6_Warming_impact.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6. Investigating the impact of warming on model predictions\n",
    "## Forecasting Crop Yields on a national scale (FACYnation)\n",
    "\n",
    "### by Raphael Shirley (University of Sussex)\n",
    "\n",
    "In this notebook we take the previous model and investigate increases in temperature.\n",
    "\n",
    "## Investigating impact of temperature increases\n",
    "\n",
    "In this notebook we take the two dimensional Gaussian fitted to the regional data and check for the impact of temperature increases which should depend on where current temperatures lie with respect to the peak yield response.\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Rerun model from notebook 3\n",
    "First lets generate the same set of samples from the posterior."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import pylab as plt\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import pystan\n",
    "\n",
    "import time\n",
    "\n",
    "from numpy import exp,arange\n",
    "from pylab import meshgrid,cm,imshow,contour,clabel,colorbar,axis,title,show\n",
    "\n",
    "from xidplus.stan_fit import stan_utility\n",
    "\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compute the productivity trend and find a final baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "yields_for_comp = pd.read_table('./Crop_data_files/Maize_yield_obs_timeseries.csv', sep=',')\n",
    "anoms_for_comp = pd.read_table('./Crop_data_files/Maize_median_yield_anoms.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "states = ['Indiana', 'Illinois', 'Ohio', 'Nebraska', 'Iowa', 'Minnesota']\n",
    "state = 'Nebraska'\n",
    "\n",
    "years = np.arange(1960, 2008)\n",
    "anoms_by_year = [anoms_for_comp[anoms_for_comp['Region'] == 'Maize_Spring_USA_{}'.format(state)][str(year)].iloc[0] \n",
    "                 for year in years]\n",
    "yields_by_year = [yields_for_comp[yields_for_comp['Region'] == 'Maize_Spring_USA_{}'.format(state)][str(year)].iloc[0] \n",
    "                 for year in years]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1a277128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "\n",
    "for state in states:\n",
    "    anoms_by_year = [anoms_for_comp[anoms_for_comp['Region'] == 'Maize_Spring_USA_{}'.format(state)][str(year)].iloc[0] \n",
    "                 for year in years]\n",
    "    yields_by_year = [yields_for_comp[yields_for_comp['Region'] == 'Maize_Spring_USA_{}'.format(state)][str(year)].iloc[0] \n",
    "                 for year in years]\n",
    "    \n",
    "    plt.scatter(years, anoms_by_year, s = 1.0, c ='b')\n",
    "    plt.scatter(years, yields_by_year, s = 1.0, c ='r')\n",
    "    \n",
    "plt.scatter([-1], [0], s = 2.0, c ='r', label = 'actual values')\n",
    "plt.scatter([-1], [0], s = 2.0, c ='b', label = 'anomalies')\n",
    "\n",
    "ax.set_xlabel('year')\n",
    "ax.set_ylabel('Yield [tonnes ha$^{-1}$]')\n",
    "ax.set_xlim(1960, 2007)\n",
    "plt.legend()\n",
    "\n",
    "plt.savefig(\"./figs/yields_vs_anoms_noline.png\")\n",
    "plt.savefig(\"./figs/yields_vs_anoms_noline.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_anoms = np.array([])\n",
    "all_yields = np.array([])\n",
    "all_years = np.array([])\n",
    "for state in states:\n",
    "    anoms_by_year = [anoms_for_comp[anoms_for_comp['Region'] == 'Maize_Spring_USA_{}'.format(state)][str(year)].iloc[0] \n",
    "                 for year in years]\n",
    "    yields_by_year = [yields_for_comp[yields_for_comp['Region'] == 'Maize_Spring_USA_{}'.format(state)][str(year)].iloc[0] \n",
    "                 for year in years]\n",
    "    all_yields = np.hstack([all_yields, yields_by_year])\n",
    "    all_anoms = np.hstack([all_anoms, anoms_by_year])\n",
    "    all_years = np.hstack([all_years, years])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1a381470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "\n",
    "import scipy\n",
    "slope, intercept, rval, pval, sterr = scipy.stats.linregress(all_years, all_yields)\n",
    "\n",
    "x = np.array([1950., 2020.])\n",
    "y = slope * x + intercept\n",
    "plt.plot(x, y, c ='r')\n",
    "   \n",
    "plt.scatter( all_years,  all_yields, s = 15.0, alpha=0.5, c ='r', label = 'Actual values')\n",
    "plt.scatter( all_years,  all_anoms, s = 15.0, alpha=0.5, c ='b', label = 'Anomalies')\n",
    "\n",
    "    \n",
    "#plt.scatter([-1], [0], s = 5.0, c ='b', label = 'anomalies')\n",
    "#plt.scatter([-1], [0], s = 5.0, c ='r', label = 'actual values')\n",
    "\n",
    "\n",
    "ax.set_xlabel('Year')\n",
    "ax.set_ylabel('Yield [tonnes ha$^{-1}$]')\n",
    "ax.set_xlim(1960, 2007)\n",
    "ax.set_ylim(-4, 12)\n",
    "plt.legend()\n",
    "\n",
    "plt.savefig(\"./figs/yields_vs_anoms.png\")\n",
    "plt.savefig(\"./figs/yields_vs_anoms.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "predicted value in 2019 is 11.151293266210757\n",
      "latest value in 2007 is 9.748564345238123\n"
     ]
    }
   ],
   "source": [
    "print('predicted value in 2019 is {}'.format(slope * 2019 + intercept) )\n",
    "print('latest value in 2007 is {}'.format(slope * 2007 + intercept) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_months = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']\n",
    "months = all_months[3:9]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def take_mean_anom(table):\n",
    "    means_by_year = []\n",
    "    for year in np.arange(1980, 2008):\n",
    "        means_by_year.append(np.mean(\n",
    "            [table[str(month)][table['Unnamed: 0'] ==  year].iloc[0]  for month in months ]\n",
    "        ) )\n",
    "    return np.array(means_by_year)\n",
    "    \n",
    "precip_mean_anoms = np.array([])\n",
    "temp_mean_anoms   = np.array([])\n",
    "yields_from_1980  = np.array([])\n",
    "\n",
    "for state in states:\n",
    "    precip_month_anoms = pd.read_table(\n",
    "        'Crop_data_files/maize_met_anoms/Maize_Spring_USA_{}_precip_anom_real.csv'.format(state))\n",
    "    temp_month_anoms = pd.read_table(\n",
    "        'Crop_data_files/maize_met_anoms/Maize_Spring_USA_{}_temp_anom_real.csv'.format(state))\n",
    "\n",
    "    precip_mean_anoms = np.hstack([precip_mean_anoms,\n",
    "                                   take_mean_anom(precip_month_anoms)])\n",
    "    temp_mean_anoms = np.hstack([temp_mean_anoms,\n",
    "                                 take_mean_anom(temp_month_anoms)])\n",
    "    yields_from_1980 = np.hstack([ yields_from_1980,\n",
    "                                  ([yields_for_comp[\n",
    "                                      yields_for_comp['Region'] == 'Maize_Spring_USA_{}'.format(state)\n",
    "                                  ][str(year)].iloc[0] \n",
    "                 for year in np.arange(1980, 2008)]) \n",
    "                                 ])\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1fce3710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Scatter plot of the sources.\n",
    "fig, axis = plt.subplots()\n",
    "\n",
    "\n",
    "im = axis.scatter(temp_mean_anoms, \n",
    "                  precip_mean_anoms, \n",
    "                  c=yields_from_1980,  \n",
    "                  cmap=\"viridis\",\n",
    "                  s=50.0,\n",
    "                 alpha = 1.)#vmin=0.0, vmax=360.,\n",
    "axis.set_xlabel(\"Average monthly temperature anomaly [K]\")\n",
    "axis.set_ylabel(\"Average monthly precipitation anomaly [mm]\")\n",
    "#axis.axis('equal')\n",
    "\n",
    "cbar = fig.colorbar(im) # adding the colobar on the right\n",
    "cbar.set_label('Yield [tonnes ha$^{-1}$]')\n",
    "# Optionally add a colorbar\n",
    "#cax, _ = mpl.colorbar.make_axes(ax)\n",
    "#cbar = mpl.colorbar.ColorbarBase(cax, cmap=cmap)\n",
    "\n",
    "\n",
    "plt.savefig('./figs/real_data_scatter.pdf', bbox_inches='tight')\n",
    "plt.savefig('./figs/real_data_scatter.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a2021bc50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "\n",
    "test = np.full([10, 10], np.nan)\n",
    "\n",
    "ts = np.linspace(-2,2, 10)\n",
    "ps = np.linspace(-60,60, 10)\n",
    "\n",
    "t_width = np.absolute(ts[1] - ts[0])\n",
    "p_width = np.absolute(ps[1] - ps[0])\n",
    "\n",
    "X,Y = meshgrid(ts, ps) # grid of point\n",
    "\n",
    "def z_func(X,Y):\n",
    "    in_bin = temp_mean_anoms > (X - t_width/2)\n",
    "    in_bin = np.bitwise_and(in_bin , temp_mean_anoms < (X + t_width/2) )\n",
    "    in_bin = np.bitwise_and(in_bin , precip_mean_anoms > (Y - p_width/2) )\n",
    "    in_bin = np.bitwise_and(in_bin , precip_mean_anoms < (Y + p_width/2) )\n",
    "    yields_temp = in_bin*yields_from_1980\n",
    "    yields_temp[np.isclose(yields_temp , 0)] = np.nan\n",
    "    return np.nanmean(yields_temp)\n",
    "\n",
    "\n",
    "z_func = np.vectorize(z_func)\n",
    "#Lets use the median\n",
    "\n",
    "Z = z_func(X, Y)\n",
    "\n",
    "im = imshow(np.flip(Z, axis=0),cmap=\"viridis\", \n",
    "            extent=[ts[0],ts[-1] ,ps[0], ps[-1]],\n",
    "            aspect=\"auto\") # drawing the function\n",
    "# adding the Contour lines with labels\n",
    "#cset = contour(Z,arange(-1,1.5,0.2),linewidths=2,cmap=cm.Set2)\n",
    "#clabel(cset,inline=True,fmt='%1.1f',fontsize=10)\n",
    "cbar = colorbar(im) # adding the colobar on the right\n",
    "cbar.set_label('Yield [tonnes ha$^{-1}$]')\n",
    "# latex fashion title\n",
    "#title('$z=(1-x^2+y^3) e^{-(x^2+y^2)/2}$')\n",
    "\n",
    "ax.set_xlabel('Monthly mean $T_{inc}$ [K]')\n",
    "ax.set_ylabel('Monthly mean $P_{inc}$ [mm]')\n",
    "\n",
    "plt.savefig('./figs/real_data_hist.pdf', bbox_inches='tight')\n",
    "plt.savefig('./figs/real_data_hist.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Indiana & 1.93 & 35.25 & 1.68 \\\\ \n",
      "Illinois & 2.02 & 36.39 & 1.78 \\\\ \n",
      "Ohio & 1.87 & 31.05 & 1.67 \\\\ \n",
      "Nebraska & 2.21 & 30.9 & 1.95 \\\\ \n",
      "Iowa & 2.23 & 33.55 & 1.92 \\\\ \n",
      "Minnesota & 2.39 & 29.06 & 2.06 \\\\ \n"
     ]
    }
   ],
   "source": [
    "\n",
    "for state in states:\n",
    "    precip_month_anoms = pd.read_table(\n",
    "        'Crop_data_files/maize_met_anoms/Maize_Spring_USA_{}_precip_anom_real.csv'.format(state))\n",
    "    temp_month_anoms = pd.read_table(\n",
    "        'Crop_data_files/maize_met_anoms/Maize_Spring_USA_{}_temp_anom_real.csv'.format(state))\n",
    "    \n",
    "    yields_by_year = [yields_for_comp[yields_for_comp['Region'] == 'Maize_Spring_USA_{}'.format(state)\n",
    "                                     ][str(year)].iloc[0] \n",
    "                 for year in years]\n",
    "    \n",
    "    #temp_mean =  np.nanmean(np.array(temp_month_anoms[all_months]))\n",
    "    temp_variance = np.nanstd(np.array(temp_month_anoms[all_months])) \n",
    "    #precip_mean =  np.nanmean(np.array(precip_month_anoms[all_months]))\n",
    "    precip_variance = np.nanstd(np.array(precip_month_anoms[all_months])) \n",
    "    yield_variance = np.nanstd(np.array(yields_by_year)) \n",
    "    \n",
    "    print(\"{} & {} & {} & {} \\\\\\\\ \".format(state, \n",
    "                                 #round(temp_mean, 2), \n",
    "                                 round(temp_variance, 2), \n",
    "                                 #round(precip_mean, 2),\n",
    "                                 round(precip_variance, 2),\n",
    "                                 round(yield_variance, 2)\n",
    "                                     )\n",
    "         )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[3.38904,\n",
       " 4.048019999999999,\n",
       " 3.7342199999999997,\n",
       " 4.33044,\n",
       " 3.64008,\n",
       " 3.82836,\n",
       " 4.76976,\n",
       " 4.51872,\n",
       " 5.146319999999999,\n",
       " 5.46012,\n",
       " 5.3346,\n",
       " 5.20908,\n",
       " 5.83668,\n",
       " 5.83668,\n",
       " 3.89112,\n",
       " 4.3932,\n",
       " 3.70284,\n",
       " 6.276,\n",
       " 6.52704,\n",
       " 6.276,\n",
       " 6.08772,\n",
       " 6.9036,\n",
       " 7.091880000000001,\n",
       " 5.27184,\n",
       " 6.715319999999999,\n",
       " 7.2174,\n",
       " 7.65672,\n",
       " 7.97052,\n",
       " 4.64424,\n",
       " 7.845,\n",
       " 7.78224,\n",
       " 7.5312,\n",
       " 7.15464,\n",
       " 4.3932,\n",
       " 8.91192,\n",
       " 7.468439999999999,\n",
       " 7.845,\n",
       " 8.28432,\n",
       " 9.60228,\n",
       " 9.414,\n",
       " 9.1002,\n",
       " 8.1588,\n",
       " 9.85332,\n",
       " 9.16296,\n",
       " 9.97884,\n",
       " 10.92024,\n",
       " 10.10436,\n",
       " 9.16296]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yields_by_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  7.04357383,  -3.49763475, -22.64754188, -55.05743983,\n",
       "        -29.40544654,  -9.16025199, -43.60732175,  81.16157186,\n",
       "         14.75415008, -25.40226811, -24.22793364,  -7.81111636],\n",
       "       [ -6.54554422,  18.68405776,  -5.26810127,  -0.30902153,\n",
       "        -36.18021167,  23.29038607,  14.99584592,  25.03589283,\n",
       "        -48.88323717,  17.31911413, -14.13875667,  -3.36095884],\n",
       "       [  2.00493247,  -6.95511034,  -8.00935947, -33.54022235,\n",
       "         38.44451158, -58.33524483,   0.67849683,  -8.13566772,\n",
       "         15.20244632,  34.96313155,  33.06978919,  26.98527072],\n",
       "       [  2.03839958,  -2.64086805,  14.88356838, -27.39703391,\n",
       "         -3.60715643,  16.24437972, -15.62137768,  -9.96020125,\n",
       "          9.31834654,   3.89175307,  33.48001108,  -7.91115338],\n",
       "       [ -2.1079126 ,  11.12650188, -18.08212931,  24.28584145,\n",
       "        -19.45655228,  68.7699285 , -29.59731297, -32.1375665 ,\n",
       "        -22.40299604,  73.01998064,  -2.55428164,  14.99461405],\n",
       "       [ -5.7040543 ,  -4.09335594,  11.874451  ,   8.80617254,\n",
       "        -14.70573446, -34.17884995, -47.96945729,  18.57806363,\n",
       "         55.36402558,  -8.74401536,  -3.90670877,  -7.12208408],\n",
       "       [ -3.40560376,  -0.61774834,   1.31103744,  63.4656789 ,\n",
       "         -8.90233536,  -5.08726273,  48.96892448,  -1.61237929,\n",
       "         86.58590879, -11.54183567, -10.93050596,  -9.84169114],\n",
       "       [ -4.15836572,  -6.71425729,   4.79028476, -52.59022404,\n",
       "        -25.49366018, -56.61337919,  54.70578359, -13.93234807,\n",
       "        -12.49674634, -38.38678204,  17.09396852,   4.09912591],\n",
       "       [ -2.40176868,  -8.00699925,  -6.19206869, -28.09353731,\n",
       "        -45.98124793, -85.44932333, -59.4859999 ,   1.46997565,\n",
       "         18.38361778, -47.97961072,   9.43895255,  -1.30182617],\n",
       "       [ -1.99751097,  -7.29079259, -12.278793  ,  -9.613727  ,\n",
       "        -29.37173244, -46.55093801,  -6.19778198,  -7.78544928,\n",
       "        -24.80071354, -33.8437902 , -13.30675675, -10.48440791],\n",
       "       [ -2.63784874,  -5.3850222 ,  37.41608491,   4.56498924,\n",
       "          5.65415291,  58.07565308,  28.5308457 ,  -5.21990852,\n",
       "        -38.78532599,  -5.01007339, -19.71865498,  -5.48442814],\n",
       "       [ -4.54175055,  -4.52482245,  23.49242797,  39.14810033,\n",
       "         44.6515592 ,   7.55171355,  24.54930304, -22.12599802,\n",
       "         30.18227118, -22.01187118,  29.32396214,   1.94581618],\n",
       "       [  5.70675661,  -0.59381936,   4.61743821,  -8.00992441,\n",
       "        -49.40701616, -22.99392439,  35.47367798, -23.84809342,\n",
       "         -0.31163542,  -8.64407345,  21.13008259,   3.02194143],\n",
       "       [  1.43595366,  -5.70014493,  -4.08774657,   2.80818225,\n",
       "         60.01740731,  76.08374453,  62.3502346 ,  55.98234613,\n",
       "        -15.2965957 , -38.84864448,   2.37456641,  -3.00518491],\n",
       "       [ -1.67180053,  -5.85035022, -21.83242592,   9.66013985,\n",
       "        -48.5555558 ,   2.46144657,   8.66862114,   2.11456225,\n",
       "         15.58227606,   5.67772655,  -2.57240084,  -9.04603181],\n",
       "       [  0.98192307,  -5.03897748,  18.51931461,   3.12186093,\n",
       "         -1.00985841, -48.26956531, -17.6849993 ,  14.84388029,\n",
       "         -9.81252468,  20.705713  , -13.86907136,  -5.57636909],\n",
       "       [ 11.15770242,  -2.21893844, -15.57026557, -49.34505274,\n",
       "          3.16922076, -12.53960481,  -1.2792466 , -40.59865237,\n",
       "        -28.12809261,  16.228598  ,  39.80588832,  -5.33256937],\n",
       "       [  6.77899287,  -6.0541219 , -13.04805513, -34.13730653,\n",
       "        -39.97628024, -19.89684719, 105.65713272,  -2.88417799,\n",
       "         -4.37659596,  -6.74803185, -19.72349618,  -9.71680896],\n",
       "       [  0.81259856,  11.73196166,  44.69577543, -10.05532557,\n",
       "          9.55456249,  37.43044719, -26.33006405,  16.27169598,\n",
       "        -43.39577843,  23.15993765,   6.36174619,  -4.19245836],\n",
       "       [  1.2900568 ,  -2.89729262, -14.17903379,  26.79675885,\n",
       "         42.61489544,   2.20561539,  20.7864218 ,  -2.16318882,\n",
       "         -9.18973562, -36.34876909, -17.51339766,  -9.4960667 ],\n",
       "       [ -3.83823099,   7.86261502,  -6.10853295, -37.38308385,\n",
       "         20.49272622,   3.86216082,  15.84486488, -40.31526978,\n",
       "        -33.33733062, -18.29579976,  33.89347411,  -9.10477209],\n",
       "       [  9.29416439,   7.8169259 , -18.74466988,  83.75188118,\n",
       "         27.15080382,   7.17648789, -34.49623016, -47.27851288,\n",
       "         16.27205054, -10.73538715,  21.48785826,  -1.64313744],\n",
       "       [ -2.57187561,  -1.0869639 , -16.65775473,  -0.72093792,\n",
       "        -19.86850366,  50.04533256,  32.48188709,  66.55750948,\n",
       "          2.73484302,  26.98704302, -23.72523798,  -4.82430777],\n",
       "       [ -5.03603548,  -2.17041306, -10.86526421,  -8.55047381,\n",
       "         40.24751013,   0.24885107, -21.74131417, -71.39279091,\n",
       "          2.14768248, -35.31722401,  -4.48301633,  -5.45873129],\n",
       "       [ -4.93364889,  -1.16045354,  22.02821313, -26.36630398,\n",
       "         71.32567562, -37.323074  ,   5.91342988, -51.95503294,\n",
       "         33.3575045 ,   6.71748896,  -4.28726433,   1.86308121],\n",
       "       [  5.89813927,   8.89580687, -16.84429935,  -3.66284426,\n",
       "         -8.64130507,  20.48066633, -17.10293684,  13.28808219,\n",
       "         18.13342343,  44.07809791,  21.21582072,  -1.24346472],\n",
       "       [  3.70004447,  -5.72258397,   7.51018726,  48.40629089,\n",
       "        -40.30206588, -47.02213762, -48.061555  ,  52.55423689,\n",
       "         21.03108004, -35.04547899,  -9.78589982,  31.99353765],\n",
       "       [ -4.78018118,   4.02554522,  29.33150531,  -4.93160034,\n",
       "        -11.02624665, -25.74132846, -29.65650663, 105.03688978,\n",
       "         58.56263958,  44.15497629, -25.82334757,   9.96836307],\n",
       "       [ -4.73838756,  -4.1887247 ,  -9.49130864,  15.7377958 ,\n",
       "        -16.93683128,  22.86663064,  -5.92553757, -16.22418038,\n",
       "          5.50780647,  27.35955498,   3.06823653,  -3.84762775],\n",
       "       [ -4.24659199,   5.78859661,  19.50729816, -22.90303491,\n",
       "        -56.98953793, -19.55360611, -24.75196432,  24.95728314,\n",
       "        -24.66762823,  89.44698624, -10.99931024,  10.07441769],\n",
       "       [  8.9217537 ,  -6.31158079,  -1.38279544, -25.77680717,\n",
       "        -25.32039261,  61.09322424,  40.61937247,  26.102415  ,\n",
       "        100.50710791, -17.81422476,   0.90601152,   6.45457213],\n",
       "       [ -4.08897118,   2.8521789 ,   3.22076795,   6.27840916,\n",
       "         20.03006949,  19.34901494,   9.37542766, -18.03066259,\n",
       "        -61.10388004, -35.07509207, -15.56807004,   1.49048361],\n",
       "       [ -2.92183489,  18.09093633,  -3.05612847,  25.33797257,\n",
       "         78.59657378, -38.2634747 , -14.88839552, -67.54099971,\n",
       "        -71.92893029, -21.03740467,  -6.64227247,   5.62395147],\n",
       "       [  6.55188654,   3.37370344,   0.42841733,  26.80035357,\n",
       "         52.40257795,  16.6967937 , -34.14358797, -57.01923473,\n",
       "        -33.30615192,  42.56343236, -13.97901758,   0.59012477],\n",
       "       [ -1.2889604 ,  -1.52785346, -19.2804976 ,  49.47347395,\n",
       "         16.78542428,  73.04633583, -31.0586801 ,  36.20591006,\n",
       "        -21.40328171, -19.44315741, -14.89496731,   6.6998964 ]])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "precip_month_anoms = pd.read_table(\n",
    "        'Crop_data_files/maize_met_anoms/Maize_Spring_USA_{}_precip_anom_real.csv'.format(state))\n",
    "np.array(precip_month_anoms[all_months])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Rerun the full models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read in climate temperatures\n",
    "clim_temp_maize=pd.read_table('./Crop_data_files/clim_file/temp_climatology_Maize.csv')\n",
    "clim_temp_maize.rename(columns = {'Unnamed: 0':'Crop_season_location'}, inplace = True)\n",
    "# Read in climate precipitation\n",
    "clim_precip_maize=pd.read_table('./Crop_data_files/clim_file/precip_climatology_Maize.csv')\n",
    "clim_precip_maize.rename(columns = {'Unnamed: 0':'Crop_season_location'}, inplace = True)\n",
    "# Read in Yields\n",
    "yields=pd.read_table('./Crop_data_files/Maize_median_yield_anoms.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "states=['Indiana','Illinois', 'Ohio','Nebraska', 'Iowa','Minnesota']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Read in and add back mean temperature to get real temperature values\n",
    "temp_states=[]\n",
    "for i,s in enumerate(states):\n",
    "    maize_temp=pd.read_table('./Crop_data_files/maize_met_anoms/Maize_Spring_USA_'\n",
    "                             +s+'_temp_anom_real.csv')\n",
    "    maize_temp.rename(columns = {'Unnamed: 0':'Year'}, inplace = True)\n",
    "    tmp=maize_temp.iloc[:,1:].add(clim_temp_maize[\n",
    "        clim_temp_maize['Crop_season_location']== 'Maize_Spring_USA_'+states[0]\n",
    "                                                 ].iloc[0,1:,])\n",
    "    temp_states.append(tmp)\n",
    "temp_states=pd.concat(temp_states,keys=states)\n",
    "\n",
    "#Read in and add back mean precipitation to get real precipitation values\n",
    "precip_states=[]\n",
    "for i,s in enumerate(states):\n",
    "    maize_precip=pd.read_table('./Crop_data_files/maize_met_anoms/Maize_Spring_USA_'\n",
    "                               +s+'_precip_anom_real.csv')\n",
    "    maize_precip.rename(columns = {'Unnamed: 0':'Year'}, inplace = True)\n",
    "    tmp=maize_precip.iloc[:,1:].add(clim_precip_maize[\n",
    "        clim_precip_maize['Crop_season_location']== 'Maize_Spring_USA_'+states[0]\n",
    "                                                     ].iloc[0,1:,])\n",
    "    precip_states.append(tmp)\n",
    "precip_states=pd.concat(precip_states,keys=states)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_years=np.array(yields[yields['Region']=='Maize_Spring_USA_Indiana'].iloc[0,22:]).size\n",
    "data2={\n",
    "    'n_regions':len(states),\n",
    "    'n_years':n_years,\n",
    "    'd_temp':np.array(temp_states.iloc[:,3:9]).reshape(\n",
    "                     len(states),\n",
    "                     np.int(np.array(temp_states.iloc[:,3:9]).shape[0]/len(states)),6\n",
    "                                                      ).astype(float),\n",
    "    'd_precip':np.array(precip_states.iloc[:,3:9]).reshape(\n",
    "                     len(states),\n",
    "                     np.int(np.array(precip_states.iloc[:,3:9]).shape[0]/len(states)),6\n",
    "                                                      ).astype(float),\n",
    "    'd_yields':np.array(yields[yields[\"Region\"].isin(\n",
    "                     ['Maize_Spring_USA_'+s for s in states]\n",
    "                                                    )].iloc[:,22:]).astype(float)+9.75,\n",
    "    'n_gf':40,\n",
    "    'temp':np.arange(0,40,1),\n",
    "    'precip':np.arange(0,200,5)\n",
    "\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:pystan:COMPILING THE C++ CODE FOR MODEL anon_model_889b1084947e1d379c9ef129719d74d2 NOW.\n"
     ]
    }
   ],
   "source": [
    "\n",
    "gm2 = pystan.StanModel(file='./stan/2d-gaussian.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fit=gm2.sampling(data=data2,chains=4,iter=1000,verbose=True,seed=1308)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Inference for Stan model: anon_model_889b1084947e1d379c9ef129719d74d2.\n",
       "4 chains, each with iter=1000; warmup=500; thin=1; \n",
       "post-warmup draws per chain=500, total post-warmup draws=2000.\n",
       "\n",
       "                    mean se_mean     sd   2.5%    25%    50%    75%  97.5%  n_eff   Rhat\n",
       "mu_t               18.66    0.03   0.78  17.04  18.15  18.68  19.22  20.12    891    1.0\n",
       "sigma_t             7.16    0.02   0.58   6.06   6.74   7.13   7.55   8.32    997    1.0\n",
       "mu_p              114.47    0.09   3.38 108.07 112.23  114.4 116.66 121.26   1413    1.0\n",
       "sigma_p            75.01    0.07   2.62  70.18  73.14  74.85  76.87  80.21   1505    1.0\n",
       "norm                 2.2  1.8e-3   0.05    2.1   2.16    2.2   2.24   2.31    974    1.0\n",
       "fdy1[0]             0.08  1.8e-3   0.06   0.01   0.04   0.07   0.11   0.23    999    1.0\n",
       "fdy1[1]             0.12  2.3e-3   0.07   0.02   0.06    0.1   0.15    0.3    993    1.0\n",
       "fdy1[2]             0.16  2.9e-3   0.09   0.03   0.09   0.14   0.21   0.38    986    1.0\n",
       "fdy1[3]             0.21  3.5e-3   0.11   0.05   0.13   0.19   0.27   0.47    980    1.0\n",
       "fdy1[4]             0.28  4.1e-3   0.13   0.08   0.19   0.26   0.36   0.58    975    1.0\n",
       "fdy1[5]             0.36  4.8e-3   0.15   0.12   0.25   0.34   0.45    0.7    969    1.0\n",
       "fdy1[6]             0.46  5.5e-3   0.17   0.17   0.34   0.44   0.57   0.84    965    1.0\n",
       "fdy1[7]             0.58  6.1e-3   0.19   0.24   0.45   0.56    0.7   0.99    960    1.0\n",
       "fdy1[8]             0.71  6.6e-3    0.2   0.34   0.57    0.7   0.85   1.15    954    1.0\n",
       "fdy1[9]             0.86  7.4e-3   0.21   0.46   0.72   0.85   1.01   1.31    833    1.0\n",
       "fdy1[10]            1.03  7.6e-3   0.22    0.6   0.89   1.02   1.17   1.48    823    1.0\n",
       "fdy1[11]             1.2  7.6e-3   0.22   0.76   1.06    1.2   1.35   1.63    816    1.0\n",
       "fdy1[12]            1.38  7.3e-3   0.21   0.95   1.25   1.38   1.52   1.78    812    1.0\n",
       "fdy1[13]            1.56  6.6e-3   0.19   1.16   1.43   1.56   1.69   1.91    813    1.0\n",
       "fdy1[14]            1.72  5.8e-3   0.17   1.37   1.61   1.73   1.84   2.03    823    1.0\n",
