[2615] | 1 | #include "sopnamsp.h"
|
---|
[729] | 2 | #include "machdefs.h"
|
---|
[2322] | 3 | #include <iostream>
|
---|
[729] | 4 | #include <math.h>
|
---|
| 5 | #include <complex>
|
---|
| 6 | #include "sphericaltransformserver.h"
|
---|
| 7 | #include "tvector.h"
|
---|
[3510] | 8 | #include "stsrand.h"
|
---|
[729] | 9 | #include "nbmath.h"
|
---|
[1683] | 10 | #include "timing.h"
|
---|
| 11 | //#include "spherehealpix.h"
|
---|
[729] | 12 |
|
---|
[1683] | 13 |
|
---|
[2808] | 14 | /*!
|
---|
| 15 | \ingroup Samba
|
---|
| 16 | \class SOPHYA::SphericalTransformServer
|
---|
| 17 |
|
---|
| 18 | \brief Analysis/synthesis in spherical harmonics server.
|
---|
[729] | 19 |
|
---|
[1218] | 20 | Class for performing analysis and synthesis of sky maps using spin-0 or spin-2 spherical harmonics.
|
---|
| 21 |
|
---|
[2808] | 22 | Maps must be SOPHYA SphericalMaps (SphereHEALPix or SphereThetaPhi or SphereECP).
|
---|
[3508] | 23 | When generating map contents (synthesis), specify PixelSizeIndex=-1 if you want to keep
|
---|
| 24 | the map pixelisation scheme (resolution, layout ...)
|
---|
[1218] | 25 |
|
---|
| 26 | Temperature and polarization (Stokes parameters) can be developped on spherical harmonics :
|
---|
| 27 | \f[
|
---|
| 28 | \frac{\Delta T}{T}(\hat{n})=\sum_{lm}a_{lm}^TY_l^m(\hat{n})
|
---|
| 29 | \f]
|
---|
| 30 | \f[
|
---|
| 31 | 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)
|
---|
| 32 | \f]
|
---|
| 33 | \f[
|
---|
| 34 | 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)
|
---|
| 35 | \f]
|
---|
| 36 | \f[
|
---|
| 37 | \left(Q \pm iU\right)(\hat{n})=\sum_{lm}a_{\pm 2lm}\, _{\pm 2}Y_l^m(\hat{n})
|
---|
| 38 | \f]
|
---|
| 39 |
|
---|
| 40 | \f[
|
---|
| 41 | Y_l^m(\hat{n})=\lambda_l^m(\theta)e^{im\phi}
|
---|
| 42 | \f]
|
---|
| 43 | \f[
|
---|
| 44 | _{\pm}Y_l^m(\hat{n})=_{\pm}\lambda_l^m(\theta)e^{im\phi}
|
---|
| 45 | \f]
|
---|
| 46 | \f[
|
---|
| 47 | W_{lm}(\hat{n})=\frac{1}{N_l}\,_{w}\lambda_l^m(\theta)e^{im\phi}
|
---|
| 48 | \f]
|
---|
| 49 | \f[
|
---|
| 50 | X_{lm}(\hat{n})=\frac{-i}{N_l}\,_{x}\lambda_l^m(\theta)e^{im\phi}
|
---|
| 51 | \f]
|
---|
| 52 |
|
---|
| 53 | (see LambdaLMBuilder, LambdaPMBuilder, LambdaWXBuilder classes)
|
---|
| 54 |
|
---|
| 55 | power spectra :
|
---|
| 56 |
|
---|
| 57 | \f[
|
---|
| 58 | 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
|
---|
| 59 | \f]
|
---|
| 60 | \f[
|
---|
| 61 | 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
|
---|
| 62 | \f]
|
---|
| 63 | \f[
|
---|
| 64 | 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
|
---|
| 65 | \f]
|
---|
| 66 |
|
---|
| 67 | \arg
|
---|
| 68 | \b Synthesis : Get temperature and polarization maps from \f$a_{lm}\f$ coefficients or from power spectra, (methods GenerateFrom...).
|
---|
| 69 |
|
---|
| 70 | \b Temperature:
|
---|
| 71 | \f[
|
---|
| 72 | \frac{\Delta T}{T}(\hat{n})=\sum_{lm}a_{lm}^TY_l^m(\hat{n}) = \sum_{-\infty}^{+\infty}b_m(\theta)e^{im\phi}
|
---|
| 73 | \f]
|
---|
| 74 |
|
---|
| 75 | with
|
---|
| 76 | \f[
|
---|
| 77 | b_m(\theta)=\sum_{l=\left|m\right|}^{+\infty}a_{lm}^T\lambda_l^m(\theta)
|
---|
| 78 | \f]
|
---|
| 79 |
|
---|
| 80 | \b Polarisation
|
---|
| 81 | \f[
|
---|
| 82 | Q \pm iU = \sum_{-\infty}^{+\infty}b_m^{\pm}(\theta)e^{im\phi}
|
---|
| 83 | \f]
|
---|
| 84 |
|
---|
| 85 | where :
|
---|
| 86 | \f[
|
---|
| 87 | b_m^{\pm}(\theta) = \sum_{l=\left|m\right|}^{+\infty}a_{\pm 2lm}\,_{\pm}\lambda_l^m(\theta)
|
---|
| 88 | \f]
|
---|
| 89 |
|
---|
| 90 | or :
|
---|
| 91 | \f[
|
---|
| 92 | Q = \sum_{-\infty}^{+\infty}b_m^{Q}(\theta)e^{im\phi}
|
---|
| 93 | \f]
|
---|
| 94 | \f[
|
---|
| 95 | U = \sum_{-\infty}^{+\infty}b_m^{U}(\theta)e^{im\phi}
|
---|
| 96 | \f]
|
---|
| 97 |
|
---|
| 98 | where:
|
---|
| 99 | \f[
|
---|
| 100 | 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)
|
---|
| 101 | \f]
|
---|
| 102 | \f[
|
---|
| 103 | 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)
|
---|
| 104 | \f]
|
---|
| 105 |
|
---|
| 106 | 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.
|
---|
| 107 |
|
---|
| 108 |
|
---|
| 109 | \arg
|
---|
| 110 | \b Analysis : Get \f$a_{lm}\f$ coefficients or power spectra from temperature and polarization maps (methods DecomposeTo...).
|
---|
| 111 |
|
---|
| 112 | \b Temperature:
|
---|
| 113 | \f[
|
---|
| 114 | a_{lm}^T=\int\frac{\Delta T}{T}(\hat{n})Y_l^{m*}(\hat{n})d\hat{n}
|
---|
| 115 | \f]
|
---|
| 116 |
|
---|
| 117 | approximated as :
|
---|
| 118 | \f[
|
---|
| 119 | a_{lm}^T=\sum_{\theta_k}\omega_kC_m(\theta_k)\lambda_l^m(\theta_k)
|
---|
| 120 | \f]
|
---|
| 121 | where :
|
---|
| 122 | \f[
|
---|
| 123 | C_m (\theta _k)=\sum_{\phi _{k\prime}}\frac{\Delta T}{T}(\theta _k,\phi_{k\prime})e^{-im\phi _{k\prime}}
|
---|
| 124 | \f]
|
---|
| 125 | 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.
|
---|
| 126 |
|
---|
| 127 | \b polarisation:
|
---|
| 128 |
|
---|
| 129 | \f[
|
---|
| 130 | a_{\pm 2lm}=\sum_{\theta_k}\omega_kC_m^{\pm}(\theta_k)\,_{\pm}\lambda_l^m(\theta_k)
|
---|
| 131 | \f]
|
---|
| 132 | where :
|
---|
| 133 | \f[
|
---|
| 134 | C_m^{\pm} (\theta _k)=\sum_{\phi _{k\prime}}\left(Q \pm iU\right)(\theta _k,\phi_{k\prime})e^{-im\phi _{k\prime}}
|
---|
| 135 | \f]
|
---|
| 136 | or :
|
---|
| 137 |
|
---|
| 138 | \f[
|
---|
| 139 | 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)
|
---|
| 140 | \f]
|
---|
| 141 | \f[
|
---|
| 142 | 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)
|
---|
| 143 | \f]
|
---|
| 144 |
|
---|
| 145 | where :
|
---|
| 146 | \f[
|
---|
| 147 | C_m^{Q} (\theta _k)=\sum_{\phi _{k\prime}}Q(\theta _k,\phi_{k\prime})e^{-im\phi _{k\prime}}
|
---|
| 148 | \f]
|
---|
| 149 | \f[
|
---|
| 150 | C_m^{U} (\theta _k)=\sum_{\phi _{k\prime}}U(\theta _k,\phi_{k\prime})e^{-im\phi _{k\prime}}
|
---|
| 151 | \f]
|
---|
| 152 |
|
---|
| 153 | */
|
---|
| 154 |
|
---|
[3510] | 155 | //! Default constructor - Creates a non thread-safe RandomGenerator to be used by GenerateFromCl
|
---|
| 156 | template<class T>
|
---|
| 157 | SphericalTransformServer<T>::SphericalTransformServer()
|
---|
| 158 | : rg_(1, false)
|
---|
| 159 | {
|
---|
| 160 | fftIntfPtr_=new FFTPackServer(true); // preserveinput = true
|
---|
| 161 | fftIntfPtr_->setNormalize(false);
|
---|
| 162 | }
|
---|
| 163 |
|
---|
| 164 | //! Constructor with the specification of a RandomGenerator object to be used by GenerateFromCl
|
---|
| 165 | template<class T>
|
---|
| 166 | SphericalTransformServer<T>::SphericalTransformServer(RandomGenerator const & rg)
|
---|
| 167 | : rg_(rg)
|
---|
| 168 | {
|
---|
| 169 | fftIntfPtr_=new FFTPackServer(true); // preserveinput = true
|
---|
| 170 | fftIntfPtr_->setNormalize(false);
|
---|
| 171 | }
|
---|
| 172 |
|
---|
| 173 | template<class T>
|
---|
| 174 | SphericalTransformServer<T>::~SphericalTransformServer()
|
---|
| 175 | {
|
---|
| 176 | if (fftIntfPtr_!=NULL) delete fftIntfPtr_;
|
---|
| 177 | }
|
---|
| 178 |
|
---|
| 179 | /*!
|
---|
| 180 | Set a fft server. The constructor sets a default fft server (fft-pack).
|
---|
| 181 | So it is not necessary to call this method for a standard use.
