demo_lensit_wcurl.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Iterative lensing estimation demo (CMB-S4).\n",
    "\n",
    "This notebook goes through an example of joint lensing gradient and curl iterative reconstruction, from polarization Q and U CMB data on 650 square degrees, under ideal conditions (no masking, homogeneous noise, etc), with pixel size 1.5 arcmin. Specifically, we proceed as follows\n",
    "1. We build the quadratic estimator and initial curvature estimates\n",
    "2. We setup the iterator instance\n",
    "3. We perform the iterations\n",
    "4. We have a quick look at the quality of the lensing map reconstruction\n",
    "\n",
    "It should take or minute or so to perform all iterations from the starting point.\n",
    "\n",
    "Before starting, set the environment variable LENSIT to somewhere that is safe to write. A bunch of stuff might get cached."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import pylab as pl\n",
    "import numpy as np\n",
    "import lensit as li\n",
    "import os\n",
    "EXP = 'S4_SPDP'\n",
    "nlev_t, nlev_p, beam, ellmin, ellmax = li.get_config(EXP)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. We build the quadratic estimator and initial curvature estimates:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " I will be using data from ell=14 to ell=4096 only\n",
      " The sky band-limit is ell=5120\n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:00])   inverse len Pmats  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:00])   Fxx , part 1  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:00])   Fyy , part 1  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:01])   Fxy , part 1  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:01])   Fxx , part 2  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:01])   Fyy , part 2  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:02])   Fxy , part 2  curvpOlm \n",
      "     [00:00:02]  (total [00:00:02])   get_qlm_resplm:: get curvature matrices  \n",
      "     [00:00:00]  (total [00:00:02])   get_qlm_resplm:: inversion  \n",
      "     _get_pmati:: cached /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/Covs/S4_SPDP/LD10HD11/QU_Pmatinv_len_00.npy\n",
      "     _get_pmati:: cached /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/Covs/S4_SPDP/LD10HD11/QU_Pmatinv_len_10.npy\n",
      "     _get_pmati:: cached /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/Covs/S4_SPDP/LD10HD11/QU_Pmatinv_len_11.npy\n",
      "     [00:00:00]  (total [00:00:00])   get_qlms::mult with len Pmat  \n",
      "     [00:00:00]  (total [00:00:00])   get_qlms::cartesian gradients  \n",
      "     [00:00:00]  (total [00:00:00])   get_qlms::rotation to phi Omega  \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:00])   inverse unl Pmats  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:00])   Fxx , part 1  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:00])   Fyy , part 1  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:01])   Fxy , part 1  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:01])   Fxx , part 2  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:01])   Fyy , part 2  curvpOlm \n",
      "  lensit.ffs_covs.ffs_cov   [00:00:00]  (total [00:00:02])   Fxy , part 2  curvpOlm \n",
      "     [00:00:02]  (total [00:00:02])   get_qlm_resplm:: get curvature matrices  \n",
      "     [00:00:00]  (total [00:00:02])   get_qlm_resplm:: inversion  \n"
     ]
    }
   ],
   "source": [
    "def get_starting_point(idx): \n",
    "    \"\"\"\n",
    "    This returns initial data for simulation index 'idx' from a CMB-S4 simulation library.\n",
    "    On first call the simulation library will generate all simulations phases, hence might take a little while.\n",
    "    \"\"\" \n",
    "    sims = li.get_maps_lib(EXP, 10, HDres=11, wrotation=True, nsims=1) # Simulation-library for configuration 'S4'. \n",
    "                                         # Parameters 10, 11 produces data on 645 sq. deg, \n",
    "                                         # with lensed CMB's generated at 0.75 arcmin resolution, \n",
    "                                         # but data collected at 1.5 arcmin resolution. \n",
    "    \n",
    "    isocov = li.get_isocov(EXP, 10, 11) # Isotropic filtering instance, that can used for Q.E. calculation\n",
    "                                         # and other things.isocov.lib_datalm defines the mode-filtering applied \n",
    "                                         # the data, and isocov.lib_skyalm the band-limits of the unlensed sky.\n",
    "    print(\" I will be using data from ell=%s to ell=%s only\"%(isocov.lib_datalm.ellmin, isocov.lib_datalm.ellmax))\n",
    "    print(\" The sky band-limit is ell=%s\"%(isocov.lib_skyalm.ellmax))\n",
    "    \n",
    "    lib_qlm = isocov.lib_skyalm #: This means we will reconstruct the lensing potential for all unlensed sky modes.\n",
    "    def cli(cl):\n",
    "        ret = np.zeros_like(cl)\n",
    "        ret[np.where(cl > 0)] = 1./cl[np.where(cl > 0)]\n",
    "        return ret\n",
    "    \n",
    "    # We now build the Wiener-filtered quadratic estimator. We use lensed CMB spectra in the weights.\n",
    "    datalms = np.array([isocov.lib_datalm.map2alm(m) for m in sims.get_sim_qumap(idx)]) \n",
