{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TXS 0506+056 time-integrated and time-dependent\n",
    "\n",
    ".. contents:: :local:\n",
    "\n",
    "In this tutorial, we'll reproduce some TXS 0506+056 results."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy import stats\n",
    "\n",
    "import histlite as hl\n",
    "import csky as cy\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "cy.plotting.mrichman_mpl()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll set up a timer:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "timer = cy.timing.Timer()\n",
    "time = timer.time"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's grab the \"PS + GFU\" sample, using `version-002-p01` for each, and disabling the angular error floor (which was not included in the original analysis):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "ana_dir = cy.utils.ensure_dir('/data/user/mrichman/csky_cache/ana')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up Analysis for:\n",
      "IC40, IC59, IC79, IC86_2011, IC86_2012_2014, GFU_2015_2017\n",
      "Setting up IC40...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC40_corrected_MC.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC40_exp.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC40_GRL.npy ...\n",
      "<- /data/user/mrichman/csky_cache/ana/IC40.subanalysis.version-002-p01.npy    \n",
      "Setting up IC59...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC59_corrected_MC.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC59_exp.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC59_GRL.npy ...\n",
      "<- /data/user/mrichman/csky_cache/ana/IC59.subanalysis.version-002-p01.npy    \n",
      "Setting up IC79...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC79b_corrected_MC.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC79b_exp.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC79b_GRL.npy ...\n",
      "<- /data/user/mrichman/csky_cache/ana/IC79.subanalysis.version-002-p01.npy    \n",
      "Setting up IC86_2011...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86_corrected_MC.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86_exp.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86_GRL.npy ...\n",
      "<- /data/user/mrichman/csky_cache/ana/IC86_2011.subanalysis.version-002-p01.npy    \n",
      "Setting up IC86_2012_2014...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86-2012_corrected_MC_v2.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86-2012_exp_v2.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86-2013_exp_v2.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86-2014_exp_v2.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86-2012_GRL.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86-2013_GRL.npy ...\n",
      "Reading /data/ana/analyses/ps_tracks/version-002-p01/IC86-2014_GRL.npy ...\n",
      "<- /data/user/mrichman/csky_cache/ana/IC86_2012_2014.subanalysis.version-002-p01.npy    \n",
      "Setting up GFU_2015_2017...\n",
      "Reading /data/ana/analyses/gfu/version-002-p01/SplineMPEmax.MuEx.MC.npy ...\n",
      "Reading /data/ana/analyses/gfu/version-002-p01/SplineMPEmax.MuEx.IC86-2015.npy ...\n",
      "Reading /data/ana/analyses/gfu/version-002-p01/SplineMPEmax.MuEx.IC86-2016.npy ...\n",
      "Reading /data/ana/analyses/gfu/version-002-p01/SplineMPEmax.MuEx.IC86-2017.npy ...\n",
      "Reading /data/ana/analyses/gfu/version-002-p01/SplineMPEmax.MuEx.GRL.npy ...\n",
      "<- /data/user/mrichman/csky_cache/ana/GFU_2015_2017.subanalysis.version-002-p01.npy    \n",
      "Done.\n",
      "\n",
      "0:00:52.551711 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('ana setup'):\n",
    "    ana = cy.get_analysis(\n",
    "        cy.selections.repo,\n",
    "        'version-002-p01', cy.selections.PSDataSpecs.ps_7yr,\n",
    "        'version-002-p01', cy.selections.GFUDataSpecs.gfu_3yr,\n",
    "        dir=ana_dir, min_sigma=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll also look at the 2012-2014 configuration alone:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up Analysis for:\n",
      "IC86_2012_2014\n",
      "Setting up IC86_2012_2014...\n",
      "<- /data/user/mrichman/csky_cache/ana/IC86_2012_2014.subanalysis.version-002-p01.npy    \n",
      "Done.\n",
      "\n",
      "0:00:02.544739 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('ana_12_14 setup'):\n",
    "    ana_12_14 = cy.get_analysis(\n",
    "        cy.selections.repo,\n",
    "        'version-002-p01', cy.selections.PSDataSpecs.ps_7yr[-1],\n",
    "        dir=ana_dir, min_sigma=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From here on, we'll use `ana` and `mp_cpus=10` wherever relevant, unless otherwise specified:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "cy.CONF['ana'] = ana\n",
    "cy.CONF['mp_cpus'] = 10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Time-integrated analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can grab the coordinates from a (currently very short) list of pre-defined sources:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "src = cy.sources(*cy.selections.Coordinates.txs_0506_056[:2])\n",
