVector<T>& dataq, | 
|---|
| 1009 | const TVector<T>& datau, | 
|---|
| 1010 | Alm<T>& alme, | 
|---|
| 1011 | Alm<T>& almb) const | 
|---|
| 1012 |  | 
|---|
| 1013 | Compute polarized Alm's as : | 
|---|
| 1014 | \f[ | 
|---|
| 1015 | a_{lm}^E=-\frac{1}{2}\sum_{slices}{\omega_{pix}\left(\,_{+}\lambda_l^m\tilde{P^+}+\,_{-}\lambda_l^m\tilde{P^-}\right)} | 
|---|
| 1016 | \f] | 
|---|
| 1017 | \f[ | 
|---|
| 1018 | a_{lm}^B=\frac{i}{2}\sum_{slices}{\omega_{pix}\left(\,_{+}\lambda_l^m\tilde{P^+}-\,_{-}\lambda_l^m\tilde{P^-}\right)} | 
|---|
| 1019 | \f] | 
|---|
| 1020 |  | 
|---|
| 1021 | where \f$\tilde{P^{\pm}}=\tilde{Q}\pm\tilde{U}\f$  computed by FFT (method CFromFourierAnalysis, called by present method) from the Stokes parameters,\f$Q\f$ and \f$U\f$ . | 
|---|
| 1022 |  | 
|---|
| 1023 | \f$\omega_{pix}\f$ are solid angle of each pixel. | 
|---|
| 1024 |  | 
|---|
| 1025 | dataq, datau : Stokes parameters. | 
|---|
| 1026 |  | 
|---|
| 1027 | */ | 
|---|
| 1028 | template<class T> | 
|---|
| 1029 | void SphericalTransformServer<T>::almFromPM(int_4 nph, int_4 nlmax, | 
|---|
| 1030 | int_4 nmmax, | 
|---|
| 1031 | r_8 phi0, r_8 domega, | 
|---|
| 1032 | r_8 theta, | 
|---|
| 1033 | const TVector<T>& dataq, | 
|---|
| 1034 | const TVector<T>& datau, | 
|---|
| 1035 | Alm<T>& alme, | 
|---|
| 1036 | Alm<T>& almb) const | 
|---|
| 1037 | { | 
|---|
| 1038 | TVector< complex<T> > phasep(nmmax+1); | 
|---|
| 1039 | TVector< complex<T> > phasem(nmmax+1); | 
|---|
| 1040 | TVector<complex<T> > datain(nph); | 
|---|
| 1041 | for (int i=0;i< nmmax+1;i++) | 
|---|
| 1042 | { | 
|---|
| 1043 | phasep(i)=0; | 
|---|
| 1044 | phasem(i)=0; | 
|---|
| 1045 | } | 
|---|
| 1046 | int kk; | 
|---|
| 1047 | for(kk=0; kk<nph; kk++) datain(kk)=complex<T>(dataq(kk),datau(kk)); | 
|---|
| 1048 |  | 
|---|
| 1049 | phasep = CFromFourierAnalysis(nmmax,datain,phi0); | 
|---|
| 1050 |  | 
|---|
| 1051 | for(kk=0; kk<nph; kk++) datain(kk)=complex<T>(dataq(kk),-datau(kk)); | 
|---|
| 1052 | phasem = CFromFourierAnalysis(nmmax,datain,phi0); | 
|---|
| 1053 | LambdaPMBuilder lpmb(theta,nlmax,nmmax); | 
|---|
| 1054 |  | 
|---|
| 1055 | for (int m = 0; m <= nmmax; m++) | 
|---|
| 1056 | { | 
|---|
| 1057 | r_8 lambda_p=0.; | 
|---|
| 1058 | r_8 lambda_m=0.; | 
|---|
| 1059 | complex<T> im((T)0.,(T)1.); | 
|---|
| 1060 | lpmb.lam_pm(m, m, lambda_p, lambda_m); | 
|---|
| 1061 |  | 
|---|
| 1062 | alme(m,m) +=   -( (T)(lambda_p)*phasep(m) + (T)(lambda_m)*phasem(m)  )*(T)(domega*0.5); | 