    "    H0lens =  [cli(isocov.get_N0cls('QU', lib_qlm, use_cls_len=True)[0]), cli(isocov.get_N0cls('QU', lib_qlm, use_cls_len=True)[1])]\n",
    "    POlm0 = 0.5 * isocov.get_qlms('QU',  isocov.get_iblms('QU', datalms, use_cls_len=True)[0], lib_qlm, \n",
    "                                 use_cls_len=True)\n",
    "    \n",
    "    # Normalization and Wiener-filtering:\n",
    "    cpp_prior = li.get_fidcls(wrotationCls=True)[0]['pp'][:lib_qlm.ellmax+1]\n",
    "    coo_prior = li.get_fidcls(wrotationCls=True)[0]['oo'][:lib_qlm.ellmax+1]\n",
    "    lib_qlm.almxfl(POlm0[0], cli(H0lens[0] + cli(cpp_prior)), inplace=True)\n",
    "    lib_qlm.almxfl(POlm0[1], cli(H0lens[1] + cli(coo_prior)), inplace=True)\n",
    "    \n",
    "    # Initial likelihood curvature guess. We use here N0 as calculated with unlensed CMB spectra:\n",
    "    H0unls =  [cli(isocov.get_N0cls('QU', lib_qlm, use_cls_len=False)[0]), cli(isocov.get_N0cls('QU', lib_qlm, use_cls_len=False)[1])]\n",
    "    return POlm0, lib_qlm, datalms, isocov.lib_datalm, H0unls, H0lens\n",
    "\n",
    "POlm0, lib_qlm, datalms, lib_datalm, H0s, H0len = get_starting_point(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. We setup the iterator instance:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " I see t 0.35355339059327373\n",
      " I see q 0.5\n",
      " I see u 0.5\n",
      "ffs iterator : This is 0 trying to setup /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000\n",
      "++ ffs_QU_iterator: Caching qlm_norms and N0s/Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000\n",
      "     cached /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_P_H0.dat\n",
      "     cached /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_O_H0.dat\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it000.npy\n",
      "++ ffs_<class 'type'> masked iterator : setup OK\n"
     ]
    }
   ],
   "source": [
    "from lensit.ffs_iterators.ffs_iterator_wcurl import ffs_iterator_cstMF\n",
    "from lensit.misc.misc_utils import gauss_beam\n",
    "from lensit.qcinv import ffs_ninv_filt_ideal, chain_samples\n",
    "from lensit.ffs_covs import ell_mat\n",
    "from lensit import LMAX_SKY\n",
    "\n",
    "def get_itlib(lib_dir, POlm0, lib_qlm,  datalms, lib_datalm, H0s, beam_fwhmamin=beam, NlevT_filt=nlev_t, NlevP_filt=nlev_p):\n",
    "    \"\"\"\n",
    "    This returns an iterator instance from the input data maps, lensing map starting point,\n",
    "    likelihood curvature guess and choice of filtering parameters (ideally close to those of the data).\n",
    "    \"\"\"\n",
    "\n",
    "    if not os.path.exists(lib_dir): \n",
    "        os.makedirs(lib_dir)\n",
    "    # Prior on lensing power spectrum, and CMB spectra for the filtering at each iteration step.\n",
    "    cls_unl = li.get_fidcls(LMAX_SKY, wrotationCls=True)[0]\n",
    "    cpp_priors = [np.copy(cls_unl['pp'][:LMAX_SKY+1]), np.copy(cls_unl['oo'][:LMAX_SKY+1])]\n",
    "    \n",
    "    \n",
    "    lib_skyalm = ell_mat.ffs_alm_pyFFTW(lib_datalm.ell_mat, filt_func=lambda ell:ell<=LMAX_SKY)\n",
    "                            #: This means we perform here the lensing of CMB skies at the same resolution \n",
    "                            #  than the data with the band-limit of 6000.\n",
    "    transf = gauss_beam(beam_fwhmamin / 180. / 60. * np.pi, lmax=LMAX_SKY) #: fiducial beam\n",
    "    \n",
    "    # Anisotropic filtering instance, with unlensed CMB spectra as inputs. Delfections will be added by the iterator.\n",
    "    filt = ffs_ninv_filt_ideal.ffs_ninv_filt(lib_datalm, lib_skyalm, cls_unl, transf, NlevT_filt,NlevP_filt)\n",
    "    \n",
    "    # Description of the multi-grid chain to use: (here the simplest, diagonal pre-conditioner) \n",
    "    chain_descr = chain_samples.get_isomgchain(filt.lib_skyalm.ellmax,filt.lib_datalm.shape,\n",
    "                                                        tol=1e-6,iter_max=200)\n",
    "    \n",
    "    # We assume no primordial B-modes, the E-B filtering will assume all B-modes are either noise or lensing:\n",
    "    opfilt =  li.qcinv.opfilt_cinv_noBB\n",
    "    opfilt._type = 'QU' # We consider polarization only\n",
    "    \n",
    "    # With all this now in place, we can build the iterator instance:\n",
    "    iterator = ffs_iterator_cstMF(lib_dir, 'QU', filt, datalms, lib_qlm, \n",
    "              POlm0, H0s, POlm0 * 0., cpp_priors,  chain_descr=chain_descr, opfilt=opfilt, verbose=True)\n",
    "               # We use here an iterator instance that uses an analytical approximation \n",
    "               # for the mean-field term at each step.\n",
    "    return iterator\n",
    "\n",
    "\n",
    "lib_dir = os.path.join(os.environ['LENSIT'], 'temp', 'iteratorwcurl_%s_sim%03d'%(EXP, 0))\n",
    "itlib = get_itlib(lib_dir, POlm0, lib_qlm, datalms, lib_datalm, H0s)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. We perform the iterations:\n",
    "This should take about ~ 1 min. to finish. A lot of print-out in verbose mode."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rank 0 caching displacement comp. for it. 5\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0 inverting displacement it. 5\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.18e+00 amin\n",
