mapping_main.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pickle\n",
    "import time\n",
    "import numpy as np\n",
    "import matplotlib.pylab as plt\n",
    "%matplotlib inline\n",
    "import scipy.stats as ss\n",
    "import matplotlib.style as style\n",
    "style.use('ggplot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "host = 'SYNPAI' #'habanero'\n",
    "if host == 'habanero':\n",
    "    resultdir ='/rigel/issa/users/Tahereh/Results/'\n",
    "    neuraldir = '/rigel/issa/users/Tahereh/models/neural_features/'\n",
    "    modeldir = '/rigel/issa/users/Tahereh/models/'\n",
    "    datadir = '/rigel/issa/users/Tahereh/Data/DiCarlo/'\n",
    "elif host =='Mac':\n",
    "    resultdir ='/Users/tahereh/Documents/Results/invertibility/error coding in visual cortex/'\n",
    "    neuraldir = '/Users/tahereh/Documents/Results/invertibility/error coding in visual cortex/models/neural_features/'\n",
    "    modeldir ='/Users/tahereh/Documents/Results/invertibility/error coding in visual cortex/models/'\n",
    "    datadir = '/Users/Tahereh/Documents/Data/DiCarlo'\n",
    "elif host == 'SYNPAI':\n",
    "    resultdir = '/home/tahereh/Documents/Research/Results/Neural-Dynamics/'\n",
    "    neuralfeaturesdir = '/home/tahereh/Documents/Research/features/neural_features/'\n",
    "    modeldir =  '/home/tahereh/Documents/Research/features/'\n",
    "    datadir =  '/home/tahereh/Documents/Research/Data/DiCarlo/'\n",
    "    # %pylab inline\n",
    "    # pylab.rcParams['xtick.color'] = 'white'\n",
    "    # pylab.rcParams['ytick.color'] = 'white'\n",
    "    # # pylab.rcParams['figsize'] = [6,4]\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n  from ._conv import register_converters as _register_converters\n"
     ]
    }
   ],
   "source": [
    "import MappingV36\n",
    "from MappingV36 import MappingV36 as Mapping\n",
    "from ReadMeta import ReadMeta\n",
    "from ReadData import ReadData\n",
    "from ReadModel import ReadModel\n",
    "\n",
    "import pickle\n",
    "import h5py\n",
    "import numpy as np\n",
    "import matplotlib.pylab as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import importlib\n",
    "Mapping = Mapping()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['PLS', 'ridge', 'ridge']"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg_methods"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "ename": "FileNotFoundError",
     "evalue": "[Errno 2] No such file or directory: \"/home/tahereh/Documents/Research/Results/Neural-Dynamics/MappingMtoD_['PLS', 'ridge']_[5, [20, -10, 10]]_[-1, 5], ni5120_nf168_nt46_SBFalse_5cmp_masterplot.pickle\"",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mFileNotFoundError\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-34-aac34356e380>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     50\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     51\u001b[0m     file = open(resultdir + 'MappingMtoD_%s_%s_%s, ni%d_nf%d_nt%d_SB%s_%dcmp_%s.pickle' % (\n\u001b[0;32m---> 52\u001b[0;31m               reg_methods, reg_params_list, PCA_ncomponents_list, ni, nf, nt,  spearman_brown, nc, purpose_of_this_run), 'rb')\n\u001b[0m\u001b[1;32m     53\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     54\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: \"/home/tahereh/Documents/Research/Results/Neural-Dynamics/MappingMtoD_['PLS', 'ridge']_[5, [20, -10, 10]]_[-1, 5], ni5120_nf168_nt46_SBFalse_5cmp_masterplot.pickle\""
     ],
     "output_type": "error"
    }
   ],
   "source": [
    "import scipy.stats as ss\n",
    "data_unit_indices = range(168)\n",
    "purpose_of_this_run = 'masterplot' # 'sweepPLSridgesitefit_%dsites' % len(data_unit_indices) #'3regswithridgesitefits_168sites'\n",
    "\n",
    "ni = 5120\n",
    "nf = 168\n",
    "nt = 46\n",
    "reg_methods = ['PLS', 'ridge']\n",
    "n_components_range = [5, 12, 25, 35, 50,80,100,128, 168]#np.arange(nf,1,-1)\n",
    "\n",
    "VEarraysite = np.zeros((len(n_components_range), len(reg_methods)))\n",
    "Narraysite = np.zeros_like(VEarraysite)\n",
    "Larraysite = np.zeros_like(VEarraysite)\n",
    "Rarraysite = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraypop = np.zeros_like(VEarraysite)\n",
    "Narraypop = np.zeros_like(VEarraysite)\n",
    "Larraypop = np.zeros_like(VEarraysite)\n",
    "Rarraypop = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraypopStd = np.zeros_like(VEarraysite)\n",
    "NarraypopStd = np.zeros_like(VEarraysite)\n",
    "LarraypopStd = np.zeros_like(VEarraysite)\n",
    "RarraypopStd = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraysiteStd = np.zeros_like(VEarraysite)\n",
    "NarraysiteStd = np.zeros_like(VEarraysite)\n",
    "LarraysiteStd = np.zeros_like(VEarraysite)\n",
    "RarraysiteStd = np.zeros_like(VEarraysite)\n",
    "\n",
    "\n",
    "for inc, nc in enumerate(n_components_range):\n",
    "    trainfraci = 0.8  # image trainfrac\n",
    "    splitfract = 0.5  # trial splitfrac\n",
    "    nfoldi = 2\n",
    "    nfoldt = 2\n",
    "    \n",
    "    spearman_brown = False\n",
    "    \n",
    "    reg_params_list = [nc, [20, -10, 10]]\n",
    "    report_popfit = [True, True]  # [False, True, True]\n",
    "    report_sitefit = [False, False]  # [True, True, True]\n",
    "    PCA_ncomponents_list = [ -1,  nc]\n",
    "\n",
    "    # file = open(resultdir+'unit_test_%s_ni%d_nf%d_nt%d_collilnearity%s_%s_SB%s_VariousSd%s_noisymap%s.pickle' % (\n",
    "    #     reg_methods, ni, nf, nt, Collinearity, noise_dist, spearman_brown, various_unit_stds, noisy_map), 'rb')\n",
    "    \n",
    "    # file = open(resultdir+'unit_test_%s_%s_%s, ni%d_nf%d_nt%d_collinearity%s_%s_SB%s_VariousSd%s_noisymap%s_statsfromHvM%s_%dcmp.pickle' % (\n",
    "    #     reg_methods, reg_params_list, PCA_ncomponents_list , ni, nf, nt, Collinearity, noise_dist, spearman_brown, various_unit_stds, noisy_map,stats_from_data, nc), 'rb')\n",
    "    \n",
    "    file = open(resultdir + 'MappingMtoD_%s_%s_%s, ni%d_nf%d_nt%d_SB%s_%dcmp_%s.pickle' % (\n",
    "              reg_methods, reg_params_list, PCA_ncomponents_list, ni, nf, nt,  spearman_brown, nc, purpose_of_this_run), 'rb')\n",
    "\n",
    "    \n",
    "    data_list = pickle.load(file)\n",
    "    file.close()\n",
    "    r12_reg,r11_reg,r22_reg,r12_reg_sitfit,r11_reg_sitfit,r22_reg_sitfit = data_list\n",
    "    \n",
    "    for ir, r in enumerate(reg_methods):\n",
    "        \n",
    "        VEpop = np.nanmedian(r12_reg[:,ir]/np.sqrt(r11_reg[:, ir]*r22_reg[:, ir]))\n",
    "        Npop = np.median(r12_reg[:, ir])\n",
    "        Lpop = np.median(r11_reg[:, ir])\n",
    "        Rpop = np.median(r22_reg[:, ir])\n",
    "        \n",
    "        VEarraypop[inc, ir] = VEpop\n",
    "        Narraypop[inc, ir] = Npop\n",
    "        Larraypop[inc, ir] = Lpop\n",
    "        Rarraypop[inc, ir] = Rpop\n",
    "        \n",
    "        VEarraypopStd[inc, ir] = ss.sem(r12_reg[:, ir]/np.sqrt(r11_reg[:, ir]*r22_reg[:, ir]), axis=None)\n",
    "        NarraypopStd[inc, ir] = ss.sem(r12_reg[:, ir], axis=None)\n",
    "        LarraypopStd[inc, ir] = ss.sem(r11_reg[:, ir], axis=None)\n",
    "        RarraypopStd[inc, ir] = ss.sem(r22_reg[:, ir], axis=None)\n",
    "        \n",
    "        VEsite = np.nanmedian(r12_reg_sitfit[:,ir]/np.sqrt(r11_reg_sitfit[:, ir]*r22_reg_sitfit[:, ir]))\n",
    "        Nsite = np.median(r12_reg_sitfit[:, ir])\n",
    "        Lsite = np.median(r11_reg_sitfit[:, ir])\n",
    "        Rsite = np.median(r22_reg_sitfit[:, ir])\n",
    "        \n",
    "        VEarraysite[inc, ir] = VEsite\n",
    "        Narraysite[inc, ir] = Nsite\n",
    "        Larraysite[inc, ir] = Lsite\n",
    "        Rarraysite[inc, ir] = Rsite\n",
    "        \n",
    "        VEarraysiteStd[inc, ir] = ss.sem(r12_reg_sitfit[:, ir]/np.sqrt(r11_reg_sitfit[:, ir]*r22_reg_sitfit[:, ir]), axis=None)\n",
    "        NarraysiteStd[inc, ir] = ss.sem(r12_reg_sitfit[:, ir], axis=None)\n",
    "        LarraysiteStd[inc, ir] = ss.sem(r11_reg_sitfit[:, ir], axis=None)\n",
    "        RarraysiteStd[inc, ir] = ss.sem(r22_reg_sitfit[:, ir], axis=None)\n",
    "\n",
    "\n",
    "# Plot figures\n",
    "\n",
    "\n",
    "reg_methods_labels = ['PLS','PCR']\n",
    "colors = ['gray','DarkRed']            \n",
    "n_regs = len(reg_methods)\n",
    "fig, axes = plt.subplots(nrows=3, ncols=4, figsize=[20, 14])\n",
    "\n",
    "for iax, ax in enumerate(np.ravel(axes).tolist()):\n",
    "    ax.set_ylim([-0.0, 1.02])\n",
    "\n",
    "\n",
    "for iax, ax in enumerate(axes[0,:len(reg_methods)].tolist()):\n",
    "    \n",
    "    ax.errorbar(n_components_range, VEarraypop[:,iax], yerr=VEarraypopStd[:,iax], marker='o', color='r', label='VE')\n",
    "    ax.errorbar(n_components_range, Narraypop[:,iax], yerr=NarraypopStd[:,iax], marker='o', color='salmon', label='N')\n",
    "    ax.errorbar(n_components_range, Larraypop[:,iax], yerr=LarraypopStd[:,iax], marker='o', color='b', label='L')\n",
    "    ax.errorbar(n_components_range, Rarraypop[:,iax], yerr=RarraypopStd[:,iax], marker='o', color='g', label='R')\n",
    "    ax.set_title(reg_methods_labels[iax]+ ' popfit')\n",
    "\n",
    "for iax, ax in enumerate(axes[1, :len(reg_methods)].tolist()):\n",
    "    \n",
    "    ax.errorbar(n_components_range, VEarraysite[:,iax], yerr=VEarraysiteStd[:,iax], marker='o', color='r', ls='--', label='VE')\n",
    "    ax.errorbar(n_components_range, Narraysite[:,iax], yerr=NarraysiteStd[:,iax], marker='o', color='salmon', ls='--', label='N')\n",
    "    ax.errorbar(n_components_range, Larraysite[:,iax], yerr=LarraysiteStd[:,iax], marker='o', color='b', ls='--', label='L')\n",
    "    ax.errorbar(n_components_range, Rarraysite[:,iax], yerr=RarraysiteStd[:,iax], marker='o', color='g', ls='--', label='R')\n",
    "    ax.set_title(reg_methods_labels[iax] + ' sitefit')\n",
    "    # ax.set_xlabel('ni/nf')\n",
    "\n",
    "axes[0,0].legend()\n",
    "axes[0,1].legend()\n",
    "\n",
    "#         \n",
    "axes[2,0].set_title('VE')\n",
    "\n",
    "#[axes[2,0].errorbar(n_components_range, VEarraysite[itrace, :], yerr=VEarraysiteStd[itrace, :], marker='o', ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [1]]\n",
    "if nc == 2:\n",
    "    axes[2,0].legend()\n",
    "[axes[2,0].errorbar(n_components_range, VEarraypop[:, itrace], yerr=VEarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "axes[2,1].set_title('N')\n",
    "\n",
    "#[axes[2,1].errorbar(n_components_range, Narraysite[itrace, :], yerr= NarraysiteStd[itrace, :], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [1]]\n",
    "\n",
    "[axes[2,1].errorbar(n_components_range, Narraypop[:, itrace], yerr=NarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "\n",
    "axes[2,2].set_title('L')\n",
    "\n",
    "#[axes[2,2].errorbar(n_components_range, Larraysite[itrace, :], yerr=LarraysiteStd[itrace, :], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [1]]\n",
    "\n",
    "[axes[2,2].errorbar(n_components_range, Larraypop[:, itrace], yerr=LarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "axes[2,3].set_title('R')\n",
    "\n",
    "#[axes[2,3].errorbar(n_components_range, Rarraysite[itrace, :], yerr=RarraysiteStd[itrace, :], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [1]]\n",
    "\n",
    "[axes[2,3].errorbar(n_components_range, Rarraypop[:, itrace], yerr=RarraypopStd[:, itrace],  marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "[ax.set_xlabel('# of components') for ax in axes[2, :]]\n",
    "[ax.legend() for ax in axes[2, :]]\n",
    "\n",
    "\n",
    "title = 'Model %s, Data %s'%('conv5', 'IT' )\n",
    "axes[0,0].text(1,1.2, title, fontsize=15)\n",
    "\n",
    "fig.savefig(resultdir+'MappingMtoD_SweepNumComponents_%s_%s_SB%s_%s.png'%(reg_methods, reg_params_list, spearman_brown, purpose_of_this_run), dpi=300)\n",
    "fig.savefig(resultdir+'MappingMtoD_SweepNumComponents_%s_%s_SB%s_%s.pdf'%(reg_methods, reg_params_list, spearman_brown, purpose_of_this_run), dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:81: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:86: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:91: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:96: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:103: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:113: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:118: RuntimeWarning: invalid value encountered in true_divide\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/numpy/lib/function_base.py:4033: RuntimeWarning: All-NaN slice encountered\n  r = func(a, **kwargs)\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:128: RuntimeWarning: invalid value encountered in true_divide\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:118: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:128: RuntimeWarning: invalid value encountered in sqrt\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import scipy.stats as ss\n",
    "data_unit_indices = range(168)\n",
    "purpose_of_this_run = 'testpinvsite' # 'testMPCA'  #'sweepPLSridgesitefit_%dsites' % len(data_unit_indices) #'3regswithridgesitefits_168sites'\n",
    "\n",
    "ni = 5120\n",
    "nf = 168\n",
    "nt = 46\n",
    "reg_methods = ['PLS', 'ridge','ridge']\n",
    "n_components_range = [5, 12, 25, 35, 50,80,100,128, 168]#np.arange(nf,1,-1)\n",
    "\n",
    "VE = np.zeros((len(n_components_range)))\n",
    "N = np.zeros_like(VE)\n",
    "L = np.zeros_like(VE)\n",
    "R = np.zeros_like(VE)\n",
    "\n",
    "VEStd = np.zeros((len(n_components_range)))\n",
    "NStd = np.zeros_like(VE)\n",
    "LStd = np.zeros_like(VE)\n",
    "RStd = np.zeros_like(VE)\n",
    "\n",
    "\n",
    "\n",
    "VE_site = np.zeros((len(n_components_range)))\n",
    "N_site = np.zeros_like(VE)\n",
    "L_site = np.zeros_like(VE)\n",
    "R_site = np.zeros_like(VE)\n",
    "\n",
    "VEStd_site = np.zeros((len(n_components_range)))\n",
    "NStd_site = np.zeros_like(VE)\n",
    "LStd_site = np.zeros_like(VE)\n",
    "RStd_site = np.zeros_like(VE)\n",
    "\n",
    "VEarraysite = np.zeros((len(n_components_range), len(reg_methods)))\n",
    "Narraysite = np.zeros_like(VEarraysite)\n",
    "Larraysite = np.zeros_like(VEarraysite)\n",
    "Rarraysite = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraypop = np.zeros_like(VEarraysite)\n",
    "Narraypop = np.zeros_like(VEarraysite)\n",
    "Larraypop = np.zeros_like(VEarraysite)\n",
