sncosmo_vs_snana.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sncosmo\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import sncosmo_vs_snana as ss\n",
    "from sedfit.utils import register_bands_for_sncosmo\n",
    "import glob\n",
    "from scipy.interpolate import interp1d\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "register_bands_for_sncosmo()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "salt2par_sncosmo = ss.sncosmo_fitres_to_df('data/fitres_sncosmo/salt2pars_all_sncosmo.dat')\n",
    "salt2par_snana = ss.snana_fitres_to_df(glob.glob('data/fitres_snana/results/*.FITRES.TEXT'))\n",
    "mjdrange_snana = ss.get_snana_fitmjd_range(glob.glob('data/fitres_snana/results/*.LCPLOT.TEXT'))\n",
    "salt2par_snana = salt2par_snana.merge(mjdrange_snana,on=['CID','Survey'])\n",
    "# salt2par_sncosmo['Name_upper'] = ['SN'+x.upper() for x in salt2par_sncosmo['Name']]\n",
    "salt2par_snana['Name_upper'] = [x.upper() for x in salt2par_snana['CID']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "salt2par_sncosmo.to_csv('data/salt2par_all_sncosmo.csv')\n",
    "salt2par_snana.to_csv('data/salt2par_all_snana.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "common_list = ss.commonElements([salt2par_sncosmo.Name_upper.unique(),salt2par_snana.Name_upper.unique()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(796, 812, 796)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(common_list),len(salt2par_sncosmo.Name_upper.unique()),len(salt2par_snana.Name_upper.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "difflist = set(list(salt2par_sncosmo.Name_upper.unique())) ^ set(list(salt2par_snana.Name_upper.unique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "difflist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# print('SN1999BH' in salt2par_sncosmo.Name_upper.unique())\n",
    "# print('SN1999BH' in salt2par_snana.Name_upper.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plotlist = ['SN2005KU']\n",
    "plotlist = common_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "from matplotlib.backends.backend_pdf import PdfPages\n",
    "import gc\n",
    "meta_df = pd.read_csv('data/snmeta_all.csv',index_col=0) \n",
    "lc_df = pd.read_csv('data/snlc_all.csv',index_col=0)\n",
    "sn_per_pdf = 100\n",
    "\n",
    "pdfs = {}\n",
    "j = 0\n",
    "f = open('plots/plotlist.txt','w')\n",
    "for i,name in enumerate(np.sort(plotlist[0:])): #common_list:\n",
    "    if i==0 and j == 0:\n",
    "#         pdfs[0] = PdfPages('plots/plots_all_{:d}.pdf'.format(j))\n",
    "#         pdf = pdfs[0]\n",
    "        pdf = PdfPages('plots/plots_all_{:d}.pdf'.format(j))\n",
    "    if i//sn_per_pdf>0 and i%sn_per_pdf == 0:\n",
    "        pdf.close()\n",
    "        j += 1\n",
    "#         pdfs[j] = PdfPages('plots/plots_all_{:d}.pdf'.format(j))\n",