      "           mean x, y rms increment : 2.84e-02, 2.74e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.57e-02 amin\n",
      "           mean x, y rms increment : 6.25e-05, 6.06e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 1.63e-04 amin\n",
      "           mean x, y rms increment : 1.37e-07, 1.32e-07 amin\n",
      " [00:00:00] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it005.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it005.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it4.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:01] (0, 0.0095713802)\n",
      "(1024, 5120) [00:00:01] (1, 0.0009723153)\n",
      "(1024, 5120) [00:00:02] (2, 0.0001726545)\n",
      "(1024, 5120) [00:00:02] (3, 0.0000470249)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000149307)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000050369)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000016570)\n",
      "(1024, 5120) [00:00:05] (7, 0.0000005724)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it5.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it005.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n",
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy\n",
      " [00:00:01]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it006.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it005_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it005_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it005_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it005_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0005_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0005_libdir\n",
      "rank 0 caching displacement comp. for it. 6\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0 inverting displacement it. 6\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.18e+00 amin\n",
      "           mean x, y rms increment : 2.83e-02, 2.72e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.46e-02 amin\n",
      "           mean x, y rms increment : 6.10e-05, 5.90e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 9.16e-05 amin\n",
      "           mean x, y rms increment : 1.28e-07, 1.24e-07 amin\n",
      " [00:00:00] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it006.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it006.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it5.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:01] (0, 0.0087053592)\n",
      "(1024, 5120) [00:00:01] (1, 0.0008791582)\n",
      "(1024, 5120) [00:00:02] (2, 0.0001482900)\n",
      "(1024, 5120) [00:00:03] (3, 0.0000402557)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000129424)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000044242)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000015075)\n",
      "(1024, 5120) [00:00:05] (7, 0.0000005350)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it6.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 21) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it006.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy\n",
      " [00:00:01]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it007.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it006_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it006_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it006_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it006_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0006_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0006_libdir\n",
      "rank 0 caching displacement comp. for it. 7\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0 inverting displacement it. 7\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.14e+00 amin\n",
      "           mean x, y rms increment : 2.86e-02, 2.75e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.47e-02 amin\n",
      "           mean x, y rms increment : 6.18e-05, 5.98e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 8.00e-05 amin\n",
      "           mean x, y rms increment : 1.28e-07, 1.24e-07 amin\n",
      " [00:00:00] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it007.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it007.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it6.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:00] (0, 0.0058521079)\n",
      "(1024, 5120) [00:00:01] (1, 0.0005849342)\n",
      "(1024, 5120) [00:00:02] (2, 0.0000994096)\n",
      "(1024, 5120) [00:00:02] (3, 0.0000262633)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000083892)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000028699)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000009712)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it7.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it007.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n",
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_7_po.npy\n",
      " [00:00:01]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it008.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it007_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it007_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it007_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it007_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0007_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0007_libdir\n",