|
---|
| 182 | \warning The FFTServerInterface object should NOT overwrite the input arrays
|
---|
| 183 | */
|
---|
| 184 | template<class T>
|
---|
| 185 | void SphericalTransformServer<T>::SetFFTServer(FFTServerInterface* srv)
|
---|
| 186 | {
|
---|
| 187 | if (fftIntfPtr_!=NULL) delete fftIntfPtr_;
|
---|
| 188 | fftIntfPtr_=srv;
|
---|
| 189 | fftIntfPtr_->setNormalize(false);
|
---|
| 190 | }
|
---|
| 191 |
|
---|
| 192 |
|
---|
[1218] | 193 | /*! \fn void SOPHYA::SphericalTransformServer::GenerateFromAlm( SphericalMap<T>& map, int_4 pixelSizeIndex, const Alm<T>& alm) const
|
---|
| 194 |
|
---|
| 195 | synthesis of a temperature map from Alm coefficients
|
---|
| 196 | */
|
---|
[729] | 197 | template<class T>
|
---|
| 198 | void SphericalTransformServer<T>::GenerateFromAlm( SphericalMap<T>& map, int_4 pixelSizeIndex, const Alm<T>& alm) const
|
---|
| 199 | {
|
---|
| 200 | /*=======================================================================
|
---|
[1756] | 201 | computes a map from its alm for the HEALPIX pixelisation
|
---|
[729] | 202 | map(theta,phi) = sum_l_m a_lm Y_lm(theta,phi)
|
---|
| 203 | = sum_m {e^(i*m*phi) sum_l a_lm*lambda_lm(theta)}
|
---|
| 204 |
|
---|
| 205 | where Y_lm(theta,phi) = lambda(theta) * e^(i*m*phi)
|
---|
| 206 |
|
---|
| 207 | * the recurrence of Ylm is the standard one (cf Num Rec)
|
---|
| 208 | * the sum over m is done by FFT
|
---|
| 209 |
|
---|
| 210 | =======================================================================*/
|
---|
| 211 | int_4 nlmax=alm.Lmax();
|
---|
| 212 | int_4 nmmax=nlmax;
|
---|
| 213 | int_4 nsmax=0;
|
---|
[1756] | 214 | // le Resize est suppose mettre a zero
|
---|
[729] | 215 | map.Resize(pixelSizeIndex);
|
---|
[2291] | 216 | string sphere_type=map.TypeOfMap();
|
---|
[2984] | 217 | int premiereTranche = 0;
|
---|
| 218 | int derniereTranche = map.NbThetaSlices()-1;
|
---|
| 219 |
|
---|
[729] | 220 | Bm<complex<T> > b_m_theta(nmmax);
|
---|
| 221 |
|
---|
| 222 | // pour chaque tranche en theta
|
---|
[2991] | 223 | for (int_4 ith = premiereTranche; ith <= derniereTranche;ith++) {
|
---|
| 224 | int_4 nph;
|
---|
| 225 | r_8 phi0;
|
---|
| 226 | r_8 theta;
|
---|
| 227 | TVector<int_4> pixNumber;
|
---|
| 228 | TVector<T> datan;
|
---|
| 229 |
|
---|
| 230 | map.GetThetaSlice(ith,theta,phi0, pixNumber,datan);
|
---|
| 231 | nph = pixNumber.NElts();
|
---|
| 232 | if (nph < 2) continue; // On laisse tomber les tranches avec un point
|
---|
| 233 | // -----------------------------------------------------
|
---|
| 234 | // for each theta, and each m, computes
|
---|
| 235 | // b(m,theta) = sum_over_l>m (lambda_l_m(theta) * a_l_m)
|
---|
| 236 | // ------------------------------------------------------
|
---|
| 237 | // ===> Optimisation Reza, Mai 2006
|
---|
| 238 | /*--- Le bout de code suivant est remplace par l'appel a la nouvelle fonction
|
---|
| 239 | qui calcule la somme au vol
|
---|
[729] | 240 | LambdaLMBuilder lb(theta,nlmax,nmmax);
|
---|
| 241 | // somme sur m de 0 a l'infini
|
---|
[2991] | 242 | for (int_4 m = 0; m <= nmmax; m++) {
|
---|
| 243 | b_m_theta(m) = (T)( lb.lamlm(m,m) ) * alm(m,m);
|
---|
| 244 | for (int l = m+1; l<= nlmax; l++)
|
---|
| 245 | b_m_theta(m) += (T)( lb.lamlm(l,m) ) * alm(l,m);
|
---|
| 246 | }
|
---|
[2958] | 247 | ------- Fin version PRE-Mai2006 */
|
---|
[2991] | 248 | LambdaLMBuilder::ComputeBmFrAlm(theta,nlmax,nmmax, alm, b_m_theta);
|
---|
| 249 | //Fin Optimisation Reza, Mai 2006 <====
|
---|
[2958] | 250 |
|
---|
[729] | 251 | // obtains the negative m of b(m,theta) (= complex conjugate)
|
---|
[2991] | 252 | for (int_4 m=1;m<=nmmax;m++)
|
---|
| 253 | b_m_theta(-m) = conj(b_m_theta(m));
|
---|
| 254 | // ---------------------------------------------------------------
|
---|
| 255 | // sum_m b(m,theta)*exp(i*m*phi) -> f(phi,theta)
|
---|
| 256 | // ---------------------------------------------------------------*/
|
---|
[729] | 257 |
|
---|
[2991] | 258 | /* ----- Reza, Juin 2006 :
|
---|
| 259 | En verifiant la difference entre deux cartes
|
---|
| 260 | cl -> map -> alm -> map2 et mapdiff = map-map2
|
---|
| 261 | je me suis apercu qu'il y avait des differences importantes - dans les
|
---|
| 262 | deux zones 'polar cap' de HEALPix - qui utilisait RfourierSynthesisFromB
|
---|
| 263 | TF complex -> reel . Le probleme venant de l'ambiguite de taille, lie
|
---|
| 264 | a la partie imaginaire de la composante a f_nyquist , j'ai corrige et
|
---|
| 265 | tout mis en TF complexe -> reel
|
---|
| 266 | */
|
---|
| 267 | TVector<T> Temp = RfourierSynthesisFromB(b_m_theta,nph,phi0);
|
---|
| 268 | // Si on peut acceder directement les pixels d'un tranche, on le fait
|
---|
| 269 | T* pix = map.GetThetaSliceDataPtr(ith);
|
---|
| 270 | if (pix != NULL)
|
---|
| 271 | for (int_4 i=0;i< nph;i++) pix[i] = Temp(i);
|
---|
| 272 | else
|
---|
| 273 | for (int_4 i=0;i< nph;i++) map(pixNumber(i))=Temp(i);
|
---|
| 274 | }
|
---|
[729] | 275 | }
|
---|
| 276 |
|
---|
| 277 |
|
---|
| 278 |
|
---|
[1218] | 279 | /*! \fn TVector< complex<T> > SOPHYA::SphericalTransformServer::fourierSynthesisFromB(const Bm<complex<T> >& b_m, int_4 nph, r_8 phi0) const
|
---|
| 280 |
|
---|
| 281 | \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)
|
---|
| 282 |
|
---|
| 283 | The object b_m (\f$b_m\f$) of the class Bm is a special vector which index goes from -mmax to mmax.
|
---|
| 284 | */
|
---|
[729] | 285 | template<class T>
|
---|
| 286 | TVector< complex<T> > SphericalTransformServer<T>::fourierSynthesisFromB(const Bm<complex<T> >& b_m, int_4 nph, r_8 phi0) const
|
---|
| 287 | {
|
---|
| 288 | /*=======================================================================
|
---|
| 289 | dataout(j) = sum_m datain(m) * exp(i*m*phi(j))
|
---|
| 290 | with phi(j) = j*2pi/nph + kphi0*pi/nph and kphi0 =0 or 1
|
---|
| 291 |
|
---|
| 292 | as the set of frequencies {m} is larger than nph,
|
---|
| 293 | we wrap frequencies within {0..nph-1}
|
---|
| 294 | ie m = k*nph + m' with m' in {0..nph-1}
|
---|
| 295 | then
|
---|
| 296 | noting bw(m') = exp(i*m'*phi0)
|
---|
| 297 | * sum_k (datain(k*nph+m') exp(i*k*pi*kphi0))
|
---|
| 298 | with bw(nph-m') = CONJ(bw(m')) (if datain(-m) = CONJ(datain(m)))
|
---|
| 299 | dataout(j) = sum_m' [ bw(m') exp (i*j*m'*2pi/nph) ]
|
---|
| 300 | = Fourier Transform of bw
|
---|
| 301 | is real
|
---|
| 302 |
|
---|
| 303 | NB nph is not necessarily a power of 2
|
---|
| 304 |
|
---|
| 305 | =======================================================================*/
|
---|
| 306 | //**********************************************************************
|
---|
| 307 | // pour une valeur de phi (indexee par j) la temperature est la transformee
|
---|
| 308 | // de Fourier de bm (somme sur m de -nmax a +nmmax de bm*exp(i*m*phi)).
|
---|
| 309 | // on demande nph (nombre de pixels sur la tranche) valeurs de transformees, pour nph valeurs de phi, regulierement reparties sur 2*pi. On a:
|
---|
| 310 | // DT/T(j) = sum_m b(m) * exp(i*m*phi(j))
|
---|
| 311 | // sommation de -infini a +infini, en fait limitee a -nmamx, +nmmax
|
---|
| 312 | // On pose m=k*nph + m', avec m' compris entre 0 et nph-1. Alors :
|
---|
| 313 | // DT/T(j) = somme_k somme_m' b(k*nph + m')*exp(i*(k*nph + m')*phi(j))
|
---|
| 314 | // somme_k : de -infini a +infini
|
---|
| 315 | // somme_m' : de 0 a nph-1
|
---|
| 316 | // On echange les sommations :
|
---|
[2625] | 317 | // DT/T(j) = somme_m' (exp(i*m'*phi(j)) somme_k b(k*nph + m')*exp(i*(k*nph*phi(j))
|
---|
[729] | 318 | // mais phi(j) est un multiple entier de 2*pi/nph, la seconde exponentielle
|
---|
| 319 | // vaut 1.
|
---|
| 320 | // Il reste a calculer les transformees de Fourier de somme_m' b(k*nph + m')
|
---|
| 321 | // si phi0 n'est pas nul, il y a juste un decalage a faire.