       "fdy1[15]            1.87  4.8e-3   0.14   1.58   1.78   1.88   1.97   2.13    845    1.0\n",
       "fdy1[16]            1.99  3.7e-3   0.11   1.77   1.92    2.0   2.07    2.2    892    1.0\n",
       "fdy1[17]            2.08  2.6e-3   0.08   1.92   2.03   2.08   2.14   2.24    999    1.0\n",
       "fdy1[18]            2.13  1.8e-3   0.06   2.02   2.09   2.13   2.17   2.25   1170    1.0\n",
       "fdy1[19]            2.14  1.6e-3   0.05   2.05    2.1   2.14   2.18   2.25   1093    1.0\n",
       "fdy1[20]            2.11  1.9e-3   0.06    2.0   2.07   2.11   2.15   2.24    980    1.0\n",
       "fdy1[21]            2.04  2.2e-3   0.07   1.91   1.99   2.04   2.08   2.18    979    1.0\n",
       "fdy1[22]            1.93  2.6e-3   0.08   1.78   1.88   1.92   1.98   2.09    894    1.0\n",
       "fdy1[23]            1.79  2.8e-3   0.09   1.62   1.73   1.78   1.85   1.97    943    1.0\n",
       "fdy1[24]            1.63  2.9e-3   0.09   1.45   1.56   1.63   1.69   1.81   1027    1.0\n",
       "fdy1[25]            1.45  2.9e-3    0.1   1.26   1.38   1.45   1.52   1.64   1124    1.0\n",
       "fdy1[26]            1.27  2.8e-3    0.1   1.07    1.2   1.27   1.34   1.47   1301    1.0\n",
       "fdy1[27]            1.09  2.9e-3    0.1   0.89   1.02   1.09   1.16   1.29   1306    1.0\n",
       "fdy1[28]            0.92  2.9e-3   0.11   0.72   0.84   0.91   0.99   1.13   1286    1.0\n",
       "fdy1[29]            0.76  2.9e-3    0.1   0.57   0.68   0.75   0.82   0.97   1254    1.0\n",
       "fdy1[30]            0.61  2.8e-3    0.1   0.43   0.54   0.61   0.68   0.82   1218    1.0\n",
       "fdy1[31]            0.49  2.7e-3   0.09   0.32   0.42   0.48   0.55   0.68   1183    1.0\n",
       "fdy1[32]            0.38  2.5e-3   0.08   0.23   0.32   0.37   0.43   0.56   1153    1.0\n",
       "fdy1[33]            0.29  2.2e-3   0.08   0.16   0.24   0.29   0.34   0.45   1128    1.0\n",
       "fdy1[34]            0.22  2.0e-3   0.06   0.11   0.17   0.21   0.26   0.36   1106    1.0\n",
       "fdy1[35]            0.16  1.7e-3   0.05   0.08   0.12   0.16   0.19   0.29   1089    1.0\n",
       "fdy1[36]            0.12  1.4e-3   0.05   0.05   0.09   0.11   0.14   0.22   1074    1.0\n",
       "fdy1[37]            0.08  1.1e-3   0.04   0.03   0.06   0.08    0.1   0.17   1062    1.0\n",
       "fdy1[38]            0.06  8.9e-4   0.03   0.02   0.04   0.05   0.07   0.13   1052    1.0\n",
       "fdy1[39]            0.04  6.9e-4   0.02   0.01   0.03   0.04   0.05   0.09   1044    1.0\n",
       "fdy2[0]             0.67  1.8e-3   0.07   0.55   0.62   0.67   0.72   0.82   1452    1.0\n",
       "fdy2[1]             0.74  1.8e-3   0.07   0.61   0.69   0.74   0.79   0.89   1444    1.0\n",
       "fdy2[2]             0.82  1.9e-3   0.07   0.68   0.76   0.81   0.86   0.96   1435    1.0\n",
       "fdy2[3]             0.89  1.9e-3   0.07   0.76   0.84   0.89   0.94   1.04   1424    1.0\n",
       "fdy2[4]             0.97  2.0e-3   0.07   0.84   0.92   0.97   1.02   1.12   1412    1.0\n",
       "fdy2[5]             1.05  2.0e-3   0.07   0.92    1.0   1.05    1.1   1.21   1397    1.0\n",
       "fdy2[6]             1.14  2.0e-3   0.08    1.0   1.09   1.14   1.19   1.29   1381    1.0\n",
       "fdy2[7]             1.23  2.0e-3   0.07   1.09   1.17   1.22   1.27   1.38   1361    1.0\n",
       "fdy2[8]             1.31  2.0e-3   0.07   1.18   1.26   1.31   1.36   1.46   1333    1.0\n",
       "fdy2[9]              1.4  2.0e-3   0.07   1.26   1.35    1.4   1.45   1.55   1302    1.0\n",
       "fdy2[10]            1.48  2.0e-3   0.07   1.35   1.43   1.48   1.53   1.63   1270    1.0\n",
       "fdy2[11]            1.57  2.0e-3   0.07   1.44   1.52   1.57   1.62   1.71   1235    1.0\n",
       "fdy2[12]            1.65  2.0e-3   0.07   1.52    1.6   1.65    1.7   1.79   1200    1.0\n",
       "fdy2[13]            1.73  2.0e-3   0.07    1.6   1.68   1.73   1.77   1.87   1164    1.0\n",
       "fdy2[14]             1.8  2.0e-3   0.07   1.68   1.76    1.8   1.85   1.94   1128    1.0\n",
       "fdy2[15]            1.87  1.9e-3   0.06   1.75   1.83   1.87   1.91    2.0   1094    1.0\n",
       "fdy2[16]            1.93  1.9e-3   0.06   1.82   1.89   1.93   1.97   2.06   1063    1.0\n",
       "fdy2[17]            1.99  1.9e-3   0.06   1.88   1.95   1.99   2.03   2.12   1035    1.0\n",
       "fdy2[18]            2.04  1.9e-3   0.06   1.93   1.99   2.03   2.08   2.17   1011    1.0\n",
       "fdy2[19]            2.08  1.9e-3   0.06   1.97   2.04   2.07   2.12    2.2    993    1.0\n",
       "fdy2[20]            2.11  1.9e-3   0.06    2.0   2.07   2.11   2.15   2.24    980    1.0\n",
       "fdy2[21]            2.13  1.9e-3   0.06   2.02   2.09   2.13   2.17   2.26    973    1.0\n",
       "fdy2[22]            2.15  1.9e-3   0.06   2.04   2.11   2.14   2.18   2.27    968    1.0\n",
       "fdy2[23]            2.15  1.9e-3   0.06   2.04   2.11   2.15   2.19   2.28    968    1.0\n",
       "fdy2[24]            2.14  1.9e-3   0.06   2.03    2.1   2.14   2.18   2.27    975    1.0\n",
       "fdy2[25]            2.13  1.9e-3   0.06   2.01   2.09   2.13   2.17   2.26    983    1.0\n",
       "fdy2[26]             2.1  2.0e-3   0.06   1.98   2.06    2.1   2.14   2.23    996    1.0\n",
       "fdy2[27]            2.07  2.0e-3   0.06   1.95   2.03   2.07   2.11    2.2   1014    1.0\n",
       "fdy2[28]            2.03  2.0e-3   0.07    1.9   1.99   2.03   2.07   2.17   1034    1.0\n",
       "fdy2[29]            1.98  2.1e-3   0.07   1.85   1.93   1.98   2.02   2.12   1057    1.0\n",
       "fdy2[30]            1.92  2.1e-3   0.07   1.79   1.88   1.92   1.97   2.06   1080    1.0\n",
       "fdy2[31]            1.86  2.2e-3   0.07   1.72   1.81   1.85    1.9    2.0   1104    1.0\n",
       "fdy2[32]            1.79  2.2e-3   0.07   1.64   1.74   1.78   1.84   1.93   1127    1.0\n",
       "fdy2[33]            1.71  2.3e-3   0.08   1.56   1.66   1.71   1.76   1.86   1149    1.0\n",
       "fdy2[34]            1.63  2.3e-3   0.08   1.48   1.58   1.63   1.68   1.79   1170    1.0\n",
       "fdy2[35]            1.55  2.3e-3   0.08    1.4    1.5   1.55    1.6   1.71   1188    1.0\n",
       "fdy2[36]            1.47  2.4e-3   0.08   1.31   1.41   1.46   1.52   1.63   1205    1.0\n",
       "fdy2[37]            1.38  2.4e-3   0.08   1.22   1.32   1.38   1.44   1.55   1221    1.0\n",
       "fdy2[38]            1.29  2.4e-3   0.08   1.13   1.23   1.29   1.35   1.47   1234    1.0\n",
       "fdy2[39]            1.21  2.4e-3   0.08   1.05   1.15    1.2   1.26   1.38   1246    1.0\n",
       "pred_yields[0,0]    9.62    0.02   1.01    7.7   8.95   9.61  10.29  11.61   2000    1.0\n",
       "pred_yields[1,0]     9.5    0.02   1.03   7.46   8.82   9.51  10.19  11.58   2000    1.0\n",
       "pred_yields[2,0]    9.56    0.02   1.05   7.47   8.85   9.53  10.25  11.68   2000    1.0\n",
       "pred_yields[3,0]    8.14    0.02   1.01   6.13   7.47   8.15   8.83  10.08   1941    1.0\n",
       "pred_yields[4,0]    8.77    0.02   1.01   6.75   8.09   8.79   9.46  10.74   2000    1.0\n",
       "pred_yields[5,0]    9.77    0.02    1.0    7.8   9.09   9.79  10.45  11.72   2000    1.0\n",
       "pred_yields[0,1]   10.44    0.02   1.05   8.42   9.74  10.44  11.14   12.5   2000    1.0\n",
       "pred_yields[1,1]   10.27    0.02   1.05   8.18   9.55  10.26  10.97  12.26   2000    1.0\n",
       "pred_yields[2,1]   10.53    0.02   1.02   8.52   9.87  10.54  11.22  12.53   2000    1.0\n",
       "pred_yields[3,1]   10.32    0.02   0.99   8.33   9.64  10.34  10.99  12.16   1990    1.0\n",
       "pred_yields[4,1]   10.77    0.02   0.98   8.87  10.11  10.75  11.42  12.72   1826    1.0\n",
       "pred_yields[5,1]   10.42    0.02   0.99   8.51   9.77  10.43  11.08  12.39   2000    1.0\n",
       "pred_yields[0,2]   10.02    0.02   1.02   8.05   9.32  10.01  10.69  12.05   1917    1.0\n",
       "pred_yields[1,2]    9.99    0.02   1.03   8.01   9.28   10.0  10.68  11.97   2000    1.0\n",
       "pred_yields[2,2]    9.77    0.02   0.99   7.82   9.09   9.76  10.43   11.7   1887    1.0\n",
       "pred_yields[3,2]    9.03    0.02   1.01   7.03   8.34   9.04   9.72  11.05   2000    1.0\n",
       "pred_yields[4,2]    9.09    0.02   1.01   7.04   8.39   9.13   9.76   11.0   1962    1.0\n",
       "pred_yields[5,2]    9.52    0.02   1.02   7.53   8.82    9.5  10.22  11.46   2000    1.0\n",
       "pred_yields[0,3]     7.8    0.03   1.03   5.82   7.08   7.79   8.53   9.79   1694    1.0\n",
       "pred_yields[1,3]    8.08    0.02   1.02   6.12   7.37   8.08   8.75  10.18   2000    1.0\n",
       "pred_yields[2,3]    8.46    0.02   0.99   6.51   7.79   8.45   9.14   10.4   1877    1.0\n",
       "pred_yields[3,3]    7.55    0.02   0.98   5.64    6.9   7.54   8.22   9.45   1886    1.0\n",
       "pred_yields[4,3]    8.31    0.02   1.02   6.28   7.62    8.3   8.98  10.36   1985    1.0\n",
       "pred_yields[5,3]    9.09    0.03   1.05   7.01   8.38   9.09    9.8  11.15   1759    1.0\n",
       "pred_yields[0,4]    9.39    0.02   1.02   7.42   8.73   9.39  10.07  11.41   2000    1.0\n",
       "pred_yields[1,4]    9.38    0.02   0.99    7.4   8.74   9.36  10.05  11.25   1840    1.0\n",
       "pred_yields[2,4]    9.63    0.02    1.0   7.69   8.96   9.63  10.29  11.66   2000    1.0\n",
       "pred_yields[3,4]     7.6    0.02   1.03   5.63   6.91    7.6   8.28    9.6   1723    1.0\n",
       "pred_yields[4,4]    8.17    0.02    1.0   6.28   7.49   8.17   8.86  10.06   2000    1.0\n",
       "pred_yields[5,4]    9.04    0.02   1.01   6.99   8.36   9.03   9.72  11.02   2000    1.0\n",
       "pred_yields[0,5]   10.52    0.02   0.99   8.61   9.83  10.51  11.21  12.42   1941    1.0\n",
       "pred_yields[1,5]   10.79    0.02   1.02    8.9  10.09  10.77  11.47  12.84   1713    1.0\n",
       "pred_yields[2,5]   10.51    0.02   1.01   8.55   9.83  10.51   11.2  12.46   1932    1.0\n",
       "pred_yields[3,5]   11.66    0.02   1.01   9.69  10.96  11.67  12.36  13.58   2000    1.0\n",
       "pred_yields[4,5]   11.05    0.03   1.02   9.12  10.36  11.02  11.75  13.07   1658    1.0\n",
       "pred_yields[5,5]   11.36    0.02   1.03   9.33  10.67  11.34  12.08  13.34   1866    1.0\n",
       "pred_yields[0,6]   10.41    0.02    1.0   8.48   9.73  10.41   11.1  12.36   1960    1.0\n",
       "pred_yields[1,6]     9.7    0.02   0.98   7.79   9.04    9.7  10.37  11.52   2000    1.0\n",
       "pred_yields[2,6]   10.72    0.02   1.02   8.77  10.02  10.69   11.4  12.71   1822    1.0\n",
       "pred_yields[3,6]   10.36    0.02   1.03   8.31   9.67  10.39  11.04  12.37   1789    1.0\n",
       "pred_yields[4,6]    10.1    0.02   1.01   8.12   9.45   10.1  10.77  12.15   2000    1.0\n",
       "pred_yields[5,6]   10.07    0.02   1.01   8.15   9.38  10.07  10.71  12.08   2000    1.0\n",
       "pred_yields[0,7]    9.55    0.02   1.01   7.65   8.86   9.59  10.25   11.5   2000    1.0\n",
       "pred_yields[1,7]    9.54    0.02   1.02   7.49   8.87   9.58  10.24  11.51   2000    1.0\n",
       "pred_yields[2,7]    9.89    0.02   1.02   7.81   9.21   9.91  10.59  11.82   2000    1.0\n",
       "pred_yields[3,7]   10.14    0.02    1.0    8.2   9.45  10.16   10.8  12.13   1840    1.0\n",
       "pred_yields[4,7]    9.53    0.02    1.0   7.55   8.88   9.52  10.21   11.5   1929    1.0\n",
       "pred_yields[5,7]    9.46    0.02   1.04   7.46   8.74   9.45  10.15  11.59   1763    1.0\n",
       "pred_yields[0,8]    7.98    0.02   1.02   5.96   7.32   7.98   8.66   10.0   2000    1.0\n",
       "pred_yields[1,8]    7.37    0.02   1.01   5.41   6.71   7.39   8.06   9.32   1912    1.0\n",
       "pred_yields[2,8]    8.36    0.02   1.02   6.37   7.69   8.38   9.01  10.34   2000    1.0\n",
       "pred_yields[3,8]    9.38    0.02   1.02   7.34   8.66   9.35  10.05  11.45   1816    1.0\n",
       "pred_yields[4,8]    7.88    0.02   1.01   5.88   7.23   7.89   8.57   9.79   2000    1.0\n",
       "pred_yields[5,8]    8.21    0.02    1.0   6.25   7.55   8.24   8.86  10.19   1848    1.0\n",
       "pred_yields[0,9]   10.58    0.03   1.01   8.66    9.9  10.56  11.26  12.67   1588    1.0\n",
       "pred_yields[1,9]   10.49    0.02   1.02   8.49   9.81  10.48  11.15  12.53   1856    1.0\n",
       "pred_yields[2,9]   10.15    0.02   1.01   8.18   9.46  10.16  10.85  12.08   2000    1.0\n",
       "pred_yields[3,9]   10.48    0.02    1.0   8.52   9.81  10.48  11.16  12.47   1785    1.0\n",
       "pred_yields[4,9]   10.18    0.02   1.01   8.12   9.52  10.18  10.85   12.2   2000    1.0\n",
       "pred_yields[5,9]    9.67    0.02   1.03   7.66   8.98   9.66   10.4   11.7   1804    1.0\n",
       "pred_yields[0,10]   9.64    0.02   0.99   7.72   8.98   9.64  10.32  11.63   1934    1.0\n",
       "pred_yields[1,10]   8.48    0.02    1.0   6.48   7.82    8.5   9.13  10.46   1709    1.0\n",
       "pred_yields[2,10]  10.16    0.02    1.0   8.12   9.55  10.16  10.83   12.2   1935    1.0\n",
       "pred_yields[3,10]   8.98    0.02   1.02   6.99   8.27   8.96   9.67  10.99   1782    1.0\n",
       "pred_yields[4,10]   8.67    0.02   0.97   6.81   7.99   8.66   9.35  10.58   2000    1.0\n",
       "pred_yields[5,10]    9.6    0.02   1.02   7.66   8.91   9.59  10.28  11.62   2000    1.0\n",
       "pred_yields[0,11]   9.49    0.02   1.02   7.49   8.82   9.52  10.16  11.47   1953    1.0\n",
       "pred_yields[1,11]   9.47    0.02   1.01   7.55   8.82   9.47  10.14  11.53   1813    1.0\n",
       "pred_yields[2,11]   9.65    0.02    1.0   7.69   8.97   9.65  10.34  11.58   2000    1.0\n",
       "pred_yields[3,11]   9.52    0.02   1.02   7.52   8.86   9.55   10.2  11.53   1818    1.0\n",
       "pred_yields[4,11]   9.67    0.02    1.0    7.7   9.02   9.68  10.32  11.59   1911    1.0\n",
       "pred_yields[5,11]  10.06    0.02   0.99    8.1   9.38  10.08  10.76  11.97   1911    1.0\n",
       "pred_yields[0,12]   9.82    0.02   1.02   7.84   9.11   9.82  10.51  11.79   2000    1.0\n",
       "pred_yields[1,12]   8.26    0.02   1.04   6.23   7.56   8.22   8.95  10.37   1745    1.0\n",
       "pred_yields[2,12]  10.23    0.02    1.0    8.3   9.55  10.19  10.92  12.18   1677    1.0\n",
       "pred_yields[3,12]   9.91    0.02   1.02    7.9   9.22   9.92  10.61  11.85   2000    1.0\n",
       "pred_yields[4,12]   8.51    0.02   1.03   6.51    7.8   8.51   9.22  10.54   2000    1.0\n",
       "pred_yields[5,12]  10.73    0.02    1.0    8.7  10.08  10.73  11.42  12.65   1612    1.0\n",
       "pred_yields[0,13]   9.92    0.02   1.01   7.94   9.23   9.92  10.62  11.93   2000    1.0\n",
       "pred_yields[1,13]   8.42    0.02   1.03   6.39   7.69   8.43   9.14  10.39   1857    1.0\n",
       "pred_yields[2,13]   9.81    0.02   1.01   7.85   9.12   9.82  10.46  11.84   1865    1.0\n",
       "pred_yields[3,13]   8.21    0.03   1.05   6.22   7.51   8.22   8.91  10.34   1670    1.0\n",
       "pred_yields[4,13]   7.44    0.02   1.01   5.43   6.79   7.45   8.17   9.39   1730    1.0\n",
       "pred_yields[5,13]   8.35    0.02   1.05   6.37   7.62   8.36   9.09  10.36   1907    1.0\n",
       "pred_yields[0,14]   9.78    0.02   1.01   7.87    9.1   9.75  10.46   11.8   1847    1.0\n",
       "pred_yields[1,14]  10.01    0.02    1.0   8.08   9.33   9.99  10.69  12.01   2000    1.0\n",
       "pred_yields[2,14]   9.98    0.02   0.99   8.02   9.34   9.98  10.64  11.96   2000    1.0\n",
       "pred_yields[3,14]  10.28    0.02   0.98   8.39   9.61  10.29  10.96  12.17   2000    1.0\n",
       "pred_yields[4,14]  10.45    0.02   1.01   8.54   9.79  10.43  11.08  12.57   2000    1.0\n",
       "pred_yields[5,14]   11.0    0.02   1.01   9.08  10.32  10.98  11.72  12.98   1935    1.0\n",
       "pred_yields[0,15]   8.11    0.02    1.0   6.08   7.45   8.14   8.81  10.01   1931    1.0\n",
       "pred_yields[1,15]    7.0    0.02   1.01   5.08   6.31   6.98   7.65    9.1   2000    1.0\n",
       "pred_yields[2,15]   8.93    0.02   1.02   6.96   8.22   8.94   9.63  10.88   1770    1.0\n",
       "pred_yields[3,15]   7.71    0.02   1.01   5.79   7.03   7.71    8.4    9.7   1851    1.0\n",
       "pred_yields[4,15]   7.87    0.02   1.02   5.92   7.18   7.84   8.57   9.81   1970    1.0\n",
       "pred_yields[5,15]   9.09    0.02   0.99   7.25   8.41   9.08   9.78  11.06   2000    1.0\n",
       "pred_yields[0,16]   8.68    0.02    1.0   6.76   8.01   8.69   9.35  10.58   2000    1.0\n",
       "pred_yields[1,16]   9.13    0.02    1.0   7.17   8.44   9.14   9.79   11.1   2000    1.0\n",
       "pred_yields[2,16]   9.15    0.03   1.01   7.07   8.48   9.18   9.79  11.14   1609    1.0\n",
       "pred_yields[3,16]   9.61    0.02   1.01   7.66   8.96   9.58  10.27  11.65   1878    1.0\n",
       "pred_yields[4,16]   9.05    0.02    1.0   7.06   8.38   9.07   9.72  10.99   1668    1.0\n",
       "pred_yields[5,16]   9.36    0.02   1.03   7.27   8.69   9.39  10.06  11.36   1860    1.0\n",
       "pred_yields[0,17]    9.4    0.03   1.03   7.39   8.68   9.36  10.11  11.52   1525    1.0\n",
       "pred_yields[1,17]   9.28    0.02   1.03   7.25   8.58   9.29   9.97  11.37   2000    1.0\n",
       "pred_yields[2,17]   9.45    0.02   1.02   7.46   8.76   9.46  10.13  11.52   2000    1.0\n",
       "pred_yields[3,17]   9.58    0.02    1.0   7.58   8.92   9.61  10.26   11.5   2000    1.0\n",
       "pred_yields[4,17]   9.61    0.02   1.01   7.67   8.92    9.6  10.29  11.67   2000    1.0\n",
       "pred_yields[5,17]   8.82    0.02   1.01   6.77   8.17   8.85   9.51  10.72   2000    1.0\n",
       "pred_yields[0,18]   9.54    0.02    1.0   7.65   8.82   9.56  10.24  11.51   2000    1.0\n",
       "pred_yields[1,18]   9.76    0.02   1.03   7.79   9.07   9.76  10.45  11.75   2000    1.0\n",
       "pred_yields[2,18]   9.31    0.02   1.01   7.32    8.6   9.32   9.98  11.34   1800    1.0\n",
       "pred_yields[3,18]   9.74    0.02    1.0   7.77   9.09   9.74   10.4  11.63   1993    1.0\n",
       "pred_yields[4,18]   9.63    0.02   1.01    7.7   8.94    9.6  10.32  11.62   1620    1.0\n",
       "pred_yields[5,18]  10.09    0.02    1.0   8.11   9.45  10.08  10.77  12.11   2000    1.0\n",
       "pred_yields[0,19]   9.08    0.02   1.02   7.14   8.39   9.07   9.78  11.07   1949    1.0\n",
       "pred_yields[1,19]   9.83    0.02   1.01   7.75   9.15   9.84  10.53  11.78   2000    1.0\n",
       "pred_yields[2,19]   9.19    0.02    1.0   7.24   8.51   9.21   9.86  11.11   2000    1.0\n",
       "pred_yields[3,19]   9.36    0.02   1.03   7.34   8.65   9.37  10.07  11.33   2000    1.0\n",
       "pred_yields[4,19]   9.73    0.02   1.02   7.69   9.06   9.71   10.4  11.75   2000    1.0\n",
       "pred_yields[5,19]  10.16    0.02    1.0   8.21   9.49  10.15  10.82  12.11   2000    1.0\n",
       "pred_yields[0,20]  10.72    0.02    1.0   8.71  10.05   10.7  11.44  12.63   1747    1.0\n",
       "pred_yields[1,20]   10.3    0.02   1.03   8.33    9.6  10.32  11.01  12.26   2000    1.0\n",
       "pred_yields[2,20]  11.01    0.02   0.98   9.08  10.37  10.99  11.68  12.92   2000    1.0\n",
       "pred_yields[3,20]   9.39    0.02   1.01   7.38   8.71   9.39  10.07   11.3   2000    1.0\n",
       "pred_yields[4,20]   9.79    0.02   0.99   7.86   9.12    9.8  10.46  11.71   2000    1.0\n",
       "pred_yields[5,20]    9.8    0.02   1.01   7.87    9.1   9.78  10.48  11.75   1841    1.0\n",
       "pred_yields[0,21]  10.93    0.02   1.01   8.87  10.26  10.93  11.61  12.95   2000    1.0\n",
       "pred_yields[1,21]  11.36    0.02   1.02   9.29  10.68  11.36  12.01  13.39   2000    1.0\n",
       "pred_yields[2,21]  10.75    0.02   0.99    8.8   10.1  10.72  11.39  12.72   2000    1.0\n",
       "pred_yields[3,21]  10.11    0.02   1.01   8.11   9.47  10.09  10.82  12.12   1920    1.0\n",
       "pred_yields[4,21]   9.75    0.02   0.98   7.82   9.11   9.74  10.42  11.66   2000    1.0\n",
       "pred_yields[5,21]   9.22    0.02   0.99   7.29   8.55   9.22   9.89  11.11   2000    1.0\n",
       "pred_yields[0,22]   9.04    0.02   1.03    7.1   8.34   9.03   9.72  11.02   2000    1.0\n",
       "pred_yields[1,22]   9.13    0.02   1.02   7.15   8.42   9.13   9.85  11.13   2000    1.0\n",
       "pred_yields[2,22]   9.09    0.02   1.02   7.09    8.4    9.1   9.77  11.09   2000    1.0\n",
       "pred_yields[3,22]   9.45    0.02    1.0   7.45   8.79   9.46   10.1  11.46   2000    1.0\n",
       "pred_yields[4,22]   9.88    0.02    1.0   7.97   9.19   9.91  10.57  11.83   1839    1.0\n",
       "pred_yields[5,22]   9.38    0.02   1.01   7.44   8.69   9.34  10.05  11.35   2000    1.0\n",
       "pred_yields[0,23]   9.68    0.02   1.02   7.69   9.02   9.69  10.33  11.69   1999    1.0\n",
       "pred_yields[1,23]  10.65    0.02   1.02   8.58   9.96  10.69  11.34  12.57   1925    1.0\n",
       "pred_yields[2,23]  10.12    0.02   0.98   8.18   9.44  10.12  10.77  12.03   2000    1.0\n",
       "pred_yields[3,23]  10.31    0.02   1.01   8.34   9.64  10.31  11.05  12.28   1955    1.0\n",
       "pred_yields[4,23]   10.0    0.02   1.02   8.02   9.29   10.0  10.68  12.04   2000    1.0\n",
       "pred_yields[5,23]   9.38    0.02   1.02    7.4    8.7    9.4  10.06  11.34   1934    1.0\n",
       "pred_yields[0,24]   8.68    0.02   1.03   6.63    8.0   8.69   9.37  10.72   1756    1.0\n",
       "pred_yields[1,24]   8.85    0.02   1.01   6.79   8.17   8.85   9.51  10.88   2000    1.0\n",
       "pred_yields[2,24]   9.78    0.02   0.99   7.75   9.14   9.78  10.45  11.62   2000    1.0\n",
       "pred_yields[3,24]   11.1    0.02   1.01   9.12  10.39  11.11  11.81  12.98   1929    1.0\n",
       "pred_yields[4,24]  10.34    0.02   1.01   8.26   9.68  10.37  11.01  12.32   2000    1.0\n",
       "pred_yields[5,24]    9.6    0.02    1.0   7.55   8.96   9.61  10.27  11.56   2000    1.0\n",
       "pred_yields[0,25]   9.56    0.02   1.03   7.52   8.86   9.55  10.27  11.52   2000    1.0\n",
       "pred_yields[1,25]   9.23    0.02   0.98   7.32   8.57   9.23   9.87  11.21   1965    1.0\n",
       "pred_yields[2,25]   9.58    0.02    1.0   7.61    8.9    9.6  10.27  11.49   1800    1.0\n",
       "pred_yields[3,25]  10.31    0.02   0.99   8.43   9.62  10.29  10.97  12.28   2000    1.0\n",
       "pred_yields[4,25]  10.13    0.02   1.01   8.23   9.42  10.17  10.81  12.09   2000    1.0\n",
       "pred_yields[5,25]  10.47    0.02   1.03   8.43   9.75  10.46  11.18  12.46   2000    1.0\n",
       "pred_yields[0,26]  10.95    0.02   1.01   9.06  10.24  10.96  11.62  12.96   2000    1.0\n",
       "pred_yields[1,26]  10.69    0.02    1.0   8.68  10.04   10.7  11.38  12.66   1887    1.0\n",
       "pred_yields[2,26]  10.65    0.02   1.01   8.74   9.96  10.64   11.3  12.71   1837    1.0\n",
       "pred_yields[3,26]    9.5    0.02   1.02   7.48   8.81    9.5  10.16  11.53   1927    1.0\n",
       "pred_yields[4,26]   9.43    0.02   0.99   7.49   8.75   9.44  10.14  11.25   2000    1.0\n",
       "pred_yields[5,26]    9.8    0.02   1.05   7.79   9.09   9.79  10.53  11.83   1898    1.0\n",
       "pred_yields[0,27]   8.51    0.02   0.99   6.64   7.83   8.48   9.23  10.37   2000    1.0\n",
       "pred_yields[1,27]   9.21    0.02   0.98   7.24   8.54   9.19   9.89  11.14   2000    1.0\n",
       "pred_yields[2,27]   9.07    0.02   1.03   7.11   8.38   9.08   9.76  11.02   2000    1.0\n",
       "pred_yields[3,27]   9.01    0.02   1.01   7.09   8.31   9.02   9.67  11.05   2000    1.0\n",
       "pred_yields[4,27]   9.26    0.02   1.01   7.25    8.6   9.29   9.95   11.2   1977    1.0\n",
       "pred_yields[5,27]   9.15    0.02   0.97   7.24   8.48   9.16   9.82  10.97   2000    1.0\n",
       "pred_yields[0,28]   9.51    0.02    1.0   7.57   8.84   9.51  10.18  11.45   1869    1.0\n",
       "pred_yields[1,28]   8.06    0.02   1.04   5.99   7.35   8.05   8.74  10.15   2000    1.0\n",
       "pred_yields[2,28]    9.7    0.02    1.0   7.73   9.05    9.7  10.38  11.61   1899    1.0\n",
       "pred_yields[3,28]  10.13    0.02   1.01   8.12   9.45  10.13  10.82   12.1   1747    1.0\n",
       "pred_yields[4,28]   8.99    0.02   1.02   6.93   8.31   8.99   9.67   11.0   1948    1.0\n",
       "pred_yields[5,28]  10.26    0.02   0.99   8.31   9.61  10.26  10.93  12.18   2000    1.0\n",
       "pred_yields[0,29]  10.61    0.02   1.01    8.7   9.91  10.61  11.28  12.59   1882    1.0\n",
       "pred_yields[1,29]  10.58    0.02   0.97   8.72   9.92  10.57  11.23   12.5   2000    1.0\n",
       "pred_yields[2,29]  11.04    0.02   0.99   9.03  10.36  11.06   11.7  12.96   1923    1.0\n",
       "pred_yields[3,29]  10.73    0.02   1.01   8.79  10.03  10.74  11.41   12.7   1986    1.0\n",
       "pred_yields[4,29]  11.18    0.02   1.01   9.14  10.52  11.23  11.81  13.23   1935    1.0\n",
       "pred_yields[5,29]  10.45    0.02   1.04   8.42   9.75  10.45  11.18  12.49   1942    1.0\n",
       "pred_yields[0,30]   8.17    0.02   0.99   6.25   7.49   8.19   8.84  10.18   1836    1.0\n",
       "pred_yields[1,30]    9.4    0.02   1.02   7.34   8.71   9.41  10.08  11.47   1858    1.0\n",
       "pred_yields[2,30]   8.74    0.02   1.06   6.64   8.05   8.77   9.46   10.7   2000    1.0\n",
       "pred_yields[3,30]   9.43    0.02   1.04   7.47    8.7   9.41  10.16  11.41   2000    1.0\n",
       "pred_yields[4,30]    8.7    0.02   1.03   6.74    8.0    8.7   9.36  10.84   1930    1.0\n",
       "pred_yields[5,30]   9.36    0.02    1.0   7.45   8.68   9.37  10.06  11.27   2000    1.0\n",
       "pred_yields[0,31]   8.68    0.02   1.03   6.73   7.98   8.68   9.37  10.67   1862    1.0\n",
       "pred_yields[1,31]   9.19    0.02   1.01   7.23   8.53   9.19   9.88  11.12   2000    1.0\n",
       "pred_yields[2,31]   8.62    0.02   1.02   6.64   7.92   8.61   9.27   10.7   2000    1.0\n",
       "pred_yields[3,31]   8.58    0.02   1.01   6.54   7.93   8.55   9.26  10.54   1961    1.0\n",
       "pred_yields[4,31]   8.58    0.02   0.99    6.6   7.93   8.59   9.23  10.52   1863    1.0\n",
       "pred_yields[5,31]   8.76    0.02   1.01   6.84   8.07   8.75   9.41  10.78   1725    1.0\n",
       "pred_yields[0,32]   8.82    0.02    1.0   6.85   8.16   8.84   9.48  10.79   2000    1.0\n",
       "pred_yields[1,32]   8.95    0.02    1.0   6.98   8.27   8.95   9.64   10.9   2000    1.0\n",
       "pred_yields[2,32]   9.17    0.02   1.03   7.14   8.48   9.16   9.87  11.17   2000    1.0\n",
       "pred_yields[3,32]   8.16    0.02    1.0   6.19   7.47   8.14   8.84  10.14   2000    1.0\n",
       "pred_yields[4,32]   7.88    0.02   1.04   5.85   7.15   7.89    8.6   9.89   2000    1.0\n",
       "pred_yields[5,32]   6.82    0.02   1.01   4.93   6.13   6.82   7.52   8.81   1631    1.0\n",
       "pred_yields[0,33]   9.85    0.02   0.98    8.0   9.16   9.85  10.51   11.8   1986    1.0\n",
       "pred_yields[1,33]   7.96    0.02    1.0   5.98    7.3   7.96   8.67   9.88   1901    1.0\n",
       "pred_yields[2,33]  10.19    0.02   1.01   8.15   9.53  10.19  10.85  12.15   1962    1.0\n",
       "pred_yields[3,33]   9.05    0.02    1.0   7.09   8.39   9.02   9.73  10.99   1906    1.0\n",
       "pred_yields[4,33]    7.8    0.02   0.99   5.89   7.11    7.8   8.47   9.74   2000    1.0\n",
       "pred_yields[5,33]   7.85    0.02   0.99   5.89   7.16   7.87   8.51   9.73   1977    1.0\n",
       "pred_yields[0,34]  10.57    0.02   0.99   8.74   9.86  10.57  11.26  12.58   2000    1.0\n",
       "pred_yields[1,34]   9.19    0.02   1.01   7.13   8.51    9.2   9.88  11.24   2000    1.0\n",