    "#         pdf = pdfs[j]\n",
    "        pdf = PdfPages('plots/plots_all_{:d}.pdf'.format(j))\n",
    "    print(i,name,j)\n",
    "    f.write('{} {} {}\\n'.format(i,name,j))\n",
    "    name = name.upper()\n",
    "    meta,lc = ss.get_photometry_single_sn(name,df_meta=meta_df,df_lc=lc_df)\n",
    "    for s in ss.commonElements([salt2par_sncosmo.loc[salt2par_sncosmo.Name_upper==name].Survey.unique(),\n",
    "                                salt2par_snana.loc[salt2par_snana.Name_upper==name].Survey.unique()]):\n",
    "        try:\n",
    "            data = ss.convert_lc_for_sncosmo(lc.loc[lc.Survey==s],select_filts='BVRIgriz')\n",
    "        except:\n",
    "            print(\"failed to get photometry for {}, {}\".format(name,s))\n",
    "            continue\n",
    "        psncosmo = salt2par_sncosmo.loc[(salt2par_sncosmo.Name_upper==name) & (salt2par_sncosmo.Survey==s)]\n",
    "        psnana = salt2par_snana.loc[(salt2par_snana.Name_upper ==name) & (salt2par_snana.Survey==s)]\n",
    "        models = []\n",
    "        text = ','.join([name,s])\n",
    "        text += '\\n'\n",
    "        for pname,source in zip([psncosmo,psnana],['sncosmo','snana']):\n",
    "            pmap = ss.get_lcpar_map(source=source)\n",
    "            model = ss.update_sncosmo_model(pname,pmap)\n",
    "            models.append(model)\n",
    "            text += '[{}] '.format(source)\n",
    "            for p,v in zip(model.param_names,model.parameters):\n",
    "                if p=='t0':\n",
    "                    text += '{}={:.2f}, '.format(p,v) \n",
    "                else:\n",
    "                    text += '{}={:.4f}, '.format(p,v)\n",
    "            if source == 'sncosmo':\n",
    "                text += 'fitbands={}, '.format(pname['bands'].values[0].strip())\n",
    "                text += '\\nchisq={:.2f}, '.format(pname['chisq'].values[0])\n",
    "                text += 'ndof={}, '.format(pname['ndof'].values[0])\n",
    "                text += 'reduced_chisq={:.2f}, '.format(pname['chisq'].values[0]/pname['ndof'].values[0])\n",
    "            elif source == 'snana':\n",
    "                text += '\\nchisq={:.2f}, '.format(pname['FITCHI2'].values[0])\n",
    "                text += 'ndof={}, '.format(pname['NDOF'].values[0])\n",
    "                text += 'reduced_chisq={:.2f}, '.format(pname['FITCHI2'].values[0]/pname['NDOF'].values[0])\n",
    "            text += 'first_fitmjd={:.2f}, '.format(pname['first_fitmjd'].values[0])\n",
    "            text += 'last_fitmjd={:.2f}'.format(pname['last_fitmjd'].values[0])\n",
    "            text += '\\n'\n",
    "\n",
    "        data_plot = data[[x.split('-')[1][0] in 'BVRIgriz' for x in data['Filter']]]           \n",
    "        fig = ss.compare_model_sncosmo(data=data_plot,models=models,model_label=['sncosmo','snana'],figtext=text,\n",