|
---|
| 322 | //**********************************************************************
|
---|
| 323 |
|
---|
| 324 | TVector< complex<T> > bw(nph);
|
---|
| 325 | TVector< complex<T> > dataout(nph);
|
---|
| 326 | TVector< complex<T> > data(nph);
|
---|
| 327 |
|
---|
| 328 |
|
---|
| 329 | for (int kk=0; kk<bw.NElts(); kk++) bw(kk)=(T)0.;
|
---|
[833] | 330 | int m;
|
---|
| 331 | for (m=-b_m.Mmax();m<=-1;m++)
|
---|
[729] | 332 | {
|
---|
| 333 | int maux=m;
|
---|
| 334 | while (maux<0) maux+=nph;
|
---|
| 335 | int iw=maux%nph;
|
---|
| 336 | double aux=(m-iw)*phi0;
|
---|
| 337 | bw(iw) += b_m(m) * complex<T>( (T)cos(aux),(T)sin(aux) ) ;
|
---|
| 338 | }
|
---|
[833] | 339 | for (m=0;m<=b_m.Mmax();m++)
|
---|
[729] | 340 | {
|
---|
| 341 | // int iw=((m % nph) +nph) % nph; //between 0 and nph = m'
|
---|
| 342 | int iw=m%nph;
|
---|
| 343 | double aux=(m-iw)*phi0;
|
---|
| 344 | bw(iw)+=b_m(m) * complex<T>( (T)cos(aux),(T)sin(aux) );
|
---|
| 345 | }
|
---|
| 346 |
|
---|
| 347 | // applies the shift in position <-> phase factor in Fourier space
|
---|
| 348 | for (int mprime=0; mprime < nph; mprime++)
|
---|
| 349 | {
|
---|
| 350 | complex<double> aux(cos(mprime*phi0),sin(mprime*phi0));
|
---|
| 351 | data(mprime)=bw(mprime)*
|
---|
| 352 | (complex<T>)(complex<double>(cos(mprime*phi0),sin(mprime*phi0)));
|
---|
| 353 | }
|
---|
| 354 |
|
---|
| 355 | //sortie.ReSize(nph);
|
---|
| 356 | TVector< complex<T> > sortie(nph);
|
---|
| 357 |
|
---|
| 358 | fftIntfPtr_-> FFTBackward(data, sortie);
|
---|
| 359 |
|
---|
| 360 | return sortie;
|
---|
| 361 | }
|
---|
| 362 |
|
---|
| 363 | //********************************************
|
---|
[1218] | 364 | /*! \fn TVector<T> SOPHYA::SphericalTransformServer::RfourierSynthesisFromB(const Bm<complex<T> >& b_m, int_4 nph, r_8 phi0) const
|
---|
| 365 |
|
---|
| 366 | same as fourierSynthesisFromB, but return a real vector, taking into account the fact that b(-m) is conjugate of b(m) */
|
---|
[729] | 367 | template<class T>
|
---|
| 368 | TVector<T> SphericalTransformServer<T>::RfourierSynthesisFromB(const Bm<complex<T> >& b_m, int_4 nph, r_8 phi0) const
|
---|
| 369 | {
|
---|
| 370 | /*=======================================================================
|
---|
| 371 | dataout(j) = sum_m datain(m) * exp(i*m*phi(j))
|
---|
| 372 | with phi(j) = j*2pi/nph + kphi0*pi/nph and kphi0 =0 or 1
|
---|
| 373 |
|
---|
| 374 | as the set of frequencies {m} is larger than nph,
|
---|
| 375 | we wrap frequencies within {0..nph-1}
|
---|
| 376 | ie m = k*nph + m' with m' in {0..nph-1}
|
---|
| 377 | then
|
---|
| 378 | noting bw(m') = exp(i*m'*phi0)
|
---|
| 379 | * sum_k (datain(k*nph+m') exp(i*k*pi*kphi0))
|
---|
| 380 | with bw(nph-m') = CONJ(bw(m')) (if datain(-m) = CONJ(datain(m)))
|
---|
| 381 | dataout(j) = sum_m' [ bw(m') exp (i*j*m'*2pi/nph) ]
|
---|
| 382 | = Fourier Transform of bw
|
---|
| 383 | is real
|
---|
| 384 |
|
---|
| 385 | NB nph is not necessarily a power of 2
|
---|
| 386 |
|
---|
| 387 | =======================================================================*/
|
---|
| 388 | //**********************************************************************
|
---|
| 389 | // pour une valeur de phi (indexee par j) la temperature est la transformee
|
---|
| 390 | // de Fourier de bm (somme sur m de -nmax a +nmmax de bm*exp(i*m*phi)).
|
---|
| 391 | // on demande nph (nombre de pixels sur la tranche) valeurs de transformees, pour nph valeurs de phi, regulierement reparties sur 2*pi. On a:
|
---|
| 392 | // DT/T(j) = sum_m b(m) * exp(i*m*phi(j))
|
---|
| 393 | // sommation de -infini a +infini, en fait limitee a -nmamx, +nmmax
|
---|
| 394 | // On pose m=k*nph + m', avec m' compris entre 0 et nph-1. Alors :
|
---|
| 395 | // DT/T(j) = somme_k somme_m' b(k*nph + m')*exp(i*(k*nph + m')*phi(j))
|
---|
| 396 | // somme_k : de -infini a +infini
|
---|
| 397 | // somme_m' : de 0 a nph-1
|
---|
| 398 | // On echange les sommations :
|
---|
[2313] | 399 | // DT/T(j) = somme_m' (exp(i*m'*phi(j)) somme_k b(k*nph + m')*exp(i*(k*nph*phi(j))
|
---|
[729] | 400 | // mais phi(j) est un multiple entier de 2*pi/nph, la seconde exponentielle
|
---|
| 401 | // vaut 1.
|
---|
[2313] | 402 | // Il reste a calculer les transformees de Fourier de somme_k b(k*nph + m')
|
---|
[729] | 403 | // si phi0 n'est pas nul, il y a juste un decalage a faire.
|
---|
| 404 | //**********************************************************************
|
---|
| 405 | TVector< complex<T> > bw(nph);
|
---|
| 406 | TVector< complex<T> > data(nph/2+1);
|
---|
| 407 |
|
---|
| 408 | for (int kk=0; kk<bw.NElts(); kk++) bw(kk)=(T)0.;
|
---|
[833] | 409 | int m;
|
---|
[2991] | 410 | for (m=-b_m.Mmax();m<=-1;m++) {
|
---|
| 411 | int maux=m;
|
---|
| 412 | while (maux<0) maux+=nph;
|
---|
| 413 | int iw=maux%nph;
|
---|
| 414 | double aux=(m-iw)*phi0;
|
---|
| 415 | bw(iw) += b_m(m) * complex<T>( (T)cos(aux),(T)sin(aux) ) ;
|
---|
| 416 | }
|
---|
| 417 | for (m=0;m<=b_m.Mmax();m++) {
|
---|
| 418 | // int iw=((m % nph) +nph) % nph; //between 0 and nph = m'
|
---|
| 419 | int iw=m%nph;
|
---|
| 420 | double aux=(m-iw)*phi0;
|
---|
| 421 | bw(iw)+=b_m(m) * complex<T>( (T)cos(aux),(T)sin(aux) );
|
---|
| 422 | }
|
---|
[729] | 423 |
|
---|
| 424 | // applies the shift in position <-> phase factor in Fourier space
|
---|
[2991] | 425 | for (int mprime=0; mprime <= nph/2; mprime++)
|
---|
| 426 | data(mprime)=bw(mprime)*complex<T>((T)cos(mprime*phi0),(T)sin(mprime*phi0));
|
---|
| 427 | TVector<T> sortie(nph);
|
---|
| 428 | // On met la partie imaginaire du dernier element du data a zero pour nph pair
|
---|
| 429 | if (nph%2 == 0) data(nph/2) = complex<T>(data(nph/2).real(), (T)0.);
|
---|
| 430 | // et on impose l'utilisation de la taille en sortie pour FFTBack (..., ..., true)
|
---|
| 431 | fftIntfPtr_-> FFTBackward(data, sortie, true);
|
---|
[729] | 432 | return sortie;
|
---|
| 433 | }
|
---|
| 434 | //*******************************************
|
---|
| 435 |
|
---|
[1218] | 436 | /*! \fn Alm<T> SOPHYA::SphericalTransformServer::DecomposeToAlm(const SphericalMap<T>& map, int_4 nlmax, r_8 cos_theta_cut) const
|
---|
| 437 |
|
---|
[1756] | 438 | \return the Alm coefficients from analysis of a temperature map.
|
---|
[1218] | 439 |
|
---|
| 440 | \param<nlmax> : maximum value of the l index
|
---|
| 441 |
|
---|
| 442 | \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.
|
---|
[1683] | 443 |
|
---|
[1756] | 444 | */
|
---|
[729] | 445 | template<class T>
|
---|
[1756] | 446 | void SphericalTransformServer<T>::DecomposeToAlm(const SphericalMap<T>& map, Alm<T>& alm, int_4 nlmax, r_8 cos_theta_cut) const
|
---|
| 447 | {
|
---|
| 448 | DecomposeToAlm(const_cast< SphericalMap<T>& >(map), alm, nlmax, cos_theta_cut, 0);
|
---|
| 449 | }
|
---|
| 450 | //*******************************************
|
---|
| 451 |
|
---|
| 452 | /*! \fn Alm<T> SOPHYA::SphericalTransformServer::DecomposeToAlm(const SphericalMap<T>& map, int_4 nlmax, r_8 cos_theta_cut, int iterationOrder) const
|
---|
| 453 |
|
---|
| 454 | \return the Alm coefficients from analysis of a temperature map. THE MAP CAN BE MODIFIED (if iterationOrder >0)
|
---|
| 455 |
|
---|
| 456 | \param<nlmax> : maximum value of the l index
|
---|
| 457 |
|
---|
| 458 | \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.
|
---|
| 459 |
|
---|
| 460 | \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 ! */
|
---|
| 461 | template<class T>
|
---|
[1683] | 462 | void SphericalTransformServer<T>::DecomposeToAlm(SphericalMap<T>& map, Alm<T>& alm, int_4 nlmax, r_8 cos_theta_cut, int iterationOrder) const
|
---|
[729] | 463 | {
|
---|
[1683] | 464 | int_4 nmmax = nlmax;
|
---|
| 465 | // PrtTim("appel carteVersAlm");
|
---|
| 466 | carteVersAlm(map, nlmax, cos_theta_cut, alm);
|
---|
| 467 | // PrtTim("retour carteVersAlm");
|
---|
| 468 | if (iterationOrder > 0)
|
---|
| 469 | {
|
---|
| 470 | TVector<int_4> fact(iterationOrder+2);
|
---|
| 471 | fact(0) = 1;
|
---|
[1715] | 472 | int k;
|
---|
| 473 | for (k=1; k <= iterationOrder+1; k++)
|
---|
[1683] | 474 | {
|
---|
| 475 | fact(k) = fact(k-1)*k;
|
---|
| 476 | }
|
---|
| 477 | Alm<T> alm2(alm);
|
---|
| 478 | T Tzero = (T)0.;
|
---|
| 479 | complex<T> complexZero = complex<T>(Tzero, Tzero);
|
---|
| 480 | alm = complexZero;
|
---|
| 481 | int signe = 1;
|
---|
| 482 | int nbIteration = iterationOrder+1;
|
---|
[1715] | 483 | for (k=1; k <= nbIteration; k++)
|
---|
[1683] | 484 | {
|
---|
| 485 | T facMult = (T)(0.5*signe*fact(iterationOrder)*(2*nbIteration-k)/(fact(k)*fact(nbIteration-k)));
|
---|
| 486 | for (int m = 0; m <= nmmax; m++)
|
---|
| 487 | {
|
---|
| 488 | for (int l = m; l<= nlmax; l++)
|
---|
| 489 | {
|
---|