       "pred_yields[2,34]  10.84    0.02   1.03    8.8  10.13  10.84  11.54  12.87   2000    1.0\n",
       "pred_yields[3,34]   9.51    0.02   1.01   7.58    8.8   9.49  10.19  11.51   2000    1.0\n",
       "pred_yields[4,34]   8.83    0.02   1.02   6.81   8.16   8.84   9.52  10.75   2000    1.0\n",
       "pred_yields[5,34]   9.18    0.02   1.02   7.16   8.52   9.18   9.88  11.15   2000    1.0\n",
       "lp__              -161.3    0.06   1.59 -165.3 -162.1 -161.0 -160.1 -159.1    718   1.01\n",
       "\n",
       "Samples were drawn using NUTS at Wed Apr 24 14:51:53 2019.\n",
       "For each parameter, n_eff is a crude measure of effective sample size,\n",
       "and Rhat is the potential scale reduction factor on split chains (at \n",
       "convergence, Rhat=1)."
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "fit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0 of 2000 iterations ended with a divergence (0.0%)\n",
      "0 of 2000 iterations saturated the maximum tree depth of 10 (0.0%)\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# carry out some diagnostic checks on fit\n",
    "\n",
    "stan_utility.check_div(fit)\n",
    "stan_utility.check_energy(fit)\n",
    "stan_utility.check_treedepth(fit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "samples=fit.extract()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a218c29e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.figure(figsize=(10,10))\n",
    "for i in range(0,2000,10):\n",
    "    plt.plot(fit.data['temp'],samples['fdy1'][i,:],alpha=0.1, c='b')\n",
    "plt.xlabel('Temperature [$^\\circ$C]')\n",
    "plt.ylabel(r'$dy$ [tonnes ha$^{-1}$]')\n",
    "\n",
    "plt.savefig('./figs/2d_Gauss_temp_post_sample_growth_curve.pdf')\n",
    "plt.savefig('./figs/2d_Gauss_temp_post_sample_growth_curve.png')\n",
    "#plt.title('Growth Curve')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEKCAYAAADTgGjXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAIABJREFUeJzsvXeQ5Ol5HvZ+nXNPTw47s2l293b3di8t7nDAEcCZAkEeAZIiTRdpmiXRpGHRJkt0uSjKsgzRkmkVLVss0hTDlQ2GYnDJJcJUWQQFmEEIBxxweXOasJNz59z9+Y9nHr6/mdu929mbmZ2Z/Z6qru7+9S919/e9z/dmY60VBwcHBweHB4HvYd+Ag4ODg8P+hSMRBwcHB4cHhiMRBwcHB4cHhiMRBwcHB4cHhiMRBwcHB4cHhiMRBwcHB4cHhiMRBwcHB4cHhiMRBwcHB4cHhiMRBwcHB4cHRuBh38BOo7u72x45cuRh34aDg4PDvsLrr7++bK3teb/9DjyJHDlyRF577bWHfRsODg4O+wrGmMn72c+ZsxwcHBwcHhiORBwcHBwcHhiORBwcHBwcHhiORBwcHBwcHhiORBwcHBwcHhiORBwcHBwcHhiORBwcHBwcHhgHPk/EweGgwFqRdvvdz+22SLOJh98vEgiI+Hx4zWcRvHZw2G44EnFw2IOwVqTREKnX8dxoiFQqIq2WPppNEEirhf3vBWOUTGIxkWhUJB4XCYdBOCQZB4cHgSMRB4c9gHYbJEHSIEHUavqamoXPJxIKgQC8zz6fkgqf+Wi3ce5yWaRQwLZAQCQYBJnEYiCWSATbHBzuF7tOIsaYYRH5fRHpF5G2iLxsrf3VTfv8mIj8wvrbooj8tLX27fXPJkSkICItEWlaay/s0q07OGwr2m2RahXkUathGwU+TVXhsGoNkYhqDptNU9YqYfh8G8mDpq5WC1pIswlCqdVEikWQVquF83Z06CMWc4Ti8P54GJpIU0T+W2vtG8aYpIi8boz5srX2imefcRH5uLV2zRjzPSLysog85/n8RWvt8i7es4PDtsDajcRhrRKCtRDkfj80i3AYj1Do3edpNlWrWFsTWV3FeekjIRHRjEXiocZCE5cx6kOpVkUmJtS3kkqJZDJ4pFIgs1AIxzg4ELtOItbaORGZW39dMMZcFZEhEbni2ecVzyHfFJFDu3qTDg7bjEYDq/5qVckiFFJNQQSaRiwG4vBqGo0GCGN5WUmD52o0lBCopQQCSg7GqCmLBLZ532AQ+8XjIum07l+piExOgljCYWgnXV0i3d1KcE5TcXioPhFjzBEReUpEXn2P3X5SRL7oeW9F5EvGGCsiv22tfXnHbtDB4QOi0YDgr1YhqGMxPDca0ES4LR6HQBeBaalchhBfWhJZWcFztQrCoR8kFgPxUKDTD0LTVbuN69DpvtlHQngd75u1lVgMx1sL8iqVRObnQTYdHbg+taZo1EWAPYp4aCRijEmIyL8RkZ+z1ubvsc+LAhJ5wbP5o9baWWNMr4h82RhzzVr7lU3HfVZEPisiMjIysiP37+DwXqjXQR61GgRrIgGhXKlsNBfFYvi82RTJ5UAUq6sQ2EtL2L/dxvE9PRDaIjjeGPVvZLOqmfBRryuhtFo4LhjE9fjsFfp0sjcaqp2IqGO/3QaJRKM4X7Wq2hOJxKtNOTwaMPa9YgN36qLGBEXk/xWRf2+t/Zf32Oe8iHxBRL7HWnvjHvv8oogUrbX/672udeHCBev6iTjsFmo1kEe9ruQRCIjk86pFxOMQxCIQxKUSCGRuDgRCokkmIczjcTWBtdswZRWLOCabxXOxqD6QYBDXCYXUbBaP6/HeCC5qGcZoqDDJJZnUh8+H+yyVlGQCAfhL0mnsE43q9RhOHIu5EOL9CmPM6/cTuPQworOMiPyfInL1PQhkRET+RER+3Esgxpi4iPjWfSlxEfkuEfmnu3DbDg7viWYTAr1eh9BMpyFMCwUQiN8v0tmJlTqJgOQxOwuTVbMJ4ZxKYSVfr2Pf2Vl1nlOQl8u4rjE4JpHAcfRRUIsgYayu4j21BZrEvGHBNJeRAOfmRBYX9RqdnfCJiOC7ZrM4bySCa3d3iwwMqB8mn8f3p3ZCLcrhYOFhmLM+KiI/LiIXjTFvrW/7RyIyIiJirf0tEfmciHSJyG+Ac/4mlLdPRL6wvi0gIn9krf3z3b19BweFtRCUxSIEbzqNFXmxCEe4iEY2NZsgg0oFwndhAYLYu7JvNmHGajY1BJfnZ25HKiVy+PBGLaFahdBmUiLzS/jszW4X2egHiUZxfT6MwfmqVRBKq4Xz3LgBIshkQCahELavrcH5nkiIDA6K9PXp/fl8OEe1inMnk45MDhoeijlrN+HMWQ47hWoVmkSrhZV2KgWhms/rtmQS+3JVvrwMAqH/QwT7Vioq1I2BtkGzVbutfgZ+3mrhs2wW16xUcC5jNOqKkVf0i3jzRehw9742BvfL7xKPYxt9K5WKhhEHAvi8owP3Ra2q2cT2/n6R3l6Ro0fVAe8NX6a25bB3sWfNWQ4O+x2tlpJAIAAzjt8PcqjXISQzGQjwYhEEsrSERy6nGeONhhJEZ6dGQK2u4rh6XUmB2kitpmYumpliMVyPoGOdeSR8z2NElEBENpq+cjnc78oKNJR0GvfW2QmhX6tt9I0Uizi2uxtayPIyzjExge87PY176+sTGRrCOY3B90wk1MTnsH/hSMTB4T5hLYRnoYD3XK0zospazfQul7FteRnCdHkZ5iueJxSCSaijA/veuQPfR6kEkqBpiT4Lvx/HxePQFryaQT6vIcE0PTEDnc7yzQYHb+a7MaqJ8H2rhftYXcW9BQK4bn8/CCMW2xgUUC6DIHp6QBrZLH4nEmuphNeDg3r80hJ+k3gc2zo6XIjwfoQjEQeH+wCd040GVs7pNAQeBWgoBCHYbsMZnc2COBYXsRrnPl1dELThMPa5dAmfLy/j3Cw1Qh8GBfzKykZNhERBYU/fxeZIKJKDtTiv349rUxPanL1OxzyJh853a3G/a2v6XTMZNVNRU6pW8f36+/H54qIGEbA2WC6nZBIIYPvEhJIQa3i5zPj9AUciDg7vg0oFAtQYCMZoFAKYEVX0I+TzIJqVFTwmJiB0w2EIzcFBjXq6dQvPPEckgvPE4+rcLhSwWvcKYEZSUfOoVDaG6fLBGlreOlskA5GNWe7e4+hL8T7y61lc4TC0L79f74dRV6EQ7pMRXf39+L5Hj4JYZmbwXSsV3HsuB5/J0BBMWvE4ts/NafJlNOrqd+0HOBJxcLgHrN2oaWQyEKD0c/j9WE1bC+G3sAASmZzE60AAQnJwEJrD9evQOubnQUrWqhM7nYaQbjZxbvoWmANSLqtTu9HA/RmzsXov8z5ENCdEZKNJzNtrhIRELUREHfDeRMRAAOdlBj3zTjIZJZJEAu+TSdzz2BjMYH19+P5DQ/gNFhfx2zDqbG1NZHgYhJJIaCQZfS6lkjrxme3vsLfgSMTB4S5oNkEIzSaEWzIJobuyAmEYjUL4F4sQinNzII/ZWRwzOIjVeLMJjWRsDORB8qHg5Uo7n4d2Qv9JuaxRUSQHagrhsJZOqVTUae6F18/BZxKJiBKKt9aWiPphuJ0FF2m2ElGT1OqqCniSSk8PwntrNXzfuTl8p2QSn3V1qemN33F+Hr/VmTPYh6Y27/UZkJBIaNSYw96AIxEHh02gucUYCD2WAlldhSDu6IBwm58HgUxPi9y+De2ipwcJd34/PqfmsbwMYZ5IqJnGWpxzYgLmnrU1zSSnkGRILM1N9TqIy5v7wf2oMXj7iXgrBIts1Fb42luwcXNXRNbz4vUZ1utNpqSDfW0Nv8fMDH6HoSGQ0uKi5sfQn+St++X3g2xmZ0WefFLk5EnVaIpFvR7Dp4tFRyZ7CS5PxMFhHXQeVyoQbpmMJvKtreF1ZyeE2ewsBN/4uMjUFD7r7cXn+TwE58oKCKRUgjCNRNT2n83iHNROGG5LLYMC3e/H9Ws1NWNRs2DEFrUN+jzoEyG8DnoR/cxLQl5th9cneTE3ZbOJKxZTM5y1WicsENDw4HQav0swiO9RLuN89boSal8fSKNUgtbS1yfy3HPQaJjxHwhAu4vFNtYkoxbkorq2H/ebJ+JIxMFBNkZfpVIQcCLq/2BE0uIiBP+dO3jQVDMwoOauQgH7LC1hWzQKIUiTDMN+SyXNQqeZKhrVEvH0hXjzQpiwJ/LufiGbneQi+hwIqL+DTaj4oPbCz9jUiufxdlT0ZrV7yY75Ktx3czFGOsrTaS39wkTJQAAa3/Hj+C3Hx/GdT58WeeYZkPncHMg9k4EPpd3eWB2ZQQlOM9k+uGRDB4f7hNdURfOVVyuh+WliAsRB81OjAVt+KqVO4qUlCDyW+fCWQFlYAHGwp0gopOVOGNbKfVmRl9oBw3OpFZAwvI2lvETiDd31mqy85U68BMLSJHywRS/9Ezy22dQkS6+jXgT3Ho3ie7DxVrGI34clWMplHNPfj8itQgHEPDGB3+f0aZGzZ0HGV6/itzx/XuTxx3E+hgyPjEDrY6l95spQM3LYPThNxOGRhtdU1dUFAddua5Ic6z/duiVy8yZMV8vLIBVvvsfiIlbQ5bJmoVcqEIYUfOXyxtV8JKKax8qK5op4NQCSBkmEWgAJiO/ZyMqbXLi5TtbdVulezYWRXyyPQsJgHgpzU+ig92o5Ihs1olBIKxUHAlpzi5n8LAtz7BjMV2NjSszDwzBnRSIg7WJR5MgRaCXd3divVsP/degQrlmradkV+l1c9eAPBqeJODi8D7ymqs5O7euxsgJBmclAc7hxA487d0AEXV1YSbfbMFtNTOCZjmOfT/MiCgXtCUITD6vuhsMgGGoeTAT0agws6R6J4Bj2/PBqFtQWRJRMvI5xrhOtfXcxRmoi1oIoeB6ar+j4JinQN8PzUNPhuanhsNQKtbpyGb91LgfTXyKBc128CN/QsWP4XScmQMYrK3CynzqF11NT+L/OnME2+k8KBWglqRRInRUFFhdxDfZxcdg5OE3E4ZGDN/8jGoWvwxgItdVVCEba4W/cgFllfh5Cd2QEwiqbhbCbmMAx9G3QaZ7PawVdESWCVArXovZCmz79CtQ2/H7tab651zobTnmr8nrb3Hr9GF6fxWZzFzUIL8nQ5FQuQyCzNHytptnyJCE6t9tt1YJ4Hn4Pb0gyzV7hMH7D3l78/jT9jY5C08hmofnV69BKnn8e515exrGHD4ucOIHfcn4e+3mj4lot/P6VilY9dvW5tg7nWF+HIxEHL9ptmK9qNS2lLqJmLXYcvHULyYHM3QiFsFqORqFljI/juVRSU9DKikYgkTxovmIEUbGo+4loXw8R9Y3Qr0BzFaOZvKYqai2xGPZPJrVIIo9hmC9Jhy15azXVIniPNEGxiZYIjmXvklwOvw+bYNFvQuc8v683zJi+GJ6b34f3H4/jnlk80hhEYB06hP1YTyyREHnqKWh/rBzQ1wefyuAgflMWdBweVsLwmrjo1HdRXPcPRyLrcCTiQLRaEOCtFrQP2uxZ1oS2+ytXtKbV6iqE19AQhObYGB5LSxrJxPyFbBbnYyQVQ13ZFXBtTVfy7P7HmlbRqCYf8j78fk0opGkrmYQwpPmGx7TbKtjpu6BTXmSjX4TkUqupz8Pbe8Tn0zIszOVot/V3qlS0MjHfk5y8bXhFNprH+D3plLcW52ZnREZzpVLQKpJJHH/7Nu7z5Ek8SIK9vdBcDh3ShESfD/9VOq1hzcyO9/tBWNToHN4bzifi4OABfR3WakMlETU/hUL47JVXYKdfXoawOnZMcz8uXoQGUq2qL4CrdNaSoumIK2/2P2c4rzcj3OfTFrJeYWrMRkc8m0ANDOBeQiElC0aEeU1UXvOVt7cICUREfR68D2od3u6J2aySFzUcVi6mFlEua9Z+Nov39JuQRElYdNCTLOgQX17GObq61JdSq4Ekk0lEbE1OQivM5+EXCYVwHDUtRsmxIjIJhtWHGQCxvKwVCJyvZHvgSMThwKPZ1C6DXV26SmYZ83AYxPD1r4tcvqy29FOnINjv3BF56y2Yr+j7YCl2+lZYvp3RVo2GlkPnatzbfzwS0cKFmYyGq/JcTGzs74cwJfmwjS1b8NJZH43qtb1hv4wCo8OduSDekikM600kcK12W0ubMHSW4b6850hEa4pRe2DNL35v9opnMACvzWrErIdFDajZhIbI5Er6XQoF/G8sH//mm/BNZTIamm2tRn9Vq/A31WogEvqWeno0C75Ww74BJwE/MNxP6HCgwWq7LGFCocHILK5Qv/IVONEbDQjnwUEInkuXQCAkIa6s2eKWbXFphmLxQFbYpXlnM3FEIhrl1Wjg/PW6ZnAPDmI17U1CZCjv5pBd7zmp1VBzYI0tEpO3xEk0qpoJo7MqFQjtXE7NTdWqBgtQo4rHcf/0O1CbSyQ0tJkaBvvBk9RodiPB8juVy3jf0bGxL0o6jeum0zBdra2pRpjJqDZC8193N86Xz4OYMhntVcIui9ksCImalcODY9dJxBgzLCK/LyL9ItIWkZettb+6aR8jIr8qIi+JSFlE/q619o31z/6OiPzj9V3/J2vt7+3WvTvsL9TrEGbeHBARCK9CAcJkehoEMjkJgdrbCyHeaol885vwj5RKSgLeKrvsYkifAc04fCaB0A/A0N7ubkQY1euagBgOI+LoyBGYWpjERxNUpaLmHzqmaaqq1XAdagv0efD63l4hFPTsq84QYpJVMonvT0KkuWppCU7u2Vn8pqur2pmwtxe/7+ysJhwmk9i+ugqtYHUV98me7tSoqHnRB+TVSPh7V6s419KSki8rClerGuElApJZXsZ9BYO4JnvWd3aqv6m3d2OQAEvcOGwdu+5YN8YMiMiAtfYNY0xSRF4XkR+w1l7x7POSiPysgESeE5FftdY+Z4zpFJHXROSCiNj1Y5+x1q7d63rOsf5oguG6fj+EDhPPWMAvHEb01Ve/qr6Szk74HQoFka99DeG7LIPCHhoLC/jcWo2iolOZZiLmUHi1AQqus2fx2Z072mtkeBi+F2+2NUmCQpXOeGoc1D64H/0bBBMS+ZoERG2Dzn/mnFBD8d4zNRqvX2VtDYEF165tJI1MBqTCniLMTG828XutrOA75/MbAwuYnCmi5VEokhg0EIngv6EDvVJREx/Di43BPkePggQTCW0SxgCIjg6NYCPo06L50PUuUexZx7q1dk5E5tZfF4wxV0VkSESueHb7fhH5fQuG+6YxpmOdfD4hIl+21q6KiBhjviwi3y0if7yLX8Fhj4PhuqzJxBUmCSQQEPnWt/CoVCDQUikI+fFxkVdfBVkEg9AaRLBqnZ6GIKZGQYFGwUwtwJuzQcf4mTM4ZmwM5wqHIfBOnMB2rzbA3h0UrtGoFoRMpVTQeYsqMi/DG0rrdbR774llTthel74LajLN5sacFBGNIgsEUILkiSewwr9yBWbA5WX8vp2dqIHFPBiSMAliYQEPfr9EAvfi7ZeSSGhEG/NU6LA/cgRku7ioEV3JpJaer1a1FL8xIKFQCNtzOa3CTBMWS/GzlTErDDvcPx6qT8QYc0REnhKRVzd9NCQiU5730+vb7rXdwUFEIJzW1iAYvARSLOLRaIj85V+q/yMchtDq6hJ57TWRt9+GsGHf70YDK+v5ebxmHgcd6zS3cLXLlb0I9jtyBEl04+PQfEIhkMfRo1qniwLbG8XE0h29vdpKltdicUQWPaQfhFnj1Ii8meNekiPh0OxEYuH3oEnOu4rn6r3ZBEHw2i+8IPKhD+H3fOMNCGKWi+/qAtEsLuLeGC3FJMHVVSXxeBykQY2OxNpsahXk2VmNmBsYwLnpe+K5czk43hsNON+pBUajOA+d8PR98Xfs6dE8GBKfi966Pzw0EjHGJETk34jIz1lr85s/vssh9j22bz73Z0XksyIiIyMjH/BOHfYLqIGwlzmFQKmkeRxf+xr8HxQu8TgExle+opFZAwNY9edyGzPS43EIUwp6rnqZW+EtRdLfr0UDX38dxw8Oipw7h3vjuazVFThNb319ePT341wMuaXA4wqcuRY8h4iG5DI/xWvq8makl0qqjfA4b8Z8IqGaEAmTIcde4snlsP+xY8jPuHULpq7lZTVDpVIQ+OXyxh4iyaSWi/H7tRQKTWBesx19QcvLuPdiETkj/G/Hx+Fn6ujAGHj9dZz35Ek1ayWT6g8j4bLcDf1m1FYbDe1k6fDeeCgkYowJCgjkD621f3KXXaZFZNjz/pCIzK5v/8Sm7X+9+WBr7csi8rIIfCLbctMOexr3IhDavOfmRL7xDYTp0qxDB/Zf/AW0hHZbzSXMSi+VcJ5kEseVy7pqZ8QTixEaA4H12GMQnDdv4viuLuQ6jIxoiDG1CjZ66u2FzZ+hrKwsTHJgyRRvgypvKRF+33ZbTUCbux2KbMxMZ8iwN8y30dD744o9GFTHM6O36nXcdzqtJBYOIyy6p0c7PRaLOL6jA9fmKj+TgXkwEAC5s4tkJKK/L39jan9MqLQW/xeJZHAQJrLxcVx7eBjnu3gR/9f587i+CK5NsxYz672+EJoLGb3FvByHe+NhONaNiPyeiKxaa3/uHvt8r4j8jKhj/destc+uO9ZfF5Gn13d9Q+BYX73X9Zxj/eCDTvTNPhCG4Y6NIUx3fl4Fb1cXhOHrr2P1HAzCxBQMakVZ5mPQ/0FzCwUsQSf6yAhWwysrEELRKHwep0+DhJg/0m5rQcVMBloHe5hQy2m3N2aoMwx2K2CNK5qxSC5egvFGbYmoAKd/hDXF+L0ZTGCtdis0RutTUfDTmb6yAmFcLGpCI1f57Tay0aemtHkXkxR5HwwAYL5LIrGxjH4mAxI5elS1zUQC5MLeLb29qArc26tkzETKdBrn2exw97ZHTqcfzTDgPVv2xBjzgoh8VUQuCkJ8RUT+kYiMiIhYa39rnWh+XeA0L4vIT1hrX1s//j9f319E5Jestb/zXtdzJHKwwTBevx92cRJIrQZb/NWrMK8sLGA780UWFmA7n5qCkBgdhaC7cQNCr9XS/Aza5Vm/iiYO5iYkEjDn+HwalTQyglU5E9oYqsus9GRSHeUkvlpNI6kYmbRTdnk61711tVhPi4RFHw1/t0YDq/hGQ5t0iUB4UzuhwKWpjL6PQkHzagoFPCeT0L5WV+GPmpzEaxGNTGOuCFv9xmJ6XTrO2WHxySexfWpKCbzdhsbT0YFCjocO4fvR7GWtkrS3FI6INs5iImYqtTP/xV7FniWR3YYjkYMLJultJpB6HZrExYvwabDDoLUQcMvL0EwWF3HcyZMQdlevQkiKaPY3e5rX69jOcifMUGf1WBYg7O8HgXDVy0ztZBKfxeN43d2NY9mnnOejv2O3QX8JI7ZIKNRQvL6fVgu/CfNnOjrwPZl0yO+bSmHfpSUIdmaxU1NbXdWii6kUHPOXLoHgm00lX2o33oAClq5pNvGb0o9z5gz+j4kJ7M+imZOTuKfnn8f/Ew5D+8hmlfxY0XlzdFYuh+/qrfj8KMCRyDociRxMeDPRu7tVO6jXYY66dElb1FLAGwMCuXhR+1qMjmL/Gzcg5OijYG9zhpl6mzaxgu7gIASRCN739uKc/f0b8y64XzwO0xVNJyxXTifzXgktpY+DDyY3kkBJJqxKXCziONaoyufxaDSwraMD33NsDMRNLWVhAf9HuYz9aAr86lfh32A4NbPZveXv/X783rGYJhwynPfECTzm5/Gf9vfjMzYT+9CHQC4M4c7ncQ4Sibc9MnG33jMHHXs2T8TB4YOCxRQ3E0ijAYfr5csQGKurWsmWhf4uXdoYJnrtGuzypRIEPwV5sahhseyZwbawHR0QTJkMzs1kuyNH8EwfSXc3rsNjMhkIS9bsEsH7vdY4id8pGlVCoYbCkikkk1gM34GVfdmiNpFQwTs1BcF8+rQ61BlBFotB2K+saGTcD/8wouXefBPH+/3YjyY/lrlfWMC1urvxm3I9fP067vXMGfwP8/PYp6MDv/u3vgVCOnkSiwxvLhDBcitEIoHrsoijN4H1UYfTRBz2FbzVeJkbwO2XLoFAVle17Lq32u3ly9jvzBmYjG7dwuq4VtNyGHT+suESCYg5JT09IAuanGIxrIgPH1YbfV8fBBQJiQTC0FW2z00m95cgajaVXKmdbG7DW69rXkcsht+TlZLbbfxn7Tay12lmzGZhfmJ9snQav/GdO8jpmZpSYmbOize6iq2K222tNGwtCOnxx7HP4qJGxjE35cwZRNJFo7gvNuNKpbRPy2bzVb2ufhtvMc+DCGfOWocjkYODdhuCh+XcvT0rLl9G9vTiohb2K5W0eOCNGzju9Gk837oFOzmLHrLxk7dooohGCEUiyIM4fFjDetNp2Nf7+iBge3qQB9LTA2HLENh4HCvqeh2C0Jt1vh/RbmvuCsuOMF9FRLUU+nlYEr9YhCbA/JxiEVoCne23b6uPIhRSH9Kf/RlMkMykZ0Y//yc6+vv6VFOidtfdDaIIBrGQYCFGhiSzR0kqhX0ZWZZM4trsDe8lks0tlA9q10RnznI4ULBWJ66XQGo1OMSvXsVKlgRSKGBVSeEkAmFC2/z8vCYQMv+BOSF0qHvDV9kWl36Onh6U92B13XPnIIxo9mLHPjbCErm703Y/ggmHiYQWaKRmIKKCnGYnloNPp/HbFotYzUejiJYqFDQLnj1DqlVoIqmUyGc+g9/um9/Ef0QtjqVQGDo9O4v/hf9nJIIxceUKfF+ZDM6dTmvhyps38R+NjmqjK59PzY3UtLy5R4EACGd1FY+D8r8+KByJOOx5WIvJ2mhsTP4ql6FhMEOavT0ohAoFCCJjoIHkciAUdjckAVAQiuDcrIfFon6srGuMah/9/Rrmevo0VsHFooabJhIbtY+Ojv1lurofMAghHlfCZk4KkyFF1OxFDYEmJe4fiUD4M6T6xg39vysVLBCefBL/xZe/jN+VNbNYGJOEtbSE35r/VzCo2uvx47jXQkGTJMtlXWQcP65E0tGhPhIuYLxEwojA1VXsx8i/RxGORBz2PHI5rBpZHlwEgmRiAk7UpSUNw6SdvVjEytTvBwksL0MDyeUgCBIJtfEzP4PRRiIavjs0pH3H6Q9h+O7p3RhwAAAgAElEQVTgIBznbNBE3weFDjPYH4VVKn0IjNbyRlFRG2FJFW9VYJI9w4LpbL91C74Q9jMZGwNR/+APivy7f6ddDalJlkoaALG2poUWSWCbiaRYxH/FnJXbtzea5fr6QBreplebiYSLjLU17b+yOarrUYAjEYc9DUb8MATWWqz8ZmdBIPSBsOc3Henz8xBYAwMQOOPj2IcEwpUzBQRNICIQZGyvmkrhPfusDw7i/fAwXovgml1dGnnlLaB40LSP9wJLycRiGr5Mpzsr6bLLIcvIkEz4HzKqLRqFsB8b03PPz+M3/cEfFPniFxHlRbMVtQoSCX0rXV34D+t1jBURmK5iMQh/5rLk8yAuJlm22/jPu7vV4S/y3kSSz+O4Ry0p0ZGIw54FC+3RBt5ua5Oj69cR4kmnLKvsMpSXNbRWVkAgLAEfCKhT2Jv/wRyERALCo7dXM8sHB2G7Z5+KoSF8TudsZ6d2y3uUtI97we/HKt8bUMByJqwyzArCkYhqEcEg/q9qFYL4/Hn8jjdvaoHLtTX8xy+9JPJXf4X/tlzW6snF4sZGXiyemU7j2ouL+Hx0FPdQqWgflLU1aCQk/nYb/31XFz7zlt73EomI1hYrFjUL/lGBIxGHPQn2f4hEtAorO+rduIECf8xDYJ+PVguCnD2/V1ZgEmF+B5sgsZw6s6+bTa3yytLrLA9PB3pXlzqCMxkVPrStsz2sq/yqCIWwkmcOCf1Mfv9GExYz96mZLC3hv0skELAQiWDRkM1qn/SlJZGPfQz/9dWr+E+52GCVXmbhM9y7pwcENj+Pz06cwH/KpEYRjK+bN/GaxSlHRjb6SIjNRMLS8iQSlmc56HAk4rDnUK9rRV5vhFMuBwKZmdGInjt3NAGNjagiEaw4Z2e1iCJ7bnv7fXvbyVKYDQ9Ds+js1HwPVpEludTrWveKmdQsKb+Xkgb3CthUK5fToAV2hKzXNas9EsF/6PPheXUVn584gf+QQREdHVpc8amn8Nnly/gv2MuEpMUwYCaTdnRg+9wcrnP4sLYvZgvk5eWNFZInJ7Ef2wO8l0bCMUAiexTKpDgScdhTYPVUvx+CnATCKBqWF89mYcqgmSSX01axy8taXoMJanS8M8eDTZusVQJhGC/Ll5w4oU2jhoa0MKHXVGMt3nsL9zm8G+zwGI2qv4KZ797Iu4EBCHLmlywsQNMcGsL+U1PYRmd6LgeitxZEUq2q9sr+6ay6nM3iP2fi59wc/s+BAS1Vby2OX1lRjYRBAseOgSToW2NI+GYiYWSYd3wcZCJxJOKwZ0CTlQgmJgmE0TljY1pifGJCy25wsvp8WtiP+QP1umogXF1S++AE7+rCSrOnR8nkxAnNdxgZ0WKE6bReh6XIA24W3TfYa55aSa2m/iQ6whn9xp7yMzMQ+MzTCQQwBhjqXauhFLzPh5wQEgnzQUolrb7MhUQmg+vNzuqY6OrSSLtQCNfwRmzdugXTZjqNMSWiPpjOzo1EwWTHXA5a1UEmEjf8HfYEuCJlYhc1kkYDQuTmTQgaOspbLUxUCiJrMaGXlzXDvFrF/ixrQmFBEwdNVnSad3VBSDDxzOeDeYslyJNJ3Fe5rPkgB1Uw7CQ2ayV0tNdqWsGXmeDs4Dg9DSLp6sL/xerHPh/+82oVZG8t8oa8PhIRkEelguswiov9W+7c0bHR349j8nlsW13V8jfGYCFDjYQFO6mZbCaKeBzvs9mDTSSORBz2BHI59TUw1r/VguC4dg0TMZuFSYt9OVgYkPWZmETIiB/a1FmBlo5Sn08J5PBhrax78iQEBKO22KM7kdAMaeYfPMrRV9sFr1ZSLmstrLk5dYQPDOC3DgRgylxe1t7z/G8CAZBPrYY8HmvhO6NGkkziP2fYMet+iSipTE1Bs5yfx3hgv5FmU30kgYA20jp+XIkkGNxIJF4wLP0gaySORBweOopFzQUR0QJ3S0soqri0hIl8+zaEC30l7E/BzoZsr3o3AmEYL/uYsw4WBcaZM1jh0lY/PIx943EIApoturv3d92rvQZqJV7HuzGa88P/h76p8XHsx66QhQIEut8P01S9DiIRQUQXzZps8EU/CUFyyOVAUqOjILGhIc0RqVbhYwsG9dy3b2PBkUxqNFYup90SvWAmO/utbCaa/Q5HIg4PFdUqVnEsgsjonKUlOEqXljDxxsawL1dyXL02GuowjUQwoVdWtDc4NRBW5e3s1Da2/f14f/o0Iq9KJQic4WGtQstzsoWqC9/dGcRi+I1XVzXpk34SaiPHjuH/uXUL4yQeB7nU6xD+gQDMXtUqfCTGKJGwtL2XSBhuLILX9MedPg0Tal8ftCEGabACAs1mExMYR+xpwrHI2mJexOM4hr6Ug0QkjkQcHhoaDQ3lDYUwAf1+JZC5OU0Aq1Q0y5n5IpyUrO5aLGrZE29HPmogrIN15AgERG+v1r0qFJRAwmE1oZRKd6/k6rD9YKmZbFbNW5UKNARGUDEk9+pVjAOWRanVUGDT78f+1SpIp9WCP61chvYSiWi7Y5q2OK5IJLduYVywpXJPD8YVO2b6/dBU6HA/dAj3xCZe9KNsrqXFkig0fR2UPBJHIg4PBa2WTrZIBKvDQADmA+aCsN6VN7Fvbk7rJlETCYdBAouLqpGI4DP6QDIZCJXhYZDG0BAERX+/ZrgPDWntJhFcl+XcHXYH1BZDIS3ZTp9FswlTZF8f/uOLF/HfJRLacvfMGQj5iQkcNzqK427fxrnYnlhEqwCzXEsohAXI7CzG4pkzGFPNJu6JGszMDD7v69NM9qEhjJt6Xf1z7C/jRSIB4ikU8P4gEMmuk4gx5vMi8mkRWbTWPn6Xz39eRH5s/W1ARE6LSI+1dtUYMyEiBRFpiUjzfmrdO+w9eCOxGKJLU8b4OKJl5uYgOFj6u7MT2xYXtdseCYTd5hoNraPETHQKpdFRTPS+PtjQT5yAzXtpSQkkkQCJsA0ue4E47D4SCfwvFNLFIoR3s4n/MJ0WuXBB5J13MCbSaYyfpSVkuYfD0ECKRfzXrZZ2sEyltL1voaA+kkxGe59MTuLzxx9Xc2lXF8ZGuYx7Ya/3lRUQV1+f+l5YyJPhyl6QxBjdtd8XKQ9DE/ldEfl1Efn9u31orf0XIvIvRESMMZ8Rkf/GWrvq2eVFa+3y3Y512B9YW1OBTz9ELgfyGB8HeczOYpKxd/ncHB6MsPLmF6ysQLiwnWurBUFAAjl5EuYQJhCeOoUV4NISzn/okPYHZwmUri68d3h4CIVgSqKfLJ/HGGg24cOKRJCxfvUqhHq7rYuMs2dxzK1bGFvUSO7cgYBnQiAbbNGh39mpRDIxgXOcPo1jqNHm8zpefT4dSz4fxg0XIzyGZe69SCY1LJ09a/Yr7otEjDGd97Fb21qbfb+drLVfMcYcuZ/risiPisgf3+e+DvsAXPmxDDhNWdPTWDmOjcEWncthMg8OYoXJVWijoR3zqIHUahpKSRIIBLDCO3kSJNTfD2Fw/DgEwcIChBQJJBLBuY3BpHcJhHsDXAgwyi6XA1GwplUwiP81GMQYYivcsTEUcPT74Vy3FmOh2cQCJZ/XoolMPmX+CImkVsN5/H5cg/1P2F89l8OCR0THFHNKGNXHelt9fe+O6mMbX5LNfi0jf79TZXb98V6uRb+IjHzgO1qHMSYmIt8tIj/j2WxF5EvGGCsiv22tfXm7ruew82DvCAp7Rj7NzsLUcPs2BEA2CyF+6BAm+82b2F8EEzsc1iiseh2TlT0rmCSYTmPid3fDVHXmjGYbLyzgHCMjagv3hv+6CKy9BTqhmTDKqgSNBqKwgkH8t6zSLKKtAp58Eu+vXcN52H1ybg4kwDppIlrdmTW2mENy4wau8dhj0Iqsxbjy+dT3wTFHIqHDPpnUgJHe3nePrXQa90Nn+340n94viVy11j71XjsYY97chvvx4jMi8vVNpqyPWmtnjTG9IvJlY8w1a+1X7nIvnxWRz4qIjIxsG685fAAwEovNiZhRPjeH1d61ayAQFlE8dAgT+OZNrY9VKmEyl0oba2O121oHi+aFkyc1H+TcOQgbTnL2A/ESCO3bLLrnsPfAxEISSS6H8cHGYMPD+JyReZOTGBNPrUuu69chqI8fx3hZWsJ4YyVmRoNRoKdSOk4vX8Z5H3sMCYnLyyCfRAKvGU5urUZ1iWgbYZZQYfteLzo6NuaZ7Lc6bPdLIs9v0z5bwY/IJlOWtXZ2/XnRGPMFEXlWRN5FIusayssiIhcuXLDbfF8OWwT7gLDabjSq5UwmJ0EgCwvqRO3rA9lcu6Z1jMplTPxqVcMtg0HtgU3E47B/d3dDqHgJZG4OpEMCoe2bvUdcCO/eRySiWgBNSjdvghhiMZgt2UHRGGi31oo884w610VUc11ZgVZLDYCmMq/TmwuNd97B6+PH1T/X0wOSWFhQguNiRgSvUync2/KyarvescZM95UVnQObnfF7GfdFItbav8nxNMb8grX2l99rnw8KY0xaRD4uIv+ZZ1tcRHzW2sL66+8SkX+6Xdd02DmsrWnhO5qeZmZgw758Gc/ZLLbTdvzWW5jc0SgmNKvukkBCISUQdqMjgfT0IJ/gzBntYre4qD6QVAoTns59lwOyvxAMKpGwGOOtW9BI2AeG2q4xcLwbAyIRgeZrDP5/Ee1cSGc3Q8ULBZwnmdRgjTfewDVPnoRjfX5eI7Dm5zX0N5/XFgQkI2opbNrlhTG47+VlzJeuLtWo9jrel0SMMf/a+1ZEnhSRd5HI/cIY88ci8gkR6TbGTIvIPxGRoIiItfa31nf72yLyJWttyXNon4h8wWC2B0Tkj6y1f/6g9+GwO2AkC/tPG4PJNzuLyT05ic+ZAxCPY8XHjoa5nBIIfSBeB6XPp/08Tp7EBD5yBKGZ7F63tITJPTQEEwQJJBp9NPo9HET4/SASOtwXFzeWIuHi4umnMbZu3MD//uyz+Gx8HD4xBmuw8yXDbRcWsB/zOag9l0oi3/42rjs6iggu1tvifQQC2jzLq5nQ/zE/j+M3R2R5iWRlZf+U2LkfTSRvrf0pvjHG/OYHuaC19kfvY5/fFYQCe7eNicgTH+TaDruLSkVLkDCBbGICE/TWLTyyWW3e09Ul8vbbWIlFo5iwjYaaw6pVnZQi6qD3aiAjI0ogsZgSCPtl+3w4p0si3P9gSC5X/PPzIJIjR7QbpojICy/gP2cP9Q99CNsnJmDqbDRgEiuV1BfSaOjYZfRUOKy93L/9bQj40VGYXZeWoJHU62q2Yu91OtN9Ppw7m4U5jLXZvGCYsJdI9nqk4P24EX9p0/v/fiduxOFgodHAqqxc1rLfU1OY6BMT6PtAAmEy4dWrmDzsfc0KvayhRNMAH+029j16VDWQJ5+ERhKN6mTu7cVDBOdLJh2BHCSk09rQir1n6DCPRrF4+eQn4Qu7fh1E88wz2q1wdBRmTr9fOy92d2vUFjWQRgPn8vuhYX/zmzDFnjgBollc1LHLmm+1GsZhuYznQgHnZahx9S5OAPpNRLQy9V7G+5KItXZcRMQY073+fvW9j3B41NFua64Hk69mZkAgMzPwdywuqgMxnYZZa3YWhFKraSMh5pW027riZEJhJALNo78fz+fOQSCwM53PB2HAsuF0cm4ujuew/5FMajQe8zvYSjccxlj49KdBJDSjPvkk3qfTiLoaHMQ4Kxa1QRYrRnNMsrAnK/9+7WsY62fPKpHE4yCNtTUtFMryKcvLIKDubpxzZgbX2oxAQDPkWd1hr2IrAY2f37G7cDgwoPNwZQUTMJWC6j4/j4n01ltaDTUUwufLy9BOGE5JHwlLxDMsmGasdlsTEVnS/fHHVQOhXTqTgYbCY1Op/ZvQ5fD+iMchnA8dgrAfH8fYYvdJY0R+4Acwbq5cwbg8f167Wp45gwUHQ31bLYzhjg5tRtZoaMFPvx/axte/jrF66hTG/dISFkOFAgigXMaD2fbLy9i/uxvENDt7d22DlaMZHm/3aJzpVkjEuR8d3herqyCRdFprU83MYPtbb2GFyJVcOIxJdPs2tjWbIBBjdOXH+lciKghCIZguRkYgME6fxkoyFtO2qamUhns6Anl0EI2CCIaHIZgnJjAGaR4SEfn+78fi4soVLG5On8ZipL8fRNLRAe2A5eJ7erSrZT6veUneyr9//dfYfuqURlgFgyAL9klhXlSziXuqVNQBPzd3d20jHAYJ1mqa/b7XsBUS2aM86LBXUCzCRhyPY0KurMAPks0ilPf6ddVAWDH1xg0c6/NpN8NKRUtM0P/h92tCYSaD1SMr8Z49C5JYXcXETSYhSBjZ4gjk0UI4rEEW1iIakERCs+j3fR/2uXED4/TUKYyZoSEdT9Q66MCPRkEujNhi33ZmwP/FX+Dax4+DnNi+t1jEa5a4Z8MtJsx2doJ4lpbu/n2iUdxPpQLC2WtwmojDtqBWg/nA74d2kM3C7lwqIfLl9dexXySiSVnXr2udq6UlTPh6XTUQht5yxccormPHcI1Tp+AHSac1a5grURa8cwTyaCIU0oRTa7GYWVnBAoTh3S+9hPdjYxDyp05py+RTpzCGmORKYopEVJgzUpBJr3NzIn/5lzjnkSMwU9FRzw6cuRyuxdB1+gZTKRDP2trdv08igcVZqaQVgvcKtkIi/92O3YXDvkarhYloLQR8oaD9HMbHEQ7JictWtePjeGb9ITaPYo8HEdVajMFniQQisQ4fxmrviSewimNzIlb8DYd1YjoCeXTBJldeIsnlMC6qVSw+vvM7sTCZnISgHx2FFn38OMay36/jMRoFMdEMWyqpJsKE1zt3QCSDgyCkmRmMffYiabVAFPSVVCogD3ZenJ6+N0kw2oy94vcK7ptErLWXdvJGHPYnrAUhVKsQ7uUyJlKthgnxjW9gsiWTGi8/P4/9jMGEqtW0iipDHtn7momBsRgI5OhRmCmeegoTemVFfSxdXUpU6bQjEAeMC/pI2m3VjhlB1d8v8tGPagRhqwXy6OtDsMbhwxjjLNnDaMJgEIulSkXzR0SwGBofh4/k5EkQ1PQ0PveG666tqTmLTveODozdiYl7k0RHh3YBJbk9bGy53JwxJmOMedYY8zE+duLGHPYHpqawMjp0CBNkelqb9rz6KsxMJJBGAyswOs+zWZAGo2FYhkIEEyUQwEQJhWAeGB2FMHjmGS0R7/dr1VUW6HPNpBy8YK4QiYSLHFaDPnwYDa6iUW05cPQojnn8cRBKo4GxGgphfKXTWpq+Wt1YNLFWg6n2a1+DfyUSAUmIwMRlLe5pdRULKhHcx9ycRodNTm6cDwT9M4GAVjJ+2NgSiRhjfkpQ8PDfi8j/uP78i9t/Ww77AYuLIImeHkyu2Vntc/7665isqRQma60GolhdxUQoFLAi9BIIJ0QwqL2zWdH3+HEQx9NPQxgwkTASAUmlUjgulXIE4vBusPPgyAgWO5OTIAtWhR4dRbhvPK7+ucOHMfZoNq1WtW4biypaq03WWFG63cZ+V69iIXXunPpMuNBiZ0UGnxiD+2AtrloN8+duEVvsecLjH3Yy4lY1kb8vIh8SkUlr7Ysi8pSI3COmwOEgY20NpNHRgcf8PDSLtTWE8l67hgnpLaDIQneMVmESFxtNGYNJnUioWau3V0Mwz52DqYGVTqNRTFwmlCUSzoTlcG9sJpI7dzQCsFbDQuXUKYwhOriHhrBoOXMG22s17aKZToNIWA3YWi0MyjynK1fwOH9e50C1CiJh+fiVFRAMe6+vrGBhls+DYO6WH7I5q/1hJiNulUSqrNZrjAlba6+JyKntvy2HvYxCAZOAjsaFBazeqlWsvt5+W8MSSSCrqzqxCgWNta/XVQPx+UA81aoWZDx/HhrImTMwDTARMZHQ3iGRCI7bzy1GHXYHJJLhYYyjqSkdi8ao3y2dhhCnJnzsGHwcwSDMtSJaPodVFlZXtZ8JNZJiUeTSJZjJTp/WyC42Y2Ol4aWljUUfi0UNGvH2J/GCCbXN5sNNRtwqiUwbYzpE5P8RNIX6U0HHQ4dHBJUKCMTvh1NyaQmPdhsrrm99CxMtk4F6XqthMrJHw9qaEgcfjNKKx1Ur6eyE83xgABP4iSe0BEomg2NounLFFB22Ar8f44pEwozxeh2Ln5ERLFwyGYy5QADbzp6Fb87v18gsNp2KRjHWV1YwHgMBdcjncohQLJVwPE25a2vQ4GMxXHtpCfvW65gzzaZWfLhX6C81cV7nYWBL9SGttX97/eUvGmP+SkTSIvLFbb8rhz2JWk0n3MAAVl4rK/js+nUQSK0GDYKZupychYJm+1Ldr9fxHAioGapWw4R84glM5OFhOD1rNRBIT4+GZ3Z0YAJ2dDy838Rhf8Lnw/iyFouixUVo1dUqxtPAgEYILi1Buxgexuflsvr/UimMxUZDSWN5GXOgUNBS87kc/CMf/zjOTYc6TWD9/ZhPU1Pwz4hoafhwGKa3UOju/r5Y7N39T3YTWyIRY0xYRH5IRI54jn1SXHOoA49GA2p1paI1f5aXMQlu3kRF03xeW31ms2przuXwWb2OydtsYnu9jkkai0F7yeXwfO4cJmx/v8hzz+H6pRImVKWCScIMYkcgDg8Knw8+DzZJW1vDeKpUMPa8Ca/LyxDix45h7HJcl8sYh9SOazVsCwRAMMWihuJmsyKvvCLysY9h7LMV9NycNkdbWUGI74kTOA+d6LkcQodPnrx7syqWZaHmtJstdrdqzvpTEfl+EWmKSMnzcDjAYEdBkgRVb58PK6c33tCic8EgXhuDVVs2CwJgHSI601stJRA6343B5BkdBWE8/zwmF8tnVyrYt6sLx7mOhA4fFD4ffB4DAxDa+bwmAQ4MQMvo64MGXK9DKzh5EqG/bFRVrapGzOx1hv7Sd1et6sLr1Vdx7mQSGlC1Ck2jXMZ11ta0+2KphHtiGZaxsXs70Rlgsts5JFttd3LIWvvdO3InDnsS7TZWR8z0ZWlqY7CCeu01hEvSuT0/r+aqtTV1EjI2ntEtjK5itdN2W1vadnejcVA6jQlBc0E4rN0POzsdgThsD/x+jdhaXNSqCiSQVkufl5Yw1ln6/fJlNcnG4yCaVgvjdWkJGk08jn3ZVG1+HhGMTzwB7YLRhtRAurtV8zl8WBdYqRTmIZtpbR7/7NXOApC71dBqq5rIK8aYcztyJw57Dsyyzech8BkTT3PVG2+gWxyT/VhYjp/n81hJ0XFerYJEjNGQXG7r7ER/h+5umLMGBjBh2FKXkViplCMQh+2H3w+nd3c3xh2r5gYC2MbeNZ2dGM+JBJJeR0ZwfKOhiYB0rLNdrrWYP+02xnulAg3++nVoQdRcKhUsyNgVcWFBq/uyv04shjk4M3P378HOiCK7F/p7XyRijLlojHlHRF4QkTeMMdeNMe94tt83jDGfN8YsGmPuWkbFGPMJY0zOGPPW+uNzns++e/3at4wx/3Ar13XYGqyFxsHWoMZo9Ee5DAK5cgXvmaBVq0Gdz+dVC6nXMSmqVW0uFQ7rqq1UAjl8+MNYtT32GMxZpRImbiiECUryYJKVg8N2IxCAz6OzU6OjGG3FUifHj6v/I52GybWnBwKeApv+Oo5TluaJxzXMnR0YZ2cx7hkiXCphYRaLgaimp6FZMGKLJX4WFzWoZTOYQ9Ju3zuqa1t/t/vc79PbeM3fFZFfF5Hff499vmqt3XBNY4xfRP6ViHxSRKZF5NvGmH9rrb2yjffmsA46Da3dSCDNJgjk4kXtb86mO3SOsw1os6nZ6iSQSERXZYUCjr9wAZrHoUMwE9AxmUhgwiUSmKhdXVp/y8FhJxAIgCjabWgiiYT2OmcwiPfz7m6RZ58V+Q//AWM1HMYYzWQ2JtMuL8MkJqK5UoEAknJZWoVVfAMBBKs89hiuNzGhjnKee3VVI7buFo3FMPvdmC/3taaz1k6+12MrF7TWfkVEHqTF7rMicstaO2atrYvI/yVw8jtsM2iGYgIWCcTvB4G8/TZU72AQ+zBmvlLBZGGZ60BAV14imGBcoRWL+PzMGS149/TTahZIpzHRwmGs1Lq6dse+6+AQCkEbTqUwtlmuh10yo1F8ztwkFgTlfPD7MW47O/HMGnCsNs0Onqwdd+0a5gmz1Fnhd2wMFRt8PpAKkxfzecwHdm+8lxM9EtGeOjuJvWoYeN4Y87Yx5ovGmLPr24ZEZMqzz/T6NodtBPsVkEAYqhuJgDwuXsQgZgw9c0HolGQyYTCIycTckGAQKrrPp019Tp0CiSSTIBAmaDFbmC1wu7t3ZzI4OBDhMIgikdAIqUIBQr3Vwpg9cgRj1e/X7pqsm8Uw21RKtYFKBedJp0FU1aoGnty4oX4/lg/KZmHOGhjQWlzWYn+21y2X0Rl0P5U92Q28ISKHrbVPiMj/LsiOF7l7U6y7JvobYz5rjHnNGPPa0r3ahTm8C+xvLoIBzVDdVAr+j3feUTtsKKRhjIEAnICrq1gVBQKYOLkcBncopKsimrUOHYIjPZPBcyKhvg9qLn19GxtMOTjsJqJRmK5iMQhrLpAyGTyzDheLjD71FPKbWAfO79d6blwgscMhC4YWi5q9fusW5gnbPLML4sICrlMsYp9WC68bDSWd8fGH9ztttYrvDxtjkuuv/7Ex5k+MMU9v5w1Za/PW2uL66z8TkaAxplugeQx7dj0k9yi5Yq192Vp7wVp7oaenZztv78Ci0dBQQ2vh8GOi061bIm++qQ12QiHN/QiH1cnH5KhQSH0idCiGQtrlrasLSYTpNFZvg4PYN5nUEOD+fmyPRB72L+PwKCMeh7mVvr+lJSWHSgXCfWhICzI+95yWS/H7MR9YnocaST6PcZ5Mai5IuQxT8NQUttN53mjAV7K6CoJaWUEEV72ORVoohHtZXITW8jCwVU3kf7DWFowxL4jIp0Tk90TkN7fzhowx/cYggNMY8+z6Pa6IyLdF5IQx5qgxJiQiPyIi/3Y7r/2ootXC4GQo7ewsXvf0YFC/9pugMloAACAASURBVBqem02tG0SfSC6HicVY9nhcwxUZOx8IYNDXahjwzz4LchodRbx7vY6JQ0dkdzcmZiz2sH8ZBweQw+HDqpEsLyspeLshhsPQTp57DqRSLGI/YzDu6XQX0fDbdBrjPp/Homx2FoTQ2Yl5wUZZU1O4NkumsODp2ppqNTMz947Y2klslURYuf57ReQ3rbV/KiJ3ScK/N4wxfywi3xCRU8aYaWPMTxpj/p4x5u+t7/Ifi8glY8zbIvJrIvIjFmiKyM8IephcFZF/ba29vMX7d9gEJhOKQNjPzoIsensxWV5/HTZXaghMpmIfkNlZ1WBiMc34pQZijJY5CYVQlXdoCI9Tp3CeZFKdhuk0VlyupLvDXkJXFzRjJscuLMCURB/ImTMgkGAQGe1PPomxXSpByLOZVTColRu4cEsmcQ5mmk9N4XVvr84xnw8mK2OwyBofV1NYPg8SY5vq0i7XENlqvMuMMea3BWG2v7xeS2tLRGSt/dH3+fzXBSHAd/vsz0Tkz7ZyPYd7g7kgTGKansYEGR7GYP7WtxA5wtBGNr9hCZPpaQx2Eay8KhUcHwxiYrA0NVt9jo6COLq7MekCAZirGO6bTMJXksns/m/h4PB+GBhQ81KhoI7w1VWQxOnTWDAtLcE/srYGZ7jPp/Mhk8F2LqzW1kBQiQTIwO/H3JqchD+muxuaSTiM69++rde5eRPzKJ/HNYaHQSLcfrcaWzuBrWoi/4lAE/iUtTYrIp0i8vPbflcOu4JsFuSQSGBltbCAFU0wiIKK165pkiBLOzCcl6ozG/Q0Gtp/gRpIIKDNpUZGkOGbycAPkslob+pSCecYHMTqy8Fhr2J4WKMFWVg0nYZWEI3q2I5ERD7yEYzpQgHjPBZTfwoTeKtVkFAshkelgvlQKCAPpNHQUiZsfHXjht7D9euYg/k8jh0eBjExJHg3sFUSqYhIXESoTQRFJLutd+SwK+CgSya1BHV3N2yx3/gGVlDch5nodJwvLoJwWi1NgspmMcDjcY2mYv/0nh4QSCqF2kCHDumKi5V9+/th4nLlTBz2Mti4qqtLe6yXyxjzLBQ6Oop51dEh8sILIBkSRTismnoohPNRg0+lsPDK57WVAk1Z4TBel0ogjevXMV9aLRCGNzBmYADzc2JidxpVbZVEfkNEPixKIgVBFrnDPgJzQRhxcusWBvyRI6gwevkyBj3LkjDrnH0RZmeVWCIRDHr6NmIxrICaTW2q89RT0DAOH4a9mFpKvY4HP3PlTBz2A4xBxFZXF8Z9Pq9+wmIRwv3oUSywjhyBoz0UgpmrqwvzJhTC54EAzlcsYi6k0xqxVamAGJaXNUl3fBzXqlTwemQEc3hqCttYpqWzEyQyuwstA7c6bZ+z1v7XIlIVEbHWrskWHesODxfMBWHo7NWrGKAnT8KJfukSBnulAu2CPaNLJQzyyUn1e0SjWPEw6oq5HqxiGgqhjMnQELSRs2cxEaJRbe7T3Q3br8tGd9hPYHmUTAbjnyG9lQoeR4/CtBQIoFrv+fOYP6urmA/0B0ajGnyyuqrmq1YLBFIqYY6xinazCb9IOIzP5+ZgMltaAtkUi3jd14f5mN0FO9FWSaSxXsPKiogYY3pE5CHmSjpsBWyEEwpBY7h0CQRx9ixU4osXsXIplzFomfNBx93YmBJQKqXJV9EoBmyzqZ0MfT4QE52D58+rqYs23+5urOhcMqHDfkQohPHd2amRUqzI0G4jiGRgAAuu7/gO7MsOn319mtUeiYCA6Ff0hv7mckokjPQqFGDOSiZBIoUCzjczo/extAQt6Nixnf8dtkoivyYiXxCRXmPML4nI10Tkf972u3LYdrRaWOn4fBigFy9ixXT2LAboa69B/a1UtPYVI098Pmgg1EwiEW02FQ5jYLdaSiIi2huksxPPPT1KIKUSto+M7H4rTweH7UQ0Cq0jk9F5I6KLtdOnMdYjEZFPfQrzYHkZRMGmaswh8fmwgGMJela6zudx7sVFTcpdXYVGkk7DAU/iGRtTH8vy8h4qwEhYa/9QRP6BiPxzEZkTkR+w1v7fO3FjDtsHby5IVxdWMaurcHJbC0f67dsYwCz+1teHidBqQTuZn1dTFGsABQI6ERoNJRGG8HZ2YiV09CgmCe2+8bjWxHJw2O9IpbDqTyYxZ9gmd2UFBHH2LJ67ukS+8zsx/mdncRyTbOlo9/l0IReLaY2tYhHnY0fReByax+ws9rt9WyO8bt/WOXqvviPbiS27Mq2116y1/8pa++vW2qs7cVMO2wdvLkhnJxxws7Ow13Z2ooT1rVtqwsrnsVoqFNQvMjOjIYrNJvZjI6pgEPu12zoZzp2Ds3xoCGYslklhbxFGYjk4HBR0dUGzjkZ1fhQKeJ3JQCMJhWDi/ehHtbU0fRetFp5ZaJTBLx0dWkaIZVdoHYhEQBi5HI67eVNzrCYnMS93I2l3q7WzwsaY/9QY84+MMZ/jY6duzuGDg7kgLNR28ybI4/hxkb/+a7wvFDQKJJPRDobVKgZjo4HByNa4fj9WUSwDYa2G+54+DS2jvx8tbiMRLXliLcjlyBEXyutw8NDfj8WZiAafrKxgHg0MYM4Zg/yRJ5/EnGCV3lBIe+iwpDzDh1kpmNUglpZAMNRcrl/H3LIW5qy+Ppx7dhZzeqexVU3kTwU9PJoiUvI8HPYgmOeRSoEILl+GGnzunMgrr6hZi/bWeFy7FFoLgqlUQBbGaGvcREJLNZBA2A96ZASrsmef1d7oNGP19MC05RpLORxEGAMNe2QEGjsrNczPY54cPYoFVKMh8tJLIBWWmO/vV+2C9ebYNrdaxeLO58P++bwWPI1EcL4bN/C6VILloKtL2/zuNLYaWHnIWvvdO3InDtuKchnCm1myb7+NQf7EE3CqX7mihRPn5jThb3ISA+/2bc3CDQS0VEMiAVKq1fBot9VENToKorhwAdoInYKNBraPjLiqvA4HG34/xjk1AUYtzs5ikXXiBObm9LTI932fyB/8AUiGCYyLi5pDUq1i7hQKIJFYTJMQg0E82CphZQWLwvPnoX0Eg7A+5PM7/523qom8Yow5tyN34rBtqNVgumInwatXse30aQzYt97CoK5UsGppNrFCmp+H4GdNrEgE52Dp6kgE2oW12I9+kJ4e2Hp7elD24dQpzaClKW1oSDvBOTgcZASDCCjp7obQDwZ1sRaPYx729uL1pz8NopmcBFEwtDeRwHyzVnO7QiHt057NgjiyWdVg5udBJD09uFa1ioXdTmOrJPKCiLxhjLlujHnHGHPRGPPOTtyYw4Oh0YCJiqUVbt/G+6NHMSBfeQUhgaWSZrkePYrBmMth8C0uagw7Q3LZszkQwEqKGbodHSCQri5MnAsXQC75vJZ47+93kVgOjxYikY2dEUMh7bvT0QEiSaWwuPr4x0EEt25Bg2dobyyGOdhuq2UhEtH2usUiwvPZajoUwtyemsJ8nJtD2P5OY6vmrO/Zkbtw2BZ4c0FSKWgUc3