      "rank 0 caching displacement comp. for it. 8\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0 inverting displacement it. 8\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.15e+00 amin\n",
      "           mean x, y rms increment : 2.91e-02, 2.80e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.58e-02 amin\n",
      "           mean x, y rms increment : 6.37e-05, 6.16e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 8.56e-05 amin\n",
      "           mean x, y rms increment : 1.32e-07, 1.28e-07 amin\n",
      " [00:00:00] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it008.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it008.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it7.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:00] (0, 0.0048643191)\n",
      "(1024, 5120) [00:00:01] (1, 0.0004962416)\n",
      "(1024, 5120) [00:00:02] (2, 0.0000844926)\n",
      "(1024, 5120) [00:00:02] (3, 0.0000221685)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000071136)\n",
      "(1024, 5120) [00:00:03] (5, 0.0000024124)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000008115)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it8.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (20, 22) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it008.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_7_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_7_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_7_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_8_po.npy\n",
      " [00:00:01]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it009.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it008_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it008_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it008_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it008_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0008_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0008_libdir\n",
      "rank 0 caching displacement comp. for it. 9\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 inverting displacement it. 9\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.19e+00 amin\n",
      "           mean x, y rms increment : 2.96e-02, 2.85e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.68e-02 amin\n",
      "           mean x, y rms increment : 6.56e-05, 6.35e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 9.14e-05 amin\n",
      "           mean x, y rms increment : 1.36e-07, 1.32e-07 amin\n",
      " [00:00:00] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it009.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it009.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it8.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:00] (0, 0.0052355090)\n",
      "(1024, 5120) [00:00:01] (1, 0.0005227634)\n",
      "(1024, 5120) [00:00:02] (2, 0.0000866531)\n",
      "(1024, 5120) [00:00:02] (3, 0.0000228284)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000074249)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000025959)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000008965)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it9.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it009.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_8_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_7_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_7_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_8_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_8_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n",
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_9_po.npy\n",
      " [00:00:01]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it010.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it009_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it009_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it009_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it009_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0009_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0009_libdir\n",
      "rank 0 caching displacement comp. for it. 10\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 inverting displacement it. 10\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.22e+00 amin\n",
      "           mean x, y rms increment : 3.00e-02, 2.89e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.71e-02 amin\n",
      "           mean x, y rms increment : 6.69e-05, 6.46e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 9.26e-05 amin\n",
      "           mean x, y rms increment : 1.38e-07, 1.34e-07 amin\n",
      " [00:00:00] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it010.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it010.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it9.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:00] (0, 0.0044314909)\n",
      "(1024, 5120) [00:00:01] (1, 0.0004459392)\n",
      "(1024, 5120) [00:00:02] (2, 0.0000745972)\n",