    "                                       fname=pdf, format='pdf')\n",
    "        plt.clf()\n",
    "        plt.close('all')\n",
    "        gc.collect()\n",
    "pdf.close()\n",
    "f.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mi/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/hist.py:309: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "  ax.hist(values, bins=bins, **kwds)\n",
      "/home/mi/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/hist.py:309: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "  ax.hist(values, bins=bins, **kwds)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mi/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/hist.py:309: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "  ax.hist(values, bins=bins, **kwds)\n",
      "/home/mi/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/hist.py:309: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "  ax.hist(values, bins=bins, **kwds)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mi/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:16: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "  app.launch_new_instance()\n",
      "/home/mi/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:16: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "  app.launch_new_instance()\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mi/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:30: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "/home/mi/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:30: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mi/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:30: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "/home/mi/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:30: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mi/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:35: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "/home/mi/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:36: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f2776468860>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for p in ['x1','c']:\n",
    "    for label,salt2p in zip(['sncosmo','snana'],[salt2par_sncosmo,salt2par_snana]): \n",
    "        if p == 'x1':\n",
    "            bins = np.linspace(-5,5,20)\n",
    "        else:\n",
    "            bins = np.linspace(-1,2,20)\n",
    "        salt2p[p].hist(bins=bins,normed=True,histtype='step',lw=2,label=label)\n",
    "    plt.title(p)\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "for label,salt2p in zip(['sncosmo','snana'],[salt2par_sncosmo,salt2par_snana]): \n",
    "    if label == 'sncosmo':\n",
    "        reduced_chisq = salt2p['chisq']/salt2p['ndof']\n",
    "    else:\n",
    "        reduced_chisq = salt2p['FITCHI2']/salt2p['NDOF']\n",