Nw3HV2inzpSxhk5TI0kWwWg5atbbNZ7M+y7gw1DIUw8CMRDFgWdotENNlqZARRJ2yZS19KX5/aex0cHiUkEtDML16EJh+NYv7FYlhUnTwJs/Izz4BcXn0VGsnQEJ5pPmbIMMPqmZdVLMJ8trCgpVTabfhH4nFcZ25u57/nVqf2vIh8VER+TET+jufh8JDBUN52GwJ/ZQUrks5O2GK//nXU2snntYAi1eepKQzIqSktokhbbCgEbSKVAvmwJHwwCOLo7UVEynd8h1YzZVIiCcT1R3d4VEFNnZUcAgEErPh8WMAdPYr58YlPaEZ7Not54+2MGApp86pyGXOXEVu5HExZ7Tb2t1b7sT+9rX1n7w4XnXUAQAJh3Z1sFiuZWAwmpjfegKqcz2O/O3dADMPD2K9YxLZ6Xft7sOZOPI5z0slnLSbCwACOHxwU+fCHUZm3UMD5w2GstHp6XHdCB4eeHhBEowHB326jzUIsBhIZHISW8tJLWHgtLmrvEWtxDDsp0oxVrWqxRm9Gu4h2H718GT6SnYaLzjoAYCgfNYHpaaxIjhwBeVy6BM1kbQ3aCGv+TE3h2JkZLWctsrH/AR3p5bKudDIZkEZXF3wgZ8/iPCsrGMBdXdBQnCPdwQFz5tAhCP7paRAGq2ePjsLkVa1C0/jUp0S+8AXMyaNHtZ5WJILPSyXMdSYixuNai45+kY4OvF5ZwTV2/PttcX8XnbXHUCxiQDJvY34eA29gAKF+77wD8sjnMaCsRZghm9zMzeGzUAhqNVc5kQgGo98Pgmi1NGrk2DEQyfnz0EKKRWgg1oJAenpgRnMJhQ4OgN8PUuju1vm1tATfZEcHFmKdndBKPvYx7D81pb7IZlP9HHS0M+AlGtUuiYuL+MwYLAp3o7jpg0Rnve6is/YGWB+HERxLSxhY3d14/9Zb8H2QQCoVTXaamQGB0BEfDOLYchnnS6dBLCxpQl/JkSN4PnUKdYAYTlyrgVjY2MolFDo4bEQohAUcfY7hMKKnslksvk6dwtw5exa+jFYLc7q3V30i7NnTakH7ICGFQpqXxUKN7fbuLOR2PTrLGPN5Efm0iCxaax+/y+c/JiK/sP62KCI/ba19e/2zCUEjrJaINK21Fz7o/exX1OvqwPb7oVnkcjrIXn0VqnM+r/V1Dh0CObDp1MoKBhkzXSsVEEgqhYHOrobMSB8exmcnT0LtDgRw3XxeTVgdHa60u4PDvRCLgSzYhoE1r86fhzO9XAYBPPcc5uiNG5hnvb3QWljFt9GADCgUMIeTScgDFlENBKDV7EaL3K1W8Z0UkQ4R+cz6o2N921bwuyLyXn6VcRH5uLX2vIj8MxF5edPnL1prn3yUCYQNbPx+bU7DPh8dHTBhTUzAzDQ5CRMXVeXr1yH4Gc3BqryVCjQSVgKt17HKabcx0AcGQBTDwyKf/CRMVtksVj0dHRjkyaQr7e7g8H7o6IAvhIuzdhvRVKEQIh7pb/z4x0EsLDHf0wPCYH5WMKhE0mphgefzYd6urUGL2XOl4I0xf19E/lBEetcff2CM+dmtnMNa+xURWX2Pz19Z75goIvJNETm0lfMfdHjLusfjEORsGtXZiZULy7rPzYFE0mmo0WNjGFjz85qBzg6DLC+dSmGg0tFnDM7b1weiePFFTIC1NZjKOjpgworH8drBweH90deH0Hu2RiiVELGVSmGu9vaCQD7+cfghl5ZAGqkUjvf5sGDz+TSHxBici6au1dW9WQr+JwUtcj9nrf2coN/6f7H9t7Xhel/0vLci8iVjzOvGmM/u4HX3JLy5IIkEVijFoq5M7tyBakwCuX4dmsrp0xhM8/N4lMtqY61WcW52K2RGeqOhavLQED778IdRmXd1FYM6EsFgT6VANC6h0MHh/mAMtI6+Psy3eBwWglu3tHQ8y8t/5CMgkMXFjeVQIhG8N0YbUYXDWpSR/pGdxlZ9IkbgjyBa69u2HcaYFwUk8oJn80ettbPGmF4R+bIx5tq6ZrP52M+KyGdFREZGRnbi9h4KWIsqmcSAKRY1YmplBVpIsQgCuXIFg+vMGZDNnTvYzhIJfr8eH4thEIdCWNXU6xrdcfgwPnvmGTjSGYnVbsPElUpB03EJhQ4OW4PfD62ebatjMfg9olGQy9mzMF3X65jfb72FuceOoyyNwux11uniApHH7TS2unb8HRF51Rjzi8aYXxSYmz6/3TdljDkvIv+HiHy/tfZvfgZr7ez686KgTe+zdzveWvuytfaCtfZCT0/Pdt/eQwFLQlP1ZR8Pdgy8eVMTjq5exQAaHdXe6IuL2jfd59N4cxZ142CkGSsYhA8lHAYRvfSSllUplUAgrCzqEgodHB4MoZDW2LJW618VCtDyT5yAWevJJxFaT/Mzzc4ieM0wYCYJx2KYw3uuKZW19l+KyE8IfBprIvIT1tpf2c4bMsaMiMifiMiPW2tveLbHjTFJvhaR7xKRS9t57b2KYhGCOxbDIKpUMODabait4+Pa8ezGDZDF8DBWLNevw4TFFUkwiHOwlk8shgHo82nfj0AAfo5kEkmJn/kMXq+s4DE4CFU7HHYJhQ4OHxSJBELnIxFoJ80mFn7tNpzsR47ApPzss5iX+byWN2L4fTqtRMKw/XAY83mnsSVzljHml621vyAib9xl2/2e449F5BMi0m2MmRaRfyIiQRERa+1vicjnRKRLRH7DgGoZytsnIl9Y3xYQkT+y1v75Vu5/P6Ja1WTARkN9GPRZjI1p6fbr16EO9/fDDHX5MpzfXgIpl0EggQAGHVcsTCgMBuHf6OqCtvHSSyCk+XloMwMDGMiBgEsodHDYLvT0YHE4PY05WShgbp84AQ2EScDPP4+W1tks5mEioSbqVEo7kq6t4fNjx3b+3rfqE/mkaA4H8T132XZPWGt/9H0+/ykR+am7bB8TkSfu9zoHAbSVBoPaV0AEJBAMYpDNzOD9xAQisTo6EIc+Pg4CWVxU0mg08PCG8obDGpsuAtMWEwZffBFdEBcXoel0dkLFZll4l1Do4LB9GB7GHF9chOa/vIzF47FjcLQzFP+pp5AHtrqKxV6zieMYvZXLaTZ7obDz931fJGKM+WkR+a9E5NimDPWkiHx9J27sUYc3F4Qht+02iCUYBEnMzGD77dsID4xGIfTn5lAyYWFBexnU6yAQEewXj4NAmLRkLbb390M1/shHUH6BzW9CIajU4TAGuEsodHDYXhgDwqBDPBYDoUQimJdnz2K+s0nVxYt4TqUgL1gapdkEebC6707jfjWRPxKE2v5zEfmHnu0Fa+09cz4cHgztNghEROPA221tRXvnDgiC/pDLl7HfuXMYOBMTGDz5PPav1zcSRTyufpB8HoMuHIamEYvBife3/ha2Ly/j+cQJHBuJuIRCB4edQiAAImk2tX/IzIxWxj5zRv2ixSLmP2vn0SRNIqlU9lCIr7U2JyI5EXlPU5TDBwdzQeifYBtafjYzA9NStQqyuHgR+5w9i2Noxspm1V/RaOCzaBQPFltk+fhwWEN1z55FSRM2tCkWNXrE73cJhQ4OO41YDD7NsTGNopye1soRjz0Ggjh/Hp8vLWmUJRtVpVIavbnTuK/oLGPMG9uxj8P7g7kgJBBroTE0m1hVLC5CVZ2ehgZSLGLApVIwa7GoojE4jucIhaBFBIMYbGtr6luJxaCFHDuGXJBDh6DJrK7i3B0dmrnuEgodHHYemQxMWAx8KZUQNNNqYU6ePIkoyaefVi2ErXNbLV3wDQ/v/L3erznr9PtU6zUi4oI9PyCYC0ICoSbBSrk0Ua2ugkDm5zGQDh9GaO/CgpqfaAZrNqESx2K6WimVsD0UwvaeHqxwXnwRmgg1mYEBrdfDWj0ODg67g8FB9YWKaMO3/n6Yl8tlaCRnz4q8+SY+z2Q032u3TM/3SyKP3cc+rfffxeFeYC4Io6h8Ps0DKZXgB2GZk0uXEImVycDUNDGBVcrCAojD78czO6nRlxGNYoCx2GI4jHNkMqjRc+ECBuLyMjSbwUGtxxONPuxfyMHh0YIxyBHxmrQXFjBvOzrgHykWNTjmnXcgK5jR3mxqMM1O4n59Ilut1OuwBbAviM+nqqi1Wrrg1i2YsBoNlDO5cQOC/cwZaCMzMxqJZQyOaTZxHobxBoMgqGwW52GtrI4O1MR6/nmNColE0EDH74e2wqJvDg4Ouwu/H2bmGzdAJK0WFow0Sz/xhLZxYNuHUgk+TFb33Wk4C/dDBk1V1DxY6rnVgrC/fh2Dot2GBnLxIgbW6dMYNCSQXE41F28kVjis5ixvE5tYDBrIk08inDcS0d7Ox45hkLIVrksodHB4eGANO1aXaDbh+6zXMT+fegp+TJZIYT5ZLLY7PkxHIg8RjYa2lWX/cmojrRYIhBrKtWsib7+N7SdO4JjpaQymlRW8F9Hy7XSy+XzQJIpFbToVjUIDOX1a5IUXkP+xsIBBefw4PrfWJRQ6OOwVpFLwUXJhWCphzrbbSAA+dw6fnz2LeVssYt/dKB241X4i/58x5pHKGt8psJghhb/PB42h2QQJ3LyJCCq/H6rsG2/AkXb0KNRYtrclgVirxwaDIBBjMKDqda3wyXDe0VEQyOgozlMuY7WTTuNcqZRLKHRw2Evo7dXePcGgJgL7/SgZ/8QTII2zZzH/s9k9FOLrwT8QkV8xxvyOMWZgJ27oUQAbSzWbGsLLYml+P+yai4sglclJlIBeW9Noqfl5rEKYcEhbKdttJhI4NpHAdlbvDYexbXgYBHL6NIisXIYa3NOj+SS7Uf3TwcHh/mEMrAbptOZtLS/DTM1qwGfOgGgeewwy4LXXdv6+tlQ7y1r7hoj8R8aYHxKRPzfG/ImI/C/W2sqO3N0BBJMJGw28ZvQUI7MmJ1GyxOcDWbz5JsxWvb0Q/isrcKzNzGiFzlZrYyhvIKB5IUxcZMXevj70b378cY38SqcRiUUScgmFDg57E34/5ADnfi6HxeShQ1gknj+v7XTrdcz1ncaWfSIGZXSvi8hvisjPishNY8yPb/eNHVQwyY8lTBIJaAKBALSPW7ewX6kEDYSdzg4fhklqagokwn4izEYnGfn9IIJ0GoOpXFYC6ezEoHr6aSUYRmIZowmFzpHu4LB3EQ6DSKJRmJ2rVRCJtZjnFy6AVIaHd6eK71Z9Il8TkVkR+RURGRKRvyso6/6sMebl7b65g4ZsFs5tCv1EAmTBGlZXrqhj/M038T4eh72z0YBGwoRDhv+SjJgL4verY61YxLZQCNd6+mm0t81kQCCsEBoIOEe6g8N+QjIJ8zZ9nGyFy5Inzz4L60W5vPP3cr9VfL9HRN4Wkf9SRK5YS3fw3+BnjTFXt/vmDhIKBRBGo6G5F/yDazVEXlWrIINvfxvv/X6sKPx+daSvrakpjBFddKSTCJpNLSEfi4FAzp0T+ehHYc5iWZSjR7VMgnOkOzjsL3R2YiE5P6+ljMJhlIfv6cGC8dChnb+P+/WJ/JCI/DNBY6hrxpi3BKTylohctda2ROR7d+YW9z/KZU0WJIGwDa3PB62jUMD2N98EiTQa0EDicRRYYz4IHel0yLN0iTE4PhCAuYuJhpEIwnZfeAHOc7bPHB5Wx7tzpDs47D8YA23DG1yztKTtGg4dAqHsNO7LnGWt/an17oL/m4jcEJFxEXlRRL4lIpPr+4zt1E3uZ1SrENwsqphKaWn2UAiksbYGDeL6dURTFApQVTs6oKayNwgz0ZmUSE2DeSHJpJY+oXlraEjkO74DPhU64/r7sYppt0E6zpHu4LA/4ffDutDRgYWgzwcZUK3CL7ob5Yq22tnwJ6y1f5MnYoz5DRH5+e29pYMDdias1UAYyaQWR4vHkfsxO4sBMDaGbmXeFrT1uvYGqVa1sKLfjweTCdlrYGEBPhc2jerqggZy4gTup9nEfn190GScI93BYf+DRRlbLSwMs1nIgqGhvZmxnjfGPMM31trXReTkVk5gjPm8MWbRGHPpHp///+2de3Cc5XXGn2dv0motybo6+IIt7nWwwY5iICSUSbmZEJxLp4FmGpK0zWQmaZJmMlPadAhDZtokvU3TpE1Jy+RaYJImjUsJlwIJKeCA7BgMGBvH4SJssLCRLNvy6rKnfzzfO98noZXtRdqV5fOb2dnd97u9+nb1Pnve855zSPKrJHeQfILk6sS260g+Gz2uO8a+V5VQmXBoKI7dSKXkF2loALZulUA0NMgE3bBBU1ZtbTJRAdUG2bVL1khIZRL8ICGteyYjYXjtNVkwIaVJY6PmRM85J87o29wsgQopVlpadLzjOMc3hYLGjUJB/+cHDyqGpBoicqxDyEcBfI/k0wA2AlgB4FjzRH4LwNcAfKfM9rUATo8e50FLic8j2QrgCwC6ARiAjSTXm9lrx3j9GWdsTPEchw7FApLLaaDP5xULsn17nPJ9wwYFGDY26hdFqJ++a5dWWIVfGOELERzppZIEZGgoXm3V3KxtISdWOq39mps1R5rNag7VHemOM7dobh7vH+nv12Kc006b2esek06Z2bMA3gbgTsjJvhXAlcd4jgcBTFVSdx2A75jYAGB+FB1/OYB7zWxfJBz3ArjiWK5dDUI0eggeLBRkGfT3a5Dfs0dJFENg3y9/qdogob55oaAPPqR+DzElIa9WSOlupmmwVErnTDrWzzgDeOc79bpU0jkXLtT1R0bcke44cxFSMxltbbFPNJeb+ese82RGtBLrB9FjJlgE4MXE+96orVz7rCEIyOCgLIBCQdZFyG9z8KAc6cWifBEbNug9GTvS9+zRNNb+/bJkkhZIECVA521u1pTY2Ji+ONmsHOiXXqovESmxCBXSQsldd6Q7ztwkldLsRJi9qEbc12zM4juZm9emaH/9CciPkewh2dPX1zetnStHSGcyOKgPMgzyISajVNLS3f37JSBPPAE8/LB8J296k0RgYEAR6vv2SXBCLYAQhR4EJDjSX3pJDvfgSO/oUH30sPJq3jzNkzY2uiPdcU4UstnxFslMMxtFpBdAsjLwYihKvlz76zCzm82s28y6O6qQC9lM/o79+/W6sVFO69dei5fRPvqoLJLOTmXoffBB+Ts6OzXVdOiQlvgGIQoZeTOZuLpgCCxcuDCOXJ83T9dratIU1qJFEp/mZglNqI9eKklAPCLdceY++bzGoGpMZ81GEVkP4EPRKq3zAQyY2W4AdwO4jGQLyRYAl0VtNae/Xw8zDd6trXFkeYhAf/llDeovvAD87GcSi/Z2TWONjqpeSF9fXP8c0IAfYkAyGb1ftEirLvbu1RRVcJCfd55SQBeLamtv15co1Cdpbq7OF8pxnNlBU1N1fJ9VX+BJ8lYo31Y7yV5oxVUWAMzsG5DT/koAOwAcAvCRaNs+kl8E8Fh0qpvMbCoHfVUYGJCAjI3pV39ra+wQLxSAjRsVLNjUpIH//vvj2JAlSyQOW7bIFxLKXCYtkOAcIzXtFZbu5XIyVzMZicfbZckTJAAAFwFJREFU365tYYnw/PmxIz049x3HcaabqouImV17hO0G4BNltt0C4JaZ6FclDA7Kohgelni0taktVBB86ik5yevrJSr336/lvfPmyQE+b55yZL3ySnwcoCkrM1kYyeJSqZQEqFTSlBYALFsGrF0rAWlqktCEZb6jozpHc3PNbpHjOHOc2TiddVxw8KAsi8OHNcC3tant4EEJyHPPyccBSBQeeECxIfm8Eh82N2v7yy/LkjlwQMKRy8XPyZVYra2aChsdVcQ5IEf6VVdJxBoaNNXV2KhrjI7GGX0dx3FmCheRCjh0SP6LQ4c0uIeAvxAx3turKarhYQ3qDz6o1VikLJCODlkoIRZkcFDbgoBkszquVJIgdHZq/6EhHRuCCt/9bglFLheXti0U4pK7ra3ViVh1HOfExYeYY2RoKPZftLVJQIpFiUFdncRlyxZZFq2tWpXV06OBvatLjvSdO2WpDAzoEVZwmcWxIMEfsnixlvIODMTTVA0NwCWXxFNcXV0SnYaG2JHe2uopTRzHmXlcRI6BUEHswAGJR3u7po36+2U99PcrqeK+fZpy2rIFeOghObeXLpUf4/nnZVUMDGiJbqkkSyIkVgy1k0ntv2ePps0aG3W9TAa44AJFpZdK8ok0NUl4Mhldq7nZU5o4jlMdXESOkmIxFpCODg3opZIEI5XS1FYQkM5OBQ0+8ID2P+kk1QZJRqP398d10YNoFAoSI0AWyIED8pnU1+scpZJqKL/1rdoW6gUUChKi4WG9Dr4Ux3GcmcZF5CgI1cMGB2MLxEwWQiql7Zs3SyTa2rSC6p57FCty0klKgLZ3r5Is7t8fr+gKAmIWC0FYeVUqyWcSYkNGRlRc6tJL1Y9Fi/RoaJDVUSz6SizHcaqPi8gRGBlRQsT9+yUeIQB+7149j43JaR5iQfr7gbvuktXS0SEB6e/XyqwQUxIEJJWSBZLP61EqyYqpr5clYyYLZnRUYnTVVepHZ6cEJZeLl/JmMr4Sy3Gc6uMiMgVBQAYGYgEhJSBhBVSIBZk3T+Kwfr0EpbUVOOsstW3frumnV1/VMxmnH6mri5MjtrbKkti+XcIQBKSlRQJy8GB83lRKx5VKOk9bm6/Echyn+viwU4bRUQlIf7/Eo7MzFpBQG33rVuXBCiVq77hDjvOmJkWRj40BTz+twb+vLxaQsJS3rk7iUyrJcd7WJgEpFuXvCIkcL79cotPUBKxYofaQWM1zYjmOU0tcRMowNhb7QBYs0MC9b19conbbNj0yGYnCnXdKUAoFVRNMp4Enn5Rw7N4dp4cPxaRCrEcy2+7zz8tB/6Y3xelOLrpIKUzy+bhKYUODrjs6Gqc3cRzHqQUeSVCGXE7WQGOjrIekM/zZZyUgpAb0e+7RtFZdnSoKptNKZzI0FK/oCrXQh4clEM3NEqr6eonGrl2yetradE0AWL1aS4OzWVk2uZyul83KWgnR6Y7jOLXCLZEykBroSa2yKhYlIDt2aBorlJh98EEVlkqnNejX18vRfuiQBKS/X+cLU1j19fJxkBKDBQs0RfbqqzpfW5usk+XLgbPP1nmXL5e1EqyTYjGuo+44jlNLXESOQH+/ggzTaUWab90qa2L+fFUmfOQR7bdqlaayNm/WIL93r44l48C/VEpCEZzyCxZIbHbtikvYjowokLC7W9c84wz5POrr40SOvpTXcZzZgovIFPT3a0oqnZa/YutWDeLz5yvF+89/Lqth1SqJw6ZNEpg9e/QYHY39FaEmOikxCXVEXnxRorB4sc69bJmCCdNpLQ9esEAWS5gKy2a9OqHjOLMHF5EyDA/LSkinFfT3zDNxuvUtW5TWvVjUaqnWVvlARkY0hdXXJ39H8FeEsrQhLqSjQ6+ff16iEARk4UJgzRpZHKedpvZsVu/DijAXEMdxZhMuImXI5eSH6O2VgAwMyAexbRtw990SmJUr5RTftCkWkGCB5PPxwN/cHBeZam+XMOzcqfeLFslKWbBAU1hNTbJGFi3SMWE1l5n8Jb6U13Gc2YSLSBmGh+VEDwLS1KSB/447tNrq3HM10AcL5NVXYwGpq9PAT+q4kKG3pUWi8MIL4wtLNTVpSqytTauxliyJLRAgjgXxrLyO48w2XETKMDKi6aa9e+X0fuEFCcjBg7JAFi+WBVIsavXW7t06JpSyDT6QUKWwpUUrql58UfstXBjXDVm5UoK0dKnOGyoaptMeC+I4zuzGRaQMhw5JQObNk0D85CfKW7VypQb7nh5ZK/39qvdRLMYCUirF6djHxmIB6e3VMQsWyKqoqwPOPFP1QJYs0SOfl4Bks9q3qcljQRzHmb3URERIXkFyG8kdJK+fZPs/kNwcPbaT7E9sG0tsWz9TfWxt1WDf1wf8+McSkBUrlM9q40YN8Pv3y7I4fDiuhQ7EBaKCNVIoKAvw4cNyqgdL49RTlQfr5JMlIA0N42NBCgWJmOM4zmyl6rPsJNMAvg7gUgC9AB4jud7Mng77mNmfJvb/EwCrEqcYMrNzZ7qfBw5oOutHP9J01YoVcniHZbxDQ9peLI63FBoaNPgHaySfj0vptrfHItHVpXMGCyQISF2dxCaf91gQx3FmP7WwRNYA2GFmO81sGMBtANZNsf+1AG6tSs8S7NsH3HqrprTOPlvWQvCBHD6s1Cch8C8QLAczTUPV10uAQqncQkH7L12qcy5ZotcNDfE0VrBq5s+v9l/sOI5z7NRCRBYBeDHxvjdqex0klwLoAnB/ormeZA/JDSTfU+a4j0X79PT19VXUyVCB8KyzZIFs3iwLpFiUgAwPSxDCFFY+LwEh45xW/f0SkODXyGQUZLhihaayli2TYBQKcTR6LhenRXEcx5nt1GLR6GTDo5XZ9xoAPzSzsUTbyWa2i+QpAO4nucXMfj3uZGY3A7gZALq7u8ude0re8hZg7VoN7Js2aZXURAFJpTRt1dAQWyDB2ti/Xyu5CgWJSjotATnnHOCUU7QKq75e24KAZDIeTOg4zvFFLSyRXgBLEu8XA9hVZt9rMGEqy8x2Rc87AfwM4/0l00appBVSGzfGAhJqfeRycVnbYEmEFO35vPwlg4N6H+JEFi7Uyq4zzogtkLCCa2RE5/PCUo7jHG/UYsh6DMDpJLtI5iCheN0qK5JnAmgB8EiirYVkXfS6HcCFAJ6eeOx08NJLSq5YKsXVCYOApFKxgDQ2xkWigkUxOKjXTU0ShwUL5AM580wJSKglkstJoEgXEMdxjk+qPp1lZqMkPwngbgBpALeY2VMkbwLQY2ZBUK4FcJuZJaejfgvAv5IsQQL4peSqrumElIUxMBBPYSUtkHw+tkDy+VgQBgbiLLtheurNb9bjlFPUNn9+HEMCeDoTx3GOX2qSSMPM7gRw54S2Gya8v3GS4x4GsGJGOxfR2amVUtu3x8WoMpnxPhBAYpDLaUpqogXS0qJaIGF5cEigGBIqmmnZr6czcRzneMWHrzJs2CB/SLEo30g2GwtIsECy2dgCOXgwrpmeSsm6WL5chaq6umKfR7BASiW9z2Zr/Zc6juNUjotIGcxkXUwUkLBUN5eLLZPDh7VPoaC29nYJyKpVr7dAzOKEip4Py3Gc4x0XkTKE4MDDhzXoJ3Na1dXFFsTQUFxrPZuVv2P5chWWOvnkuJZIPi8BGRvT+2SQouM4zvGKi0gZSqXYiZ7MaRUsEFJTWEFAcjmJw4oVsYAAskoaGvTaBcRxnLmGi0gZBgfj2iC5XPycyUhcDhzQfg0N2tbSIgE5/3wFFYYprJBXK2TzdQFxHGcu4SJShlNPlQiEqPN0Wo9MRm3BL5LNajXW2WfHApJOx9NhgMQo+EQcx3HmEi4iZThwII5CT6UkGqlUXHe9vl5tTU2qcvi2t6lUbrBAXEAcxzkR8BjpMmQyWq4bLJBUSvEi6bSmpIKArF4tAensjJ3oIYZkdDQuies4jjMXcUukDGNjsQ8kk1G8CKDX2azEYtUqYM0aCUiYwmpslDM+CIhXJXQcZy7jIlKGpiY9p1JaxptKxQ729natwFq1Ko44DwJSKsVxIG6BOI4z13ERKcPoqKyLkZE4Oj2bVTLFNWuUKn7+fG1raXm9gPgqLMdxTgRcRMpQLEpAQm2Pujqlc7/wQq3ECgISLJBkMkWPRHcc50TBRaQMwaEeCk0tWgRcfLHSuYcMvcECSQqI58JyHOdEwkWkDKFSYT6vBIoXX6xU7qHIVLBAkvVAPBuv4zgnGj7slSHUBenqAi66CFi6VKKRzUowGhu15Ddk7HUBcRznRMSHvjIsXSrxWLJEj3nz5Otoa9PrEDPiBaUcxzmRcREpQ329Agnz+Tg/VkimGMrktrZ6SVvHcU5sXESmoKMjTgPf3h7XUK+vl1M9rNxyHMc5UanJ72iSV5DcRnIHyesn2f5hkn0kN0ePP0psu47ks9HjupnqY8iP1dAgMamr0xRWoSALxAXEcRynBpYIyTSArwO4FEAvgMdIrjezpyfseruZfXLCsa0AvgCgG4AB2Bgd+9pM9LW+Xv6PTEYxI01NcV4sx3EcpzaWyBoAO8xsp5kNA7gNwLqjPPZyAPea2b5IOO4FcMVMdDKXi3NilUqavnIBcRzHGU8tRGQRgBcT73ujtom8n+QTJH9IcskxHvuGGR3V9JXZ+OJSjuM4TkwtRGQyb4JNeP/fAJaZ2UoA/wvg28dwLEh+jGQPyZ6+vr6KOhlqiLS3ex4sx3GcctRCRHoBLEm8XwxgV3IHM9trZlHydXwTwFuO9tjo+JvNrNvMujs6OirqZAgi9DQmjuM45amFiDwG4HSSXSRzAK4BsD65A8mTEm+vBrA1en03gMtItpBsAXBZ1OY4juPUgKqvzjKzUZKfhAb/NIBbzOwpkjcB6DGz9QA+RfJqAKMA9gH4cHTsPpJfhIQIAG4ys33V/hscx3EcQbPXuRTmFN3d3dbT01PrbjiO4xxXkNxoZt1H2s+TdjiO4zgV4yLiOI7jVIyLiOM4jlMxLiKO4zhOxbiIOI7jOBUz51dnkewD8PwbOEU7gFenqTvTjfetMrxvleF9q4zjtW9LzeyI0dpzXkTeKCR7jmaZWy3wvlWG960yvG+VMdf75tNZjuM4TsW4iDiO4zgV4yJyZG6udQemwPtWGd63yvC+Vcac7pv7RBzHcZyKcUvEcRzHqRgXkTKQvILkNpI7SF5f474sIfkAya0knyL56aj9RpIvkdwcPa6sUf+eI7kl6kNP1NZK8l6Sz0bPLTXo15mJe7OZ5H6Sn6nVfSN5C8k9JJ9MtE16nyi+Gn3/niC5ugZ9+xuSz0TX/zHJ+VH7MpJDifv3jRr0rexnSPLPo/u2jeTlNejb7Yl+PUdyc9Re7ftWbtyY3u+cmfljwgNKUf9rAKcAyAF4HMDyGvbnJACro9eNALYDWA7gRgCfmwX36zkA7RPavgLg+uj19QC+PAs+05cBLK3VfQNwEYDVAJ480n0CcCWAn0LVPM8H8Msa9O0yAJno9ZcTfVuW3K9G923SzzD6v3gcQB2Aruj/OF3Nvk3Y/ncAbqjRfSs3bkzrd84tkclZA2CHme00s2EAtwFYV6vOmNluM9sUvR6EinTNSG35aWQd4rLG3wbwnhr2BQB+B8CvzeyNBJ6+IczsQag+TpJy92kdgO+Y2ABg/oRibTPeNzO7x8xGo7cboEqiVafMfSvHOgC3mVnRzH4DYAf0/1z1vpEkgN8DcOtMXX8qphg3pvU75yIyOYsAvJh434tZMmiTXAZgFYBfRk2fjEzPW2oxZRRhAO4huZHkx6K2BWa2G9CXGUBnjfoWuAbj/5lnw30Dyt+n2fYd/Cj0KzXQRfJXJH9O8h016tNkn+Fsum/vAPCKmT2baKvJfZswbkzrd85FZHI4SVvNl7GRnAfgPwF8xsz2A/gXAKcCOBfAbsh0rgUXmtlqAGsBfILkRTXqx6RQZZivBvCDqGm23LepmDXfQZKfh6qMfj9q2g3gZDNbBeCzAP6DZFOVu1XuM5w19w3AtRj/w6Um922ScaPsrpO0HfHeuYhMTi+AJYn3iwHsqlFfAAAks9AX4ftm9iMAMLNXzGzMzEoAvokZNNunwsx2Rc97APw46scrwRSOnvfUom8RawFsMrNXgNlz3yLK3adZ8R0keR2AqwB80KKJ82iqaG/0eiPkdzijmv2a4jOcLfctA+B9AG4PbbW4b5ONG5jm75yLyOQ8BuB0kl3Rr9hrAKyvVWeiudV/B7DVzP4+0Z6cr3wvgCcnHluFvhVINobXkDP2Seh+XRftdh2An1S7bwnG/SKcDfctQbn7tB7Ah6IVM+cDGAhTENWC5BUA/gzA1WZ2KNHeQTIdvT4FwOkAdla5b+U+w/UAriFZR7Ir6tuj1exbxCUAnjGz3tBQ7ftWbtzAdH/nqrVS4Hh7QCsVtkO/Fj5f4768HTIrnwCwOXpcCeC7ALZE7esBnFSDvp0CrYZ5HMBT4V4BaANwH4Bno+fWGt27BgB7ATQn2mpy3yAh2w1gBPrV94fl7hM0tfD16Pu3BUB3Dfq2A5ojD9+5b0T7vj/6rB8HsAnAu2vQt7KfIYDPR/dtG4C11e5b1P4tAB+fsG+171u5cWNav3Mese44juNUjE9nOY7jOBXjIuI4juNUjIuI4ziOUzEuIo7jOE7FuIg4juM4FeMi4swJSI5FmVGfJPkDkg3TcM5ukl89wj4LSf4wen0ujyIj8MT9SF7NacgUncwS+0bPVeb83ye5j+TvzsT5neMTFxFnrjBkZuea2dkAhgF8PLkxCqA6pu+7mfWY2aeOsM8uMwuD6rnQOvwjMW4/M1tvZl86lr5Nwa/N7NxpOtc4zOyDqGHQrTM7cRFx5iK/AHBa9Mt8K8l/hoK7lpC8jOQjJDdFFss8ACD5VpIPk3yc5KMkG0leTPKOaPuNJL9L8n6qDsMfR+3LIusnB+AmAB+ILKIPkFwTnfNX0fOZZfb7MMmvRedbSvK+KLHgfSRPjtq/RdV6eJjkzqOxBqK+PUPy36I+fp/kJSQfiv6GNYm/7dsk76HqX7yP5FeoGjF3RakzHGdSXEScOUWUs2gtFHELAGdC6a1XATgI4C8BXGJKGNkD4LPRwH47gE+b2TlQyoqhSU6/EsC7AFwA4AaSC8MGU8mAGwDcHllEtwN4BsBF0bVvAPBXZfZL8rWovyuhhIfJ6bSToCjkqwAcreVyGoB/jPp+FoDfj87xOQB/kdjv1OhvWwfgewAeMLMV0X1411FeyzkBydS6A44zTeQTvoBfQDmDFgJ43lQbAVChneUAHlJaIeQAPAIJzW4zewwALMp0Gu2T5CdmNgRgiOQDUNK/qfwPzQC+TfJ0KP3E0fyivwBK3AcotcdXEtv+y5Rw8GmSC47iXADwGzPbAgAknwJwn5kZyS1QkaTAT81sJGpPA7grap+4n+OMw0XEmSsMTfQFRCJwMNkE4F4zu3bCfitxdOnCJ+5zpGO+CP2ify9Vz+FnR3GNqa5ZTLyeLG33ZCSPKSXelzD+/78IAGZWIjlicT6kifs5zjh8Oss5kdgA4EKSpwEAyQaSZ0DTTgtJvjVqb4ymxSayjmQ9yTYAF0PZnpMMQmVIA80AXopef3iK/ZI8DGWNBoAPAvi/o/i7HKdmuIg4Jwxm1gcN5reSfAISlbMiP8UHAPwTyccB3AugfpJTPArgf6LjvmhRHZUEDwBYHhzm0FTUX5N8CJoiKrdfkk8B+EjUvz8A8OnK/2LHmXk8i6/jHAUkbwRwwMz+ttZ9KUc0ZXZHtMx5pq7xregaP5ypazjHF26JOM7cYQxA80wGGwL4bQCHZ+L8zvGJWyKO4zhOxbgl4jiO41SMi4jjOI5TMS4ijuM4TsW4iDiO4zgV4yLiOI7jVIyLiOM4jlMx/w9fhnASAnVFOwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a207d8d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(0,2000,10):\n",
    "    plt.plot(fit.data['precip'],samples['fdy2'][i,:],alpha=0.1, c='b')\n",
    "plt.xlabel('Precipitation [mm]')\n",
    "plt.ylabel(r'$dy$ [tonnes ha$^{-1}$]')\n",
    "plt.savefig('./figs/2d_Gauss_precip_post_sample_growth_curve.pdf')\n",
    "plt.savefig('./figs/2d_Gauss_precip_post_sample_growth_curve.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compute a sample of region averaged yield anomalies for each posterior sample\n",
    "\n",
    "We have a sample from the posterior on the Gaussian parameters $\\boldsymbol \\theta_i$. For each sample, $i$, we want to compute the mean yield anomaly for all the regions combined at a given temperature increment $\\Delta T_j$ applied to every tru temperature. After this procedure we will have a sample of mean yield anomalies $Y_i_j$ for each temperature increase."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.630036266199637"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def yield_anomaly(temp_6m, precip_6m, mu_t, sigma_t, mu_p, sigma_p, norm):\n",
    "    \"\"\"Take six months of T and P and return yield for given params.\n",
    "    \n",
    "    This should be identical to the function in the STAN model\n",
    "    \"\"\"\n",
    "    if len(norm) == 1:\n",
    "        norm = norm * np.ones(6)\n",
    "    dy = np.zeros(6)\n",
    "    for month in np.arange(6):\n",
    "        dy[month] = norm[month]*np.exp(-0.5 *( np.square((temp_6m[month]    - mu_t)/sigma_t) \n",
    "                                             + np.square((precip_6m[month] - mu_p)/sigma_p) )\n",
    "                                      )\n",
    "    return np.sum(dy)\n",
    "\n",
    "\n",
    "def compute_mean_yield_anomaly(T_inc, data, mu_t, sigma_t, mu_p, sigma_p, norm):\n",
    "    \"\"\" Compute mean yield anomaly for a model over regions and years\n",
    "    \n",
    "    The function yield_anomaly returens the yield anomaly for a year \n",
    "    and and region. Here we loop over the regions and years to create\n",
    "    and overall mean.\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    yield_anomalies = np.full((6, 35), np.nan)\n",
    "    #loop over states\n",
    "    for state in np.arange(6):\n",
    "        #loop over years\n",
    "        for year in np.arange(35):\n",
    "            temp_6m = data['d_temp'][state, year, :] + T_inc\n",
    "            precip_6m = data['d_precip'][state, year, :]\n",
    "            yield_anomalies[state, year] = yield_anomaly(temp_6m, precip_6m, mu_t, sigma_t, mu_p, sigma_p, norm)\n",
    "    \n",
    "    return np.nanmean(yield_anomalies)\n",
    "\n",
    "\n",
    "compute_mean_yield_anomaly(1., data2, 19.71, 5.23, 114.69, 62.41, [1.56])    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "16.7 ms ± 1.92 ms per loop (mean ± std. dev. of 7 runs, 100 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "compute_mean_yield_anomaly(1., data2, 20., 5., 120., 65., [1.5]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "46.666666666666664"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(200000* 14.e-3)/(60)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are six regions each with 35 years with six months. We want the total average for each parameter set and temperature increase."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "T_incs = np.linspace(-5, 5, 50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2000, 50)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.zeros((len(samples['mu_t']), len(T_incs))).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_yield_samples =  np.zeros((len(samples['mu_t']), len(T_incs)))\n",
    "\n",
    "#Loop over the samples and for each parameter realisation\n",
    "#TODO: Vectorise to speed up\n",
    "\n",
    "for n in np.arange(len(samples['mu_t'])):\n",
    "    \n",
    "    mu_t    = samples['mu_t'][n]\n",
    "    sigma_t = samples['sigma_t'][n]\n",
    "    mu_p    = samples['mu_p'][n]\n",
    "    sigma_p = samples['sigma_p'][n]\n",
    "    norm    = [samples['norm'][n]]\n",
    "    for m, T_inc in enumerate(T_incs):\n",
    "        mean_yield_samples[n, m] = compute_mean_yield_anomaly(T_inc, data2, mu_t, sigma_t, mu_p, sigma_p, norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2000"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(mean_yield_samples.T[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_means   = [np.mean(s) for s in mean_yield_samples.T]\n",
    "sample_medians = [np.median(s) for s in mean_yield_samples.T]\n",
    "sample_25      = [np.percentile(s, 25.) for s in mean_yield_samples.T]\n",
    "sample_75      = [np.percentile(s, 75.)  for s in mean_yield_samples.T]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAELCAYAAAAoUKpTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xd8m+W1wPHf0ba8d2I7ibN3CMGElBkIM0BaCi1woe0tt1DoYJe2lwsFSqEUCr3QFgodwIVCyw6UFUqBEgppEpyQveM4w/FesmSN5/4hx9iJLVmJJdnO+X4++tiW3vfVo8TW0bPOEWMMSimlVCSWZDdAKaXUwKfBQimlVFQaLJRSSkWlwUIppVRUGiyUUkpFpcFCKaVUVBoslFJKRaXBQimlVFQaLJRSSkVlS3YD+kteXp4pLS1NdjOUUmpQWbZsWY0xJj/acUMmWJSWlrJ06dJkN0MppQYVEdnel+N0GEoppVRUGiyUUkpFpcFCKaVUVENmzkKpoczv91NZWYnX6012U9Qg5XK5KCkpwW63H9T5GiyUGgQqKytJT0+ntLQUEUl2c9QgY4yhtraWyspKRo8efVDX0GEopQYBr9dLbm6uBgp1UESE3NzcQ+qZarBQapDQQKEOxaH+/miwUEopFZXOWaiDFgiGCIQM7cEQoZAhGDIEjSEUouNruL67xSJYRbBYwCqC1SKd91ktgkUEW8d9SqmBSYOF6pUxBl8g1HEL4vOH8AdD+IMGfzCEMf37fCJgtQh2qwWH1YLDZsFuFRy28PdOm7V/n1Ap1Wc6DKUACIUMrb4A1c0+Kmo9bKxqZvWuJjZWtVBR66Gq0UeDx0+rL0h7oP8DBYAxEAga2tqDNLb5qW72savBy7YaDxv2tLB6VyNbqlvY0+ilsc1PeyDU/41Qvdq2bRuTJk3iW9/6FtOmTeOSSy7hnXfe4bjjjmP8+PEsWbKE1tZWLrvsMo4++miOPPJIXnnllc5zTzjhBGbNmsWsWbP46KOPAHjvvfeYO3cuF1xwAZMmTeKSSy7BdPxy/ehHP2LKlCnMmDGDG2+8MWmvW4Vpz+Iw5Q+GaPUFaG0P0tYewOuPTwDoT6EQtPqCtPqCnffZrEKa0xa+uWzYrUP/88/tr65mza6mfr3mlKIMfnLu1KjHbdq0ieeee45HH32Uo48+mj//+c98+OGHLFy4kLvuuospU6Zwyimn8Mc//pGGhgZmz57NqaeeSkFBAYsWLcLlcrFx40Yuvvjizlxun376KatXr6aoqIjjjjuOxYsXM2XKFF566SXWrVuHiNDQ0NCvr1fFToPFYcQXCNLUFqDJ68fT5Q13MAsEDQ0ePw0ePwAuu4V0l500l41Uh1VXEPWz0aNHM336dACmTp3KvHnzEBGmT5/Otm3bqKysZOHChdx3331AeMlvRUUFRUVFfO9736O8vByr1cqGDRs6rzl79mxKSkoAmDlzJtu2bWPOnDm4XC6+9a1vcfbZZ3POOeck/sWqbjRYDHFt7UGavH6a2vx4/UN/2MbrD+H1+6hu9mGxQLrTTkaKjXSXHesQmUDvSw8gXpxOZ+f3Foul82eLxUIgEMBqtfLCCy8wceLEbufddtttFBYWsmLFCkKhEC6Xq8drWq1WAoEANpuNJUuW8Pe//51nn32WX//617z77rtxfnUqEg0WQ9C+Mf/DfVw/FKLz30GkDbfDSrorHDx0sjw+zjjjDB566CEeeughRIRPP/2UI488ksbGRkpKSrBYLDzxxBMEg5F7ti0tLXg8HubPn8+cOXMYN25cgl6B6o0GiyHC6w/S4NEA0RtjPp/v2NMIKQ4rWW47mSn2w2KeI1FuueUWrr32WmbMmIExhtLSUl577TW+853vcP755/Pcc89x8sknk5qaGvE6zc3NfPGLX8Tr9WKM4YEHHkjQK1C9ETPQZzX7qKyszByOxY+avH5qW9pp8QaS3ZRBSQTcDitZbgeZKQN3qGrt2rVMnjw52c1Qg1xPv0cisswYUxbtXO1ZDEKhkKHe005tazu+QTIPEQiGqG7xsbvRy55GLzUtPpq9AZp9AVq8/o6vAVrbA4RCYDCEDGAg1PGBxmW34nZYSXXawl8dNtxOKzluBwXpTvLTXRSkOynIcOJ29O1X+/MeRxu7GtrITLGTk+og1al/Gkp1pX8Rg4g/GKKmxUddazuhARojgiFDZb2HzdUtbNzbQmV9G7sb26hu9hHq0om1WoT0juWu6U4buakOSnNScTutWEUQEUTAIiCEP+17/UFa2wN42oO0+gLUtPhoqQ1Q72nHH+zeQ051WinKTGFMXiqj89MYnZdKaa47YhAxhs6VVU67hWy3g2y3HZsOUymlwWIwaA+EP5XXt7YPuL0Q9a3trKhsYENVM5uqW9lS3YKvY87EabMwMsfNxMJ0TppQwPAMF8MyXQzPdJGd6sDST8taQ8bQ6PGzt9nH3mYve5t9VDV52Vnfxkeba3lrTVXnscMzXYzNT2NaUQbTijMZmePucXmtzx9iT6OXqiYvmSl28tKcpDh0UlwdvjRYDGC+QJDq5vDO6YESJNrag6ze1Uj5jgbKdzSwvc4DgMNmYWxeKqdPKWRcQRrjCtIpzkpJyByARYTsVAfZqQ4mDkvv9pgxhpqWdrbWtLK1poWtNa2s29PMh5tqAMhw2ZhWnMn0jtv+waNrb8PttJKX5iQz5eCKxyg1mGmwGIB8gSB7m3w0tg2MIFHT4uOjzTV8tLmWdXuaCYYMdqswtSiTuRMLOKIkkzH5aQNyclhEyE93kp/uZPboHCAcQKqafayqbOSzXY18trORjzbXAlCQ7mTOmFzmjM5hSlFmt9fk8QWp8Hlw2CzkpTnIdjs0+aE6bGiwGEACwRB7m8NzEskOEnubvXy0qZbFm2tYt6cZgNJcN1+aWcyRI7KYNDx90O5VEBGGZbgYNsXFqVMKAahq8lK+o4FPttbyxqrdLFyxi3SnjaNLc5gzJodZo7I7X297IMSuBi97mrzkpjrJTXPo8ls15GmwGABCIUNNi4/qFl9SJ65bfQHe21DNu+uq2FDVAsCYvFS+NmcUx43Nozg7JXmNi7PCDBdnTB3GGVOH0dYe5NMd9XyypY4l2+p4d/1eUh1Wjh+XxymTC5k8LB0RIRSC6mYfNS0+MlPs5Kc7cdkTE0A/q2zs1+tNL8mM+Pj69eu58MILO3/esmULd9xxB9deey233XYbjz32GPn5+QDcddddzJ8/n/Lycnbt2sX8+fOB8C7utLS0g04KWF5ezlVXXUVTUxNWq5Wbb765s03/+Z//yfvvv09mZvh1PP7448ycOZP33nsPh8PBscce23ncOeecwwUXXHBQbUi0l19+mQkTJjBlyhQAbr31Vk488UROPfVU5s6dy3333UdZWdRVr/1Cg0USGWOo9/ipavISCCanK2GMYd2eZt5cvYcPN9XQHghRmuvm63NGcdy4PIqyhm6A6E2Kw8qxY/M4dmwewZDhs52N/GPdXt7fWM1ba6ooynRxyqQCTp5YQEGGq9u8RprLRn66k7QhtvR24sSJlJeXAxAMBikuLua8887rfPy66647IAiUl5ezdOnSzmBxqNxuN08++STjx49n165dHHXUUZxxxhlkZWUBcO+99x4QBN577z3S0tI6g0Ui7EtX0h9efvllzjnnnM5gcccdd/TLdQ/G0PqNHkRafQF2N7bR1p6crkSLN8Df11Xx1poqdtR5SLFbOXlCPmdMHca4grQBlYBPJJwg0GaxdNa7CH8NF0/aN68ghJfb7jsHwhPUxoT3bYS/QjBowvU5Oup0RMq4a7UIM0dkMXNEFle2j+WjzTW8u24vT31SwVOfVDBrZBbnzihi1qhsLCK0eMP7RVIcFvLTXGS6h95k+N///nfGjh3LqFGjej2mvb2dW2+9lba2Nj788EN+/OMfA7BmzRrmzp1LRUUF1157LVdffXWfn3fChAmd3xcVFVFQUEB1dXVnsNjftm3beOSRR7BarTz11FM89NBDAHzwwQfcf//97Nmzh1/84hcHBJht27Zx5plncswxx/Dpp58yYcIEnnzySdxuN8uWLeP666+npaWFvLw8Hn/8cYYPH87cuXM59thjWbx4MQsWLODSSy/lyiuvZMuWLQA8/PDDHHvssTz11FM8+OCDtLe3c8wxx/Db3/4Wq9VKWloa11xzDa+99hopKSm88sorbN68mYULF/L+++9z55138sILL/DTn/60x57R22+/zU9+8hN8Ph9jx47lT3/6E2lpaX3+t+2LhA+0isg1IrJKRFaLyLU9PD5XRBpFpLzjdmui2xhPgWCIHXUetlS3JiVQ7G328tg/t/DNJ5bw+w+3kmK38L2Tx/H4N4/me6eMZ3xhetIChcUCaS4buWkOirJcjM5PZeKwdKYVZzKuIJ3SvFRG5LgZlukiP91JlttBusuO22HD7bCR4rDisodvTlv45rJbSXFYcTtspHakMs902ynIcDEix824gvD1Jw5LZ1Sem8IMJ+kuW4+T9SkOK/MmF/Kz86bz+6+X8R+zR7Kt1sPtr63hqqeW8eqKXXjawzvp29pDVNR52FDV3LHkeQCsVOgnzz77LBdffHG3+379618zY8YMLrvsMurr63E4HNxxxx1ceOGFlJeXdw4XrVu3jrfeeoslS5Zw++234/f7D6oNS5Ysob29nbFjx3bed/PNNzNjxgyuu+46fD4fpaWlXHnllVx33XWUl5dzwgknALB7924+/PBDXnvtNX70ox/1eP3169dzxRVXsHLlSjIyMvjtb3+L3+/n+9//Ps8//zzLli3jsssu4+abb+48p6Ghgffff58bbriBq6++mpNOOokVK1awfPlypk6dytq1a/nLX/7C4sWLO7PvPv300wC0trYyZ84cVqxYwYknnshjjz3Gsccey4IFC7j33nspLy/v9lq7qqmp4c477+Sdd95h+fLllJWVcf/99x/Uv2skCe1ZiMg04HJgNtAOvCkifzPGbNzv0H8aY4ZUTmJjDLWt7VQ1eZMyL7G1ppUXP63kgw3VAJw4IZ8vzSxmbH7/fvqIhdUipDrDO7JTHTZcdkvSAtW+anwZrs97Au2BEG3tQTz+AM3eQLfd8oUZLi6ePZKvHFXC4s21vLpiF4/+cwv/9/F2TptSyLkzihiW6cLnD1FZ30ZVs5e8NCc5g3wFVXt7OwsXLuTuu+/uvO+qq67illtuQUS45ZZbuOGGG/jjH//Y4/lnn302TqcTp9NJQUEBVVVVnenJ+2r37t187Wtf44knnsBiCX/evfvuuxk2bBjt7e1cccUV3HPPPdx6a8+fM7/0pS9hsViYMmUKVVVVPR4zYsQIjjvuOAAuvfRSHnzwQc4880xWrVrFaaedBoSH44YPH955Ttc5nXfffZcnn3wSCGfSzczM5P/+7/9YtmwZRx99NABtbW0UFBQA4HA4OtOwH3XUUSxatKjP/x4ff/wxa9as6Wxve3s7X/jCF/p8fl8lehhqMvCxMcYDICLvA+cBv0hwOxLK0x5gV0Nyhpw+29nI88t2sLyiAZfdwrkzilgws4iCdFf0k/uZCKQ6baS7wp/wEzUZfLD2BZBM7AzPDO8gD6d7D9DWHs6aarNaOGlCPidNyGdDVTMLV+zib5/t5rWVu5g7oYCvlJVQku3GHzDsbvBS3ewjP91JbqpjQA319dUbb7zBrFmzKCws7Lyv6/eXX355xNoTPaUj7+qll17i9ttvB+D3v//9AZO3TU1NnH322dx5553MmTOn8/59b9pOp5NvfvObnfU0orWhtx7f/v83IoIxhqlTp/Kvf/2rx3OiJUc0xvCNb3yjW6Ddx263dz5nT/8u0a572mmn8cwzz/T5nIOR6GGoVcCJIpIrIm5gPjCih+O+ICIrROQNEUle8v5DFAoZdje2JWXIaUNVM7e8sor/fukzttS08vU5o/jTN2bzrRPGJDRQOGwWctMcjMpzM2V4BqPzUslLS9yqof7kslspSHcxriCNScPTKcpyddvVPaEwnRtPn8gfvl7GgiOK+HBzDd/983LufWs9FR2bFwPBcNBYX9XcsUR6cA1PPfPMMwcMQe3evbvz+5deeolp06YBkJ6eTnNzc0zXP++88ygvL6e8vPyAQNHe3s55553H17/+db7yla/02AZjDC+//PIhtQGgoqKiMyg888wzHH/88UycOJHq6urO+/1+P6tXr+7x/Hnz5vHwww8D4R5IU1MT8+bN4/nnn2fv3r0A1NXVsX379ojt6Ev758yZw+LFi9m0aRMAHo+nW3Gp/pLQnoUxZq2I3AMsAlqAFcD+IXQ5MMoY0yIi84GXgfE9XU9ErgCuABg5cmTc2n0wWn0Bdja0JTzR3/baVp76ZDsfb6kjw2Xjv44fzVnThiV0T4TdJmSlOMhy2wdlUOgLu9VCbpqT3DQnnvYAtS3tnZsoc9Oc/NfxYzh/Vgkvl+/ib5/t4p8bqzl2XB4Xlo1gdF4q/oBhZ304Z1ZBupPsVEdMzx9tqWs8eDweFi1axO9+97tu9990002Ul5cjIpSWlnY+fvLJJ/Pzn/+cmTNndk5wH4q//vWvfPDBB9TW1vL4448Dny+RveSSS6iursYYw8yZM3nkkUcAOPfcc7ngggt45ZVXOie4+2Ly5Mk88cQTfPvb32b8+PFcddVVOBwOnn/+ea6++moaGxsJBAJce+21TJ164OfZ//3f/+WKK67gD3/4A1arlYcffpgvfOEL3HnnnZx++umEQiHsdju/+c1vIi4UuOiii7j88st58MEHef7553s8Jj8/n8cff5yLL74Yn88HwJ133tltQUB/SGqKchG5C6g0xvw2wjHbgDJjTE2kaw2UFOWhkGFPk5falvaEPu+eRi9PL9nO++urSXFYOe/IYhYcUdTn7KuHymKBzBQ7WW7HkFs22leBYIi61nA24K5LoRvb/CxcsYtXV+zC6w8yd2I+l84Z1a2H57RbKMxw9ZpKRFOUJ862bds455xzWLVqVbKb0u8GVYpyESkwxuwVkZHAl4Ev7Pf4MKDKGGNEZDbhobLaRLfzYLT6AlTWtyW0+JCnPcBfl+7glfJdWCzCl2cV8+UjS8hIUP6iVKeV3FQnGSm2QTkG359sVgsFGeGVWo1t4cSGPn+IzBQ7X5szivNmFvPC8koWrtjFh5tqWHBEERccNYI0pw2fP0RFrQe308qwDJemSFcDTjJ+I18QkVzAD3zXGFMvIlcCGGMeAS4ArhKRANAGXGQGwcDu3mYve5t8CUvTETKG99ZX8/hHW6n3+Dl1cgGXHjOK3DRn9JMPkdUiZKfayXY7