|---|
| 1063 | almb(m,m) +=  im*( (T)(lambda_p)*phasep(m) - (T)(lambda_m)*phasem(m) )*(T)(domega*0.5); | 
|---|
| 1064 | for (int l = m+1; l<= nlmax; l++) | 
|---|
| 1065 | { | 
|---|
| 1066 | lpmb.lam_pm(l, m, lambda_p, lambda_m); | 
|---|
| 1067 | alme(l,m) +=  -( (T)(lambda_p)*phasep(m) + (T)(lambda_m)*phasem(m)  )*(T)(domega*0.5); | 
|---|
| 1068 | almb(l,m) += im* ( (T)(lambda_p)*phasep(m) - (T)(lambda_m)*phasem(m) )*(T)(domega*0.5); | 
|---|
| 1069 | } | 
|---|
| 1070 | } | 
|---|
| 1071 | } | 
|---|
| 1072 |  | 
|---|
| 1073 |  | 
|---|
| 1074 | /*! \fn void SOPHYA::SphericalTransformServer::mapFromWX(int_4 nlmax, int_4 nmmax, | 
|---|
| 1075 | SphericalMap<T>& mapq, | 
|---|
| 1076 | SphericalMap<T>& mapu, | 
|---|
| 1077 | const Alm<T>& alme, | 
|---|
| 1078 | const Alm<T>& almb, bool healpix) const | 
|---|
| 1079 |  | 
|---|
| 1080 | synthesis of Stokes parameters following formulae : | 
|---|
| 1081 |  | 
|---|
| 1082 | \f[ | 
|---|
| 1083 | Q=\sum_{m=-mmax}^{mmax}b_m^qe^{im\varphi} | 
|---|
| 1084 | \f] | 
|---|
| 1085 | \f[ | 
|---|
| 1086 | U=\sum_{m=-mmax}^{mmax}b_m^ue^{im\varphi} | 
|---|
| 1087 | \f] | 
|---|
| 1088 |  | 
|---|
| 1089 | computed by FFT (method fourierSynthesisFromB called by the present one) | 
|---|
| 1090 |  | 
|---|
| 1091 | with : | 
|---|
| 1092 |  | 
|---|
| 1093 | \f[ | 
|---|
| 1094 | b_m^q=-\frac{1}{\sqrt{2}}\sum_{l=|m|}^{lmax}{\left(\,_{w}\lambda_l^ma_{lm}^E-i\,_{x}\lambda_l^ma_{lm}^B\right) } | 
|---|
| 1095 | \f] | 
|---|
| 1096 | \f[ | 
|---|
| 1097 | b_m^u=\frac{1}{\sqrt{2}}\sum_{l=|m|}^{lmax}{\left(i\,_{x}\lambda_l^ma_{lm}^E+\,_{w}\lambda_l^ma_{lm}^B\right) } | 
|---|
| 1098 | \f] | 
|---|
| 1099 | */ | 
|---|
| 1100 | template<class T> | 
|---|
| 1101 | void SphericalTransformServer<T>::mapFromWX(int_4 nlmax, int_4 nmmax, | 
|---|
| 1102 | SphericalMap<T>& mapq, | 
|---|
| 1103 | SphericalMap<T>& mapu, | 
|---|
| 1104 | const Alm<T>& alme, | 
|---|
| 1105 | const Alm<T>& almb, bool healpix) const | 
|---|
| 1106 | { | 
|---|
| 1107 | int i; | 
|---|
| 1108 |  | 
|---|
| 1109 | Bm<complex<T> > b_m_theta_q(nmmax); | 
|---|
| 1110 | Bm<complex<T> > b_m_theta_u(nmmax); | 
|---|
| 1111 |  | 
|---|
| 1112 | for (int_4 ith = 0; ith < mapq.NbThetaSlices();ith++) | 
|---|
| 1113 | { | 
|---|
| 1114 | int_4 nph; | 
|---|
| 1115 | r_8 phi0; | 
|---|
| 1116 | r_8 theta; | 
|---|
| 1117 | TVector<int_4>  pixNumber; | 
|---|
| 1118 | TVector<T> datan; | 
|---|
| 1119 |  | 
|---|