      "(1024, 5120) [00:00:02] (3, 0.0000197750)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000065581)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000023459)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000008208)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it10.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it010.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_9_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_7_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_7_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_8_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_8_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_9_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_9_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n",
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_10_po.npy\n",
      " [00:00:01]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it011.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it010_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it010_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it010_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it010_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0010_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0010_libdir\n",
      "rank 0 caching displacement comp. for it. 11\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 inverting displacement it. 11\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.23e+00 amin\n",
      "           mean x, y rms increment : 3.04e-02, 2.92e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.72e-02 amin\n",
      "           mean x, y rms increment : 6.78e-05, 6.55e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 9.37e-05 amin\n",
      "           mean x, y rms increment : 1.39e-07, 1.34e-07 amin\n",
      " [00:00:01] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it011.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it011.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it10.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:00] (0, 0.0043028434)\n",
      "(1024, 5120) [00:00:01] (1, 0.0004375614)\n",
      "(1024, 5120) [00:00:02] (2, 0.0000729218)\n",
      "(1024, 5120) [00:00:03] (3, 0.0000188345)\n",
      "(1024, 5120) [00:00:04] (4, 0.0000061745)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000021912)\n",
      "(1024, 5120) [00:00:05] (6, 0.0000007585)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it11.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it011.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_10_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_7_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_7_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_8_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_8_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_9_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_9_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_10_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_10_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_11_po.npy\n",
      " [00:00:02]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it012.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it011_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it011_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it011_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it011_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0011_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0011_libdir\n",
      "rank 0 caching displacement comp. for it. 12\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 inverting displacement it. 12\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.23e+00 amin\n",
      "           mean x, y rms increment : 3.08e-02, 2.95e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.77e-02 amin\n",
      "           mean x, y rms increment : 6.89e-05, 6.64e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 9.67e-05 amin\n",
      "           mean x, y rms increment : 1.40e-07, 1.35e-07 amin\n",
      " [00:00:00] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it012.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it012.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it11.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:00] (0, 0.0047488887)\n",
      "(1024, 5120) [00:00:01] (1, 0.0004847463)\n",
      "(1024, 5120) [00:00:02] (2, 0.0000801758)\n",
      "(1024, 5120) [00:00:02] (3, 0.0000204146)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000067034)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000023759)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000008215)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it12.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 22) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it012.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_11_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_7_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_7_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_8_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_8_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_9_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_9_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_10_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_10_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_11_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_11_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n",