    "    plt.hist(reduced_chisq[np.isfinite(reduced_chisq)],bins=np.linspace(0,20,20),normed=True,histtype='step',lw=2,label=label)\n",
    "plt.title('reduced chisq')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "for p in ['x1','c']:\n",
    "    for label,salt2p in zip(['sncosmo','snana'],[salt2par_sncosmo,salt2par_snana]): \n",
    "        if label == 'sncosmo':\n",
    "            perr = salt2p[p+'_err']\n",
    "        else:\n",
    "            perr = salt2p[p+'ERR']\n",
    "        if p == 'x1':\n",
    "            bins = np.linspace(0,1,20)\n",
    "        else:\n",
    "            bins = np.linspace(0,0.2,20)\n",
    "        plt.hist(perr,bins=bins,normed=True,histtype='step',lw=2,label=label)\n",
    "    plt.title(p+'_error')\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "bins = np.linspace(0,10,20)\n",
    "plt.hist(salt2par_sncosmo['t0_err'],label='sncosmo',bins=bins,histtype='step',normed=True,lw=2)\n",
    "plt.hist(salt2par_snana['PKMJDERR'],label='snana',bins=bins,histtype='step',normed=True,lw=2)\n",
    "plt.title('t0_error')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = salt2par_sncosmo.merge(salt2par_snana,on=['Name_upper','Survey'],suffixes=['_sncosmo','_snana'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAVsUlEQVR4nO3de5Bed33f8fcH2djUGGzHsiNsgUxQUyxaLlXNLWQAN7Vjx5E7U7diklRN3XEzYzIw9ZDKaUpoGk3d9BJIwWFcoIgJxaMJEGu4C4EnbRxsZLCN5QsW2FiKhCXMxZiAqZ1v/3h+gsfrXe1Z7T67y8/v18wzzzm/yznfPTr67NE5u49SVUiS+vKUpS5AkrTwDHdJ6pDhLkkdMtwlqUOGuyR1yHCXpA4Z7pLUIcNdkjpkuEvzlORZSfYtdR3SOMNdRyXJMUPaniQuAD4x6Z14zDUXhnvnkqxO8qEkh5I8mOTts4z/l0nuTPKtJJ9M8pyxvkpyeZJ7gHtmapthu0nyh0kOJvlOktuSvKD1vTfJO5J8NMl3k9yY5GfG5q5LsiPJN5M8kOS3W/txSd6aZH97vTXJca3v1CQfSfLtNu//JHlK67svyZtaDd9L8u4kpyf5eNv/p5OcPLb/X06yu23r+iTPn/LlXQB87GiO9zI+5m9LsjfJQ0luTvKqsb63JNmW5H1t7u4k68f6Nyf5Suu7I8k/nu0YaAKqylenL2AFcCvwh8AJwPHAzx1h/MXAHuD5wDHA7wA3jPUXsAM4BXjaTG0zbPs84GbgJCBtH6ta33uBbwLntP2+H7i29Z0IHACuaPWfCLy09f0e8DngNGAlcAPwH1vffwLeCRzbXq8C0vrua/NOB84ADgJfAF4MHAd8BvjdNvZvA98DfqFt57faMXpq6z8W+Eara07He7ke89b/q8BPtb4rgK8Dx7e+twA/YPRNbUU71p8bm3sJ8CxGF4//rB2/VUv99+HJ9lryAnxN8A8XXg4cAo4ZOP7jwKVj608B/hp4Tlsv4LVT5jyhbYZtvxb4MvAy4ClT+t4LvGts/QLgrrb8OuCLM2zzK8AFY+vnAfe15d8DrgOeN828+4BfGVv/IPDHY+u/CfxZW/73wLYpx+SvgFe39XOBnUdzvJfrMZ9hW98CXtiW3wJ8eqzvbOD7R5h7C7Bhqf8+PNle3pbp22rga1X16MDxzwHe1m4/fJvRlV0YXd0etneaedO1PU5VfQZ4O/AO4IEk1yR5xtiQr48t/zXw9LGv4SszbPZZwNfG1r/W2gD+C6Mr4k8l+WqSzVPmPjC2/P1p1g/v/3H7qKq/YfT1Hj4mP7olw9yPNyzPY06SK9qtou+0up4JnHqEuccfvv+f5J8nuWXsa3rBlLlaBIZ73/YCz87wh257gX9dVSeNvZ5WVTeMjZnuM6IHfW50Vf1RVf19YB2j2x1vGljTz8zQt59ROB727NZGVX23qq6oqucCFwH/Jsm5Q+o80j6ShFGI/1VrugD46Fitczneh+csq2Pe7q//W+CfAidX1UnAdxh905lt7nOA/wm8HvipNvf2IXO1sAz3vt3E6H71VUlOSHJ8klceYfw7gSuTrANI8swklyxEIUn+QZKXJjmW0T3YHwCPDZj6EeCnk7yxPUA9MclLW98HgN9JsjLJqcCbgT9p+/ulJM9rYfxQ29eQ/U21Dbgwybmt9iuAR4AbkpwFHFdVd7Wxcz3esDyP+YnAo7RbTEneDDzjyFN+5ARG33gOtRp+ndGVuxaZ4d6xqnqM0VXr84D7gX2MHnDNNP7DwH8Grk3yEKMrrl9coHKeweiK7luMbnM8CPzX2SZV1XcZPcy8iNGtgHuA17Tu3wd2AbcBX2L0UPT3W99a4NPAw8BfAldX1fVzLbqq7mb0cPF/MHpwehFwUVX9ELiQH9+SmfPxbnOW3TEHPsnoWcCX27wfMOA2EEBV3QH8N0bH/AHg7wJ/MdfCNX+Hf3pA0hwl+Rjw9qr62KyDpUXmlbt09K4HPrvURUjTMdyfZJK8M8nD07zeuQDbftUM2354IWpfbqrqD6rq+7ON85hrKXhbRpI6tCw+l+LUU0+tNWvWLHUZkvQT5eabb/5GVa2crm9ZhPuaNWvYtWvXUpchST9Rknxtpj7vuUtShwx3SeqQ4S5JHTLcJalDhrskdchwl6QOGe6S1CHDXZI6ZLhLUoeWxW+oSsvBms0fnX1Qc99VF06wEmn+vHKXpA4Z7pLUIcNdkjpkuEtShwx3SeqQ4S5JHTLcJalDhrskdchwl6QOGe6S1CHDXZI6ZLhLUocMd0nq0KBwT3JSkj9NcleSO5O8PMkpSXYkuae9nzw2/soke5LcneS8yZUvSZrO0Cv3twGfqKq/A7wQuBPYDOysqrXAzrZOkrOBjcA64Hzg6iQrFrpwSdLMZg33JM8Afh54N0BV/bCqvg1sALa2YVuBi9vyBuDaqnqkqu4F9gDnLHThkqSZDblyfy5wCPhfSb6Y5F1JTgBOr6oDAO39tDb+DGDv2Px9re1xklyWZFeSXYcOHZrXFyFJerwh4X4M8BLgj6vqxcD3aLdgZpBp2uoJDVXXVNX6qlq/cuXKQcVKkoYZEu77gH1VdWNb/1NGYf9AklUA7f3g2PjVY/PPBPYvTLmSpCFmDfeq+jqwN8nPtqZzgTuA7cCm1rYJuK4tbwc2JjkuyVnAWuCmBa1aknREQ/+D7N8E3p/kqcBXgV9n9I1hW5JLgfuBSwCqaneSbYy+ATwKXF5Vjy145ZKkGQ0K96q6BVg/Tde5M4zfAmyZR12SpHnwN1QlqUOGuyR1yHCXpA4Z7pLUIcNdkjpkuEtShwx3SeqQ4S5JHTLcJalDhrskdchwl6QOGe6S1CHDXZI6ZLhLUocMd0nqkOEuSR0y3CWpQ4a7JHXIcJekDhnuktQhw12SOmS4S1KHDHdJ6tCgcE9yX5IvJbklya7WdkqSHUnuae8nj42/MsmeJHcnOW9SxUuSpjeXK/fXVNWLqmp9W98M7KyqtcDOtk6Ss4GNwDrgfODqJCsWsGZJ0izmc1tmA7C1LW8FLh5rv7aqHqmqe4E9wDnz2I8kaY6GhnsBn0pyc5LLWtvpVXUAoL2f1trPAPaOzd3X2h4nyWVJdiXZdejQoaOrXpI0rWMGjntlVe1PchqwI8ldRxibadrqCQ1V1wDXAKxfv/4J/ZKkozfoyr2q9rf3g8CHGd1meSDJKoD2frAN3wesHpt+JrB/oQqWJM1u1nBPckKSEw8vA/8IuB3YDmxqwzYB17Xl7cDGJMclOQtYC9y00IVLkmY25LbM6cCHkxwe/7+r6hNJPg9sS3IpcD9wCUBV7U6yDbgDeBS4vKoem0j1kqRpzRruVfVV4IXTtD8InDvDnC3AlnlXJ0k6Kv6GqiR1yHCXpA4Z7pLUIcNdkjpkuEtShwx3SeqQ4S5JHTLcJalDhrskdchwl6QOGe6S1CHDXZI6ZLhLUocMd0nqkOEuSR0y3CWpQ4a7JHXIcJekDhnuktQhw12SOmS4S1KHDHdJ6tDgcE+yIskXk3ykrZ+SZEeSe9r7yWNjr0yyJ8ndSc6bROGSpJnN5cr9DcCdY+ubgZ1VtRbY2dZJcjawEVgHnA9cnWTFwpQrSRpiULgnORO4EHjXWPMGYGtb3gpcPNZ+bVU9UlX3AnuAcxamXEnSEEOv3N8K/BbwN2Ntp1fVAYD2flprPwPYOzZuX2t7nCSXJdmVZNehQ4fmXLgkaWazhnuSXwIOVtXNA7eZadrqCQ1V11TV+qpav3LlyoGbliQNccyAMa8EfjnJBcDxwDOS/AnwQJJVVXUgySrgYBu/D1g9Nv9MYP9CFi1JOrJZr9yr6sqqOrOq1jB6UPqZqvpVYDuwqQ3bBFzXlrcDG5Mcl+QsYC1w04JXLkma0ZAr95lcBWxLcilwP3AJQFXtTrINuAN4FLi8qh6bd6WSpMHmFO5VdT1wfVt+EDh3hnFbgC3zrE2SdJT8DVVJ6pDhLkkdMtwlqUOGuyR1yHCXpA4Z7pLUIcNdkjpkuEtShwx3SeqQ4S5JHTLcJalDhrskdchwl6QOGe6S1CHDXZI6ZLhLUocMd0nqkOEuSR0y3CWpQ4a7JHXIcJekDhnuktQhw12SOjRruCc5PslNSW5NsjvJf2jtpyTZkeSe9n7y2Jwrk+xJcneS8yb5BUiSnmjIlfsjwGur6oXAi4Dzk7wM2AzsrKq1wM62TpKzgY3AOuB84OokKyZRvCRperOGe4083FaPba8CNgBbW/tW4OK2vAG4tqoeqap7gT3AOQtatSTpiAbdc0+yIsktwEFgR1XdCJxeVQcA2vtpbfgZwN6x6fta29RtXpZkV5Jdhw4dms/XIEmaYlC4V9VjVfUi4EzgnCQvOMLwTLeJabZ5TVWtr6r1K1euHFatJGmQOf20TFV9G7ie0b30B5KsAmjvB9uwfcDqsWlnAvvnXakkabAhPy2zMslJbflpwD8E7gK2A5vasE3AdW15O7AxyXFJzgLWAjctdOGSpJkdM2DMKmBr+4mXpwDbquojSf4S2JbkUuB+4BKAqtqdZBtwB/AocHlVPTaZ8iVJ05k13KvqNuDF07Q/CJw7w5wtwJZ5VydJOir+hqokdchwl6QOGe6S1CHDXZI6ZLhLUocMd0nqkOEuSR0y3CWpQ4a7JHXIcJekDhnuktQhw12SOmS4S1KHDHdJ6pDhLkkdMtwlqUOGuyR1yHCXpA4Z7pLUIcNdkjpkuEtShwx3SerQrOGeZHWSzya5M8nuJG9o7ack2ZHknvZ+8ticK5PsSXJ3kvMm+QVIkp5oyJX7o8AVVfV84GXA5UnOBjYDO6tqLbCzrdP6NgLrgPOBq5OsmETxkqTpzRruVXWgqr7Qlr8L3AmcAWwAtrZhW4GL2/IG4NqqeqSq7gX2AOcsdOGSpJnN6Z57kjXAi4EbgdOr6gCMvgEAp7VhZwB7x6bta22SpEUyONyTPB34IPDGqnroSEOnaatptndZkl1Jdh06dGhoGZKkAQaFe5JjGQX7+6vqQ635gSSrWv8q4GBr3wesHpt+JrB/6jar6pqqWl9V61euXHm09UuSpjHkp2UCvBu4s6r++1jXdmBTW94EXDfWvjHJcUnOAtYCNy1cyZKk2RwzYMwrgV8DvpTkltb228BVwLYklwL3A5cAVNXuJNuAOxj9pM3lVfXYglcuSZrRrOFeVf+X6e+jA5w7w5wtwJZ51CVJmgd/Q1WSOmS4S1KHDHdJ6pDhLkkdMtwlqUOGuyR1yHCXpA4Z7pLUIcNdkjpkuEtShwx3SerQkA8OkzTFms0fHTTuvqsunHAl0vS8cpekDhnuktQhw12SOmS4S1KHDHdJ6pDhLkkdMtwlqUOGuyR1yHCXpA4Z7pLUIcNdkjo0a7gneU+Sg0luH2s7JcmOJPe095PH+q5MsifJ3UnOm1ThkqSZDblyfy9w/pS2zcDOqloL7GzrJDkb2Aisa3OuTrJiwaqVJA0ya7hX1Z8D35zSvAHY2pa3AhePtV9bVY9U1b3AHuCcBapVkjTQ0d5zP72qDgC099Na+xnA3rFx+1rbEyS5LMmuJLsOHTp0lGVIkqaz0A9UM01bTTewqq6pqvVVtX7lypULXIYkPbkdbbg/kGQVQHs/2Nr3AavHxp0J7D/68iRJR+Now307sKktbwKuG2vfmOS4JGcBa4Gb5leiJGmuZv1v9pJ8AHg1cGqSfcDvAlcB25JcCtwPXAJQVbuTbAPuAB4FLq+qxyZUuyRpBrOGe1W9boauc2cYvwXYMp+iJEnz42+oSlKHDHdJ6pDhLkkdMtwlqUOGuyR1yHCXpA4Z7pLUoVl/zl36Sbdm80eXugRp0XnlLkkdMtwlqUOGuyR1yHCXpA4Z7pLUIcNdkjpkuEtShwx3SeqQ4S5JHTLcJalDhrskdcjPltGy0tvnwMzl67nvqgsnWImebLxyl6QOGe6S1CHDXZI6NLFwT3J+kruT7EmyeVL7kSQ90UQeqCZZAbwD+AVgH/D5JNur6o5J7E/LW28PSSdlEsfJh7RPXpP6aZlzgD1V9VWAJNcCG4CJhPvQvxSTONEn8dMQS71N9aO3P/e5/B2eRC4sZdbMVapq4Tea/BPg/Kr6V23914CXVtXrx8ZcBlzWVn8WuHseuzwV+MY85k+Kdc2Ndc2Ndc1Nj3U9p6pWTtcxqSv3TNP2uO8iVXUNcM2C7CzZVVXrF2JbC8m65sa65sa65ubJVtekHqjuA1aPrZ8J7J/QviRJU0wq3D8PrE1yVpKnAhuB7RPalyRpionclqmqR5O8HvgksAJ4T1XtnsS+mgW5vTMB1jU31jU31jU3T6q6JvJAVZK0tPwNVUnqkOEuSR1a1uE+20cYZOSPWv9tSV4ydO6E6/qVVs9tSW5I8sKxvvuSfCnJLUl2LXJdr07ynbbvW5K8eejcCdf1prGabk/yWJJTWt8kj9d7khxMcvsM/Ut1fs1W11KdX7PVtVTn12x1Lfr5lWR1ks8muTPJ7iRvmGbMZM+vqlqWL0YPYr8CPBd4KnArcPaUMRcAH2f0c/UvA24cOnfCdb0COLkt/+Lhutr6fcCpS3S8Xg185GjmTrKuKeMvAj4z6ePVtv3zwEuA22foX/Tza2Bdi35+Daxr0c+vIXUtxfkFrAJe0pZPBL682Pm1nK/cf/QRBlX1Q+DwRxiM2wC8r0Y+B5yUZNXAuROrq6puqKpvtdXPMfo5/0mbz9e8pMdritcBH1igfR9RVf058M0jDFmK82vWupbo/BpyvGaypMdrikU5v6rqQFV9oS1/F7gTOGPKsImeX8s53M8A9o6t7+OJB2emMUPmTrKucZcy+u58WAGfSnJzRh/BsFCG1vXyJLcm+XiSdXOcO8m6SPK3gPOBD441T+p4DbEU59dcLdb5NdRin1+DLdX5lWQN8GLgxildEz2/lvN/szfrRxgcYcyQuUdr8LaTvIbRX76fG2t+ZVXtT3IasCPJXe3KYzHq+gKjz6J4OMkFwJ8BawfOnWRdh10E/EVVjV+FTep4DbEU59dgi3x+DbEU59dcLPr5leTpjL6ZvLGqHpraPc2UBTu/lvOV+5CPMJhpzCQ//mDQtpP8PeBdwIaqevBwe1Xtb+8HgQ8z+ifYotRVVQ9V1cNt+WPAsUlOHTJ3knWN2ciUfzJP8HgNsRTn1yBLcH7NaonOr7lY1PMrybGMgv39VfWhaYZM9vxa6AcJC/Vi9K+KrwJn8eOHCuumjLmQxz+QuGno3AnX9WxgD/CKKe0nACeOLd/A6NMzF6uun+bHv7h2DnB/O3ZLerzauGcyum96wmIcr7F9rGHmB4SLfn4NrGvRz6+BdS36+TWkrqU4v9rX/T7grUcYM9Hza8EO7iRejJ4mf5nRk+N/19p+A/iNsQP4jtb/JWD9keYuYl3vAr4F3NJeu1r7c9sf1K3A7iWo6/Vtv7cyehD3iiPNXay62vq/AK6dMm/Sx+sDwAHg/zG6Wrp0mZxfs9W1VOfXbHUt1fl1xLqW4vxidKusgNvG/pwuWMzzy48fkKQOLed77pKko2S4S1KHDHdJ6pDhLkkdMtwlqUOGuyR1yHCXpA79f0uT5P8X2XyhAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for p in ['x1','c']:\n",