    "cy.CONF['src'] = src"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We construct the single-source trial runner:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "tr = cy.get_trial_runner()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's the result:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(18.663818740946425, {'gamma': 2.0291608496153426, 'ns': 15.332733637279862})\n",
      "\n",
      "0:00:00.083892 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('unblind time integrated'):\n",
    "    print(tr.get_one_fit(TRUTH=True, flat=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We should also check the p-value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing 1000 background trials using 10 cores:\n",
      "       1000/1000 trials complete.   \n",
      "\n",
      "0:00:09.737622 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('bg trials time integrated'):\n",
    "    bg = cy.dists.Chi2TSD(tr.get_many_fits(1000, seed=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.05949407168683"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bg.sf_nsigma(tr.get_one_fit(TRUTH=True)[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We know that in practice more background scrambles are required to really pin down a significance at this level, but this is pretty close to the accepted value."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Untriggered single flare — Gaussian"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's the untriggered Gaussian single-flare fit for 2012-2014:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "conf_gauss = {\n",
    "    'time': 'utf',\n",
    "    'seeder': cy.seeding.UTFSeeder(),\n",
    "    'fitter_args': dict(_log_params='dt', _fmin_method='minuit'),\n",
    "    'concat_evs': True\n",
    "}\n",
    "tr_gauss = cy.get_trial_runner(conf_gauss, ana=ana_12_14)\n",
    "tr_gauss_ps_gfu = cy.get_trial_runner(conf_gauss, ana=ana)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TS                  23.136275598269947\n",
      "ns                  13.046011428453836\n",
      "t0                  57006.184382641426\n",
      "dt                  54.8228829919159\n",
      "gamma               2.1318043821995247\n",
      "\n",
      "0:00:00.801333 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('unblind gaussian flare'):\n",
    "    print(tr_gauss.format_result(tr_gauss.get_one_fit(TRUTH=True)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is similar to, though not exactly the same as, the official result.  We get better agreement with the original result (from psLab) if we disable detailed accounting for the GRL (in the signal time PDF):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TS                  22.023591770597093\n",
      "ns                  12.577074639465641\n",
      "t0                  57004.84556952451\n",
      "dt                  55.534056983245726\n",
      "gamma               2.1167311263655715\n",
      "\n",
      "0:00:00.559732 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('unblind gaussian flare'):\n",
    "    print(tr_gauss.format_result(tr_gauss.get_one_fit(TRUTH=True, use_grl=False)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We find the same flare in a full PS+GFU (i.e. multi-dataset) analysis:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TS                  20.796921427881177\n",
      "ns                  13.030879169522892\n",
      "t0                  57005.575320121075\n",
      "dt                  53.82572764881969\n",
      "gamma               2.1315463983861895\n",
      "\n",
      "0:00:19.714746 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('unblind gaussian flare (ps+gfu)'):\n",
    "    print(tr_gauss_ps_gfu.format_result(tr_gauss_ps_gfu.get_one_fit(TRUTH=True)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that `to_E2dNdE`, etc., return fluences rather than fluxes for time-dependent analyses:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.155e-04 TeV/cm2\n",
      "2.108e-04 TeV/cm2 (neglecting GRL)\n"
     ]
    }
   ],
   "source": [
    "fit_params = tr_gauss.get_one_fit(TRUTH=True, flat=False)[1]\n",
    "print('{:.3e} TeV/cm2'.format(tr_gauss.to_E2dNdE(E0=100, unit=1e3, **fit_params)))\n",
    "fit_params = tr_gauss.get_one_fit(TRUTH=True, flat=False, use_grl=False)[1]\n",
    "print('{:.3e} TeV/cm2 (neglecting GRL)'.format(tr_gauss.to_E2dNdE(E0=100, unit=1e3, **fit_params)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check the p-value:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing 1000 background trials using 10 cores:\n",
      "       1000/1000 trials complete.   \n",
      "\n",
      "0:01:06.775784 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('bg trials gaussian flare'):\n",
    "    bg_gauss = cy.dists.Chi2TSD(tr_gauss.get_many_fits(1000, seed=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5.8009972504182815e-05, 3.8543870175381363)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bg_gauss.sf(22.67), bg_gauss.sf_nsigma(22.67)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To match the official result, we need a trial factor 9.5yr/3yr:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.5624595771843457"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.norm.isf(bg_gauss.sf(22.67) * 9.5/3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, let's use a `SkyScanTrialRunner` to map out the region nearby region:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "RA, DEC = cy.trial.SkyScanTrialRunner.get_rectangular_grid(\n",