| 490 | alm(l,m) += facMult*alm2(l,m);
|
---|
| 491 | }
|
---|
| 492 | }
|
---|
| 493 | if (k == nbIteration) break;
|
---|
| 494 | signe = -signe;
|
---|
| 495 | for (int k=0; k< map.NbPixels(); k++) map(k) = (T)0.;
|
---|
| 496 | // synthetize a map from the estimated alm
|
---|
| 497 | // PrtTim("appel GenerateFromAlm");
|
---|
| 498 | GenerateFromAlm( map, map.SizeIndex(), alm2);
|
---|
| 499 | // PrtTim("retour GenerateFromAlm");
|
---|
| 500 | alm2 = complexZero;
|
---|
| 501 | // analyse the new map
|
---|
| 502 | // PrtTim("appel carteVersAlm");
|
---|
| 503 | carteVersAlm(map, nlmax, cos_theta_cut, alm2);
|
---|
| 504 | // PrtTim("retour carteVersAlm");
|
---|
| 505 | }
|
---|
| 506 | }
|
---|
| 507 | }
|
---|
| 508 |
|
---|
| 509 | template<class T>
|
---|
| 510 | void SphericalTransformServer<T>::carteVersAlm(const SphericalMap<T>& map, int_4 nlmax, r_8 cos_theta_cut, Alm<T>& alm) const
|
---|
| 511 | {
|
---|
[729] | 512 |
|
---|
| 513 | /*-----------------------------------------------------------------------
|
---|
| 514 | computes the integral in phi : phas_m(theta)
|
---|
| 515 | for each parallele from north to south pole
|
---|
| 516 | -----------------------------------------------------------------------*/
|
---|
| 517 | TVector<T> data;
|
---|
| 518 | TVector<int_4> pixNumber;
|
---|
| 519 | int_4 nmmax = nlmax;
|
---|
| 520 | TVector< complex<T> > phase(nmmax+1);
|
---|
[1683] | 521 |
|
---|
[729] | 522 | alm.ReSizeToLmax(nlmax);
|
---|
[746] | 523 | for (int_4 ith = 0; ith < map.NbThetaSlices(); ith++)
|
---|
[729] | 524 | {
|
---|
| 525 | r_8 phi0;
|
---|
| 526 | r_8 theta;
|
---|
[1683] | 527 | // PrtTim("debut 1ere tranche ");
|
---|
[729] | 528 | map.GetThetaSlice(ith,theta,phi0,pixNumber ,data);
|
---|
[1683] | 529 | phase = complex<T>((T)0.,(T)0.);
|
---|
[729] | 530 | double cth = cos(theta);
|
---|
| 531 |
|
---|
| 532 | //part of the sky out of the symetric cut
|
---|
[1428] | 533 | bool keep_it = (fabs(cth) >= cos_theta_cut);
|
---|
[1683] | 534 |
|
---|
| 535 | // PrtTim("fin 1ere tranche ");
|
---|
| 536 |
|
---|
[729] | 537 | if (keep_it)
|
---|
| 538 | {
|
---|
[1683] | 539 | // phase = CFromFourierAnalysis(nmmax,data,phi0);
|
---|
| 540 | // PrtTim("avant Fourier ");
|
---|
| 541 | CFromFourierAnalysis(nmmax,data,phase, phi0);
|
---|
| 542 | // PrtTim("apres Fourier ");
|
---|
[729] | 543 |
|
---|
| 544 | }
|
---|
| 545 |
|
---|
[1683] | 546 | // ---------------------------------------------------------------------
|
---|
| 547 | // computes the a_lm by integrating over theta
|
---|
| 548 | // lambda_lm(theta) * phas_m(theta)
|
---|
| 549 | // for each m and l
|
---|
| 550 | // -----------------------------------------------------------------------
|
---|
[2958] | 551 |
|
---|
| 552 | // ===> Optimisation Reza, Mai 2006
|
---|
| 553 | /*--- Le bout de code suivant est remplace par l'appel a la nouvelle fonction
|
---|
| 554 | qui calcule la somme au vol
|
---|
[1683] | 555 | // PrtTim("avant instanciation LM ");
|
---|
[729] | 556 | LambdaLMBuilder lb(theta,nlmax,nmmax);
|
---|
[1683] | 557 | // PrtTim("apres instanciation LM ");
|
---|
[729] | 558 | r_8 domega=map.PixSolAngle(map.PixIndexSph(theta,phi0));
|
---|
[1683] | 559 |
|
---|
| 560 | // PrtTim("avant mise a jour Alm ");
|
---|
| 561 | complex<T> fi;
|
---|
| 562 | T facteur;
|
---|
| 563 | int index;
|
---|
[729] | 564 | for (int m = 0; m <= nmmax; m++)
|
---|
| 565 | {
|
---|
[1683] | 566 | fi = phase(m);
|
---|
| 567 | for (int l = m; l<= nlmax; l++)
|
---|
[729] | 568 | {
|
---|
[1683] | 569 | index = alm.indexOfElement(l,m);
|
---|
| 570 | // facteur = (T)(lb.lamlm(l,m) * domega);
|
---|
| 571 | facteur = (T)(lb.lamlm(index) * domega);
|
---|
| 572 | // alm(l,m) += facteur * fi ;
|
---|
| 573 | alm(index) += facteur * fi ;
|
---|
[729] | 574 | }
|
---|
| 575 | }
|
---|
[2958] | 576 | ------- Fin version PRE-Mai2006 */
|
---|
| 577 | r_8 domega=map.PixSolAngle(map.PixIndexSph(theta,phi0));
|
---|
| 578 | phase *= complex<T>((T)domega, 0.);
|
---|
| 579 | LambdaLMBuilder::ComputeAlmFrPhase(theta,nlmax,nmmax, phase, alm);
|
---|
| 580 | //Fin Optimisation Reza, Mai 2006 <====
|
---|
[1683] | 581 |
|
---|
| 582 |
|
---|
| 583 |
|
---|
| 584 | //
|
---|
| 585 | //
|
---|
| 586 | // PrtTim("apres mise a jour Alm ");
|
---|
[729] | 587 | }
|
---|
| 588 | }
|
---|
[1218] | 589 | /*! \fn TVector< complex<T> > SOPHYA::SphericalTransformServer::CFromFourierAnalysis(int_4 nmmax, const TVector<complex<T> >datain, r_8 phi0) const
|
---|
| 590 |
|
---|
| 591 | \return a vector with mmax elements which are sums :
|
---|
| 592 | \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.
|
---|
| 593 | */
|
---|
[729] | 594 | template<class T>
|
---|
[746] | 595 | TVector< complex<T> > SphericalTransformServer<T>::CFromFourierAnalysis(int_4 nmmax, const TVector<complex<T> >datain, r_8 phi0) const
|
---|
[729] | 596 | {
|
---|
| 597 | /*=======================================================================
|
---|
| 598 | integrates (data * phi-dependence-of-Ylm) over phi
|
---|
| 599 | --> function of m can be computed by FFT
|
---|
| 600 |
|
---|
| 601 | datain est modifie
|
---|
| 602 | =======================================================================*/
|
---|
| 603 | int_4 nph=datain.NElts();
|
---|
| 604 | if (nph <= 0)
|
---|
| 605 | {
|
---|
| 606 | throw PException("bizarre : vecteur datain de longueur nulle (CFromFourierAnalysis)");
|
---|
| 607 | }
|
---|
| 608 | TVector<complex<T> > transformedData(nph);
|
---|
[3003] | 609 | // Il faut avoir instancie le serveur de FFT avec l'option preserveinput=true
|
---|
| 610 | fftIntfPtr_-> FFTForward(const_cast<TVector< complex<T> > &>(datain), transformedData);
|
---|
[729] | 611 |
|
---|
| 612 | TVector< complex<T> > dataout(nmmax+1);
|
---|
| 613 |
|
---|
| 614 | int im_max=min(nph,nmmax+1);
|
---|
[833] | 615 | int i;
|
---|
[1683] | 616 | dataout = complex<T>((T)0.,(T)0.);
|
---|
| 617 | // for (i=0;i< dataout.NElts();i++) dataout(i)=complex<T>((T)0.,(T)0.);
|
---|
[833] | 618 | for (i=0;i<im_max;i++) dataout(i)=transformedData(i);
|
---|
[729] | 619 |
|
---|
| 620 |
|
---|
| 621 | for (int kk=nph; kk<dataout.NElts(); kk++) dataout(kk)=dataout(kk%nph);
|
---|
[833] | 622 | for (i = 0;i <dataout.NElts();i++){
|
---|
[729] | 623 | dataout(i)*= (complex<T>)(complex<double>(cos(-i*phi0),sin(-i*phi0)));
|
---|
| 624 | }
|
---|
| 625 | return dataout;
|
---|
| 626 | }
|
---|
| 627 |
|
---|
| 628 | //&&&&&&&&& nouvelle version
|
---|
[1218] | 629 | /* \fn TVector< complex<T> > SOPHYA::SphericalTransformServer::CFromFourierAnalysis(int_4 nmmax, const TVector<T> datain, r_8 phi0) const
|
---|
| 630 |
|
---|
| 631 | same as previous one, but with a "datain" which is real (not complex) */
|
---|
[729] | 632 | template<class T>
|
---|
[1683] | 633 | void SphericalTransformServer<T>::CFromFourierAnalysis(int_4 nmmax, const TVector<T> datain, TVector< complex<T> >& dataout, r_8 phi0) const
|
---|
[729] | 634 | {
|
---|
| 635 | //=======================================================================
|
---|
| 636 | // integrates (data * phi-dependence-of-Ylm) over phi
|
---|
| 637 | // --> function of m can be computed by FFT
|
---|
| 638 | // ! with 0<= m <= npoints/2 (: Nyquist)
|
---|
| 639 | // ! because the data is real the negative m are the conjugate of the
|
---|
| 640 | // ! positive ones
|
---|
| 641 |
|
---|
| 642 | // datain est modifie
|
---|
| 643 | //
|
---|
| 644 | // =======================================================================
|
---|
| 645 | int_4 nph=datain.NElts();
|
---|
| 646 | if (nph <= 0)
|
---|
| 647 | {
|
---|
| 648 | throw PException("bizarre : vecteur datain de longueur nulle (CFromFourierAnalysis)");
|
---|
| 649 | }
|
---|
[1756] | 650 | // if (nph%2 != 0 )
|
---|
| 651 | // {
|
---|
| 652 | // throw PException("SphericalTransformServer<T>::CFromFourierAnalysis : longueur de datain impair ?");
|
---|
| 653 | // }
|
---|
[729] | 654 | TVector<complex<T> > transformedData;
|
---|
| 655 |
|
---|
[1683] | 656 | // la taille du vecteur complexe retourne est nph/2+1 (si la taille
|
---|
| 657 | // du vecteur reel entre est nph)
|
---|
[1756] | 658 | // cout << " longueur de datain = " << nph << endl;
|
---|
[3003] | 659 | // Il faut avoir instancie le serveur de FFT avec l'option preserveinput=true
|
---|
| 660 | fftIntfPtr_-> FFTForward(const_cast< TVector<T> &>(datain), transformedData);
|
---|
[1756] | 661 | // cout << " taille de la transformee " << transformedData.Size() << endl;
|
---|
[1683] | 662 | // TVector< complex<T> > dataout(nmmax+1);
|
---|
| 663 | dataout.ReSize(nmmax+1);
|
---|
[729] | 664 |
|
---|
| 665 | // on transfere le resultat de la fft dans dataout.