    "    plt.hist(df[p+'_sncosmo']-df[p+'_snana'],label=p,bins=np.linspace(-0.1,0.1,30))\n",
    "    plt.title('{}_sncosmo-{}_snana'.format(p,p))\n",
    "    plt.show()\n",
    "    plt.hist(df[p+'_err']/df[p+'ERR'],bins=np.linspace(0,2,30))\n",
    "    plt.title('{}_err_sncosmo/{}_err_snana'.format(p,p))\n",
    "    plt.show()\n",
    "    plt.hist((df[p+'_sncosmo']-df[p+'_snana'])/df[p+'_err'],bins=np.linspace(-1,1,30),label='err = sncosmo_err')\n",
    "    plt.hist((df[p+'_sncosmo']-df[p+'_snana'])/df[p+'ERR'],bins=np.linspace(-1,1,30),label='err = snana_err')\n",
    "    plt.title('({}_sncosmo-{}_snana)/err'.format(p,p))\n",
    "    plt.legend()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.errorbar(df['c_sncosmo'],df['c_snana'],xerr=df['c_err'],yerr=df['cERR'],fmt='o')\n",
    "x = np.linspace(-0.5,1,100)\n",
    "plt.plot(x,x,color='g',lw=2,ls='--')\n",
    "plt.xlabel('c_sncosmo')\n",
    "plt.ylabel('c_snana')\n",
    "plt.xlim(-0.5,1.1)\n",
    "plt.ylim(-0.5,1.1)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Unnamed: 0', 'c_sncosmo', 'c_c_cov', 'c_err', 'c_mwebv_cov',\n",
       "       'c_t0_cov', 'c_x0_cov', 'c_x1_cov', 'c_z_cov', 'chisq',\n",
       "       ...\n",
       "       'x0ERR', 'COV_x1_c', 'COV_x1_x0', 'COV_c_x0', 'NDOF', 'FITCHI2',\n",
       "       'FITPROB', 'x0_offset', 'first_fitmjd_snana', 'last_fitmjd_snana'],\n",
       "      dtype='object', length=103)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Name</th>\n",
       "      <th>mwebv</th>\n",
       "      <th>MWEBV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [Name, mwebv, MWEBV]\n",
       "Index: []"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[(np.absolute(df['mwebv']-df['MWEBV']))>1.e-5][['Name','mwebv','MWEBV']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.loc[df['c_err']>0.5,'Name_upper'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.loc[np.absolute(df['c_sncosmo']-df['c_snana'])>0.2,'Name_upper'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plotlist = ['SN2005KU']\n",
    "# plotlist = df.loc[df['c_err']>0.5,'Name_upper'].values\n",
    "# plotlist = df.loc[np.absolute(df['c_sncosmo']-df['c_snana'])>0.2,'Name_upper'].unique()\n",
    "plotlist = df.loc[(df['t0_err']>2) | (df['PKMJDERR']>2),'Name_upper'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib.backends.backend_pdf import PdfPages\n",
    "pdf = PdfPages('plots/large_t0_err.pdf')\n",
    "meta_df = pd.read_csv('data/snmeta_all.csv',index_col=0) \n",
    "lc_df = pd.read_csv('data/snlc_all.csv',index_col=0)\n",
    "for i,name in enumerate(plotlist): #common_list:\n",
    "    print(i,name)\n",