    "    src.ra[0], src.dec[0], np.radians(6), np.radians(2/2**3))\n",
    "sstr = cy.get_sky_scan_trial_runner(conf_gauss, src=src, ana=ana_12_14, scan_ra=RA, scan_dec=DEC)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scanning 625 locations using 10 cores:\n",
      "        625/625 coordinates complete.   \n",
      "\n",
      "0:00:41.679475 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('sky scan gaussian flare'):\n",
    "    scan = sstr.get_one_scan(TRUTH=True, mp_cpus=10, logging=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are plots of the best fit:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x1000 with 11 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, oaxs = plt.subplots(2, 3, figsize=(10,5), dpi=200)\n",
    "axs = np.ravel(oaxs)\n",
    "labels = r'TS $n_s$ $T$ $\\sigma_T$ $\\gamma$'.split()\n",
    "for (i, (ax, m)) in enumerate(zip(axs, scan[1:])):\n",
    "    pc = ax.pcolormesh(np.degrees(RA), np.degrees(DEC), m)\n",
    "    cb = fig.colorbar(pc, ax=ax)\n",
    "    cb.set_label(labels[i])\n",
    "    ax.set_xlim(ax.get_xlim()[::-1])\n",
    "    ax.set_aspect('equal')\n",
    "for ax in oaxs[-1]:\n",
    "    ax.set_xlabel(r'$\\alpha\\ \\ [^\\circ]$')\n",
    "for ax in oaxs[:,0]:\n",
    "    ax.set_ylabel(r'$\\delta\\ \\ [^\\circ]$')\n",
    "axs[-1].set_visible(False)\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Untriggered single flare — step function"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By default, untriggered flare analysis fits a Gaussian signal time PDF.  We can also perform a box-window search:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "conf_box = {\n",
    "    'time': 'utf',\n",
    "    'box': True,\n",
    "    'seeder': cy.seeding.UTFSeeder(),\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "tr_box = cy.get_trial_runner(conf_box, ana=ana_12_14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TS                  28.48249947602982\n",
      "ns                  13.45388195814707\n",
      "t0                  57013.62927017009\n",
      "dt                  151.6205902344591\n",
      "gamma               2.158717128472718\n",
      "\n",
      "0:00:00.865713 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('unblind box flare'):\n",
    "    print(tr_box.format_result(tr_box.get_one_fit(TRUTH=True)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.150e-04 TeV/cm2\n"
     ]
    }
   ],
   "source": [
    "fit_params = tr_box.get_one_fit(TRUTH=True, flat=False)[1]\n",
    "print('{:.3e} TeV/cm2'.format(tr_box.to_E2dNdE(E0=100, unit=1e3, **fit_params)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is slightly different from the official result, though the specific reasons have not yet been determined.  According to our implementation, this really is a slightly better fit than the official result:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(28.354866951770635, {'gamma': 2.1846017881006956, 'ns': 13.992756991949692})\n"
     ]
    }
   ],
   "source": [
    "print(tr_box.get_one_fit(TRUTH=True, t0=57004+13, dt=159, seeder=None, flat=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multiflare"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "multiflare analysis was recently added to csky.  For this we get a `MultiflareTrialRunner`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "mtr = cy.get_multiflare_trial_runner(src=src, ana=ana, max_dt=2000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    3 /     3 ... 0:00:00.013713 elapsed.\n",
      "  378 /   378 ... 0:00:01.559281 elapsed.\n",
      "   36 /    36 ... 0:00:00.075493 elapsed.\n",
      "  105 /   105 ... 0:00:00.290528 elapsed.\n",
      " 1275 /  1275 ... 0:00:04.851777 elapsed.\n",
      " 2278 /  2278 ... 0:00:10.272435 elapsed.\n",
      "\n",
      "0:00:17.267262 elapsed.\n"
     ]
    }
   ],
   "source": [
    "with time('unblind multiflare'):\n",
    "    result_multiflare = mtr.get_one_fit(TRUTH=True, logging=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ts                  66.96751539769558\n",
      "ns_total            41.32806989887902\n",
      "gamma_mean          2.2522205167780407\n",
      "ts_no_prior         101.13580074401915\n",
      "n_flare             8\n",
      "n_flare_tested      4075\n",
      "Flares(\n",
      "    src, TS, ns, gamma, TS0 =   0,  28.482,  13.45, 2.16  32.37 | [56937.818975 -> 57089.439565] ( 151.620590)\n",
      "    src, TS, ns, gamma, TS0 =   0,  14.041,   3.27, 1.67  14.73 | [57391.443843 -> 58018.871186] ( 627.427342)\n",
      "    src, TS, ns, gamma, TS0 =   0,   8.973,   1.36, 1.55  17.87 | [58018.871186 -> 58029.256078] (  10.384892)\n",
      "    src, TS, ns, gamma, TS0 =   0,   7.188,   4.27, 2.27  15.16 | [55204.982082 -> 55211.545663] (   6.563581)\n",
      "    src, TS, ns, gamma, TS0 =   0,   3.607,   7.23, 2.81   5.49 | [55808.334098 -> 55938.004387] ( 129.670289)\n",
      "    src, TS, ns, gamma, TS0 =   0,   2.041,   3.21, 2.03   5.52 | [57236.013093 -> 57391.443843] ( 155.430751)\n",
      "    src, TS, ns, gamma, TS0 =   0,   1.706,   1.54, 1.53   6.11 | [54666.328255 -> 54707.987934] (  41.659679)\n",