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_12_po.npy\n",
      " [00:00:02]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it013.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it012_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it012_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it012_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it012_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0012_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0012_libdir\n",
      "rank 0 caching displacement comp. for it. 13\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0 inverting displacement it. 13\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.24e+00 amin\n",
      "           mean x, y rms increment : 3.11e-02, 2.98e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.84e-02 amin\n",
      "           mean x, y rms increment : 6.99e-05, 6.74e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 1.00e-04 amin\n",
      "           mean x, y rms increment : 1.41e-07, 1.36e-07 amin\n",
      " [00:00:01] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it013.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it013.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it12.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:01] (0, 0.0037303156)\n",
      "(1024, 5120) [00:00:01] (1, 0.0003838162)\n",
      "(1024, 5120) [00:00:02] (2, 0.0000636987)\n",
      "(1024, 5120) [00:00:02] (3, 0.0000162271)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000053410)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000019030)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000006573)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it13.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it013.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_12_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_7_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_7_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_8_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_8_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_9_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_9_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_10_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_10_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_11_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_11_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_12_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_12_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n",
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_13_po.npy\n",
      " [00:00:02]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it014.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it013_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it013_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it013_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it013_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0013_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0013_libdir\n",
      "rank 0 caching displacement comp. for it. 14\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0 inverting displacement it. 14\n",
      "NR iter 1: max. increment size in NR deflection inverse 1.25e+00 amin\n",
      "           mean x, y rms increment : 3.13e-02, 3.00e-02 amin\n",
      "NR iter 2: max. increment size in NR deflection inverse 1.90e-02 amin\n",
      "           mean x, y rms increment : 7.06e-05, 6.80e-05 amin\n",
      "NR iter 3: max. increment size in NR deflection inverse 1.02e-04 amin\n",
      "           mean x, y rms increment : 1.43e-07, 1.37e-07 amin\n",
      " [00:00:00] ffs_deflect::calculating inverse displ. field > 00%\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "almxfl, 2 comp\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpopri_it014.npy\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpodet_it014.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0 loading /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it13.npy\n",
      "This is calc prep for QU W. Filtering cinv_noBB\n",
      "(1024, 5120) [00:00:01] (0, 0.0020047042)\n",
      "(1024, 5120) [00:00:01] (1, 0.0002054289)\n",
      "(1024, 5120) [00:00:02] (2, 0.0000343615)\n",
      "(1024, 5120) [00:00:02] (3, 0.0000088461)\n",
      "(1024, 5120) [00:00:03] (4, 0.0000029251)\n",
      "(1024, 5120) [00:00:04] (5, 0.0000010419)\n",
      "(1024, 5120) [00:00:04] (6, 0.0000003628)\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/MAPlms/Mlik_po_it14.npy\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0, ffs_deflect::buffers size, chk_shape (21, 23) [1024 