huww06EQEbLcDrLcjo5suF68/hBpLhvfOLaUs6YP4+mPK3hx+U7eXlPFRUeP4Kxpw7FbLXh8QbZUt5LusjEs06X/vklQWlo6JHsVh0or5R2iYMiwo85DcwIr1W2saubRf25h3Z5mJhSmccUJYw/IthoPKQ4LuanhrKuDeflnMjS2+alu9nZb6LCluoU/fbSN8h0NDMtwcfkJo5k9OrfzcRHIctspzHCxacN6Jk2adNj33tTBM8awbt26gx6G0mBxCNrag1TUeRI27NTU5ueJf21j0ZoqMlPsfOPYUk6ZVNBvdSF6k+IID6903YOgDk5jm589jd5uvzPLt9fzh8VbqajzMLs0h8tPHMOwjM/nMywWsLXWkJOVQV5engYMFTNjDLW1tTQ3NzN69Ohuj2mwiLO61nZ2NbQlZNjJGMPizbX87v3NNPsCnDujiIuOHhH3cW2XPRwktH5D/zLGUN3i6zZsGQiGWLhiF8/8u4JQCL5aVsKXZ5V0ZrM1oQDiqcMS8sf9w4EamlwuFyUlJdjt3f+eNVjESShk2NnQRoPn4NIUxKq2xccjH2zm4y11jMtP4+p54xmdF3nzz6Fy2i0Upg/NvEYDiT8YrsbX9XeppsXH7z/cyuJNNRRluvj2SWOZNTK723kZKTaGZ6bgsGladHXoNFjEgdcfZEedB28C9k4YY1i0too/frgVf9BwyTEj+eLM4rgWGLJZhcIMF9luuw51JFB4h7+3c1c4wPKKen73/mZ2NXo5aUI+V5wwptsKNxEoyHCSn+bU/yt1SDRY9LNGj5/KBk9C8jpVNXl58N2NrKxsZFpRBt8/ZXxcU4VbLJCf5iQvzakT10lU3VE7fN+fpD8Y4vlllfx16Q7SnDaumjuWY8fmdTsnxWGhJNutq6bUQdNg0U+MMexuTMwmO2MM722o5pH3N2MMXHbcaE6fWhi3MWoRyE51UJjuxKaV3gYErz9IZX1bt17G1ppW/vfvG9hc3coJ4/P49olju80jiYTLweanay9DxU6DRT/wB8Nppj2+YPSDD1GLL8DD723ig401TBmewfWnTaAwI345nNJcNobrOv4BqbcJ8Bc+3cmzSypIddq48qSxHD+uey/DZQ/3Mrrmq1IqGg0Wh6jVF2B7rYdgKP7/Pqt2NnL/OxuobfHxH8eM4oJZJXGbm7BZhaLMFJ28HgTCvQxPt70Z22tb+dXfN7JpbwvHjcvje3PHkeb6fFWcCOSlOSnM0F6G6hsNFoegvrWdnQlYFusPhnhmSQXPL6tkWKaLG0+fyITC+GyuE4GcVAeFGa64TpKr/mWMoarJR3Wzr/O+YMjw4vJKnl5SQbbbwQ2nTWBacffEgtrLUH2lweIg7Wn0dvvDjJe9zV5+8eZ61lc1c9qUQi4/fkzc/rBTHFaKs1L0jWMQa/b6qaxv65agcENVM798ez27G71ccFQJ/zF7ZLe5J+1lqL7QYBGjUMiwo95DU1v803Ys317PfYvWEwgarp43/oCx5/4iEq7olp8e/3xRKv78wXDVvZYuqWXa2oM89s8tLFpbxYTCNG48fSLDM7uvnNNehopEg0UM/MEQ22vjX6AoZAx/+fcOnllSwcgcNz8+azLF2fFZEqtLKoeunpJWLt5Uw0P/2EgoBFeeNIZTJhV2O0cE8tOdFOiKKbWfAZuifCDaXN2CPxDfoNnY5uf+RetZXtHAyRPz+c7ccXF5I9c3haGvIN1FmtNGRZ2n8/f2uHF5TChM5/5F63ngnY2s3tXEFSeO6Sx6ZQzsbfLR4gswItutu79VzPQ3BrqNA8fDhqpmrv1LOSsrG/nu3HFcd+qEuAQKh83CmPxUCjNcGiiGOLfDxrj8NFKdn/8e5ac7ufNL0/lq2QjeXlPFTS+sZE+jt9t5Hl+QTXtbaGxLTLoaNXRosIizd9dV8cMXVmIRuPeCIzhz2rC4vJFnp9oZX5CWsMp4KvlsVku4pnn65yVZrRbha3NGces5U9jb5OPav3zKJ1u71w4LhgwVtZ6OFX9DYxhaxZ8GizgJhgyPf7SVB97ZyJSiDB746kzGFaT1+/OIQHF2CiXZbk3VcRgSEYZnpjAiJ4Wun0GOLs3hVxfOZHhmCnf+bS2Pf7T1gD1DdS3tbNrbgtcf/02navDTYBEHnvYAP3t9DS8s38lZ04Zx+7lT41Lm1G4TxuSnkpPqiH6wGtKy3A7GFaR1m4sozHBxz/kzOGvaMF5YvpP/efkzGjzd09Z4/aHwsFSCsiirwUuDRT/b0+jlB8+vZNn2eq48aSzfmTsuLnmXUp1WxuXrsJP6nMtuZVxBWrcd3Q6bhe90zJNtqGrh+udWsKW6pdt5xkBFnYfdjTospXqnwaIffbazkeufK6eutZ07Fkzj7OnD4/I8eekORuelavI/dQCrRSjNdZOT1r23ecqkAu45fwahkOGmF1ayeFPNAefWNLezpaYVfzAxlR/V4KLvNv3kH+v3cusrq8hw2fnlV47giBFZ/f4cIjAyx83wzBRd7aR6JSIUZ6UwPMvVbR5jXEEaD3x1JqW5qfz8zXU8/cl2Qvv1JPatlmr1Ja6mvBocNFgcImMMzy3bwf2LNjB5eAb3feWIuNSesFnD8xOaAFD1VV6ak1G5bixd/sqzUx3c/eXpzJtUwLP/3sHP31jXLR06hJeSb61pTUjaGzV4aLA4BMGQ4ZEPtvDkv7Zz4vg8bl8wlbQ41MVOcVgYq/MT6iCku+yMze8+8W23Wrhm3nj+6/jRfLK1lpteWMHepu77MYwJz7/tqPMQSkDmZTXwabA4SL5AkJ+/uZbXP9vNl48s5obTJ2KPwxxCRoqNMXlpuuNWHTSX3crY/NRuuaFEhC/NLOYn50ylutnHjc+vYNPelgPObfD4dR5DARosDkpTm5//eXkVn2yp4/ITxvDN40bHpZpdfrqTUbmpun9CHTKb1cKYvFTSXd17p7NGZXPP+TOwWS386MWVLNlad8C5be3heQxPu85jHM40WMRob5OXm15YyebqFn545iQWHFHU788hAiXZKQzLjF+lPHX4sViEUblusvab9xqVm8ovLziCEdlufvb6Gv722e4Dzg0EDVuqWw/Yp6EOHxosYlBZ7+GHL66koa2dn35xGsfFIbW4xQKlealk60Y7FQciwogc9wFp6/dNfJeNyuGR9zfzx8VbD1gpZQzsqGtjd2NbIpusBog+BQsRyenDrc9rRUXkGhFZJSKrReTaHh4XEXlQRDaJyEoRmRXLi4qHLdUt/OjFzwgEDXefN52pRZnRT4qR3SaMzU+LyyS5Ul0Ny3QxPKt7z9Vlt/Lf8ydzzvThvPTpTn7x5jp8gQNTgdQ0t7OtplUnvg8zfX1X2tVxizR4bgVGRruQiEwDLgdmA+3AmyLyN2PMxi6HnQWM77gdAzzc8TUp1u5u4vZXV5PisHHnF6fFpQZFisPCqNzUuEySK9WTvDQnNotQWf95CWGrRbjixDEUZrr444dbqfes5pZzphzwAabZG2BzdQujclN18cVhoq//y2uNMWOMMaN7uwG1Ua8SNhn42BjjMcYEgPeB8/Y75ovAkybsYyBLROKzHTqK8h0N3PLKKjJT7Nxz/vS4BIo0l43ReWkaKFTCZbkdjMp1d9u8t2+l1E1nTmJDVTM/fnElda0HzlV4/SE2V+vE9+Gir+9OX+inYwBWASeKSK6IuIH5wIj9jikGdnT5ubLjvoT615Zabn91NcMzXfz8/BkUpPf/hHN2qp3SXDdWXfGkkiTdZac0L5X9F/QdPy6Pn5w7lT1NXm56YQW7Gg6cq9CJ78NHn4KFMaZzx44FJNb1AAAgAElEQVSI/DDaMVGutRa4B1gEvAmsAPb/aNLTO+cBA6QicoWILBWRpdXV1X15+j57b/1efv7GWsbmp3HXedPJdvf/hHN+upOSbLem7lBJl+a0MTovtdtub4CZI7L42Zem42kP8sMXVx6QhBA+n/iuaurTW4AapKIGCxH5a5fbc8C3DvVJjTF/MMbMMsacCNQBG/c7pJLuvY0SwnMm+1/nUWNMmTGmLD8//1Cb1em99Xt54J0NTBmewR1fnEq6q/9TbBRmOHVprBpQUp3hDaD793InFKaH92JYLPz4pc/4bGdjj+fvbfKxo86jmWuHqL70LJqMMV/tuH0FeOdQn1RECjq+jgS+DDyz3yELga93rIqaAzQaYw5c/B0H/+gIFFOLMvnJuVPjkmJjWKaLggwNFGrgSXFYGZOfekDAGJHt5t4LZpCb6uAnC1fx8ZaepygbPH621WqKkKGoL8HiZ/v9fHM/PO8LIrIGeBX4rjGmXkSuFJErOx5/HdgCbAIeA77TD88Z1T/W7+VX72xgWlEmt54zJS51souyXAescVdqIHHZwwHDZu0eMPLSnPz8yzMYk5fG3W+s5YMNPQ/9tngDbKlpJaApQoYU6WuXUUTyjDEHJsEfIMrKyszSpUsP6txVOxt5d91eHli0genFmdwSh0AhAsVZKbrZTg0avkCQLdWtBIL7pTFvD3DHa2tYs6uJq08Zz6lTCns832m3UKpLawc8EVlmjCmLdlws/4t/PIT2DGiJCBQjst0aKNSg4rRZO4psde9huB02bjt3KkeMyOJ/393YY3oQAJ8/xJYarfE9VMQSLIbkkp0Xl1eGA0VJfAOF1qFQg5HL3nPAcNmt3HL2FGaXhtODvPzpzh7P9wcMm6u1mNJQEEuwGHIzVsYY3lq9Jxwozo7PHEVJdooGCjWo7QsY+096O2wWfnTWJI4bl8cfFm/lL/+u6PH8UAi21rTSogFjUItlqc+Q61mICA9dPIuVlQ04bf0fKIqzU8iKw/4MpRJt36T3lupWgl1WOtmtFn5w+kQcVuGpTyrwBUJ8bc6oA/YOGQPbaloZmesmIw5L0VX8xRIsfhy3ViSRw2bBZbfS30vDi7Jc5OgchRpCegsYVotw7akTcFgtPLesEqDXgFFR69Fh2UGqz8HCGLMqng0ZSoZnuchN0+WxaujZFzA2V7cQ6rIy1iLCd04eByI8t6wSEeHSY0b2HDDqPBSbFP0wNcjEvONMRLIJZ4Pt3FVmjPmgPxs1mBVmOsnTQKGGsH1zGFuqW7v1yC0ifGfuWIwx/HXpDiwClxwzqsdr7KxvI2SM/q0MIjEFCxH5FnAN4fQb5cAc4F/AKf3ftMGnMMMZl2SDSg00boeN0rxUttUcGDC+e/I4jIFn/70DiwgXz+65csHuBi8hY/RvZpCIdbfMNcDRwHZjzMnAkUD/ZvAbpAoznJrCQx1W0pw2Ru6X3hzCAeN7p4xj3qQC/rykgmeW9LxKCqCq0cfeZk1AOBjEOgzlNcZ4RQQRcRpj1onIxLi0bBAp0EChDlMZLjsl2SnsqOuevtwiwvdPGY8B/rykAovAhUf33MOoavQhiKbBGeBiDRaVHeVTXwYWiUg9PWSDPZwUZDgp1EChDmNZbgchE56H6MpqEa4+ZTzGGJ76pAK71cKXZ5X0eI09jeHehQaMgSumYGGM2VfR7jYR+QeQCbzR760aJDRQKBWWk+ogGDKdb/r7WC3CNfMmEAgZ/vTRNpx2K2dP77no5Z5GLyLopPcAFesEtxM4Hyjtcu5M4I7+bdbAl5+ugUKprvLTnYSMYW+Tr9v9Votw/akT8PlDPPL+Zlw2C/Mm95x8cHdDONhowBh4Yp3gfoVwfewA0NrldljJT9fCRUr1pDDDRXbqgRvubFYLPzxzEjNHZPHguxtZvKn3BNa7G7zUtvh6fVwlR6xzFiXGmDPj0pJBIstt10ChVATFWSkEQ4amtu65oBw2CzfPn8ytC1dz39vrcdoslJXm9HiNXQ1eLCKaqXkAibVn8ZGITI9LSwaBdJeNkuyUZDdDqQFNRBiZ4ybVeWC+NZfdyk/OmcKoXDd3v7GOlZUNvV5nZ0MbjW3+eDZVxaBPwUJEPhORlcDxwHIRWS8iK7vcP+SlOKyMzHEfkL5AKXUgEWFUbiopjgPfYlKdNm5fMI1hmS5++rc1rN/T3OM1jIEddR6avRowBoK+9izOAc4FzgLGAad3/Lzv/iEtXPHLjcWigUKpvrJawgGjp0p5mSl2fvrFaWS7Hdz+6moq6jw9XsMY2F7rwdOu6c2TrU/BwhizPdIt3o1MJptVGJXrxmbV0pBKxcputVCa5z6geBKEl9vesWAaNqvwk4Wret3JbUy4HoZW3EsufQeMwGKB0tzUuNS6UOpw4bRZKc1NPSAtCMCwTBe3L5hGW3uQW19Z3escxb4CSr6ABoxk0WDRCxE6xlw1UCh1qFIcVkbmunt8bHReKrecM4XqZh+3vbq61yGnQNCwtaaV9kCox8dVfMUULETkKyKS3vH9/4jIiyIyKz5NS67irBTSnDFncFdK9SLDZWd4Vs/LzqcWZfLDMyexpbqFu99Yhz/Yc0DwBwzbalsJ9PK4ip9Yexa3GGOaReR44AzgCeDh/m9WchVkOHV9t1JxkJfmJDet57+t2aNzuGbeeMp3NPDLRRu6VePryucPsa3WQ6iXx1V8xBos9g0Yng08bIx5BRhS76pZbrum8VAqjoqyUkh39dxrP2VSIZcdV8riTTU8+s8tmF7qHbe1B9le5+n1cdX/Yg0WO0Xkd8CFwOsduaKGzLxHqtOqm+6USoCROe4e92AAnHdkCV8+spjXP9vN88sre71GizdwQGp0FT+xvtF/FXgLOMMY0wDkAD/o91YlgdNuYVRuqm66UyoBLB17MOy2nv/evnFsKSeOz+fJf23n3XV7e71OY5ufnQ0aMBIh1mDRBqQCF3f8bAd6368/SNisQmluKlbddKdUwtitFkpzU7H08C5kEeHaU8czoziTB9/dSPmO3t9m6lraqWrSanvxFmuw+C3hutv7gkUz8JtYn1RErhOR1SKySkSeERHXfo//p4hUi0h5x+1bsT5HLEbl9LzLVCkVXy67lRE5B5ZmhXAw+fH8yZRkpXDX62vZWtPS63X2Nvmo0Uy1cRXrO+QxxpjvAl4AY0w9MU5wi0gxcDVQZoyZBliBi3o49C/GmJkdt9/H2M6Y6F4KpZInw9V7Juc0p43bFkwl1WnltlfXRKzXvbvBS6NH80jFS6zBwi8iVsAAiEg+cDALnm1AiojYADeHeWlWpQ53eWlOcnpZUpuX5uS2c6fi8we57dU1tHh7zxO1o95Di0/zSMVDrMHiQeAloEBEfgZ8CNwVywWMMTuB+4AKYDfQaIx5u4dDz+/IbPu8iIyIsZ1KqUGmKNNFWi9LakflpvLf8yezu6GNn72+ptdNe+HEg620tWtakP4WU7AwxjwN3ATcTfiN/kvGmOdiuYaIZBOutjcaKAJSReTS/Q57FSg1xswA3iG8+a+na10hIktFZGl1dXUszVBKDTD76mC47D2/Lc0oyeKaeeNZtauJh97d2OseC80jFR8xz+oaY9YZY35jjPm1MWbtQTznqcBWY0y1McYPvAgcu99z1Bpj9s1WPQYc1UtbHjXGlBljyvLz8w+iKUqpgWRfWvPeVibOnVjAJceM5B/rq3n23zt6vU4wZNhW4+m1B6JiF1Pyo45NeOcDpV3PNcbcEcNlKoA5IuImvBR3HrB0v+cZbozZ3fHjAuBggpJSahBy2MJpzbdUt9JT5+HCshHsbvDy5yUVDM90MXdiQY/XaQ+E2F7byui8NF0W3w9i7Vm8QngIKQC0drn1mTHmE+B5YDnwWUcbHhWRO0RkQcdhV3csrV1BeOXUf8bYTqXUIOZ22CjO6jmbgojwvVPGMa0og//9+0ZW72rs9Tpt7eGAoWlBDp3E8o8oIqs6lrsOOGVlZWbp0qXRD1RKDRq7G9uoaW7v8bFmr58fPL+SJq+f+y44gqJegguEc76NyOk5RfrhTkSWGWPKoh0Xa8/iIxGZfpBtUkqpmAzL6H2FVLrLzk/OnQLA7a+upqmXwkkADR4/exp1l/ehiDVYHA8sF5H1HctaPxORlfFomFJK7Vsh5exlhdTwzBT+5+wp7G32cdcbayNOaFc3+6jVXd4HLdZgcRYwDjgdOBc4p+OrUkrFhdUSDhg95ZACmDI8g2tPncDqXU08/N7miPMTuxu9NHl1l/fBiLUU3B56WA0FxLIaSimlYuKyWxmZ42Z7rafHFVInTchnR52HvyzdwcgcN186srjH6xgDFbUexuSn4nZoJcxYJHw1lFJKHYx0V+TCZP9xzEiOHZvLnz7aytJtdb0eZwxsq/Hopr0Y6WoopdSgsqPOQ0MvCQO9/iA/fHEluxu83HvBDEblpvZ6Hafdwpi8VGzWwzvjtK6GUkoNScVZKb1W2XPZrdxy9hRcdgs//dsaGiOskPL5Q1RoadY+O5jVUMt0NZRSKlksFmFkTu8pQfLSnNw8fwp1re3cHWWFVKsvSGW9Vtrri4NZDTUeXQ2llEoih83CyNyeiyYBTByWztWnjGf1riYeeT/yCqkGj5+9Wmkvqlizzm4HsggHiHOBrI77lFIqodKctogT3nMnFvDVshG8vaaKV1fu7vU4gKomHw2enneKq7CYgoWIXAM8DRR03J4Ske/Ho2FKKRVNfrqTLLe918cvOWYkx4zO4Q8fbmFFZe91vAEq69to1cJJvYp1GOq/CJdWvdUYcyvhetyX93+zlFKqbyJNeFtEuP60CRRnu7nnzXXsiTDcFC6c5MHr1yW1PYk1WAjQ9V8y2HGfUkolRbQJb7fDxv/Mn0zIGH72tzURg0EwZNhe6yGgdTAOEGuw+BPwiYjcJiK3AR8Df+z3VimlVAwcNgsjcnrPOluUlcJNp0+ios7Dr/7ee5U9CNfB0CW1B4p1gvt+4JtAHVAPfNMY80A8GqaUUrFId9kpyHD2+visUdl84wulLN5Uw3PLKiNeq9UXZJdmqe0m1kp59xhjfki4cNH+9ymlVFIVZrjwtAdp8fY8UX3ekcVsrWnlqY+3U5qbyuzROb1eq66lHZfNQm5a7wHocBLrMNRpPdx3Vn80RCml+sOI7BTstp7nL/ZV2RuTn8p9b69nR70n4rV2N3pp0RVSQB+DhYhcJSKfARM7dm7vu20FdAe3UmrAsFktjMzpfcOe02blv+dPxmGzcNfra/G09x4M9mWp1aSDfe9Z/JnwJryFfL4h71zgKGPMpXFqm1JKHRS3w8awzN437BWku/jhGRPZ1dDGr97ZSCjCZPa+FVLB0OE94d2nYGGMaTTGbDPGXGyM2d7l1nseYKWUSqK8tMgb9qaXZPHN40bzry21vBBlwluTDvZ9GGp5fxyjlFKJVJyV0mtJVoAvHlHEiePz+b+Pt7O8oj7itVq8AaqaDt+yrH1dDTU5SnZZATL7oT1KKdVvLB0lWTftbemxwp6I8P1TxlFR18p9b63n/gtnMixCvqnqZh8pDiuZKb33WIaqvgaLSX04RmeAlFIDjstupTgrpddU5C57eML7ur+Wc/fra7nn/Bm47NZer1dZ78FpS4t4zFDU1zmL7X24RR70U0qpJMlOdUScvxiemcKNp09ka00rv31vU8S5iVAIKuoOvwnvw7ueoFLqsFGclYIrwvxF2agc/uOYkfxjfTV/+yxySnOfP0RllD0aQ40GC6XUYcFiEUZE2H8B8NWyEcwuzeEPH25l3e6miNdraguwt/nwSQmiwUIpddjYN3/RG4sI1502gbw0Jz9/c13EGt4AVY0+mr2Rjxkq+jTBLSLXR3q8I8Fgn4nIdcC3AAN8RjghobfL407gSeAooBa40BizLZbnUEqpnmSnOmjxBWjw9Pwmn+a08eOzJvGD51dy71vruH3BtF7TnwPsqGtjbIEFp21oT3j3tWeR3nErA64CijtuVwJTYnlCESkGrgbKjDHTACtw0X6H/RdQb4wZBzwA3BPLcyilVCTR5i/G5Kdx1UljWVHZyJ+XVES8VjBk2HEYbNjr62qo240xtwN5wCxjzA3GmBsIf/IvOYjntQEpImID3MCu/R7/IvBEx/fPA/NEIo00KqVU3/Vl/uLUKYWcNqWQvy7dwb+3RU5W0dYeYmdDz0tzh4pY5yxGAl2rmrcDpbFcwBizE7gPqAB2A43GmLf3O6wY2NFxfABoBHL3v5aIXCEiS0VkaXV1dSzNUEod5qLNXwB8+8QxjMlP5ZeL1kcsyQpQ3+qnvrU94jGDWazB4v+AJV0q5S0hPLfQZyKSTbjnMBooAlJFZP9khD3F+wP6eMaYR40xZcaYsvz8/FiaoZRSUfdfOG1WfnzmZADufmMt7YHI5VZ3NrQN2RresVbK+xmfV8qrA75hjLkrxuc8FdhqjKk2xviBF4Fj9zumEhgB0DFUldnxfEop1a+i5Y8aluni+lMnsqW6ld99sDnitYxhyGao7etqqGa6f7KXLo8ZY0xGDM9ZAcwRETfQBswDlu53zELgG8C/gAuAd81Qnz1SSiVFtPxRALNH5/CVo0p4blklU4syOGVSYa/Xaw+EN+yNyk2NU4uTo68T3OnGmIwut/Qut1gCBcaYTwhPWi8nvGzWAjwqIneIyIKOw/4A5IrIJuB64EexPIdSSsXCZbcyPEL9C4BLjhnFtKIMfvveZirqIu/ebmoLUN08tDLUylD5wF5WVmaWLt2/g6KUUn1XUeuJuBGvrrWda579lHSXjfu/OjNiMkERGJ2XSqqzr/lak0NElhljyqIdF9OchYRdKiK3dvw8UkRmH2wjlVJqICnOTsFh6/1tMSfVwY2nT6Syvi1qwkFjwgkH/cHIk+KDRayroX4LfAG4uOPnZuA3/doipZRKEmvH/EWk/RdHjMji4tnhhINvr6mKeL1A0AyZCnuxBotjjDHfBbwAxph6wNHvrVJKqSRJcVgpjFAACcIJB2eOyOLRD7awtaYl4rEeXzDqHo3BINZg4RcRKx0ro0QkHxgafSyllOqQn+4k3dX7XIPVItxw2gTSnDZ+/sY6PO2BiNeraW6nsZdcVINFrMHiQeAloFBEfgYsBu7u91YppVSSlWSnYLf1Ph6V5XbwgzMmsqfJy0PvRp6/ANhR7xnUG/Zi3ZT3NHATcBfhfE4LjDF/jUfDlFIqmWxWCyOyI89fTCvO5NI5o/hwUw1vrt4T8Xr7JrxDg3TDXl835X1ojDm+y+a8Q9mUp5RSg0Kq00ZBhpOqxt73TJw/q4RVOxt57J9bmDQsndF5ab0eG66w18bIXHc8mhtXfd2Ud3zH1/RD3ZSnlFKDSUG6i7QI8xcWEa47dQLpTjv3vLmetvbIQ02Nbf5BuWGvT8FCRF4XkdL4NkUppQamEdkp2KyR5y9uPH0Cuxvb+O370ecvqpq8tPoiT4oPNH2ds3gceFtEbhaR3lM0KqXUEGSzWqLWv5heksVFR4/kvfXV/H3d3ojXMyY84R0YRBv2+joM9VfgSCADWCoiN4rI9ftucW2hUkoNAGlOG/npzojHfLVsBNOLM3nk/ej5o/wBw476wVMwKZbVUH6gFXDyeZnVfTellBryCtKdpDh6zwe1b/+Fy27lF2+ui7pUtsUbYO8g2bDX1zmLM4FywiVQZxljfrKv1GpHuVWllBryRIQROSlYIrxz5qY5uf7UCWyv8/D7f26Jes29zT5aBsH8RV97FjcDXzHG/MgYE7lvpZRSQ5jTZqUoM3I51lmjsrlgVglvraninxsjl3w2BnYMgoSDfZ2zOMEYszrejVFKqcEgO9VBZkrktT6XHDOSScPS+fU/NkXNDRUIGnZEmeNItljTfSillCKczjxSOhCb1cKNp09EgPveWh915VOrL0jVAJ6/0GChlFIHwWoRRmRH3oldmOHie6eMZ31VM09/UhH1mnubfDR7B2bCQQ0WSil1kFL7sJz2+HF5nDGlkBeWV1K+oyHqNXfUtQ3I+QsNFkopdQgKMyIvpwX41gljKMlO4f5F62nwtEc8NhgKz18MtIJJGiyUUuoQ7FtOG2l3t8tu5aYzJtHiC/DAOxsJRQkErb4gewdY/igNFkopdYicNivFWZGX05bmpfJfx49heUU9C8t3Rb3mQJu/0GChlFL9oC/LaedPG8acMTk88a9tbKxqjnrNgTR/ocFCKaX6SbTltCLC1aeMJ8tt5763o6czH0jzFxoslFKqn+xbThtp/iLdZef60yayu9HLY31IBzJQ5i80WCilVD/qy3La6cWZfKVsBIvWRk8HAgNj/kKDhVJK9bNo2WkBLj56BBML0/nNPzb1KfNsZX1bUutfJDxYiMhEESnvcmsSkWv3O2auiDR2OebWRLdTKaUOVl+y09qsFm44fQIhA79ctIFgKPK8RCBoqExi/YuEBwtjzHpjzExjzEzgKMADvNTDof/cd5wx5o7EtlIppQ5NX7LTDs9M4cqTxrJmdxPPLdsR9ZrN3gA1LcmZv0j2MNQ8YLMxZnuS26GUUv2uL8tpT56Yz0kT8nlmSQXrdjdFveaeRm/UVVTxkOxgcRHwTC+PfUFEVojIGyIyNZGNUkqp/lKU5cJmjbyc9qqTxpKX5uTet9fTGqUQ0r763aEow1b9LWnBQkQcwALguR4eXg6MMsYcATwEvNzLNa4QkaUisrS6OvqKAqWUSjSb1UJxduThqFSnjR+cPpGaFh8Pv7856jV9/hC7GhM7f5HMnsVZwHJjTNX+DxhjmowxLR3fvw7YRSSvh+MeNcaUGWPK8vPz499ipZQ6CBkuO9mpkYejJg3P4KKjR/L+hmreW7836jXrW/00ehK3nDaZweJiehmCEpFhIuFtLSIym3A7axPYNqWU6ldFmSk4bJHfcr9aNoLJw9J5+P3NfVtO2+ChPZCY5bRJCRYi4gZOA17sct+VInJlx48XAKtEZAXwIHCRGQj73ZVS6iBZLEJJlOEoq0W4/vSJGAP3vxN9OW0oBC1R5jj6S1KChTHGY4zJNcY0drnvEWPMIx3f/9oYM9UYc4QxZo4x5qNktFMppfpTqtNGXroj4jHDMlxcedIYVu9q4oXllQlqWXTJXg2llFKHlWEZLlz2yG+9J08s4ITxefx5SUWfstMmggYLpZRKoPDu7sjJBkWE75w0jmy3g/veXo/Xn/h9FfvTYKGUUgnmslspyIicbDDNZeP6U8ezu9HL7/uQnTbeNFgopVQS5KdFTzY4vSSLL88q4a01VfxrS3IXhGqwUEqpJBAJr46KNBwFcMkxIxmbn8pD726krrU9MY3rgQYLpZRKEpfdyrBMV8Rj7FYLN5w+EV8gxIPvbkxa1TwNFkoplUR5aU5SnZGHo0Zku7ns2FKWba/n9VV7EtSy7jRYKKVUkpVkuyPWvgCYP304s0Zm88fFW6ms9ySmYV1osFBKqSRz2CwMj1L7QkS4Zt54nDYLv1y0IeFV8zRYKKXUAJCT6iDdZYt6zPdOHsemvS08++/oxZL6kwYLpZQaIIqzU7BaIi+POnZsHvMmFfDcsh2s7UOxpP6iwUIppQYIu9VCUVbk1VEAV5w4hrw0J/cv2hC1WFJ/0WChlFIDSJY7eilWt8PG9adNYG+zl1++vT4h7dJgoZRSA0xRlivqcNTUokzOn1XCy+W7WLGjIe5tijybopRSKuH2lWKtqI28RPbi2SM5fnweR4zIinubtGehlFIDUGaKnSx35OEou9XCCeMTU1Jag4VSSg1QRVkp2KxRkkcliAYLpZQaoKwWoThKKdZE0WChlFIDWIYr+nBUImiwUEqpAa4oKwW7LbnDURoslFJqgLNahOKs5A5HabBQSqlBIN1lJzs1ecNRGiyUUmqQGJ6ZvOEoDRZKKTVIJHM4SoOFUkoNIskajtJgoZRSg0wyhqM0WCil1CCTjOGohAcLEZkoIuVdbk0icu1+x4iIPCgim0RkpYjMSnQ7lVJqIEv0cFTCs84aY9YDMwFExArsBF7a77CzgPEdt2OAhzu+KqWU6jA8MwVP++FR/GgesNkYs32/+78IPGnCPgayRGR44punlFIDl9UipLsS07tIdrC4CHimh/uLga7VyCs77utGRK4QkaUisrS6ujpOTVRKKZW0YCEiDmAB8FxPD/dwnzngDmMeNcaUGWPK8vMTk9NdKaUOR8nsWZwFLDfGVPXwWCUwosvPJcCuhLRKKaXUAZIZLC6m5yEogIXA1ztWRc0BGo0xuxPXNKWUUl0lpQa3iLiB04Bvd7nvSgBjzCPA68B8YBPgAb6ZhGYqpZTqkJRgYYzxALn73fdIl+8N8N1Et0sppVTPkr0aSiml1CCgwUIppVRUGiyUUkpFJeHpgcFPRKqB/XeCD3R5QE2yG5Fg+poPD/qaB49RxpioG9WGTLAYjERkqTGmLNntSCR9zYcHfc1Djw5DKaWUikqDhVJKqag0WCTXo8luQBLoaz486GseYnTOQimlVFTas1BKKRWVBosBQkRuFBEjInnJbku8ici9IrKuo2TuSyKSlew2xYuInCki6ztKBP8o2e2JNxEZISL/EJG1IrJaRK5JdpsSQUSsIvKpiLyW7LbEiwaLAUBERhBOrFiR7LYkyCJgmjFmBrAB+HGS2xMXHWWDf0M4Hf8U4GIRmZLcVsVdALjBGDMZmAN89zB4zQDXAGuT3Yh40mAxMDwA3EQPBZ6GImPM28aYfYWDPyZcr2Qomg1sMsZsMca0A88SLhk8ZBljdhtjlnd830z4DfSAKpdDiYiUAGcDv092W+JJg0WSicgCYKcxZkWy25IklwFvJLsRcdKn8sBDlYiUAkcCnyS3JXH3K8If9kLJbkg8JSVF+eFGRN4BhvXw0M3AfwOnJ7ZF8RfpNRtjXuk45mbCwxZPJ7JtCdSn8sBDkYikAS8A1xpjmpLdnngRkXOAvcaYZSIyN9ntiScNFglgjMYi5cYAAAMtSURBVDm1p/tFZDowGlghIhAejlkuIrONMXsS2MR+19tr3kdEvgGcA8wzQ3f99mFZHlhE7IQDxdPGmBeT3Z44Ow5YICLzAReQISJPGWMuTXK7+p3usxhARGQbUGaMGYzJyPpMRM4E7gdOMsZUJ7s98SIiNsIT+POAncC/gf8wxqxOasPiSMKfep4A6owx1ya7PYnU0bO40RhzTrLbEg86Z6GS4ddAOrBIRMpF5JFoJwxGHZP43wPeIjzR+9ehHCg6HAd8DTil4/+2vONTtxrktGehlFIqKu1ZKKWUikqDhVJKqag0WCillIpKg4VSSqmoNFgopZSKSoOFUkqpqDRYKNWFiOR22R+wR0R2dvnZ0cPxH/XT8w7fl95aROZ2TXUtIneKyFsi4hSRZ0VkfH88p1Kx0HQfSnVhjKkFZgKIyG1AizHmvgjHH9tPT3098Nj+d3bkzzoOmG+M8YnIw4ST1l3eT8+rVJ9oz0KpQyAiLR1f54rIeyLyfEdhp6c7Ul8gIkeLyEciskJElohIeg+XOh94c79r3wDMB841xrR13P1P4NSOVCJKJYz+winVf44EphJOFrgYOE5ElgB/AS40xvxbRDKAtq4nichooN4Y4+ty93HAROAoY0zLvjuNMSER2QQcASyL66tRqgvtWSjVf5YYYyqNMSGgHCgl/Ia/2xjzb/j/9u5QpYIoiMP4928Wk0G0KmaLT+ATaDGYfAS7b2E2aResIj6AySYXzaIgiCAmGcNecb1cOAZXLvL96pyFKbvDMOwcqKqX3sVPn5aAyYWKt3Qrzqetr38Eln8zcanFzkL6Pf3O4J3u/QrtOyze6NZb9z0Au8BFkqequuzF5pjoTqSh2VlIw7oBlpNsACSZnzJvGNF1Id9U1QjYBk6SrPdCa8B/316rGWOxkAY0vnt7BzhMcg2cM9FFVNUrcJdkdcrzV8AecJZkJcki8FZV98NnL31xRbk0A5Js0Q2zDxrn9oGXqjr6m8ykjjMLaQZU1WmShR8cfQaOh85HmmRnITWMP+IXU0Kb45/4pH/PYiFJanLALUlqslhIkposFpKkJouFJKnJYiFJavoA3qo643hFRoQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a214a5b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.plot(T_incs, sample_means, label='means')\n",
    "#plt.plot(T_incs, sample_medians, label='medians')\n",
    "#plt.plot(T_incs, sample_25, label='25th percentile')\n",
    "#plt.plot(T_incs, sample_75, label='75th percentile')\n",
    "plt.fill_between(T_incs, sample_25, sample_75, label='75th - 25th percentile', alpha=0.2)\n",
    "plt.legend()\n",
    "plt.xlabel('T_inc (K)')\n",
    "plt.ylabel('Yield [tonnes ha$^{-1}$]')\n",
    "plt.savefig('./figs/temperature_impact.pdf')\n",
    "plt.savefig('./figs/temperature_impact.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  2.,   9.,  16.,  30.,  61.,  96., 158., 171., 235., 254., 231.,\n",
       "        208., 182., 133.,  95.,  62.,  27.,  20.,   7.,   3.]),\n",
       " array([8.48109329, 8.52893474, 8.5767762 , 8.62461765, 8.67245911,\n",
       "        8.72030056, 8.76814202, 8.81598347, 8.86382493, 8.91166638,\n",
       "        8.95950784, 9.00734929, 9.05519075, 9.1030322 , 9.15087366,\n",
       "        9.19871511, 9.24655657, 9.29439802, 9.34223948, 9.39008093,\n",
       "        9.43792239]),\n",
       " <a list of 20 Patch objects>)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a214b2eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(mean_yield_samples.T[33], 20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  2.,   3.,  16.,  31.,  60.,  83., 166., 211., 236., 267., 252.,\n",
       "        223., 176., 119.,  86.,  39.,  19.,   9.,   0.,   2.]),\n",
       " array([ 7.2101904 ,  7.38397627,  7.55776214,  7.73154801,  7.90533387,\n",
       "         8.07911974,  8.25290561,  8.42669148,  8.60047735,  8.77426322,\n",
       "         8.94804908,  9.12183495,  9.29562082,  9.46940669,  9.64319256,\n",
       "         9.81697843,  9.99076429, 10.16455016, 10.33833603, 10.5121219 ,\n",
       "        10.68590777]),\n",
       " <a list of 20 Patch objects>)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a228f7390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(mean_yield_samples.T[0], 20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This clearly shows that the model predicts reductions in yields as a consequence of raising temperatures."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make a two dimensional plot of the function on one posterior\n",
    "\n",
    "We want to see how the function "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Posterior sample\n",
    "#                    mean se_mean     sd   2.5%    25%    50%    75%  97.5%  n_eff   Rhat\n",
    "#mu_t               19.71    0.02   0.68  18.17  19.29  19.77  20.19  20.87   1205    1.0\n",
    "#sigma_t             5.23    0.02   0.58   4.26    4.8   5.17    5.6   6.49    795    1.0\n",
    "#mu_p              114.69    0.11   4.26 106.07 111.81 114.87 117.65 122.71   1571    1.0\n",
    "#sigma_p            62.41    0.07   2.76  57.16  60.51  62.41  64.24   67.9   1633    1.0\n",
    "#norm                1.56  2.3e-3   0.07   1.43   1.51   1.56    1.6    1.7    862    1.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a22a96cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "t1, t2 = 5., 35.\n",
    "p1, p2 = 0., 250.\n",
    "x = arange(t1,t2,.1)\n",
    "y = arange(p1,p2,1.)\n",
    "X,Y = meshgrid(x, y) # grid of point\n",
    "def yield_equal(temp, precip, mu_t, sigma_t, mu_p, sigma_p, norm):\n",
    "    \"\"\"Take six months of T and P and return yield for given params.\n",
    "    \n",
    "    This should be identical to the function in the STAN model\n",
    "    \"\"\"\n",
    "    temp_6m = np.full(6, temp)\n",
    "    precip_6m = np.full(6, precip)\n",
    "    if len(norm) == 1:\n",
    "        norm = norm * np.ones(6)\n",
    "    dy = np.zeros(6)\n",
    "    for month in np.arange(6):\n",
    "        dy[month] = norm[month]*np.exp(-0.5 *( np.square((temp_6m[month]    - mu_t)/sigma_t) \n",
    "                                             + np.square((precip_6m[month] - mu_p)/sigma_p) )\n",
    "                                      )\n",
    "    return np.sum(dy)\n",
    "def z_func(X,Y):\n",
    "    return yield_equal(X, Y, 19.77, 5.17, 114.87, 62.41, [1.56])\n",
    "z_func = np.vectorize(z_func)\n",
    "#Lets use the median\n",
    "\n",
    "Z = z_func(X, Y) # evaluation of the function on the grid\n",
    "\n",
    "im = imshow(np.flip(Z, axis=0),cmap=\"viridis\", extent=[t1,t2,p1,p2], aspect=\"auto\") # drawing the function\n",
    "# adding the Contour lines with labels\n",
    "#cset = contour(Z,arange(-1,1.5,0.2),linewidths=2,cmap=cm.Set2)\n",
    "#clabel(cset,inline=True,fmt='%1.1f',fontsize=10)\n",
    "cbar = fig.colorbar(im) # adding the colobar on the right\n",
    "cbar.set_label('Yield [tonnes ha$^{-1}$]')\n",
    "# latex fashion title\n",
    "#title('$z=(1-x^2+y^3) e^{-(x^2+y^2)/2}$')\n",
    "\n",
    "ax.set_xlabel('T [$^{\\circ}$C]')\n",
    "ax.set_ylabel('Monthly precipitation (mm)')\n",
    "\n",
    "plt.savefig('./figs/2d_yield_response.pdf')\n",
    "plt.savefig('./figs/2d_yield_response.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "t1, t2 = 5., 35.\n",
    "p1, p2 = 0., 250.\n",
    "x = arange(t1,t2,.1)\n",
    "y = arange(p1,p2,1.)\n",
    "X,Y = meshgrid(x, y) # grid of point\n",
    "def yield_equal(temp, precip, mu_t, sigma_t, mu_p, sigma_p, norm):\n",
    "    \"\"\"Take six months of T and P and return yield for given params.\n",
    "    \n",
    "    This should be identical to the function in the STAN model\n",
    "    \"\"\"\n",
    "    temp_6m = np.full(6, temp)\n",
    "    precip_6m = np.full(6, precip)\n",
    "    if len(norm) == 1:\n",
    "        norm = norm * np.ones(6)\n",
    "    dy = np.zeros(6)\n",
    "    for month in np.arange(6):\n",
    "        dy[month] = norm[month]*np.exp(-0.5 *( np.square((temp_6m[month]    - mu_t)/sigma_t) \n",
    "                                             + np.square((precip_6m[month] - mu_p)/sigma_p) )\n",
    "                                      )\n",
    "    return np.sum(dy)\n",
    "def z_func(X,Y):\n",
    "    return yield_equal(X, Y, 19.77, 5.17, 114.87, 62.41, [1.56])\n",
    "z_func = np.vectorize(z_func)\n",
    "#Lets use the median\n",
    "\n",
    "Z = z_func(X, Y) # evaluation of the function on the grid\n",
    "\n",
    "im = imshow(np.flip(Z, axis=0),cmap=\"viridis\", extent=[t1,t2,p1,p2], aspect=\"auto\") # drawing the function\n",
    "# adding the Contour lines with labels\n",
    "#cset = contour(Z,arange(-1,1.5,0.2),linewidths=2,cmap=cm.Set2)\n",
    "#clabel(cset,inline=True,fmt='%1.1f',fontsize=10)\n",
    "cbar = fig.colorbar(im) # adding the colobar on the right\n",
    "cbar.set_label('Yield [tonnes ha$^{-1}$]')\n",
    "# latex fashion title\n",
    "#title('$z=(1-x^2+y^3) e^{-(x^2+y^2)/2}$')\n",
    "\n",
    "CS = ax.contour(X, Y, Z)\n",
    "ax.clabel(CS, inline=1, fontsize=10)\n",
    "\n",
    "\n",
    "ax.set_xlabel('T [$^{\\circ}$C]')\n",
    "ax.set_ylabel('Monthly precipitation (mm)')\n",
    "\n",
    "plt.savefig('./figs/2d_yield_response_contours.pdf')\n",
    "plt.savefig('./figs/2d_yield_response_contours.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Include correlation\n",
    "\n",
    "We originally ran a zero correlated gaussian, now let's incldue correlation\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:pystan:COMPILING THE C++ CODE FOR MODEL anon_model_2df05908660c977a9152ef2888fc2e93 NOW.\n"
     ]
    }
   ],
   "source": [
    "gm3 = pystan.StanModel(file='./stan/2d-gaussian_with_correlation.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "fit2=gm3.sampling(data=data2,chains=4,iter=1000,verbose=True,seed=1308)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0 of 2000 iterations ended with a divergence (0.0%)\n",
      "0 of 2000 iterations saturated the maximum tree depth of 10 (0.0%)\n"
     ]
    }
   ],
   "source": [
    "# carry out some diagnostic checks on fit\n",
    "\n",
    "stan_utility.check_div(fit2)\n",
    "stan_utility.check_energy(fit2)\n",
    "stan_utility.check_treedepth(fit2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "samples2=fit2.extract()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Inference for Stan model: anon_model_2df05908660c977a9152ef2888fc2e93.\n",
       "4 chains, each with iter=1000; warmup=500; thin=1; \n",
       "post-warmup draws per chain=500, total post-warmup draws=2000.\n",
       "\n",
       "                    mean se_mean     sd   2.5%    25%    50%    75%  97.5%  n_eff   Rhat\n",
       "mu_t               18.64    0.02   0.81  16.95  18.12  18.71  19.17  20.15   1303    1.0\n",
       "sigma_t             7.13    0.02   0.58   6.03   6.72    7.1   7.51   8.28   1206    1.0\n",
       "mu_p              115.15    0.09   3.42 108.26 112.88 115.18 117.41  121.6   1372    1.0\n",
       "sigma_p            74.98    0.06   2.58  69.99  73.28  74.98  76.69   80.0   1933    1.0\n",
       "rho                -0.05  1.4e-3   0.06  -0.15  -0.08  -0.05-8.3e-3   0.06   1573    1.0\n",
       "norm                 2.2  1.6e-3   0.06    2.1   2.17    2.2   2.24   2.32   1241    1.0\n",
       "fdy1[0]             0.08  1.7e-3   0.06   0.01   0.04   0.07   0.11   0.23   1213    1.0\n",
       "fdy1[1]             0.11  2.1e-3   0.07   0.02   0.06    0.1   0.15   0.29   1210    1.0\n",
       "fdy1[2]             0.16  2.6e-3   0.09   0.03   0.09   0.14   0.21   0.37   1206    1.0\n",
       "fdy1[3]             0.21  3.1e-3   0.11   0.05   0.13   0.19   0.27   0.47   1202    1.0\n",
       "fdy1[4]             0.28  3.7e-3   0.13   0.07   0.18   0.25   0.35   0.58   1198    1.0\n",
       "fdy1[5]             0.36  4.4e-3   0.15   0.11   0.25   0.34   0.45    0.7   1195    1.0\n",
       "fdy1[6]             0.46  5.0e-3   0.17   0.16   0.33   0.44   0.56   0.84   1191    1.0\n",
       "fdy1[7]             0.57  5.5e-3   0.19   0.23   0.44   0.56    0.7   0.99   1189    1.0\n",
       "fdy1[8]             0.71  6.0e-3   0.21   0.33   0.56    0.7   0.84   1.15   1187    1.0\n",
       "fdy1[9]             0.86  6.3e-3   0.22   0.45   0.71   0.85   1.01   1.32   1187    1.0\n",
       "fdy1[10]            1.03  6.4e-3   0.22    0.6   0.87   1.02   1.17   1.48   1188    1.0\n",
       "fdy1[11]             1.2  6.4e-3   0.22   0.77   1.05    1.2   1.35   1.64   1193    1.0\n",
       "fdy1[12]            1.38  6.1e-3   0.21   0.95   1.24   1.38   1.53    1.8   1203    1.0\n",
       "fdy1[13]            1.56  5.5e-3   0.19   1.15   1.43   1.56   1.69   1.93   1219    1.0\n",
       "fdy1[14]            1.72  4.8e-3   0.17   1.37   1.62   1.72   1.84   2.05   1246    1.0\n",
       "fdy1[15]            1.87  4.0e-3   0.14   1.57   1.78   1.87   1.97   2.14   1290    1.0\n",
       "fdy1[16]            1.99  3.1e-3   0.11   1.77   1.92   1.99   2.07   2.21   1361    1.0\n",
       "fdy1[17]            2.08  2.2e-3   0.08   1.91   2.03   2.08   2.14   2.24   1469    1.0\n",
       "fdy1[18]            2.13  1.6e-3   0.06   2.02   2.09   2.13   2.17   2.26   1553    1.0\n",
       "fdy1[19]            2.14  1.4e-3   0.05   2.04   2.11   2.14   2.18   2.25   1411    1.0\n",
       "fdy1[20]            2.11  1.6e-3   0.06    2.0   2.07   2.11   2.15   2.23   1242    1.0\n",
       "fdy1[21]            2.03  2.0e-3   0.07    1.9   1.99   2.03   2.08   2.17   1224    1.0\n",
       "fdy1[22]            1.92  2.2e-3   0.08   1.77   1.87   1.92   1.97   2.09   1285    1.0\n",