|
---|
[1683] | 666 |
|
---|
| 667 | int maxFreqAccessiblesParFFT = min(nph/2,nmmax);
|
---|
[833] | 668 | int i;
|
---|
[1683] | 669 | for (i=0;i<=maxFreqAccessiblesParFFT;i++) dataout(i)=transformedData(i);
|
---|
[729] | 670 |
|
---|
| 671 |
|
---|
[1683] | 672 | // si dataout n'est pas plein, on complete jusqu'a nph+1 valeurs (a moins
|
---|
[729] | 673 | // que dataout ne soit plein avant d'atteindre nph)
|
---|
[1683] | 674 | if (maxFreqAccessiblesParFFT != nmmax )
|
---|
[729] | 675 | {
|
---|
[1683] | 676 | int maxMfft = min(nph,nmmax);
|
---|
| 677 | for (i=maxFreqAccessiblesParFFT+1; i<=maxMfft; i++)
|
---|
[729] | 678 | {
|
---|
| 679 | dataout(i) = conj(dataout(nph-i) );
|
---|
| 680 | }
|
---|
| 681 | // on conplete, si necessaire, par periodicite
|
---|
[1683] | 682 | if ( maxMfft != nmmax )
|
---|
[729] | 683 | {
|
---|
[1683] | 684 | for (int kk=nph+1; kk <= nmmax; kk++)
|
---|
| 685 | {
|
---|
| 686 | dataout(kk)=dataout(kk%nph);
|
---|
| 687 | }
|
---|
[729] | 688 | }
|
---|
| 689 | }
|
---|
[1683] | 690 | for (i = 0;i <dataout.NElts();i++)
|
---|
| 691 | {
|
---|
| 692 | dataout(i)*= (complex<T>)(complex<double>(cos(-i*phi0),sin(-i*phi0)));
|
---|
| 693 | }
|
---|
| 694 | // return dataout;
|
---|
[729] | 695 | }
|
---|
| 696 |
|
---|
[1218] | 697 | /*! \fn void SOPHYA::SphericalTransformServer::GenerateFromAlm(SphericalMap<T>& mapq,
|
---|
| 698 | SphericalMap<T>& mapu,
|
---|
| 699 | int_4 pixelSizeIndex,
|
---|
| 700 | const Alm<T>& alme,
|
---|
| 701 | const Alm<T>& almb) const
|
---|
| 702 |
|
---|
| 703 | synthesis of a polarization map from Alm coefficients. The spheres mapq and mapu contain respectively the Stokes parameters. */
|
---|
[729] | 704 | template<class T>
|
---|
| 705 | void SphericalTransformServer<T>::GenerateFromAlm(SphericalMap<T>& mapq,
|
---|
| 706 | SphericalMap<T>& mapu,
|
---|
| 707 | int_4 pixelSizeIndex,
|
---|
| 708 | const Alm<T>& alme,
|
---|
| 709 | const Alm<T>& almb) const
|
---|
| 710 | {
|
---|
| 711 | /*=======================================================================
|
---|
| 712 | computes a map form its alm for the HEALPIX pixelisation
|
---|
| 713 | map(theta,phi) = sum_l_m a_lm Y_lm(theta,phi)
|
---|
| 714 | = sum_m {e^(i*m*phi) sum_l a_lm*lambda_lm(theta)}
|
---|
| 715 |
|
---|
| 716 | where Y_lm(theta,phi) = lambda(theta) * e^(i*m*phi)
|
---|
| 717 |
|
---|
| 718 | * the recurrence of Ylm is the standard one (cf Num Rec)
|
---|
| 719 | * the sum over m is done by FFT
|
---|
| 720 |
|
---|
| 721 | =======================================================================*/
|
---|
| 722 | int_4 nlmax=alme.Lmax();
|
---|
| 723 | if (nlmax != almb.Lmax())
|
---|
| 724 | {
|
---|
| 725 | cout << " SphericalTransformServer: les deux tableaux alm n'ont pas la meme taille" << endl;
|
---|
| 726 | throw SzMismatchError("SphericalTransformServer: les deux tableaux alm n'ont pas la meme taille");
|
---|
| 727 | }
|
---|
| 728 | int_4 nmmax=nlmax;
|
---|
| 729 | int_4 nsmax=0;
|
---|
| 730 | mapq.Resize(pixelSizeIndex);
|
---|
| 731 | mapu.Resize(pixelSizeIndex);
|
---|
[2291] | 732 | string sphere_type=mapq.TypeOfMap();
|
---|
| 733 | if (sphere_type != mapu.TypeOfMap())
|
---|
[729] | 734 | {
|
---|
| 735 | cout << " SphericalTransformServer: les deux spheres ne sont pas de meme type" << endl;
|
---|
| 736 | cout << " type 1 " << sphere_type << endl;
|
---|
| 737 | cout << " type 2 " << mapu.TypeOfMap() << endl;
|
---|
| 738 | throw SzMismatchError("SphericalTransformServer: les deux spheres ne sont pas de meme type");
|
---|
| 739 |
|
---|
| 740 | }
|
---|
[2313] | 741 | bool healpix = true;
|
---|
[2291] | 742 | if (sphere_type.substr(0,4) == "RING")
|
---|
[729] | 743 | {
|
---|
| 744 | nsmax=mapq.SizeIndex();
|
---|
| 745 | }
|
---|
| 746 | else
|
---|
| 747 | // pour une sphere Gorski le nombre de pixels est 12*nsmax**2
|
---|
| 748 | // on calcule une quantite equivalente a nsmax pour la sphere-theta-phi
|
---|
| 749 | // en vue de l'application du critere Healpix : nlmax<=3*nsmax-1
|
---|
| 750 | // c'est approximatif ; a raffiner.
|
---|
[2313] | 751 | healpix = false;
|
---|
[2291] | 752 | if (sphere_type.substr(0,6) == "TETAFI")
|
---|
[729] | 753 | {
|
---|
| 754 | nsmax=(int_4)sqrt(mapq.NbPixels()/12.);
|
---|
| 755 | }
|
---|
| 756 | else
|
---|
| 757 | {
|
---|
| 758 | cout << " unknown type of sphere : " << sphere_type << endl;
|
---|
| 759 | throw IOExc(" unknown type of sphere ");
|
---|
| 760 | }
|
---|
| 761 | cout << "GenerateFromAlm: the spheres are of type : " << sphere_type << endl;
|
---|
| 762 | cout << "GenerateFromAlm: size indices (nside) of spheres= " << nsmax << endl;
|
---|
| 763 | cout << "GenerateFromAlm: nlmax (from Alm) = " << nlmax << endl;
|
---|
| 764 | if (nlmax>3*nsmax-1)
|
---|
| 765 | {
|
---|
| 766 | cout << "GenerateFromAlm: nlmax should be <= 3*nside-1" << endl;
|
---|
[2291] | 767 | if (sphere_type.substr(0,6) == "TETAFI")
|
---|
[729] | 768 | {
|
---|
| 769 | cout << " (for this criterium, nsmax is computed as sqrt(nbPixels/12))" << endl;
|
---|
| 770 | }
|
---|
| 771 | }
|
---|
| 772 | if (alme.Lmax()!=almb.Lmax())
|
---|
| 773 | {
|
---|
| 774 | cout << "GenerateFromAlm: arrays Alme and Almb have not the same size ? " << endl;
|
---|
| 775 | throw SzMismatchError("SphericalTransformServer: arrays Alme and Almb have not the same size ? ");
|
---|
| 776 | }
|
---|
[2313] | 777 | mapFromWX(nlmax, nmmax, mapq, mapu, alme, almb, healpix);
|
---|
[729] | 778 | // mapFromPM(nlmax, nmmax, mapq, mapu, alme, almb);
|
---|
| 779 | }
|
---|
[1756] | 780 | /*! \fn void SOPHYA::SphericalTransformServer::DecomposeToAlm(const SphericalMap<T>& mapq,
|
---|
| 781 | const SphericalMap<T>& mapu,
|
---|
| 782 | Alm<T>& alme,
|
---|
| 783 | Alm<T>& almb,
|
---|
| 784 | int_4 nlmax,
|
---|
| 785 | r_8 cos_theta_cut) const
|
---|
[729] | 786 |
|
---|
[1756] | 787 | analysis of a polarization map into Alm coefficients.
|
---|
[729] | 788 |
|
---|
[1756] | 789 | The spheres \c mapq and \c mapu contain respectively the Stokes parameters.
|
---|
| 790 |
|
---|
| 791 | \c a2lme and \c a2lmb will receive respectively electric and magnetic Alm's
|
---|
| 792 | nlmax : maximum value of the l index
|
---|
| 793 |
|
---|
| 794 | \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.
|
---|
| 795 |
|
---|
| 796 |
|
---|
| 797 | */
|
---|
| 798 | template<class T>
|
---|
| 799 | void SphericalTransformServer<T>::DecomposeToAlm(const SphericalMap<T>& mapq,
|
---|
[1218] | 800 | const SphericalMap<T>& mapu,
|
---|
| 801 | Alm<T>& alme,
|
---|
| 802 | Alm<T>& almb,
|
---|
| 803 | int_4 nlmax,
|
---|
| 804 | r_8 cos_theta_cut) const
|
---|
[1756] | 805 | {
|
---|
| 806 | DecomposeToAlm(const_cast< SphericalMap<T>& >(mapq), const_cast< SphericalMap<T>& >(mapu), alme, almb, nlmax, cos_theta_cut);
|
---|
| 807 | }
|
---|
[1218] | 808 |
|
---|
[1756] | 809 | /*! \fn void SOPHYA::SphericalTransformServer::DecomposeToAlm(const SphericalMap<T>& mapq,
|
---|
| 810 | const SphericalMap<T>& mapu,
|
---|
| 811 | Alm<T>& alme,
|
---|
| 812 | Alm<T>& almb,
|
---|
| 813 | int_4 nlmax,
|
---|
| 814 | r_8 cos_theta_cut,
|
---|
| 815 | int iterationOrder) const
|
---|
| 816 |
|
---|
[1218] | 817 | analysis of a polarization map into Alm coefficients.
|
---|
| 818 |
|
---|
| 819 | The spheres \c mapq and \c mapu contain respectively the Stokes parameters.
|
---|
| 820 |
|
---|
| 821 | \c a2lme and \c a2lmb will receive respectively electric and magnetic Alm's
|
---|
| 822 | nlmax : maximum value of the l index
|
---|
| 823 |
|
---|
| 824 | \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.
|
---|
[1756] | 825 |
|
---|
| 826 | \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 !