    "    name = name.upper()\n",
    "    meta,lc = ss.get_photometry_single_sn(name,df_meta=meta_df,df_lc=lc_df)\n",
    "    for s in ss.commonElements([salt2par_sncosmo.loc[salt2par_sncosmo.Name_upper==name].Survey.unique(),\n",
    "                                salt2par_snana.loc[salt2par_snana.Name_upper==name].Survey.unique()]):\n",
    "        try:\n",
    "            data = ss.convert_lc_for_sncosmo(lc.loc[lc.Survey==s],select_filts='BVRIgriz')\n",
    "        except:\n",
    "            print(\"failed to get photometry for {}, {}\".format(name,s))\n",
    "            continue\n",
    "        psncosmo = salt2par_sncosmo.loc[(salt2par_sncosmo.Name_upper==name) & (salt2par_sncosmo.Survey==s)]\n",
    "        psnana = salt2par_snana.loc[(salt2par_snana.Name_upper ==name) & (salt2par_snana.Survey==s)]\n",
    "        models = []\n",
    "        text = ','.join([name,s])\n",
    "        text += '\\n'\n",
    "        for pname,source in zip([psncosmo,psnana],['sncosmo','snana']):\n",
    "            pmap = ss.get_lcpar_map(source=source)\n",
    "            model = ss.update_sncosmo_model(pname,pmap)\n",
    "            models.append(model)\n",
    "            text += '[{}] '.format(source)\n",
    "            for p,v in zip(model.param_names,model.parameters):\n",
    "                if p=='t0':\n",
    "                    text += '{}={:.2f}, '.format(p,v) \n",
    "                else:\n",
    "                    text += '{}={:.4f}, '.format(p,v)\n",
    "            if source == 'sncosmo':\n",
    "                text += 'fitbands={}, '.format(pname['bands'].values[0].strip())\n",
    "                text += '\\nchisq={:.2f}, '.format(pname['chisq'].values[0])\n",
    "                text += 'ndof={}, '.format(pname['ndof'].values[0])\n",
    "                text += 'reduced_chisq={:.2f}, '.format(pname['chisq'].values[0]/pname['ndof'].values[0])\n",
    "            elif source == 'snana':\n",
    "                text += '\\nchisq={:.2f}, '.format(pname['FITCHI2'].values[0])\n",
    "                text += 'ndof={}, '.format(pname['NDOF'].values[0])\n",
    "                text += 'reduced_chisq={:.2f}, '.format(pname['FITCHI2'].values[0]/pname['NDOF'].values[0])\n",
    "            text += 'first_fitmjd={:.2f}, '.format(pname['first_fitmjd'].values[0])\n",
    "            text += 'last_fitmjd={:.2f}'.format(pname['last_fitmjd'].values[0])\n",
    "            text += '\\n'\n",
    "            \n",
    "            data_chisq = data[[x.split('-')[1][0] in 'BVRIgriz' for x in data['Filter']]]\n",
    "\n",
    "        fig = ss.compare_model_sncosmo(data=data,models=models,model_label=['sncosmo','snana'],figtext=text,\n",
    "                                       fname=pdf, format='pdf')\n",
    "pdf.close()"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}