    "Rarraypop = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraypopStd = np.zeros_like(VEarraysite)\n",
    "NarraypopStd = np.zeros_like(VEarraysite)\n",
    "LarraypopStd = np.zeros_like(VEarraysite)\n",
    "RarraypopStd = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraysiteStd = np.zeros_like(VEarraysite)\n",
    "NarraysiteStd = np.zeros_like(VEarraysite)\n",
    "LarraysiteStd = np.zeros_like(VEarraysite)\n",
    "RarraysiteStd = np.zeros_like(VEarraysite)\n",
    "\n",
    "\n",
    "for inc, nc in enumerate(n_components_range):\n",
    "    trainfraci = 0.8  # image trainfrac\n",
    "    splitfract = 0.5  # trial splitfrac\n",
    "    nfoldi = 2\n",
    "    nfoldt = 2\n",
    "    \n",
    "    spearman_brown = False\n",
    "    \n",
    "    reg_params_list = [nc, [20, -10, 10], [20, -10, 10]]\n",
    "    report_popfit = [True, True, True]  # [False, True, True]\n",
    "    report_sitefit = [False, False, False]  # [True, True, True]\n",
    "    PCA_ncomponents_list = [ -1,  nc, -1]\n",
    "\n",
    "    # file = open(resultdir+'unit_test_%s_ni%d_nf%d_nt%d_collilnearity%s_%s_SB%s_VariousSd%s_noisymap%s.pickle' % (\n",
    "    #     reg_methods, ni, nf, nt, Collinearity, noise_dist, spearman_brown, various_unit_stds, noisy_map), 'rb')\n",
    "    \n",
    "    # file = open(resultdir+'unit_test_%s_%s_%s, ni%d_nf%d_nt%d_collinearity%s_%s_SB%s_VariousSd%s_noisymap%s_statsfromHvM%s_%dcmp.pickle' % (\n",
    "    #     reg_methods, reg_params_list, PCA_ncomponents_list , ni, nf, nt, Collinearity, noise_dist, spearman_brown, various_unit_stds, noisy_map,stats_from_data, nc), 'rb')\n",
    "    \n",
    "    file = open(resultdir + 'MappingMtoD_%s_%s_%s, ni%d_nf%d_nt%d_SB%s_%dcmp_%s.pickle' % (\n",
    "              reg_methods, reg_params_list, PCA_ncomponents_list, ni, nf, nt,  spearman_brown, nc, purpose_of_this_run), 'rb')\n",
    "\n",
    "    \n",
    "    data_list = pickle.load(file)\n",
    "    file.close()\n",
    "    r12,r11,r22,r12_sitefit,r11_sitefit,r22_sitefit,r12_reg,r11_reg,r22_reg,r12_reg_sitfit,r11_reg_sitfit,r22_reg_sitfit = data_list\n",
    "    \n",
    "    VE[inc] = np.nanmedian(r12/np.sqrt(r11*r22))\n",
    "    N[inc] = np.median(r12)\n",
    "    L[inc] = np.median(r11)\n",
    "    R[inc] = np.median(r22)\n",
    "    \n",
    "    VEStd[inc] = ss.sem(r12/np.sqrt(r11*r22), axis=None)\n",
    "    NStd[inc] = ss.sem(r12, axis=None)\n",
    "    LStd[inc] = ss.sem(r11, axis=None)\n",
    "    RStd[inc] = ss.sem(r22, axis=None)\n",
    "    \n",
    "    VE_site[inc] = np.nanmedian(r12_sitefit/np.sqrt(r11_sitefit*r22_sitefit))\n",
    "    N_site[inc] = np.median(r12_sitefit)\n",
    "    L_site[inc] = np.median(r11_sitefit)\n",
    "    R_site[inc] = np.median(r22_sitefit)\n",
    "    \n",
    "    VEStd_site[inc] = ss.sem(r12_sitefit/np.sqrt(r11_sitefit*r22_sitefit), axis=None)\n",
    "    NStd_site[inc] = ss.sem(r12_sitefit, axis=None)\n",
    "    LStd_site[inc] = ss.sem(r11_sitefit, axis=None)\n",
    "    RStd_site[inc] = ss.sem(r22_sitefit, axis=None)\n",
    "    \n",
    "    for ir, r in enumerate(reg_methods):\n",
    "        \n",
    "        VEpop = np.nanmedian(r12_reg[:,ir]/np.sqrt(r11_reg[:, ir]*r22_reg[:, ir]))\n",
    "        Npop = np.median(r12_reg[:, ir])\n",
    "        Lpop = np.median(r11_reg[:, ir])\n",
    "        Rpop = np.median(r22_reg[:, ir])\n",
    "        \n",
    "        VEarraypop[inc, ir] = VEpop\n",
    "        Narraypop[inc, ir] = Npop\n",
    "        Larraypop[inc, ir] = Lpop\n",
    "        Rarraypop[inc, ir] = Rpop\n",
    "        \n",
    "        VEarraypopStd[inc, ir] = ss.sem(r12_reg[:, ir]/np.sqrt(r11_reg[:, ir]*r22_reg[:, ir]), axis=None)\n",
    "        NarraypopStd[inc, ir] = ss.sem(r12_reg[:, ir], axis=None)\n",
    "        LarraypopStd[inc, ir] = ss.sem(r11_reg[:, ir], axis=None)\n",
    "        RarraypopStd[inc, ir] = ss.sem(r22_reg[:, ir], axis=None)\n",
    "        \n",
    "        VEsite = np.nanmedian(r12_reg_sitfit[:,ir]/np.sqrt(r11_reg_sitfit[:, ir]*r22_reg_sitfit[:, ir]))\n",
    "        Nsite = np.median(r12_reg_sitfit[:, ir])\n",
    "        Lsite = np.median(r11_reg_sitfit[:, ir])\n",
    "        Rsite = np.median(r22_reg_sitfit[:, ir])\n",
    "        \n",
    "        VEarraysite[inc, ir] = VEsite\n",
    "        Narraysite[inc, ir] = Nsite\n",
    "        Larraysite[inc, ir] = Lsite\n",
    "        Rarraysite[inc, ir] = Rsite\n",
    "        \n",
    "        VEarraysiteStd[inc, ir] = ss.sem(r12_reg_sitfit[:, ir]/np.sqrt(r11_reg_sitfit[:, ir]*r22_reg_sitfit[:, ir]), axis=None)\n",
    "        NarraysiteStd[inc, ir] = ss.sem(r12_reg_sitfit[:, ir], axis=None)\n",
    "        LarraysiteStd[inc, ir] = ss.sem(r11_reg_sitfit[:, ir], axis=None)\n",
    "        RarraysiteStd[inc, ir] = ss.sem(r22_reg_sitfit[:, ir], axis=None)\n",
    "\n",
    "VEarraysite = np.concatenate((VE_site[:, np.newaxis], VEarraysite), axis=1)\n",
    "VEarraypop = np.concatenate((VE[:, np.newaxis], VEarraypop), axis=1)\n",
    "\n",
    "VEarraysiteStd = np.concatenate((VEStd_site[:, np.newaxis], VEarraysiteStd), axis=1)\n",
    "VEarraypopStd = np.concatenate((VEStd[:, np.newaxis], VEarraypopStd), axis=1) \n",
    "\n",
    "\n",
    "NarraysiteStd = np.concatenate((NStd_site[:, np.newaxis], NarraysiteStd), axis=1)\n",
    "NarraypopStd  = np.concatenate((NStd[:, np.newaxis], NarraypopStd), axis=1)\n",
    "\n",
    "Narraysite = np.concatenate((N_site[:, np.newaxis], Narraysite), axis=1)\n",
    "Narraypop = np.concatenate((N[:, np.newaxis], Narraypop), axis=1)\n",
    "\n",
    "\n",
    "LarraysiteStd = np.concatenate((LStd_site[:, np.newaxis], LarraysiteStd), axis=1)\n",
    "LarraypopStd = np.concatenate((LStd[:, np.newaxis], LarraypopStd), axis=1)\n",
    "\n",
    "Larraysite = np.concatenate((L_site[:, np.newaxis], Larraysite), axis=1)\n",
    "Larraypop = np.concatenate((L[:, np.newaxis], Larraypop), axis=1)\n",
    "\n",
    "\n",
    "RarraysiteStd = np.concatenate((RStd_site[:, np.newaxis], RarraysiteStd), axis=1)\n",
    "RarraypopStd = np.concatenate((RStd[:, np.newaxis], RarraypopStd), axis=1)\n",
    "\n",
    "Rarraysite = np.concatenate((R_site[:, np.newaxis], Rarraysite), axis=1)\n",
    "Rarraypop = np.concatenate((R[:, np.newaxis], Rarraypop), axis=1)\n",
    "\n",
    "\n",
    "# Plot figures\n",
    "\n",
    "\n",
    "reg_methods_labels = ['Pinv','PLS','PCR','Ridge']\n",
    "colors = ['k','gray','DarkRed','red']            \n",
    "n_regs = len(reg_methods)\n",
    "fig, axes = plt.subplots(nrows=1, ncols=4, figsize=[20, 4])\n",
    "\n",
    "for iax, ax in enumerate(np.ravel(axes).tolist()):\n",
    "    ax.set_ylim([0., 1.02])\n",
    "\n",
    "#         \n",
    "axes[0].set_title('VE')\n",
    "\n",
    "[axes[0].errorbar(n_components_range, VEarraysite[:, itrace], yerr=VEarraysiteStd[:, itrace], marker='o', ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [0,3]]\n",
    "if nc == 2:\n",
    "    axes[0].legend()\n",
    "[axes[0].errorbar(n_components_range, VEarraypop[:, itrace], yerr=VEarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs+1)]\n",
    "\n",
    "\n",
    "\n",
    "axes[1].set_title('N')\n",
    "\n",
    "[axes[1].errorbar(n_components_range, Narraysite[:, itrace], yerr= NarraysiteStd[:, itrace], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [0,3]]\n",
    "\n",
    "[axes[1].errorbar(n_components_range, Narraypop[:, itrace], yerr=NarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs+1)]\n",
    "\n",
    "\n",
    "axes[2].set_title('L')\n",
    "\n",
    "[axes[2].errorbar(n_components_range, Larraysite[:, itrace], yerr=LarraysiteStd[:, itrace], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [0,3]]\n",
    "\n",
    "[axes[2].errorbar(n_components_range, Larraypop[:, itrace], yerr=LarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs+1)]\n",
    "\n",
    "axes[3].set_title('R')\n",
    "\n",
    "[axes[3].errorbar(n_components_range, Rarraysite[:, itrace], yerr=RarraysiteStd[:, itrace], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [0,3]]\n",
    "\n",
    "[axes[3].errorbar(n_components_range, Rarraypop[:, itrace], yerr=RarraypopStd[:, itrace],  marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs+1)]\n",
    "\n",
    "[ax.set_xlabel('# of components') for ax in axes[ :]]\n",
    "[ax.legend() for ax in axes[ :]]\n",
    "\n",
    "\n",
    "title = 'Model %s, Data %s'%('conv5', 'IT')\n",
    "axes[0].text(1,1.2, title, fontsize=15)\n",
    "\n",
    "fig.savefig(resultdir+'UnitTest_summary_SweepNumComponents_%s_%s_SB%s_%s.png'%(reg_methods, reg_params_list, spearman_brown, purpose_of_this_run), dpi=300)\n",
    "fig.savefig(resultdir+'UnitTest_summary_SweepNumComponents_%s_%s_SB%s_%s.pdf'%(reg_methods, reg_params_list, spearman_brown, purpose_of_this_run), dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.75300119],\n       [0.91221882],\n       [0.89937096],\n       [0.96590286],\n       [0.93007857],\n       [0.95936731],\n       [0.96806372],\n       [0.94630328],\n       [0.96276552],\n       [0.9691001 ],\n       [0.97132871],\n       [0.81240535],\n       [0.95703787],\n       [0.86782119],\n       [0.96431105],\n       [0.93057635],\n       [0.97747484],\n       [0.91968839],\n       [0.97764396],\n       [0.93380028],\n       [0.97739581],\n       [0.96258085],\n       [0.96575594],\n       [0.97595019],\n       [0.72341788],\n       [0.92493238],\n       [0.82029154],\n       [0.65567361],\n       [0.90257538],\n       [0.84390088],\n       [0.62567017],\n       [0.81052268],\n       [0.87948234],\n       [0.91264086],\n       [0.77360014],\n       [0.82091576],\n       [0.84636636],\n       [0.88818915],\n       [0.77890521],\n       [0.96781206],\n       [0.94356092],\n       [0.89499031],\n       [0.98294864],\n       [0.90491038],\n       [0.98356125],\n       [0.95657573],\n       [0.76159243],\n       [0.72046969],\n       [0.81193873],\n       [0.67810274],\n       [0.89511429],\n       [0.92291192],\n       [0.41910667],\n       [0.10006264],\n       [0.89606417],\n       [0.76009926],\n       [0.93910754],\n       [0.75816964],\n       [0.74824248],\n       [0.93234279],\n       [0.98127809],\n       [0.93606899],\n       [0.94456125],\n       [0.95902738],\n       [0.94567719],\n       [0.93877257],\n       [0.9210982 ],\n       [0.85971631],\n       [0.87187455],\n       [0.98118625],\n       [0.91234818],\n       [0.84639683],\n       [0.94355991],\n       [0.97684002],\n       [0.92168764],\n       [0.98706516],\n       [0.95071218],\n       [0.96023517],\n       [0.97963902],\n       [0.8935616 ],\n       [0.90060838],\n       [0.94283147],\n       [0.85788109],\n       [0.94800309],\n       [0.97062545],\n       [0.97909858],\n       [0.92727117],\n       [0.87534099],\n       [0.97416105],\n       [0.91093503],\n       [0.956885  ],\n       [0.92389051],\n       [0.94192218],\n       [0.11433664],\n       [0.37580945],\n       [0.79330864],\n       [0.35207712],\n       [0.78475264],\n       [0.43104913],\n       [0.70338206],\n       [0.26033779],\n       [0.89506008],\n       [0.8105673 ],\n       [0.97717151],\n       [0.30770324],\n       [0.3488612 ],\n       [0.31829481],\n       [0.35399234],\n       [0.31397932],\n       [0.46554448],\n       [0.16078392],\n       [0.6382613 ],\n       [0.50394242],\n       [0.27683532],\n       [0.74028236],\n       [0.12918762],\n       [0.53199097],\n       [0.7219472 ],\n       [0.59087536],\n       [0.28799309],\n       [0.37676617],\n       [0.24765051],\n       [0.86252669],\n       [0.9376622 ],\n       [0.97627182],\n       [0.38890955],\n       [0.52658753],\n       [0.76067279],\n       [0.83452957],\n       [0.90981011],\n       [0.9033674 ],\n       [0.38112334],\n       [0.85951945],\n       [0.95211354],\n       [0.97002645],\n       [0.96309074],\n       [0.9611357 ],\n       [0.95264471],\n       [0.98858099],\n       [0.98725454],\n       [0.98745275],\n       [0.98491609],\n       [0.63833569],\n       [0.67448137],\n       [0.93087795],\n       [0.9498354 ],\n       [0.83891641],\n       [0.94733695],\n       [0.97471339],\n       [0.93145109],\n       [0.99316268],\n       [0.93393229],\n       [0.96364222],\n       [0.97660289],\n       [0.98189919],\n       [0.90749359],\n       [0.9768065 ],\n       [0.96695693],\n       [0.97249853],\n       [0.98373489],\n       [0.9729701 ],\n       [0.99259519],\n       [0.98815378],\n       [0.99059827],\n       [0.97029435],\n       [0.98705291],\n       [0.9168833 ],\n       [0.96489656]])"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r11"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:72: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:77: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:84: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:94: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:99: RuntimeWarning: invalid value encountered in true_divide\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/numpy/lib/function_base.py:4033: RuntimeWarning: All-NaN slice encountered\n  r = func(a, **kwargs)\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:109: RuntimeWarning: invalid value encountered in true_divide\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:99: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:109: RuntimeWarning: invalid value encountered in sqrt\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(1,1.2,'Model conv5, Data IT')"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1008 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import scipy.stats as ss\n",
    "data_unit_indices = range(168)\n",
    "purpose_of_this_run = 'masterplot' #'masterplotwithpinv' \n",
    "\n",
    "ni = 5120\n",
    "nf = 168\n",
    "nt = 46\n",
    "reg_methods = ['PLS', 'ridge', 'ridge']\n",
    "n_components_range = [5, 12, 25, 35, 50,80,100,128, 168]#np.arange(nf,1,-1)\n",
    "\n",
    "VE = np.zeros((len(n_components_range)))\n",
    "N = np.zeros_like(VE)\n",
    "L = np.zeros_like(VE)\n",
    "R = np.zeros_like(VE)\n",
    "\n",
    "VEstd = np.zeros((len(n_components_range)))\n",
    "Nstd = np.zeros_like(VE)\n",
    "Lstd = np.zeros_like(VE)\n",
    "Rstd = np.zeros_like(VE)\n",
    "\n",
    "\n",
    "\n",
    "VEarraysite = np.zeros((len(n_components_range), len(reg_methods)))\n",
    "Narraysite = np.zeros_like(VEarraysite)\n",
    "Larraysite = np.zeros_like(VEarraysite)\n",