| 1120 | mapq.GetThetaSlice(ith,theta,phi0, pixNumber,datan); | 
|---|
| 1121 | nph =  pixNumber.NElts(); | 
|---|
| 1122 | //       ----------------------------------------------------- | 
|---|
| 1123 | //              for each theta, and each m, computes | 
|---|
| 1124 | //              b(m,theta) = sum_over_l>m (lambda_l_m(theta) * a_l_m) | 
|---|
| 1125 | //              ------------------------------------------------------ | 
|---|
| 1126 | LambdaWXBuilder lwxb(theta,nlmax,nmmax); | 
|---|
| 1127 | //      LambdaPMBuilder lpmb(theta,nlmax,nmmax); | 
|---|
| 1128 | r_8 sqr2inv=1/Rac2; | 
|---|
| 1129 | int m; | 
|---|
| 1130 | for (m = 0; m <= nmmax; m++) | 
|---|
| 1131 | { | 
|---|
| 1132 | r_8 lambda_w=0.; | 
|---|
| 1133 | r_8 lambda_x=0.; | 
|---|
| 1134 | lwxb.lam_wx(m, m, lambda_w, lambda_x); | 
|---|
| 1135 | complex<T>  zi_lam_x((T)0., (T)lambda_x); | 
|---|
| 1136 |  | 
|---|
| 1137 | b_m_theta_q(m) =  ( (T)(lambda_w) * alme(m,m) - zi_lam_x * almb(m,m))*(T)sqr2inv ; | 
|---|
| 1138 | b_m_theta_u(m) =  ( (T)(lambda_w) * almb(m,m) + zi_lam_x * alme(m,m))*(T)sqr2inv; | 
|---|
| 1139 |  | 
|---|
| 1140 |  | 
|---|
| 1141 | for (int l = m+1; l<= nlmax; l++) | 
|---|
| 1142 | { | 
|---|
| 1143 |  | 
|---|
| 1144 | lwxb.lam_wx(l, m, lambda_w, lambda_x); | 
|---|
| 1145 | zi_lam_x= complex<T>((T)0., (T)lambda_x); | 
|---|
| 1146 |  | 
|---|
| 1147 | b_m_theta_q(m) += ((T)(lambda_w)*alme(l,m)-zi_lam_x *almb(l,m))*(T)sqr2inv; | 
|---|
| 1148 | b_m_theta_u(m) += ((T)(lambda_w)*almb(l,m)+zi_lam_x *alme(l,m))*(T)sqr2inv; | 
|---|
| 1149 |  | 
|---|
| 1150 | } | 
|---|
| 1151 | } | 
|---|
| 1152 | //        obtains the negative m of b(m,theta) (= complex conjugate) | 
|---|
| 1153 | for (m=1;m<=nmmax;m++) | 
|---|
| 1154 | { | 
|---|
| 1155 | b_m_theta_q(-m) = conj(b_m_theta_q(m)); | 
|---|
| 1156 | b_m_theta_u(-m) = conj(b_m_theta_u(m)); | 
|---|
| 1157 | } | 
|---|
| 1158 | if (healpix) | 
|---|
| 1159 | { | 
|---|
| 1160 | TVector<T> Tempq = RfourierSynthesisFromB(b_m_theta_q,nph,phi0); | 
|---|
| 1161 | TVector<T> Tempu = RfourierSynthesisFromB(b_m_theta_u,nph,phi0); | 
|---|
| 1162 | for (i=0;i< nph;i++) | 
|---|
| 1163 | { | 
|---|
| 1164 | mapq(pixNumber(i))=Tempq(i); | 
|---|
| 1165 | mapu(pixNumber(i))=Tempu(i); | 
|---|
| 1166 | } | 
|---|
| 1167 | } | 
|---|
| 1168 | else | 
|---|
| 1169 | // pour des pixelisations quelconques (autres que HEALPix | 
|---|
| 1170 | //  nph n'est pas toujours pair | 
|---|
| 1171 | // ca fait des problemes pour les transformees de Fourier | 