       "fdy1[23]            1.78  2.4e-3   0.09    1.6   1.73   1.78   1.84   1.96   1375    1.0\n",
       "fdy1[24]            1.62  2.5e-3    0.1   1.42   1.56   1.62   1.68   1.82   1466    1.0\n",
       "fdy1[25]            1.44  2.6e-3    0.1   1.24   1.37   1.44   1.51   1.65   1535    1.0\n",
       "fdy1[26]            1.26  2.7e-3   0.11   1.05   1.19   1.26   1.33   1.48   1567    1.0\n",
       "fdy1[27]            1.08  2.8e-3   0.11   0.87    1.0   1.08   1.15    1.3   1562    1.0\n",
       "fdy1[28]            0.91  2.8e-3   0.11    0.7   0.83    0.9   0.98   1.14   1533    1.0\n",
       "fdy1[29]            0.75  2.8e-3   0.11   0.55   0.67   0.74   0.82   0.98   1493    1.0\n",
       "fdy1[30]             0.6  2.7e-3    0.1   0.42   0.53    0.6   0.67   0.83   1451    1.0\n",
       "fdy1[31]            0.48  2.6e-3    0.1   0.31   0.41   0.47   0.54   0.69   1413    1.0\n",
       "fdy1[32]            0.37  2.3e-3   0.09   0.22   0.31   0.37   0.43   0.56   1379    1.0\n",
       "fdy1[33]            0.29  2.1e-3   0.08   0.16   0.23   0.28   0.34   0.46   1350    1.0\n",
       "fdy1[34]            0.21  1.9e-3   0.07   0.11   0.17   0.21   0.26   0.37   1267    1.0\n",
       "fdy1[35]            0.16  1.6e-3   0.06   0.07   0.12   0.15   0.19   0.29   1247    1.0\n",
       "fdy1[36]            0.12  1.3e-3   0.05   0.05   0.08   0.11   0.14   0.22   1231    1.0\n",
       "fdy1[37]            0.08  1.0e-3   0.04   0.03   0.06   0.08    0.1   0.17   1217    1.0\n",
       "fdy1[38]            0.06  8.2e-4   0.03   0.02   0.04   0.05   0.07   0.13   1205    1.0\n",
       "fdy1[39]            0.04  6.4e-4   0.02   0.01   0.02   0.04   0.05    0.1   1196    1.0\n",
       "fdy2[0]             0.66  1.7e-3   0.07   0.53   0.62   0.66   0.71    0.8   1566    1.0\n",
       "fdy2[1]             0.73  1.7e-3   0.07    0.6   0.69   0.73   0.78   0.87   1553    1.0\n",
       "fdy2[2]             0.81  1.8e-3   0.07   0.67   0.76    0.8   0.85   0.95   1540    1.0\n",
       "fdy2[3]             0.88  1.9e-3   0.07   0.74   0.83   0.88   0.93   1.03   1527    1.0\n",
       "fdy2[4]             0.96  1.9e-3   0.07   0.82   0.91   0.96   1.01   1.11   1513    1.0\n",
       "fdy2[5]             1.04  1.9e-3   0.07    0.9    1.0   1.04   1.09    1.2   1499    1.0\n",
       "fdy2[6]             1.13  1.9e-3   0.07   0.99   1.08   1.13   1.18   1.28   1485    1.0\n",
       "fdy2[7]             1.22  1.9e-3   0.07   1.07   1.17   1.21   1.26   1.37   1471    1.0\n",
       "fdy2[8]              1.3  1.9e-3   0.07   1.16   1.25    1.3   1.35   1.45   1457    1.0\n",
       "fdy2[9]             1.39  1.9e-3   0.07   1.25   1.34   1.39   1.44   1.54   1443    1.0\n",
       "fdy2[10]            1.48  1.9e-3   0.07   1.33   1.43   1.48   1.52   1.62   1429    1.0\n",
       "fdy2[11]            1.56  1.9e-3   0.07   1.42   1.51   1.56   1.61   1.71   1415    1.0\n",
       "fdy2[12]            1.64  1.8e-3   0.07   1.51    1.6   1.64   1.69   1.79   1401    1.0\n",
       "fdy2[13]            1.72  1.8e-3   0.07   1.59   1.67   1.72   1.76   1.86   1387    1.0\n",
       "fdy2[14]            1.79  1.8e-3   0.07   1.67   1.75   1.79   1.84   1.94   1372    1.0\n",
       "fdy2[15]            1.86  1.7e-3   0.06   1.74   1.82   1.86   1.91    2.0   1350    1.0\n",
       "fdy2[16]            1.93  1.7e-3   0.06   1.81   1.89   1.93   1.97   2.06   1327    1.0\n",
       "fdy2[17]            1.98  1.7e-3   0.06   1.87   1.94   1.98   2.02   2.11   1305    1.0\n",
       "fdy2[18]            2.03  1.7e-3   0.06   1.93   1.99   2.03   2.07   2.16   1283    1.0\n",
       "fdy2[19]            2.08  1.7e-3   0.06   1.97   2.04   2.08   2.11    2.2   1261    1.0\n",
       "fdy2[20]            2.11  1.6e-3   0.06    2.0   2.07   2.11   2.15   2.23   1242    1.0\n",
       "fdy2[21]            2.13  1.7e-3   0.06   2.03   2.09   2.13   2.17   2.25   1224    1.0\n",
       "fdy2[22]            2.15  1.7e-3   0.06   2.04   2.11   2.15   2.18   2.26   1208    1.0\n",
       "fdy2[23]            2.15  1.7e-3   0.06   2.04   2.11   2.15   2.19   2.27   1188    1.0\n",
       "fdy2[24]            2.15  1.7e-3   0.06   2.04   2.11   2.15   2.19   2.27   1171    1.0\n",
       "fdy2[25]            2.13  1.8e-3   0.06   2.02   2.09   2.13   2.17   2.26   1158    1.0\n",
       "fdy2[26]            2.11  1.8e-3   0.06   1.99   2.07   2.11   2.15   2.24   1151    1.0\n",
       "fdy2[27]            2.08  1.9e-3   0.06   1.95   2.03   2.07   2.12   2.21   1150    1.0\n",
       "fdy2[28]            2.04  2.0e-3   0.07   1.91   1.99   2.03   2.08   2.18   1154    1.0\n",
       "fdy2[29]            1.99  2.0e-3   0.07   1.85   1.94   1.99   2.03   2.13   1164    1.0\n",
       "fdy2[30]            1.93  2.1e-3   0.07   1.79   1.88   1.93   1.98   2.09   1178    1.0\n",
       "fdy2[31]            1.87  2.1e-3   0.07   1.72   1.82   1.87   1.92   2.03   1196    1.0\n",
       "fdy2[32]             1.8  2.2e-3   0.08   1.65   1.75    1.8   1.85   1.96   1217    1.0\n",
       "fdy2[33]            1.72  2.2e-3   0.08   1.57   1.67   1.72   1.77   1.89   1241    1.0\n",
       "fdy2[34]            1.65  2.3e-3   0.08   1.49   1.59   1.64    1.7   1.82   1267    1.0\n",
       "fdy2[35]            1.56  2.3e-3   0.08   1.41   1.51   1.56   1.62   1.74   1294    1.0\n",
       "fdy2[36]            1.48  2.3e-3   0.08   1.32   1.42   1.48   1.54   1.66   1322    1.0\n",
       "fdy2[37]            1.39  2.3e-3   0.09   1.23   1.33   1.39   1.45   1.57   1350    1.0\n",
       "fdy2[38]            1.31  2.3e-3   0.09   1.15   1.25   1.31   1.37   1.49   1377    1.0\n",
       "fdy2[39]            1.22  2.3e-3   0.09   1.06   1.16   1.22   1.28    1.4   1405    1.0\n",
       "pred_yields[0,0]    9.54    0.02    1.0   7.51   8.86   9.56   10.2  11.47   1722    1.0\n",
       "pred_yields[1,0]    9.47    0.02    1.0   7.62   8.78   9.45  10.17  11.45   1918    1.0\n",
       "pred_yields[2,0]     9.5    0.02   0.99   7.66    8.8   9.48  10.16   11.5   2000    1.0\n",
       "pred_yields[3,0]    8.09    0.02   1.03   6.13   7.42   8.07   8.78  10.15   1745    1.0\n",
       "pred_yields[4,0]    8.67    0.02    1.0   6.73   8.01   8.66   9.35  10.61   1975    1.0\n",
       "pred_yields[5,0]    9.71    0.02    1.0   7.64   9.05   9.71   10.4  11.65   2000    1.0\n",
       "pred_yields[0,1]    10.5    0.02   0.99   8.56   9.82  10.52  11.17  12.41   2000    1.0\n",
       "pred_yields[1,1]   10.24    0.02    1.0   8.25   9.58  10.24  10.89  12.26   2000    1.0\n",
       "pred_yields[2,1]   10.56    0.02   1.02   8.59   9.86  10.55  11.27  12.57   1826    1.0\n",
       "pred_yields[3,1]   10.22    0.02   0.99   8.25   9.58  10.22   10.9  12.16   1913    1.0\n",
       "pred_yields[4,1]   10.73    0.02    1.0   8.81  10.05  10.73  11.42  12.67   2000    1.0\n",
       "pred_yields[5,1]   10.41    0.02    1.0   8.43   9.75  10.43  11.11  12.35   2000    1.0\n",
       "pred_yields[0,2]    9.98    0.02    1.0   8.01   9.31  10.01  10.66  11.96   1858    1.0\n",
       "pred_yields[1,2]    9.92    0.02   0.99   7.94   9.26   9.93   10.6  11.88   1967    1.0\n",
       "pred_yields[2,2]    9.75    0.02   1.01   7.76   9.07   9.75  10.46  11.72   1977    1.0\n",
       "pred_yields[3,2]    8.96    0.02    1.0   7.04   8.26   8.98   9.63  10.96   1935    1.0\n",
       "pred_yields[4,2]    9.01    0.02   1.01    7.0   8.31   9.03    9.7  11.05   2000    1.0\n",
       "pred_yields[5,2]    9.52    0.02   0.99   7.58   8.88   9.49  10.17  11.53   1961    1.0\n",
       "pred_yields[0,3]    7.87    0.02    1.0   5.78    7.2   7.88   8.53    9.8   2000    1.0\n",
       "pred_yields[1,3]    8.17    0.02    1.0   6.18   7.51   8.15   8.91  10.08   1959    1.0\n",
       "pred_yields[2,3]     8.5    0.02    1.0   6.61   7.78   8.52   9.19  10.47   1756    1.0\n",
       "pred_yields[3,3]    7.55    0.02   1.01   5.52   6.87   7.55   8.27   9.51   1796    1.0\n",
       "pred_yields[4,3]    8.28    0.02    1.0   6.31   7.63    8.3   8.97  10.27   1901    1.0\n",
       "pred_yields[5,3]    9.05    0.02   1.01   7.09   8.38   9.06   9.72  11.03   1941    1.0\n",
       "pred_yields[0,4]    9.44    0.02   1.03   7.38   8.72   9.42  10.13   11.5   1735    1.0\n",
       "pred_yields[1,4]    9.42    0.02   1.05   7.36   8.73   9.42  10.11  11.48   1899    1.0\n",
       "pred_yields[2,4]    9.63    0.02   0.97   7.68   8.98   9.62  10.29  11.55   1954    1.0\n",
       "pred_yields[3,4]    7.63    0.02   0.98   5.68   6.98   7.64   8.31   9.54   2000    1.0\n",
       "pred_yields[4,4]    8.21    0.02   0.99   6.24   7.51   8.22    8.9   10.1   1851    1.0\n",
       "pred_yields[5,4]    9.07    0.02   1.01   7.12   8.38   9.07   9.77  11.01   2000    1.0\n",
       "pred_yields[0,5]    10.5    0.02   0.98   8.51   9.85  10.49  11.18  12.39   2000    1.0\n",
       "pred_yields[1,5]   10.69    0.02   1.04   8.64  10.01  10.68  11.37  12.85   2000    1.0\n",
       "pred_yields[2,5]   10.48    0.02   1.01   8.54   9.78  10.48  11.15  12.49   1983    1.0\n",
       "pred_yields[3,5]   11.66    0.02   1.04   9.67   11.0  11.63  12.33  13.72   1937    1.0\n",
       "pred_yields[4,5]   11.03    0.02   1.01   9.05  10.37  11.03  11.68  13.03   2000    1.0\n",
       "pred_yields[5,5]   11.34    0.02   1.03   9.24  10.67  11.34  12.04  13.39   1831    1.0\n",
       "pred_yields[0,6]   10.29    0.02   1.02   8.33   9.57  10.29  10.97  12.33   1987    1.0\n",
       "pred_yields[1,6]    9.67    0.02   1.02   7.63   9.02   9.67  10.34  11.75   2000    1.0\n",
       "pred_yields[2,6]    10.7    0.02   1.01   8.71  10.03  10.72  11.38   12.7   2000    1.0\n",
       "pred_yields[3,6]   10.44    0.02   1.01   8.41   9.74  10.44  11.16  12.39   1830    1.0\n",
       "pred_yields[4,6]   10.15    0.02   0.99    8.2   9.47  10.18   10.8  12.07   2000    1.0\n",
       "pred_yields[5,6]   10.13    0.02   1.02   8.16   9.45   10.1  10.84  12.12   2000    1.0\n",
       "pred_yields[0,7]     9.5    0.02   1.02   7.51    8.8   9.48  10.17  11.57   1939    1.0\n",
       "pred_yields[1,7]    9.46    0.02   1.03   7.45   8.76   9.44  10.13  11.45   1945    1.0\n",
       "pred_yields[2,7]    9.82    0.02   1.02   7.78   9.16   9.83   10.5  11.75   2000    1.0\n",
       "pred_yields[3,7]   10.09    0.02   0.98   8.15   9.43   10.1  10.78  12.01   1734    1.0\n",
       "pred_yields[4,7]    9.47    0.02    1.0   7.53   8.83   9.43  10.14  11.44   2000    1.0\n",
       "pred_yields[5,7]    9.41    0.02   1.04   7.42   8.71   9.39  10.14  11.46   1893    1.0\n",
       "pred_yields[0,8]    7.95    0.02   1.03   5.93   7.27   7.93   8.63   9.93   2000    1.0\n",
       "pred_yields[1,8]    7.38    0.02   1.03   5.46   6.69   7.38   8.08    9.4   1922    1.0\n",
       "pred_yields[2,8]    8.37    0.02   1.01   6.42   7.69   8.36   9.03  10.43   2000    1.0\n",
       "pred_yields[3,8]     9.4    0.02   0.99   7.41   8.74    9.4   10.1  11.38   1935    1.0\n",
       "pred_yields[4,8]    7.93    0.02   1.02   5.85   7.26   7.92   8.63   9.88   1738    1.0\n",
       "pred_yields[5,8]    8.27    0.02   0.97   6.37   7.62   8.28   8.93  10.05   1914    1.0\n",
       "pred_yields[0,9]   10.66    0.02   0.99   8.68   9.98  10.68  11.32  12.58   1960    1.0\n",
       "pred_yields[1,9]   10.49    0.02   1.02    8.5   9.84  10.48  11.18   12.5   1772    1.0\n",
       "pred_yields[2,9]   10.17    0.02    1.0   8.22   9.48  10.17  10.83  12.08   1843    1.0\n",
       "pred_yields[3,9]    10.4    0.02   0.98   8.51   9.73  10.41  11.04  12.33   1990    1.0\n",
       "pred_yields[4,9]   10.12    0.02   0.99   8.22   9.45  10.13   10.8  12.01   2000    1.0\n",
       "pred_yields[5,9]    9.62    0.02   1.01   7.64   8.94   9.61  10.31  11.67   2000    1.0\n",
       "pred_yields[0,10]   9.59    0.02   0.99   7.67   8.88   9.56  10.27  11.55   2000    1.0\n",
       "pred_yields[1,10]   8.47    0.02   1.02   6.48    7.8   8.48   9.17  10.52   2000    1.0\n",
       "pred_yields[2,10]  10.17    0.02   1.01   8.25   9.46  10.18  10.85  12.15   1971    1.0\n",
       "pred_yields[3,10]   9.02    0.02    1.0   7.07   8.35   9.02   9.72  10.93   2000    1.0\n",
       "pred_yields[4,10]   8.65    0.02   1.02   6.72   7.98   8.64   9.32  10.65   2000    1.0\n",
       "pred_yields[5,10]   9.59    0.02   1.03    7.6   8.87   9.58  10.25  11.63   2000    1.0\n",
       "pred_yields[0,11]   9.54    0.02   1.01   7.61   8.86   9.53  10.22  11.52   2000    1.0\n",
       "pred_yields[1,11]   9.47    0.03   1.01   7.45   8.79   9.48  10.17  11.36   1405    1.0\n",
       "pred_yields[2,11]   9.66    0.02   1.01   7.64   8.99   9.69  10.34  11.67   1762    1.0\n",
       "pred_yields[3,11]   9.53    0.02   1.04   7.41   8.85   9.52  10.23  11.47   1941    1.0\n",
       "pred_yields[4,11]   9.77    0.02    1.0   7.74   9.14   9.77  10.47  11.67   2000    1.0\n",
       "pred_yields[5,11]  10.06    0.02   1.03   8.11   9.34  10.05  10.74  12.07   2000    1.0\n",
       "pred_yields[0,12]    9.7    0.02   1.02   7.66   9.03    9.7  10.41  11.77   2000    1.0\n",
       "pred_yields[1,12]   8.15    0.02   1.01   6.13   7.45   8.16   8.81  10.11   1938    1.0\n",
       "pred_yields[2,12]  10.24    0.02   1.02   8.21   9.59  10.24  10.92  12.28   1859    1.0\n",
       "pred_yields[3,12]   9.85    0.02   1.01   7.88   9.17   9.85  10.54   11.8   1914    1.0\n",
       "pred_yields[4,12]   8.46    0.02   1.01    6.4   7.82   8.47   9.13  10.45   2000    1.0\n",
       "pred_yields[5,12]  10.71    0.02   1.01   8.72  10.01  10.73  11.39  12.67   2000    1.0\n",
       "pred_yields[0,13]   9.86    0.02   1.02   7.82   9.18   9.88  10.55  11.73   2000    1.0\n",
       "pred_yields[1,13]   8.38    0.03   1.03   6.54   7.64   8.36   9.07  10.53   1618    1.0\n",
       "pred_yields[2,13]    9.8    0.02   1.01   7.77   9.12   9.78  10.49  11.78   1734    1.0\n",
       "pred_yields[3,13]   8.22    0.02   1.01   6.25   7.51   8.23    8.9  10.21   2000    1.0\n",
       "pred_yields[4,13]   7.38    0.02   1.06   5.35   6.64   7.39   8.08   9.51   2000    1.0\n",
       "pred_yields[5,13]   8.29    0.02   1.03   6.33   7.59   8.31   8.98  10.31   1820    1.0\n",
       "pred_yields[0,14]   9.81    0.02   1.01   7.86   9.11    9.8   10.5  11.75   2000    1.0\n",
       "pred_yields[1,14]  10.06    0.02   0.99   8.14   9.41  10.05  10.73   12.0   1858    1.0\n",
       "pred_yields[2,14]   9.97    0.02   1.04   7.88   9.26   9.97  10.66  12.09   2000    1.0\n",
       "pred_yields[3,14]  10.24    0.02   0.99   8.29   9.57  10.24  10.91  12.25   2000    1.0\n",
       "pred_yields[4,14]  10.37    0.02   1.01   8.39   9.69  10.38  11.02  12.42   2000    1.0\n",
       "pred_yields[5,14]  11.01    0.02   1.01   9.07  10.31   11.0  11.71  12.95   2000    1.0\n",
       "pred_yields[0,15]   8.12    0.02   0.99   6.17   7.44   8.11   8.79  10.02   2000    1.0\n",
       "pred_yields[1,15]   7.07    0.02    1.0   5.19   6.39   7.06   7.77   9.08   1907    1.0\n",
       "pred_yields[2,15]   8.95    0.02   1.01   6.99   8.26   8.94   9.64  10.92   1871    1.0\n",
       "pred_yields[3,15]   7.78    0.02   1.03   5.81   7.07   7.78    8.5   9.83   1951    1.0\n",
       "pred_yields[4,15]    8.0    0.02    1.0    6.0   7.32    8.0   8.69   9.91   1894    1.0\n",
       "pred_yields[5,15]   9.14    0.02   1.01   7.14   8.44   9.13   9.81   11.1   2000    1.0\n",
       "pred_yields[0,16]   8.72    0.02    1.0   6.71   8.05   8.73   9.36  10.72   1995    1.0\n",
       "pred_yields[1,16]   9.13    0.02   1.01   7.13   8.43   9.16   9.81  11.11   1715    1.0\n",
       "pred_yields[2,16]    9.2    0.02   1.04   7.22   8.49   9.19   9.94  11.16   2000    1.0\n",
       "pred_yields[3,16]   9.59    0.02    1.0   7.62   8.89   9.57  10.24  11.62   2000    1.0\n",
       "pred_yields[4,16]   9.11    0.02    1.0   7.11   8.41    9.1   9.79  11.06   2000    1.0\n",
       "pred_yields[5,16]   9.35    0.02    1.0   7.41   8.69   9.38  10.01  11.32   1934    1.0\n",
       "pred_yields[0,17]   9.39    0.02   1.04   7.34   8.68   9.38  10.11  11.44   2000    1.0\n",
       "pred_yields[1,17]   9.27    0.02   1.02    7.3   8.57   9.26   9.95  11.36   2000    1.0\n",
       "pred_yields[2,17]   9.44    0.02    1.0   7.42   8.77   9.45  10.13  11.34   2000    1.0\n",
       "pred_yields[3,17]   9.54    0.02   1.02   7.59   8.84    9.5  10.26  11.55   2000    1.0\n",
       "pred_yields[4,17]   9.58    0.02   1.02   7.56    8.9   9.57  10.29  11.52   2000    1.0\n",
       "pred_yields[5,17]   8.78    0.02   1.01    6.8   8.04   8.76   9.47  10.75   1910    1.0\n",
       "pred_yields[0,18]   9.55    0.02   0.99   7.64   8.89   9.55  10.22  11.53   2000    1.0\n",
       "pred_yields[1,18]   9.74    0.02   1.01   7.78   9.06   9.75  10.43   11.7   1735    1.0\n",
       "pred_yields[2,18]   9.36    0.02    1.0   7.47   8.67   9.35  10.03  11.36   1889    1.0\n",
       "pred_yields[3,18]   9.76    0.02    1.0   7.78   9.06   9.76  10.45  11.66   1774    1.0\n",
       "pred_yields[4,18]   9.58    0.02   1.04   7.59   8.86   9.58  10.28  11.72   2000    1.0\n",
       "pred_yields[5,18]  10.11    0.02   0.99   8.15   9.45   10.1  10.78  12.09   2000    1.0\n",
       "pred_yields[0,19]   9.13    0.02   1.04   7.17   8.41   9.12   9.84  11.16   2000    1.0\n",
       "pred_yields[1,19]   9.86    0.02   1.01   7.89   9.16   9.87  10.53  11.82   1881    1.0\n",
       "pred_yields[2,19]   9.28    0.02   1.03    7.2   8.59   9.29   9.98  11.25   2000    1.0\n",
       "pred_yields[3,19]   9.44    0.02   1.01   7.42   8.74   9.44  10.11  11.45   2000    1.0\n",
       "pred_yields[4,19]   9.67    0.02   0.97   7.82    9.0   9.65  10.32  11.65   2000    1.0\n",
       "pred_yields[5,19]  10.18    0.02   1.01   8.24   9.49   10.2  10.88  12.13   2000    1.0\n",
       "pred_yields[0,20]  10.71    0.02   1.01   8.73  10.04  10.71  11.36  12.72   1936    1.0\n",
       "pred_yields[1,20]  10.31    0.02   1.01   8.28   9.66  10.32   11.0   12.2   2000    1.0\n",
       "pred_yields[2,20]  11.05    0.02    1.0   9.04  10.38  11.06   11.7  13.07   1995    1.0\n",
       "pred_yields[3,20]   9.35    0.02    1.0   7.46   8.65   9.35  10.03  11.28   1819    1.0\n",
       "pred_yields[4,20]   9.78    0.02   1.02    7.8   9.05   9.77  10.47  11.77   2000    1.0\n",
       "pred_yields[5,20]   9.75    0.02   1.01   7.78   9.07   9.75  10.44  11.76   2000    1.0\n",
       "pred_yields[0,21]  10.91    0.02   1.02   8.88  10.25  10.92  11.59  12.85   2000    1.0\n",
       "pred_yields[1,21]  11.34    0.02   1.01   9.41  10.65  11.33   12.0  13.42   2000    1.0\n",
       "pred_yields[2,21]  10.72    0.02   1.01   8.72  10.04  10.75  11.41  12.66   1867    1.0\n",
       "pred_yields[3,21]  10.18    0.02   1.02   8.17    9.5  10.18  10.86  12.22   1867    1.0\n",
       "pred_yields[4,21]   9.86    0.02    1.0   7.89    9.2   9.89  10.53  11.83   2000    1.0\n",
       "pred_yields[5,21]   9.35    0.02   1.02   7.25   8.68    9.4  10.01   11.3   2000    1.0\n",
       "pred_yields[0,22]   9.12    0.02   0.99   7.21   8.44   9.12   9.81  11.11   2000    1.0\n",
       "pred_yields[1,22]   9.19    0.02   1.02   7.12   8.51   9.19   9.86  11.22   2000    1.0\n",
       "pred_yields[2,22]   9.22    0.02    1.0    7.3   8.52   9.22   9.87  11.25   2000    1.0\n",
       "pred_yields[3,22]   9.45    0.02    1.0   7.43   8.79   9.46  10.12  11.36   1951    1.0\n",
       "pred_yields[4,22]   9.84    0.02    1.0   7.86   9.15   9.84  10.55  11.76   2000    1.0\n",
       "pred_yields[5,22]   9.38    0.02   1.01   7.39   8.66   9.37  10.06  11.33   2000    1.0\n",
       "pred_yields[0,23]   9.67    0.02    1.0   7.79    9.0   9.67  10.35  11.62   2000    1.0\n",
       "pred_yields[1,23]   10.7    0.02   1.01   8.72  10.05  10.69  11.39  12.63   1901    1.0\n",
       "pred_yields[2,23]  10.08    0.02   1.01   8.08   9.43  10.08  10.73  12.02   2000    1.0\n",
       "pred_yields[3,23]   10.4    0.02   1.03   8.35   9.73  10.43  11.09  12.42   1969    1.0\n",
       "pred_yields[4,23]  10.13    0.02   1.02   8.12   9.42  10.14  10.82  12.12   1764    1.0\n",
       "pred_yields[5,23]   9.39    0.02   0.99   7.42   8.74    9.4  10.04  11.35   2000    1.0\n",
       "pred_yields[0,24]   8.65    0.02    1.0   6.66   7.97   8.65   9.35   10.5   1923    1.0\n",
       "pred_yields[1,24]   8.84    0.02   0.99   6.85   8.18   8.86   9.49  10.76   1966    1.0\n",
       "pred_yields[2,24]   9.76    0.02   1.01   7.85   9.07   9.76  10.44  11.78   1883    1.0\n",
       "pred_yields[3,24]  11.04    0.02   1.04   8.97  10.35  11.01  11.74  13.09   1781    1.0\n",
       "pred_yields[4,24]   10.3    0.02    1.0   8.38   9.63  10.26  10.98  12.31   2000    1.0\n",
       "pred_yields[5,24]   9.61    0.02   1.01    7.6   8.93   9.61   10.3  11.58   1971    1.0\n",
       "pred_yields[0,25]   9.53    0.02    1.0   7.53   8.86   9.52  10.18  11.55   2000    1.0\n",
       "pred_yields[1,25]   9.21    0.02   0.99   7.29   8.53   9.21   9.85  11.13   2000    1.0\n",
       "pred_yields[2,25]   9.53    0.02   1.02   7.54   8.83   9.52  10.21  11.59   1880    1.0\n",
       "pred_yields[3,25]  10.31    0.02   1.02   8.37    9.6  10.33  10.99  12.39   1921    1.0\n",
       "pred_yields[4,25]  10.16    0.02    1.0   8.19    9.5  10.16  10.82  12.11   1995    1.0\n",
       "pred_yields[5,25]  10.44    0.02   1.03   8.33   9.78  10.46  11.16  12.37   1725    1.0\n",
       "pred_yields[0,26]  10.91    0.02   1.01   8.89  10.22  10.92  11.59  12.87   2000    1.0\n",
       "pred_yields[1,26]  10.66    0.02   0.97   8.77   9.98  10.68  11.29  12.58   1895    1.0\n",
       "pred_yields[2,26]  10.68    0.02   1.02   8.71   10.0  10.68  11.38  12.73   2000    1.0\n",
       "pred_yields[3,26]   9.52    0.02   1.01   7.57   8.84    9.5  10.18   11.5   2000    1.0\n",
       "pred_yields[4,26]   9.52    0.02   1.04   7.47   8.84   9.51  10.23  11.49   1943    1.0\n",
       "pred_yields[5,26]   9.84    0.02   0.98   7.92   9.16   9.84  10.52  11.76   1890    1.0\n",
       "pred_yields[0,27]   8.48    0.02   1.03   6.45   7.79   8.46   9.16  10.55   1861    1.0\n",
       "pred_yields[1,27]    9.2    0.02   1.03   7.22   8.52   9.19   9.93  11.23   1986    1.0\n",
       "pred_yields[2,27]   9.08    0.02   0.97   7.24   8.43   9.06   9.75  11.02   1791    1.0\n",
       "pred_yields[3,27]   9.02    0.02    1.0   7.15   8.34   9.02   9.67   11.0   1906    1.0\n",
       "pred_yields[4,27]   9.26    0.02   1.01   7.26   8.58   9.26   9.95  11.11   1923    1.0\n",
       "pred_yields[5,27]   9.16    0.02    1.0   7.23   8.46   9.14   9.84  11.15   1829    1.0\n",
       "pred_yields[0,28]   9.46    0.02   1.02   7.58   8.75   9.42  10.13  11.47   2000    1.0\n",
       "pred_yields[1,28]   8.09    0.02   1.03   6.05   7.45   8.11   8.76  10.16   2000    1.0\n",
       "pred_yields[2,28]   9.64    0.02    1.0   7.65   8.96   9.65  10.32  11.57   2000    1.0\n",
       "pred_yields[3,28]   10.2    0.02   1.03   8.17    9.5  10.22  10.89  12.31   1928    1.0\n",
       "pred_yields[4,28]   8.99    0.02   0.99   6.96   8.33    9.0   9.67  10.96   1960    1.0\n",
       "pred_yields[5,28]  10.25    0.02    1.0   8.24   9.59  10.27   10.9  12.16   2000    1.0\n",
       "pred_yields[0,29]  10.65    0.02    1.0   8.59  10.01  10.68  11.33  12.56   1922    1.0\n",
       "pred_yields[1,29]  10.63    0.02   1.01   8.58   9.96  10.66  11.31  12.57   2000    1.0\n",
       "pred_yields[2,29]  11.12    0.02   1.01   9.07  10.44  11.15  11.83  13.05   1900    1.0\n",
       "pred_yields[3,29]   10.7    0.02    1.0   8.75  10.05  10.74  11.36  12.55   2000    1.0\n",
       "pred_yields[4,29]  11.14    0.02   1.03   9.12  10.44  11.14  11.85  13.14   2000    1.0\n",
       "pred_yields[5,29]  10.44    0.02   1.02   8.31   9.77  10.45  11.12  12.39   1893    1.0\n",
       "pred_yields[0,30]   8.19    0.02   1.03   6.15   7.47   8.21   8.89  10.12   1867    1.0\n",
       "pred_yields[1,30]   9.34    0.02   1.03   7.31   8.64   9.37  10.02  11.35   2000    1.0\n",
       "pred_yields[2,30]   8.72    0.02   0.99   6.75   8.06   8.74    9.4  10.57   2000    1.0\n",
       "pred_yields[3,30]    9.4    0.02   1.02   7.31   8.73   9.41   10.1  11.37   1967    1.0\n",
       "pred_yields[4,30]   8.68    0.02   1.01   6.73   7.97   8.65   9.38  10.67   2000    1.0\n",
       "pred_yields[5,30]   9.24    0.02   1.01    7.3   8.56   9.24   9.94  11.25   1866    1.0\n",
       "pred_yields[0,31]   8.77    0.02   1.02   6.78   8.07   8.77   9.46  10.78   2000    1.0\n",
       "pred_yields[1,31]   9.26    0.02   1.01    7.3   8.58   9.25   9.95  11.21   1895    1.0\n",
       "pred_yields[2,31]   8.69    0.02   1.02   6.74   8.01   8.69    9.4  10.67   2000    1.0\n",
       "pred_yields[3,31]   8.59    0.02    1.0   6.63    7.9    8.6   9.26  10.55   1962    1.0\n",
       "pred_yields[4,31]   8.57    0.02   1.02   6.66   7.85   8.57   9.28  10.56   1852    1.0\n",
       "pred_yields[5,31]    8.8    0.02   0.98   6.84   8.13    8.8   9.46  10.72   2000    1.0\n",
       "pred_yields[0,32]   8.83    0.02    1.0   6.86   8.13   8.82    9.5  10.78   2000    1.0\n",
       "pred_yields[1,32]   8.95    0.02    1.0   6.94   8.27   8.93   9.64  10.92   1941    1.0\n",
       "pred_yields[2,32]   9.22    0.02   1.01   7.22   8.53   9.24   9.88  11.23   1904    1.0\n",
       "pred_yields[3,32]   8.23    0.02   1.02   6.16   7.57   8.25   8.92  10.21   1896    1.0\n",
       "pred_yields[4,32]    8.0    0.02   1.01   6.03   7.31    8.0    8.7   9.96   2000    1.0\n",
       "pred_yields[5,32]   6.94    0.02   1.01   4.98   6.26   6.95   7.63   8.96   1915    1.0\n",
       "pred_yields[0,33]   9.95    0.02   1.03   7.91   9.23   9.93  10.67  11.98   1896    1.0\n",
       "pred_yields[1,33]   8.05    0.02   1.01   6.11   7.36   8.07   8.74   9.98   1960    1.0\n",
       "pred_yields[2,33]  10.11    0.02    1.0   8.09   9.48  10.14  10.82  11.93   2000    1.0\n",
       "pred_yields[3,33]   9.09    0.02    1.0   7.17   8.43   9.06   9.76  11.07   2000    1.0\n",
       "pred_yields[4,33]   7.83    0.02   1.01   5.82   7.18   7.82   8.53   9.74   1862    1.0\n",
       "pred_yields[5,33]   7.95    0.02   1.01   5.95   7.26   7.95   8.65   9.94   2000    1.0\n",
       "pred_yields[0,34]  10.57    0.02   1.02   8.53   9.87  10.57  11.26  12.56   1951    1.0\n",
       "pred_yields[1,34]   9.21    0.02   1.03   7.26    8.5   9.21   9.93  11.31   2000    1.0\n",
       "pred_yields[2,34]  10.84    0.02   0.98   8.93  10.18  10.85  11.48  12.76   1973    1.0\n",
       "pred_yields[3,34]   9.44    0.02   1.01   7.47   8.75   9.44  10.15  11.35   1911    1.0\n",
       "pred_yields[4,34]   8.85    0.02   1.03   6.83   8.17   8.84   9.51  10.98   1988    1.0\n",
       "pred_yields[5,34]    9.2    0.02   1.03   7.17    8.5   9.26    9.9  11.11   1892    1.0\n",
       "lp__              -161.4    0.06   1.77 -165.8 -162.3 -161.0 -160.1 -158.9    869    1.0\n",
       "\n",
       "Samples were drawn using NUTS at Wed Apr 24 15:19:26 2019.\n",
       "For each parameter, n_eff is a crude measure of effective sample size,\n",
       "and Rhat is the potential scale reduction factor on split chains (at \n",
       "convergence, Rhat=1)."