    "Rarraysite = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraypop = np.zeros_like(VEarraysite)\n",
    "Narraypop = np.zeros_like(VEarraysite)\n",
    "Larraypop = np.zeros_like(VEarraysite)\n",
    "Rarraypop = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraypopStd = np.zeros_like(VEarraysite)\n",
    "NarraypopStd = np.zeros_like(VEarraysite)\n",
    "LarraypopStd = np.zeros_like(VEarraysite)\n",
    "RarraypopStd = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraysiteStd = np.zeros_like(VEarraysite)\n",
    "NarraysiteStd = np.zeros_like(VEarraysite)\n",
    "LarraysiteStd = np.zeros_like(VEarraysite)\n",
    "RarraysiteStd = np.zeros_like(VEarraysite)\n",
    "\n",
    "\n",
    "for inc, nc in enumerate(n_components_range):\n",
    "    trainfraci = 0.8  # image trainfrac\n",
    "    splitfract = 0.5  # trial splitfrac\n",
    "    nfoldi = 2\n",
    "    nfoldt = 2\n",
    "    \n",
    "    spearman_brown = False\n",
    "    \n",
    "    reg_params_list = [nc, [20, -10, 10], [20, -10, 10]]\n",
    "    report_popfit = [True, True, True]  # [False, True, True]\n",
    "    report_sitefit = [False, False, True]  # [True, True, True]\n",
    "    PCA_ncomponents_list = [ -1,  nc, -1]\n",
    "\n",
    "    # file = open(resultdir+'unit_test_%s_ni%d_nf%d_nt%d_collilnearity%s_%s_SB%s_VariousSd%s_noisymap%s.pickle' % (\n",
    "    #     reg_methods, ni, nf, nt, Collinearity, noise_dist, spearman_brown, various_unit_stds, noisy_map), 'rb')\n",
    "    \n",
    "    # file = open(resultdir+'unit_test_%s_%s_%s, ni%d_nf%d_nt%d_collinearity%s_%s_SB%s_VariousSd%s_noisymap%s_statsfromHvM%s_%dcmp.pickle' % (\n",
    "    #     reg_methods, reg_params_list, PCA_ncomponents_list , ni, nf, nt, Collinearity, noise_dist, spearman_brown, various_unit_stds, noisy_map,stats_from_data, nc), 'rb')\n",
    "    \n",
    "    file = open(resultdir + 'MappingMtoD_%s_%s_%s, ni%d_nf%d_nt%d_SB%s_%dcmp_%s.pickle' % (\n",
    "              reg_methods, reg_params_list, PCA_ncomponents_list, ni, nf, nt,  spearman_brown, nc, purpose_of_this_run), 'rb')\n",
    "\n",
    "    \n",
    "    data_list = pickle.load(file)\n",
    "    file.close()\n",
    "    \n",
    "    r12,r11,r22,r12_reg,r11_reg,r22_reg,r12_reg_sitfit,r11_reg_sitfit,r22_reg_sitfit = data_list\n",
    "    \n",
    "    VE[inc] = np.nanmedian(r12/np.sqrt(r11*r22))\n",
    "    N[inc] = np.median(r12)\n",
    "    L[inc] = np.median(r11)\n",
    "    R[inc] = np.median(r22)\n",
    "    \n",
    "    VEstd[inc] = ss.sem(r12/np.sqrt(r11*r22), axis=None)\n",
    "    Nstd[inc] = ss.sem(r12, axis=None)\n",
    "    Lstd[inc] = ss.sem(r11, axis=None)\n",
    "    Rstd[inc] = ss.sem(r22, axis=None)\n",
    "    \n",
    "    for ir, r in enumerate(reg_methods):\n",
    "        \n",
    "        VEpop = np.nanmedian(r12_reg[:,ir]/np.sqrt(r11_reg[:, ir]*r22_reg[:, ir]))\n",
    "        Npop = np.median(r12_reg[:, ir])\n",
    "        Lpop = np.median(r11_reg[:, ir])\n",
    "        Rpop = np.median(r22_reg[:, ir])\n",
    "        \n",
    "        VEarraypop[inc, ir] = VEpop\n",
    "        Narraypop[inc, ir] = Npop\n",
    "        Larraypop[inc, ir] = Lpop\n",
    "        Rarraypop[inc, ir] = Rpop\n",
    "        \n",
    "        VEarraypopStd[inc, ir] = ss.sem(r12_reg[:, ir]/np.sqrt(r11_reg[:, ir]*r22_reg[:, ir]), axis=None)\n",
    "        NarraypopStd[inc, ir] = ss.sem(r12_reg[:, ir], axis=None)\n",
    "        LarraypopStd[inc, ir] = ss.sem(r11_reg[:, ir], axis=None)\n",
    "        RarraypopStd[inc, ir] = ss.sem(r22_reg[:, ir], axis=None)\n",
    "        \n",
    "        VEsite = np.nanmedian(r12_reg_sitfit[:,ir]/np.sqrt(r11_reg_sitfit[:, ir]*r22_reg_sitfit[:, ir]))\n",
    "        Nsite = np.median(r12_reg_sitfit[:, ir])\n",
    "        Lsite = np.median(r11_reg_sitfit[:, ir])\n",
    "        Rsite = np.median(r22_reg_sitfit[:, ir])\n",
    "        \n",
    "        VEarraysite[inc, ir] = VEsite\n",
    "        Narraysite[inc, ir] = Nsite\n",
    "        Larraysite[inc, ir] = Lsite\n",
    "        Rarraysite[inc, ir] = Rsite\n",
    "        \n",
    "        VEarraysiteStd[inc, ir] = ss.sem(r12_reg_sitfit[:, ir]/np.sqrt(r11_reg_sitfit[:, ir]*r22_reg_sitfit[:, ir]), axis=None)\n",
    "        NarraysiteStd[inc, ir] = ss.sem(r12_reg_sitfit[:, ir], axis=None)\n",
    "        LarraysiteStd[inc, ir] = ss.sem(r11_reg_sitfit[:, ir], axis=None)\n",
    "        RarraysiteStd[inc, ir] = ss.sem(r22_reg_sitfit[:, ir], axis=None)\n",
    "\n",
    "\n",
    "# Plot figures\n",
    "\n",
    "\n",
    "reg_methods_labels = ['PLS', 'PCR', 'Ridge']\n",
    "n_regs = len(reg_methods)\n",
    "colors = ['gray', 'darkred', 'red']            \n",
    "fig, axes = plt.subplots(nrows=3,ncols=len(reg_methods)+1, figsize=[20, 14])\n",
    "\n",
    "for iax, ax in enumerate(np.ravel(axes).tolist()):\n",
    "    ax.set_ylim([-0.0, 1.02])\n",
    "\n",
    "axes[0, 0].errorbar(n_components_range, VE[:], yerr=VEstd[:], label='VE', marker='o', color='r')\n",
    "axes[0, 0].errorbar(n_components_range, N[:], yerr=Nstd[:], label='N', marker='o', color='salmon')\n",
    "axes[0, 0].errorbar(n_components_range, L[:], yerr=Lstd[:], label='L', marker='o', color='b')\n",
    "axes[0, 0].errorbar(n_components_range, R[:], yerr=Rstd[:], label='R', marker='o', color='g')\n",
    "\n",
    "axes[0,0].legend()\n",
    "axes[0,0].set_title('pinv')\n",
    "\n",
    "for iax, ax in enumerate(axes[0,1:].tolist()):\n",
    "    \n",
    "    ax.errorbar(n_components_range, VEarraypop[:,iax], yerr=VEarraypopStd[:,iax], marker='o', color='r')\n",
    "    ax.errorbar(n_components_range, Narraypop[:,iax], yerr=NarraypopStd[:,iax], marker='o', color='salmon')\n",
    "    ax.errorbar(n_components_range, Larraypop[:,iax], yerr=LarraypopStd[:,iax], marker='o', color='b')\n",
    "    ax.errorbar(n_components_range, Rarraypop[:,iax], yerr=RarraypopStd[:,iax], marker='o', color='g')\n",
    "    ax.set_title(reg_methods_labels[iax]+ ' popfit')\n",
    "\n",
    "for iax, ax in enumerate(axes[1, 1:].tolist()):\n",
    "    \n",
    "    ax.errorbar(n_components_range, VEarraysite[:,iax], yerr=VEarraysiteStd[:,iax], marker='o', color='r', ls='--')\n",
    "    ax.errorbar(n_components_range, Narraysite[:,iax], yerr=NarraysiteStd[:,iax], marker='o', color='salmon', ls='--')\n",
    "    ax.errorbar(n_components_range, Larraysite[:,iax], yerr=LarraysiteStd[:,iax], marker='o', color='b', ls='--')\n",
    "    ax.errorbar(n_components_range, Rarraysite[:,iax], yerr=RarraysiteStd[:,iax], marker='o', color='g', ls='--')\n",
    "    ax.set_title(reg_methods_labels[iax] + ' sitefit')\n",
    "    # ax.set_xlabel('ni/nf')\n",
    "#         \n",
    "axes[2,0].set_title('VE')\n",
    "\n",
    "[axes[2,0].errorbar(n_components_range, VEarraysite[:, itrace], yerr=VEarraysiteStd[:, itrace], marker='o', ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "if nc == 2:\n",
    "    axes[2,0].legend()\n",
    "[axes[2,0].errorbar(n_components_range, VEarraypop[:, itrace], yerr=VEarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "axes[2,1].set_title('N')\n",
    "\n",
    "[axes[2,1].errorbar(n_components_range, Narraysite[:, itrace ], yerr= NarraysiteStd[:, itrace], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "[axes[2,1].errorbar(n_components_range, Narraypop[:, itrace], yerr=NarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "\n",
    "axes[2,2].set_title('L')\n",
    "\n",
    "[axes[2,2].errorbar(n_components_range, Larraysite[:, itrace ], yerr=LarraysiteStd[:, itrace, ], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "[axes[2,2].errorbar(n_components_range, Larraypop[:, itrace], yerr=LarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "axes[2,3].set_title('R')\n",
    "\n",
    "[axes[2,3].errorbar(n_components_range, Rarraysite[:, itrace], yerr=RarraysiteStd[:, itrace], marker='o', ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "[axes[2,3].errorbar(n_components_range, Rarraypop[:, itrace], yerr=RarraypopStd[:, itrace],  marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "[ax.set_xlabel('# of components') for ax in axes[2, :]]\n",
    "[ax.legend() for ax in axes[2, :]]\n",
    "\n",
    "title = 'Model %s, Data %s'%('conv5', 'IT' )\n",
    "\n",
    "axes[0,0].text(1,1.2, title, fontsize=15)\n",
    "\n",
    "# fig.savefig(resultdir+'unit_test_SweepNumComponents_nitonf%d_nt%d_Collin%s_%s_SB%s_%s_statsfromHvM%s_%s.png'%(imag_feat_ratio, nt, Collinearity, noise_dist, spearman_brown, noisy_map,stats_from_data, purpose_of_this_run), dpi=300)\n",
    "# fig.savefig(resultdir+'unit_test_SweepNumComponents_nitonf%d_nt%d_Collin%s_%s_SB%s_%s_statsfromHvM%s_%s.pdf'%(imag_feat_ratio, nt, Collinearity, noise_dist, spearman_brown, noisy_map,stats_from_data, purpose_of_this_run), dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9, 3)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "VEarraysite.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:61: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:71: RuntimeWarning: invalid value encountered in sqrt\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:76: RuntimeWarning: invalid value encountered in true_divide\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/numpy/lib/function_base.py:4033: RuntimeWarning: All-NaN slice encountered\n  r = func(a, **kwargs)\n/home/tahereh/anaconda3/envs/pytorch_tensorflow/lib/python3.5/site-packages/ipykernel_launcher.py:86: RuntimeWarning: invalid value encountered in true_divide\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIkAAANcCAYAAADb5v9KAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3Xmcm2W9///XlWT2ycy0M51pZ7rSAtIiKrRlUyi4HNxAa42i5xw53/MVz1c9KAKlIALiVgF/R1FQQT3o8QhEGBUUF0RaFoUuIkILSKEtTPeZtrNvyX39/riTTJJJZjJrkpn38/HII3fu3MvnTjKfyf3JdV23sdYiIiIiIiIiIiLTmyfbAYiIiIiIiIiISPapSCQiIiIiIiIiIioSiYiIiIiIiIiIikQiIiIiIiIiIoKKRCIiIiIiIiIigopEIiIiIiIiIiJCjhaJjDHXG2OsMealNM/viDx//Tjtr3mk24rE2Dwe+58sxphVkdct+bZ+FNu6KG59xxjTaoz5uzHmm8aYxaOM72JjzPtGs26a7W0wxtwbmd6V5tjjbxeN175FRERERERE8o0v2wEMoQdYZIxZbq3dEp1pjFkBLIg8L6PzUeCVuMd7xrCtc4FuoBxYBnwC+LgxZo219rcj3NbFwHPAL8cQTzrvB4riHv8OuBf4Qdy8lydgvyIiIiIiIiJ5IZeLRJ3AX4EPA1vi5n8Y+BNwSjaCmiL+bq19bpy2tdla2xGZ/qMx5nvAr4GfGWMWWmtbx2k/Y2KtfTr+sTEmBDRZa5/MUkgiIiIiIiIiOSUnu5vFuRsIGGMMQOQ+EJk/iDEmYIx51hjTa4x5zRjzFWOML2mZs4wxzxhjeowxW40xZ6TZ1gXGmC2R5fYbY240xhSM9ACMMScZYx4wxhw1xnQYYzYZY94e9/wiY8wvjTFtxpj2yLJLkrZhjTGfMcZ81RhzyBhz0BhzqzGmKG4b1hjzrqT1vJHYvzTSuEfLWtsL/CdQBVwYF8tlxpjNkW5pB5KP0xizAbfw97Hk7l/GmH81xjxujDlsjDlijHnEGLN8so5JREREREREZDrI9SJRI1AHvDny+C3ALOAXyQsaY94B3IPb+ugC4NvA5cB34papB34LHAbWAN8H/hcoTdpWILLvTcD5wBdxu0J9bSTBG2NeBzwBzAH+A7fL0y+AeZHni4CHgROAjwMXAYuAjcaYmUmbuwyoB/4ZuAm3W9dnAKy1OyOxfihpnbNxX797kub/yRgTjozTc40xxjuS4xqOtfYFoAk4LW72XNz34gLcY/UCTxhjKiPPfxJ4AXgQOD1y+03kuYXAT4APAh+JbPtRY8wx4xm3iIiIiIiIyHSWy93NsNYeNcb8DreL2WOR+99F5icvfgOwwVr7scjj30WW+Zox5svW2ibgs7hjGb3bWtsFYIzpBH4a3UiktdJNwE+stZ+Mm98L3GqM+Zq1tiXDQ7gOaAXeYq3tjsx7KO75fwPmA8dZa1+J7Ocp3PGCPkFiUWqXtfaiyPTvjTFnAquBGyPz7gauN8YURVrzgFs02h7XtawVWI/7WvYB78EtgM0iUnAaR024BSoArLWXRqcjRamHgIO4RaOfWGu3R96LQ8ldwKy1N8St64msuwK3YHYDIiIiIiIiIjJmud6SCNzix5pIq5s1pOhqFik6nAz8POmpe3CP8fTI45XAQ9ECUURj0jrH4RZugsYYX/SGOw5SMXDiCGI/F7gnrkCUbCXw12iBCCBSzHqCgdZTUX9Ierwdt3VOVBDwA+cBRGJeTdzrZa192lp7lbX2QWvtH621nwVuBj5pjKkZwXFlIqGKZ4w5zRjzkDGmBQgBXbiDXR837IaMOcEY8wtjzAEgDPQDx2eyroiIiIiIiIhkJh+KRPfjFhO+ApQBD6RYpgYoAA4kzY8+jnbdmo3beiUmUsDpiJsVLZY8iFuMiN52RubPG0Hs1cC+IZ6fkyJmIvOSu5sdTXrch1u0AsBauwd4nIEuZ2/FPZaU4zfFuRe3RdlJwyw3Ug1Ejs0YMx+3yGVwW0ididsS6CBxx5CKMcYfWXce8DncLocrgGeGW1dEREREREREMpfT3c0ArLWdxphfA5cCP7fWdqZYrBm3kFObND/a3elw5H5/8jLGmBLcIhRJy14MJFwRK2JninnptOAWgtLZh3vZ+GR1cXGMxD3A+sgxfQh42lr7Uobr2lHsLyVjzAm4rZz+Epl1Hu64TxdE379IS6fkQlgqp0e29fbIWEfRfVSmX0VERERERERERiofWhIBfBe3BdH3Uj1prQ0DW3EHNo4XABwGihWbgbcbY+IHql6dtM6LwB5gobV2S4pbpuMRgTsodcAYk67Fy1PAKcaYRdEZxpgG4AzcVkEj9XOgBHeA7PczfCsigA/gdv/6+yj2N0ikW+AtuC2fovsvwX0fQnGLBhhcpExoHRW3LkB0nCUiV6RbOB7xioiIiIiIiIgr51sSAVhrNwAbhlnsOtwBnf8btzjxeuBLwB2RcX4Avgl8Cvi1Meb/w71a2FVAbMwga61jjLkM+B9jTAXu1dD6gGOA9wFrksY0GsoXcQtTjxpjvoHbsuhNQIu19kfAncCVwG+NMdfijrdzPW7LqO9nuI8Ya+3ByKXkb8a9BH0w/nljzHeBQ5GY+oB3AZ8Gvhlf/DLG3AmsstYuzGC3K4wx3bgthU7E7U62EPd1ao0s8yfcq5n9tzHmh7itpy5ncBe6F4B/Msb8E+5rtRN4Erc74B3GmBtxWxVdj1vIExEREREREZFxki8tiYZlrf0D7tXPluO2Ovos8A3cIkh0mT24hZEa4D7cy67/M+4gyvHbugf3qltvxG2d0xhZ9q+4xZVMY3oRdwDqZuAHwC9wB9/eHXm+F3gbbnHkh8CPI8+tstaOprsZuAWyOcCT1tpdSc89j1vo+hnuWE9vAy4DrkharpSksZuG8Cfcllq/wr20/R+Bk6y1v40uYK19FvdKbqcCv8a9jP0Hca+2Fu/LkRiDuIWs91prD0SWnR3Zx2eB/wB2ZBifiIiIiIiIiGTAWDtuQ9HIFGGM2Q1ca639cbZjEREREREREZHJMWVaEsn4MMbU414p7q5sxyIiIiIiIiIik0ctiURERERERERERC2JRERERERERERERSIREREREREREUFFIhERERERERERQUUiERERERERERFBRSIREREREREREUFFIhERERERERERQUUiERERERERERFBRSIREREREREREUFFIhERERERERERQUUiERERERERERFBRSIREREREREREUFFIhERERERERERQUUiERERERERERFBRSIREREREREREUFFIhERERERERERQUUiERERERERERFBRSIREREREREREUFFIhERERERERERQUUiERERERERERFBRSIREREREREREUFFIslTxpivGmMOGGOsMeYiY8z1xpgd2Y5LRKYn5SQRERERmQpUJJpijDF3Rk5SrDEmZIzZbYz5njGmOm4Za4z55yG28WZjzB+MMYeMMT2RbdxrjFkwOUcxNGPMqcBVwMXAHOAe4GbgtLhlrjHG7MpKgCLTXCZ5KLLc6caYxkhxpccY87Ix5qfGmJPjlrFxty5jzHZjzOcm/6jSU04SERERkalCRaKp6THcE5WFwCXAB4CfZLKiMeYE4CHgJeBtwAnARcAuoGLcIx2dYwHHWvsra+1+a223tbbDWtuc7cBEJGbIPGSM+bfIMv3AR3FzzYdwc823krb16ci2lkWe+7ox5uIJjX5klJNEREREZEpQkWhq6oucqDRZa38FfBM4zxhTksG6/wR0WGs/Za19xlq701r7iLX2cmvts+lWinatMMZ8xBjzSqRVwB+NMYuSlvtYpCVArzGmyRjzZWOML+75DcaYHxlj1htjmo0xbcaYH0RjN8bcCfwP4Im2Lojff2T6IuBLwIK4FgjXj+D1E5GxS5uHjDH1wHeBH1hrP2St/WMk12yx1l4DnJ+0rdbItnZaa78P/B03V6WlnCQiIiIiMnIqEk0P3bjvtW+4BYF9wAxjzDtHsZ85wCdxWwO8BfADvzTGGABjzLuBH+GeUL0euAz4FHBd0nbWANWRbXwU94Tx65HnPgN8FghH9jcnRRz3RJZvilvm5lEcj4iMn/g8FACKgC+nWtBaeyTVfON6K26ro74M9qmcJCIiIiIyApkUDSSPGWOW4p70PGWtbc9glZ8Dbwd+Y4w5AmwGHgF+Zq19bZh1S4GLrLXRX8//BXgReCvwR2AdcJ+19muR5f9hjJkNrDfGfMlaGz3pOwz8h7U2DDxvjLkG+LYx5iprbasxphXAWrs/VRDW2m5jTAcQTreMiEye5DxkjDkOaLPWNmW4iR8YY76HW1jyAV24LZOGo5wkIiIiIjICakk0Na0yxnQYY7qB54BXgI9ksqK11rHW/l+gHncckO3AJ3BPjFYNs/qh6MlYZFv/AJqBpZFZy4BHk9bZCBQDi+PmbYqcjEU9ARQmLSMiuW2oPGRGuK3PA28EzsHNB9dYa5/KYD3lJBERERGREVBLoqnpKeBjQAjYZ63tHekGIr923wXcZYxZBzyN2wVjwwg3lXwyaNM8nzx/qG2ISO4bKg+9CFQYY+Zm2JroQKTYs8MY837gJWPM09baDaOISzlJRERERCQNtSSamrqttTustbtGUyBKFuly8QpQO8yis4wxsV/WI11KqoHnI7O2AWcnrXMW7lglr8TNW2GM8cY9Ph13/JGXRxB2H+AddikRmShD5aGfA73ANalWNMbMSLdRa+0h4DbgW9GxhYagnCQiIiIiMgJqSTR9zTfGvDFp3l7g/cDJQCPuCVAB7iCt7wTWD7PNLuC/jTGX4v7S/m3gWdyxPwC+BjwQaZnUiNt95HrgG3Fjf4B7EnerMeZbwDG4VwW6w1rbOYLj2wnMNsacDrwEdFlru0awvohMEGvtHmPMp4HvG2OqgDtw881M4ALcbmVnDbGJW4DP4XZf+98hllNOEhEREREZAbUkmr6+gtuFLP72f4BNuIPD3op7mek/416J6LPAtcNscx9wO3Af7pgd3cD7rbUWwFr7YGQfH8Mdo+S/cFsEfDFpO/cC7cDjwN3Ag8DaER7fL3FbK/wGODSK9UVkAllrf4DbiqcYt2vri7h/+4uAS4ZZdz/uFcluMMYUDLGocpKIiIiIyAiYyHdlkTExxlwP/LO1dskYt7MB2BEZPFtEZFSUk0RERERERk4tiUREREREREREREUiERERERERERFRdzMREREREREREUEtiUREREREREREBPBlcd9qwiQydZhsBzAGykUiU4dykYjkAuUiEckFo8pFwxaJAoHAj4D3AAeDweCJKZ43wLeAdwFdwEXBYPCvmex87969semamhqam5szDHviKA7Fkcsx5GIc9fX1k7I/5SLFoThyK45ciCE+jsnKRTBx+Sg+F+WjXPlMjFa+xw/5fwz5Hj+gXJQDpsLnKN+PId/jh/w/hrHkoky6m90JnDfE8+8Ejo3cLga+O+poRETSuxPlIhHJDXeifCQi2XcnykUiMs6GLRIFg8FHgcNDLHIB8JNgMGiDweCTQFUgEJgzXgGKiIBykYjkDuUjEckFykUiMhHGY0yiBuC1uMdNkXn7khcMBAIX41axCQaD1NTUDATi8yU8zhbFoThyOQbFMSTlIsWhOKZZDLkUR5KM8tFQuSgf5eh7kbF8jx/y/xjyPf4cpFyUp/L9GPI9fpgaxzBa41EkSjUYUsoBz4LB4O3A7dFl4vv45UqfP8WhOHI5hlyMYzL73g9DuUhxKI5pFkN8HDmUiyDDfDRULspHufKZGK18jx/y/xjyPX6Y3DGJMqBclKfy/RjyPX7I/2OY6DGJhtMEzIt7PBfI79HORCQfKReJSK5QPhKRXKBcJCIjNh4tie4HPh0IBO4GTgVag8HgoO4dIiITTLlIRHKF8pGI5ALlIhEZsWGLRIFA4C5gFVATCASagOuAAoBgMPg94EHcyyruwL204r9NVLAiMn0pF4lIrlA+EpFcoFwkIhNh2CJRMBi8cJjnLfCpcYtIRCQF5SIRyRXKRyKSC5SLRGQijMeYRCIiIiIiIiIikudyukjU2FjCypW1zJ07h5Ura2lsLMl2SCIiIiIiIiIiU1LOFokaG0tYu7aSPXt8WGvYs8fH2rWVKQtFjY0lLFo0h4YGFZNEREREREREREZjPK5uNiHWr/fT3Z1Yw+ru9rBuXSW7dnmprnaornbYts3H97/vp6/PAMSKSQCrV3dPetwiIlPFmjXVFBT4uOuubEciIiIiIiKTIWdbEu3d6005v7PT8I1vVHD11VV84hMzueWWCnp7TcIy3d0e1q/3T0aYIiIx49Wqcc2aat7+9pHV8Me7RWVjYwlbtxby6KNGLTQZeH2Ligpy7vVYs6aaNWuqsx2GiIiIiEwBOVskqq8Pp5zf0BBm1669PP30fh566CDG2JTL7dnj5Y47yti7N2cPUUTyRCYFmGgXWbdV49BdZIfb10iLM+O174naXr7L5dcj+nn5y18Kc654JSIiIiL5J2e7m61b187atZUJXc5KShzWrWunoABqax1qax3q68Ps2TP4MAoK4PrrK7n++kpWruzl/PO7efe7e6itdSbzMEQkzyUWCAa6tPb3w1vf2ktHh6Gjw/DFL1ak7CJ7/fUV1NQ4+HwWnw98PktBAXi97n384z/8oZgbbqhI2Nfll1exa5eX5cv76Ojw0N5uaG9376OPGxtLUu770kur+M53ynEcCIcNjkNkGhwn9eNwGLq6DNYObqF5xRWVPP54EdXV4ViX35oa9zZzpjuvuHgC34wssBa+9rXU3Z9vuKGCRYtC+Hzu+5d4P3iezwcez8DzxqTZaYbSfTZhanS3jraOuvfelixHIiIiIjJ95GyRKPoFd/16P3v3eqmvD7NuXfugL77pikk33tjKG97Qx/33l/DrX5dwzTVVXHut5bTT+njve92CUXV1YsFI42+ISLJ046N97nMzMlq/pcXLhReOvitQb6/bxTaV4mIHv9/S3Z262hAKweLFITwe8HjcYoUxAwWM6Pzoc9HpO+4oS7m9nh7Dxo1FtLR46O9PvU+/34kVkKqrw9TUOAkFpWiByS0sORQUjO51AbdIctllVfT1ua1MU/2PGEpHh+HAAQ8HDngjN0/C/f79Xg4e9NDVlbpF6qFDXt7znlmjjj9aPIovHEWLiV6vjbxPifPc5aGkxMfmzVWxAlFUdOy+bdsKKCmxFBcPvqWaH51XUgJFRe5nIZuiLaT6+mDlytoRv7ciIiIiMjo5WyQCt1A03JfC4YpJl17awaWXdvDiiz4eeKCE++8v5qqrqrjmGsuZZ/Zy/vk9nHdeN488UqwvpCIySLrx0cDyla+0Ul5uKS+3XHllJc3Ng5edNSvMHXccob8fQiFDf7/bWqe/3xAKufNCIffxlVdWAqmKL5Zf/KKF8nK3KFRe7lBebmMFlpUra1O2qGxocPc9Ug8+WJx2e5s2HcRaaG83NDd7aGnx0NLijZt2b83NXl57zcff/uY+DodTF5WqqgYXjtzWSeG4aQfHcVv1eCMv8VCtaM47r4f9+91Cz8GDHvbvTy4CudOdnYMrISUlDnV1DrNnhznppH7q6sIEg6W0tg5etqYmzDe+cRTHGXgfw+Hoffz0wLxQyCQ9N7BudDvx86LLxq8bCrmFor6+1O9fZ6fhzjvL6OkZfVOloqLhC0qVlV6MqYzMZ8hlh56X2KpqqreQEhEREcllOV0kylQmxaTjjw9x/PHtXHZZO9u3uwWjBx4o4fLLq7jiikqMcbtbwMi+kI71l2wRyW3purQ2NIS56KKu2OOuLpOyVeO117axYkWas/kkt9xSnnZfK1em38ZQ3XNHY7jtGQMVFZaKijDHHBMG+ofcnuNAa6uhpcUbKSC5t8OH3WJSdN6OHT5aWtz5yd3d3P3OYcYMt2i0a5cvZSuaSy6pSrlucbGlri5MXV2YZcv6OffcMHV1Tmze7NkOtbVh/H47qBvYSSf1p3w9rruujbe9rXe4l3NC1NTUcMwxnmGLeT09bguw6K272yQ8HnoeKecfOeKhp8fQ1+ehq6s4tn7yRSRGIr5odOjQ4KJi9IIU+v8qIiIiMrGmRJFoJIyBZctCLFvWzpVXtvPsswV88IPVdHQM7k5y+eWVbN5cSENDmLlzwzQ0hJg7N0xtrYPXq187RaaDTAsw0b/5sRSNR1vsybR7bqbGe3seD8yYYZkxI8SSJcMvHw7D0aMDxaSWFg89PRW8+mp37PE//pH635e1cPXVbdTWDhR/6urCVFQMLv5kajze24mQSTGvpMRt4QOpL/IwFjU1NTQ3N8ceO47bPbK7m5SFp8yKVHD33aUp95e+VZ+IiIiIjJdpVySKZ4z7C3FnZ+ozh95ewwMPFHPkSOIX04ICy5w5Yfbv96b8JfurX/Xzvvd1j2pMh8SWSer2JpJtIymYZNKqMZN9jaYYMdZ9p9teciFgMni9xMYxOv54d15NTTnNzQPFsqG62H3qUx3jHlM2X490cq145Y6V5I5rNJai1GOPFaV8b9Nd9VRERERExs+0LhJFDdWdZNOmg3R2Gvbs8dLU5E24/8UvUl9qeN8+HwsXzqG21h3Xwu3O4ER+1U7s4jBjxsCv22qZJJKbxrsAk8m+cqkYkYvGu4tdvprMz+Zk0XsrIiIikj0qEjH8F9KyMstxx4U47rhQwnqbNhWmLC5VVYX513/tig2Ounu3j6ee8gxqkQRQWGhjRaRt23wpr6KkcRhERBLlWisaGT96b0VERESyR0UiRj/+Rrri0pe+1JZy3Z4e95LJ+/d7Y1feOXBg4Mo76S5jvWePl3e8YxYLFoRYsCDM/PkhFi507xsawmO6hLSISL6aiq1oxKX3VkRERCQ7VCSKGE0Xj5H+2llcDPPmhZk3L/W4CunG2Cgrs8yeHeYf//Dx8MPFCVeQ8XotDQ3hQcWjhQtDzJ/vDtYqIiIiIiIiIjIcFYnGaDx/7UzXMmn9+tbYPhwH9u/38OqrPnbv9rJ798D9b39bzOHDiV3aqqqcWMEo2hIpej97dhjvKC8Ws2ZNNQD33tsyug2IiIiIiIiISE5RkSiHZNIyyeOB+nqH+vo+Tjtt8Dba2w27d3sHFZH+/vcCHnywmFBooBVSYaFl7txwyiLS/PlhSktTt0JqbCxh69ZC+vrc1k8TPVbEmjXVFBT4uOuuCdvFqKhQJiIiIiIiIlOJikQ5ZqxXNvL7LSeeGOLEE0ODnguFYO9eL7t2DS4ibd5cSHt74qDZtbVhFi821NdXxbqz7d7t5bvfLZ+0K7BNdkFKREREREREZLpSkWga8flg/vww8+eHgb6E56yFI0fMoOLRvn0l/OUvhTQ2erE29cDa3d0eLr+8igcfLKa83OL3O5F7S1mZg99vKS937/1+h7IyG5tXVAQm9WZpbCxh7drKSStIjYSKVyIiIiIiIjLVqEgkgFuomTnTMnNmP298Y39sfk1NAc3NzfT2wmuveTn77FpgcFWntxd27vTR3m7o6PDQ3m5wnDTVnzgFBW6xqLzcJhWYHB56qDhhfCZwC1LXXFNJW5vB5wOv1x282522eL1uMczjcefFT8c/n2p6YDupp22k910uF69ERERERERERktFIslIUREsWRKmoSGc8gpsDQ1hHn74UOyxtdDTY2hvN7HCUUfHQAEpcdq97+w0tLd7aGnxsGuXj66u1EWm1lYPn/981YQd61C83jmEw5BcKOvu9nDllZW89JKP6mqHmhqH6uowNTXu9IwZDj79tYmIiIiIiEgO02mrjEi6K7CtW9eesJwxUFJiKSmx1NYChEe8r5Ura1MWpObMCfG73zUTDrvjLIXDhnDYvQ+F3HmOYxKeSzXtLgehkElaZ+D5+HWKispoa+vi298uTxlvV5fh1lvLCYcHF7eMscyYES0eOZHiUZjq6vjHA4WligqbthveRNOA3CIiIiIiItOTikQyIplcgW28pCtIXX11OzU1zrjvbzg1NcU0N7fT2FiStjXVk08e5OhRQ0uLl+ZmD83Nbsuo5mZvbLqlxcP27T5aWoo4etSTYk9uNzy3eBROKCxVVzssXOihqKgoobBUUjI+x6ixlkRERERERKYvFYlkxKJXYJuM/cDkFKRGYqjWVB5PdGynEMceO/y2+vrg8OFoMSl9Yenll30cOuShpye6z+qE7ZSVxbdScgtLM2cObqEUnZ+q65vGWhIREREREZneVCSSnBYtSNXU1NDc3JztcIDxLV4VFsLs2Q6zZztAaNjlu7oM4XA1L73Umraw9NprPp55xp2XqusbwIwZ4bgCknt/330lKQcKX7/eryKRiIiIiIjINKAikcgoTFZrqmSlpZaaGvD7+4dd1nGgtTWx61uqwtILL/hobvbS3p6669vevd7xPgwRERERERHJQSoSiUxRHg/MmGGZMSPEkiXDL59uoPD6+pEPOi4iIiIiIiL5J3XTARGZdtata6ekJHFA8FRXrhMREREREZGpSS2JRATI3YHCRUREREREZHKoSCQiMatXd/Ozn5UCcO+9LVmORkRERERERCaTikQikkDFIRERERERkelJYxKJiIiIiIiIiIiKRCIiIiIiIiIioiKRiIiIiIiIiIiQ4ZhEgUDgPOBbgBf4QTAYXJ/0/Hzgx0BVZJl1wWDwwXGOVUSmOeUiEckFykUikguUi0RkIgzbkigQCHiBW4F3AkuBCwOBwNKkxa4BgsFg8E3Ah4HbxjtQEZnelItEJBcoF4lILlAuEpGJkkl3s5XAjmAw+EowGOwD7gYuSFrGAhWR6Upg7/iFKCICKBeJSG5QLhKRXKBcJCITIpPuZg3Aa3GPm4BTk5a5HvhDIBD4T6AMeFuqDQUCgYuBiwGCwSA1NTUDgfh8CY+zRXEojlyOYZrHoVykOLIdhuLIsRiyFMek5KJ8lCufidHK9/gh/48h3+OfZMpFaUyFz1G+H0O+xw9T4xhGK5MikUkxzyY9vhC4MxgMfiMQCJwO/E8gEDgxGAw68QsFg8Hbgduj22hubo49V1NTQ/zjbFEciiOXY8jFOOrr6ydrl8pFiiPbYSiOHIshPo6plovyUa5qBNelAAAgAElEQVR8JkYr3+OH/D+GfI8fUC7KAVPhc5Tvx5Dv8UP+H8NYclEm3c2agHlxj+cyuKnivwNBgGAw+BegGJieZTcRmSjKRSKSC5SLRCQXKBeJyITIpEi0GTg2EAgsCgQChbiDnt2ftMyrwFsBAoHACbgJ6NB4Bioi017WclFJYyO1K1cyZ+5caleupKSxcaybHHf5EKPIFKHvRSKSC5SLRGRCDFskCgaDIeDTwO+B591ZwW2BQOCGQCBwfmSxy4CPBwKBZ4C7gIuCwWByc0cRkVHLVi4qaWykcu1afHv2YKzFt2cPlWvXTlgRJlrsKSguzrjYM9kxikxn+l4kIrlAuUhEJoqxNmt5wu7dO9AiMlf6/CkOxZHLMeRiHJH+rqn6xeeLIXNR7cqV+PbsGbRSqKGBg5s2jWsg0WKPp7t7ILiiIjo+9jFCy5bhOXoUT2srprUVz5EjeFpb8bS2UvC3v2FCoUHbC1dX0/yb3xCeOxfM6N6iXPu8ZZviyK0Y4uOYarkoH+XKZ2K08j1+yP9jyPf4AeWiHDAVPkf5fgz5Hj/k/zGMJRdlMnC1iCQpaWzEv3493r17CdfX075uHd2rV2c7LJkA3jRflNLNHy3T1kbFddclFIgATG8v/ttvT5jnVFTgVFbiVFZiKyshRYEIwNvSQt1pp+FUVtK/dCn9y5bFbqFjj4XCwnE9BhERERERyW8qEuWY+OJDrYoPOSm5tUe0aw+g92oKCtfXp2xJhLWU/9d/0fHJT0JR0cg37DgUbNtG0SOPULRhA4VbtmDC4ZSLWmM4+NhjblGoogJ8iak7XWun8KxZtF92GQXbtlHw3HOU/vSneHp63G0WFtJ/3HGE4gpH/UuXuttHuShfqGAtIiIiIuNJRaIcouJDjurpwbdzJ+bQIcqffpry73xnUGsPT3c3lVdfjenoIDx7Ns6cOe59dTV4MhkfXnJV+7p1g7qAOcXFhJYupeLmmyltbOTo+vX0nXnmsNvytLRQ9OijbmFo40a8kSasfSedRMenPkXpXXfhPTR4PMlwfT3hRYtGFmNJCW3XXpuYO8JhfDt34tu2LVY4KvrjHym9557YIqEFC3CqqijYti3Whc23Zw+VV1yB6eig+33vwxYWuq2QsvzZnu4FEv3PEBEREZHxpiJRDvGvX5+y+FDx5S/Te/bZOBUVUFCQpeimOMfBs28fvpdfxvfKK+59ZNrb1ISJjN1VAKQbxcu0t1N11VUJ82xBAeHa2ljRKDx7NuE5c3Ci05EbxcUTe3wyatGT7VTFiKING6i8+mpqAgG6PvAB+k45hfJbbx1Ybu1aQgsWULxhA0UbNlDwzDMYawnPnEnvqlXu7ayzcGbNAiB07LEpiz3t69aNOsYEXi+hJUsILVlCzwUXuPOsxXPwoFs0ihSOin/720Gtmjw9PVRddVXCZ9z6fNiCAigqwhYUuMWjggJs5DGFhQPzItPxy8TPp7BwYBvRZeO2bWpqKOrpic0veuIJ/N/6Fqa3F5iiBZJwGM+hQ3j37cO7d697i0z7Dh2iavPmwe9TdzdV//mfVK5bhy0ujt2Im7bFxdiSksTHKeYzzLL4fNDb6xYMRznmVTrTvQAoIiIiki0qEmVbby+FTz9N0RNP4E3VpQXwHjjA7JNOAsApLcVWVOBUVeFUVLjTcWOTRMcqiU1XVGCjy5aXj8sv//n85d20tw8uBL38Mt6dOxNPzEtLCS1eTN8ppxD+4AcJLV5M+Smn0DxzJrPOOSd1156GBpp/9Su8+/e7J3L79+OJmy7Yvp2iP/0JT1fX4HVnzHALR9EWSPFFpMg8O2PGuJ+ISWa6V69O+RnvXbWKgw8/jP+WWyj/zncoue++2Ohwvj17qPrMZzCA9XjoO+UU2i+/nN5zzqH/9a9P+beYcbFnBDEOyxicujp66+roPfdcAObMnZtyUQu0XXstpq8P+vsxvb2Y/v7E6b4+TORGf7873dWFOXoUE1mO/v7E6b4+d91hVA/zvKe7m8orrqDg6acJNzQQnjs3du/U1OTW34/j4GluTij8ePfuxROdjuSN5AHJbXEx4TlzYMECSNM9EaDrIx/B9PQMutHTg6etDdPdPfj5NGNbDaUe9/Nti4pixaRBBalURaikwlP8OgVbt1L+wx+6nyGmaAFQREREJEepSBQxaeNv9PdT8MwzFP35zxT9+c8UbN6Mp6cHa4zbSijFiVJ4xgw6Pvc594pGra3uF/y2NjxHj+Ldtw/fiy/G5g/FejxuUSlaSIovMEXnV1UlLMPChXjCYXe6uJiSX/wi97s3hEJ4X301oTVQtCjkPXgwtpj1eAjPm0do8WJ6zziD0OLF7u2YY3Bmzx50QllWU4Ntbk7btad93TqcOXNw5syh/01vSh2btZj29pRFpOjjgr//PdYNKWHV4mLCdXV45s2jqro6oagU6+JWW6vBiCdbSQntV15J6T334D1wIOEpA4RnzuTgY49hq6oy2ly02JPNKyqkG4cp3NBA5yc+MTE7tdYtMEWKRtFiU3TejNJSjh48GCsozfzoR1NersH09FD685/jaW9P3HxxMeH6ekJz57rFo/p69z56mz17yJaaI/ofYS2elpZYsceT1ArIu3evWwBKyve2qMj9m54zh75TT3Wn6+sH7uvrY8XimpoaPMcck/Z9arv++uFe8cFCocGFo0gxiRRFpXKvl66WlpQFJ+KmPc3NCduK3TIoDEZ5urvxr1+fO/9nRERERKYoFYmY4HEdwmEKtm2j8M9/puiJJyh86ik8nZ0A9J9wAl0f/Sh9Z55J76mnUvynP6UeV+SGGzKLIxzGdHS4l8hua4tdHnuoad/LLw9cUjupq1vU7Mi9LSx0TyIcJ+F5T3c3lVddhXf3brfA5PenvR9pd7m0J2bW4jl8OLE1ULQgtHt3wslHeMYMwosX07tq1UAhaPFiQgsWjGrA4bG09sAYbEUFoYoKQscfn365vj68Bw+6rQoiBaRoEam4uZnCv/3NPcmMdLWJF66pSRgXKVUXN1tRkbJVRT63Ess2T1wBMmH+kSMZF4hyxVCF0AljjNu1rKgoZZdOW1NDf1zRLNzQkLZAcnDTJkxrK96mJrx79uBraopNe/fsoeChhwaN/WQ9HvfvI9r6KNoSae5cfM8/j/8b34gN+u3bs4eqyy/H9+yzhBcsGCj+xN1HW8HEtl9QECv29K1YkVD4cSLTzsyZI2rtNO7vk8+HLS93W51moLSmho6xFDLD4cTCUqSINOu882JdfOON9xUFRURERGQwFYkYYiyga6+NfYkPz56dsoXGoJPqK6+kf+lStyD05z9T9OSTeFpbAehfvJjuD3yA3jPOoO+MM9xBjeOMqfgA4PViKysJV1aO4lUA+vrcVkqRVkme1lYqraWjqSlWVCq/9daUq5qODipuvnnYXTjFxYMLSH6/2x0u/t7vp2D7dsp+8pOELgdVn/0s/ptvxnP0aOx1BbeAFVq4kNCxx9Jz3nmEjjnGvS1ejJ05c3SvxxBG3bUnU4WFsRPU5N/aYy1MrMUcOTKoiBRrmdTURMGWLXiPHBm0eae0dFCXNu+BA5Tcf7+6eIxS2tY39fVZiGZsxpyLJsFwBRJbWUmospLQsmUMLqUC3d3u2D579gwUkCL3hVu24H3ggSG7X5neXvy33+7uy+cbaAH0pjfhvOtdA/87ogWgCRjEPh/epyF5vdiyMmxZWcLsqfS3JCIiIpJvVCQi/a+T3iNHqPnABwD3EtRObe3AF/+GBjzNzZT8+teJRYxLLol1gQgtWED3u97lthQ6/XS3C9MwstrVpLDQHbejpoboSBdOTQ1dcXGU/PKX6X+9f+IJTHs7nvZ2976tzZ1Ovo88Z9rb3eLT3r0Dz6dpzRRlwmG8+/fTFQgktAoKz50LXu94vhq5zxjszJmEZs4ktHRp+uV6evAeOJBYRIproVS4aRPeAwdSdv1QF4/MZaX1zQTKhW5vQxlzgaSkhPDixYQXL079fDiMZ/9+fHv2UP3+96fs2maN4cDWre7A41m60tuEF6yzYKr9LYmIiIjkk2lfJPI995z75T7FAKDhujqOfvObeCK/NkfvfS+8QNHDD8e6HsQzuN2bmn/3O7dwMcUM+eW9oAA7cybhsbTc6e93i0cdHdSecUbKLgf09dG6fv3o9zHdFBcTXrCA8IIF6ZdxHObMn68uHmOQ96068tCEFki8XpyGBvoiXc/StWxx6uomZv/TmP6WRERERLJn+haJrKX0f/+XymuvdZu69/YmjO/ilJTQds019J51Vtr158ybl/Kk2nP06JQsEMEkfHmPKzSpy8Ek8nj0eo+DqdiqQ9SyJRv0tyQiIiKSHdlpH59lprOTqksuoerKK+k97TQOPv44R2++mVBDA9YYQg0NtN5449BfUI1Je/I81U+qu1ev5uCmTexrauLgpk0T9kW+fd06nJKShHk6MZs4er1FUutevZrWG28c2f8IEREREZE8NO1aEvlefJEZF1+M75VXaLviCjouuQQ8nlGNv6FflyeWuhxMLr3eIunl+hhNIiIiIiLjYVoViUqCQSqvugrr99Ny1130vfnNY9qeTqonnk7MJpe6eIiIiIiIiExf06JIZLq7qbjmGsruvpve00/nyG234dTWjsu2dVItIiIiIiIiIlPBlC8SeXfsYOZ//Ae+F16g/TOfof1znwPflD9sEREREREREZERmdLVkuJf/YqqK67AFhVx+Kc/pXfVqmyHJCIiIiIiIiKSk6Zmkainh8ovfpGyn/yE3hUr3O5lU/yKYyIiIiIiIiIiYzElikQljY2xwaOd2lpsQQG+piY6/t//o+3KK6GgINshioiIiIiIiIjktLwvEpU0NiZcht574AAWaL/4YtqvuSa7wYmIiIiIiIiI5AlPtgMYK//69bECUZQBSn7zm+wEJCIiIiIiIiKSh/K+SOTdu3dE80VEREREREREZLC8LxI5lZUp54c1ULWIiIiIiIiISMbyukhU+OijeFpbsZ7Ew3BKSmhfty5LUYmIiIiIiIiI5J+8LRL5XnqJmZ/4BKHjj+fo179OqKEBawyhhgZab7yR7tWrsx2iiIiIiIiIiEjeyMurm3kOH2bmxz6GLSzk8I9/THjuXLo/8pFshyUiIiIiIiIikrfyr0jU28uMf/93vPv30/zznxOeOzfbEYmIiIiIiIiI5L38KhJZS9XatRRt2sTh226j/5RTsh2RiIiIiIiIiMiUkFdjEpV/+9uU3nsvbZdfTs8FF2Q7HBERERERERGRKSNvikTFDzxAxde/Ttfq1XR89rPZDkdEREREREREZErJ6e5mJY2N+Nevx7t3L1hLaNEijt50ExiT7dBERERERERERKaUnG1JVNLYSOXatfj27MFYiwE8+/ZR8uCD2Q5NRERERERERGTKydkikX/9ejzd3QnzPD09+Nevz1JEIiIiIiIiIiJTV84Wibx7945ovoiIiIiIiIiIjF7OFonC9fUjmi8iIiIiIiIiIqOX0cDVgUDgPOBbgBf4QTAYHNTnKxAIBIDrAQs8EwwGPzKWwNrXraNy7dqELmdOSQnt69aNZbN5Yc2v11BQUMBd/3RXtkMRySnZyEXTmXKRSGrKRSKSC5SLRGQiDNuSKBAIeIFbgXcCS4ELA4HA0qRljgWuAs4MBoPLgDFfo7579Wpab7yRUEMD1hhCDQ203ngj3atXj3XTOa1xRyNbD2zl0VcfZeVdK2nc0ZjtkERyQrZy0XSlXCSSmnKRiOQC5SIRmSiZtCRaCewIBoOvAAQCgbuBC4Dtcct8HLg1GAweAQgGgwfHI7ju1aunfFEoXuOORtY+tpY+pw+APR17WPvYWgBWL5k+r4NIGlnLRdONcpHIkJSLRCQXKBeJyITIpEjUALwW97gJODVpmeMAAoHAE7jNHa8PBoO/S95QIBC4GLgYIBgMUlNTMxCIz5fweKTe/r9vB+Chjz406m2MNo7x2vfX7/463aHEK7p1h7q5aetNXHzaxWPa9miN9X2ZSnHkQgzTPA7lokna94333KhcpDjyIoYsxTEpuSgf5cpnYrTyPX7I/2PI9/gnmXJRGlPhc5Tvx5Dv8cPUOIbRyqRIZFLMsym2cyywCpgLPBYIBE4MBoNH4xcKBoO3A7dHt9Hc3Bx7rqamhvjHI9Xf3w8wpm2MJo7GHY082fQkfU4fx3z7GNatWJf2l3ZrLc3dzexq38Xutt0Dt3b3/lD3oZTrvdr2Kt//y/dZXrec+f75GJPqLZkYY31fplIcuRBDLsZRP3mDySsXDWGkuehI7xF2t+3m1fZX2dU2kJN2te9if+f+lOu92vYqtzx+CyvqVrC4ajEeM3nXPsi1z3225UIcuRBDfBxTLRflo1z5TIxWvscP+X8M+R4/oFyUA6bC5yjfjyHf44f8P4ax5KJMikRNwLy4x3OB5OvQNwFPBoPBfmBnIBB4ETchbR51ZCMQHTujz+lj5V0rhzw5Gu/9DuqS8ehaDnUd4vgZx8eKP/GFoK5QV2x9g6G+vJ75/vm8bf7b+M3O39DW1zZoPwbDJRsuAaC2pJbls5ezvHY5K2av4MTqEyn0Fk74sWZTrg6eu+bXawC49z33ZjmSaUO5aIj9pspFh7sP87qZr0uZi5JzTV1pHQv8C3hL/Vv4/e7fp8xFHjxc8dgVAFQVVXFK7SmsmL2CFXUreMOsN1DiK5nwY80m5SKJyPlcJCLTgnKRiEyITIpEm4FjA4HAImAP8GEgeVT8XwIXAncGAoEa3KaNr4xnoOmM59gZIz0B+Nqmrw3ukhHu5oanbog9LvYWM98/nwUVCziz/kwW+BewoMK9zfPPo8hbFFv2jPozWPvY2oRtlvhKWH/mepZWL2Xzgc1sObCFLQe28ODOB2Pbf8OsN7CibgXL65ZzSt0pzCyeOaLjlpHLVjFgmlMuSiNdLrruyetijws8Bczzz2OBfwGn1J7CgooFLKxYyAL/AuZXzE8o8ESPJTkX3fjmGzlp1klsPbCVzQc2s/nAZh5+7WEAfMbH62tez/K65Syvc4vYdaV1IzpuGTnloqzI6VwkItOGcpGITIhhi0TBYDAUCAQ+Dfwety/rj4LB4LZAIHADsCUYDN4fee4dgUBgOxAGrggGgy0TGXjUV576SsqxM77w5y+wwL+AE2aeQGlB6bjsy7EO21q28chrj7CxaSN7O5OL9QMa39PIgooF1JbWZtwlI/rF/rKNl9Hn9NFQ3pDwhX9p9VI+tvRjAOzv3B8rGG05sIXv/f17hGwIgCVVS2JFo+V1y1lcuXhSu6iNp1w8AZrqg/rmaquEXM9FX9301ZS56JonrmFO2RyWzlxKZVHluOzLWsuOozvY0LSBDU0bhsxF97zrHhZWLGRO2Ry8Hm9G2x8uFy2pWsKHjv8QAId7DrP1wNZYLvqf5/+HO567A4D5/vmxPLS8bjmvm/G6jGPINcpFk0+5SEQkPeUiEZkoxtrkrquTxu7dO3Bik67PX7oviUd7j3LL07fw/We/P+RODIZFlYtYVr0s4VZbUptQOGnc0ZjyhKi5u5mNTRvZ0LSBjU0baelx8+qJ1Seyu2037f3tg/bZUN7Apgs3Zf5KpDjmkXZp6A5188yhZ9xf9/dvZuvBrRztdbsbzyia4f6yX7eCFbNXcFLNSRT7ijPabjb7YqZtzfCWGxNOgKy1WCyOdQjbMI51sDbpMZawE8bBwbGDb2EbTlyHwduoqKzg8JHDfOLhT9DcPfg1qSut4/7z76fQW0iRtyh2P97jtkzke5Lu72CoOCL9XfOzCukaUy7qDnVzx7N38PUtXx92R/PK58Vy0NLqpSyrXsbc8rkZ5aLW3lYe3/s4G5s28shrj8QKQ0uqlrC/cz8d/R2D9peNXNQX7uO5lufYcmCL2/px/xYOdrsXU/EX+Dm59mS3aDR7OSfPOpnywvKMtpsPuQhIm1ccHMJOeFCuGvJG0jbitumv9POh+z6UNhf96vxfxXJQkbeIQk/huBfolIvGXUIuykf5Pn5DvscP+X8M+R4/oFyUA6bC5yjfjyHf44f8P4ax5KK8KxL1hnu5c9ud3PK3W2jtbaXEV5Iwzk/U7NLZfO3NX2Nbyza2tWxje8t2drfvjj1fXVwdO1nr6u/i7n/cTW+4N/a8z/iYXTabpo6m2PJnzz2bVXNXcVbDWcwqnTWiE4eRGuuH0rEOLx99OdZFbfOBzbzS6rYuLfAU8Pqa18daG62oW8Gs0lkjimMifuFt7W1lV9sudrXt4pXWV7jtmdtSvrcAhZ7C2EmUY51xi2EiFHgKEopGsZO2aDHJU0ixr5hCTyFFvoETuiJfEUWeIop8kceR9aqrqunr6otto9hbPKgwlbyPTIpVI/08T9UTs0w/82EnzL077uWmLTexr3Mfxd5iesI9g9abUzqHm866yc1Dh7ezrWUbLx99GRsZW7KisCJWNOrp7