|---|
| 1172 | // car le server de TF ajuste la longueur du vecteur reel | 
|---|
| 1173 | // en sortie de TF, bref, la securite veut qu'on prenne une | 
|---|
| 1174 | // TF complexe | 
|---|
| 1175 | { | 
|---|
| 1176 | TVector<complex<T> > Tempq = fourierSynthesisFromB(b_m_theta_q,nph,phi0); | 
|---|
| 1177 | TVector<complex<T> > Tempu = fourierSynthesisFromB(b_m_theta_u,nph,phi0); | 
|---|
| 1178 | for (i=0;i< nph;i++) | 
|---|
| 1179 | { | 
|---|
| 1180 | mapq(pixNumber(i))=Tempq(i).real(); | 
|---|
| 1181 | mapu(pixNumber(i))=Tempu(i).real(); | 
|---|
| 1182 | } | 
|---|
| 1183 | } | 
|---|
| 1184 | } | 
|---|
| 1185 | } | 
|---|
| 1186 | /*! \fn void SOPHYA::SphericalTransformServer::mapFromPM(int_4 nlmax, int_4 nmmax, | 
|---|
| 1187 | SphericalMap<T>& mapq, | 
|---|
| 1188 | SphericalMap<T>& mapu, | 
|---|
| 1189 | const Alm<T>& alme, | 
|---|
| 1190 | const Alm<T>& almb) const | 
|---|
| 1191 |  | 
|---|
| 1192 | synthesis of polarizations following formulae : | 
|---|
| 1193 |  | 
|---|
| 1194 | \f[ | 
|---|
| 1195 | P^+ = \sum_{m=-mmax}^{mmax} {b_m^+e^{im\varphi} } | 
|---|
| 1196 | \f] | 
|---|
| 1197 | \f[ | 
|---|
| 1198 | P^- = \sum_{m=-mmax}^{mmax} {b_m^-e^{im\varphi} } | 
|---|
| 1199 | \f] | 
|---|
| 1200 |  | 
|---|
| 1201 | computed by FFT (method fourierSynthesisFromB called by the present one) | 
|---|
| 1202 |  | 
|---|
| 1203 | with : | 
|---|
| 1204 |  | 
|---|
| 1205 | \f[ | 
|---|
| 1206 | b_m^+=-\sum_{l=|m|}^{lmax}{\,_{+}\lambda_l^m \left( a_{lm}^E+ia_{lm}^B \right) } | 
|---|
| 1207 | \f] | 
|---|
| 1208 | \f[ | 
|---|
| 1209 | b_m^-=-\sum_{l=|m|}^{lmax}{\,_{+}\lambda_l^m \left( a_{lm}^E-ia_{lm}^B \right) } | 
|---|
| 1210 | \f] | 
|---|
| 1211 | */ | 
|---|
| 1212 | template<class T> | 
|---|
| 1213 | void SphericalTransformServer<T>::mapFromPM(int_4 nlmax, int_4 nmmax, | 
|---|
| 1214 | SphericalMap<T>& mapq, | 
|---|
| 1215 | SphericalMap<T>& mapu, | 
|---|
| 1216 | const Alm<T>& alme, | 
|---|
| 1217 | const Alm<T>& almb) const | 
|---|
| 1218 | { | 
|---|
| 1219 | Bm<complex<T> > b_m_theta_p(nmmax); | 
|---|
| 1220 | Bm<complex<T> > b_m_theta_m(nmmax); | 
|---|
| 1221 | for (int_4 ith = 0; ith < mapq.NbThetaSlices();ith++) | 
|---|
| 1222 | { | 
|---|
| 1223 | int_4 nph; | 
|---|
| 1224 | r_8 phi0; | 
|---|
| 1225 | r_8 theta; | 
|---|
| 1226 | TVector<int_4> pixNumber; | 
|---|
| 1227 | TVector<T> datan; | 
|---|
| 1228 |  | 
|---|
| 1229 | mapq.GetThetaSlice(ith,theta,phi0, pixNumber,datan); | 