|
---|
| 827 |
|
---|
| 828 | THE INPUT MAPS CAN BE MODIFIED (only if iterationOrder >0)
|
---|
| 829 |
|
---|
[1218] | 830 | */
|
---|
[729] | 831 | template<class T>
|
---|
[1683] | 832 | void SphericalTransformServer<T>::DecomposeToAlm(SphericalMap<T>& mapq,
|
---|
| 833 | SphericalMap<T>& mapu,
|
---|
| 834 | Alm<T>& alme,
|
---|
| 835 | Alm<T>& almb,
|
---|
| 836 | int_4 nlmax,
|
---|
| 837 | r_8 cos_theta_cut,
|
---|
| 838 | int iterationOrder) const
|
---|
| 839 | {
|
---|
| 840 | int_4 nmmax = nlmax;
|
---|
| 841 | carteVersAlm(mapq, mapu, alme, almb, nlmax, cos_theta_cut);
|
---|
| 842 | if (iterationOrder > 0)
|
---|
| 843 | {
|
---|
| 844 | TVector<int_4> fact(iterationOrder+2);
|
---|
| 845 | fact(0) = 1;
|
---|
[1715] | 846 | int k;
|
---|
| 847 | for (k=1; k <= iterationOrder+1; k++)
|
---|
[1683] | 848 | {
|
---|
| 849 | fact(k) = fact(k-1)*k;
|
---|
| 850 | }
|
---|
| 851 | Alm<T> alme2(alme);
|
---|
| 852 | Alm<T> almb2(almb);
|
---|
| 853 | T Tzero = (T)0.;
|
---|
| 854 | complex<T> complexZero = complex<T>(Tzero, Tzero);
|
---|
| 855 | alme = complexZero;
|
---|
| 856 | almb = complexZero;
|
---|
| 857 | int signe = 1;
|
---|
| 858 | int nbIteration = iterationOrder+1;
|
---|
[1715] | 859 | for (k=1; k <= nbIteration; k++)
|
---|
[1683] | 860 | {
|
---|
| 861 | T facMult = (T)(0.5*signe*fact(iterationOrder)*(2*nbIteration-k)/(fact(k)*fact(nbIteration-k)));
|
---|
| 862 | for (int m = 0; m <= nmmax; m++)
|
---|
| 863 | {
|
---|
| 864 | for (int l = m; l<= nlmax; l++)
|
---|
| 865 | {
|
---|
| 866 | alme(l,m) += facMult*alme2(l,m);
|
---|
| 867 | almb(l,m) += facMult*almb2(l,m);
|
---|
| 868 | }
|
---|
| 869 | }
|
---|
| 870 | if (k == nbIteration) break;
|
---|
| 871 | signe = -signe;
|
---|
| 872 | for (int k=0; k< mapq.NbPixels(); k++)
|
---|
| 873 | {
|
---|
| 874 | mapq(k) = (T)0.;
|
---|
| 875 | mapu(k) = (T)0.;
|
---|
| 876 | }
|
---|
| 877 | // synthetize a map from the estimated alm
|
---|
| 878 | GenerateFromAlm(mapq,mapu,mapq.SizeIndex(),alme2,almb2);
|
---|
| 879 | alme2 = complexZero;
|
---|
| 880 | almb2 = complexZero;
|
---|
| 881 | // analyse the new map
|
---|
| 882 | carteVersAlm(mapq, mapu, alme2, almb2, nlmax, cos_theta_cut);
|
---|
| 883 | }
|
---|
| 884 | }
|
---|
| 885 | }
|
---|
| 886 |
|
---|
| 887 | template<class T>
|
---|
| 888 | void SphericalTransformServer<T>::carteVersAlm(const SphericalMap<T>& mapq,
|
---|
[729] | 889 | const SphericalMap<T>& mapu,
|
---|
| 890 | Alm<T>& alme,
|
---|
| 891 | Alm<T>& almb,
|
---|
| 892 | int_4 nlmax,
|
---|
| 893 | r_8 cos_theta_cut) const
|
---|
| 894 | {
|
---|
| 895 | int_4 nmmax = nlmax;
|
---|
| 896 | // resize et remise a zero
|
---|
| 897 | alme.ReSizeToLmax(nlmax);
|
---|
| 898 | almb.ReSizeToLmax(nlmax);
|
---|
| 899 |
|
---|
| 900 |
|
---|
| 901 | TVector<T> dataq;
|
---|
| 902 | TVector<T> datau;
|
---|
| 903 | TVector<int_4> pixNumber;
|
---|
| 904 |
|
---|
[2291] | 905 | string sphere_type=mapq.TypeOfMap();
|
---|
| 906 | if (sphere_type != mapu.TypeOfMap())
|
---|
[729] | 907 | {
|
---|
| 908 | cout << " SphericalTransformServer: les deux spheres ne sont pas de meme type" << endl;
|
---|
| 909 | cout << " type 1 " << sphere_type << endl;
|
---|
| 910 | cout << " type 2 " << mapu.TypeOfMap() << endl;
|
---|
| 911 | throw SzMismatchError("SphericalTransformServer: les deux spheres ne sont pas de meme type");
|
---|
| 912 |
|
---|
| 913 | }
|
---|
| 914 | if (mapq.NbPixels()!=mapu.NbPixels())
|
---|
| 915 | {
|
---|
| 916 | cout << " DecomposeToAlm: map Q and map U have not same size ?" << endl;
|
---|
| 917 | throw SzMismatchError("SphericalTransformServer::DecomposeToAlm: map Q and map U have not same size ");
|
---|
| 918 | }
|
---|
[746] | 919 | for (int_4 ith = 0; ith < mapq.NbThetaSlices(); ith++)
|
---|
[729] | 920 | {
|
---|
| 921 | r_8 phi0;
|
---|
| 922 | r_8 theta;
|
---|
| 923 | mapq.GetThetaSlice(ith,theta,phi0, pixNumber,dataq);
|
---|
| 924 | mapu.GetThetaSlice(ith,theta,phi0, pixNumber,datau);
|
---|
| 925 | if (dataq.NElts() != datau.NElts() )
|
---|
| 926 | {
|
---|
| 927 | throw SzMismatchError("the spheres have not the same pixelization");
|
---|
| 928 | }
|
---|
| 929 | r_8 domega=mapq.PixSolAngle(mapq.PixIndexSph(theta,phi0));
|
---|
| 930 | double cth = cos(theta);
|
---|
| 931 | //part of the sky out of the symetric cut
|
---|
[1428] | 932 | bool keep_it = (fabs(cth) >= cos_theta_cut);
|
---|
[729] | 933 | if (keep_it)
|
---|
| 934 | {
|
---|
[1328] | 935 | // almFromPM(pixNumber.NElts(), nlmax, nmmax, phi0, domega, theta, dataq, datau, alme, almb);
|
---|
[746] | 936 | almFromWX(nlmax, nmmax, phi0, domega, theta, dataq, datau, alme, almb);
|
---|
[729] | 937 | }
|
---|
| 938 | }
|
---|
| 939 | }
|
---|
| 940 |
|
---|
| 941 |
|
---|
[1218] | 942 | /*! \fn void SOPHYA::SphericalTransformServer::almFromWX(int_4 nlmax, int_4 nmmax,
|
---|
| 943 | r_8 phi0, r_8 domega,
|
---|
| 944 | r_8 theta,
|
---|
| 945 | const TVector<T>& dataq,
|
---|
| 946 | const TVector<T>& datau,
|
---|
| 947 | Alm<T>& alme,
|
---|
| 948 | Alm<T>& almb) const
|
---|
| 949 |
|
---|
| 950 | Compute polarized Alm's as :
|
---|
| 951 | \f[
|
---|
| 952 | 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)}
|
---|
| 953 | \f]
|
---|
| 954 | \f[
|
---|
| 955 | 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)}
|
---|
| 956 | \f]
|
---|
| 957 |
|
---|
| 958 | 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.
|
---|
| 959 |
|
---|
| 960 | \f$\omega_{pix}\f$ are solid angle of each pixel.
|
---|
| 961 |
|
---|
| 962 | dataq, datau : Stokes parameters.
|
---|
| 963 |
|
---|
| 964 | */
|
---|
[729] | 965 | template<class T>
|
---|
[746] | 966 | void SphericalTransformServer<T>::almFromWX(int_4 nlmax, int_4 nmmax,
|
---|
[729] | 967 | r_8 phi0, r_8 domega,
|
---|
| 968 | r_8 theta,
|
---|
| 969 | const TVector<T>& dataq,
|
---|
| 970 | const TVector<T>& datau,
|
---|
| 971 | Alm<T>& alme,
|
---|
| 972 | Alm<T>& almb) const
|
---|
| 973 | {
|
---|
| 974 | TVector< complex<T> > phaseq(nmmax+1);
|
---|
| 975 | TVector< complex<T> > phaseu(nmmax+1);
|
---|
| 976 | // TVector<complex<T> > datain(nph);
|
---|
| 977 | for (int i=0;i< nmmax+1;i++)
|
---|
| 978 | {
|
---|
| 979 | phaseq(i)=0;
|
---|
| 980 | phaseu(i)=0;
|
---|
| 981 | }
|
---|
| 982 | // for(int kk=0; kk<nph; kk++) datain(kk)=complex<T>(dataq(kk),0.);
|
---|
| 983 |
|
---|
[1683] | 984 | // phaseq = CFromFourierAnalysis(nmmax,dataq,phi0);
|
---|
| 985 | CFromFourierAnalysis(nmmax,dataq,phaseq, phi0);
|
---|
[729] | 986 |
|
---|
[1683] | 987 | // phaseu= CFromFourierAnalysis(nmmax,datau,phi0);
|
---|
| 988 | CFromFourierAnalysis(nmmax,datau,phaseu, phi0);
|
---|
[729] | 989 |
|
---|
| 990 | LambdaWXBuilder lwxb(theta,nlmax,nmmax);
|
---|
| 991 |
|
---|
| 992 | r_8 sqr2inv=1/Rac2;
|
---|
| 993 | for (int m = 0; m <= nmmax; m++)
|
---|
| 994 | {
|
---|
| 995 | r_8 lambda_w=0.;
|
---|
| 996 | r_8 lambda_x=0.;
|
---|
| 997 | lwxb.lam_wx(m, m, lambda_w, lambda_x);
|
---|
| 998 | complex<T> zi_lam_x((T)0., (T)lambda_x);
|
---|
| 999 | alme(m,m) += ( (T)(lambda_w)*phaseq(m)-zi_lam_x*phaseu(m) )*(T)(domega*sqr2inv);
|
---|
| 1000 | almb(m,m) += ( (T)(lambda_w)*phaseu(m)+zi_lam_x*phaseq(m) )*(T)(domega*sqr2inv);
|
---|
| 1001 |
|
---|
| 1002 | for (int l = m+1; l<= nlmax; l++)
|
---|
| 1003 | {
|
---|
| 1004 | lwxb.lam_wx(l, m, lambda_w, lambda_x);
|
---|
| 1005 | zi_lam_x = complex<T>((T)0., (T)lambda_x);
|
---|
| 1006 | alme(l,m) += ( (T)(lambda_w)*phaseq(m)-zi_lam_x*phaseu(m) )*(T)(domega*sqr2inv);
|
---|
| 1007 | almb(l,m) += ( (T)(lambda_w)*phaseu(m)+zi_lam_x*phaseq(m) )*(T)(domega*sqr2inv);
|
---|
| 1008 | }
|
---|
| 1009 | }
|
---|
| 1010 | }
|
---|
| 1011 |
|
---|
| 1012 |
|
---|
[1218] | 1013 | /*! \fn void SOPHYA::SphericalTransformServer::almFromPM(int_4 nph, int_4 nlmax,
|
---|
| 1014 | int_4 nmmax,
|
---|
| 1015 | r_8 phi0, r_8 domega,
|
---|
| 1016 | r_8 theta,
|
---|
| 1017 | const TVector<T>& dataq,
|
---|
| 1018 | const TVector<T>& datau,
|
---|
| 1019 | Alm<T>& alme,
|
---|
| 1020 | Alm<T>& almb) const
|
---|
| 1021 |
|
---|
| 1022 | Compute polarized Alm's as :
|
---|
| 1023 | \f[
|
---|
| 1024 | a_{lm}^E=-\frac{1}{2}\sum_{slices}{\omega_{pix}\left(\,_{+}\lambda_l^m\tilde{P^+}+\,_{-}\lambda_l^m\tilde{P^-}\right)}
|
---|
| 1025 | \f]
|
---|
| 1026 | \f[
|
---|
| 1027 | a_{lm}^B=\frac{i}{2}\sum_{slices}{\omega_{pix}\left(\,_{+}\lambda_l^m\tilde{P^+}-\,_{-}\lambda_l^m\tilde{P^-}\right)}
|
---|
| 1028 | \f]
|
---|
| 1029 |
|
---|
| 1030 | 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$U\f$ .
|
---|
| 1031 |
|
---|
| 1032 | \f$\omega_{pix}\f$ are solid angle of each pixel.
|
---|
| 1033 |
|
---|
| 1034 | dataq, datau : Stokes parameters.