+HeHfcOykXz/PN4tf1VwjaMv8DPWxreEstHc/1zczoXWWt5rf21WPe0LQe28MLhF7BYPMbDCTNPcAvYkXzUUN6QsuXjZOaizv5OdrXtYmfrTna17eJbT39rSuQin/EN5JNofonPRUn5ZLgcU11ZTV93X+q8Fl0vsq/4PDdcsUq5KH/l+xfqfI8f8v8Y8j1+UJEoF0yFz1G+H0O+xw/5fwxjyUWZjEmUNcnN+98272386bU/8VrHa6yau4rPr/w8Lxx5IeWXyc+f+nneseAdvGPBO2Lz2/raeL7l+YHC0eHt/GjbjxJOyKJCNsTBroNcufxKzpl3Dsuqlw06yR6uS0Y2eYyHY2ccy7EzjuUjr3O7J7d0t7D14FY273dP1O7cfmesJdbCioWxQWiX1y3n+BnHT8iVi1p7W9nZtpNdrbvY2baTna073cdtuzjcczjj7Xz89R/HYzyxm9d4MZjEx8bEpj3GgzEmNu3Bk7B+8i22DgPbqKqqor2tnU8/8umUv97PKJrBF079Aj3hHvrCffSGe2P30Vvy4+i8lu4W+pw+ekI99DmD1x0PwxWrnj30LL1O4r66Q92s37w+Jz7T2ZSciy445gI2NG1g++HtvHHWG/nOOd9hb+felLno6lOv5px553DOvHNi87v6u3jhyAtsb9key0c/e+Fng7qrgZuLmjqa+PQbP805c8/hTbVvwudJTN25nIuMMcyvmM/8ivl84NgPAG4u/uuBv7Ll4BY2799M8B9B7tx+JwCzy2YnFLCXVi+lwFMw7nF19ncm5KLYfdsuDnQdyHg7H3/9x0ecW4wxePDg9XhjyydsI/4WWS45v1VVVfEvv/yX9LnotC/QG+p180nkPlVuSp53pPfIoHwVn4vsoAvnjNxwxapnm5+NdaGLUi4SERERmRw5WyRKNdbCj5//MfVl9dz1zrs4a+5ZgDtOD2R2clRRWMGpc07l1Dmnxub1O/0s+uGilF98+51+LnnTJUPGuXrJan72ws+A3Bs3IVl1SXVC4aw33Muzzc/GxhLZuGcj9+24D3Bfq1NqT+GsRWex1L+Uk2tPjo3tNNzYHEd7jyb8Cv9K6yuxx0d6jyTEVF9Wz8KKhbxz4TtZVLGIRZWL3MF0KxZw9s/PZk/HnkHH0VDewNUrr56olymtaDX5utOuS1kMuOGMGybkBMZaS5/TFztRK6ssY9+hfWkLTtH7jIpVTi+9od7YdCp7O/L716SxSpWLbvv7bcwsmslt597Ge495b0JBNZNcVFpQysm1J3Ny7cmxeWEnzIIfLkiZi0JOiLXL1w4Z54WcwAXO/6GWEiwVdHIC41NeHH8VhRWsmreKVfNWAe7xvXD4hYSWjw+88gDg/m29cdYbOXvR2SyrWMbJtSdTVVQFDJ+LOvo63NwTKUhHWyrubN0Z6wIXVVtSy6LKRayau4qFFQtZVLmIRRVuPjr3vnNzKheBm4+ykYv6nf4R5aJojhmuWJWwblKBKGq65yIRERGRyZCzRaL1m9en/FXdYGIFoqjVS1aP+gtxgaeA+vL6lCcA9eX1GW0j14tD6RR5i2IDyoJ7ArC7fTeb97vdQrYe2MoNj92AxeI1XpZWL2Vm0Uz+su8vCSfMl264lB9v+zFhwuxq3ZWyELSochHvWvQujqk8xj0Bq1g06IpKydatWJfyBGjdinUT8GpkbrJbbRhjYq1+/PipqaihpG/8LzW+8q6VY/o7mKrS5aJiXzEXLL4gYd5YcpHX46WhaBZNvQcHPddQlLo7aFTRjufxP/YHKogM0t/Rjv+xPwDQu+SEUcUzmXweHyfWnMiJNSfyb8v+DXALAtEC9uYDm7npLzcRtmEAjp9xPNXF1Ww+sJl+px+I5KKNl/LT53+KtZadbTs51H0oYT91pXVu0WfeuSysdPNQ9L6soCxtfMpFLmMMhd5CCr2FlFNOTUUNxX2ZjW83EspFIiIiItmTs0WidL8YDnUVn9HK1ROAyWaMYWHFQhZWLOSDx30QAF+5j4eefyg2IPajex4d1NIhZEM8fehpTp9zOu9e9O7YL/CLKhcxzz9vyELQUHK5C00+tSDLlP4OUkuXi/Z17hs0r2jH8/g3/h6cME65n84Vb8msSGMtpruLLzln8Cl+TReh2FOl+Phy30r8jzwI1gHHYqwDjuOu5zgU7GvCOOGETZpQiPI//wmnuASnzI9T5scWFmZ83PHHMnMkxzJO6svrOb/8fM5ffD4AJRUl/PGFP8YG5n/ktUcG5yInxOYDm1lZt5K3zntrrGVi9H6oQtBQcjkX5VMLskwpF4mIiIhkT84WicbaumckcukEINsnZsmqiqsSxlOZe8fclMs51uGed98z7vuPtszIxYHDpkpxKCqX/g5ySaa5KNqaJ1qs8Sa15jG9PXjbW/G0t+Jtb4ubdh+bcIiPcgwe3s4XeJzXaGcefr7Em/mwPQ7nwF6sxwPGAx4P1hjweMAYSCoQRXl6e6j67X2xx05BIU5ZOU6Zn3Dk3r2VEymSr1YAACAASURBVI7c26Jiil5+YchjyYaywjLOrD+TM+vPBNLnImst9733vpTPjUUu5qKif2zD//gf87YFWTrKRSIiIiLZk7NFosn+JTHaMmSkl3seT8OdZOaCySzeyeSbii2kxmrdinWs3Xg53XFjNpV4igblorLNj2FCoYR5JhTCv/F3lD/xMJ6+xPYdTmERYX8F4aqZ9M1bRNhfSdlf/8KFPSdwIYl/7+FyP4c//H/TxjjzrtvxdrQPmh8uLaf93Hfj6WzH09mBp7Mdb+S+8EgLnu5OTNIVLq3XB47jtlZKOpaypzbSu+g48I7vpdRHYyrnItPXh6e7E09X5NbdmfDYdHdCTw/+zvZBl6wwoRD+R39Pwb4mwv4KHH8l4cjNFpe4RcU8oFwkIiIikh05WyTKxi+J977n3uz9SmwtZU89mvIks2zzYzlTJFI3gKlPJ2SJLuQE/PbtXMvGWOueG+zZnN9ej/OPbfgOH8LXcghPiiINAI5D75LXxU7UoyfttmjwWC62qNgtFMflAevz0bniLUPG2LniLanXO/Us+uekbnETjc3T1RkrInk72/F0tlPy7NaUi3u7Oqn50Tfdlkf+SpzySsIVFYTLK3H8Fe68Mr/bwmmCTXYuGnMrT8fB0901UOxJKAJ14enqiNx3YkL9g1a3xoNTWopTUua+xvMWwt82p95XOEzRrpfw9CSOpWV9PsLllXHFo4qB99Ff4X4mc6iIpFwkIiIiMvlytkgEYxsEdrKMaAySUH+sm4m37ajb1aTtaORxKyYcSrmap6OdsqcepX9WHaGaOhx/Zda+yKsbgGvUY89I3inb/BgfscfxEY4bmGmBLU+4k14foRnVWF9BypN7p9xPx5lvy2hf0c9Q2ebH8HS0Z/zZGu16eDw45X6ccn/C7KKd/0jZMskpKqZ72ZvwtrfhaW+lYP9rFL3ckdAayRrjdmfzVya0ZHHKIwWJ0rJxKSJNZi5K28rTWvoWLE4s+kQLPl2JBSHT3TWo1Q+4Lcqc0jKckjL6Z82OTTulZXHTpdiixFZANTU1ODteSP0+lfs5fOHFboukjrbEbo4d7nTBgb2DW7cVFCYVjyoS3kdbWDSur6uIiIiI5J6cLhLlupQnDo/+Ac/RwziVMxIKQJ72VrxdHQnrW18B4YpKwhVV9DUspPil5/D0phhy1OOh5LmtlDpu9w+nqJhQdS2hGrdo1D9rcgtH+VC8m0j50C1Qxk+6FkIWOLLmIsKVM8DjGfhcjLAVULLeJSfQu+QEampqODyCVo3R9cZDupZJHWecO3gf4bDbjS1ShPC0t+KNFCYKm3YPznseD+HyioGWR+UDRYmwvxJbUppxLruQE7iYzwBhHPzjM2izE8bT083/z97dR1l21nWi/+6u6rekO+kkRQKdBIwSHVpwgUAYlhPwBTCZwQS4mQeCOiBomDui4yUOE0eGwTizFsJSh7sGxchIlDuAD5loWIxMYAbioAKGiIoQ0LwA6SSEdBLy2m9Vte8f53R3daWq+3S97l39+axVq/vss88+3+fUrl/3+dWzn9Ps3p11ex7Luj27s+Uv/vfclxLe8LE5Gz/tunUHGz1TW0/K/tOfdFjDZ/qELYO/bz4hGV/4P8PzziAbnnPthg2ZOnUiU6dOzPn4Q+tkDRtJjzx08N+sDXd+43FNz+mNmwYNo+H3LE/cng3rxjM1bCZl/foFjwUAgG7QJBoa+VKC6enBf6Qf/Ha2/MUnH//GYWoyW77w2SSDN5HTJ27N1EknZ99ZT8n01m2DptDWQWNo9voQk084Y87/8D98/kuy95xzM37/fRnfdc/w65vZ/Hc3pelI42ita/btS+67N+vv3jn3971jlwWydKa3bJ13tsbUKacdvH3ge78WZpgd08yksbFMn7Qt0ydty+PnUSWZnMzYIw8dakDMaCRt/PqtWbf7scN2b8fGh42IQ7NYpraenOzbnWaqPVg3R2rWtm0yuT/rdu/Our27s273Y2n27M66PYMG0MxG0MHt+46tzfTI8144bAAdaP6csGKXbS14BtlQu3FTJjduSibOmOPONs3e3Ye+XzMWXB/79n3ZcMftab54U06e8ZDpTZtnXFY5/P4duBRxy0nH1BAzWxMAYHVoEmW+GUHXZ3zXtwZvEB98YHB52EPfHnwK0awFXWdrkzzwz39q8JvWsdFf4qP9h3/yCWdk8gkz/jM/NXnkxtGGjQebRhpHc2jbNPv3DdcFeWTuPx99NOt2P5J1+wdvf7cd4XDrHnk4J338jzN56hMydcppgz+Hs0zor6PN1phpKWfzrLaFzmh6nPHxTG07NVPbTp2nibT/8E97O3h51EPZeO83s27vnoO7TmQ4A3PrSXNeontg0ebNf/v5YSNo97yX8bbr1mV60wlpN23O9OZBc2N60+bB7U0nDP6+efD3kz92TcYefeRxx5jesjW7v+85C39tlsCynXNNk3bTCZncdELyhCc+/v62zcTmTfn2128/1Px7ZNBMGt91T8a+9g8H/y06YOqELTOaR7PXRdqarBssiG62JgDA6tEkyjyfSjQ1lRO++Pkkg3Uapk4+JZMTp2fvd35Ppoa/Nd/6yf/xuEspkuEMg22nLijLMb0xGxsfoXF0T68bRwtaLPZg8+eRYZPn0cEnO81qAo3Nt0Ds+PjgzcwJJ2Zy4vRMn3BOpk/YkhPPeGIenGqz9YY/ydhjj875uLEHH8iGb9x2cI2Wdt1YpradmslTJzJ5yuCyj8lTJwYLz3b0Nedwi52twVGMr8/UKacdNitrpmbf3qx7+KGcsq7No3fuPLS2zgP3zX28qalMn7AlU6c9IdObNg+/hs2gTZsHM302bU67fsPIP4OPnveCJbmUcE1pmmTL1kyesT2TZ8zxiXLDRdFnXsZ24Hu3/pt3ZuOjX5ljLastmdpycsZ33WO2JgDAKjm+m0Rtm/V3feOIa47c95P/9+MWDD3g0ed18I3DohpHp2dy4olHbRyt1GUA86759OjDmZw4Y86ZP2PDv89+g5EMZyAcaP484Yzs27wl0yeeOLhE5MAisSdsmffN44kTE9m/a1cefd4L516v5fyXDF6HqcmMffv+jN+/K+P378rYA7uy/u6d2XTLzQf3n96w8eBso8lTJzJ1yqB5NNcnXrH61tIMob5pN2zM1GlPSCYmsvuU0w9uP/WDV817GeBDF7x8STNoFC7APIuiHzQ9fXAtqwNrIh2YUTZX8z6Zf30wAACWzvHZJJqazMZbv5oTvnhTxu+/N23TDNaumGV6y9a0m06Y9zC9eeMwcuPorx7XOMrZ35GNJ24dfOrO1pOz8davLPwygOnpNPv3Db727Uuzf2+affvT7N+bdfsObN978P6Nt94895pPf/npww+7fv3BRs/BTwc62Pg59Ge7YcMiX8gcNs55v+9j45k67fRMnXb6YYvoNnv3ZOyBQeNo/IFdGbt/Vzbe+pVsvvnQXlMnbjnYMDo482jbaYta3BbWomO5DHApLNnldwysW5fprScPfiExy5EagAAALK/j6p1ns2d3Nt/8N9n0pb/O2O5HM3nKaXn4BS9JmyZbZ316zahvNno7w2DOxtFUxh/YlfF7DzWOcuOf56RhQ2h6w8Y0U5NppqYOO1QzOZktf/a/sv6eu2Y0gGY0fPbvGzSB5vnt8Gzt+PpMb9gw52ygZDDD68GXluEnBS1d8+dYLOT73m7clMknnpXJJ541Y2ObdY8+crBpNJh5dG82f+mOg69z2zSDyx2HTaM85TuzbnxDpk/a5pI1jltraaFwDrfSDUAAAA45LppEY9++P5v/7q+y6e+/lGZqMvvO+o48/IwLsv/Mpxx6kz021v0ZQcttbOzgWkUHTJyyLQ/8w1cPNo42feVv53xos39fNt76lbTrN6TdsCHt+g2DtT9O2nbYtnbDxsF9GzakXb9xxvbhn+s3HFzo+Ui/Td7/pLOX5zVYaU2T6S1bs2/L1uTscw5tn57O2EMPHGwcjd+/K+vv+1Y23v73aW76i5yWwZumyW2nDS5XG162NnnKxDF9hDj0WW+b9ByRBiAAwOpZu02its36u+/I5i/elI3fuC3t2Fj2PHVHdj/9+wezMWZxKcE8xsYPaxxt2Hn7vI2b+y+9bEmf+rj+bfK6dZnadlqmtp2Wfd/5PYe279+fiXYyj9x+y6CB9MCubLzj9qz7+y8d3GV60+ZMDi9Zmxo2jiZPnUjWjzbj6uSP/mGS5MGXvnJJhwQwKg1AAIDVsSaaRIctpHzi1ux98jlZ/627s/6+ezO9aXMe/f7nZ/eOZw5mWLAoK9m46c2aTytp/fpk4knZs2HzYZub3Y8dttbR+AO7svmrf3fYJX5TW08+fK2jUyYyte2Ugx87nQx+ltbfc/fg0+Q+eJXXGwAA4DjS+ybR4z4B69GHc8LNf5upzSfm4fNfkj1PfZpFf5fQSl8GYIbXaNrNJ2T/mU/O/jOfPGNjm3UPPzhYZ+r+Q2sebfjGbQc/erpdty5T207N5CkTyfR0Nn791oUtSg4AAEDv9b57cuKNn557geOxddnzj56x8oGOAy4D6ImmyfRJ27LvpG3Z95SnHto+NZmxbz8wbB7dm7H7d2X9PXfOeRlhMzmZE2/8tO83AADAcaD3TaJ1c7yxPdJ2OO6NjWfqtCdk6rQnZG8ONX8mfvfXM9dy136WAAAAjg/rVjvAYjS7H5v3U5ymt2xd4TTQb/P9zPhZAgAAOD70t0k0uT8nf/yPkyTt2Nhhdx03n4AFS+jR556fdtb6XX6WAAAAjh/9vNysbXPSDR/L+LfuzkMv+rE0U1M+AQsWyafJAQAAHN962SQ68XN/mo23/0Meed4Ls++c707i05dgKViUHAAA4PjVu8vNNn3pCznhizdl945nZvcznr3acQAAAADWhF41iTZ8/dZs+cynsvfJ35lHnv9D8y5aDQAAAMCx6U2TaPzeb+akT340k6ednod++KXJut5EBwAAAOi8Tq9JtPGWm7P1T69PpqeSpkm7YWMe/NGXJ+vXr3Y0AAAAgDWls9NxNt5yc7Z++uNppqfSJGnaNs3k/my46xurHQ0AAABgzelsk+jEGz+dZnLysG0HPuoeAAAAgKXV2SbRukcePqbtAAAAACxcZ5tE01u2HtN2AAAAABaus02iR597ftrxw9fVbsfH8+hzz1+lRAAAAABr10ifblZKuSDJu5KMJXlvrfXt8+x3SZIPJ3lurfXziwm296lPSzJYm2jdIw9nesvWPPrc8w9uB44/q1GLAGZTi4AuUIuA5XDUmUSllLEk705yYZIdSS4tpeyYY7+tSX4+yeeWKtzepz4t9196WXb9zOW5/9LLNIjgOLaatQjgALUI6AK1CFguo1xudl6SW2qtt9Va9yX5UJKL59jvV5O8I8meJcwHcIBaBHSBWgR0gVoELItRLjc7M8kdM27vTPK8mTuUUp6V5Oxa60dLKb8434FKKZcluSxJaq2ZmJg4FGR8/LDbq0UOObqc4TjPoRbJsdox5OhYhlXKsSK1qI+6ck4sVN/zJ/0fQ9/zrzC1aB5r4Tzq+xj6nj9ZG2NYqFGaRM0c29oDfymlrEvym0lee7QD1VqvSnLVgWPs2rXr4H0TExOZeXu1yCFHlzN0Mcf27dtX6inVIjlWO4YcHcswM8daq0V91JVzYqH6nj/p/xj6nj+JWtQBa+E86vsY+p4/6f8YFlOLRrncbGeSs2fcPivJXTNub03y9CQ3lFK+luQfJ/lIKeU5C04F8HhqEdAFahHQBWoRsCxGmUl0Y5JzSynnJLkzyauSvPrAnbXWB5McnIdVSrkhyS9aOR9YYmoR0AVqEdAFahGwLI46k6jWOpnkjUmuT3LzYFP9UinlylLKRcsdECBRi4BuUIuALlCLgOXStG179L2WR3vXXYdmRHblmj855Ohyhi7mGF7vOtd18X2hFskhR08zzMyx1mpRH3XlnFiovudP+j+GvudPohZ1wFo4j/o+hr7nT/o/hsXUolHWJAIAAABgjdMkAgAAAECTCAAAAABNIgAAAACiSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAaBIBAAAAEE0iAAAAAKJJBAAAAEA0iQAAAABIMr7aAWZq2zZ79uzJ9PR0mqZZlQz33HNP9u7de8yPa9s269aty6ZNm1YtOwAAAMBCdapJtGfPnqxfvz7j46sXa3x8PGNjYwt67OTkZPbs2ZPNmzcvcSoAAACA5dWpy82mp6dXtUG0WOPj45menl7tGAAAAADHrFNNooVcpnXaJZfktEsuWYY0C+NSMwAAAKCPOtUkAgAAAGB19LpJtPnaa7Phppuy4TOfyennnZfN1167qONdcskl+dSnPnXYtt/93d/NT/7kT+a7vuu78uIXv/jg14c//OFFPRcAAABAl/R2AaDN116bk9/85jT79iVJxu+8Mye/+c1Jkt2veMWCjnnxxRfnj//4j3P++ecf3Hbdddfl3//7f58777wzn/jEJxYfHAAAAKCDOtskOumtb836L3953vs33HTTwQbRAet27862yy/PCR/4wJyP2b9jRx668sp5j/nP/tk/yzvf+c7s3bs3GzduzB133JF77rknT3rSkxY2CAAAAICe6O/lZrMaREfdPoJTTz01z3rWs3LDDTckGcwiuuiii9I0Tb7+9a8fdrnZ5z73uQU/DwAAAEDXdHYm0ZFm/CTJ6eedl/E773zc9qkzz8x911yz4Od9+ctfnuuuuy4/+qM/muuuuy6/8Ru/kSR5ylOe4nIzAAAAYM3q7Uyih6+4ItObNx+2bXrz5jx8xRWLOu6FF16YP/uzP8sXv/jF7NmzJ894xjMWdTwAAACAPujsTKKjObA49bbLL0/27cvUmWfm4SuuWPCi1QeceOKJef7zn583velNednLXrYUUQEAAAA6r7dNomTQKDqwSPViLjGb7WUve1l++qd/Or/92799cNuBNYkOeNWrXpXXv/71S/acAAAAAKup102iZGmbQwdceOGFuXPGekdnn312br311iV/HgAAAICu6O2aRAAAAAAsHU0iAAAAADSJAAAAANAkAgAAACBroEl08kf/MCd/9A9XOwYAAABAr/W+SQQAAADA4vW6SbTxlpuz/p67s/7unTn1g1dl4y03L/qYZ5xxRn7lV37l4O33vOc9+fVf//VFHxcAAACgy8ZH2amUckGSdyUZS/LeWuvbZ93/piQ/nWQyyb1JXldr/foSZz3MxltuztZPfzzN9FSSZOyRh7P10x9Pkux96tMWftyNG/Oxj30sP/dzP5dTTz11SbICS6OLtQg4/qhFQBeoRcByOOpMolLKWJJ3J7kwyY4kl5ZSdsza7QtJnlNr/b4k1yR5x2KDnfiZTx1cb2iur61/en2aycnDHtNMTmbrn14/72NO/Mynjvq8Y2Nj+fEf//FcddVVix0CsIRWqxYBzKQWAV2gFgHLZZSZROcluaXWeluSlFI+lOTiJF8+sEOtdWb35bNJfmIpQ85pOINo5O3H4LWvfW1e9KIX5V/9q3+16GMBS6abtQg43qhFQBeoRcCyGKVJdGaSO2bc3pnkeUfY//VJPjbXHaWUy5JcliS11kxMTBwKMj6ejRs3Znx8EGnv+S/O3iM8yUnv/+2MPfLQ47ZPbzkpj77sx+d93CgDPuWUU1JKydVXX51NmzZl3bp1B3MdzcaNGw8b10KNj48vyXHkWFsZjvMcK1aLjtPXVw45epdhlXKsSC3qo66cEwvV9/xJ/8fQ9/wrTC2ax1o4j/o+hr7nT9bGGBZqlM5HM8e2dq4dSyk/keQ5SV441/211quSHLiOq921a9fB+yYmJrJ3796MjY2NECl59Ln/ZLAm0YxLztrx8Tz63H+SyVmXoR2rycnJvO51r8sFF1yQV77ylWnbduRj7t27NzPHtVATExNLchw51laGLubYvn37Sj3litWiLr2+q00OObqcYWaOtVaL+qgr58RC9T1/0v8x9D1/ErWoA9bCedT3MfQ9f9L/MSymFo3y6WY7k5w94/ZZSe6avVMp5UVJfjnJRbXWI00CWhJ7n/q0PHz+S9KuG0ubZGrL1jx8/ksWtWj1TKecckp+7Md+LB/84AeX5HjAonWyFgHHHbUI6AK1CFgWo8wkujHJuaWUc5LcmeRVSV49c4dSyrOS/E6SC2qt31rylPPY+9SnZdNX/jZJ8uBLX7nkx3/DG96Q973vfUt+XGBBOluLgOOKWgR0gVoELIujziSqtU4meWOS65PcPNhUv1RKubKUctFwt3cm2ZLkw6WUvy6lfGTZEs/y4EtfuaQNottvv/3g35/whCfk1ltvzeWXX75kxwcWpuu1CDg+qEVAF6hFwHJp2nbOS1dXQnvXXYdmRE5MTOQb3/hGTjjhhNXKk2SwQNVi1jR67LHHlmQMXbkGUo5uZehijuH1rnNdF98Xj6tFXXp9V5sccnQ5w8wca60W9VFXzomF6nv+pP9j6Hv+JGpRB6yF86jvY+h7/qT/Y1hMLRplTSIAAAAA1jhNIgAAAAA0iQAAAABYA02iSy45LZdcctpqxwAAAADotd43iQAAAABYvF43ia69dnNuumlDPvOZDTnvvNNz7bWbF33Mc845ZwmSAQAAAPTL+GoHWKhrr92cN7/55OzbN/hUtzvvHM+b33xykuQVr9i9mtEAAAAAeqezTaK3vvWkfPnL6+e9/6abNhxsEB2we/e6XH75tnzgAyfM+ZgdO/bnyisfWtKcAAAAAGtBby8327fv2LYDAAAAML/OziQ62oyf8847PXfe+fj4Z545lWuuuW+5YgEAAACsSb2dSXTFFQ9n8+bpw7Zt3jydK654eJUSAQAAAPRXZ2cSHc2Bxakvv3xb9u0bzCC64oqHLVoNAAAAsAC9bRIlg0bRgUWql+oSs927d+fZz372wduXXXZZ3vCGNyzJsQEAAAC6qtdNomTpmkMHfPOb38zk5OSSHhMAAACg63q7JhEAAAAAS0eTCAAAAABNIgAAAAA0iQAAAADIGmgSXfLRS3LJRy9Z7RgAAAAAvdb7JhEAAAAAi9frJtG1t1ybm+65KZ+5+zM574Pn5dpbrl30MZ/0pCflxS9+cX74h384r3nNa/Lggw8uQVIAAACAbuttk+jaW67Nmz/95uyb3pckufORO/PmT7950Y2iTZs25ROf+EQ++clPZtu2bbn66quXIC0AAABAt42vdoD5vPUzb82X7/vyvPffdM9NBxtEB+ye3J3L//TyfOArH5jzMTtO25Ern3/lyBme/exn5+abbx55fwAAAIC+6u1MotkNoqNtP1ZTU1P5sz/7s7zkJS9ZkuMBAAAAdFlnZxIdbcbPeR88L3c+cufjtp+55cxc89JrFvy8e/bsyYtf/OLs3Lkzz3jGM/KCF7xgwccCAAAA6IveziS64rlXZPP45sO2bR7fnCuee8WijntgTaLPfe5z2b9/vzWJAAAAgONCb5tEr3jqK/KO89+RDes2JBnMIHrH+e/IK576iiU5/kknnZRf/dVfzXve857s379/SY4JAAAA0FW9bRIlg0bRs894dp7/pOfnLy/9yyVrEB3w9Kc/PTt27Mh11123pMcFAAAA6JrOrkk0qsWsPzSX22+/PZOTkwdv//7v//6SHh8AAACgi3o9kwgAAACApaFJBAAAAEC3mkRt2652hEVbC2MAAAAAjj+dahKtW7fusPWA+mZycjLr1nXqJQUAAAAYSacWrt60aVP27NmTvXv3pmmaVcmwcePG7N2795gf17Zt1q1bl02bNi1DKgAAAIDl1akmUdM02bx586pmmJiYyK5du1Y1AwAAAMBKG6lJVEq5IMm7kowleW+t9e2z7t+Y5A+SPDvJfUleWWv92tJGBY53ahHQBWoR0AVqEbAcjrqATillLMm7k1yYZEeSS0spO2bt9vokD9Ran5rkN5P82lIHBY5vahHQBWoR0AVqEbBcRlll+bwkt9Rab6u17kvyoSQXz9rn4iS/P/z7NUl+pJSyOosKAWuVWgR0gVoEdIFaBCyLUS43OzPJHTNu70zyvPn2qbVOllIeTHJaksMW9ymlXJbksuF+2b59+2EHmX17tchxODm6lSE5bnOoRatEjsPJ0a0MydqtRX3U9zH0PX/S/zH0Pf8KUouOwBhWX9/zJ2tjDAsxykyiubrN7QL2Sa31qlrrc2qtzxk+5uBXKeWm2dtW40sOObqcocM5VsJcz6MWySHHcZxhjhwrYa7nWXQt6srruZivvo+h7/nXwhj6nn/GGFZCM8c2tWhtnUernuN4zb8WxrCYWjRKk2hnkrNn3D4ryV3z7VNKGU9ycpL7FxoKYA5qEdAFahHQBWoRsCxGudzsxiTnllLOSXJnklclefWsfT6S5DVJPpPkkiSfrLU+rksNsAhqEdAFahHQBWoRsCyOOpOo1jqZ5I1Jrk9y82BT/VIp5cpSykXD3f5rktNKKbckeVOSKxaQ5aoFPGY5yHE4OQ7pQobkOM2hFq0aOQ4nxyFdyJCsnVrUlddzMfo+hr7nT/o/hr7nT1ZoDGrRERnD6ut7/qT/Y1hw/qZtNZMBAAAAjnejrEkEAAAAwBqnSQQAAADASAtXL7tSygVJ3pVkLMl7a61vX6HnPTvJHyR5YpLpJFfVWt9VSnlbkp9Jcu9w139Xa/2TZczxtSQPJ5lKMjn8+MlTk/xhku9I8rUkpdb6wDJm+J7h8x3wnUnemmRblvm1KKX8XpKXJvlWrfXpw21zjr+U0mRwrvzTJI8leW2t9a+WMcc7k/xYkn1Jbk3yU7XWb5dSviOD67+/Onz4Z2ut/3IZc7wt83wfSim/lOT1GZw/P19rvX4Zc/xhku8Z7rItybdrrc9cztdjJR3vtWiY5WtZxXqkFqlFI+bobS06Wp0ppWzMoB48O8l9SV5Za/3aSuc8khHG8KYkP51kMoNzryHwIQAAIABJREFU5XW11q+veNB5jFrrSymXJPlwkufWWj+/ghGPapQxlFJKkrdl8JHnf1Nrnb2o8aoZ4Rx6cpLfz+DneyzJFcv979+xmKsuzbp/2f6NWCpq0epTi1afWjS3VZ9JVEoZS/LuJBcm2ZHk0lLKjhV6+skkl9dan5bkHyf52RnP/Zu11mcOv1biRPih4XM9Z3j7iiT/u9Z6bpL/nYUtwDuyWutXD4w3g38MHkvyR8O7l/u1uDrJBbO2zTf+C5OcO/y6LMlvL3OOTyR5eq31+5L8fZJfmnHfrTNel6V8EzJXjmSO78PwfH1Vku8dPua3hj9Ty5Kj1vrKGefJf09y7Yy7l+v1WBFq0WFWrR6pRfPmUItm6GstGrHOvD7JA7XWpyb5zSS/trIpj2zEMXwhyXOG5+s1Sd6xsinnN2qtL6VsTfLzST63sgmPbpQxlFLOzaBO/ECt9XuT/MKKB53HiN+Dt2SwEPOzMqgtv7WyKY/q6sxdHw9Yzn8jFk0tWn1q0epTi+a36k2iJOcluaXWelutdV+SDyW5eCWeuNZ694FOWq314Qx++3jmSjz3CC7OoGuZ4Z8vW8Hn/pEM/pO9Ip32Wuv/SXL/rM3zjf/iJH9Qa21rrZ9Nsq2U8qTlylFr/XgdfHpEknw2yVlL8VzHmuMILk7yoVrr3lrr7UluyeBnallzDLvSJckHl+K5OkItmt9q1SO16NA2tWgOPaxFo9SZmefcNUl+ZDjOrjjqGGqtn6q1Pja8uSLn6zEYtdb/agZvKPesZLgRjTKGn0ny7jqc9Vlr/dYKZzySUfK3SU4a/v3kJHetYL6jGqE+Ltu/EUtELVp9atHqU4vm0YUm0ZlJ7phxe2dW4c3RcIr6s3KoS/vGUsrfllJ+r5RyyjI/fZvk46WUm0oplw23nVFrvTsZvIFMcvoyZ5jpVTn8P9wr+VocMN/4V/N8eV2Sj824fU4p5QullD8tpZy/As8/1/dhtV6P85PcU2v9hxnbVvr1WGpq0UCX6pFaNDe16JC+1aJRXqeD+wwbgw8mOW1F0o3mWL