|---|
| 1230 | nph =  pixNumber.NElts(); | 
|---|
| 1231 |  | 
|---|
| 1232 | //       ----------------------------------------------------- | 
|---|
| 1233 | //              for each theta, and each m, computes | 
|---|
| 1234 | //              b(m,theta) = sum_over_l>m (lambda_l_m(theta) * a_l_m) | 
|---|
| 1235 | //------------------------------------------------------ | 
|---|
| 1236 |  | 
|---|
| 1237 | LambdaPMBuilder lpmb(theta,nlmax,nmmax); | 
|---|
| 1238 | int m; | 
|---|
| 1239 | for (m = 0; m <= nmmax; m++) | 
|---|
| 1240 | { | 
|---|
| 1241 | r_8 lambda_p=0.; | 
|---|
| 1242 | r_8 lambda_m=0.; | 
|---|
| 1243 | lpmb.lam_pm(m, m, lambda_p, lambda_m); | 
|---|
| 1244 | complex<T> im((T)0.,(T)1.); | 
|---|
| 1245 |  | 
|---|
| 1246 | b_m_theta_p(m) =  (T)(lambda_p )* (-alme(m,m) - im * almb(m,m)); | 
|---|
| 1247 | b_m_theta_m(m) =  (T)(lambda_m) * (-alme(m,m) + im * almb(m,m)); | 
|---|
| 1248 |  | 
|---|
| 1249 |  | 
|---|
| 1250 | for (int l = m+1; l<= nlmax; l++) | 
|---|
| 1251 | { | 
|---|
| 1252 | lpmb.lam_pm(l, m, lambda_p, lambda_m); | 
|---|
| 1253 | b_m_theta_p(m) +=  (T)(lambda_p)*(-alme(l,m)-im *almb(l,m)); | 
|---|
| 1254 | b_m_theta_m(m) +=  (T)(lambda_m)*(-alme(l,m)+im *almb(l,m)); | 
|---|
| 1255 | } | 
|---|
| 1256 | } | 
|---|
| 1257 |  | 
|---|
| 1258 | //        obtains the negative m of b(m,theta) (= complex conjugate) | 
|---|
| 1259 | for (m=1;m<=nmmax;m++) | 
|---|
| 1260 | { | 
|---|
| 1261 | b_m_theta_p(-m) = conj(b_m_theta_m(m)); | 
|---|
| 1262 | b_m_theta_m(-m) = conj(b_m_theta_p(m)); | 
|---|
| 1263 | } | 
|---|
| 1264 |  | 
|---|
| 1265 | TVector<complex<T> > Tempp = fourierSynthesisFromB(b_m_theta_p,nph,phi0); | 
|---|
| 1266 | TVector<complex<T> > Tempm = fourierSynthesisFromB(b_m_theta_m,nph,phi0); | 
|---|
| 1267 |  | 
|---|
| 1268 | for (int i=0;i< nph;i++) | 
|---|
| 1269 | { | 
|---|
| 1270 | mapq(pixNumber(i))=0.5*(Tempp(i)+Tempm(i)).real(); | 
|---|
| 1271 | mapu(pixNumber(i))=0.5*(Tempp(i)-Tempm(i)).imag(); | 
|---|
| 1272 | } | 
|---|
| 1273 | } | 
|---|
| 1274 | } | 
|---|
| 1275 |  | 
|---|
| 1276 |  | 
|---|
| 1277 | /*! \fn void SOPHYA::SphericalTransformServer::GenerateFromCl(SphericalMap<T>& sphq, | 
|---|
| 1278 | SphericalMap<T>& sphu, | 
|---|
| 1279 | int_4 pixelSizeIndex, | 
|---|
| 1280 | const TVector<T>& Cle, | 
|---|
| 1281 | const TVector<T>& Clb, | 
|---|
| 1282 | const r_8 fwhm) const | 
|---|
| 1283 |  | 
|---|