1024]\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/qlm_gradpolik_it014.npy\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_13_po.npy\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_0_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_0_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_1_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_1_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_2_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_2_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_3_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_3_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_4_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_4_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_5_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_5_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_6_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_6_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_7_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_7_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_8_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_8_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_9_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_9_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_10_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_10_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_11_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_11_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_12_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_12_po.npy  to Hessian\n",
      "Linked y vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_yn_13_po.npy  to Hessian\n",
      "Linked s vector  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_13_po.npy  to Hessian\n",
      "rank 0 calculating descent direction\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "almxfl, 2 comp\n",
      "almxfl, 2 comp\n",
      "rank 0 caching  /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/Hessian/rlm_sn_14_po.npy\n",
      " [00:00:02]  Exec. time for descent direction calculation\n",
      "rank 0 caching /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/polm_it015.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it014_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_po_it014_dy.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it014_dx.npy\n",
      "     removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_po_it014_dy.npy\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/f_0014_libdir\n",
      "Removed : /Users/jcarron/OneDrive - unige.ch/LensIt/temp/temp/iteratorwcurl_S4_SPDP_sim000/finv_0014_libdir\n"
     ]
    }
   ],
   "source": [
    "itlib.soltn_cond = True\n",
    "itmax = 15\n",
    "for i in range(itmax + 1):\n",
    "    itlib.iterate(i, 'p')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. We have a quick look at the quality of the lensing map reconstruction:\n",
    "We plot the binned power spectrum of the residual map\n",
    "$$\n",
    "\\hat \\phi - \\phi^{\\rm input}\n",
    "$$\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ell, = np.where(itlib.lib_qlm.get_Nell()[:3001])\n",
    "w = lambda ell : ell ** 2 * (ell + 1.) ** 2 /2./np.pi * 1e7\n",
    "edges = np.arange(0, 3100, 100)\n",
    "bl = edges[:-1];bu = edges[1:] \n",
    "\n",
    "sims = li.get_maps_lib(EXP, 11, 11, wrotation=True, nsims=1)\n",
    "plm_input = itlib.lib_qlm.udgrade(sims.lencmbs.lib_skyalm, sims.lencmbs.get_sim_plm(0))\n",
    "olm_input = itlib.lib_qlm.udgrade(sims.lencmbs.lib_skyalm, sims.lencmbs.get_sim_olm(0))\n",
    "s = itlib.lib_qlm.alm_size\n",
    "bnd_cl = lambda cl : li.misc.misc_utils.binned(cl, ell, bl, bu, w=w, return_err=True)\n",
    "pl.figure(figsize= (10,5))\n",
    "for it in range(itmax + 1):\n",
    "    cl, err = bnd_cl(itlib.lib_qlm.alm2cl(itlib.get_POlm(it)[0:s] - plm_input)) \n",
    "    pl.errorbar(0.5 * bl + 0.5 * bu, cl, yerr=err, label='Iter. %s'%it + (it == 0) * ' (Q.E.)')\n",
    "pl.plot(ell, li.get_fidcls()[0]['pp'][ell] * w(ell), c='k', label=r'$C_L^{\\phi\\phi, \\rm fid}$')\n",
    "pl.ylabel(r'$10^7\\: L^2(L + 1)^2 C^{\\hat \\phi - \\phi^{\\rm in}}_L /2 \\pi$', fontsize=20)\n",
    "pl.xlabel(r'$L$', fontsize=14)\n",
    "pl.legend(ncol=3, fontsize=14)\n",
    "pl.semilogy()\n",
    "pl.figure(figsize= (10,5))\n",
    "for it in range(itmax + 1):\n",
    "    cl, err = bnd_cl(itlib.lib_qlm.alm2cl(itlib.get_POlm(it)[s:2*s] - olm_input)) \n",
    "    pl.errorbar(0.5 * bl + 0.5 * bu, cl, yerr=err, label='Iter. %s'%it + (it == 0) * ' (Q.E.)')\n",
    "pl.plot(ell, li.get_fidcls(wrotationCls=True)[0]['oo'][ell] * w(ell), c='k', label=r'$C_L^{\\Omega \\Omega, \\rm fid}$')\n",
    "pl.ylabel(r'$10^7\\: L^2(L + 1)^2 C^{\\hat \\phi - \\phi^{\\rm in}}_L /2 \\pi$', fontsize=20)\n",
    "pl.xlabel(r'$L$', fontsize=14)\n",
    "pl.legend(ncol=3, fontsize=14)\n",
    "pl.semilogy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'cross-correlation to curl')"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from lensit.misc.misc_utils import binned\n",
    "\n",
    "coo_fid = li.get_fidcls(wrotationCls=True)[0]['oo']\n",
    "s = lib_qlm.alm_size\n",
    "bls = 50 * np.arange(40)\n",
    "bus = 50 * np.arange(1, 41)\n",
    "bcs = 0.5 * (bls + bus)\n",
    "clx = np.zeros(lib_qlm.ellmax + 1)\n",
    "ls, = np.where(lib_qlm.get_Nell()[:3001]> 1)\n",
    "\n",
    "for it in np.unique([0, itmax // 2, itmax]):\n",
    "    olm = itlib.get_POlm(it)[s:2*s]\n",
    "    clx[ls] = lib_qlm.alm2cl(olm, olm_input)[ls] /(np.sqrt(coo_fid[ls] * lib_qlm.alm2cl(olm)[ls]))\n",
    "    #pl.plot(ls, clx[ls])\n",
    "    pl.plot(bcs, binned(clx, ls, bls, bus), label='iter ' + str(it))\n",
    "pl.legend()\n",
    "pl.title('cross-correlation to curl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}