|
---|
| 1035 |
|
---|
| 1036 | */
|
---|
[729] | 1037 | template<class T>
|
---|
[1218] | 1038 | void SphericalTransformServer<T>::almFromPM(int_4 nph, int_4 nlmax,
|
---|
| 1039 | int_4 nmmax,
|
---|
[729] | 1040 | r_8 phi0, r_8 domega,
|
---|
| 1041 | r_8 theta,
|
---|
| 1042 | const TVector<T>& dataq,
|
---|
| 1043 | const TVector<T>& datau,
|
---|
| 1044 | Alm<T>& alme,
|
---|
| 1045 | Alm<T>& almb) const
|
---|
| 1046 | {
|
---|
| 1047 | TVector< complex<T> > phasep(nmmax+1);
|
---|
| 1048 | TVector< complex<T> > phasem(nmmax+1);
|
---|
| 1049 | TVector<complex<T> > datain(nph);
|
---|
| 1050 | for (int i=0;i< nmmax+1;i++)
|
---|
| 1051 | {
|
---|
| 1052 | phasep(i)=0;
|
---|
| 1053 | phasem(i)=0;
|
---|
| 1054 | }
|
---|
[833] | 1055 | int kk;
|
---|
| 1056 | for(kk=0; kk<nph; kk++) datain(kk)=complex<T>(dataq(kk),datau(kk));
|
---|
[729] | 1057 |
|
---|
[746] | 1058 | phasep = CFromFourierAnalysis(nmmax,datain,phi0);
|
---|
[729] | 1059 |
|
---|
[833] | 1060 | for(kk=0; kk<nph; kk++) datain(kk)=complex<T>(dataq(kk),-datau(kk));
|
---|
[746] | 1061 | phasem = CFromFourierAnalysis(nmmax,datain,phi0);
|
---|
[729] | 1062 | LambdaPMBuilder lpmb(theta,nlmax,nmmax);
|
---|
| 1063 |
|
---|
| 1064 | for (int m = 0; m <= nmmax; m++)
|
---|
| 1065 | {
|
---|
| 1066 | r_8 lambda_p=0.;
|
---|
| 1067 | r_8 lambda_m=0.;
|
---|
| 1068 | complex<T> im((T)0.,(T)1.);
|
---|
| 1069 | lpmb.lam_pm(m, m, lambda_p, lambda_m);
|
---|
| 1070 |
|
---|
| 1071 | alme(m,m) += -( (T)(lambda_p)*phasep(m) + (T)(lambda_m)*phasem(m) )*(T)(domega*0.5);
|
---|
| 1072 | almb(m,m) += im*( (T)(lambda_p)*phasep(m) - (T)(lambda_m)*phasem(m) )*(T)(domega*0.5);
|
---|
| 1073 | for (int l = m+1; l<= nlmax; l++)
|
---|
| 1074 | {
|
---|
| 1075 | lpmb.lam_pm(l, m, lambda_p, lambda_m);
|
---|
| 1076 | alme(l,m) += -( (T)(lambda_p)*phasep(m) + (T)(lambda_m)*phasem(m) )*(T)(domega*0.5);
|
---|
| 1077 | almb(l,m) += im* ( (T)(lambda_p)*phasep(m) - (T)(lambda_m)*phasem(m) )*(T)(domega*0.5);
|
---|
| 1078 | }
|
---|
| 1079 | }
|
---|
| 1080 | }
|
---|
| 1081 |
|
---|
| 1082 |
|
---|
[1218] | 1083 | /*! \fn void SOPHYA::SphericalTransformServer::mapFromWX(int_4 nlmax, int_4 nmmax,
|
---|
| 1084 | SphericalMap<T>& mapq,
|
---|
| 1085 | SphericalMap<T>& mapu,
|
---|
| 1086 | const Alm<T>& alme,
|
---|
[2313] | 1087 | const Alm<T>& almb, bool healpix) const
|
---|
[1218] | 1088 |
|
---|
| 1089 | synthesis of Stokes parameters following formulae :
|
---|
| 1090 |
|
---|
| 1091 | \f[
|
---|
| 1092 | Q=\sum_{m=-mmax}^{mmax}b_m^qe^{im\varphi}
|
---|
| 1093 | \f]
|
---|
| 1094 | \f[
|
---|
| 1095 | U=\sum_{m=-mmax}^{mmax}b_m^ue^{im\varphi}
|
---|
| 1096 | \f]
|
---|
| 1097 |
|
---|
| 1098 | computed by FFT (method fourierSynthesisFromB called by the present one)
|
---|
| 1099 |
|
---|
| 1100 | with :
|
---|
| 1101 |
|
---|
| 1102 | \f[
|
---|
| 1103 | 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) }
|
---|
| 1104 | \f]
|
---|
| 1105 | \f[
|
---|
| 1106 | 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) }
|
---|
| 1107 | \f]
|
---|
| 1108 | */
|
---|
[729] | 1109 | template<class T>
|
---|
| 1110 | void SphericalTransformServer<T>::mapFromWX(int_4 nlmax, int_4 nmmax,
|
---|
| 1111 | SphericalMap<T>& mapq,
|
---|
| 1112 | SphericalMap<T>& mapu,
|
---|
| 1113 | const Alm<T>& alme,
|
---|
[2313] | 1114 | const Alm<T>& almb, bool healpix) const
|
---|
[729] | 1115 | {
|
---|
[2313] | 1116 | int i;
|
---|
| 1117 |
|
---|
[729] | 1118 | Bm<complex<T> > b_m_theta_q(nmmax);
|
---|
| 1119 | Bm<complex<T> > b_m_theta_u(nmmax);
|
---|
| 1120 |
|
---|
[746] | 1121 | for (int_4 ith = 0; ith < mapq.NbThetaSlices();ith++)
|
---|
[729] | 1122 | {
|
---|
| 1123 | int_4 nph;
|
---|
| 1124 | r_8 phi0;
|
---|
| 1125 | r_8 theta;
|
---|
| 1126 | TVector<int_4> pixNumber;
|
---|
| 1127 | TVector<T> datan;
|
---|
| 1128 |
|
---|
| 1129 | mapq.GetThetaSlice(ith,theta,phi0, pixNumber,datan);
|
---|
| 1130 | nph = pixNumber.NElts();
|
---|
| 1131 | // -----------------------------------------------------
|
---|
| 1132 | // for each theta, and each m, computes
|
---|
| 1133 | // b(m,theta) = sum_over_l>m (lambda_l_m(theta) * a_l_m)
|
---|
| 1134 | // ------------------------------------------------------
|
---|
| 1135 | LambdaWXBuilder lwxb(theta,nlmax,nmmax);
|
---|
| 1136 | // LambdaPMBuilder lpmb(theta,nlmax,nmmax);
|
---|
| 1137 | r_8 sqr2inv=1/Rac2;
|
---|
[833] | 1138 | int m;
|
---|
| 1139 | for (m = 0; m <= nmmax; m++)
|
---|
[729] | 1140 | {
|
---|
| 1141 | r_8 lambda_w=0.;
|
---|
| 1142 | r_8 lambda_x=0.;
|
---|
| 1143 | lwxb.lam_wx(m, m, lambda_w, lambda_x);
|
---|
| 1144 | complex<T> zi_lam_x((T)0., (T)lambda_x);
|
---|
| 1145 |
|
---|
| 1146 | b_m_theta_q(m) = ( (T)(lambda_w) * alme(m,m) - zi_lam_x * almb(m,m))*(T)sqr2inv ;
|
---|
| 1147 | b_m_theta_u(m) = ( (T)(lambda_w) * almb(m,m) + zi_lam_x * alme(m,m))*(T)sqr2inv;
|
---|
| 1148 |
|
---|
| 1149 |
|
---|
| 1150 | for (int l = m+1; l<= nlmax; l++)
|
---|
| 1151 | {
|
---|
| 1152 |
|
---|
| 1153 | lwxb.lam_wx(l, m, lambda_w, lambda_x);
|
---|
| 1154 | zi_lam_x= complex<T>((T)0., (T)lambda_x);
|
---|
| 1155 |
|
---|
| 1156 | b_m_theta_q(m) += ((T)(lambda_w)*alme(l,m)-zi_lam_x *almb(l,m))*(T)sqr2inv;
|
---|
| 1157 | b_m_theta_u(m) += ((T)(lambda_w)*almb(l,m)+zi_lam_x *alme(l,m))*(T)sqr2inv;
|
---|
| 1158 |
|
---|
| 1159 | }
|
---|
| 1160 | }
|
---|
| 1161 | // obtains the negative m of b(m,theta) (= complex conjugate)
|
---|
[833] | 1162 | for (m=1;m<=nmmax;m++)
|
---|
[729] | 1163 | {
|
---|
| 1164 | b_m_theta_q(-m) = conj(b_m_theta_q(m));
|
---|
| 1165 | b_m_theta_u(-m) = conj(b_m_theta_u(m));
|
---|
| 1166 | }
|
---|
[2313] | 1167 | if (healpix)
|
---|
[729] | 1168 | {
|
---|
[2313] | 1169 | TVector<T> Tempq = RfourierSynthesisFromB(b_m_theta_q,nph,phi0);
|
---|
| 1170 | TVector<T> Tempu = RfourierSynthesisFromB(b_m_theta_u,nph,phi0);
|
---|
| 1171 | for (i=0;i< nph;i++)
|
---|
| 1172 | {
|
---|
| 1173 | mapq(pixNumber(i))=Tempq(i);
|
---|
| 1174 | mapu(pixNumber(i))=Tempu(i);
|
---|
| 1175 | }
|
---|
[729] | 1176 | }
|
---|
[2313] | 1177 | else
|
---|
| 1178 | // pour des pixelisations quelconques (autres que HEALPix
|
---|
| 1179 | // nph n'est pas toujours pair
|
---|
| 1180 | // ca fait des problemes pour les transformees de Fourier
|
---|
| 1181 | // car le server de TF ajuste la longueur du vecteur reel
|
---|
| 1182 | // en sortie de TF, bref, la securite veut qu'on prenne une
|
---|
| 1183 | // TF complexe
|
---|
| 1184 | {
|
---|
| 1185 | TVector<complex<T> > Tempq = fourierSynthesisFromB(b_m_theta_q,nph,phi0);
|
---|
| 1186 | TVector<complex<T> > Tempu = fourierSynthesisFromB(b_m_theta_u,nph,phi0);
|
---|
| 1187 | for (i=0;i< nph;i++)
|
---|
| 1188 | {
|
---|
| 1189 | mapq(pixNumber(i))=Tempq(i).real();
|
---|
| 1190 | mapu(pixNumber(i))=Tempu(i).real();
|
---|
| 1191 | }
|
---|
| 1192 | }
|
---|
[729] | 1193 | }
|
---|
| 1194 | }
|
---|
[1218] | 1195 | /*! \fn void SOPHYA::SphericalTransformServer::mapFromPM(int_4 nlmax, int_4 nmmax,
|
---|
| 1196 | SphericalMap<T>& mapq,
|
---|
| 1197 | SphericalMap<T>& mapu,
|
---|
| 1198 | const Alm<T>& alme,
|
---|
| 1199 | const Alm<T>& almb) const
|
---|
| 1200 |
|
---|
| 1201 | synthesis of polarizations following formulae :
|
---|
| 1202 |
|
---|
| 1203 | \f[
|
---|
| 1204 | P^+ = \sum_{m=-mmax}^{mmax} {b_m^+e^{im\varphi} }
|
---|
| 1205 | \f]
|
---|
| 1206 | \f[
|
---|
| 1207 | P^- = \sum_{m=-mmax}^{mmax} {b_m^-e^{im\varphi} }
|
---|
| 1208 | \f]
|
---|
| 1209 |
|
---|
| 1210 | computed by FFT (method fourierSynthesisFromB called by the present one)
|
---|
| 1211 |
|
---|
| 1212 | with :
|
---|
| 1213 |
|
---|
| 1214 | \f[
|
---|
| 1215 | b_m^+=-\sum_{l=|m|}^{lmax}{\,_{+}\lambda_l^m \left( a_{lm}^E+ia_{lm}^B \right) }
|
---|
| 1216 | \f]
|
---|
| 1217 | \f[
|
---|
| 1218 | b_m^-=-\sum_{l=|m|}^{lmax}{\,_{+}\lambda_l^m \left( a_{lm}^E-ia_{lm}^B \right) }