| 1284 | synthesis of a polarization  map from  power spectra electric-Cl and magnetic-Cl (Alm's are generated randomly, following a gaussian distribution). | 
|---|
| 1285 | \param fwhm FWHM in arcmin for random generation of Alm's (eg. 5) | 
|---|
| 1286 | */ | 
|---|
| 1287 | template<class T> | 
|---|
| 1288 | void SphericalTransformServer<T>::GenerateFromCl(SphericalMap<T>& sphq, | 
|---|
| 1289 | SphericalMap<T>& sphu, | 
|---|
| 1290 | int_4 pixelSizeIndex, | 
|---|
| 1291 | const TVector<T>& Cle, | 
|---|
| 1292 | const TVector<T>& Clb, | 
|---|
| 1293 | const r_8 fwhm) const | 
|---|
| 1294 | { | 
|---|
| 1295 | if (Cle.NElts() != Clb.NElts()) | 
|---|
| 1296 | { | 
|---|
| 1297 | cout << " SphericalTransformServer: les deux tableaux Cl n'ont pas la meme taille" << endl; | 
|---|
| 1298 | throw SzMismatchError("SphericalTransformServer::GenerateFromCl :  two Cl arrays have not same size"); | 
|---|
| 1299 | } | 
|---|
| 1300 |  | 
|---|
| 1301 | //  Alm<T> a2lme,a2lmb; | 
|---|
| 1302 | //  almFromCl(a2lme, Cle, fwhm); | 
|---|
| 1303 | //  almFromCl(a2lmb, Clb, fwhm); | 
|---|
| 1304 | //  Alm<T> a2lme = almFromCl(Cle, fwhm); | 
|---|
| 1305 | // Alm<T> a2lmb = almFromCl(Clb, fwhm); | 
|---|
| 1306 | Alm<T> a2lme(Cle, fwhm); | 
|---|
| 1307 | Alm<T> a2lmb(Clb, fwhm); | 
|---|
| 1308 |  | 
|---|
| 1309 | GenerateFromAlm(sphq,sphu,pixelSizeIndex,a2lme,a2lmb); | 
|---|
| 1310 | } | 
|---|
| 1311 | /*! \fn void SOPHYA::SphericalTransformServer::GenerateFromCl(SphericalMap<T>& sph, | 
|---|
| 1312 | int_4 pixelSizeIndex, | 
|---|
| 1313 | const TVector<T>& Cl, | 
|---|
| 1314 | const r_8 fwhm)  const | 
|---|
| 1315 |  | 
|---|
| 1316 | synthesis of a temperature  map from  power spectrum Cl (Alm's are generated randomly, following a gaussian distribution). */ | 
|---|
| 1317 | template<class T> | 
|---|
| 1318 | void SphericalTransformServer<T>::GenerateFromCl(SphericalMap<T>& sph, | 
|---|
| 1319 | int_4 pixelSizeIndex, | 
|---|
| 1320 | const TVector<T>& Cl, | 
|---|
| 1321 | const r_8 fwhm)  const | 
|---|
| 1322 | { | 
|---|
| 1323 |  | 
|---|
| 1324 | Alm<T> alm(Cl, fwhm); | 
|---|
| 1325 | GenerateFromAlm(sph,pixelSizeIndex, alm ); | 
|---|
| 1326 | } | 
|---|
| 1327 |  | 
|---|
| 1328 |  | 
|---|
| 1329 |  | 
|---|
| 1330 | /*! \fn TVector<T>  SOPHYA::SphericalTransformServer::DecomposeToCl(SphericalMap<T>& sph, int_4 nlmax, r_8 cos_theta_cut, int iterationOrder) const | 
|---|
| 1331 |  | 
|---|
| 1332 | \return power spectrum from analysis of a temperature map. THE MAP CAN BE MODIFIED (if iterationOrder >0) | 