      "    src, TS, ns, gamma, TS0 =   0,   0.929,   6.99, 4.00   3.88 | [56156.855129 -> 56398.468652] ( 241.613523)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "print(mtr.format_result(result_multiflare))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can plot the neutrino \"lightcurve\" like so:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "lc = result_multiflare[-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAApUAAADkCAYAAADJsN/qAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAASdAAAEnQB3mYfeAAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3df3gb5YEv+u/YVhz/SiSnSSkFtpHNKU1CKTYp0MKlELmlnFKyp3Jyur/u6dlil/bZs3vvFruce7ew2bNt7dvu3nbvsrWzP+7e7mkbLNps6fYHdvCmkKdgYgUSO6WNLQOGGAy2ZP2WJc17/3gzE8uWbcmWNNLo+3meeZ5kZqR53++8kl/Nj3cUIYQAEREREdEmVBhdACIiIiIqfexUEhEREdGmsVNJRERERJvGTiURERERbRo7lURERES0aexUEhEREdGmsVNJRERERJvGTiURERERbRo7lURERES0aexUEhEREdGmsVNJRERERJvGTiURERFRifF4POjv74fP5zO6KDpFCCGMLgQRERERZa61tRVutxuTk5Ow2+1GFwcAj1QSERERlZT+/v6i6UguxSOVRERERCXEZrPhxIkTaG1t5ZFKIiIiIsped3c3brrpJrS0tBhdlBWqjC6AEXw+H06ePImrr74a1dXVRheHiIiIciwWi2F6ehp33HEHrFarIWW4ePEivF5v2mXJZBKVlZUr5u/cuRO7du1K+xqfz4fe3l6Mjo7mtJy5UpadypMnT+LgwYNGF4OIiIjy7Pjx47jvvvsKvt2LFy/iuvdejUBQTbt8y5YtWFxcXDH/4YcfxiOPPJL2Nffffz+cTmdRHqUEyrRTefXVVwOQDa25uTmr1waDQYyMjOCDH/wg6uvr81G8ksAcmIGGOUjMQWIOEnOQjMxhYmICBw8e1P/mF5rX60UgqGLgH3ehabclZdnkVBzt/3U2bT9k586dad/P7XbD5XKteuSzGJRlp1I75d3c3Iy9e/dm9VpVVfH+978f1dXVqKgo30tSmQMz0DAHiTlIzEFiDlIx5GD0ZW67d1fhuvemdipVyHuks+mHHDt27NL77U6Z39TUhJaWlqI4JV5ULb2zsxM2mw02mw3t7e0rBvTs7+/Xl3d2dhpSxoqKCtTU1JT1lwTAHABmoGEOEnOQmIPEHCTmACSEirhIpkwJkf6U+Fp6enrg9XoxNTWFqakpnDhxAgAwODio/9toRbOXW1tbAQAnTpzA6OgofD4fDhw4oC93uVzo7u7Wl58+fRrt7e0FL2c4HMazzz6LcDhc8G0XE+bADDTMQWIOEnOQmIPEHIAEVMSXTQlk36kEAKvVmjIBgN1uN+xGpOWK5vR3Z2cnOjo69P93d3ejra0NPp8PVqsV3d3d6Onp0S9OHRgYQFNTk768UBRFgcVigaIoBdtmMWIOzEDDHCTmIDEHiTlIzEGe6tZOdy+dZ0ZF06lc2qEE5OFcrSfu8/ng8XjgcDj05VrPfGhoCE6ns2DlrKmp0Y+qljPmwAw0zEFiDhJzkJiDxByAhBCIL3vOTCIHz52x2+0otufXFM3p76WGhobQ29uLhx56CIB8aDqAFSPG2+12fVmhJJNJ+P1+JJPJgm632DAHZqBhDhJzkJiDxBwk5gDEhcDisml5J9MsiqpT6Xa7oSgK2tra4HA40NXVBQCYn59Pu35jYyPm5ubWfM/Z2VmMj4+nTBMTEwCQco2H3+9HPB4HACwuLiIQCOjLAoGAPpaUz+fD8PAwQqEQADlcQiwWAwDE43H4/X79l0MoFEI0GgUAJBIJ+P1+qKqqbzsSiQBY+aGLRCJ62VRVhd/vRyKRAABEo1F920KIlHLHYjEEg8Gs66SVW5NJnUKhEIaHhzE7O2uaOmW7n+bm5vS2YJY6bWQ/aW3hrbfeMk2dNrKftO8Gn89nmjptZD+9/fbb+ufCLHXayH7SPhfa3ygz1Gkj+2lhYQHDw8Pwer0Fr9PS7RkpAQXxZVMC5rwcoKg6lS0tLfB6vfpt8W1tbZt+z0cffRT79u1LmbSBz8+dO6evd/LkSczMzAAApqencerUKX3ZqVOnMD09DUA26IqKCtTV1QEARkZGMDU1BeByJ0P7YLrdbly4cAEA9A+W9mE4e/Yszp8/DwD6l4/2YTx//jzOnj0LQH65DA8PY2FhAQBw4cIFuN1uAPIDvfRLa2pqCiMjI1nXaWZmBidPntSXZVKnuro63HjjjRgZGTFNnbLdT6+++ip27NiBuro609RpI/uprq4O73nPezA+Pm6aOm1kP83Pz+POO+9EMBg0TZ02sp9+/etf46qrrkJNTS3m5r2mqNNG9lNdXR3e+c53YnJy0jR12sh+ikajuPPOOzE7O1vwOi3dnpFUkX4yI0UU2wn5S3w+H2w2GwYGBmC329Ha2rri2oHW1lYcPnxYP6KZzuzsbMoRFODygKgjIyPYv38/APmrqaamBhaLBYuLi4jFYmhoaAAgfzVVV1djy5YtiMfjiEQi2LZtGwDZybRYLKiurtaXNTQ0QFEUhEIhVFZWYuvWrUgkEgiHw6ivr0dFRQXC4TAURUFNTY3+q7aurg6VlZWIRCIQQqC2thaqqiIYDKK2thZVVVWIRqNIJpOoq6uDEAKBQEAvdywWQzwe1weYZZ1YJ9aJdSpknVQh8G+nx/H46fM48/IMVCHwjvpafOrmffj0hz6A2iql5Opkxv1ULnUaHR3FLbfcgrGxsazHpM6F8fFx7Nu3Dz8Y3IXmZeNUTvw6jv/UNmtY2fKlaDuVAGCz2dDR0YGHHnoINpsNk5OTKddV2mw29PT0rLjJZz3ajt7IzoxEIjh//jz27NmDmpqarF5rJsyBGWiYg1TuOSwmEvjT7/wbhs+nXueuABAAttduxbc/+9u4/uorDClfoZV7e9AYmcNm/tbncvuPDV6Bpv+w7Ik6v4njUNsbputUFsXpb4/Hs+KGG5/PB5/Ph6amJlitVtjtdrhcrpTX+Hw+HDp0qKBlFUIgHo8X3R1XhcYcmIGGOUjlnsNXjg+v6FAC0AdO8Yej+Nzf/xBv+0OFLZhByr09aJgDkISCJCqWTea8prIohhRqbGxEa2srOjs74XA4YLVa0dnZCbvdrh+F7O7uTlne3t4Op9NZ8AE/a2trccsttxR0m8WIOTADDXOQyjmHN3wBPD4yvuY6AsBCOIpjz57FFz56a2EKZqBybg9LMQcgISoQFxUr5plRUdTKarVidHQUzz//PA4cOIDW1lZ9nqajowM9PT04cOAAmpqaYLfbMTAwUPCyqqqKSCSiXxhcrpgDM9AwB6mcc/jX0fMZHYlSADz+3Fj+C1QEyrk9LMUcgISoRFxUpUwJUWl0sfKiKDqVgOxYDgwMwOv1wuv1YmBgYMVRyK6uLni9XgghDOlQAvIi4SeffLJohiowCnNgBhrmIJVLDkeOHMHevXtTpm8d/aeMXisAzPqD2LtvX8rrjxw5kt9CG6Bc2sN6mAOgQrl0CvzypJr09HfRdCpLRW1tLW677TbU1tYaXRRDMQdmoGEOUrnkcOzYMX2sX52a+cDWQqjAkqOaExMTOHbsWK6KVzTKpT2shzkAcVGZdjKjorimspRUVVVhx44dRhfDcMyBGWiYg1ROOTQ3N6eMS/qzF3+NL/7LTzJ67Qebr8E/LXmtme58Xaqc2sNamIM8/b0oqpbNSxhUmvzikcosRaNRnDt3Th+1v1wxB2agYQ5SOedwYG8zdtTXZnRC79MfuiHv5SkG5dwelmIOQFxUpDlSac7ulzlrlUfJZBI+n6+sn2MKMAeAGWiYg1TOOViqKvHn7Q5AUdbsWN61twkH9jUXrFxGKuf2sBRzANQVwwlVQDVp98uctcqjuro63H777fpjGssVc2AGGuYglXsOH9nThL/5L/eisX7ltXMVigLnzfvw9d+7B5UV5fFnp9zbg4Y5AAmkufsbvKaSIAdyVVUVFRUVUBRz3r2VCebADDTMQWIOsmM5+N9/C0NjFzA6dRGJpIp3N27HJ1vfhyusDUYXr6DYHiTmoA0pVLlinhmVx0/GHAoEAvjxj3+MQCBgdFEMxRyYgYY5SMxBikbCiL/6G/yxYz/+vL0NHQc+WHYdSoDtQcMcgKSoSDuZEY9UZqmmpgY333xzWT/LFWAOADPQMAeJOUjMQWIOEnPQnqiz/EglO5UEwGKx4IorrjC6GIZjDsxAwxwk5iAxB4k5SMyBp79pDbFYDC+99BJisZjRRTEUc2AGGuYgMQeJOUjMQWIOuPQUneV3gJvz+lJ2KrMUj8dx8eJFxONxo4tiKObADDTMQWIOEnOQmIPEHC4fqVw6mfVIJU9/Z6m+vh533XWX0cUwHHNgBhrmIDEHiTlIzEFiDkACaa6pNOkxPXYqiYiIiPIkoVYhrlatmGdG5uwq55Hf78cTTzwBv99vdFEMxRyYgYY5SMxBYg4Sc5CYA6ACUKEsm8zJnF3lPKqpqcENN9xQ1sMjAMwBYAYa5iAxB4k5SMxBYg5AQq1EXK1cMc+M2KnMksViwTXXXGN0MQzHHJiBhjlIzEFiDhJzkJhDeY1Tac5a5dHi4iImJyexuLhodFEMxRyYgYY5SMxBYg4Sc5CYA6CiIu1kRuasVR7FYjFcuHChrMfcApgDwAw0zEFiDhJzkJiDxBwuHalUUyezHqnk6e8sNTQ04O677za6GIZjDsxAwxwk5iAxB4k5SMyhvE5/s1NJRERElCeqUKAKZcU8MzJnVzmPAoEAfvaznyEQCBhdFEMxB2agYQ4Sc5CYg8QcJOYAJEUFEpeeoqNNSR6pJACorq7Gtddei+rqaqOLYijmwAw0zEFiDhJzkJiDxBwuPaZx+ZBCfEwjAcCWLVvQ1NRkdDEMxxyYgYY5SMxBYg4Sc5CYg7x+cvk1lGa9ptKctcqjeDyOV199FfF43OiiGIo5MAMNc5CYg8QcJOYgMQdAiMvXVWqTEEaXKj/YqcxSJBLBiy++iEgkYnRRDMUcmIGGOUjMQWIOEnOQmIN8ek66yYx4+jtL27Ztw7333mt0MQzHHJiBhjlIzEFiDhJzkJiDdqNOxYp5ZsROJREREVGeqFg5hJBqTFHyzpxd5TwKBoN46qmnEAwGjS6KoZgDM9AwB4k5SMxBYg4ScwCSagUSy6akas7ulzlrlUcWiwVXXnklLBaL0UUxFHNgBhrmIDEHiTlIzEFiDpdPfy+dNnP6u7OzEzabDTabDe3t7fD5fDks7eawU5ml6upqXHfddWU95hbAHABmoGEOEnOQmIPEHCTmACRUZcWRyoS6sSfqtLa2AgBOnDiB0dFR+Hw+HDhwIJfF3RR2KrMUj8fxxhtvlPXwCABzAJiBhjlIzEFiDhJzkJgDIKCsHFIIG+tUdnZ2oq+vDy0tLbDb7eju7obb7S6ao5XsVGYpEongueeeK+vhEQDmADADDXOQmIPEHCTmIDEHefo73bQRHR0dKf8fHByE1WqF1WrNRVE3rWg6lT6fL+U6gc7OzhXr9Pf3r7m8EBoaGvCJT3wCDQ0Nhmy/WDAHZqBhDhJzkJiDxBwk5gCol27MWTqpObhRZ2hoCL29vXjooYdyUMrcKJohhdrb22G32zE6Ogq32437778fHo8Hg4ODAACXy4Xu7m6cOHECVqsV7e3taG9vx8DAQEHLqSgKKivNOWhpNpgDM9AwB4k5SMxBYg4ScwBUkWZIoUtP1JmYmFix/s6dO7Fr165V38/tduvXVjocDnR1deWusJtUFEcqPR4PTp8+jb6+PtjtdjidTpw4cQJDQ0Nwu90AgO7ubvT09OjXEQwMDMDlchX8OoJQKISnn34aoVCooNstNsyBGWiYg8QcJOYgMQeJOQCqSHOk8tLp74MHD2Lfvn0p06OPPrrm+7W0tMDr9WJ0dBQA0NbWlvc6ZKoojlTa7XYcPXo0ZV5LSwsA4PTp07Db7fB4PHA4HCmvsVqtGBoagtPpLFhZKysrYbVay/6XF3NgBhrmIDEHiTlIzEFiDkBSKEguO1Kp/f/48eNobm5OWbZz585139NqtaKlpQUDAwOw2WxwuVwF7Qutpig6lQBWDaOxsREejweA7EgupXU2C2nr1q24/vrrC7rNYsQcmIGGOUjMQWIOEnOQmAMghAKxrFOp/b+5uRl79+7d8HtrN+k8//zzRdGpLIrT3+n09/fDarXC6XRifn4+7TqNjY2Ym5tb831mZ2cxPj6eMmnXMITDYX09v9+vD3mwuLiIQCCgLwsEAlhcXAQg72Sbnp5GIpEAIJ8WEIvFAMihE/x+P4SQF0uEQiFEo1EAQCKRgN/vh6qq+ra1u+GSyST8fj+SyaS+Da1sqqrC7/fr24tGo/ppBCFESrljsVjKUwsyrZNWbk0mdUokEpidnYXX6zVNnbLdT4FAAK+//joSiYRp6rSR/ZRIJDAzM4OFhQXT1Gkj+ykcDmNubg6RSMQ0dVptP6mqumqd/H4/ZmZmkEgkMqqTxug65Xo/JRIJXLx4Ud+GGeq0kf0UjUYxNzeHUChU8DoVy1N8VFVBctmkbmCcSo/Hs+JAms/ng8/nQ1NTU66KuylF2an0eDzo7u7OyU04jz766IrrFQ4ePAgAOHfunL7eyZMnMTMzAwCYnp7GqVOn9GWnTp3C9PS0vsztdusfqpGREUxNTQEA5ubmMDw8rH8w3W43Lly4AABYWFjA8PCw/mE4e/Yszp8/D0B+EIaHh/UP4/nz53H27FkA8stleHhY/4N94cIF/TpTVVUxPDysd6ynpqYwMjKSdZ1mZmZw8uRJfVkmdQqHw/jlL3+JX/ziF6apU7b7aXx8HKdPn0Y4HDZNnTayn8LhMEZGRkxVp43sp6mpKTzzzDOYnp42TZ1W20+RSGTNOo2MjCAcDmdUJ+11Rtcp1/spHA7j+eefN1WdNrKfLl68iGeeeQYTExMFr9PS7RlJFQrUS3d865PIvlPZ2NiItrY29Pb2wu12w+Px6Dc5Lx9qyDCiyHi9XmG328XAwIA+b3R0VKQraktLi+jp6Vnz/d58800xNjaWMh0/flwAECMjI/p6CwsLYnFxUQghRCwWE36/X1/m9/tFLBYTQggRjUbF7OysSCaTQgghAoGAiEajQgghFhcXxcLCglBVVQghRDAYFJFIRAghRDweFwsLC/rrQqGQCIfDQgghEomEWFhYEIlEQgghRDgcFqFQSAghRDKZFAsLCyIejwshhIhEIiIYDAohhFBVNaXc0WhUBAKBrOuklVuTSZ2SyaTw+/3C5/OZpk7Z7qdgMCjm5uZEMpk0TZ02sp+SyaTwer0p5S71Om1kP0UiEREOh0U0GjVNndLtpz179ojrrrtu1ToFAgHh9XpFMpnMqE579uwRe/bsMd13RDKZFPPz83rZzFAnIbJve7FYTITDYREOhwtep2effVYAEGNjY8IIY2NjAoCwf/MB8b4fPpwy2b/5wIbK5vV6hdPpFFarVVitVuF0OoXX681TDbKnCHHpuHER8Pl8aG1tRV9fX8pNOT6fDzabDZOTkynXVdpsNvT09GTdQx8fH8e+ffswNja2qWsZiIjKjfadOT4+XpTvR6Qx+m+9tv3d//fnUX1N6hBBsVdnMfUnj5quH1I0p7+151cu71AC8kJUu90Ol8ulz/N4PPD5fDh06FBByxkOh/Hss8+mXI9ZjpgDM9AwB4k5SMxBYg4Sc8CKRzRqkxkVzd3fBw4cgMPhQGNjo35NBnB56KDu7m50dnbC4XDog587nc6CP5pIURRYLBYoijkbRKaYAzPQMAeJOUjMQWIOEnO4dPe3mv7ub7Mpik6lx+OB2+2G2+1Gb29vyrKenh50dXWho6NDP5rp8/ngdDoL/jQdAKipqdFHsi9nzIEZaJiDxBwk5iAxB4k5AEKs7EQWz4WHuVUUnUq73Y5MLu3s6uoy/HFEyWQSoVAIdXV15T2YK3NgBpcwB4k5SMxBYg4ScwBEmiGElh+5NIuiuaayVCwfXqFcMQdmoGEOEnOQmIPEHCTmcPn0d8rE098EAHV1dbjzzjtRV1dndFEMxRyYgYY5SMxBYg4Sc5CYAwBxaVo+z4TYqcxSZWUltm3bZnQxDMccmIGGOUjMQWIOEnOQmAP0o5PL55kRT39nKRKJYHR0VH+EVbliDsxAwxwk5iAxB4k5SMwBK099p+lkmkXOO5UvvPBCynM6zUYIgXg8ntGNRWbGHJiBhjlIzEFiDhJzkJgDACiAWDaBnUoAwJkzZ/DAAw+smP/1r38dlZWVaGlpgc1mw+c///mcFLDY1NbW4pZbbkFtba3RRTEUc2AGGuYgMQeJOUjMQWIOuHxN5fLJhLLuVH71q19dMe/EiRPo6urCgw8+CK/Xi5///Of4/ve/j2984xs5KWQxUVUVkUhEf4B9uWIOzEDDHCTmIDEHiTlIzAEQarpT4EaXKj+y7lQODQ2hra0tZV5fXx9aW1vxta99Ddu3b4fD4cDXvvY1fPvb385ZQYtFMBjEk08+iWAwaHRRDMUcmIGGOUjMQWIOEnOQmAMAVUk/mVDWnUqfzwe73Z4yz+Vy4fDhwynz9u/fD4/Hs7nSFaHa2lrcdttt5X0oH8wBYAYa5iAxB4k5SMxBYg7y6klFLJuMLlSeZN2pbGlpwejoqP7/EydOAACcTmfKeh6PZ0Xn0wyqqqqwY8cOVFWV92hMzIEZaJiDxBwk5iAxB4k5oGiPVD7++ON46qmncvqeWXcqv/SlL6GjowP/8A//gB/84Adob29HW1sb3vOe96SsNzg4iJaWllyVs2hEo1GcO3cO0WjU6KIYijkwAw1zkJiDxBwk5iAxBwDqKpPBPvWpT6G1tRVHjx7F3//93+dk5J6sfzo4nU589atfxYMPPgifzwen04mjR4+mrDM1NYX+/n5Tnv5OJpPw+XxIJpNGF8VQzIEZaJiDxBwk5iAxB4k5oKifqLN9+3bcf//9ePzxx3HXXXdh//79aG9vx0033bShQes3dDy6q6sLXV1dqy7fvXs3JicnVxy9NIO6ujrcfvvtRhfDcMyBGWiYg8QcJOYgMQeJOQCKUKAsO92tFMmzvx944AE0NTXB6XTi9OnTWFhYgMfjQV9fH+bm5vCOd7wDX/ziFzN+v6w7lR/72McwMDCwbg929+7d2b51SRBCQFVVVFRUQFGKo1EYgTkwAw1zkJiDxBwk5iAxBxTtkcqPfexjaGtrS+k0bt++HTfeeCNuvPHGDb3nhoYUKmeBQAA//vGPEQgEjC6KoZgDM9AwB4k5SMxBKlQOQgi8+toczr90Ea/PePO6rY1gewAUVUk7GW1wcBAOhyOn75n1kcryftQSUFNTg5tvvhk1NTVGF8VQzIEZaJiDxBwk5iDlOwdVFXjipy/g8R+58cr0nD7/vddegfaDrXB8ZE9RHBlke0DRHql0OByYmprCBz7wgZy9Z86f/W12FosFV1xxBSwWi9FFMRRzYAYa5iAxB4k5SPnMIZlU8Re9T+Cv/nYQr742l7LsNxfewP/4v/4Nf9P3VFEcBGJ7uDQupbpsMn7XoK+vD3/3d3+X0/fc0I06R48exfbt2zNa97Of/exGNlG0YrEYpqamsHv3blRXVxtdHMMwB2agYQ4Sc5CYg5TPHL7reg5P/eIlAMDyfqP238d/NIpm+y7c89Hrc7rtbLE9IP0QQkUwpNDu3bsxMDCAr3/961ndjLOWDXUqH3zwwYzWUxTFdJ3KeDyOixcv4qqrrirfDwiYA8AMNMxBYg4Sc5DylUM8noTr+Oi66ykK8NgPn8fH2/YZehqc7eHyU3SWzysG27dvz1mHEthgp9Ltdpv27u711NfX46677jK6GIZjDsxAwxwk5iAxBylfOZw5+yp8C+F11xMCmHrlbXhefgtNu3flvByZYnsAINI8QadIhhTKtQ11Ku12+4YGxSQiIiolR44cwbFjx4wuhq66zo6Gd2Q+7uOnnL+DeHQmjyXKzuHDh/HlL3/Z6GIU1qXrKJfPM6Osb9QphrvJjOT3+/HEE0/k5HFGpYw5MAMNc5CYg2S2HI4dO4aJiYmsX6eqKgKBAFQ1t70HoSayXD+e0+1na2kOExMTRdVBLxTt9PfyyYw4pFCWampqcMMNN5T38AhgDgAz0DAHiTlIZsyhubkZ4+PjWb0mHo9jZmYG73rXu3J653MwFMN/+t2/RWxx7c6logCNtjqc+NFTqKqqzNn2s7U0h1wOXVNSivRGnXzI+khlX19fWZ/6tlgsuOaaa8p6eASAOQDMQMMcJOYgMQcpXznU11XjY459664nBPDJez5gaIcSYHsAAAVpjlQaXag8ybpTef/99