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fit2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a236bd3c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.figure(figsize=(10,10))\n",
    "for i in range(0,2000,10):\n",
    "    plt.plot(fit2.data['temp'],samples['fdy1'][i,:],alpha=0.1, c='b')\n",
    "plt.xlabel('Temperature [$^{\\circ}$C]')\n",
    "plt.ylabel(r'$dy$')\n",
    "#plt.title('Growth Curve')\n",
    "plt.savefig('./figs/growth_curve.pdf')\n",
    "plt.savefig('./figs/growth_curve.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.630036266199637"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def yield_anomaly2(temp_6m, precip_6m, mu_t, sigma_t, mu_p, sigma_p, rho, norm):\n",
    "    \"\"\"Take six months of T and P and return yield for given params.\n",
    "    \n",
    "    This should be identical to the function in the STAN model\n",
    "    \"\"\"\n",
    "    if len(norm) == 1:\n",
    "        norm = norm * np.ones(6)\n",
    "    dy = np.zeros(6)\n",
    "    for month in np.arange(6):\n",
    "        dy[month] = norm[month]*np.exp(-(1/(2 - 2*rho**2)) *( np.square((temp_6m[month]    - mu_t)/sigma_t) \n",
    "                                                           + np.square((precip_6m[month] - mu_p)/sigma_p) \n",
    "                                                           - (2*rho*(temp_6m[month]    - mu_t)\n",
    "                                                                   *(precip_6m[month] - mu_p)\n",
    "                                                             )\n",
    "                                                              /(sigma_t*sigma_p)\n",
    "                                                           )\n",
    "                                      )\n",
    "                                      \n",
    "    return np.sum(dy)\n",
    "\n",
    "\n",
    "def compute_mean_yield_anomaly2(T_inc, data, mu_t, sigma_t, mu_p, sigma_p, rho, norm):\n",
    "    \"\"\" Compute mean yield anomaly for a model over regions and years\n",
    "    \n",
    "    The function yield_anomaly returens the yield anomaly for a year \n",
    "    and and region. Here we loop over the regions and years to create\n",
    "    and overall mean.\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    yield_anomalies = np.full((6, 35), np.nan)\n",
    "    #loop over states\n",
    "    for state in np.arange(6):\n",
    "        #loop over years\n",
    "        for year in np.arange(35):\n",
    "            temp_6m = data['d_temp'][state, year, :] + T_inc\n",
    "            precip_6m = data['d_precip'][state, year, :]\n",
    "            yield_anomalies[state, year] = yield_anomaly2(temp_6m, \n",
    "                                                         precip_6m, \n",
    "                                                         mu_t, \n",
    "                                                         sigma_t, \n",
    "                                                         mu_p, \n",
    "                                                         sigma_p,\n",
    "                                                         rho, \n",
    "                                                         norm)\n",
    "    \n",
    "    return np.nanmean(yield_anomalies)\n",
    "\n",
    "\n",
    "compute_mean_yield_anomaly2(1., data2, 19.71, 5.23, 114.69, 62.41, 0.0, [1.56])  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 out of 2000\n",
      "100 out of 2000\n",
      "200 out of 2000\n",
      "300 out of 2000\n",
      "400 out of 2000\n",
      "500 out of 2000\n",
      "600 out of 2000\n",
      "700 out of 2000\n",
      "800 out of 2000\n",
      "900 out of 2000\n",
      "1000 out of 2000\n",
      "1100 out of 2000\n",
      "1200 out of 2000\n",
      "1300 out of 2000\n",
      "1400 out of 2000\n",
      "1500 out of 2000\n",
      "1600 out of 2000\n",
      "1700 out of 2000\n",
      "1800 out of 2000\n",
      "1900 out of 2000\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "#Loop over the samples and for each parameter realisation\n",
    "\n",
    "T_incs_corr = np.linspace(-5, 5, 50)\n",
    "mean_yield_samples2 =  np.full((len(samples2['mu_t']), len(T_incs_corr)), np.nan)\n",
    "\n",
    "for n in np.arange(len(samples2['mu_t'])):\n",
    "    if n % 100 == 0:\n",
    "        print(\"{} out of {}\".format(n, len(samples2['mu_t'])))\n",
    "    mu_t    = samples2['mu_t'][n]\n",
    "    sigma_t = samples2['sigma_t'][n]\n",
    "    mu_p    = samples2['mu_p'][n]\n",
    "    sigma_p = samples2['sigma_p'][n]\n",
    "    rho = samples2['rho'][n]\n",
    "    norm    = [samples2['norm'][n]]\n",
    "    for m, T_inc in enumerate(T_incs_corr):\n",
    "        mean_yield_samples2[n, m] = compute_mean_yield_anomaly2(T_inc, data2, mu_t, sigma_t, mu_p, sigma_p, rho, norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('mean_yield_samples2', mean_yield_samples2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample2_means   = [np.mean(s) for s in mean_yield_samples2.T]\n",
    "sample2_medians = [np.median(s) for s in mean_yield_samples2.T]\n",
    "sample2_25      = [np.percentile(s, 25.) for s in mean_yield_samples2.T]\n",
    "sample2_75      = [np.percentile(s, 75.)  for s in mean_yield_samples2.T]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a22687fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.plot(T_incs_corr, sample2_means, label='means')\n",
    "#plt.plot(T_incs, sample_medians, label='medians')\n",
    "#plt.plot(T_incs, sample_25, label='25th percentile')\n",
    "#plt.plot(T_incs, sample_75, label='75th percentile')\n",
    "plt.fill_between(T_incs_corr, sample2_25, sample2_75, label='75th - 25th percentile', alpha=0.2)\n",
    "plt.legend()\n",
    "plt.xlabel('T$_{{inc}}$ (K)')\n",
    "plt.ylabel('Yield [tonnes ha$^{-1}$]')\n",
    "plt.savefig('./figs/temperature_impact2_m5p5.pdf')\n",
    "plt.savefig('./figs/temperature_impact2_m5p5.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "#                    mean se_mean     sd   2.5%    25%    50%    75%  97.5%  n_eff   Rhat\n",
    "#mu_t               19.67    0.02   0.72  18.09  19.24  19.74  20.18  20.93   1036    1.0\n",
    "#sigma_t             5.23    0.02   0.57   4.32   4.83   5.17   5.57   6.53    797    1.0\n",
    "#mu_p              115.07    0.11   4.12 106.99 112.34 115.12 117.78 123.13   1443    1.0\n",
    "#sigma_p            62.62    0.08   2.82  57.36  60.63  62.57   64.5  68.19   1358    1.0\n",
    "#rho                -0.02  2.0e-3   0.08  -0.17  -0.07  -0.02   0.03   0.13   1433    1.0\n",
    "#norm                1.56  2.1e-3   0.07   1.44   1.51   1.56    1.6   1.69    973    1.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rs548/anaconda/envs/herschelhelp_internal/lib/python3.6/site-packages/ipykernel/__main__.py:7: DeprecationWarning: object of type <class 'float'> cannot be safely interpreted as an integer.\n",
      "/Users/rs548/anaconda/envs/herschelhelp_internal/lib/python3.6/site-packages/ipykernel/__main__.py:8: DeprecationWarning: object of type <class 'float'> cannot be safely interpreted as an integer.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a22e2c160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "t1, t2 = 5., 35.\n",
    "p1, p2 = 0., 250.\n",
    "x = np.linspace(t1,t2,100.)\n",
    "y = np.linspace(p1,p2,100.)\n",
    "X,Y = meshgrid(x, y) # grid of point\n",
    "def yield_equal2(temp, precip, mu_t, sigma_t, mu_p, sigma_p, rho, norm):\n",
    "    \"\"\"Take six months of T and P and return yield for given params.\n",
    "    \n",
    "    This should be identical to the function in the STAN model\n",
    "    \"\"\"\n",
    "    temp_6m = np.full(6, temp)\n",
    "    precip_6m = np.full(6, precip)\n",
    "    if len(norm) == 1:\n",
    "        norm = norm * np.ones(6)\n",
    "    dy = np.zeros(6)\n",
    "    for month in np.arange(6):\n",
    "        dy[month] = norm[month]*np.exp(-(1/(2 - 2*rho**2)) *( np.square((temp_6m[month]    - mu_t)/sigma_t) \n",
    "                                                           + np.square((precip_6m[month] - mu_p)/sigma_p) \n",
    "                                                           - (2*rho*(temp_6m[month]    - mu_t)\n",
    "                                                                   *(precip_6m[month] - mu_p)\n",
    "                                                             )\n",
    "                                                              /(sigma_t*sigma_p)\n",
    "                                                           )\n",
    "                                      )\n",
    "    return np.sum(dy)\n",
    "def z_func(X,Y):\n",
    "    return yield_equal2(X, Y, 19.74, 5.17, 115.12, 62.57, -0.02, [1.56])\n",
    "z_func = np.vectorize(z_func)\n",
    "#Lets use the median\n",
    "\n",
    "Z = z_func(X, Y) # evaluation of the function on the grid\n",
    "\n",
    "im = ax.imshow(np.flip(Z, axis=0),cmap=\"viridis\", extent=[t1,t2,p1,p2], aspect=\"auto\") # drawing the function\n",
    "# adding the Contour lines with labels\n",
    "#cset = contour(Z,arange(-1,1.5,0.2),linewidths=2,cmap=cm.Set2)\n",
    "#clabel(cset,inline=True,fmt='%1.1f',fontsize=10)\n",
    "\n",
    "cbar = fig.colorbar(im) # adding the colobar on the right\n",
    "cbar.set_label('Yield [tonnes ha$^{-1}$]')\n",
    "# latex fashion title\n",
    "#title('$z=(1-x^2+y^3) e^{-(x^2+y^2)/2}$')\n",
    "\n",
    "ax.set_xlabel('T (Celsius)')\n",
    "ax.set_ylabel('Monthly precipitation (mm)')\n",
    "\n",
    "\n",
    "\n",
    "plt.savefig('./figs/2d_2.pdf')\n",
    "plt.savefig('./figs/2d_2.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rs548/anaconda/envs/herschelhelp_internal/lib/python3.6/site-packages/ipykernel/__main__.py:5: DeprecationWarning: object of type <class 'float'> cannot be safely interpreted as an integer.\n",
      "/Users/rs548/anaconda/envs/herschelhelp_internal/lib/python3.6/site-packages/ipykernel/__main__.py:6: DeprecationWarning: object of type <class 'float'> cannot be safely interpreted as an integer.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a22428278>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "t1, t2 = 5., 35.\n",
    "p1, p2 = 0., 250.\n",
    "x = np.linspace(t1,t2,100.)\n",
    "y = np.linspace(p1,p2,100.)\n",
    "X,Y = meshgrid(x, y) # grid of point\n",
    "def yield_equal2(temp, precip, mu_t, sigma_t, mu_p, sigma_p, rho, norm):\n",
    "    \"\"\"Take six months of T and P and return yield for given params.\n",
    "    \n",
    "    This should be identical to the function in the STAN model\n",
    "    \"\"\"\n",
    "    temp_6m = np.full(6, temp)\n",
    "    precip_6m = np.full(6, precip)\n",
    "    if len(norm) == 1:\n",
    "        norm = norm * np.ones(6)\n",
    "    dy = np.zeros(6)\n",
    "    for month in np.arange(6):\n",
    "        dy[month] = norm[month]*np.exp(-(1/(2 - 2*rho**2)) *( np.square((temp_6m[month]    - mu_t)/sigma_t) \n",
    "                                                           + np.square((precip_6m[month] - mu_p)/sigma_p) \n",
    "                                                           - (2*rho*(temp_6m[month]    - mu_t)\n",
    "                                                                   *(precip_6m[month] - mu_p)\n",
    "                                                             )\n",
    "                                                              /(sigma_t*sigma_p)\n",
    "                                                           )\n",
    "                                      )\n",
    "    return np.sum(dy)\n",
    "def z_func(X,Y):\n",
    "    return yield_equal2(X, Y, 19.74, 5.17, 115.12, 62.57, -0.02, [1.56])\n",
    "z_func = np.vectorize(z_func)\n",
    "#Lets use the median\n",
    "\n",
    "Z = z_func(X, Y) # evaluation of the function on the grid\n",
    "\n",
    "im = ax.imshow(np.flip(Z, axis=0),cmap=\"viridis\", extent=[t1,t2,p1,p2], aspect=\"auto\") # drawing the function\n",
    "# adding the Contour lines with labels\n",
    "#cset = contour(Z,arange(-1,1.5,0.2),linewidths=2,cmap=cm.Set2)\n",
    "#clabel(cset,inline=True,fmt='%1.1f',fontsize=10)\n",
    "\n",
    "cbar = fig.colorbar(im) # adding the colobar on the right\n",
    "cbar.set_label('Yield [tonnes ha$^{-1}$]')\n",
    "# latex fashion title\n",
    "#title('$z=(1-x^2+y^3) e^{-(x^2+y^2)/2}$')\n",
    "\n",
    "ax.set_xlabel('T (Celsius)')\n",
    "ax.set_ylabel('Monthly precipitation (mm)')\n",
    "\n",
    "CS = ax.contour(X, Y, Z, 3, colors='w')\n",
    "ax.clabel(CS, inline=1, fmt='%1.1f', fontsize=10)\n",
    "\n",
    "plt.savefig('./figs/2d_2_contour.pdf')\n",
    "plt.savefig('./figs/2d_2_contour.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[ 75.25291753,  85.83683791, 133.44429417, 109.26561872,\n",
       "         137.0134282 ,  77.79794749],\n",
       "        [149.54253603, 173.75202661, 125.11506518, 131.02297284,\n",
       "         102.20628622,  97.21814035],\n",
       "        [ 56.44104887, 123.62218878, 107.43842577,  98.07348464,\n",
       "          84.64356708,  58.22764308],\n",
       "        ...,\n",
       "        [ 70.29394894,  86.63458033,  25.41465675,  77.06708532,\n",
       "          97.02873585, 138.82704216],\n",
       "        [176.43561756, 119.985384  , 149.03634517, 105.14860626,\n",
       "          71.01551827,  65.76953295],\n",
       "        [141.11174954, 120.43960547, 166.33686622,  80.6756401 ,\n",
       "         131.67100444,  92.48618428]],\n",
       "\n",
       "       [[ 64.93835524,  81.33833423, 113.28620661,  91.90444217,\n",
       "         142.41877647, 121.70702232],\n",
       "        [142.55719903, 151.73163468, 132.87625895, 164.63142815,\n",
       "         150.76911885,  83.358952  ],\n",
       "        [ 69.62338591, 135.0633929 , 101.42119788, 156.0985889 ,\n",
       "          90.81889853,  66.76183226],\n",
       "        ...,\n",
       "        [100.319296  ,  92.822115  ,  28.17661957,  51.22126308,\n",
       "          90.71920902,  90.02742209],\n",
       "        [186.7157709 , 209.81234241, 125.49589295,  85.72452233,\n",
       "          33.91325765,  61.25773303],\n",
       "        [156.23987765, 100.38715635, 217.84798881,  67.59291882,\n",
       "         162.53372749, 102.06982818]],\n",
       "\n",
       "       [[ 84.9870375 , 105.27153162, 133.81842802, 141.62002159,\n",
       "         162.16988723,  65.9144529 ],\n",
       "        [136.30086325, 143.4771824 , 163.69731549, 109.28423753,\n",
       "          77.95951838, 112.44932683],\n",
       "        [ 52.77451795, 126.17381465, 120.16258837,  69.468664  ,\n",
       "          79.57367914,  63.41571423],\n",
       "        ...,\n",
       "        [ 83.18701935,  80.48604901,  52.02976297,  98.94152633,\n",
       "          71.2266646 , 118.43042145],\n",
       "        [131.20989985,  97.79043888, 156.48988379, 143.97626278,\n",
       "          97.76875305,  64.94344348],\n",
       "        [130.76255436, 120.76914404, 142.7121957 ,  84.59258386,\n",
       "         132.31176696,  88.63079408]],\n",
       "\n",
       "       [[ 67.61201467,  74.50384449,  70.17883272,  60.93163184,\n",
       "         135.76917757,  41.75208963],\n",
       "        [ 63.91211228, 124.90500475,  75.97418169, 147.08030032,\n",
       "         113.55199184,  57.78305001],\n",
       "        [ 72.60270942, 224.68968249,  83.15329033, 119.84256166,\n",
       "         115.24219233,  78.09236177],\n",
       "        ...,\n",
       "        [143.92563728,  94.76638714,  72.43282315,  36.83418754,\n",
       "          39.92165083,  36.37076636],\n",
       "        [119.11417988, 155.96120755,  71.51991942,  68.04368619,\n",
       "          75.18739288,  76.01871342],\n",
       "        [100.11329298, 111.2006557 , 170.95198336,  63.32323799,\n",
       "         213.7900324 , 114.00547755]],\n",
       "\n",
       "       [[ 50.30239245,  79.11217273, 103.70389214,  67.2013782 ,\n",
       "         185.88667987,  84.20836583],\n",
       "        [102.84391791,  90.06079077, 122.74002397, 150.201678  ,\n",
       "         118.833305  ,  61.76030286],\n",
       "        [ 58.6900007 , 201.3073337 ,  79.20834988, 150.28946373,\n",
       "         117.32320897,  85.69023474],\n",
       "        ...,\n",
       "        [130.41429329, 123.94247953,  50.50467336,  45.31008557,\n",
       "          45.16804326,  29.73679342],\n",
       "        [166.67224497, 183.26368802,  82.76780004,  53.30554424,\n",
       "          40.95343986,  59.61685419],\n",
       "        [139.07841087, 100.82876337, 195.65917516,  66.25529459,\n",
       "         215.87520292, 114.17282341]],\n",
       "\n",
       "       [[ 51.43796398,  97.44698805, 103.04181102,  64.3323445 ,\n",
       "         174.98435066, 102.50355773],\n",
       "        [106.18638228,  90.67222292, 135.49244908, 122.93551216,\n",
       "         118.85867163,  38.86617048],\n",
       "        [ 72.95518147, 165.29694617,  53.86681818, 108.61816308,\n",
       "          85.68711108, 102.95185397],\n",
       "        ...,\n",
       "        [131.83337639, 205.44900837,  73.93858831,  93.05127072,\n",
       "          26.28177909,  15.82047736],\n",
       "        [133.29575739, 179.25501253, 128.8988567 ,  73.79607827,\n",
       "          36.80354407,  54.44325573],\n",
       "        [155.96887776, 143.63785887, 185.24839884,  76.88098615,\n",
       "         130.02868886,  66.34612594]]])"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data2['d_precip']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Now lets plot a response in temperature and precipitation space"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "T_incs_2 = np.linspace(-10, 10, 50)\n",
    "P_incs_2 = np.linspace(-100, 100, 50)\n",
    "T_P_grid = np.meshgrid(T_incs_2, P_incs_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def yield_anomaly2(temp_6m, precip_6m, mu_t, sigma_t, mu_p, sigma_p, rho, norm):\n",
    "    \"\"\"Take six months of T and P and return yield for given params.\n",
    "    \n",
    "    This should be identical to the function in the STAN model\n",
    "    \"\"\"\n",
    "    if len(norm) == 1:\n",
    "        norm = norm * np.ones(6)\n",
    "    dy = np.zeros(6)\n",
    "    for month in np.arange(6):\n",
    "        dy[month] = norm[month]*np.exp(-(1/(2 - 2*rho*2)) *( np.square((temp_6m[month]    - mu_t)/sigma_t) \n",
    "                                                           + np.square((precip_6m[month] - mu_p)/sigma_p) \n",
    "                                                           - (2*rho*(temp_6m[month]    - mu_t)\n",
    "                                                                   *(precip_6m[month] - mu_p)\n",
    "                                                             )\n",
    "                                                              /(sigma_t*sigma_p)\n",
    "                                                           )\n",
    "                                      )\n",
    "                                      \n",
    "    return np.sum(dy)\n",
    "\n",
    "def compute_mean_yield_anomaly3(T_inc, P_inc, data, mu_t, sigma_t, mu_p, sigma_p, rho, norm):\n",
    "    \"\"\" Compute mean yield anomaly for a model over regions and years\n",
    "    \n",
    "    The function yield_anomaly returens the yield anomaly for a year \n",
    "    and and region. Here we loop over the regions and years to create\n",
    "    and overall mean.\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    yield_anomalies = np.full((6, 35), np.nan)\n",
    "    #loop over states\n",
    "    for state in np.arange(6):\n",
    "        #loop over years\n",
    "        for year in np.arange(35):\n",
    "            temp_6m = data['d_temp'][state, year, :] + T_inc\n",
    "            precip_6m = data['d_precip'][state, year, :] + P_inc\n",
    "            yield_anomalies[state, year] = yield_anomaly2(temp_6m, \n",
    "                                                         precip_6m, \n",
    "                                                         mu_t, \n",
    "                                                         sigma_t, \n",
    "                                                         mu_p, \n",
    "                                                         sigma_p,\n",
    "                                                         rho, \n",
    "                                                         norm)\n",
    "    \n",
    "    return np.nanmean(yield_anomalies)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "take_100 = np.random.choice([0,1], \n",
    "                            size=len(samples2['mu_t']), \n",
    "                            p=(1-100/len(samples2['mu_t']), 100/len(samples2['mu_t'])))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 out of 50\n",
      "1 out of 50\n",
      "2 out of 50\n",
      "3 out of 50\n",
      "4 out of 50\n",
      "5 out of 50\n",
      "6 out of 50\n",
      "7 out of 50\n",
      "8 out of 50\n",
      "9 out of 50\n",
      "10 out of 50\n",
      "11 out of 50\n",
      "12 out of 50\n",
      "13 out of 50\n",
      "14 out of 50\n",
      "15 out of 50\n",
      "16 out of 50\n",
      "17 out of 50\n",
      "18 out of 50\n",
      "19 out of 50\n",
      "20 out of 50\n",
      "21 out of 50\n",
      "22 out of 50\n",
      "23 out of 50\n",
      "24 out of 50\n",
      "25 out of 50\n",
      "26 out of 50\n",
      "27 out of 50\n",
      "28 out of 50\n",
      "29 out of 50\n",
      "30 out of 50\n",
      "31 out of 50\n",
      "32 out of 50\n",
      "33 out of 50\n",
      "34 out of 50\n",
      "35 out of 50\n",
      "36 out of 50\n",
      "37 out of 50\n",
      "38 out of 50\n",
      "39 out of 50\n",
      "40 out of 50\n",
      "41 out of 50\n",
      "42 out of 50\n",
      "43 out of 50\n",
      "44 out of 50\n",
      "45 out of 50\n",
      "46 out of 50\n",
      "47 out of 50\n",
      "48 out of 50\n",
      "49 out of 50\n"
     ]
    }
   ],
   "source": [
    "mean_yield_anom = np.full((len(T_incs_2), len(P_incs_2)), np.nan)\n",
    "\n",
    "print(\"Starting at {}\".format(time.time()))\n",
    "for n, t in enumerate(T_incs_2):\n",
    "    print(\"{} out of {}\".format(n, len(T_incs_2)))\n",
    "    for m, p in enumerate(P_incs_2):\n",
    "        mean_yield_samples = np.full(len(samples2['mu_t']), np.nan)\n",
    "        for k in np.arange(len(samples2['mu_t'])):\n",
    "            if not take_100[k]:\n",
    "                continue\n",
    "            \n",
    "            #print('k = {}'.format(k))\n",
    "            mu_t    = samples2['mu_t'][k]\n",
    "            sigma_t = samples2['sigma_t'][k]\n",
    "            mu_p    = samples2['mu_p'][k]\n",
    "            sigma_p = samples2['sigma_p'][k]\n",
    "            rho = samples2['rho'][k]\n",
    "            norm    = [samples2['norm'][k]]\n",
    "\n",
    "            mean_yield_samples[k] = compute_mean_yield_anomaly3(t, p, data2, mu_t, sigma_t, mu_p, sigma_p, rho, norm)\n",
    "        \n",