|---|
| 1333 |  | 
|---|
| 1334 | \param<nlmax> : maximum value of the l index | 
|---|
| 1335 |  | 
|---|
| 1336 | \param<cos_theta_cut> : cosinus of the symmetric cut EULER angle theta : cos_theta_cut=0 means no cut ; cos_theta_cut=1 all the sphere is cut. | 
|---|
| 1337 |  | 
|---|
| 1338 | \param<iterationOrder> : 1,2,3,4.... order of an iterative analysis. If iterationOrder is not null, the method works with SphereHEALPix but NOT WITH SphereThetaPhi maps ! | 
|---|
| 1339 |  | 
|---|
| 1340 | */ | 
|---|
| 1341 | template <class T> | 
|---|
| 1342 | TVector<T>  SphericalTransformServer<T>::DecomposeToCl(SphericalMap<T>& sph, int_4 nlmax, r_8 cos_theta_cut, int iterationOrder) const | 
|---|
| 1343 | { | 
|---|
| 1344 | Alm<T> alm; | 
|---|
| 1345 | DecomposeToAlm( sph, alm, nlmax, cos_theta_cut, iterationOrder); | 
|---|
| 1346 | // power spectrum | 
|---|
| 1347 | return  alm.powerSpectrum(); | 
|---|
| 1348 | } | 
|---|
| 1349 |  | 
|---|
| 1350 |  | 
|---|
| 1351 | /*! \fn TVector<T>  SOPHYA::SphericalTransformServer::DecomposeToCl(const SphericalMap<T>& sph, int_4 nlmax, r_8 cos_theta_cut) const | 
|---|
| 1352 |  | 
|---|
| 1353 | \return power spectrum from analysis of a temperature map. | 
|---|
| 1354 |  | 
|---|
| 1355 | \param<nlmax> : maximum value of the l index | 
|---|
| 1356 |  | 
|---|
| 1357 | \param<cos_theta_cut> : cosinus of the symmetric cut EULER angle theta : cos_theta_cut=0 means no cut ; cos_theta_cut=1 all the sphere is cut. | 
|---|
| 1358 |  | 
|---|
| 1359 |  | 
|---|
| 1360 | */ | 
|---|
| 1361 |  | 
|---|
| 1362 |  | 
|---|
| 1363 | template <class T> | 
|---|
| 1364 | TVector<T>  SphericalTransformServer<T>::DecomposeToCl(const SphericalMap<T>& sph, int_4 nlmax, r_8 cos_theta_cut) const | 
|---|
| 1365 | { | 
|---|
| 1366 | Alm<T> alm; | 
|---|
| 1367 | DecomposeToAlm( sph, alm, nlmax, cos_theta_cut); | 
|---|
| 1368 | // power spectrum | 
|---|
| 1369 | return  alm.powerSpectrum(); | 
|---|
| 1370 | } | 
|---|
| 1371 |  | 
|---|
| 1372 | #ifdef __CXX_PRAGMA_TEMPLATES__ | 
|---|
| 1373 | #pragma define_template SphericalTransformServer<r_8> | 
|---|
| 1374 | #pragma define_template SphericalTransformServer<r_4> | 
|---|
| 1375 | #endif | 
|---|
| 1376 | #if defined(ANSI_TEMPLATES) || defined(GNU_TEMPLATES) | 
|---|
| 1377 | template class SphericalTransformServer<r_8>; | 
|---|
| 1378 | template class SphericalTransformServer<r_4>; | 
|---|
| 1379 | #endif | 
|---|