+ejHCVjcXERk5OTWFxcNLoohmIOzEDDHCTmIDEHKZ85dH7mDjTt3rnmOje+/xr8TvvNOd92ttge0oxRme4aS5Pg4OdZisViuHDhAmKxmNFFMRRzYAYa5iAxB4k5SPnMob6uGt/s+TTu+ej1sCw7Erl1qwXO+1rR8+efwhbLhu7FzSm2B1x+os7yyYSMb3ElpqGhAXfffbfRxTAcc2AGGuYgMQeJOUj5zqGhfiu6/+Tj+Nx//Qiefd6DYCgK6/Za3LLfjrra4hkPku0h/ZFJsx6pZKeSiIioRG3fVoOPHdhrdDFoDdpjGpfPMyOe/s5SIBDAz372MwQCAaOLYijmwAw0zEFiDhJzkJiDxBzA09+0uurqalx77bVl+7gpDXNgBhrmIDEHiTlIzEFiDiirwc/ZqczSli1b0NTUZHQxDMccmIGGOUjMQWIOEnOQmAPkUcnlnUiTHqnk6e8sxeNxvPrqq4jHjX1KgdGYAzPQMAeJOUjMQWIOEnPgkEK0hkgkghdffBGRSMToohiKOTADDXOQmIPEHCTmIDGH8npMY1F1Kt1u96qHyfv7+2Gz2WCz2dDZ2Vngkl22bds23HvvvWX9VCGAOQDMQMMcJOYgMQeJOUjMAZcf07h82gCfz4fOzs6i6A+lUxSdyt7eXiiKgvb2dng8nhXLXS4Xuru7ceLECYyOjuL06dNob283oKREREREmcvlYxq1vs/o6CiOHj2Kxx57DG1tbTkr62YVRaeyq6sLQgj09PSkXd7d3Y2enh60tLTAbrdjYGAALpcLPp+vwCUFgsEgnnrqKQSDwYJvu5gwB1Uce2cAACAASURBVGagYQ4Sc5CYg8QcJOaQu9PfHo8Hp0+fRl9fH+x2O5xOJ06cOIGhoSG43e7cF3wDiqJTuRafzwePxwOHw6HPs9vtsFqtGBoaKnh5LBYLrrzySlgsloJvu5gwB2agYQ4Sc5CYg8QcJOaAnJ3+ttvtOHr0aMq8lpYWAMDp06c3WcjcKPohhbTT4Xa7PWW+3W5Pe6o836qrq3HdddcVfLvFhjkwAw1zkJiDxBwk5iAtzUFRqqAoWxAMRlFfv9XgkhVOLh/T6HQ6085vbGzc2BvmWNEfqZyfn087v7GxEXNzc+u+fnZ2FuPj4ynTxMQEACAcDuvr+f1+fciDxcXFlNH/A4EAFhcX9dd4PB593WAwiFgsBkAOneD3+yGEPK4dCoUQjUYBAIlEAn6/H6qq6u+j3Q2XTCbh9/uRTCYByLvltLKpqgq/349EIgEAiEajCIVCAAAhREq5Y7FYyimGTOuklVuTSZ3i8Thef/11zM/Pm6ZO2e4nv9+PV155BfF43DR12sh+isfjmJ6eTrkcpdTrtJH9FAqF8MYbbyAcDpumTqvtJ1VVV63TwsKCPoRMJnXSGF2ntfaTqqpZ76d4PI5XXnlF30ax1alQbS8SiWDwyVF86YvfxTsbD+Gdjb+Ng//xr/Cnf/wdPP2LX+W1TsV0yn21U98TExMr+iizs7MZv29/fz+sVuuqnc1CK/pO5WY9+uij2LdvX8p08OBBAMC5c+f09U6ePImZmRkAwPT0NE6dOqUvO3XqFKanpwEAr732Gs6dO6d/4EZGRjA1NQUAmJubw/DwsP7BdLvduHDhAgD5RTs8PKx/GM6ePYvz588DkB+E4eFh/cN4/vx5nD17FoD8chkeHsbCwgIA4MKFC/q1E6qqYnh4WO9cT01NYWRkJOs6zczM4OTJk/qyTOoUiURw+vRpPP3006ap00b20wsvvIBIJGKqOmW7nyKRCNxuN55//nnT1Gkj++nll1/Gc889h9dee800dVptP0UikVXrdObMGZw5cwaRSCSjOmmvM7pOa+2naDSa9X6KRCJ44YUXirZOhWp7R789iJ6//DlOP/8yFOXy7SkvvjCNP/+zH6Lv0RN49dVX81KnpRkaaa1xKg8ePLiij/Loo49m9L4ejwfd3d0YGBjIY+mzowiti18EXC4X2tvbsbRIbrcbra2tWF7M1tZWHD58GF1dXWu+5+zsLN56662UeRMTEzh48CBGRkawf/9+APKXYE1NDSwWCxYXFxGLxdDQ0ABA/mqqrq7Gli1bsLi4iHA4jO3bt0NRFASDQVgsFlRXVyMejyMSiaChoQGKoiAUCqGyshJbt25FIpFAOBxGfX09KioqEA6HoSgKampqkEwmEQqFUFdXh8rKSkQiEQghUFtbC1VVEQwGUVtbi6qqKkSjUSSTSdTV1UEIgUAgoJc7FoshHo+jvr4+qzpp5daGfMikTtXV1VhcXEQ0GkVDQ4Mp6pTtfgqHw0gmk6ivr4cQwhR12sh+qqqqQiQSQTKZ1Mtd6nXayH7asmULqqqqkEwmEY1GTVGndPtp7969UFUVL7zwQto6BYNBKIqC2tpaJJPJdet06623ApDXhBXjd8S+ffugqirOnDmT1X6qqKhAKBSCoiioq6srqjoVqu399Cdu/FXvz9f8Gw0Af/QnbbjLcV3O6zQ6OopbbrkFY2Nj2Lt377rlyLXx8XHs27cP1zm7UNN4RcqyyPwbeMnVi+PHj6O5uTll2c6dO7Fr164139vn86G1tRU9PT1Fc5QSKIFOpc/ng81mw+TkZMp1lTabDT09Pejo6Mh6O9qONqqhERGVKu07c3x8vCjfL9eKvXzFSgiBz332H+CZnMVavQxFAa54lxX//D8fQEXFRgfaSc/ov/Xa9t/3qS7U2JZ1Kr1v4FeP926obFqHsq+vL+Um5mJQ9Ke/rVYr7HY7XC6XPs/j8cDn8+HQoUMFL08oFMLTTz+tnwooV8yBGWiYg8QcJOYglXsOU563MDmxdocSAIQAZi76MH5uujAFM4AiRNppI3w+Hw4cOFCUHUqgBO7+BuQ4lZ2dnXA4HLBarWhvb4fT6YTVai14WSorK2G1WlFZWVnwbRcT5sAMNMxBYg4Sc5DKPYe33wqsv9ISb2W5finJ5d3fBw4cgMPhQGNjY8rYlNpQi0Yrik5lf39/yqOGtIt5tdPgHR0deu/c5/PB6XQadmHq1q1bcf311xuy7WLCHJiBhjlIzEFiDlK551C9NbvuRfVW845jmatOpcfjgdvthtvtRm9vb8qynp6ede8xKYSiOP3d0dEBIcSKaamuri54vV4IIQy90ymRSGBubk4fQqFcMQdmoGEOEnOQmINU7jm897orUVdXndG6Fkslrn//1XkukcHEsmkD7HZ72r6SEKIoOpRAkXQqS0k4HMYzzzyTMsZlOcp3DrF4Ao+7x+Ds+y6uP/ItXH/kWzjU/z384Mw4FovkS5ptQWIOEnOQmINU7jls3WrB3ffckNG6B9r2Ydu2mjyXyEDphhPa4OnvYlcUp79LSX19PT760Y+iujqzX2Bmlc8c3g6GcP93fohfvfEWFFz+UXfu9Tdw9vU38N2RF3H0934btjpjv4TYFiTmIDEHiTlIzAH4/f/1Noye9uDlqbdXXedd77LiDzs+UrhCGUBRBRRVrJhnRjxSmaWKigrU1NSgoqK8o8tXDklVxQPf/Vf86g05tujSj53277GLb+IL3/8RVIM/lGwLEnOQmIPEHCTmANQ3bMVffev3cfsd700Z+Fxz863N+Obf/gFstjoDSlc4ClY+USe3gycVj/Jt7RsUDofx7LPPlu0pDU2+cnh64mWce/3Ndddzv3oRz069mtNtZ4ttQWIOEnOQmIPEHKSqKoGP3fNuHP2nz+A3kz/By6/+O/whN/75f34Of/m1Q2jcUW90EfNOSaafzIinv7OkKAosFkvaX13lJF85uEbHMl53YHQMH2r6rZxuPxtsCxJzkJiDxBwk5iBpObzziu2Yfl0+irG6dg/efVWjwSUrnFwOKVTs2KnMUk1NDVpbW40uhuHylcMr8768rJsPbAsSc5CYg8QcJOYgMYdLg58vv6ayeB5mmFM8/Z2lZDIJv9+PZNKkx64zlK8cLBkOFKwAsFQa23zZFiTmIDEHiTlIzEFiDlg5nNAmhhUqduxUZikUCmF4eLhsH72lyVcOLddcmdF6AkDrNe/O6bazxbYgMQeJOUjMQWIOEnMAcOnu76UTePc3AUBdXR3uvPNO1NWZ+2619eQrh0/vf39G6ykA/nOG6+YL24LEHCTmIDEHiTlIzOHS3d7Lx6o0Z5+SncpsVVZWYtu2bWX7PFdNvnJo2rkDf/jh9a+/+dz/cjOuaTT2OadsCxJzkJiDxBwk5iAxhzTDCQl2KumSSCSC0dFRRCIRo4tiqHzm8MW22/GFj9yS9prJLZWV+OO7PoT/dtetOd9uttgWJOYgMQeJOUjMQWIOAJIi/WRCvPs7S0IIxOPxFc8mLzf5zEFRFPzRnbfidz94A3545jz+z2/8NSAEvtL1pzj4gT2w1RbH47zYFiTmIDEHiTlIzEFiDtrp7+V3fxtUmDxjpzJLtbW1uOWWW4wuhuEKkUNjXS3+8Lab8NnbvwsA+Mzx7+Z1e9liW5CYg8QcJOYgMQeJOaQ/3W3WTiVPf2dJVVVEIhGoqklHLs0Qc2AGGuYgMQeJOUjMQWIOWHHnd7pngZsFO5VZCgaDePLJJxEMBo0uiqGYAzPQMAeJOUjMQWIOEnOAHFIomTqZdUghnv7OUm1tLW677TbU1tYaXRRDMQdmoGEOEnOQmIPEHCTmANmBXN6JZKeSAKCqqgo7duwwuhiGYw6ln8GvfbN4/OUXMRP2Y2tlFT58hR0fv+p9qK7M7muh1HPIFeYgMQeJOUjMgddU0hqi0SjOnTuHaDRqdFEMxRxKNwP/YhSfffr7+I9P9uMff/Mcfvrar/DDV87hi8/9K2574pv495mJrN6vVHPINeYgMQeJOUjM4dI1lctOf/OaSgIgn2Pq8/nK+zmmYA5AaWYQTcTxmV98d9WOo28xgo5njuHpNyYzfs9SzCEfmIPEHCTmIDEHXD79vXwyIZ7+zlJdXR1uv/12o4thOOZQmhk8NvUCXpy/uOpyAQBC4M9Gf4qn7vkCKhRl3fcsxRzygTlIzEFiDhJz0E5/l8c4lTxSmSUhBJLJZFkP5AowB6D0MhBC4F8mTmO9bqIA8FrIh6ff8GT8vqWUQ74wB4k5SMxBYg6QRyWXP03HpEcq2anMUiAQwI9//GMEAgGji2Io5lAaGRw5cgR79+6V000t8ATmkOlXWcdf/h/6a48cObLqeqWQQyEwB4k5SMxBYg6AoqppJzNipzJLNTU1uPnmm1FTUxyPCjQKcyiNDI4dO4aJiUvXT6Z5lvqaKuT6ExMTOHbs2KqrlUIOhcAcJOYgMQeJOYDP/qbVWSwWXHHFFUYXw3DMoXQyaG5uxvj4OBKqiv3/+g0E4rGMXvcXf/IgDn3rRuzdu3fN9Uolh3xjDlKx5xCMxPDkmd/gtbcWYKmqxE3XXoWbrr0KSgbXD2ej2HMoFOYAKBArr6nM+JxRaWGnMkuxWAxTU1PYvXs3qqurc/7+QggIAVRU5PYLLtfynUMpKLUMqioq4Nz9AfzTb55bd93aKgvuuXpPRu9bajnkC3OQijWHpKri2z95Ft95ahTRxYQ+v++nwO532vBnn3agpfmqnG2vWHMotKU5lC1VAMtPd/OaSgKAeDyOixcvIh6P5+w9VVXgl6d+gy998Xu4x9GDj37kK/id9r/Bd/7fp+GdL85HW+Ujh1JTihn8l2v3o8FSve7NOve/91bUWzL7Q1iKOeQDc5CKMQchBP7ie0M4+rPnUjqUmpff9KLjbx7H87+Zztk2izEHIzAHlNXpb3Yqs1RfX4+77roL9fX1OXm/RCKJrxw5jj97aACnRzyIx+VYXrNv+vHP//gL/OEf9OOl86/nZFu5lOscSlEpZvDuOiv+4fZPo8GyFQDSdi5/t6kVX9iT+RAgpZhDPjAHqRhzeOb8yzj+y/FVlwsAyaSKP/vOz5HM0Q0UxZiDEZgDgEunv5dOMOnpb3YqDdb3tyfw70+dX3V5IBDBQw9+H2+/Xb53zlFutbzjKgx+/AH87/s+gqvrbLBUVKLBUo2PX/U+/MtHfg+PtNyd0fiURKXisV+8uO46AsAb3gB+MTaV/wJReUmq6ScT4jWVWfL7/Th58iTuuOMObNu2bVPv5Z0P4kfHR9dcRwggEIjiieOj+MxnP7Kp7eVSLnMoVaWcwY6tdfj8ntvw+T23bfq9iikHIQQQPwPEXwSgApV2oPp2KEr+v+qKKQcjGZ3DkSNHVoxWUHXHZ6BUVGb0+v/25a9CvfBL/f/nz6/+o38tRudQLJbmULZ4TSWtpqamBjfccENOhkcYenIMyQx+rSgK8JMnXiiqwWNzmUOpYgZSseQgYs9CzH0SYv4/QwS+ChHogfB1Qrx1F0T4B3nffrHkYDSjc0gZRgsAFCXjDiUAKJW5+QFidA7FgjkAgJBHiJZOJj39zSOVWbJYLLjmmmty8l4XL3ozWk8IwOsNIRZLYOtWS062vVm5zKFUMQOpGHIQsZMQ3s8BSPMjTX0Twv8lQHih1P1h3spQDDkUg2LIQRtGS3PXQ32YD4Qzeu3nPvMH+Pwnvq3/f6NDDRVDDsWAOQBIJuW0fJ4J8UhllhYXFzE5OYnFxcVNv5elKvNfzwBQVVU8uyuXOZQqZiAZnYMQEQjfn0J2KNP9+pfzRKAXIjGRZnluGJ1DsSjGHO67JbPhsQDgEzdnvu5aijEHIzAHXHpM47LrKXn6uzj09/fDZrPBZrOhs7Oz4NuPxWK4cOECYrHMBpBey/U3ZP7rbc/ed6Mqy05oPuUyh1LFDCTDc4j8GyD8WP90koAIfz9vxTA8hyJRjDkcuv0G1FRb1h1Kq+3Ga3HNTmtOtlmMORiBOUBeT5luMqGSOv3tcrnQ3d2NEydOwGq1or29He3t7RgYGChYGRoaGnD33Xfn5L0+9OH/gB076vH224F1T7F88rdbc7LNXMllDqWKGUj5yCHdzRar+etHEjhwm7z2eC1CAK9N/H+45/e/l9H7Hj58GF/+8pczWhdge9AUYw7vatyGb3Z8En/c9yNEFlPHS1Qgf458wH4lHvndj+Zsm8WYgxGYA5ZcR7lsngmV1JHK7u5u9PT0oKWlBXa7HQMDA3C5XPD5fEYXbUMqqyrwv3XdAwBr3oTTut+OO+9a+1F5RGay4maLNdTWZPb9rChAXW1m21/veedUej743mvw/S/9rjxqueXytenvuaIRX2q/E/1/9CnUbd1iYAnJtFT18nWV2sQjlcby+XzweDxwOBz6PLvdDqvViqGhITidzoKUIxAI4NSpU/jwhz+MhoaGTb/fLbdeC2/gF9hefzMqla0py4QQiCy+jCNf6UJlEV1PCeQ+h1LEDKR85bD8ZovVqL4uIHo8g3dU0PiO92J8/Efrrrne887TYXuQijmH39plw38/fBce/NQd8AYjsFRVwlq3NefP/QaKO4dCWppD2UqqgKKunGdCJdOp9Hg8AGRHcim73a4vK4Tq6mpce+21OX2Wayz+Oma9x/H/fOv7OON+GfHFJK5413Z8468fRFINobq6OO74XiofOZQaZiAZnYNScx9ERp1KAaXmvryVw+gcikUp5GCpqsQua36f8FIKORQCcwAgVAihrphnRiXTqZyfn087v7GxEXNzcwUrx5YtW9DU1JSHd1Zx54G9uPPA5SMkvd8I5WE7uZG/HEoHM5AMz2HLrUDVdUDipTVWUgClAaj5VP6KYXQORYI5SMxBYg64fPf38nkb5Ha70d7ejsnJyU0WLPdKplO5UbOzs3jrrbdS5mnXaoXDl8ct8/v9qKmpgcViweLiImKxmH7KIhAIoLq6Glu2bEE4HMb09DTsdjssFguCwSAsFguqq6sRj8dx33334ZVXXgEAqKoKRVH0UyuqqqKiQp7G1q6hVBQFExMTsNvtSCaTqKysRCQS0ZdPTEzgfe97X8rrhBD6/5duo5DLACCRSKCiogKVlZUr6rRWfdMtW2v7mr179+a9vtmUW1VVJBIJWCwWKIqSVZ0KtWxychLNzc3w+/2or69HRUUFwuEwFEVBTU0NkskkQqEQ6urqUtpebW0tVFVFMBjU2+GePXtWbQvxeBxVVVU5LXdTUxP8fj8aGhqgKApCoRAqKyuxdetWJBIJhMNhvU6RrX+F6tBnoYiLEFj5THOBGlTY+iCwDQG/H7W1taiqqkI0GkUymURdXR2EEAgEAvogzcs/e+uVGwCSySSqqqry3vY283na7DLt+yoWi+nfe5FIRN9PCwsLePvtt3HNNddAUZSU/ZSu7S1tX7mo08TEhN6JWeu7XCu39rSb5d/lWp000Wg0bdtb7fOkqipeeeUV7Nq1C9u2bdM/T+u1PYvFglgshng8rj8vO9O/T5nWab3PU7bfEWvVyWKx4K233kJjY6OepfbZ0vbh7t278b3vfS/nddK+vwyXTALY/DiVvb296O7uLvgZ2mwU14V6a1jaIJean5/Hjh07Vn3do48+in379qVMBw8eBACcO3dOX+/kyZOYmZkBAExPT+PUqVP6slOnTmF6ehoA8Nprr+Gll15CJBIBAIyMjGBqSj4rdm5uDm+//bb+umg0qo/NpX0wtT8G0WhUH2LBbrfjxhtv1L9gz58/j7Nnz+Lw4cNoampCKBRC8lIDXFxcRDQa1bexfJlWLm1ZIpEAIP/oL+1Eh8NhxOPyLshEIqFvGwAikciKci+vkxAC0WgU4XA4bZ1UVUUoFNL/2MZiMb3cQois6rRUPuu03n5KV6dYLKb/USvG/dTc3Ix7770Xw8PDej3Onj2rP3ouFApheHh4RdvT6jc8PIx7770Xzc3Nq9ZJCIFYLJbTOjU3N+MTn/gEhoeH9bzdbjcuXLgAAFhYWEit09jbOHfxYaCuA8B2XFaNt0N34uxrj0DZ0qrXaWFhAQBw4cIFuN1uff8ODw9jbm4Ohw8fxtVXX511naLRKOLxeEHa3kY/T5vdT83Nzbj55ptTvveW7qczZ85gbGwMkUhk5X5K0/YOHjyI5ubmnNXpne98Jz7+8Y8DWPu7fGZmBidPntSXLf8u1+pUWVmJysrK1dveKp+nSCSC8fHxFZ+n9doeAExNTWFkZEQvW6Z/nzKtE7DO5ynL74i16jQzM4MXX3wRk5OTese/ubk5pe1FIpG81GlphkYSQkCoauq0gbu/u7q6IIRAT09PHkqZI6JEeL1eAUBMTk6mzLdaraKvr2/V17355ptibGwsZTp+/LgAIEZGRvT1FhYWxOLiohBCiFgsJvx+v77M7/eLWCwmhBBicXFRLCws6MsCgYCIRqMpy1RVFUIIEQwGRSQSEUIIEY/HxcLCgkgmk0IIIUKhkAiHw0IIIRKJhFhYWBCJREIIIUQ4HBahUEgIIUQymRQLCwsiHo8LIYSIRCIiGAwKIYRQVTWl3NFoVAQCAdaJdSrvOsXDwu99USSiLwk1GTRHncy4n1gn1inPdXr22WcFADE2NiaMMDY2JgCID235hPjo1t9LmT605RMbLtvAwIAo1u6bIkTpDJbU1NSEzs5OdHV1AZA37zQ1NcHr9cJqzXzA2vHxcezbtw9jY2MbusuTiIiIipvRf+u17d9adQ/qle0py4JiAb9M/GRDZXO5XGhvb9/Q0c58K6lrKru7u9HZ2QmHw6EPfu50OrPqUALQD/FnOg7eUuFwGOfOncP111+P2toMB70zIebADDTMQWIOEnOQmINkZA7a33ijn+YTSi5ALLumMozL16svt3PnTuzatasgZcu1kupUdnR0wOfz4cCBA/D5fHA6nRt6mo52rYZ2bSURERGZ0/T0NFpaWgq+XZvNhoaGBpwNnEq7fMuWLWn7IQ8//DAeeeSRPJcuP0rq9Heu+Hw+nDx5EldffXXWY2dNTEzg4MGDOH78OJqbm/NUwuLHHJiBhjlIzEFiDhJzkIzMIRaLYXp6GnfccUfWZzRz5eLFi/B6vWmXaSO+LLfekUqe/i4yVqsV9923uUGQm5ubeT0mmAPADDTMQWIOEnOQmINkVA5GHKFc6sorr8SVV15paBkKqWSGFCIiIiKi4sVOJRERERFtGjuVREREREWuv78fiqKgvb0dAFKexlUs2KnM0s6dO/Hwww9j586dRhfFUMyBGWiYg8QcJOYgMQeJOeROR0eH/gS3pVMxKcu7v4mIiIgot3ikkoiIiIg2jZ1KIiIiIto0diqJiIiIaNPKolPZ3t6u3yWlTU1NTWnX9Xg86O/vh8/ny+r1/f39sNlssNls6OzsXPG+6y0vhPXqUYh6miGHcmkPmu7ubjQ1NUFRFLS1taUsK4f2oFkth3JoD0NDQyuWadPS8pq9PWSSQzm0B013d7dezvb29pS/m4A52gNlSZQBp9MpOjo6xOTkZMqUTktLiwCQsny91w8MDAir1SpGR0fF5OSkaGlpEU6nM+PlhbJePfJdT7PkUC7tQQghHA6HaGlpEZOTk8Lr9YrBwUF9Wbm0ByHWzqEc2oPX6xWjo6MrJrvdrmdRDu0hkxzKoT0IIURHR0fKZ6Krq0u0tLRkXI9SyYGyUzadyq6urnXX6+vrE06nM22ncq3X2+120dfXp/9/cnJSABBerzej5YWyXj3yXU+z5FAu7aGvr09YrdZVl5dLe1gvh3JpD8v19PSIjo4O/f/l0h6WW55DubQHAGJ0dDRlntYJFMI87YGyUxanvzPV3d2Nhx56KKvX+Hw+eDweOBwOfZ7dbofVasXQ0NC6y0vFZutplhzWY6Ycenp6Vv08lFN7WCuH9Zgph6U8Hg+6u7vR09MDoLzaw1LLc1iP2XKwWq0r5s3Pz5dte6AyuaYSANxuN5qamvRrP5br7u7GTTfdtOrD51d7vcfjASAb/FJ2ux0ej2fd5YW2Xg75qqdZclhvuVly0L7U7XY72tvbYbPZ0NTUBLfbDaB82sN6Oaz3erPksFx3dzc6Ojr0TkW5tIflluew3uvNlENXVxc6Ozv16yh7e3tht9vhcDhM1x4oc2XTqRwaGkJfXx9GR0fh8XhSLrT3+Xzo7e1d89fmaq+fn59Pu35jYyPm5ubWXV5oa+Ww1vLN1tMsOay33Cw5aF/c3d3d6OzsxOjoKFpaWnDgwAEA5dMe1sthvdebJYelPB4PXC4Xuru79Xnl0h6WSpfDeq83Uw49PT1oaWmBzWaDoig4duwYRkdHAZivPVDmyqJTefjwYQwODsLhcMBut6OnpwdDQ0P6H4z7778fTqdz1aOU672+VKxXD7PUcz2bzcEsOa1VD+1LvaenR18+MDAAn88Hl8tlcMlza7M5lEN7WKqvrw92u33FUSSz2GwO5dIe+vv74XK5MDg4iNHRUTQ2Nq55VJfKQ5XRBSgEp9OZ8v+bbroJgPwVdtNNN8HlcsHr9W749enMz89jx44daGxsXHN5Ia1Vj46OjrzW0yw55DunQlqrHtq1TMt/aNntdjz//PM4fPhw2vc0W3tYLwen01kW7aGjo0Of73K5Vqy72f1tlhzWe71Z2sOhQ4fQ2dmpdzoBYHBwEIqiwOVyrfqDoxTbA2WnLI5ULqcdfWhsbMSxY8cAALt379bHwwKApqYmtLa2rvt67cOz/Feox+OB1Wpdd7mRltZjveWbradZclhvuVly0Mq5fNw5ANixY0fZtIf1csj09aWeg0a7xnT56eByaQ+a1XJY7/VmyeH06dMAsKKTrP3YMnN7oHUYfft5IQwMDKz4P5YMG+T1evVpdHRUABCDg4P60AXrvd5ut4uenh59ebqhEdZaXijr1SPf9TRLDuXSHlpaWlLKKYTQPxtClE97WC+HcmkPQggxODi4Yp6mXNqDEGvnUA7twev1VQE/0QAAA7VJREFUrjqkkDYMkFnaA2XH9J3KyclJvaFrnUar1Zoyrtjy9Zd+AWTy+r6+Pv0Dlm6Q1vWWF8J69ShEPc2QQ7m0ByEu/+EcHBwUk5OTwul0CofDkbN6miGHcmoPQlzuWKT7w14u7UGI1XMop/bgdDqF3W4Xo6Ojwuv16v/PVT2LIQfKnuk7lULID4jD4RAAVvz6Sbfu8l+gmby+p6dHWK1WASBtw19veSGsV49C1NMMOZRLexBC/vG02+3CarXmZX+bIYdyag89PT1irRNc5dIe1sqhnNpDV1eXsFqt+udieSfbDO2BsqMIIcRap8eJiIiIiNZTljfqEBEREVFusVNJRERERJvGTiURERERbRo7lURERES0aexUEhEREdGmsVNJRERERJvGTiURERERbRo7lURERES0aexUEhEREdGmsVNJREXP5XJBURS0tbWtuo7b7YaiKGhqalrzvRRFQW9vr/7//v5+KIqSMtlsNrS3t8Pj8eSsDkREZsdOJRGVjNOnT8Pn86Vd1tfXB6vVuuH3FkLo04kTJwAAra2tGBoa2vB7EhGVE3YqiahkOBwOPPbYY2mXPfbYYzh06FBOttPS0oKBgQF0dHSgvb09J+9JRGR27FQSUck4fPgw+vr6VswfGhpCY2Pjuqe+s9XT0wMAKafLiYgoPXYqiahkOJ1OuN3uFdc69vX1obOzMy/bdDgcOHbsWF7em4jITNipJKKS4nQ64XK5Uua5XC44nc68bG///v28YYeIKAPsVBJRSVl+CtzlcqGlpQV2uz1v21zt5iAiIrqMnUoiKilOpxMejwdutxsAcOzYsbyd+gaAyclJtLS05O39iYjMgp1KIio5HR0d+nWOLpcrZ3d9p/PYY4/B4XDk7f2JiMyiyugCEBFlq729HZ2dndi/fz8cDkfG41Nqp7EzXb+7uxsA8NBDD22soEREZYRHKomo5DgcDszPz6O7u3vNU98ejwednZ3weDzweDy4//779devRXtdb28vBgYGNjWoOhFRuWCnkohKUkdHBzwez5p3fTc2NgIA2tra0NTUBLfbjYGBgbQ39Sx9TGNrayvm5+cxOTnJU99ERBlShBDC6EIQERERUWnjkUoiIiIi2jR2KomIiIho09ipJCIiIqJNY6eSiIiIiDaNnUoiIiIi2jR2KomIiIho09ipJCIiIqJNY6eSiIiIiDaNnUoiIiIi2jR2KomIiIho09ipJCIiIqJNY6eSiIiIiDaNnUoiIiIi2jR2KomIiIho0/5/RSmlSWvdw/AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x240 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(6,2))\n",
    "hts = lc.to_hist('ts')\n",
    "hl.plot1d(ax, hts, color='k', lw=1)\n",
    "hgamma = lc.to_hist('gamma')\n",
    "m = hts.values > 0\n",
    "sc = ax.scatter(hts.centers[0][m], hts.values[m], c=hgamma.values[m], vmin=1, vmax=4, zorder=10)\n",
    "cb = fig.colorbar(sc)\n",
    "cb.set_label(r'$\\gamma$')\n",
    "ax.set_xlabel(r'MJD')\n",
    "ax.set_ylabel(r'TS')\n",
    "ax.set_ylim(-1,32)\n",
    "ax.grid()\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Timing summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ana setup                       | 0:00:52.551711\n",
      "ana_12_14 setup                 | 0:00:02.544739\n",
      "unblind time integrated         | 0:00:00.083892\n",
      "bg trials time integrated       | 0:00:09.737622\n",
      "unblind gaussian flare          | 0:00:00.559732\n",
      "unblind gaussian flare (ps+gfu) | 0:00:19.714746\n",
      "bg trials gaussian flare        | 0:01:06.775784\n",
      "sky scan gaussian flare         | 0:00:41.679475\n",
      "unblind box flare               | 0:00:00.865713\n",
      "unblind multiflare              | 0:00:17.267262\n",
      "------------------------------------------------\n",
      "total                           | 0:03:31.780676\n"
     ]
    }
   ],
   "source": [
    "print(timer)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Remarks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The results above are not identical to the official results.  Here is a summary of known differences and their explanations:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* **Time-integrated:** In the official results, no-longer-recommended settings were used in the energy PDF ratio construction: namely, the so-called \"skylab\" (as opposed to \"classic\") method of filling empty bins in the background energy PDF, and rectangular-kernel smoothing of the energy PDFs.  Those settings induce some undesireable biases.  Here, we do not reproduce those settings, and thus the results are slightly different.\n",
    "* **Gaussian flare:** The central time of the best-fit flare obtained here is slightly different than the one originally obtained.  This may be due to slightly different settings originally used in pslab, which so far does not (as far as I know) use GRL details in the signal time PDF; nor does it bin the energy or background space PDFs in a way that necessarily matches the standard used in csky, Skylab, and SkyLLH.  \n",
    "* **Box flare:** Again we obtain a result similar, though not identical, to the official result.\n",
    "* **Multiflare:** Thanks to detailed investigations by W. Luszczak, csky's multiflare implementation appears to be reliable and ready for use.  Here we recover to good precision the result he originally obtained with Skylab."
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}