    "        mean_yield_anom[n, m] = np.nanmean(mean_yield_samples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('mean_yield_anom', mean_yield_anom)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEOCAYAAACXX1DeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAIABJREFUeJztvXu4bFdV5v17q/Y+5+SESALhEnIx0EaUSxNCRGwuHxrkkubjYotfeL7WCGrElhb6piAKNHx0iw34abciB0mLitzEkDwYCIEW0cZETiCExCSShCCHhMQQLiHJ2Zeq0X+sdUhl1ztrr3Xquvcev+epZ1eNmnOuWWutXbPmfOcYQxFBkiRJkpTozLsDSZIkyWKTA0WSJEkykhwokiRJkpHkQJEkSZKMJAeKJEmSZCQ5UCRJkiQjWaiBQtK5km6VdOWA7X6SLpb0hfrvMbVdkn5H0nWSrpB02vx6niRJsn1ZqIEC+EPgmRtsrwA+HhGnAB+vXwM8CzilfpwDvHVGfUySJNlRLNRAERGfBG7fYH4u8M76+TuB5w3Y/ygqLgGOlnTcbHqaJEmyc1iogaLAgyLiZoD67wNr+/HAlwfKHahtSZIkyQRZmncHxkDGNhSPRNI5VEtTdLq7Hrf3yAfcu0JpqNRw8+GO6GzTKluoP+7xLaZsq2AvbY41DQqdbdytNh/WlNWY9YttNLW1qF/sqwnvo36b45s33PH7hQZs/UJZ14at7z6Ar/+tuP22iHiAKd2YZ/zwkfG123uNyl52xcpFEbFx6X0h2AoDxS2SjouIm+ulpVtr+wHgxIFyJwA3bawcEfuAfQBH3feEOO2Jv3Sv93u7/VdHb9ewvb88bHPlqrKmTVO/VDbMlXHlwA92fVM/uoX63eF/ElfWDj7QeFAr1h8T+0VX+D5wZe2XX6F+pzf8IbRu2ix8N3SM3dUH6KwZ2/rwB3DlSvbumqvvv3y7q6assXVX/cnqrAx/2M5qMxuAVoY/gLMBsDpsj5XV4XIrK7Z6rA6X/ejdf/Ilf7Dm3HZ7j0svOqFR2eXjrj923ONNi62w9HQBcHb9/Gzg/AH7T9W7n54AfPPQElWSJMliEPSi3+ixyCzUjELSu4GnAsdKOgC8BvgN4H2Sfgb4R+AFdfELgTOB64C7gBdt1n50hmcQpRmBs/tZhj9Wv01ZY3czgmL9JTMjcDOKws+CxjOK4jKd+UXqyk5iRjHuckzfzAjcD1pTDqDfM7+y7SzB1w83+yjO9Fx9syRaqt8x17UzXL/bZvnVtNluSbX5TdDmV2zTldbSKpvt1d0tOlAggHWaLT0tMgs1UETECwtvnWHKBvCL0+1RkiTJ4RMEvW2QymGhBookSZLtRr/dNpCFJAeKJEmSKRFALweKJEmSZBQ5o9hqSEOCdBsxu7+rWblyWd8tvxV2+OYqbo81YqYXuAs3rBWunUBdqG+F6+ai59g44bm0Nb9hWZltsAAYuxWoC1te+0bX7Kw1F777xlYSo51wbTcumHLgxeiuVY0LHWh8vZt/DY0tcBf0gpjSrqOA1CiSJEmSMkGwljOKJEmSpEiA2VG95ciBIkmSZEoERSf/LcWOGihCw2E4WmkMJtxHSXewTnTFss30CFcOvPZg9QizPl20G5sK9WW0C+tX1SoAUgvMYnppydlpFNbWQqPo2bAevr7TI4oOc0aPsHpGyeHNak8NY5hBQWdqfvzxAz+00C6cDmDiN5U0CvWn9XUuenMPfDY+O2qgSJIkmSWBj1e41ciBIkmSZEoEsLolQuqNJgeKJEmSKdKfVtjkGbKzBgoNaxIl3wQfFHC4XDFQ327Xpp+D2jDjRo9wNoBYMuurRqOQKwd0TNlOZ7hsp1uobzWKhrpFoWyJcHqEW57u+19x/YYaRbG+C8rXGy4bBd+I3pLRM9b8scJE1G4TrDFWx8up4vUId63afBGazrbKp+EFHV+2eT6LaWkUlWd2DhRJkiRJgUD0cukpSZIkGUUuPSVJkiRFcukpSZIkGUkg1lwwty3G1v8ELagc7u5tc5noKruxOSc4I1qDF677JTHb2K1wvewFNxm7E6i7Sz7TVteI1M62ZARu8MJ31wjcnYJoPa6Y7ab2vVKGOiNS91x9I1ADrK8Pi6mubH+pJFAb4buwhN03EfjCpIGOwi4Bn42uhcOetY+dN85Ub7GG30L49k54pXan5+yQM4okSZKkSITotRkIF5SFHygkPRx474DpYcCrgaOBnwP+qbb/akRcOOPuJUmSjKS/DWYUCz/URcS1EXFqRJwKPA64Czivfvu3Dr2Xg0SSJItGJWZ3Gj02Q9LLJF0p6SpJL59+7+9h4WcUGzgDuD4ivqSS99YoBL2hoIC+qHWus8mImusOJY0inH2X0R2WCxqDcaRbMmWXCxrFUnfYvsuUXS5oFF1jd3pGZ0px+d0vtvWCw1zP2NeMbb3nHbvWzXlZM2XX1nz9XnfY3i8EW/R6hrnvO6XEQc30hJLG0ZySyOEc3ozDYUE3kC3rP6sN9tcfPtedQsxv9af1VTiZpSdJj6JaQXk8sAp8RNJfRMQXxm68AQs/o9jAWcC7B16/VNIVks6VdMy8OpUkSeIIYC26jR6b8P3AJRFxV0SsA38FPH/a/T/ElhkoJO0CngO8vza9FfhnwKnAzcCbC/XOkbRf0v71u++cSV+TJEngHs/sCSw9XQk8RdL9Je0FzgROnPoHqNlKS0/PAj4TEbcAHPoLIOntwIdcpYjYB+wD2PvAE7dBwN8kSbYS/eZLT8dK2j/wel/9/UVEXC3pjcDFwLeBzwGFzOyTZysNFC9kYNlJ0nERcXP98vlUI26SJMnCcEjMbshtEXF6sa2IdwDvAJD0X4ADY3ewIVtioKinWj8K/PyA+TclnUp1LW7c8J7FOtwVor86Z0orUJeixzonuoKY7YTr7u5h0bTkMLdr1/APi11Lzubr7zZi9u7ucP1lUw5gyaiRS53hsiWHu5Ld4ZzrmtoAVo1o6YTvNSOEAqysD9dfNWL2Ssf/a60ZR8b1TkE4N/aQce4rOsy5L6imEWELZV2pUnV3DYzo3Ka+/C1oRW6ZjEEu0i9AFK73uASyDp2Hg6QHRsStkk4Cfgz4oYk03IAtMVBExF3A/TfYfnJO3UmSJGlEBJMM4fEBSfcH1oBfjIivT6rhzdgSA0WSJMnWRBNzuIuIJ0+kocMgB4okSZIpEZAhPLYcTqMonIG+Ccrnyo7rRAdej1haNhqD0SIA9piye4xGccSSSZkG7HZ6RsfZShrHcFmnW3QKnlXdFhqFw+kJJY1izfzTOt3C6Q4AB7vDotRKb7h+Sc85uGaO1fU3oUxQP6dblDWKYZsXVps7zDknuKIfpbnc7hYoBXD0Zf2hZNqQ0SO07K9rGD1jUmTioiRJkqRIoExclCRJkowmZxRJkiRJkSpx0XS23s6SnTVQaNg/orRzzSYpMrpFFJIJuSRDnV0FP4SGesTeXV5jcHrE3uXhDDd7l0zWG2CP0RiO6A4fa7fRLcD7TCybDe/OBuP7UbhfbKXp/orzozD/yHf3vIPMnt7wObhrfThapNNowOsxLskT+IROzubvCp+jx8bOK4mtDf0gSiKJTBBNqyUU9AGnXZT8KPrLDTWKUlDAgn/FuAStPLMXlp01UCRJksyYzHCXJEmSFIlQziiSJEmS0aQfRZIkSVIk2B6pUHfUQBEaTnrlBGrwznXhyi4VxDEjZpeC+i2bbHTOie6IZS9bHtlQuD6yIGYf2R22H2FsJTF7T2e4X064LjrcjZn5zq0BrxU8Kd0OFCdwH1EI6nd31wjXRszfZTYIgM8G6GzQTuR3uLvFHanoa2Z+CTuHu37RCc7YGgrc4IXnYhDOnhGzzSXUUuFYS9MSs1UMMLmV2FEDRZIkyaxJP4okSZKkSHpmJ0mSJJvSzxnF1qOpw10Y7cFqFAWHu+7SsN1pEeD1iL1GjygF9XN6xFFLK0O2I40N4D7dYfteq1EUHP5kkhxZm//83YJ20RS3q2StkAzI5QY4aGwrhcXw3cbhbsl8rjsLn8npDuM6HJYIU3bN6AFR0AhcoDyrR5Q0CqcbON2i8D8o5/Rq2gRw/pHudis51nUKjnjjEsHEEhfNkx03UCRJksyKQKynmJ0kSZKMIj2zkyRJkiJVrKccKGaCpBuBO4AesB4Rp0u6H/Be4GTgRuAnNs0hK+hv0B5Ks0KnUdA168tGiwDvM7Gr4EfhEgftMXrEfZa9xuD0iKOWDw7XN1oEwFHd4bJ7O200iuGyzrei5C/hNIpOYeHbCYNOoyj9ijtotIe9RqO4S7ttfRvssGFQRBjfN8LpDs4G0O8PnxenR6wXNYpmbTotAryPktUtCr4RHVN24//vd8oatxXnStPpFvpasI/P9gjhsZU+wQ9HxKkRcXr9+hXAxyPiFODj9eskSZKFol/nzd7ssRmS/p2kqyRdKendkvbMoPvA1hooNvJc4J3183cCz5tjX5IkSYY4tOupyWMUko4Hfgk4PSIeBXSBs2bwEYAtsvREtdT3UVXB+N8WEfuAB0XEzQARcbOkB7qKks4BzgFYuu8xs+pvkiTJpHc9LQFHSFoD9gI3TarhJgfeCjwxIm6qB4OLJV3TtGI9qOwD2HP8idPLoJ4kSWKYRFDAiPiKpDcB/wjcDXw0Ij46dsMN2RIDRUTcVP+9VdJ5wOOBWyQdV88mjgNu3bQhwcaYcGEEamghZhfqLzmHuzZitgkqt8dknQPvSOeEaydaV2WH7Ud2jBOesQHs0XC/dtkMd4VAeUbkLgUQdMKgE65LoZ0PGuXUOeGV+urspb5OA7eDprSrxgX76xvh2gnUAD0jJocL1FcKjOnqu2Cbpax1NqhfqezwsZzAbf+vgZhaUMBWu56OlbR/4PW++ocuko6hWm5/KPAN4P2S/nVE/Mkk+1ti4TUKSUdKOurQc+DpwJXABcDZdbGzgfPn08MkSZIy/eg0egC3RcTpA499A808DfhiRPxTRKwBfw78i1l9hq0wo3gQcJ4kqPr7pxHxEUmfBt4n6WeopmMvmGMfkyRJhomJBQX8R+AJkvZSLT2dAewfXWVyLPxAERE3AI8x9q9RnawkSZKFZFKJiyLiUkl/BnwGWAc+S629zoKFHygmSTCci8XksanoGGchozt0un59ernrktkUNAqnRxiHuyMKGoWzO4c5p0UAHNW5e8h2pKnvdAvwGoWzldbyd5WiyjnM/5xNXFS4sHtMOh/nhFfUUxoG9WuTjKmUr8D9El03a+m9gsawZuzrveF7sFcIlNdfdhqF0x0K6/7rzRzmZMoBuLiOUXSYM9qJqe+0jOpYUwoKCKwXrk/rtiJeA7xmIo21ZEcNFEmSJLMmQ3gkSZIkReaVuKgOcbQZ/Yj4RpP2cqBIkiSZIpPQKA6Dm+rHqIN3gZOaNLazBgoxvCHYaBFAY5+JbkGjWDJ2p1sA7DJB5ZzuUNIonB6x1/hRlDQGp0c43cIF+gM40gQFXDZ6xHJBi3Dr/qU1/qYhm9cKm/PXjH+FO/9dFxGvQLeFxtLG58N59K4Y54JV44cDsMfUX+85W8mPwughRiMpaRRN9YiSTuiD+vmyrg2nZ5Ti85W0i7GJuS09XR0Rjx1VQNJnmza2swaKJEmSGTJJMbslPzShMsAWcLhLkiTZqhzSKJo8JnrciO9scZT0K5uV2YycUSRJkkyRUr6QaSHpfYMvgVOBN47TZg4USZIkU2QOYva3IuJnD72Q9NZxG9xxA0VsEK9LQQHdopyc6FoUs43DnRFNAXYZh7tdJqLZbhflDC8yO4HbOcGBF7ldm060rtod/lzOZvy3AL/+WQ7MPHwN3Flds1ZYc9n0wmxcKIjZTR3pSk50zhFwreP/DVe6w3YrZhfCWK92h+2rS8a27uuvdYbPQd/c7yUnOC8wN7OBF56LZZ1zXUMbQKcznS/zmI+Y/YYNr181boM7bqBIkiSZJbNeeoqILwJIOjYibouI28dtM8XsJEmSqSF6/U6jxxQ4d1IN5YwiSZJkSrTMRzFpJnbgHTdQDK17Fk6lXFBAs2brAsJBwYnM1AevXTiHtZLD225j32WC2pXqNw3q53QHgL0N9Yhd8id7ecz7uWd0g12FJleNHtHFaz+OvjkvPbO+7ZIhgdeDVjrDQQkB9naHNSGnUawUFt4Pdk2wQ3MPOj0NYMnoGb314fNX1PmMPcy5KiYPM2XbaAwb9cjK5uuX7GMTlU4xJyZ25B03UCRJksySOYXwgJxRJEmSLD7B7MXsAV45qYZyoEiSJJkaomfylM+CiLhyUm3trIFCDE/GStfQaAxuib2kUTj7UkGjcAl9lqxu4deSnd3b/Fq8szf1jajqD9v2mJO1LL8Q3DUXodNiQ17fBOVzukXVrrsGzjfDf9a++Vwu0N8e48cC3o/CaUwAu01CJedL4/Qs8L44LjDlcsEXaMWs8XeMrbRhx2kP1lbwYbA+E8VjNStbPFYpOOgEmOOMAgBJxwCnAHsO2SLik23aWPjtsZJOlPSXkq6WdJWkl9X210r6iqTL68eZ8+5rkiTJIBHVQNHkMQ0k/SzwSeAi4D/Xf1/btp2tMKNYB/5DRHxG0lHAZZIurt/7rYh40xz7liRJMpI5Z7h7GfADwCUR8cOSvo9qwGjFwg8UEXEzcHP9/A5JVwPHz7dXSZIkzZjj9liAgxFxUBKSdkfENZIe3raRhV96GkTSycBjgUtr00slXSHp3HodztU5R9J+Sft7375zRj1NkiSpmNTSk6SHDyy1Xy7pW5Jevkm1A5KOBj4IXCzpfKrMd61Y+BnFISTdB/gA8PKI+FYdEfH1VErk64E3Ay/eWC8i9gH7AHafdGI03llsyrmggM4GXszuFARWmw3OCMel+i7DmhOolwsCrTuWC35XCurnHOmccL1cCPW3rHIIwCa4feou0B9gr6sTw/ulDHvmXK2a31ulAIwHtatx2buscD18/N0mqCTAUm/4c3XNvVZyBO0Yu3M6dc6pAPa7z/00Lf1cNfWLDnMNyxYd6wrOoOMyyZzZEXEtVchwJHWBrwDnbVLn+fXT10r6S+C+wIfbHntLDBSSlqkGiXdFxJ8DRMQtA++/HfjQnLqXJEniiantejoDuD4ivjSqkKTdwL8CTuae7/tTgde1OdjCDxSSBLyDKgfsWwbsx9X6BcDzgYntGU6SJJkYzTWKYyXtH3i9r14RcZwFvLtBm+cD3wQuA4bjxzRk4QcK4InATwKfl3R5bftV4IWSTqW6DDcCPz+f7iVJkpRpMaO4LSJO36yQpF3Ac2jmeX1CRDyzaQdKbDpQSLpfg3b6EfGNcTvjiIi/wbvFXTiN443CLWOWNApnd451bXDry1W7LtCdsRXqu7JODykt7/rEQ8Mnq1tYB+7Yss33WfTMP2LJEbJvEhK5oISrJT3ItLvLtFlKcGS1o6IjpHGOM46YJe1qyTpyOt2i+Wd1DneUnNVMfaslFL5HrcZQ+s5tWrbFsSbFFHY9PQv4zODy+wg+JenREfH5cQ7YZEZxU/0YNSx2gZPG6UiSJMl2Y0qxnl7IJstOkj5fH34JeJGkG6iWngRERPzzNgdsMlBcHRGP3aRTn21z0CRJkh1BQEww1pOkvcCPsvlS+7MndlCaDRQ/NKEySZIkO48JLj1FxF3A/RuUG7kbqi2bDhQRcXASZZIkSXYe04vjNEsa73qSdDrwKuC763qHtda1lZmzK/5MKQmctmzD6K+liLBthOum9fvhnQttv9xmgFLSNifym/olgdo6vJXE5IabDFyk4apfzTIyloR/72DarByUHO5M2ZKzm223EP21oXBdFM6n+V2+Db432myPfRfwn4DPg43VnCRJkgwyPYe7Rkh6AfCROk7erwGnAf9fRHymTTttBop/iogL2jSeJEmy45nv0tOvR8T7JT0JeAbwJuCtwA+2aaTNQPEaSX8AfJwBD79DITWSJEkSw3yXng6tS/5L4K0Rcb6k17ZtpM1A8SLg+4Bl7ll6CmBrDRRNL1rDcqVppbP3p+nVk9wL58QH3ulvzVzronPgDIUqpzGUnCYd1mGuxbdWSXsYLtjS3pA2DnO+vis8h2/t+Q4UX5H0NqottW+sYz+1/iJqM1A8JiIe3fYASZIkO5Zg3ktPPwE8E3hTRHxD0nFUWnMr2owsl0h6RNsDJEmS7GSqdKibP6bE3cCRVN7cUK0ItQ631GageBJwuaRr62RBn5d0RdsDJkmS7Cii4WM6/B7wBO4ZKO4AfrdtI22WnsaOQDh32lwQM110ukNJo3DJSlyCnZLd1e8VNA5b1rRZqu/Lmj4V1qx75qT6ZED+87tjtQsK2CLxkPnp1jc2V65qd3YsoqbVWLdowURWZsZISNaq/mGgCYbwOAx+MCJOOxRmKSK+XkefbUXjgWLSLuFJkiTbnunOFpqwVmfDCwBJD+Awfu+kZ3aSJMnU0LzF7N+hSpf6QElvAH4c+LW2jaRndpIkyTSZ44wiIt4l6TKq1KkCnhcRV7dtJz2zkyRJpsmcYz1FxDXANeO0seM8s4f8lQpzI6dlurjyTogFLzCv9704udbvDtuima2yD19Ga6NU3xzLZFdbM6Jx1a7LsGcC0pU22ZljOYG5DU5ML9mdGO/D7Hnh328mKGwcMAJ16R5qWr9ctvmGiqZsh0ioM2eOA0XtYPevgJMZ+L6PiNe1aWfLe2ZLeibw21RZ9v4gIn5jnv1JkiT5DjH3XU/nA98ELmPgB35btrRndq3m/y6Ve/oB4NOSLoiIv59vz5IkSWrmu/R0QkSM7dqw1T2zHw9cFxE3RMQq8B7guXPuU5IkyaLwKUlj/8BvM6N4EnC2pC8yRpLuCXM88OWB1wfYED5X0jnAOQDdY44Z2qpWmhY6PcI60RV0B6dH9EplzbrzeiuNoqnG4S+31ziGE++sFn5XLLvgde5nlGkTYNkmOZqORuF0FlfSBQoEv+7vzkvpWvVNWWcrteE0hnWjcYF32LP3cIvAllY6Kl2qpjEFC+Va+faNG+xzir/6p+Cj2IYnAS+SdANjfG9vdc/sTcNDRsQ+YB/A7hNP3Aa5ppIk2VJMaAOApKOBPwAeRfU99+KI+NtNqj1rEsfe6p7ZB4ATB16fANw0p74kSZLcm2CSXme/TZWt7sfrMBx7G9T5KmbXE9Bq19OmGoWkTVPmNSkzJT4NnCLpofWJOwtIX48kSRYGRbPHyDak7wKeArwDICJWI6JJFNjzqXTbdeDOgUcrmswovn+TKLEC7tv2wJMgItYlvRS4iGp77LkRcVW5gtmyXxrte2Z91ukWBY3D6RGrhbXk1d6wfaU3fGlWuv5yHewvN7Np2FayLxs9weoOQLfpwnVhBt6LYa8F54cB0DHJaGxQv0Jf10zZNVeu0NmVMfWgVVN2tVDW+Ww4n5uSb4TTvtpoFO4ebhMY016CVhqHabeFnqEZJpkaSfNuHCtp/8DrffXSOcDDgH8C/qekx1Btd31ZRGz2pT+RXU9NBorva1Cm5J80dSLiQuDCeR0/SZJkJM0Hitsi4vTCe0vAacC/jYhLJf028Arg1zdp81OSHh0Rn2/ci8LBR7Kg2kSSJMnC02RZqSEHgAMRcWn9+s+oBorNeBLw0+PuVm2z6ylJkiRpywR2PUXEVyV9WdLDI+JaqiB/TRyLZ7vrKUmSJGmPcTM6XP4t8K56484NVGGVRhIRX6o1jSfXpr+OiM+1PfBhDRSSXhoR/6N+fnRD9X3uiOFpoIxoDV64DlO21ys40Rn7mhGtAVb7Rrg2trt7PjHV3u7qkO2u/u4h27IJvleyd4yY3O0U5tDuH6EzLIaXssYtm0tQEs6brveWRDPnSOeE64MFh7mDRng+GGbjgLGB32SwYmwlu7sv3MYH8E6fTgwvBSW0urPbvFGKZWTadQ6uU3O4swL3BI7Vlgm1HRGXAyUNwyLpZcDPcU9Mvj+RtC8i/nubdg53RvHdA89fCfzKYbaTJEmyfZmcRnG4/AxVOtQ7ASS9EfhboNVAcbhJeTuSniypA9z/MNtIkiTZ/kTDx3QQ955g9ziMDOGHO6P4ZeAXgJ+mcuhIkiRJHPOdUfxP4FJJ59Wvnwec27aRwxooIqIH/I/DqTtXjMNdYdneahexbgL9FTQKp0eUEhe5NeaDxnZ3x69l32W0C6c7lDUK41znAv0VEhc5x7CeWYveUzq+00PG/O9yfQJYM05ozua0iMreTI9wWkRVdvhaFY/VUM9w/Qevfbl7cL2gnbl72znnWd0CvHblbKVLbcoWhWGnR7SoP0HBebjt+aZCfYukT1BtkxXwooj4bNt2DlfMfg+VQ2sAX42IXz6cdpIkSbY9881w98aI+BXgM8bWmMPVKC6JiJ+MiJ8C3niYbSRJkmxvGsZ5muKs40eNrbVvxeFqFM+R1AMuioh/OMw2kiRJtj9zmFFI+gXg3wAP2xCr7yjgf7dtr/VAIel1wI1UeVjPknRSRPxs23bmRWf93uupfZ9Lp4q1OGQzAdWMbgGwbuwra/507+oOr90f7A6vRe/q+DX+O3vNfCZKGkVnzAXanox203FaQHM/jpJG0TE/vdy6eUmjsIH2jM9EyQ/iTuOf4nxWXLmq7LBGcZe5fgB395r5URw05cAHm7TaWUFncxqF8yVyATSh4DPhdIOSTjimxmB/pRcDEE7x23w+S09/CnwY+K/cO9THHRFxe9vGWg8UEfFqSQ8CHgs8pG39JEmSnYJz8p0FEfFNqh/zL5xEe03yUZwt6TZJt0v6I0lHRcQtEfGRiPiNrTSbSJIkmSlRzXaaPCbJpPMINZlR/DqVIPIVqlgj/6X+myRJkmzGfJaeJppHqMlA8a2Bfbe/LunSkaWTJEmSe5jPQDHRPEJNBorjJJ0DXA1cA3jlbCsQsNG3rKAPYxK80WvohAfQWzdC4pKfXx5cH74My0bMXirMT5fMh1hygf6mtFi62hn+rHtiOG/cHrlccrDL9rUwF2/4EfoFJzQvZjcL9AfeCc4J13f099j63+4N253ADXC3sTuBuyRmO0dOJ3CXHEH7Vsw2ZQtitrM7R9ZWAnUbhzlb399A283hbtJ5hJoMFK8B/jnw/wKPBu4j6ULgc8AVEfHuSXYoSZJkW7EgGVmQ2xm6AAAaoUlEQVTHoUmGu32DryWdQDVwPBo4E5jaQCHpvwH/N7AKXE/lfv4NSSdTzXCurYteEhEvmVY/kiRJDovpBvybGYezPfYAVVq+WeSpvhh4ZUSs1+FxB0OaXx8Rp86gD0mSJIfNNJe1ZsVCZ7iLiI8OvLwE+PFx2pMLCmic6Cq7sxmNYq3g2LVkHM7WfPC1lc7wZeh2hu+uNhpFt8XCqHVYM2v8/a5fyz5o9IgjOyvD5eTX0q1GYaPH+WCFDtd/8HqEc7hbLSQucs51Ts9wWgT4AI53rjd3uLtr3TjhGY0LCg53RjtbNzZo4XBX0g2M/ucd7koOe83arOzD97vTH3dKUEBJ/37U+xHxljbtHW6sp3nwYipPw0M8VNJnJf2VpCeXKkk6R9J+SfvX77pz+r1MkiQZZD75KI6qH6dTpYQ4vn68BHhE28bmPqOQ9DHgweatV0XE+XWZV1EF1XhX/d7NwEkR8TVJjwM+KOmREfGtjY3UGss+gCOOO3EbrBYmSbJlmJNGERH/GUDSR4HTIuKO+vVrgfe3bW/uA0VEPG3U+5LOBp4NnBFRBWSJiBVgpX5+maTrge8F9k+5u0mSJI0Rh5FOblR70o3AHVQ+EOsRsVkO7ZOoNgMdYhU4ue1x5z5QjELSM6nE6/8rIu4asD8AuD0iepIeBpwC3DCnbiZJkhSZgv7xwxFxW8Oyfwz83UCGu+cDf9T2gAs9UFBl0dsNXCwJ7tkG+xTgdZLWqUbWlzSKiBjQ2SBSO8e6kt0J1yWBN9aG55u9rhcNVzvDZZ2YfWdBFdOYaplzTusZ+cqJvgB7u8PCtcvEtrtTcrgbPtklMXtcmorZrhz4bHQu+mvJic4J13cagRvgjjXT7rpxwjMCN8CqEbld9NhewWnUOpOaDR2lDSEbIzVXZYfLlQXq5mWbCtclB9uSI95EmG+GuzdI+jBVhjuAsyPi8rbtLPRAERHfU7B/APjAjLuTJEnSnuYDxbGSBpfP9230Y6tb+6iqX4dvM+8DIOmODUfWwHsREd/VuFcs+ECRJEmypWmXve62BprDEyPiJkkPpFppuSYiPjl02IijWvZ0JFtpe2ySJMnWY4LbYyPipvrvrcB5wOMn3l/DjptRbNQoShnuOlaPGC4nv2xPGO2iZ7QIAJlGnBNeSYtoGuzPOdYB9JbNZzV7NdZN1jrwWdd2bzzRwJ6CRuEy3JWCApYy322klOHO6TE2w53RWMB/VmdzAf0Avm00Bqc7lOwHjR7hgkoCrLjAlMbps6xROI1huGzRadXoAR0bKNBWH/pfHVXW6xnD94qzVceanpAwKYc7SUcCnYi4o37+dOB1m9QRVZy+h0XE6ySdBDw4Iv6uzbF33ECRJEkySya46+lBwHn1xp4l4E8j4iOb1Pk9Kt/5H6EaVO6g0nd/oM2Bc6BIkiSZFhN0uIuIG4DHtKz2gxFxmqTP1m18XZKfwo4gB4okSZJpMt94EGuq1rYDvuOD1nqOs6MGCgVsXCY3OXcAcDHlbNmOX58NYw8VkumYJtbG9o0wxy9oFG7dft0IMitdf7vs7g4vJu8yC8xOtwCvUTQN/tcWFyzQaRSlZD7OP8QF73M2gIMmmVBJo3B6xF1rRqNYKwQFXB22u4Ra/UKwShr6UTg9D0qBNYdthdvCaxwFLcHrIc7WvP4kEPMJCjjA71CJ3g+S9AbgBcCvtW1kRw0USZIkM2e+DnfvknQZcEZtek5EXNO2nRwokiRJpohi9iOFpL+JiCcNON6lw12SJMlCEvNJXBQRT6r/TsTxLh3ukiRJpskc8lFIurBOGT0RdtaMIqCzIVhfp1sQo42+54ICRrfgwLNqnNgKDmtOY14vefI1xAnXvYKYvW4E3lUjZh8siNl7jJjtBO5Shr5lozp2Cv85JUe8jTiBHrwjoTsvq8aJDmDNnRcjUJfquwB+pQx1zpHOCddOtAZYN2V7q+a8OBsgc787n8myw5yp30bMblHWidzO5tqs7NP72T8nMfsPqWJCvRP4zQiThrIFO2ugSJIkmTXzSVz0Pkl/Abwa2C/pjxnYFts2FWoOFEmSJNOiXVDASbMG3EmVquEoDsN/4hA5UCRJkkwJMR8xu0769hbgAqpUqHdtUmUkO2qgcA533UKgPudIZ5e9VUh06MyFon2zp8D1qrC8avUIq1EUnMic3TnXrSwV1tKtw93wwvVSIWvMkknSVNYoJu+I6DSaksOd0x5cMqAVo1tU9uGyq8YJrrKbYxnnOKdFQHM9oph4yOhszrmu5HBn9YyGTnhQ0h1KZU27DQMFQtkRbyLMYXss8CrgBRFx1SQa21EDRZIkyayZx9JTRDx5ku3lQJEkSTItprD1dR4stB+FpNdK+oqky+vHmQPvvVLSdZKulfSMefYzSZKkhPrNHovMVphR/FZEvGnQIOkRwFnAI4GHAB+T9L0RMTq0Vx+6q/ce3l3wvso+bDO5iMo47aKgUTisblH4ZbLeb6hR9JprFOtLw4u+q2Z9Hbyesdx1GkUhGZH5LylpEdPQKOznL/hhOD1izdUvnCunR6wXrotPMuSC+hVuTKdHmLIyWgSU9AhTbtUf3vpMuPprhWvtyhaCAlo9w7TbWfP3oKaYuChnFPPjucB7ImIlIr4IXMeMUgImSZI0JkD9aPRYZLbCQPFSSVdIOlfSMbXteODLA2UO1LYkSZKFQtHsscjMfaCQ9DFJV5rHc4G3Av8MOBW4GXjzoWqmKXuqJZ0jab+k/esrd07lMyRJkhSZQ6ynSTN3jSIintaknKS3Ax+qXx4AThx4+wTgpkL7+4B9AEfe/8QFvxxJkmwnFiBx0USY+0AxCknHRcTN9cvnA1fWzy8A/lTSW6jE7FOAv9u0PetwN6LwBnzWulL9YVOUnPMa3kilQHfeuc4EJVwuZLgzYur60rBtackLgStGpF4yYvZytyAkmnPdbfHf5QRuJ1qDP9VNBe7Kbhz2jHBdEqjdhoL1gsNd32SYCydcF8RsF9TPBvoriNmywnWzQH9VWWOzYnah/hSCAhYz5E0rKGDERB3u6rSm+4GvRMSzJ9bwJiz0QAH8pqRTqf6/bwR+HiAirpL0PuDvqRyWf3HTHU9JkiRzYMIzipcBVwOtEg+Ny0IPFBHxkyPeewPwhhl2J0mSpDWT8pGQdALwL6m+9/79ZFptxkIPFEmSJFuaAJpvfT1W0v6B1/tqjfUQ/z/wy1SRYGfKzhooYtjhruQE5/QEv27ewouugNx6ul1ML2gM5kYM8ysmeoXERUa76JuyvZKzk9Eeul0TKK8QgLFjNI7SWZ2Gw52zOd0HvMbQN9el5NzoroHTIgCvPZj6TneA5gH8ig5zNijgcLluqb4ra5zgugWNwpUtO+c1c64rOdx11qa4ct38lr0tIk53b0h6NnBrRFwm6akT6lljdtZAkSRJMmMmpFE8EXhOHcZoD/Bdkv4kIv71RFrfhLn7USRJkmxrDu182uwxsol4ZUScEBEnU4Uv+l+zGiQgZxRJkiTTIxY/4F8TdtRAoQi6q/e+atHxk6quTTzkAv2Vfgm00C5cE2aNXIVfHXI+E2aNPJZ9fbtubhxM+kuF4xuNYr1rEtuXNAp3DgvnteSK0hR7ChvqFgDhAjCa8xcFjcJpDNaGTyhkfSNKiYecH0JD3aFkd3pESeMY0gMLZdvpDi2CAq4a7augRUzLj6JyuJvs/tiI+ATwiYk2ugk7aqBIkiSZOTmjSJIkSUYx6RnFPMiBIkmSZFpsgYB/TciBIkmSZGpMNtbTvNhZA0UfOiv3XjDstlFHbdGSx15z57ymWq4TrQFkRGovcBcEWiNSW1vB4c5tCAgjZlMQs3vGXrws425Kt851plzhXFu7W4MuXSvnMFcQo2XEaFe2GCivaYa6Yn1jM2K0E61L9b3Dna9vxfCSw5wTro1ArVKGuyk63C16UqIm7KyBIkmSZJbk9tgkSZJkU3LpKUmSJBnJ1h8ndtZAoQi6G9coTTKi+o1mjZZughYOcy4Zjpuulqawdt3bLLmq4DAXS2bd3ngclhzuwuXdcXpE4ZSGKVtMCDUNrBOeL2p1Inf+i9fK2EoOcw3LljQGp3G0ShzUUGMoOtw5PcLoDt2V5g53Gx1m7+nD8Mlygf6KWsRa4SROgNwemyRJkowmB4okSZKkhCJQLweKJEmSZBQ5o5gukt4LPLx+eTTwjYg4VdLJVHljr63fuyQiXrJpgwGdlXH2S48Xld0mKAJ7I1mNohQ8btnYXP3CWniYu8DpEcX67rQY7cdpEZXdBVu0RcfPE9VQjyi6a4ytJzlb86B+rv74GkUp0N6wzQYFLPlBOI3B6BHF4xs9wtmqNozPxOrwB1BBiyjZJ0IOFNMlIv6fQ88lvRn45sDb10fEqbPvVZIkSUOCDAo4KyQJ+AngR+bdlyRJkjZsh11PWyXD3ZOBWyLiCwO2h0r6rKS/kvTkeXUsSZKkTEC/3+yxwMx9RiHpY8CDzVuviojz6+cvBN498N7NwEkR8TVJjwM+KOmREfEt0/45wDkAe3bdd7KdT5IkGUWQGsUkiIinjXpf0hLwY8DjBuqsACv188skXQ98L7DftL8P2Adw370Pic7BMUQrE9xLVsn19l5J4DT2nnWiK2R9s4HmTDkjegOEEz2Nw511rCvYrRNdwbnRnsKFFbNN2YYCN5TEbF/WOtxZJ7xCfStmGzG5RVBA63DXKqhfs0x0AF2z8cQ51kFJuHYOd4UPO00xe7EnC42Y+0DRgKcB10TEgUMGSQ8Abo+InqSHAacAN8yrg0mSJCW2g0axFQaKs7j3shPAU4DXSVoHesBLIuL2mfcsSZJkMyYwUEjaA3wS2E31vf1nEfGasRtuyMIPFBHx08b2AeADs+9NkiRJCwK7ZH0YrAA/EhHflrQM/I2kD0fEJZNofDMWfqCYKBFDa5mttn21WWtskXioaVBAVw68dtFxTnQFh7m+KWt1hzYahck8VJBz7EUoBgWchkbhDlO61va6Nq/vHSkLZRs613UKCaWsw12L+l2b5MhpDKWgfsMf1gX1KzrRGT3CaRElu1bMB1j1ERBjrRAZcWxiIjuaIiKAb9cvl+vHzNa0tsr22CRJkq1JRLMHHCtp/8DjnMFmJHUlXQ7cClwcEZfO6iPsrBlFkiTJLGm39HRbRJxebCqiB5wq6WjgPEmPiogrJ9DLTckZRZIkydQIiH6zR9MWI74BfAJ45rR6vZEcKJIkSaZJ86WnIpIeUM8kkHQEtdvADHoP7LSlpwh0sJCOawN2BHVTyILC6wXOgnOeES37LRyznBjdN1nrShnurPDdxuHORo9tWK5kLwXanULmO+tcV8xw18xW2vjQRszumE0KVqAu1XfCcwsxu6lwXaxvhWuXic6fLCtQl8RsJ1wbJ7qiaF0QuccmKHvatuM44J2SulT/Xe+LiA9NouEm7KyBIkmSZNZMwI8iIq4AHjt+Zw6PHCiSJEmmxubLSluBHCiSJEmmRbDwkWGbsLMGin4MrUUWl7zNr4COy0RXXMt2WetKesbwIr0N9FfIhOb0CK9b2OqE0SP6XRPUr1VQQGcrnG1jnprDnaNNUEDrcNcsQyEUgvq1KOt0i2KGu4ZlyxnuXFA/ozuslzQGU9YF6ivpDq6s0yLA6xGrRo8sOdytrPh2J0HOKJIkSZKR5ECRJEmSFIkgeoVtaVuIHCiSJEmmyWSCAs6VnTVQRBAr9163LMWKt3Z3wQvJhDpGj3Br2QDqmbLLzXQLgI7TKJaNrRAU0GkMTk4p+1E0CwDokhlBQY+Ys0ZR9KNoGBSwVN/6RoypURT9MIzGoIa6Q7l+M98IAJl22yQTsvY2Qf1MWatbjLBPhFx6SpIkSYrEZKLHzpscKJIkSaZJziiSJEmSUUTOKJIkSZIiEZOK9TRXdtZAEX3Y4FgThfC+Mr8CvMBdqB/Lw4cvCN/OMUq9YTW4Y2wA/a4Rvo1w3TGOdVBwuDMCeZuggNZWCvTXIiigrW+y6bXBXtcxxexyhrsWYrYra/Rdd/9AwTnPCczFDHkNHeaMrVTfOcaVxGwvRpcc7obFaFe2KGavF/owCVqEEF9UFiLMuKQXSLpKUl/S6Rvee6Wk6yRdK+kZA/Zn1rbrJL1i9r1OkiQZTQDRj0aPRWZRZhRXAj8GvG3QKOkRwFnAI4GHAB+T9L31278L/ChwAPi0pAsi4u9n1+UkSZJNiNgWM4qFGCgi4moADS8hPBd4T0SsAF+UdB3w+Pq96yLihrree+qyOVAkSbJQLPpsoQkLMVCM4HjgkoHXB2obwJc32H/QNVAnKD+UpHzlom+eO5Mcs5twLHBb9gFYjH4sQh9gMfqxCH2AxejHw8dt4A6+ftHH+u87tmHxeX/eIjMbKCR9DHiweetVEXF+qZqxBV5bscN2ROwD9tV92D8qefmsWIR+LEIfFqUfi9CHRenHIvRhUfohaf+4bUTEzPJaT5OZDRQR8bTDqHYAOHHg9QnATfXzkj1JkiSZIAux62kEFwBnSdot6aHAKcDfAZ8GTpH0UEm7qATvC+bYzyRJkm3LQmgUkp4P/HfgAcBfSLo8Ip4REVdJeh+VSL0O/GJE9Oo6LwUuArrAuRFxVYND7ZvOJ2jNIvRjEfoAi9GPRegDLEY/FqEPsBj9WIQ+LASKbRCHJEmSJJkei770lCRJksyZHCiSJEmSkWy7geJwwoFsKPNQSZdK+oKk99Zi+bh9eq+ky+vHjZIuL5S7UdLn63Jjb83b0PZrJX1loB9nFspNNTSKpP8m6RpJV0g6T9LRhXITPxebfbZ608R76/cvlXTyJI470P6Jkv5S0tX1PfoyU+apkr45cJ1ePck+DBxn5PlVxe/U5+IKSadN+PgPH/iMl0v6lqSXbygzlXMh6VxJt0q6csB2P0kX1//3F0s6plD37LrMFySdPYn+bAkiYls9gO+ncpT5BHD6gP0RwOeA3cBDgeuBrqn/PuCs+vnvA78w4f69GXh14b0bgWOndF5eC/zHTcp06/PyMGBXfb4eMeF+PB1Yqp+/EXjjLM5Fk88G/Bvg9+vnZwHvnfBnPw44rX5+FPAPpg9PBT40jXugzfkFzgQ+TOXL9ATg0in2pQt8FfjuWZwL4CnAacCVA7bfBF5RP3+Fuy+B+wE31H+PqZ8fM+1rtQiPbTejiIirI+Ja89Z3woFExBeBwXAgQPUrCvgR4M9q0zuB502qb3X7PwG8e1JtTpjHU4dGiYhV4FBolIkRER+NiEOhOi+h8oGZBU0+23OprjlU98AZMnFlDpeIuDkiPlM/vwO4mnsiDSwazwX+KCouAY6WdNyUjnUGcH1EfGlK7d+LiPgkcPsG8+C1L/3fPwO4OCJuj4ivAxcD28KhbjO23UAxguMZDvux8Z/0/sA3Br7IXJlxeDJwS0R8ofB+AB+VdFkdemTSvLReRji3MLVuco4myYupfrU6Jn0umny275Sp74FvUt0TE6de1noscKl5+4ckfU7ShyU9chrHZ/PzO8t74SzKP55mcS4AHhQRN0M1oAMPNGVm/f+xMCyEH0VbNNlwIG3LjNOnFzJ6NvHEiLhJ0gOBiyVdU//6acSoPgBvBV5P9XleT7UE9uKNTZi6rfdPNzkXkl5F5RvzrkIzY50L1y1jm9j1b9UR6T7AB4CXR8S3Nrz9GaolmG/XOtIHqRxNJ81m53dW52IX8BzglebtWZ2LpszknCwiW3KgiMmHAznEbVRT7KX6F2Xj0CCb9UnSElUo9ceNaOOm+u+tks6jWi5p/OXY9LxIejvwIfNWk3M0dj9qEfDZwBlRL/6aNsY6F4Ymn+1QmQP19bovw0sUYyFpmWqQeFdE/PnG9wcHjoi4UNLvSTo2IiYaMK7B+Z3IvdCAZwGfiYhbTB9nci5qbpF0XETcXC+x3WrKHKDSTQ5xApUWuu3ZSUtPpXAg36H+0vpL4Mdr09lAaYbSlqcB10TEAfempCMlHXXoOZXoO7FItxvWl59faHvqoVEkPRP4FeA5EXFXocw0zkWTz3YB1TWH6h74X6WB7HCo9Y53AFdHxFsKZR58SBeR9Hiq/9GvTaoPdbtNzu8FwE/Vu5+eAHzz0NLMhCnOsmdxLgYYvPal//uLgKdLOqZeun16bdv+zFtNn/SD6kvwALAC3AJcNPDeq6h2vlwLPGvAfiHwkPr5w6gGkOuA9wO7J9SvPwRessH2EODCgeN+rn5cRbVMM8nz8sfA54ErqP4pjtvYh/r1mVS7ca6fdB/q9q+jWue9vH78/sZ+TOtcuM8GvI5q0ALYU1/z6+p74GET/uxPolqquGLg858JvOTQvQG8tP7Mn6MS+//FFK6BPb8b+iGq5GDX1/fN6VPox16qL/77Dtimfi6oBqabgbX6u+JnqLSojwNfqP/ery57OvAHA3VfXN8f1wEvmvQ5WdRHhvBIkiRJRrKTlp6SJEmSwyAHiiRJkmQkOVAkSZIkI8mBIkmSJBlJDhRJkiTJSHKgSJIkSUaSA0Wy5ZH0cwOhqPsDz61T2yZtnSzpbg2Egpf07YHnZ9Yhpk+SdER9nFVJx07q8yTJopF+FMm2QdLxwKci4rvHaONkqtDWjxqwfTsi7iPpDKo8yk+PiOsH3r+RyiFtGqElkmTubMlYT0lS4FFUXsQTR9KTgbcDZw4OEkmyE8iBItlOPJoJxscaYDdV7J+nRsQ1U2g/SRaa1CiS7cTQjELS6yfQ7hrwKaqYQEmy48iBItlO3GtGIenBwJKkE1Tlqv53qnJiL0l6i6Q3S/qlBu32qTIT/oCkX51S35NkYcmlp2RbIKlDFTp+cGnosVQRWh8DfDAiflvSu4BfAM6PiL9q2n5E3CXp2cBfS7olIt4xwe4nyUKTM4pku/A9wIGIWBmwnco9A8WhvAEBnAb877YHiIjbqXIk/5qkieYST5JFJmcUybYgIv4BeMQG8/dQ5Rf4HuAfal+HrwJ/DbxN0u3Af60HgFFt32fg+ZeBh06y70my6KQfRZIMIOlEKuH6axFx6iZljwD+FngA8OjNBpwk2arkQJEkSZKMJDWKJEmSZCQ5UCRJkiQjyYEiSZIkGUkOFEmSJMlIcqBIkiRJRpIDRZIkSTKSHCiSJEmSkeRAkSRJkozk/wCtDhOKbFNf9AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a2213fa20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "\n",
    "im = imshow(np.flip(mean_yield_anom.T, axis=0),cmap=\"viridis\", \n",
    "            extent=[T_incs_2[0],T_incs_2[-1], P_incs_2[0], P_incs_2[-1]], \n",
    "            aspect=\"auto\") # drawing the function\n",
    "# adding the Contour lines with labels\n",
    "#cset = contour(Z,arange(-1,1.5,0.2),linewidths=2,cmap=cm.Set2)\n",
    "#clabel(cset,inline=True,fmt='%1.1f',fontsize=10)\n",
    "cbar = fig.colorbar(im) # adding the colobar on the right\n",
    "cbar.set_label('Yield [tonnes ha$^{-1}$]')\n",
    "# latex fashion title\n",
    "#title('$z=(1-x^2+y^3) e^{-(x^2+y^2)/2}$')\n",
    "\n",
    "ax.set_xlabel('$T_{inc}$ [K]')\n",
    "ax.set_ylabel('$P_{inc}$ [mm]')\n",
    "\n",
    "plt.savefig('./figs/temp_precip_impact.pdf')\n",
    "plt.savefig('./figs/temp_precip_impact.png')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9.998284200856212"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.max(mean_yield_anom.T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9.751486932269136"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_yield_anom.T[int(len(T_incs_2)/2), int(len(P_incs_2)/2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8.776338239042223"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "0.9 * mean_yield_anom.T[int(len(T_incs_2)/2), int(len(P_incs_2)/2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a245e02e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "\n",
    "im = imshow(np.flip(mean_yield_anom.T, axis=0),cmap=\"viridis\", \n",
    "            extent=[T_incs_2[0],T_incs_2[-1], P_incs_2[0], P_incs_2[-1]], \n",
    "            aspect=\"auto\") # drawing the function\n",
    "# adding the Contour lines with labels\n",
    "#cset = contour(Z,arange(-1,1.5,0.2),linewidths=2,cmap=cm.Set2)\n",
    "#clabel(cset,inline=True,fmt='%1.1f',fontsize=10)\n",
    "cbar = fig.colorbar(im) # adding the colobar on the right\n",
    "cbar.set_label('Yield [tonnes ha$^{-1}$]')\n",
    "# latex fashion title\n",
    "#title('$z=(1-x^2+y^3) e^{-(x^2+y^2)/2}$')\n",
    "\n",
    "ax.set_xlabel('$T_{inc}$ [K]')\n",
    "ax.set_ylabel('$P_{inc}$ [mm]')\n",
    "\n",
    "CS = ax.contour(T_incs_2, P_incs_2, mean_yield_anom.T, [4., \n",
    "               6., \n",
    "               8., \n",
    "               0.9 * mean_yield_anom.T[int(len(T_incs_2)/2), int(len(P_incs_2)/2)]\n",
    "              ], colors='w')\n",
    "\n",
    "fmt = {}\n",
    "strs = ['4', '6', '8', '-10% (8.8)']\n",
    "for l, s in zip(CS.levels, strs):\n",
    "    fmt[l] = s\n",
    "ax.clabel(CS, \n",
    "          inline=1, \n",
    "          fmt=fmt, \n",
    "          fontsize=10)\n",
    "\n",
    "plt.savefig('./figs/temp_precip_impact_contour.pdf')\n",
    "plt.savefig('./figs/temp_precip_impact_contour.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot posteriors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "samples = samples2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a246f3400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df=pd.DataFrame(np.random.multivariate_normal(np.array([20,5]),np.array([[5.0,0.0],[0.0,1.0]]),2000),\n",
    "                columns=['$\\mu_T$','$\\sigma_T$'])\n",
    "g=sns.PairGrid(data=df,size=2.5,diag_sharey=False)\n",
    "g.map_diag(plt.hist,color='Red',alpha=0.5)\n",
    "g.map_lower(sns.kdeplot, cmap=\"Reds\",alpha=0.8,n_levels=10,normed=True, shade=True,shade_lowest=False)\n",
    "df=pd.DataFrame(np.vstack((samples['mu_t'],samples['sigma_t'])).T,columns=['$\\mu_T$','$\\sigma_T$'])\n",
    "g.data=df\n",
    "g.map_diag(plt.hist,color='Blue',alpha=0.5)\n",
    "g.map_lower(sns.kdeplot, cmap=\"Blues\",alpha=0.8,n_levels=10,normed=True, shade=True,shade_lowest=False)\n",
    "\n",
    "g.axes[0,1].set_axis_off()\n",
    "\n",
    "#fig = g.get_figure()\n",
    "plt.savefig(\"./figs/2d_Gauss_prior_vs_post_temp_mean_vs_sigma.png\")\n",
    "plt.savefig(\"./figs/2d_Gauss_prior_vs_post_temp_mean_vs_sigma.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a24b3ce48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df=pd.DataFrame(np.random.multivariate_normal(np.array([100,25]),np.array([[25.0,0.0],[0.0,5.0]]),2000),\n",
    "                columns=['$\\mu_p$','$\\sigma_p$'])\n",
    "g=sns.PairGrid(data=df,size=2.5,diag_sharey=False)\n",
    "g.map_diag(plt.hist,color='Red',alpha=0.5)\n",
    "g.map_lower(sns.kdeplot, cmap=\"Reds\",alpha=0.8,n_levels=10,normed=True, shade=True,shade_lowest=False)\n",
    "df=pd.DataFrame(np.vstack((samples['mu_p'],samples['sigma_p'])).T,columns=['$\\mu_p$','$\\sigma_p$'])\n",
    "\n",
    "g.data=df\n",
    "g.map_diag(plt.hist,color='Blue',alpha=0.5)\n",
    "g.map_lower(sns.kdeplot, cmap=\"Blues\",alpha=0.8,n_levels=10,normed=True, shade=True,shade_lowest=False)\n",
    "\n",
    "g.axes[0,1].set_axis_off()\n",
    "\n",
    "#fig = g.get_figure()\n",
    "plt.savefig(\"./figs/2d_Gauss_prior_vs_post_precip_mean_vs_sigma.png\")\n",
    "plt.savefig(\"./figs/2d_Gauss_prior_vs_post_precip_mean_vs_sigma.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pvals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Bayesian_Pval(yields,pred_yields):\n",
    "    import scipy.stats as st\n",
    "    Pvals=np.empty_like(yields)\n",
    "    n_reg,n_years=yields.shape\n",
    "    for r in range(0,n_reg):\n",
    "        for y in range(0,n_years):\n",
    "            ind=pred_yields[:,r,y]<yields[r,y]\n",
    "            Pvals[r,y]=st.norm.ppf(ind.sum()/pred_yields[:,r,y].size)\n",
    "    return Pvals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "Pvals=Bayesian_Pval(fit2.data['d_yields'],samples2['pred_yields'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a2212a048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.figure(figsize=(10,5))\n",
    "for s in range(0,len(states)):\n",
    "    plt.plot(Pvals[s,:],label=states[s])\n",
    "plt.xticks(5*np.arange(len(np.arange(1980,2015, 5))),np.arange(1980,2015, 5),rotation=90);\n",
    "plt.legend()\n",
    "plt.ylabel('P value')\n",
    "\n",
    "plt.savefig(\"./figs/pvalues_all.png\".format(states[s]))\n",
    "plt.savefig(\"./figs/pvalues_all.pdf\".format(states[s]))"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python (herschelhelp_internal)",
   "language": "python",
   "name": "helpint"
  },
  "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
}