|
---|
| 1219 | \f]
|
---|
| 1220 | */
|
---|
[729] | 1221 | template<class T>
|
---|
| 1222 | void SphericalTransformServer<T>::mapFromPM(int_4 nlmax, int_4 nmmax,
|
---|
| 1223 | SphericalMap<T>& mapq,
|
---|
| 1224 | SphericalMap<T>& mapu,
|
---|
| 1225 | const Alm<T>& alme,
|
---|
| 1226 | const Alm<T>& almb) const
|
---|
| 1227 | {
|
---|
| 1228 | Bm<complex<T> > b_m_theta_p(nmmax);
|
---|
| 1229 | Bm<complex<T> > b_m_theta_m(nmmax);
|
---|
[746] | 1230 | for (int_4 ith = 0; ith < mapq.NbThetaSlices();ith++)
|
---|
[729] | 1231 | {
|
---|
| 1232 | int_4 nph;
|
---|
| 1233 | r_8 phi0;
|
---|
| 1234 | r_8 theta;
|
---|
| 1235 | TVector<int_4> pixNumber;
|
---|
| 1236 | TVector<T> datan;
|
---|
| 1237 |
|
---|
| 1238 | mapq.GetThetaSlice(ith,theta,phi0, pixNumber,datan);
|
---|
| 1239 | nph = pixNumber.NElts();
|
---|
| 1240 |
|
---|
| 1241 | // -----------------------------------------------------
|
---|
| 1242 | // for each theta, and each m, computes
|
---|
| 1243 | // b(m,theta) = sum_over_l>m (lambda_l_m(theta) * a_l_m)
|
---|
| 1244 | //------------------------------------------------------
|
---|
| 1245 |
|
---|
| 1246 | LambdaPMBuilder lpmb(theta,nlmax,nmmax);
|
---|
[833] | 1247 | int m;
|
---|
| 1248 | for (m = 0; m <= nmmax; m++)
|
---|
[729] | 1249 | {
|
---|
| 1250 | r_8 lambda_p=0.;
|
---|
| 1251 | r_8 lambda_m=0.;
|
---|
| 1252 | lpmb.lam_pm(m, m, lambda_p, lambda_m);
|
---|
| 1253 | complex<T> im((T)0.,(T)1.);
|
---|
| 1254 |
|
---|
| 1255 | b_m_theta_p(m) = (T)(lambda_p )* (-alme(m,m) - im * almb(m,m));
|
---|
| 1256 | b_m_theta_m(m) = (T)(lambda_m) * (-alme(m,m) + im * almb(m,m));
|
---|
| 1257 |
|
---|
| 1258 |
|
---|
| 1259 | for (int l = m+1; l<= nlmax; l++)
|
---|
| 1260 | {
|
---|
| 1261 | lpmb.lam_pm(l, m, lambda_p, lambda_m);
|
---|
| 1262 | b_m_theta_p(m) += (T)(lambda_p)*(-alme(l,m)-im *almb(l,m));
|
---|
| 1263 | b_m_theta_m(m) += (T)(lambda_m)*(-alme(l,m)+im *almb(l,m));
|
---|
| 1264 | }
|
---|
| 1265 | }
|
---|
| 1266 |
|
---|
| 1267 | // obtains the negative m of b(m,theta) (= complex conjugate)
|
---|
[833] | 1268 | for (m=1;m<=nmmax;m++)
|
---|
[729] | 1269 | {
|
---|
| 1270 | b_m_theta_p(-m) = conj(b_m_theta_m(m));
|
---|
| 1271 | b_m_theta_m(-m) = conj(b_m_theta_p(m));
|
---|
| 1272 | }
|
---|
| 1273 |
|
---|
| 1274 | TVector<complex<T> > Tempp = fourierSynthesisFromB(b_m_theta_p,nph,phi0);
|
---|
| 1275 | TVector<complex<T> > Tempm = fourierSynthesisFromB(b_m_theta_m,nph,phi0);
|
---|
| 1276 |
|
---|
| 1277 | for (int i=0;i< nph;i++)
|
---|
| 1278 | {
|
---|
| 1279 | mapq(pixNumber(i))=0.5*(Tempp(i)+Tempm(i)).real();
|
---|
| 1280 | mapu(pixNumber(i))=0.5*(Tempp(i)-Tempm(i)).imag();
|
---|
| 1281 | }
|
---|
| 1282 | }
|
---|
| 1283 | }
|
---|
| 1284 |
|
---|
| 1285 |
|
---|
[1218] | 1286 | /*! \fn void SOPHYA::SphericalTransformServer::GenerateFromCl(SphericalMap<T>& sphq,
|
---|
| 1287 | SphericalMap<T>& sphu,
|
---|
| 1288 | int_4 pixelSizeIndex,
|
---|
| 1289 | const TVector<T>& Cle,
|
---|
| 1290 | const TVector<T>& Clb,
|
---|
| 1291 | const r_8 fwhm) const
|
---|
| 1292 |
|
---|
| 1293 | synthesis of a polarization map from power spectra electric-Cl and magnetic-Cl (Alm's are generated randomly, following a gaussian distribution).
|
---|
| 1294 | \param fwhm FWHM in arcmin for random generation of Alm's (eg. 5)
|
---|
| 1295 | */
|
---|
[729] | 1296 | template<class T>
|
---|
| 1297 | void SphericalTransformServer<T>::GenerateFromCl(SphericalMap<T>& sphq,
|
---|
| 1298 | SphericalMap<T>& sphu,
|
---|
| 1299 | int_4 pixelSizeIndex,
|
---|
| 1300 | const TVector<T>& Cle,
|
---|
| 1301 | const TVector<T>& Clb,
|
---|
| 1302 | const r_8 fwhm) const
|
---|
| 1303 | {
|
---|
| 1304 | if (Cle.NElts() != Clb.NElts())
|
---|
| 1305 | {
|
---|
| 1306 | cout << " SphericalTransformServer: les deux tableaux Cl n'ont pas la meme taille" << endl;
|
---|
| 1307 | throw SzMismatchError("SphericalTransformServer::GenerateFromCl : two Cl arrays have not same size");
|
---|
| 1308 | }
|
---|
| 1309 |
|
---|
| 1310 | // Alm<T> a2lme,a2lmb;
|
---|
| 1311 | // almFromCl(a2lme, Cle, fwhm);
|
---|
| 1312 | // almFromCl(a2lmb, Clb, fwhm);
|
---|
| 1313 | // Alm<T> a2lme = almFromCl(Cle, fwhm);
|
---|
| 1314 | // Alm<T> a2lmb = almFromCl(Clb, fwhm);
|
---|
[3510] | 1315 | Alm<T> a2lme(Cle, fwhm, rg_);
|
---|
| 1316 | Alm<T> a2lmb(Clb, fwhm, rg_);
|
---|
[729] | 1317 |
|
---|
| 1318 | GenerateFromAlm(sphq,sphu,pixelSizeIndex,a2lme,a2lmb);
|
---|
| 1319 | }
|
---|
[1218] | 1320 | /*! \fn void SOPHYA::SphericalTransformServer::GenerateFromCl(SphericalMap<T>& sph,
|
---|
| 1321 | int_4 pixelSizeIndex,
|
---|
| 1322 | const TVector<T>& Cl,
|
---|
| 1323 | const r_8 fwhm) const
|
---|
| 1324 |
|
---|
| 1325 | synthesis of a temperature map from power spectrum Cl (Alm's are generated randomly, following a gaussian distribution). */
|
---|
[729] | 1326 | template<class T>
|
---|
| 1327 | void SphericalTransformServer<T>::GenerateFromCl(SphericalMap<T>& sph,
|
---|
| 1328 | int_4 pixelSizeIndex,
|
---|
| 1329 | const TVector<T>& Cl,
|
---|
| 1330 | const r_8 fwhm) const
|
---|
| 1331 | {
|
---|
| 1332 |
|
---|
[3510] | 1333 | Alm<T> alm(Cl, fwhm, rg_);
|
---|
[729] | 1334 | GenerateFromAlm(sph,pixelSizeIndex, alm );
|
---|
| 1335 | }
|
---|
| 1336 |
|
---|
| 1337 |
|
---|
| 1338 |
|
---|
[1756] | 1339 | /*! \fn TVector<T> SOPHYA::SphericalTransformServer::DecomposeToCl(SphericalMap<T>& sph, int_4 nlmax, r_8 cos_theta_cut, int iterationOrder) const
|
---|
[1218] | 1340 |
|
---|
[1683] | 1341 | \return power spectrum from analysis of a temperature map. THE MAP CAN BE MODIFIED (if iterationOrder >0)
|
---|
[1218] | 1342 |
|
---|
| 1343 | \param<nlmax> : maximum value of the l index
|
---|
| 1344 |
|
---|
| 1345 | \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.
|
---|
[1683] | 1346 |
|
---|
[1756] | 1347 | \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 !
|
---|
[1683] | 1348 |
|
---|
[1218] | 1349 | */
|
---|
[729] | 1350 | template <class T>
|
---|
[1683] | 1351 | TVector<T> SphericalTransformServer<T>::DecomposeToCl(SphericalMap<T>& sph, int_4 nlmax, r_8 cos_theta_cut, int iterationOrder) const
|
---|
[729] | 1352 | {
|
---|
[1683] | 1353 | Alm<T> alm;
|
---|
| 1354 | DecomposeToAlm( sph, alm, nlmax, cos_theta_cut, iterationOrder);
|
---|
[729] | 1355 | // power spectrum
|
---|
| 1356 | return alm.powerSpectrum();
|
---|
| 1357 | }
|
---|
| 1358 |
|
---|
[1756] | 1359 |
|
---|
| 1360 | /*! \fn TVector<T> SOPHYA::SphericalTransformServer::DecomposeToCl(const SphericalMap<T>& sph, int_4 nlmax, r_8 cos_theta_cut) const
|
---|
| 1361 |
|
---|
| 1362 | \return power spectrum from analysis of a temperature map.
|
---|
| 1363 |
|
---|
| 1364 | \param<nlmax> : maximum value of the l index
|
---|
| 1365 |
|
---|
| 1366 | \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.
|
---|
| 1367 |
|
---|
| 1368 |
|
---|
| 1369 | */
|
---|
| 1370 |
|
---|
| 1371 |
|
---|
| 1372 | template <class T>
|
---|
| 1373 | TVector<T> SphericalTransformServer<T>::DecomposeToCl(const SphericalMap<T>& sph, int_4 nlmax, r_8 cos_theta_cut) const
|
---|
| 1374 | {
|
---|
| 1375 | Alm<T> alm;
|
---|
| 1376 | DecomposeToAlm( sph, alm, nlmax, cos_theta_cut);
|
---|
| 1377 | // power spectrum
|
---|
| 1378 | return alm.powerSpectrum();
|
---|
| 1379 | }
|
---|
| 1380 |
|
---|
[729] | 1381 | #ifdef __CXX_PRAGMA_TEMPLATES__
|
---|
| 1382 | #pragma define_template SphericalTransformServer<r_8>
|
---|
| 1383 | #pragma define_template SphericalTransformServer<r_4>
|
---|
| 1384 | #endif
|
---|
| 1385 | #if defined(ANSI_TEMPLATES) || defined(GNU_TEMPLATES)
|
---|
[2872] | 1386 | template class SOPHYA::SphericalTransformServer<r_8>;
|
---|
| 1387 | template class SOPHYA::SphericalTransformServer<r_4>;
|
---|
[729] | 1388 | #endif
|
---|