/Xr8/h5+tqO2r+Usqzkpxda/3oSgY7BqN8D747yXeXUv68lPLZ4SUVXTFK/rcl+YlSys4kf5Lk51Ym2pLpxP8pjkAtWn1q0epTi+bRhSbRXB3pdiUDlFK2ZDBd/RdqrQ9lMA3ru5I8M8ndSX59mSP8QK31+zOYDvazpZQXLPPzzauUsiHJRRlc95qs/GtxNKtyvpRSfjmDS4L+23DT3UmePJx6+KYkHyilnDTf45fAfN+H1fr5uTSHv3lf6ddjOahFA52oR2rR3NSix+lbLRrldVr1WnQUI+crpfxEkuckeeeyJjo2R8xfSlmXwaU1l69YomM3yvdgPIPLC34wg5+T95ZSti1zrlGNkv/SJFfXWs/KYC2N9w+/N32xFn6O18IYkqhFy0gt6r4F/Rx3YYA7k5w94/ZZWcFpXKWU9Rm8KftvtdZrk6TWek+tdarWOp3kd7NEU+bnU2u9a/jntzJYe+O8JPccmAo2/HOlpuZdmOSvaq33DDOt6Gsxw3zjX/HzpZTymgwWBPvxWmubJMNLKu4b/v2mDBaS/e7lynCE78NqvB7jSV6RGYsLr/TrsUyO+1o0fM6u1CO1aBa16HA9rUWjvE4H9xmO8eSMftnfShjpe11KeVGSX05yUa117wplG8XR8m9N8vQkN5TBQv7/OMlHSinPSXeMeh5dV2vdP7wE9KsZvFHrglHyvz5JTZJa62eSbEoysSLplsaq/p9iBGrR6lOLVp9aNI8ufLrZjUnOLaWck+TODC4vWJEVz4fX1f7XJDfXWn9jxvYnHbi8IMnLk/zdMmY4Mcm6WuvDw7+/JMmVST6S5DVJ3j7887rlyjDLYb+VXcnXYpb5xv+RDC51+FCS5yV5cEa+JTecEvlvk7xwxjXNKaU8Icn9tdapUsp3ZlDsblvGHPN9Hz6SwW/KfyPJ9mGOv1yuHEMvSvKVWuvOGflW9PVYJsd1LRo+X5fqkVo0g1o0pz7WolHqzIFz7jNJLknyyQNNwY446hiGl0j8TpILOrb+RHKU/LXWBzPjDUAp5YYkv1i79YlCo5xHf5zhb8BLKRMZNEu78rMwSv5vZLAu3dWllKdl8Mbs3vTHiv4bsQBq0epTi1afWjSPVW8S1VonSylvTHJ9Bh8r93u11i+t0NP/QJKfTPLFUspfD7f9uwxWNn9mBlOxvpbkDcuY4Ywkf1RKSQbfjw/UWv9nKeXGJLWU8voMTs5/vowZkiSllBOSvDiHj/cdy/1alFI+mMEUxInh9Z7/IYM3ZHON/08ymOp3SwafevRTy5zjl5JsTPKJ4ffowMcpvyDJlaWUyQw+7vlf1lqX5Lcr8+T4wbm+D7XWL5VSapIvZ3AJys/WWqeWK0et9b/m8evEJMv4eqwUtShJR+qRWqQWHS1HX2vRfHWmlHJlks/XWj+SQcP4/aWUWzL4rf2rVi/x4404hncm2ZLkw8Pz9Ru11otWLfQMI+bvtBHHcH2Sl5RSvpzBz8K/OTDLbrWNmP/yJL9bSvl/Mqg3r+1Sg2Ke+rg+SWqt78ky/huxFNSi1acWrT61aH5N23ZmjAAAAACski6sSQQAAADAKtMkAgAAAECTCAAAAABNIgAAAACiSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAaBIBAAAAEE0iVlDTNDc0TfPeJTrWM5qm+cumafY0TfO14ba2aZqfWIrjA2ubegQAAI+nSbQGNU1z9fANSts0zWTTNF9vmuY9TdOcNmOfI76BaZrmnzRN8/Gmae4dvvH5etM01zRN85RFRHtFkjfNeI7/1TTN1Qs81juSPJTkHyV57nDbk5JcM+P4k03TvHaBxwcWaJQaNNzv+U3TXNs0zT3DOnNr0zT/X9M03z9jn3bG12NN03y5aZo3Pf5Zj5l6BAAAs2gSrV2fzuBNynck+fkk/1eSPxjlgU3TPC3JJ5L8Q5IXJXlaktcm+VqSkxYaqG3b+9u2fWihj5/l3CR/2rbt19q2vXd4/G+2bbtniY4PLM4Ra1DTND813Gd/kh/PoM68MoM6865Zx3rj8FjfO7zv15qmuWwx4dQjAAB4PE2itWvf8E3KzrZtr0vyn5Nc0DTN5hEe+6NJHmnb9mfbtv2btm1vb9v2U23b/mLbtl+c70FN05zVNM1/b5pmV9M0u5umua1pmn8z4/6Dl3cMf2P/I0leM2OWwA8O7ztjOBPh3qZpHm6a5s+bpnnB8L7vaJqmTfJdSa4cPu5tw/sOzo4aXvIxluR9B45/bC8fsEjz1qCmabYn+e0k723b9pVt2/6vYZ35fNu2b0ly0axjPTg81u1t2/5Okr/NoE7NSz0CAIBjN77aAVgxuzNoCo7yPb87ySlN01zYtu3HjuE5fivJCRnMPvp2knOSPHGeff91ku8cPte/Hm67f9jE+lSSm5NcODzOK5N8ommaZyb5+wxmFNyY5L9l8MbzkTmO/9zhsS9P8ofHMAZgecysQSXJxiT/ca4d27Z9YK7tTdM0SX44g1lHf3+U51OPAADgGGkSHQeaptmR5GeTfK5t24dHeMiHk7w4yf9omuaBDN4AfSrJB9q2veMIj3tKkj9q2/avh7e/Nt+Obds+2DTNviS727b95oysr83gkrZXtm07Odz8n5qm+ZEkb2jb9heSfLNpmqkMZjt9c/axh8e/d/B+cjAD4agjBpbN7BrUNM13J3mobdudIx7ivU3TvCeDxtJ4kscyaMgciXoEAADHyOVma9cPNk3zSNM0u5P8XZLbkrx6lAe2bTvdtu1PJ9mewVogX07yhiQ3H7gEYx7/Ocm/a5rmc03T/NqBSzKO0XMz+G3/t4f5H2ma5pEk52ew7gfQD0eqQc0xHuuXkzwzyQ8l+fMkb2nb9nNHeYx6BAAAx8hMorXrc0lek2Qyyd1t2+491gMMf+P9wSQfbJrmiiRfSPIfktwwz/7va5rmfya5IIM3cx9rmuaP2rY9lo+BXpfBpR0vn+O+x47hOMDqOlIN+mqSk5qmOWvE2UT3tG17S5JbmqZ5eZJ/aJrmC23b3jDfA9QjAAA4dmYSrV2727a9ZfhpO8fcIJqtbdt9GcwEOP0o+93dtu372rb9F0len+THm6aZ7xPR9mWwmOtMn89gbZCHhvlnft11jLHnOj6wMo5Ugz6cZG+St8z1wKZpTpnvoMNPD/utJO8arlE0L/UIAACOjSbR8e3JTdM8c9bX6U3TvKFpmt9pmuZHm6Z5atM0T2ua5t9msHDrH813sKZp/kvTNP+0aZrvaprme5O8IskdSeZbB+n2JM8e7j/RNM36DBZ/vT2D9ZBeMvz0oOc1TfNLTdO87BjHd3uSH2qaZnvTNBPH+FhgmbRte2cGl7L+TNM0H2qa5keGP+vf3zTNryS57iiH+H+TfE+OcAmtegQAAMdOk+j49p8yuIRs5tfrkvxlBgvEvjuDj5r+iww+jegXkrz1CMdrMlgH5O+S/J8kJya5sG3b+T7u+deT7EryN0nuTfIDbdvuSfLCDH6D/74MPj3o2iTnJfn6MY7v8iTPzuDN2b3H+FhgGbVt+94MftY3ZXBZ61eTXJPBp5D9/FEe+80k78/gY+fXz7ObegQAAMeomf//ywAAAAAcL8wkAgAAAECTCAAAAABNIgAAAACiSQQAAABAkvFVfG4rZsPa0ax2gEVQi2DtUIuALlCLgC5YUC1azSZR7rrrroN/n5iYyK5du1YxjRxydD9DF3Ns3759taMsmlokhxz9zDAzx1qrRX3UlXNiofqeP+n/GPqeP4la1AFr4Tzq+xj6nj/p/xgWU4uO2iQqpfxekpcm+Vat9elz3N8keVeSf5rksSSvrbX+1YITAcxBLQK6Qj0CukAtApbDKGsSXZ3kgiPcf2GSc4dflyX57cXHAnicq6MWAd1wddQjYPVdHbUIWGJHbRLVWv9PkvuPsMvFSf6g1trWWj+bZFsp5UlLFRAgUYuA7lCPgC5Qi4DlsBRrEp2Z5I4Zt3cOt909e8dSymUZdLFTa83ExMShIOPjh91eLXLI0eUMchyRWiSHHMdZhi7lmGWkenSkWtRHHf1ejKzv+ZP+j6Hv+TtILeqpvo+h7/mTtTGGhVqKJtFcK2bPuSp+rfWqJFcd2GfmQlBdWRhKDjm6nKGLOTq0QKNaJIccx1mGmTk6VIuSEevRkWpRH3XlnFiovudP+j+GvudPOrdwtVrUU30fQ9/zJ/0fw2Jq0ShrEh3NziRnz7h9VpJ+L4kP9JFaBHSFegR0gVoEHLOlmEn0kSRvLKV8KMnzkjxYa33c5R0Ay0wtArpCPQK6QC0CjtlRm0SllA8m+cEkE6WUnUn+Q5L1SVJrfU+SP8ngYxVvyeCjFX9qucICxy+1COgK9QjoArUIWA5HbRLVWi89yv1tkp9dskQAc1CLgK5Qj4AuUIuA5bAUaxIBAAAA0HOaRAAAAABoEgEAAACgSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAaBIBAAAAEE0iAAAAAKJJBAAAAEA0iQAAAACIJhEAAAAA0SQCAAAAIJpEAAAAAESTCAAAAIBoEgEAAAAQTSIAAAAAokkEAAAAQDSJAAAAAIgmEQAAAADRJAIAAAAgmkQAAAAARJMIAAAAgGgSAQAAABBNIgAAAACiSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAaBIBAAAAEE0iAAAAAJKMj7JTKeWCJO9KMpbkvbXWt8+6/8lJfj/JtuE+V9Ra/2SJswLHObUI6AK1COgCtQhYDkedSVRKGUvy7iQXJtmR5NJSyo5Zu70lSa21PivJq5L81lIHBY5vahHQBWoR0AVqEbBcRrnc7Lwkt9Rab6u17kvyoSQXz9qnTXLS8O8nJ7lr6SICJFGLgG5Qi4AuUIuAZTHK5WZnJrljxu2dSZ43a5+3Jfl4KeXnkpyY5EVzHaiUclmSy5Kk1pqJiYlDQcbHD7u9WuSQo8sZjvMcapEcqx1Djo5lWKUcK1KL+qgr58RC9T1/0v8x9D3/ClOL5rEWzqO+j6Hv+ZO1MYaFGqVJ1MyxrZ11+9IkV9daf72U8vwk7y+lPL3WOj1zp1rrVUmuOnCMXbt2HbxvYmIiM2+vFjnk6HKGLubYvn37Sj2lWiTHaseQo2MZZuZYa7Woj7pyTixU3/Mn/R9D3/MnUYs6YC2cR30fQ9/zJ/0fw2Jq0SiXm+1McvaM22fl8VMVX5+kJkmt9TNJNiU5PttuwHJRi4AuUIuALlCLgGUxykyiG5OcW0o5J8mdGSx69upZ+3wjyY8kubqU8rQMCtC9SxkUOO6pRUAXqEVAF6hFwLI46kyiWutkkjcmuT7JzYNN9UullCtLKRcNd7s8yc+UUv4myQeTvLbWOnu6I8CCqUVAF6hFQBeoRcByadp21epEe9ddh2ZEduWaPznk6HKGLuYYXu8613XxfaEWySFHTzPMzLHWalEfdeWcWKi+50/6P4a+50+iFnXAWjiP+j6GvudP+j+GxdSiUdYkAgAAAGCN0yQCAAAAQJMIAAAAAE0iAAAAAKJJBAAAAEA0iQAAAACIJhEAAAAA0SQCAAAAIJpEAAAAAESTCAAAAIBoEgEAAAAQTSIAAAAAokkEAAAAQDSJAAAAAIgmEQAAAADRJAIAAAAgmkQAAAAARJMIAAAAgGgSAQAAABBNIgAAAACiSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAaBIBAAAAEE0iAAAAAKJJBAAAAEA0iQAAAACIJhEAAAAA0SQCAAAAIJpEAAAAAESTCAAAAIAk46PsVEq5IMm7kowleW+t9e1z7FOSvC1Jm+Rvaq2vXsKcAGoR0AlqEdAFahGwHI46k6iUMpbk3UkuTLIjyaWllB2z9jk3yS8l+YFa6/cm+YVlyAocx9QioAvUIqAL1CJguYxyudl5SW6ptd5Wa92X5ENJLp61z88keXet9YEkqbV+a2ljAqhFQCeoRUAXqEXAshjlcrMzk9wx4/bOJM+btc93J0kp5c8zmO74tlrr/1yShAADahHQBWoR0AVqEbAsRmkSNXNsa+c4zrlJfjDJWUk+XUp5eq312zN3KqVcluSyJKm1ZmJi4tABxscPu71a5JCjyxmO8xxqkRyrHUOOjmVYpRwrUov6qCvnxEL1PX/S/zH0Pf8KU4vmsRbOo76Poe/5k7UxhoUapUm0M8nZM26fleSuOfb5bK11f5LbSylfzaAg3Thzp1rrVUmuGt5sd+3adfC+iYmJzLy9WuSQo8sZuphj+/btK/WUapEcqx1Djo5lmJljrdWiPurKObFQfc+f9H8Mfc+fRC3qgLVwHvV9DH3Pn/R/DIupRaM0iW5Mcm4p5ZwkdyZ5VZLZq+L/cZJLk1xdSpnIYGrjbQtOBfB4ahHQBWoR0AVqEbAsjrpwda11Mskbk1yf5ObBpvqlUsqVpZSLhrtdn+S+UsqXk3wqyb+ptd63XKGB449aBHSBWgR0gVoELJembWdfurpi2rvuOjQjsivTueSQo8sZuphjOJVxruvi+0ItkkOOnmaYmWOt1aI+6so5sVB9z5/0fwx9z59ELeqAtXAe9X0Mfc+f9H8Mi6lFR51JBAAAAMDap0kEAAAAgCYRAAAAAJpEAAAAAESTCAAAAIBoEgEAAAAQTSIAAAAAokkEAAAAQDSJAAAAAIgmEQAAAADRJAIAAAAgmkQAAAAARJMIAAAAgGgSAQAAABBNIgAAAACiSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAaBIBAAAAEE0iAAAAAKJJBAAAAEA0iQAAAACIJhEAAAAA0SQCAAAAIJpEAAAAAESTCAAAAIBoEgEAAAAQTSIAAAAAokkEAAAAQDSJAAAAAIgmEQAAAADRJAIAAAAgyfgoO5VSLkjyriRjSd5ba337PPtdkuTDSZ5ba/38kqUEiFoEdINaBHSBWgQsh6POJCqljCV5d5ILk+xIcmkpZccc+21N8vNJPrfUIQHUIqAL1CKgC9QiYLmMcrnZeUluqbXeVmvdl+RDSS6eY79fTfKOJHuWMB/AAWoR0AVqEdAFahGwLEa53OzMJHfMuL0zyfNm7lBKeVaSs2utHy2l/OJ8ByqlXJbksiSptWZiYuJQkPHxw26vFjnk6HKG4zyHWiTHaseQo2MZVinHitSiPurKObFQfc+f9H8Mfc+/wtSieayF86jvY+h7/mRtjGGhRmkSNXNsaw/8pZSyLslvJnnt0Q5Ua70qyVUHjrFr166D901MTGTm7dUihxxdztDFHNu3b1+pp1SL5FjtGHJ0LMPMHGutFvVRV86Jhep7/qT/Y+h7/iRqUQeshfOo72Poe/6k/2NYTC0a5XKznUnOnnH7rCR3zbi9NcnTk9xQSvlakn+c5COllOcsOBXA46lFQBeoRUAXqEXAshhlJtGNSc4tpZyT5M4kr0ry6gN31lofTHJwHlYp5YYkv2jlfGCJqUVAF6hFQBeoRcCyOOpMolrrZJI3Jrk+yc2DTfVLpZQrSykXLXdAgEQtArpBLQK6QC0ClkvTtu3R91oe7V13HZoR2ZVr/uSQo8sZuphjeL3rXNfF94VaJIccPc0wM8daq0V91JVzYqH6nj/p/xj6nj+JWtQBa+E86vsY+p4/6f8YFlOLRlmTCAAAAIA1TpMIAAAAAE0iAAAAADSJAAAAAIgmEQAAAADRJAIAAAAgmkQAAAAARJMIAAAAgGgSAQAAABBNIgAAAACiSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAaBIBAAAAEE0iAAAAAKJJBAAAAEA0iQAAAACIJhEAAAAA0SQCAAAAIJpEAAAAAESTCAAAAIBoEgEAAAAQTSIAAAAAokkEAAAAQDSJAAAAAIgmEQAAAADRJAIAAAAgmkQAAAAARJMIAAAAgGgSAQAAABBNIgAAAACiSQQAAABAkvFRdiqlXJDkXUnGkry31vr2Wfe/KclPJ5lMcm+S19Vav77EWYHjnFoEdIFaBHSBWgQsh6POJCqljCV5d5ILk+xIcmkpZces3b6Q5Dm11u9Lck2Sdyx1UOD4phYBXaAWAV2gFgHLZZSZROcluaXWeluSlFI+lOTiJF8+sEOt9VMz9v9skp9YypAAUYuAblCLgC5Qi4BlMUqT6Mwkd8y4vTPJ846w/+uTfGyuO0oplyW5LElqrZmYmDgUZHz8sNurRQ45upzhOM+hFsmx2jHk6FiGVcqxIrWoj7pyTixU3/Mn/R9D3/OvMLVoHmvhPOr7GPqeP1kbY1ioUZpEzRzb2rl2LKX8RJLnJHnhXPfXWq9KctWBY+zatevgfRMTE5l5e7XIIUeXM3Qxx/bt21fqKdUiOVY7hhwdyzAzx1qrRX3UlXNiofqeP+n/GPqeP4la1AFr4Tzq+xj6nj/p/xgWU4tGaRLtTHL2jNtnJblr9k6llBcl+eUkL6y17l1wIoC5qUVAF6hFQBeoRcCyGKVJdGOSc0sp5yS5M8mrkrx65g6llGcl+Z0kF9Rav7XkKQHUIqAb1CKgC9QiYFkc9dPNaq2TSd6Y5PokNw821S+VUq4spVw03O2dSbYk+XAp5a9LKR9ZtsTAcUktArpALQK6QC0ClkvTtnNeuroS2rvuOjQjsivX/MkhR5czdDHH8HrXua6L7wu1SA45epphZo61Vov6qCvnxEL1PX/S/zH0PX8StagD1sJ51Pcx9D1/0v8xLKYWHXUmEQAAAABrnyYRAAAAAJpEAAAAAGgSAQAAABBNIgAAAACiSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAaBIBAAAAEE0iAAAAAKJJBAAAAEA0iQAAAACIJhEAAAAA0SQCAAAAIJpEAAAAAESTCAAAAIBoEgEAAAAQTSIAAAAAokkEAAAAQDSJAAAAAIgmEQAAAADRJAIAAAAgmkQAAAAARJMIAAAAgGgSAQAAABBNIgAAAACiSQQAAABANIkAAAAAiCYRAAAAANEkAgAAACCaRAAAAABEkwgAAACAJOOj7FRKuSDJu5KMJXlvrfXts+7fmOQPkjw7yX1JXllr/dpiw73lPe/PHzz6XzK19c6MPXxm/sWJb8x//Jc/ueD9luK5luO5l/N4y2ElM3b19ehqrsXq+rjUoqV/3EodbzmoRd3NtVhdH9dq1SKAmdQiYDkcdSZRKWUsybuTXJhkR5JLSyk7Zu32+iQP1FqfmuQ3k/zaYoO95T3vz/umrszUSTuTps3USTvzvqkr85b3vH9B+y3Fcy3V41bqeMthJTN29fXoaq7F6vq41KKlf9z/z969x8dV1/kff5255jJpeknbNG2BlrRCKeWaFpRKAbmsgrj86uGy7s91FdQVd5dFoSILiOyK+lN/CIgiKj92tXoWKzd1oUCLtaVtuAhbKIW0FEiTJk3b3JqZZC7n98dMhkkyk0wuM3OSvJ+Pxzxm5lw/58z00zOffL/fk6/t5YJykXPjGi2nH1ehcpGISCrlIhHJlWxaEi0H6izL2gNgmuavgUuB11OWuRS4LfH6YeAe0zQNy7LskQb20JF7YEpX34m+Ln4RuZXf3vEI55aex73X/QMPHbk77XIPdn2HO/hbdu9t4OL//PKA7V8x639x6zVX8fL/7OYX9q3g6x6wjYfa7+H850/iC898Y8D6/7Toc4PGuOSx+Vz18VX86rGNfPPVuwes/50zb+SS85bz49/8kbveegCA9hkvp42j95hT/erS73DKicfyjft/xa+bfztg+0986m6OPaaKG+79OY8f/v2A+Zu/9CDTp5XxpR/8iGePPDNg/s6b49v8xE0pSJ9hAAAgAElEQVT/yuauPyWnZxNjaaycF275eXz9f7uVXfaOPotPjc7i+VvvA+Cib67hHeOtPvNnx+az8Zb/m/n8dt/BHcT/onzmN75Iq7u5zyIfMJbyyNfjn9npt/89R1xtfeaf7DmdtWu+BsCyb15F2Oh7PGf4P8QvvvovABx/x/9KTjcAG+iYsguKM383L5n2Mb7zpb/P6rt31aM3DJj/ufmf4vpP/zXPPf9q2u/ejSdey99deg6PP7OdG54f+H/9vy778rC/e5D5s32o/Z7k+S4w5SLlouT0yZyLIJ6P2pWLCqUguUhEpB/lIhHJCcO2B88RpmmuBi6yLOtzifd/C6ywLOvalGV2JJapT7zfnVimpd+2rgGuAbAs67SxPBARKSgj1ztQLhKRLCgXiYgTKBeJiBOMKBdlM3B1ug33ryxlswyWZd1vWdbplmWdnlgn+TBN88X+0wrxUByKw8kxODiOfEi3H+UixaE4JnEMaeLIh3T7GXUucsr5HM1jvB/DeI9/IhzDeI8/5RjywUgzTbloYn2PCh7HZI1/IhzDaHJRNkWiemB+yvt5QEOmZUzT9ADlwKGRBiUikoZykYg4gXKRiDiBcpGI5EQ2YxLVAotM01wA7AOuAK7qt8xjwKeB54HVwLPq6yoiY0y5SEScQLlIRJxAuUhEcmLIlkSWZUWAa4EngZ3xSdZrpmnebprmxxOL/QyYYZpmHfAvwJoRxHL/CNbJBcXRl+J4nxNigEkah3JRwSiOvhTH+5wQA0ycXOSU8zka4/0Yxnv8MP6PYbzHD3k6BuWiQekYCm+8xw/j/xhGHP+QA1eLiIiIiIiIiMjEl82YRCIiIiIiIiIiMsGpSCQiIiIiIiIiIlkNXJ1zpmleBNwFuIEHLMu6M0/7nQ88BFQCMeB+y7LuMk3zNuBq4EBi0Zssy/pDDuPYC3QAUSCSuP3kdOA3wDHAXsC0LOtwDmP4QGJ/vRYCtwBTyfG5ME3z58DFQLNlWUsT09Iev2maBvHvykeBLuDvLMt6KYdxfBe4BOgBdgOfsSyr1TTNY4j3/96VWH2rZVlfyGEct5HhczBN82vAZ4l/f/7RsqwncxjHb4APJBaZCrRalnVyLs9HPk32XJSIZS8FzEfKRcpFWcYxbnPRUHnGNE0/8XxwGnAQuNyyrL35jnMwWRzDvwCfAyLEvyt/b1nWO3kPNINsc71pmquB/wJqLMt6IY8hDimbYzBN0wRuI37L81csy+o/qHHBZPEdOgr4f8T/fbuBNbn+/2840uWlfvNz9n/EWFEuKjzlosJTLkqv4C2JTNN0A/cCfwUsAa40TXNJnnYfAa63LOt44AzgSyn7/oFlWScnHvn4IpyT2NfpifdrgGcsy1oEPMPIBuDNmmVZu3qPl/h/Bl3A7xKzc30uHgQu6jct0/H/FbAo8bgGuC/HcawHllqWtQx4E/hayrzdKedlLH+EpIsD0nwOie/rFcAJiXV+lPg3lZM4LMu6POV78ltgXcrsXJ2PvFAu6qNg+Ui5KGMcykUpxmsuyjLPfBY4bFlWNfAD4Nv5jXJwWR7Dy8Dpie/rw8B38htlZtnmetM0y4B/BLblN8KhZXMMpmkuIp4nPmRZ1gnAP+c90Ayy/AxuJj4Q8ynEc8uP8hvlkB4kfX7slcv/I0ZNuajwlIsKT7kos4IXiYDlQJ1lWXssy+oBfg1cmo8dW5bV2FtJsyyrg/hfH+fmY99ZuJR41ZLE8yfyuO/ziF9k56XSblnWn4BD/SZnOv5LgYcsy7Ity9oKTDVNc06u4rAs6ykrfvcIgK3AvLHY13DjGMSlwK8ty+q2LOttoI74v6mcxpGoSpvA2rHYl0MoF2VWqHykXPT+NOWiNMZhLsomz6R+5x4Gzkscp1MMeQyWZW2wLKsr8TYv39dhyDbXf5P4D8pQPoPLUjbHcDVwr5Vo9WlZVnOeYxxMNvHbwJTE63KgIY/xDSmL/Jiz/yPGiHJR4SkXFZ5yUQZOKBLNBd5LeV9PAX4cJZqon8L7VdprTdN81TTNn5umOS3Hu7eBp0zTfNE0zWsS02ZbltUI8R+QwKwcx5DqCvpecOfzXPTKdPyF/L78PfDHlPcLTNN82TTN50zTXJmH/af7HAp1PlYCTZZlvZUyLd/nY6wpF8U5KR8pF6WnXPS+8ZaLsjlPyWUShcE2YEZeosvOcD/rz9L3+1poQ8ZvmuYpwHzLsp7IZ2DDkM1nsBhYbJrmZtM0tya6VDhFNvHfBnzKNM164A/Al/MT2phxxDXFIJSLCk+5qPCUizJwQpEoXUXazmcApmkGiDdX/2fLstqJN8M6FjgZaAS+l+MQPmRZ1qnEm4N9yTTND+d4fxmZpukDPk683yvk/1wMpSDfF9M0v068S9AvE5MagaMSTQ//BfiVaZpTMq0/BjJ9DoX693MlfX+85/t85IJyUZwj8pFyUXrKRQOMt1yUzXkqeC4aQtbxmab5KeB04Ls5jWh4Bo3fNE0X8a411+ctouHL5jPwEO9esIr4v5MHTNOcmuO4spVN/FcCD1qWNY/4WBr/kfhsxouJ8O94IhwDoFyUQ8pFzjeif8dOOMB6YH7K+3nksRmXaZpe4j/KfmlZ1joAy7KaLMuKWpYVA37KGDWZz8SyrIbEczPxsTeWA029TcESz/lqmvdXwEuWZTUlYsrruUiR6fjz/n0xTfPTxAcE+xvLsmyARJeKg4nXLxIfSHZxrmIY5HMoxPnwAJeRMrhwvs9Hjkz6XJTYp1PykXJRP8pFfY3TXJTNeUoukzjGcrLv9pcPWX3Wpml+BPg68HHLsrrzFFs2hoq/DFgKbDTjA/mfATxmmubpOEe236NHLcsKJ7qA7iL+Q80Json/s4AFYFnW80ARUJGX6MZGQa8psqBcVHjKRYWnXJSBE+5uVgssMk1zAbCPePeCvIx4nuhX+zNgp2VZ30+ZPqe3ewHw18COHMZQCrgsy+pIvL4AuB14DPg0cGfi+dFcxdBPn7/K5vNc9JPp+B8j3tXh18AKoC0lvjGXaBJ5I3B2Sp9mTNOcCRyyLCtqmuZC4sluTw7jyPQ5PEb8L+XfB6oScWzPVRwJHwHesCyrPiW+vJ6PHJnUuSixPyflI+WiFMpFaY3HXJRNnun9zj0PrAae7S0KOsSQx5DoIvET4CKHjT8BQ8RvWVYbKT8ATNPcCHzFctYdhbL5Hj1C4i/gpmlWEC+WOuXfQjbxv0t8XLoHTdM8nvgPswOMH3n9P2IElIsKT7mo8JSLMih4kciyrIhpmtcCTxK/rdzPLct6LU+7/xDwt8D/mKb5l8S0m4iPbH4y8aZYe4HP5zCG2cDvTNOE+OfxK8uy/ts0zVrAMk3zs8S/nJ/MYQwAmKZZApxP3+P9Tq7PhWmaa4k3QaxI9Pe8lfgPsnTH/wfiTf3qiN/16DM5juNrgB9Yn/iMem+n/GHgdtM0I8Rv9/wFy7LG5K8rGeJYle5zsCzrNdM0LeB14l1QvmRZVjRXcViW9TMGjhMDOTwf+aJcBDgkHykXKRcNFcd4zUWZ8oxpmrcDL1iW9RjxgvF/mKZZR/yv9lcULuKBsjyG7wIB4L8S39d3Lcv6eMGCTpFl/I6W5TE8CVxgmubrxP8tfLW3lV2hZRn/9cBPTdO8jni++TsnFSgy5EcvgGVZPyaH/0eMBeWiwlMuKjzloswM23bMMYqIiIiIiIiISIE4YUwiEREREREREREpMBWJRERERERERERERSIREREREREREVGRSEREREREREREUJFIRERERERERERQkUhERERERERERFCRSEREREREREREUJFIRERERERERERQkUhERERERERERFCRSEREREREREREUJFIRERERERERERQkUhERERERERERFCRSEREREREREREUJFIRsgwjEcNw9ieYZ7PMIwWwzDuMAxjo2EYdprHjnzHLCKTh2EYDyZyzQ/SzLMNw/hUIeISkcknkY+eLnQcIjJ5pVwX2YZhRA3DqDcM4yHDMOYWOjZxHhWJZKR+AtQYhnFSmnmXAdOABxLvfwXM6fc4Ox9BisikFgS+ZBjG4kIHIiIiIlJgm4j/DjsKuAo4BfivgkYkjqQikYzUfwPvAFenmXc1sN627b2J90Hbtvf3exzMV6AiMmltAV4EvlvoQEREREQKrCfxO2yfbdt/Au4HzjQMY0qhAxNnUZFIRsS27RjxlkKfMgyjuHe6YRjHAucQTzoiIoV2HXCJYRjnFDoQEREREScwDKMKWA1EEw+RJBWJZDR+BpQCn0yZ9jmgCXgsZdqnDcPo7Pe4J5+BisjkZNv2VuA3wPcNw9D/eSIiIjJZrUr8DusC9gEfBv6vbdtHChyXOIyn0AHI+GXbdqNhGE8Q7172kGEYHuDvgF/Yth1JWfR3wE39Vm/LT5QiIqwB3iCen35e2FBERERECmIb8GmgCDCB84F/LWhE4kgqEslo/QT4o2EYxwPHA7N5f8DqXu22bdflPTIREcC27XcSdzm7wzAMq9DxiIiIiBRAMOU32Y7EjT3uBf6+gDGJA6npvYzWU8Be4q2Jrgaetm17T0EjEhEZ6FvE/8+7sdCBiIiIiDjAbcSHBTm90IGIs6glkYyKbdsxwzB+CtwAlAGXp1ms2DCMyn7TYrZtN+c8QBERwLbtDsMw/hW4q9CxiMikFDAM4+R+00K2bb9RkGhEZNKzbfuNxNAh3yLe9UwEUEsiGRs/Jz6AdQvwaJr5VwGN/R5qbSQi+fYz4K1CByEik9IK4OV+j0cKGpGICHwH+IhhGOcVOhBxDsO27ULHICIiIiIiIiIiBaaWRCIiIiIiIiIioiKRiIiIiIiIiIioSCQiIiIiIiIiIqhIJCIiIiIiIiIigKeA+9aI2SITh1HoAEZBuUhk4lAuEhEnUC4SEScYUS4askhkmubPgYuBZsuylqaZbwB3AR8FuoC/syzrpWx23tDQkHxdUVFBS0tLlmHnjuJQHE6OwYlxVFVV5WV/ykWKQ3E4Kw4nxJAaR75yEeQuH6XmInDGOXZCDIpDcYyHOJSLcs8JcTghBsWhOLKJw+fzjXj9bLqbPQhcNMj8vwIWJR7XAPeNOBoRkcweRLlIRJzhQZSPRKTwHkS5SETG2JBFIsuy/gQcGmSRS4GHLMuyLcvaCkw1TXPOWAUoIgLKRSLiHMpHIuIEykUikgtjMSbRXOC9lPf1iWmN/Rc0TfMa4lVsLMuioqLi/UA8nj7vC0VxKA4nx6A4BqVcpDgUxySLwUlx9JNVPhosF4Ezjs0JMSgOxTEe4nBCDGlMmFzklDicEIPiUBzZxDGq9ccghnSDIaUd8MyyrPuB+3uXSe2v56T+e4pDcYw0Btu2CYVCxGIxDCM3Yxb6/X66u7tzsu1MbNvG5XJRVFSUPK5C9L0fgnKR4lAcCcpFBZdVPhosF4Ezvm9OiEFxKI7xEIdyUe45IQ4nxKA4hh9HPq6LIP/XRumui2D0YxKNRZGoHpif8n4e0JBhWZEJLRQK4fV6R129HYzH48Htduds+5lEIhFCoRDFxcV533eWlItEEpSLCk75SEScQLlIhPxcF0Fhro1ycV00FmfpMeBa0zR/DawA2izLGtC9Q2QyiMViOU8+heLxePLeamCYlItEEpSLCk75SEScQLlIBF0XDXubQy1gmuZaYBVQYZpmPXAr4AWwLOvHwB+I31axjvitFT8zphGKjCMjab74xBNPAHDxxRePdThjLpfNM4eiXCSSPeWi3FI+EhEnUC4SyY6ui4ZnyCKRZVlXDjHfBr40ZhGJiKShXCQiTqF8JCJOoFwkIrngKnQAIpNZXV0dTU1NNDY2snbtWurq6ka9zfnz53P++edz7rnncs011xAMBgFYtGhR2v2vXr2a888/n7PPPpsbbrhh1PsXkfFHuUhEREQkbrJfF03Mjnki40BdXR2bNm0iFosB0NnZyaZNmwCorq4e8XaLiopYv349ANdeey0PPfQQn//859Mue8stt3D11Vdz4YUXArBz584R71dExiflovGjrq6O2tpaOjs7CQQC1NTUjOozEhERkb50XaQikUjOPP/88xw8eDDj/KampmTy6RWJRHjuued444030q4zY8YMVq5cmXUMy5cvHzSpNDc3M2fOnOT7448/Putti8j4oFw0MfRetEYiEWDwi9a6ujqee+45YrGYikkiIiIpcnVddOaZZ2Ydg9Ovi9TdTKRA+iefoaYPVyQSYcOGDRx33HEZl7n66qsxTZNPfepT3H///bS1tY3JvkVk/FAuGh9qa2uTBaJekUiELVu2sHfvXpqammhra+ONN97gT3/604C/gI5FU3kREZGJTtdFakkkkjNDVZPXrl1LZ2fngOmBQGBUo+iHQiHOP/98AFasWMGVV2Ye0/Dyyy/n7LPPZuPGjTz55JP853/+J+vXr8fv9494/yLiLMpFE0O6zwigu7s72Xw9k0gkQm1t7bBbEz29ejVer5ez167Nep23161j2/XXE+vpoWTuXE5as4YFl102rP3mcnsTwUg+FxERidN10dBUJBIpkJqamj5dBwA8Hg81NTWj2m5qf9dsVFZWcsUVV3DFFVdw7rnnsmvXLpYtWzaqGERk/FAuGh8CgQDBLVvwPfMMRlsbdnk5Peedh/+MM7jwwgsJBoOEQiE2btyI+9VXByzXuWwZb7/9NvPnz8fjGfry7+1162h58UViPT08snx5VsWZt9etY/sNNxDr6QGga98+ticG2xxJYWestzcRjORzyVdcKuaJyESg6yIViUQKpvcvuoUcN2LDhg2cddZZeL1empubOXz4MJWVlXnbv4gUnnLR+LCgtZXdjz+OEQ4DYLS14X/8cY5dsoSKigoAbNum9kc/wn7sMYzExa3R1ob/scfoDoV4uqsLb1ER848+moXV1cw/5pi0BaOMxRnbZt4FFxDu6KCno4NwRwfhzk4iieeXvvlNoom7tfSKBoPU3ngjTZs2EYtGsWMxiMWwE6+Tj9T30SjEYhx44YVkDKnbe+Gmm+hpa6No+nSKKirwz5hBUUUFvmnTcLndY37unSLt5/LVr2JHIixYvRrDVZhRJFTME5GJRNdFKhKJFFR1dXVyALTRNF/MRjAY5LTTTku+v+aaa2hsbOSWW25JNl28+eabmTVrVk7jEBHnUS5yvv2/+EWyQNTLCId5+847aXrwQcKdnYQ7OyESwei3rhGJUPSHP8Af/gBAU+LxPIDbjcvrxe314vJ4MLxeulta4sWcFNFgkOf/8R9HFHukq4v9mzdjuFwYbnf8ud9r0szrXyDqFe7o4MWbbx44wzDwT5v2fuFoxgz8FRUU9b5OFJP8FRUUTZ+Ob+rUMSmsjFUrmkgwSKi5ma6mJoL79xNsaiLY3Jx83bxtG3a/camioRBbr7uOrdddB4YR/wzd7vjD48HV79lwu/u+Tn32eOLnPnW93u0l5vUu1zu/JBBgx89/nrY4+OItt+Byu3EXF+MpKsJdXIy7qKjPw5OY5sqidVshqIWUyOQ02a+LnJmRRSaRsU48b731Vtrp9fX1aaffdtttY7p/ERmflIucrauhIe10OxJh1hln4C0rw1Nayuv33JNxG6d+4xvYkQjRcJi2Q4c4dOAArYcO0RMO4wHKSkspKy1l/yOPZNzGKbfcgresDG8g0Pe5rIz1l15KV2PjgHVK5s7lE9u3D/uYH1m+nK59+9Ju76I//pFQSwvdBw8SamkhdPBg8nX3wYOEDh6kdedOQgcP0tPamnb7htuNP7U1Um9RqXdaosDUO887ZcqAbWTTiiYaCsWLPU1NfR5d+/cTSkzvamoinGZgUpfPR/Hs2RTPnj2gQJTqxK98BTsSIRaJYEej8edYDDsSiU+PRvs829Ho+8ulPvf0EAsG4/NT1+tdJ7HdWMo2IhnGy+o5fJjN//APGWPu81l4PPGiUb9ikie1qNRbbEpTZCqvqCAYifSZ5k7ZVv/tZtPiTC2kRCa3yXxdpCKRiIiIiMOVVFVlLJicedddyfd7f/e7jMsd97nPDZgejUapr69nz549vPPOOzSHw5Rs2ICRpmDhmz2b4z//+YwxnnTTTWz7yleIdXcnp7n8fk5as2bI40u7vTVr2H7DDX1aqbiLizlpzZpk66BsxMJhug8diheSWloIHToULyYlXne3tBBqaeHgK6/QffAg4Y6OtNtxeb0Uz5yJb9q0ZAGp/qmn0rai2faVr/D6PfcQbGpKW6Ryeb0UzZpF8ezZTKmuZvaHPhQvBs2aRXFlZbIw5Js6FcOItw0brGh24nXXZXUuxlpFRQUPLFyYNq7iykrOXbuWSDBINBRKPiK9rxPT+8wPBt+fn3iEjxwhdPBgn2m962HbI4rb5fUOLCL1Kz41/ulPaT/bV+68U0UiEZnQVCQSERERcbjBCiYjWS45z+3m6KOP5uijjyYSiVBfX89zO3bgefTRPt3bbK+Xrg9/mNdee41AIEBZWRmBQACfz5dcJrpsGd2XXIJ7/frkoNnh888nOsKBNnt/iL9y5510NTRQUlU1ou4+Lq83WXTJRrS7O9kaKVlYammh+9Ah6Oyktb6e7kOH6NizJ2Mrmlh3N2ULFjDrjDPi+66sjBeAEq/9I+jqNtzPNl8yxXXy179O+eLFOduvbdvEenqIhkKUl5RwoKGhT8GpT7EpTfEpuWzqOsEg4Y4OQgcODCgQ9crUqk9EZKJQkUhERETE4bItmIymsOLxeDjmmGNYv3Qp0VhswB3SoiecwJYtW/qs4/P5CAQCBAIBGhoaiCxdCkuX9lmmtrZ2xAN+LrjsMl7z+Zjh9XLhhReOaBvD5fb7KamqoqSqasC8iooKWlpaku8Ha93z4Z/9bEzjGqui2Vjr3X++x+4xDAO334/b76e0ooKg1zum28/42ab5XoiITCQqEomIiIiMAwsuuyyrH97ZLpdJIBCgc9kygv1aAAUCAS699FI6Ojro7Owc8IhkGDOns7OT3/72t5SWllJSUpJ8pL4vLi7GlaZlTV1dHU1NTcRiMdauXZv3O8wMJd+te3o/2/7FqkIb7XfOiZzacktEJNdUJBIpsKdXrwbgIw8/XOBIRGQyUy6SXjU1NWzatKlP0cfj8VBTU5Ms6sxO03Vr7dq1dKbpfuXxeAgEAnR1dXHw4EGCwSB2v7FkDMOguLi4TxEpFArx7rvvEkvcaa2zs5NNmzYBOKZQVKhWNJJ7+mxFJrfJfF2kIpGIiIiIJPUWYJ577jlisRiBQCCrFjyZiksrV67ss24sFiMYDNLV1cWRI0fo6urq87qzs5Pm5mZCodCAfUQiEZ577jnee+89ysrKmDJlSvK5pKQkOchzPk3EVjQSp89WRCYjFYlECujtdetoefFFYj09PLJ8+Zj8hWr+/Pkcd9xxRKNRqqurueuuuyguLqa5uZlbb72VV155BZ/Px/z587ntttvw+XysWrWKhQsXEg6HWbZsGd/73vfwjnHffhFxLuUi6a+6uprq6uphdWvKtrjkcrkoLS2ltLSUmTNnZtzeT3/607TTY7EY+/fvZ/fu3X1aJLnd7mTBKLV4VFZWRllZGR6PLntFRGRok/26SP9bihTI2+vWsf2GG4j19ADQtW8f22+4AWBUSaioqIj169cDcO211/LQQw9xzTXX8NnPfpZPfvKT3HfffQDs2LGDlpYWqqqqOProo1m/fj3RaJQrrriCxx9/nMv0lzORSUG5SMZSb3FpLAQCgbTd1wKBAFdeeSXRaJTOzk7a29vp6Oigvb09+bqhoWHAGEmlpaUDWh/1PhcVFY2qFdITTzwBwMUXXzzibYiISOHpukhFIpGcefGWWzj8+usZ5/dWp1NFg0G2XX89u3/1q7TrTFuyhBX//u9Zx7B8+XJ27tzJ5s2b8Xq9/O///b+T85Ym7j7z3nvvJae53W5OOeUU9u/fn/U+RMTZlItkvBpsbCSIf0/Ky8spLy8fsK5t24RCoQEFpPb2durr6+nq6uqzvNfrHbQVUrpBtXvle3DtJ554Am8e7/YmIjKR5Oq66LTbb886BqdfF6lIJFIg/ZPPUNOHKxKJsGHDBlatWsWuXbs48cQTh1wnFArx0ksvcfswkpyIjG/KReJUvYWW2tpaOjs7sx4bCd4fCLu4uDjtINuRSCRZPEotIrW2tvLee+8RjUb7bKu0tJQpU6Ywa9YsvF5vsph04MABtm7dmrfBtZ18tze1phKRiUDXRSoSieTMUNXkR5Yvp2vfvgHTS+bOHdUo+qFQiPPPPx+AFStWcOWVV/If//Efg67zzjvvcP755/P222/zsY99jCVLlox4/yLiLMpFMp6NZGykbHg8HqZNm8a0adMGzLNtmyNHjgwoInV0dLBr1y6CKbdETycSifDnP/+Z9vZ2fD4fXq8Xr9ebfJ367PF4Bm2llKquro5NmzY58m5vTi5eiYik0nXR0FQkEimQk9asYfsNNxBNudh0Fxdz0po1o9puan/XXosXL+b3v/99xnV6+7s2NTWxevVqnnrqKS644IJRxSEi44NykUhfhmEQCAQIBALMmTOnz7yKigoaGhqShaOnn3467TbC4TAvvvhiVvvrLSKlKySlvn711VcHjLMUiUTYunUrU6ZMwTAMXC5Xn+d009ItM5rxmJxcvBIRGS5dF6lIJFIwvQOfbbv+emI9PZTMnTsmI+enc9ZZZ/Htb3+bX/7yl/zN3/wNAH/5y18IBoPMmzcvudzs2bO56aabuPvuu/XDTGSSUC4SGR6fz8eMGTOYMWPGoINrX3755fT09BAOhwmHw8nXmZ5TX7e1tfWZlnoXt/6CwSCPPvroqI5pqIJSpnlerzfZgihVJBJhy5Ytfbr9FRcX4/f7R1WQEhHJNV0XqUgkUlALLrssOQDaaJovDsUwDB544AFuvfVW7r33Xvx+P/PmzeMb3/jGgGUvuugivve977Ft2zZWrFiRs5hExDmUi0RGZrDBtV0uF0VFReDifosAACAASURBVBQVFY1qH7ZtE41G+c1vfjNgwG2I/3X67LPPxrZtYrEYtm33eZ3peTjLZFrH4/EMKBD16u7u5tlnn+0zzTCM5DnpLRwN9trr9Ra8qKSxlkQmn8l+XaQikUiBjXXieeutt9JOr6ys5Cc/+UnaeakXcYZhZGw+LyITl3KRyPD1dqd67rnniMViwxpcO1uGYeDxeFixYkXagtSZZ57JUUcdNWb7G46KigruvvvutK2pSktLueiiiwiFQgSDQYLB4IDXBw4cIBgMEg6H027f7XZnVVDy+XxEo1HcbneuD1lEJonJfF2kIpGIiIiIyAj1Dq6dj/1AbgtSI5GpNdXy5cuZPn16VtuIRCKEQqFBC0rBYJBDhw4RCoX63H0uldfrzaqFUm/Xt6EGDdeA3CIyGalIJCIiIiIyDlRXV/PGG2/g9Xq58MILCx0OMDbFK4/HkxwsfCi2bRMOh/sUjzweT7JVUu+09vZ2mpubCYVCacd0MgwDv9+fsYjU2trKzp07NSC3iEw6KhKJjKHBBpacCCb68YlMFBP93+pEPz6RwVx88cVUVFTQ0tJS6FCS8tWaCuLFHZ/Ph8/no7y8HGDQ8xGLxeju7u5TQErXSqmlpYVQKERPT0/GfUciEWpra1UkEhlnJvp1w1gfn4pEImPI5XIRiUTweCbeP61IJDJks2wRcQblIhGROJfLlWwllI1oNEowGGTt2rVp56cbf0lEnE3XRcMz8c6SSAEVFRURCoXo7u7O2d04/H4/3d3dOdl2JrZtJ+/SIiLOp1wkIjIybrc72fUtXUEomy5xIuIs+bgugvxfG+XqukhFIpExZBhG1n+pGimnNTEXEedRLhIRGZ1MA3LX1NQUMCoRGYl8XBfBxLk2UpFIREREREQkhVPvJicikmsqEomIiIiIiPSTzwG5RUScQiM/ioiIiIiIiIiIikQiIiIiIiIiIqIikYiIiIiIiIiIkOWYRKZpXgTcBbiBByzLurPf/KOA/wdMTSyzxrKsP4xxrCIyySkXiYgTKBeJiBMoF4lILgzZksg0TTdwL/BXwBLgStM0l/Rb7GbAsizrFOAK4EdjHaiITG7KRSLiBMpFIuIEykUikivZdDdbDtRZlrXHsqwe4NfApf2WsYEpidflQMPYhSgiAigXiYgzKBeJiBMoF4lITmTT3Wwu8F7K+3pgRb9lbgOeMk3zy0Ap8JF0GzJN8xrgGgDLsqioqHg/EI+nz/tCURyKw8kxTPI4lIsUR6HDUBwOi6FAceQlF4EzzrETYlAcimM8xKFclHtOiMMJMSgOxZFNHKNaP4tljDTT7H7vrwQetCzre6Zpngn8h2maSy3LiqUuZFnW/cD9vdtoaWlJzquoqCD1faEoDsXh5BicGEdVVVW+dqlcpDgKHYbicFgMqXFMtFwEzjjHTohBcSiO8RCHclHuOSEOJ8SgOBRHNnH4fL4Rr59Nd7N6YH7K+3kMbKr4WcACsCzreaAIKHwJTUQmEuUiEXEC5SIRcQLlIhHJiWxaEtUCi0zTXADsIz7o2VX9lnkXOA940DTN44knoAOjDe7tdet45c476WpooKSqipPWrGHBZZeNdrNjajzEKDJBFCwXiYikUC4SESdQLhKRnBiyJZFlWRHgWuBJYGd8kvWaaZq3m6b58cRi1wNXm6b5CrAW+DvLsvo3dxyWt9etY/sNN9C1bx/YNl379rH9hht4e9260Wx20P09snw5Pywq4pHly7PaT75jFJnMCpWLRERSKReJiBMoF4lIrhi2XbA8YTc0vN8isn//vUeWL48XX/opmTuXT2zfPqaB9BZ7osFgcpq7uJhTvv51Zn3wg/S0tdHT2hp/tLUlH7vXru2zTq/iOXP46xdeGFVMTurPqDicE4MT40j0vU/XL368GDQXFYriUBxOj8MJMaTGMdFyETjjHDshBsWhOMZDHMpFueeEOJwQg+JQHNnEkRiTaES5aHTDXudQV0P6OzRmmj5Stm3zlzvuGFDsiQaDvHDzzRnX85aXpy0QAQQbG/ndaacx7YQT+jwCRx+N4cpmGChxurq6Ompra+ns7CQQCFBTU0N1dXWhwxIREREREREZMccWiUqqqtK2JDLcblpeeomKU08d8bbDnZ00bd5Mw4YNNG7cSLCpKeOyH/rRj/BNnRp/lJfjKy/HO2UKLrc7Y2snb3k5sz/0IQ6/9hqNGzdiR6MAeAIBpi1Z8n7haOlSyhcvxu33J9fVGEfOV1dXx+a778a9fj0lbW1Ey8vZfP758OUvq1AkIiIiIiIi45Zji0QnrVkzoAuYy+fDXVTEUx//OIv+9m85ac0afOXlQ27Ltm1aX3+dxo0badiwgQO1tdiRCJ7SUirPOotwRwc9ra0D1iuZO5ejL710WDG6i4s5/Y47koWdaChE25tvcvi11+KPHTvYY1lEjhwBwPB4KF+0iKlLlmDbNu8+/jh2OAxA1759bPvKVwAcVSiajIUs27Zpa2ujvr6e5++9F8+jj2IkPiejrQ3j0UfZ6nZT/rWvUVpaSnFxMYYxnlsai4iIiIiIyGTj2CJRb9GhfzFi3gUX8Op3vsObv/gF7/3xj5x6223YsdiA5apWraJx0yYaN2yg8bnnCDU3AzB1yRKO//znmbNqFRWnn47b58s4JtFJa9aMKMbUgom7qIjpy5Yxfdmy5DQ7FqPznXeSRaPDr71G0+bNBPfvH7CPWHc3z//TP7HjBz+IF8l8PlxeLy6fr+97vx93ynSX1xufl/re70+7bp/l+k0riUbp6ejAlVj3ncce63OuegfrTj0f41lPTw+tra20tbVx6MABWvfupf3ddznS2AiHD2O0t+N5+WWMSKTPekY4DA8/zH9v24bt9YLXi6eoCE9xMd6SEnylpfgCAYoCAYrKyiiaMoWSKVPwlZbiLipKPjxFRbiLi99/XVSEy+/PW8FpMhYARUREREREJM6xRSKIFx3S/UA97fbbWfDJT7L9xhvZ8qUvgcsFsRgQL1o8/4//CIkBuX1Tp1L54Q9TtWoVlWefTUllZdr9wODFnuHGOBjD5aJswQLKFizgqIsvpqenh/3797PpzDPTjixlx2J0TJmCKxbDZdsYwSDGkSMY0Sj0PiIRiESwIxHscJhY4mH3K2bkQjQY5IWvfx3D5aJ07lxK586laPZsXG53zvc9ErFYjNaWFg7s2cOhPXtoe+cdOvfto7upieihQxhtbbja2zESrb0AfIlnz5QphDOd02iUqVOmEAkGiba3Ez1wgGh3N5GeHrrC4fjnNRKGgdvvTxaSfKWlGIlWde5EUcmTUmgaUHRKU3hKt9y+Z57hpdtuIxoKAROvACgiIiIiIiKDc3SRaDDTTzyRCx5/nHUnnkhPW1vfmbaNt6yMc375S6affHJWxYrosmUEr7uOI52dGIEA0ZSWP2MtHA7T1NREQ0MDDQ0NtLS0YNs2xeXlGP2PBbDLy5n71a/S09NDd3d3n+dwostTJoZt4/d48Lpc+N1uvIaBN/HsAbyGgTvx2m3buBPPvQWpUr+f9sOHifX0EO3p4X/+z/9Jf0zt7fGCXYLL66V4zpx40WjevD7PJfPmUVpVhbuoaMhzNZKWLbFwmK79+zm8dy8Hd++m7Z136KivJ7R/Pz0tLdiHDkFn54CCnKukhKKZMyk++mjK5s2j/Kij4jHPmUNxVRXzly6lPRTi4VNPpSfNOFb+2bO55I9/zBhXqKuLzsOH6Th0iM7DhznS2kqwvZ2utjaCHR10d3TQfeRIvFVSJALhMEZKccll29i2je1yEevpwY5EiIRC0NGB3dNDrLs7XqAKhYiGQsR6eoY8v0OJBoO8cuedKhKJiIiIiIhMAuO2SATgcrvpaW9POy/c2UnFaadltZ26ujo2bdpEJNFCpLOzk02bNgGMyUDEkUiE5uZmGhsb2bdvHwcOHCAWi2EYBrNmzeLkk09mzpw5PLdrF/bDDyfHugGwvV5cH/0oq1atSrvtWCzWp2iUrpDU/zmYeN3d3U0s0QIrE09PD76SEnxTp+L3+4mVl+PKUMg67cc/JtrSQri5mZ6mJrqbmgg2NND05z8TbGrC7revopkz+xSQSnpfJ97ve+YZtn3lK8S6u4H3x2jqPnyYGcuW0dXYSGd9PYf37qWjvp5gYyM9zc3E2tqSLcmS8fn9uKZNwzdzJkWLFxNIFIFmHHss5UcfTcmcOXgDgUHPhS8QgFCI026+uU9cAC6/n9MGuRseQFFJCUUlJVTMnZtxmVgsRldXF0eOHEk+93/d1tWV/K722X5REaWlpZSVlFBaWkpJURFFHg/FHg9+txufYeCOxYglikiRUIhoSlFp+403po1prO8oKCIiIiIiIs40rotEkPkuaCVVVVlvo7a2dsCP7kgkwrZt2zjmmGPweDKfpnS3Ql+4cCEHDhxIthRqamoiGo1iGAYVFRUsXbqUqqoqKisr8Xq9yW0t/8IX2BwO416/HqOtDbu8nOj55/OhL3wh4/5dLhdFRUUUZdEqJ51IJJK2kNT72u1209ra+v68887D//jjAwpZ3eedx6Zdu+IT3G6oqoKqKoxTT8Xr9eJ1u/F1deFub8fd3g6trXD4MF0HD9Lx0kuEn3wyOWB3kmEMKPbEurt56ZZb+kyzvV7s8nLsKVNwL1qEf/ZsSufOpXz+fKYtXMisxYuZVlk5ZuP6jKZ74lBcLheBQIDAIAWrGTNm0NDQkLaA1Pv6wIEDhBLdxlK53W5KS0vjRaSSEkpnzky+9sycSeTAgQHr+GbNGvVxiYiIiIiIiPON+yJRpjuMDTXodKrOzs6007u6uvjFL36B3+9P/nAvLS1Nvj58+DCvvvoq0UR3oM7OTjZu3MjGjRuxE8WN6dOnc/zxx1NVVcWcOXPw+Xxp9wWJVktf/jK1Z5yRLDqdUVOT09uqezwePB4PJSUlaedXVFTQ0tKSfL+2vZ0g4HvmmWQhq+e88/CfcQYf+chHkgWmdI9wOJx2um3b8TGlurpwtbbG7xbW2opv/fr0YzQB/i9+kSlHHcW0Y45hWlUV06ZNY8qUKX2Kbrk0krGoxophGPj9fvx+P9OnT8+4XDQazdga6ciRIzQ3N3PkyJFkazL3ypVpC4A9556b82MSERERERGRwhv3RaLRtuoIhUK43e5koSeV3+9n2bJldHZ20tnZSUdHB42NjfQMMtaLbdt4PB5WrVrFnDlzht3Cp7q6murq6gHFGaeoqalhUyhEMGXMJo/Hw4oVK5g9e/awt2fbdrI1U//H89u3ZxyjafUQXbsk3mqorKyMsrKyjMvYtk13dzdHjhxhHdDNwAJg9AMfyFvMIiIiIiIiUjjjvkgEI2/V0dTUxDPPPEM0GsXlcvUZn8fj8fDBD34wbSuenp4ejhw5wsMPP5x2u5FIhAULFgw7nvGg93z072I30tZOhmHEu6N5vZSWlvaZt+2jH804RpOMDcMwkt0VA4EAncuW9SkAAoN2fRMREREREZGJY0IUiYbLtm127NjBtm3bKC0t5ROf+ARtbW1ZFz58Ph8+ny/+ozpNV7WJ/qO6t7VTro1kjCYZuZqamj4DuEO8WFpTU1PAqERERERERCRfJl2RqLu7mz/96U/s3buXo48+mrPPPhu/38/MmTOH3c1LP6pzqxBjNE1mY91KTERERERERMaXSVUkamlp4emnn6azs5MVK1Zw4oknjuqOV/pRnXtOH6NposlXKzERERERERFxnklRJLJtm507d/L8889TXFzMJZdcMqJBltPRj2oRERERERERmQgmfJGop6eHP//5z+zevZt58+ZxzjnnDPuOYyIiIiIiIiIiE92ELhIdOnSIp59+mvb2dmpqajjppJNG1b1MRERERERERGSimrBFol27drF582Z8Ph8f/ehHqaqqKnRIIiIiIiIiIiKONeGKROFwmM2bN/PWW29RVVXFOeecQ0lJSaHDEhERERERERFxtAlRJKqrq0veYczlchGLxTj11FM55ZRTcLlchQ5PRERERERERMTxxn2RqK6ujk2bNhGJRACIxWK4XC7Ky8tVIBIRERERERERydK4r6LU1tYmC0S9YrEYtbW1BYpIRERERERERGT8GfdFos7OzmFNFxERERERERGRgcZ1kai1tTXjvEAgkMdIRERERERERETGt3FbJAqFQjz55JN4PB7cbnefeR6Ph5qamgJFJiIiIiIiIiIy/ozLgauj0ShPP/00nZ2dfOxjH6OzszN5d7NAIEBNTQ3V1dWFDlNEREREREREZNwYd0Ui27b585//TGNjI6tWraKyshJARSERERERERERkVEYd93NXn31Vd58801OOeUUFi1aVOhwREREREREREQmhHFVJNq7dy/bt29n4cKFnHbaaYUOR0RERERERERkwhg3RaKWlhY2bNjAzJkzOfvsszEMo9AhiYiIiIiIiIhMGI4ek6iuri45ILVhGHi9Xi644AI8HkeHLSIiIiIiIiIy7ji2JVFdXR2bNm2is7MTiA9YHY1GaWhoKHBkIiIiIiIiIiITj2OLRLW1tUQikT7TotEotbW1BYpIRERERERERGTicmyRqLcFUbbTRURERERERERk5BxbJAoEAsOaLiIiIiIiIiIiI5fVCNCmaV4E3AW4gQcsy7ozzTImcBtgA69YlnXVaAKrqalh06ZNfbqceTweampqRrNZERnHCpGLRET6Uy4SESdQLsqv35xzDoZhYD77bKFDEcmpIVsSmabpBu4F/gpYAlxpmuaSfsssAr4GfMiyrBOAfx5tYNXV1axcuTLZcigQCLBy5Uqqq6tHu2kRGYcKlYtERFIpF4mIEygXiUiuZNOSaDlQZ1nWHgDTNH8NXAq8nrLM1cC9lmUdBrAsq3ksgquurlZRSER6FSwXiYikUC4SESdQLsqjrffdR3j/foz2dn51wgksvPZazvjiFwsdlkhOZFMkmgu8l/K+HljRb5nFAKZpbibe3PE2y7L+u/+GTNO8BrgGwLIsKioq3g/E4+nzvlAUh+JwcgyTPA7lIsVR6DAUh8NiKFAceclF4Ixz7IQYFIfiGA9xKBel96NlywD4h1dfHdH6o4ljrPb99He/y+5vfxtXOByf0NrK7m9/m6KiIj7y1a+Oatsj5YTvvOJwdhyjWj+LZYw00+w021kErALmAZtM01xqWVZr6kKWZd0P3N+7jZaWluS8iooKUt8XiuJQHE6OwYlxVFVV5WuXykWKo9BhKA6HxZAax0TLReCMc+yEGBSH4hgPcSgXpWfb8ZDG4vMZThxb77uPnsZGjPZ2fjh7dlYtf2LRKIf27uXAG29w6K23aN+7l2B9PcHt2zGi0T7LGuEwr91xB76ZM1l49tkUlZWN+LhGwgnfecXh7Dh8Pt+I18+mSFQPzE95Pw9oSLPMVsuywsDbpmnuIp6QakccmYhIX8pFIuIEykUi4gSOz0WF6qK19b770rb8ATj105+m+Y03aNm1i9Y9e+h85x2C9fWE9+/HPngQI+WmSbbLhTF1KvQrEPUyQiH+8vnP87LLhXv+fAJLlzJ7xQoWnHMOFQsX5vw4C82pA3n/5pxzALh8w4YCRzJ+ZVMkqgUWmaa5ANgHXAH0HxX/EeBK4EHTNCuIN23cM5aBisikp1wkIk6gXCQiTuDoXDRYoWYkhaLhFCT23HMPRu9+E4xwmN3/9m/svuOOPk2wbJ8PV0UFvrlzKV6xgrJjjmHqsccy87jjmLloER6/n1+dcAK0tjJAeTnH33wz+7dupf3VV2l78knaf/973gKYNo2i44+n4rTTmP/hDzO/pgaP1zvs4xYphCGLRJZlRUzTvBZ4knhf1p9blvWaaZq3Ay9YlvVYYt4Fpmm+DkSBr1qWdTCXgYvI5KJcJCJOoFwkIk7g9FyUsVDzgx+w+GMfY/pRR43p/mKxGPUvvcTu3/8eu7U1bV88bJuKK65gysKFTF+0iJnHHcfUefNwuQa/4ffCa69l97e/3ed4bK+XY7/8ZU656iq4Kl6b6wkGeef556nftIlDL71E8H/+h/otW6i/+262+P14jz2W8pNOourMM1l4zjmUTp8+hmcgv5w6kLdT4xoL+WwhZfT2Ey0Au6Hh/RaRTuq/pzgUh1NjcGIcib73af8vHieUixSH4hinMaTGMdFyETjjHDshBsWhOMZDHJMtFw32g7XhlVfY8NGPDnoS7LIyvEcdRWDxYiqWLaPq9NOpPPHEtK1ttt53H3U//CFGezvG1KnJH/4dBw7wxmOP0bBhA0defjnZ2sd2uTBisYE7nTqVq157bYjDTy9TDIOJxWI0v/EGb2/YwIHaWo689hqxhgYMwDYMXHPnUrpkCbOWL2fBOecwc/HiIQtWvQr5ne9tJTagaHbjjXktyMSi0eRjWnk5/33nnbz7wx8WNK5cfS7D/f6ljEk0olw0umGvRUREREREZNJrra9n06230v7UUxmXsQMBqv7mb2jbuZPgnj0cfuwxWn/3O+qI/6B3V1VRXF3NtBNOoPKUU2jasYN3f/jDvt3W/v3f2fPjH2MfOoQRi2H7/fiXLmX2ypUsvuQS3t6wIX0R49prR3xsZ3zxi5zxxS8OqwjgcrmoXLKEyiVLktOOtLSwe+NGGp9/nta//IWOjRvpfOop9txxB/aUKRR94APMOO005p11Fsd88IN4/P5hxZmr1iaxWIz2hgb279jB7u9/P30rsW99i71r14JtQywGto2deB7wvndayvsB83pfpz4S040MDV36V0SMcJjdd9zB7u99DzweDI8HvF4MrxfD48Hw+eKvfT5cXi8unw/D58Pt8+Hy+XD5/bj9ftw+H+6iovhrvx9PURHuoiI8fj+e4mI8RUV4i4tpnTmTYCSCp6gIX0kJnqIi/KWl8W1nWQDsb6y7bmZDRSIRERERERHJSv8uPUdffTVdTU00/+Y3EA5Tdt55TPnAB9j3k58MLNT88z/3+WHbEwyy7+WXaXzhBQ7t2MGRujo6t2zhyDPPUE/8dm0DfvjHYtiHDzNj9WqOueACqs89t08xZfZxxwH0aXlxrEO6HZVWVLBs9WqWrV4NQKSnh3e3beO9TZs4+NJLhHbupKG2loYf/5htXi+eBQuYsmwZc848k2PPPZeyWbNyFlssFqO1vp7m11/n0Jtv0rZnD13vvENPYyOx5maM7m5gkKYp0Wi84OJygWHEn10uDMOIPyem977uM9/t7rNc6jYMt/v994nXfdZ3ufD7/bz3059mjG3KqlXEuruJhcPEuruxw2FiPT3x52AQu70dOxyGcBg7EoFwOD5geSSSvlXaMNkAXi+43fEi1TCKVe0bN6Ytyu255x4ViUREJoO3161j2/XXE+vpoWTuXE5as4YFl12Ws/VEREREspWuVcPe734XA/CffDJnfPObzDv11PiyU6cOWajxFRez4IMfZMEHP5icFovFaKmro377dnbeeGP6QKJRLvrBDzLGecYXv8g7lgWVlY6+y5XH52PhypUsXLkyOe1AXR1vb9hA07ZtdO7YwaF16zj88MO8DlBZSemSJcxfuZLKFSuoPPHEZAuVbMbjicVitL73Hs2vv87BXbtof/vt9wtBBw4kC0GQ6LY3bRqeOXMInHACZQsWMLW6mte+8Q2M9vYBx2JMncrlzzyTk/M0lIqKCn74X/+VdoBxY+pULnnggRFvO9LTQzgYJBwM0tPVRSQUItzVRaS7O/kcCQaJdHfjBToOHybW00MkFCLW3U2ku5tYTw/RxPOwi1Upn0kqO91g6mNERSIREYd4e906tt9wA7GeHgC69u1j+w03AAxa8BnpeiIiIiLDkXZAauLdyFb//vd9pvd20Roul8vFrMWLmbV4MW9861sZf/gPxcnFocHMrK5mZnU1XH01AMG2NvY89xz7Nm+m7ZVX6NyyhTeefZY3iJ93/+LFGD4foRdewBWJxDfS2srub32L/Vu24J82ja53332/RVDiehEShaDp0/FUVhI48UTKjjmGaYlBvWcdd1za7m6hw4fHvDvfWMg4wPgo4/L4fHh8PorLy4dcNhdjEmW6u142/wZGSkUiERGHeOVb3yIaDPaZFg0GefFf/5XuQ4ewIxHsWAw7EiEWjcafIxHefPDBtOu9/M1vMve88/BOmRJvwjsMapk0PuhzEhGRfMp457DOzpzsL1c//MeT4vJyTvj4xznh4x8H4gM2t735JjueeIKWF18k+PrrcPDgwG550Shdzz7LkUQhyDtnDkUnnkjZggVMq65m5vHHM+sDHxj2uEe9hT+ndedzalyjVYh/AyoSiYiMkaF+sMeiUYL799P57rscefddOt97L/468Rzcvz/tdntaW3np1lvTzjM8nnhz1DRCzc08vGQJ7uJiSubMGfAonjOHkqoqSior8U+fHu/bjVomjRf6nEREJN+MqVPz2qphov7wHw2X282ilSuZdvzxyWm/nDs3bfHOBi6vqxt2IWgovd35jDlzMJ99dky3PRojbb3mZIX4N6AikYjIELJprZHuB/vW665j99q1uNxuOt97j659+4ilNtE2DErmzCFw1FFUnnUW9U8+SbijY8D+i+fM4aPr1+NK3JXBcLvfH7jPMHhk+XK69u0bsJ5/+nSWXHstXY2NBBsb6WpspGnLFoJNTdjRaJ9lXT4fxZWVlMyZw6FXX03bMumVO+9U8SGPIsEgoeZmgs3NhA4cIJjyOtrayrvr1/f9PhH/nF646SaioRCB+fMpnT+f0rlzcaW5pbCIiMhwFaJVg1MLEk4yWPFurAtEvS7fsCFnt3yXvvI9xpaKRCIig8jUWsOOxZixbBmHX3+d1tdf540HHiDWb2A5OxKheetWZpx0EtOXLeOoiy+O/3A/6igC8+dTMncubp9vwL5SCzTu4mJOvukm/NOmZYzxpDVr0q536je+kbaoE4tG6W5poStROEp9BBsbBxSIenXt28cfL7ggXniYPz9ZhOh99gYC2Z3UcWisYTVbhgAAIABJREFUunXZsRjdhw4RbGoieOAAod4CUO/7xOvQgQNpC4aGy4W/ooKyqqoBBaJe4Y4Otn/1q33WKa6sJHDUUZTOmxf//vU+z59P8Zw5uNzuYR+LiIhMPoVq2aOCxODULW/iy+cYWyoSiYgM4pU770zbqmbrP/1T8r3L6834gx3b5sInnshqX71Fh1fuvJOuhgZKqqqyKkYsuOwympqaqLv7boy2NoypUzn62mszrudyuymaNQujvJzY3Ln0tLdjdHQQbW+nu6OD2M6duNraBh6Kz0esrIzWt96iceNGoqFQn/n+adMGFJBK589PFic8xcVZnQenyaZbV6Sr6/0WPykFoGBzc/y5tyDU0jKgFReAJxCgeOZMimbNYtoJJ1A0axbFM2dSPGtW8nXR7Nn4p0/H5XZTUVHBAwsXpm1BVjJ3LuevW/d+V8b33ks+N23ZQtdvfwu2nVze8HgoqaoaUPTr/dyKZ89OdkUUEREZL3cOm0zULU/GkopEIiKD6GpoyDjvzLvuYuqSJUyprubxs85K/4O9qmpY+1tw2WUsuOyyYf21rK6ujp1FRUSuuy45bafHw/Q332T27Nl0dHTQ3t5OR0dHn9c9KXe3ACguLqasrIye887D//jjA/4a1X3xxTQsWwaA1+Nhms9HeTRKcSiEp72d6IEDHKmvp3XnTvY9/fSAllVFM2dSOm9evPjQr5BUOncu7hE0h3569WoAPvLww8NedzC2bRPu6KD70CFevv32tIXCbddfz47vf5/ggQNE0gzYabhcFCUKP8UzZzLthBP6FH2KZ8+mKFEI8pSUDDvGys98Ju1fDSs/85l4i6F589KuF+3poauhIT4uVn19n+fGDRsINjX1Wd7l81E6d26fFkil8+cTOfFEwmVlFFVUDHtgdBERGd9UHHIeFe9krKhIJCL/v727D6+zrvM8/s5D07RNm7Q5UNryVAgPIhQYTAAHFkTLwygyi8xP6jg6yIKzu47jqheg4yDDOLvAyLiswzUr6oyigvPTZVwEZ6E8lQxCKVQQoaCRltIHW9q0hbRJ83T2j3MScpKTJk3OybmbvF/X1Ss597nvc3/OfZJvy5ff73drH2YuXDjsaI3F2QYFDD/l6+Trrtuv87W0tLBixQp6e3upqamhsbGRhoaGYffv7OzkqaeeonvQ4tXd3d2sWLEiZ1tFRQWzZ89m9uzZzJ8/nzlz5jB79uz+r9Oy69bcvWcP7UDVww9TtmsX6dpaOt/7XqrPPJMLLriAbdu29f95dft2esrKoLaWaakU9WedRSqV4vj6emYD5bt2sTs7kmX3hg20rV/P9uefZ/399+cuuF1Wxoz58/OOQKo5/HDm5rnt6Np77mHbs8/S29nJT5qa9jnqqrenh84dO+jYvp29ra3s3b6dvdu309Ha2v+4/7nsn95BTbQhr9nZydyTTmJhXyNoUAOoau7cok3j6urq4rdz5rD34ouHfE5r6+o4Yx/HVlRVMfvII5l95JF5n+9ub2f3xo2Zz2xQI+n1F15gb2tr7utVV+cdgdT3tWru3DE1kYrVAJQkabKyOaRCsEkkSfsw2ubPWKeKDdTS0kJzczO9vb0AtLW10dzcTGdnJ6lUijfffJM333yTXbt29X/fMWjK12DnnHNOf2No1qxZo/qP9cbGRpo7OmjPjhoCqKyspKmpifr6eurr6znuuOMA6O3tZceOHTmNozVr1tCTnVI1bdo06uvrSTU1cdD7389xqRS1tbWke3tp/93vcqZD9X3/xtNP89pPfkI6ex0gu65OdpHvWYceStfu3WwasHDzno0beeqzn2X9T39K9UEH5TaCtm+nc9eunClWA02bM4fp8+ZRXV/PrEWLmLdkCdX19UyfN4/p8+bxi698hb3btw85buaiRZz1j/844vXcH93d3ezZs4c9e/awe/du2tvb2b17d/+2vj/9o8CWLMn5nCDzc/PMM88wb9485s6dS21tLeX7MV2scsYMahsaqB2mOdm1eze7X3+dil272PTiizmf4RvPPkvXoKmKlTU1uc2/QV+r5swZco79aQBKkiSpcGwSSdI+9P2H6WgWLe6bKjZWq1atyjsi6IknnsjZNmvWLObMmcORRx7JnDlz+OUvf5m3WVRTU8Oxxx673zn6Ri6tWrWKtra2fY5oKi8vH3Xj6Fe/+hUwoHGUSnHQQQeROv54FtfW5jSweru62LN5c38DIr19O1tffpndGzaw5d//nT2bNw/Jku7qYuODD1J90EFMr69n+ty51L3jHZnv581jen091dnGz/T6eqrr66maOzdn8fB8yiorWfn5z+dMnyufPn2/Rol1dXUNafTk+zN4CmDfNZ45cyYzZ86krq6OhQsXkkqlWLlyJXsHTekDKCsr47nnniOdbYpVVFRQV1fH3LlzmTdvXn/zaLRNw8GmzZpF3fHHk0qlmH366UOe78yOHhs8lW3366+z5Ykn6N69O2f/qrq6nJFHHa2trL/33n2uASVJkqTisEkkSSMYb/NnNDo7O2nLs65Nn/PPP79/WlhlZW7pnjVrFs3NzTkNpsrKShobG8ecp6GhgYaGhjHdSWS4xtHOnTt54403Rtc4SqWozU5dmg9Dctx16KH5RwaVlXHpc8+N+X3n07NkCXsvvpiK5cv7p3V1LV1Kz5IlQ5o/faN/Bo8E6sqzsPnA5s/cuXNZuHAhs2bN6t/W92f69OlDmjmpVIqKioq8n/vZZ5/NkUceyc6dO2ltbWXHjh20trayadMmWlpa+vetqqrqbxgNbB5NH+etcqtqa6mqrWXuiScOeS6dTtO5Y0fOYtp9X3f95jdseuSRIQuiQ2YNqOdvuskmkSRJUpHZJJKkEnrrrbd48cUXefnll4fdp6amhiOOOGLY5/dn5E+plJeX9zciBjeOtm3b1t88Gq5xtHjxYqqrq6nNjjgadq2o/VwovC9HV1cXnZ2deb+uXLmSzhNPhEFNj0eHmfdfUVGR0/xZtGjRkMbPcM2f/dH3+Q63hlUqlSKVSuUc09HR0d806msgtbS05DSxZs2aNaR5VFdXR0UB1lcqKyvrn8ZXf/LJQ55Pp9PcfdhheRuA+1pEXpIkSYVhk0iSSmDLli288MILrFu3DoCjjjqKuro6nn/++TGNCOob+XMgGdg46psWN9rGERddRPq73x1yZ6+ZH/oQv/nNb4Y0e4ZrAHV1dQ2Z4rc/Ghsbh4z+qaqqmrC7fe3v515dXc2CBQtYsGBB/7Z0Os3u3btzGketra1s3Lixf32ssrIyamtrc5pHFRUVpNPpgr7XQjcAJUmStH9sEknSBOnt7WXdunW88MILbN26laqqKk466STe+c53UlNTA8CcOXMSPSKo2IZrHEFmYe++xtEbhx5KRZ47e702bx6vPfZY/+uVlZVRVVVFVVUV06ZNo6qqihkzZlBbW9v/ON/Xgd/fe++97B60jg5kRnidcsopE3JdiqmsrIyamhpqamo4/PDD+7f3Nez6mkY7duzgjTfe4NVXXwVg+fLlVFZWMnfu3JxRR/PmzWPGjBljzlOoOwVKkiRp/9kkkqQi6+zs5JVXXuFXv/oVbW1tzJ49mzPPPJPjjjuu/7bzfcazFtBkVV5eTiqVory8vL9x9M1vfpOePHf2Aggh9Dd4Kioqxj3SpampqeBrPh0IBjbsjj766P7tnZ2d7Ny5k87OTtavX09rayvr16/n17/+df8+1dXVOesc9X0d/POeTyHuFChJkqSxsUkkSQXS0tKSsz7MSSedRFtbGy+//DJdXV3Mnz+fM844gyOOOGK/bkmuoWpqavIu9F1TU0NtbW1Bz3UgrPk0kaqqqjj44INJpVIceuihQGbKWnt7e856R62traxZs4aenp7+Y2fPnj1kvaPa2tohvw89S5aw/ZOfpLe3l7KaGnryNAMlSZJUeDaJJKkAWlpaaG5u7p8a1dbWxpNPPgnA0UcfzYknnsjBBx9cyoiTSmNj44SO7nGE176VlZX1r8m0aNGi/u29vb289dZbOWsd7dixg/Xr15POLk5dXl5OXV1df/Ooo6ODl156Ked3qbm5GWDKNuYkSZImik0iSSqAVatW5V0AeebMmZx33nklSDS5ObrnwFBeXk5tbS21tbUsXry4f3t3d3fOeketra1s3ryZlpaWvK/T3d3NqlWr/HwlSZKKzCaRJBVAvqlPAHv27JngJFPHgXhHN2VUVlaSSqVIpVI52/fu3cudd96Z95jhfsckSZJUOC6KIUnjlE6nqazM33Pvu2uZpJFNnz592N8Zf5ckSZKKzyaRJI3T6tWr6e7uHnIXralwByyp0BobG4c0Xf1dkiRJmhhON5Okcfj1r3/N6tWrOfbYY1m4cCHPPPOMa+RI4+B6U5IkSaVjk0iSxmjjxo08/vjjLFy4kLPOOouKigqOOeaYUseSDniuNyVJklQaTjeTpDFobW3loYceoq6ujqVLl1JRUVHqSJIkSZI0LjaJJGk/7dmzhwceeIDKykouvPBCqqqqSh1JkiRJksbNJpEk7Yeuri4eeOABOjo6uOCCC7zjkiRJkqRJwzWJJGkELS0trFixgt7eXioqKujp6eGCCy4glUqVOpokSZIkFYwjiSRpH1paWmhubqa3txeAnp4eysvL6ezsLHEySZIkSSosm0SStA+rVq2iu7s7Z1tvby+rVq0qUSJJkiRJKg6bRJK0D21tbfu1XZIkSZIOVDaJJGkfhluY2gWrJUmSJE02NokkaR8aGxuprMxd47+yspLGxsYSJZIkSZKk4hjV3c1CCBcCtwEVwLdijDcNs99lwI+AxhjjMwVLKUmUphY1NDQAmbWJ2traqKmpobGxsX+7pKnHfxdJSgJrkaRiGHEkUQihArgduAg4AVgWQjghz36zgU8DKwsdUpJKWYsaGhpYtmwZV111FcuWLbNBJE1h/rtIUhJYiyQVy2immzUBLTHGV2OMncAPgUvy7Pc3wC1ARwHzSVIfa5GkJLAWSUoCa5GkohjNdLNFwOsDHm8ATh+4QwjhVOCwGON9IYTPD/dCIYSrgasBYoykUqm3g1RW5jwuFXOYI8kZpngOa5E5Sh3DHAnLUKIcE1KLIBnXOAkZzGGOAyGHtaj4kpAjCRnMYY7R5BjX8aPYpyzPtnTfNyGEcuBrwJ+O9EIxxjuAO/peY9u2bf3PpVIpBj4uFXOYI8kZkphj4cKFE3VKa5E5Sh3DHAnLMDDHZKtFkIxrnIQM5jDHgZDDWlR8SciRhAzmMMdoclRVVY35+NFMN9sAHDbg8aHApgGPZwMnAo+FENYBZwD3hhDeNeZUkjSUtUhSEliLJCWBtUhSUYxmJNEq4JgQwmJgI3A58JG+J2OMu4D+MVUhhMeAz7tyvqQCsxZJSgJrkaQksBZJKooRRxLFGLuBTwEPAGsym+KLIYQbQwgfLHZASQJrkaRksBZJSgJrkaRiKUun0yPvVRzpTZveHhGZpPl75jBHUjMkMUd27n2+efEHCmuROcxxgGYYmGOy1SJIxjVOQgZzmONAyGEtKr4k5EhCBnOYYzQ5smsSjakWjWZNIkmSJEmSJE1yNokkSZIkSZJkk0iSJEmSJEk2iSRJkiRJkoRNIkmSJEmSJGGTSJIkSZIkSdgkkiRJkiRJEjaJJEmSJEmShE0iSZIkSZIkYZNIkiRJkiRJ2CSSJEmSJEkSNokkSZIkSZKETSJJkiRJkiRhk0iSJEmSJEnYJJIkSZIkSRI2iSRJkiRJkoRNIkmSJEmSJGGTSJIkSZIkSdgkkiRJkiRJEjaJJEmSJEmShE0iSZIkSZIkYZNIkiRJkiRJ2CSSJEmSJEkSNokkSZIkSZKETSJJkiRJkiRhk0iSJEmSJEnYJJIkSZIkSRI2iSRJkiRJkoRNIkmSJEmSJGGTSJIkSZIkSdgkkiRJkiRJEjaJJEmSJEmShE0iSZIkSZIkYZNIkiRJkiRJ2CSSJEmSJEkSUDmanUIIFwK3ARXAt2KMNw16/rPAfwK6gTeAT8QYXytwVklTnLVIUhJYiyQlgbVIUjGMOJIohFAB3A5cBJwALAshnDBot18A74oxLgF+DNxS6KCSpjZrkaQksBZJSgJrkaRiGc1IoiagJcb4KkAI4YfAJcBLfTvEGB8dsP9TwEcLGVKSsBZJSgZrkaQksBZJKorRNIkWAa8PeLwBOH0f+18J/Fu+J0IIVwNXA8QYSaVSbweprMx5XCrmMEeSM0zxHNYic5Q6hjkSlqFEOSakFkEyrnESMpjDHAdCDmtR8SUhRxIymMMco8kxruNHsU9Znm3pfDuGED4KvAs4J9/zMcY7gDv6XmPbtm39z6VSKQY+LhVzmCPJGZKYY+HChRN1SmuROUodwxwJyzAwx2SrRZCMa5yEDOYwx4GQw1pUfEnIkYQM5jDHaHJUVVWN+fjRNIk2AIcNeHwosGnwTiGE9wF/CZwTY9w75kSSlJ+1SFISWIskJYG1SFJRjKZJtAo4JoSwGNgIXA58ZOAOIYRTgW8AF8YYtxY8pSRZiyQlg7VIUhJYiyQVxYh3N4sxdgOfAh4A1mQ2xRdDCDeGED6Y3e3vgBrgRyGE50II9xYtsaQpyVokKQmsRZKSwFokqVjK0um8U1cnQnrTprdHRCZp/p45zJHUDEnMkZ17n29e/IHCWmQOcxygGQbmmGy1CJJxjZOQwRzmOBByWIuKLwk5kpDBHOYYTY7smkRjqkXjW/ZakiRJRZFOp2lvb6e3t5eystL8N+eWLVvYu7ewy5ik02nKy8uprq4u2fuSJEn52SSSJElKoNbWVqZNmzbuW9mOR2VlJRUVFQV/3e7ubjo6OpgxY0bBX1uSJI3diGsSSZIkaeJ1d3eXtEFUTJWVlfT29pY6hiRJGsQmkSRJ0iRx3333cd9995U6xqg41UySpOSxSSRJkiRJkiSbRJIkSZNBS0sLW7ZsYfPmzdx99920tLSM+zUXLFjA0qVLOe+887j66qtpb28H4Jhjjsl7/ssuu4ylS5dyzjnncM0114z7/JIkaWLZJJIkSTrAtbS00Nzc3L/OT1tbG83NzeNuFFVXV7N8+XIeeeQRqqqquPPOO4fd9/rrr+eqq65i+fLlrFixgiuuuGJc55YkSRNvcq6GKEmSNIk8+eSTbN++fdjnt2zZMmQh6O7ublasWMHLL7+c95j6+nrOPPPMUWdoampizZo1wz6/detWFixY0P/4He94x6hfW5IkJYMjiSRJkg5ww90prFB3EOvu7ubRRx/l+OOPH3afq666ihACH/3oR7njjjvYtWtXQc4tSZImjiOJJEmSEm6kET933303bW1tQ7bX1NTwgQ98YMzn7ejoYOnSpQCcfvrpLFu2bNh9P/zhD3POOefw2GOP8cADD/D973+f5cuXM3369DGfX5IkTSybRJIkSQe4xsZGmpub6e7u7t9WWVlJY2PjuF63b02i0TrkkEO4/PLLufzyyznvvPN45ZVXWLJkybgySJKkieN0M0mSpANcQ0MDZ599NuXlmX/a1dTUcPbZZ9PQ0DBhGR599FG6urqAzPpEO3bs4JBDDpmw80uSpPFzJJEkSdIk0NDQ0L9I9XimmI1Ge3s7p512Wv/jq6++ms2bN3P99df3Ty/70pe+xMEHH1zUHJIkqbBsEkmSJE0ShW4OrV27NmcKW58NGzbk3f+GG24o6PklSdLEcrqZJEmSJEmSbBJJkiRJkiTJJpEkSZIkSZKwSSRJkiRJkiRsEkmSJE0aD112GQ9ddlmpY0iSpAOUTSJJkiRJkiRRWeoAkiRJGr+199zDtmefpbezk580NXHyddex+NJLx/WaCxYs4Pjjj6enp4eGhgZuu+02ZsyYwdatW/nyl7/M888/T1VVFYcddhg33HADVVVVnHvuuRx11FF0dXWxZMkSbr31VqZNm1agdylJkorJkUSSJEkHuLX33MPT11xDb2cnAHs2buTpa65h7T33jOt1q6urWb58OY888ghVVVXceeedpNNprrzySs4880x+/vOf89hjj3Httdeybds2AI444giWL1/Oww8/zObNm/npT3867vcnSZImhiOJJEmSEu7Z669nx0svDft83wiigXra21n5uc/x27vuynvM3BNO4LQbbxx1hqamJtasWcMTTzzBtGnT+NjHPtb/3IknngjA66+/3r+toqKCU089ld/97nejPockSSotRxJJkiQd4AY3iEbavr+6u7t59NFHOf7443nllVc46aSTRjymo6OD1atXc+655xYkgyRJKj5HEkmSJCXcSCN+ftLUxJ6NG4dsn7loEe/78Y/HfN6Ojg6WLl0KwOmnn86yZcv43ve+t89jXnvtNZYuXcratWt5//vfzwknnDDm80uSpIllk0iSJOkAd/J11/H0NdfQ097ev61ixgxOvu66cb1u35pEAx177LHcf//9wx7TtybRli1buOyyy3jwwQc5//zzx5VDkiRNDKebSZIkHeAWX3opTbfcQnlVFZAZQdR0yy3jvrtZPmeddRadnZ384Ac/6N/23HPP8eSTT+bsN3/+fL74xS/y9a9/veAZJElScdgkkiRJmgQWX3opqdNO4+Azz+QPn366KA0igLKyMr71rW/x+OOP8+53v5v3vOc93HrrrcyfP3/IvhdeeCHt7e2sXLmyKFkkSVJhOd1MkiRpkhjP+kP5rF27lu7u7iHbDznkEL7xjW/kPeaRRx7p/76srIyHHnqooJkkSVLxOJJIkiRJkiRJNokkSZIkSZJkk0iSJEklkE6nSx1BkiQNYpNIkiQpgSorK/OuBzQZdHd3U17uP0MlSUoaF66WJElKoHnz5rFhwwb27t1LWVlZSTJMnz6dvXv3FvQ10+k05eXlVFdXF/R1JUnS+NkkkiRJSqCysjJmzJhR0gypVIpt27aVNIMkSZo4o2oShRAuBG4DKoBvxRhvGvT8dOBO4DRgO/DhGOO6wkaVNNVZiyQlgbVIUhJYiyQVw4iTwUMIFcDtwEXACcCyEMIJg3a7EtgRY2wAvgbcXOigkqY2a5GkJLAWSUoCa5GkYhnNioFNQEuM8dUYYyfwQ+CSQftcAnw3+/2PgfeGEEozeV7SZGUtkpQE1iJJSWAtklQUo5lutgh4fcDjDcDpw+0TY+wOIewC6oGcSewhhKuBq7P7sXDhwpwXGfy4VMyRyxzJygBTNoe1qETMkcscycoAk7cWQTKucRIygDkGM0euJOSwFhVfEnIkIQOYYzBzFM5oRhLl6zanx7APMcY7YozvijG+K3tM/58QwrODt5XijznMkeQMCc4xEfKdx1pkDnNM4Qx5ckyEfOcpeC0iIdc4CRnMYY4DIYe1aEp+zuYwR9JzjMlomkQbgMMGPD4U2DTcPiGESqAWaB1rKEnKw1okKQmsRZKSwFokqShGM91sFXBMCGExsBG4HPjIoH3uBT4OPAlcBjwSYxzSpZakcbAWSUoCa5GkJLAWSSqKEUcSxRi7gU8BDwBrMpviiyGEG0MIH8zu9m2gPoTQAnwWuG4MWe4YwzHFYI5c5nhbEjLAFM1hLSoZc+Qyx9uSkAEmby2CZFzjJGQAcwxmjlxJyGEtKr4k5EhCBjDHYObINa4cZem0zWRJkiRJkqSpbjRrEkmSJEmSJGmSs0kkSZIkSZKkUS1cXXQhhAuB24AK4Fsxxpsm6LyHAXcChwC9wB0xxttCCDcAVwFvZHf9YozxZ0XMsQ54C+gBumOM7wohzAP+BTgSWAeEGOOOImY4Lnu+PkcB1wN1FPlahBD+CfgAsDXGeGJ2W973H0IoI/Oz8gfAHuBPY4yri5jj74CLgU7gt8AVMcadIYQjycz/fiV7+FMxxj8rYo4bGOZzCCF8AbiSzM/Pp2OMDxQxx78Ax2V3qQN2xhhPKeb1mEhTvRZls6yjhPXIWmQtGmUOa1Fxzmstys0wpeuRtWhUOaxFxTmvtSg3g7XIWjRSjoLWopI3iUIIFcDtwFIyt2lcFUK4N8b40gScvhv4XIxxdQhhNvBsCGF59rmvxRi/OgEZ+rwnxrhtwOPrgIdjjDeFEK7LPr62WCePMb4CnAL9n8lG4F+BKyj+tfgO8A9k/jLoM9z7vwg4JvvndOAfs1+LlWM58IUYY3cI4WbgC7z9Ofw2xnhKgc49Ug7I8zmEEE4gczeLdwILgYdCCMfGGHuKkSPG+OEB574V2DVg/2JdjwlhLcpRsnpkLRo2h7XIWmQt8t9GE12P8mWwFlmLrEXWImtRrklXi5Iw3awJaIkxvhpj7AR+CFwyESeOMW7u627GGN8i02VbNBHnHoVLgO9mv/8u8IcTeO73kvlhem0iThZjfBxoHbR5uPd/CXBnjDEdY3wKqAshLChWjhjjgzFz9wiAp4BDC3Gu/c2xD5cAP4wx7o0xrgVayPxOFTVH9v8UBODuQpwrIaxFwytVPbIWvb3NWpSHtaiwrEX7NOXqkbVo9DmsRYVlLdonaxHWouFyFKoWlXwkEZlf+NcHPN5A4f5v7Khlh2KdCqwEfh/4VAjhY8AzZDrZRRtCCKSBB0MIaeAbMcY7gPkxxs2QKZQhhIOLeP7BLif3B2sir0Wf4d5/vp+XRcDmCcj0CXKHei4OIfwCeBP4Uoyxucjnz/c5LCJTGPv0XY9iOxvYEmP8zYBtE309Cs1alJGkemQtys9a9DZrUZFYi4awHg1lLXqbtahIrEVDWIuGsha9rSC1KAkjicrybEtPZIAQQg3wf4DPxBjfJDM07mgyw/o2A7cWOcLvxxh/j8wQvf8aQvgPRT7fsEIIVcAHgR9lN030tRhJSX5eQgh/SWbo6w+ymzYDh8cYTwU+C9wVQphTxAjDfQ6l+v1ZRu5fUBN9PYrBWpSRiHpkLcrPWjSEtagIrEW5rEdDWYuGsBYVgbUol7VoKGvREAWpRUloEm0ADhvw+FBg00SdPIQwjUzx+UGM8R6AGOOWGGNPjLEX+CYFGho2nBjjpuwIdePQAAAH+klEQVTXrWTmlzYBW/qG52W/bi1mhgEuAlbHGLdkM03otRhguPc/4T8vIYSPk1kc7I9jjGmA7NDB7dnvnyWzYNqxxcqwj8+hFNejEriUAR37ib4eRTLla1H2nEmpR9aiQaxFuaxFxWEtyst6NIC1KJe1qDisRXlZiwawFuUqZC1KQpNoFXBMCGFxtjt6OXDvRJw4O2fv28CaGOPfD9g+cO7kfwR+VcQMs7ILshFCmAWcnz3fvcDHs7t9HPi/xcowSE73cSKvxSDDvf97gY+FEMpCCGcAu/qGOxZDyNzV4VrggzHGPQO2H5RdOI4QwlFkFmh7tYg5hvsc7gUuDyFMDyEszuZ4ulg5st4HvBxj3DAg34RejyKZ0rUoe74k1SNr0QDWorysRQVmLRqW9SjLWpSXtajArEXDshZlWYvyKlgtKkunJ3TUYF4hhD8A/ieZ2yv+U4zxbyfovGcBzcALZG6vCPBFMr+Ap5AZErYO+GQRf8CPItOVhswaUXfFGP82hFAPROBwYD3wRzHG0S6UNdYsM8nMIz0qxrgru+17FPlahBDuBs4FUsAW4MvAT8jz/rN/afwDcCGZWyteEWN8pog5vgBMB7Znd3sqxvhnIYQPATeSGd7YA3w5xvjTIuY4l2E+h+wwy09ks3wmxvhvxcoRY/x2COE7ZK7D/x6wb9Gux0SayrUomyMR9chaZC0aKYe1qGjntRYNzTJl65G1aOQc1qKinddaNDSLtchaNGyOQteiRDSJJEmSJEmSVFpJmG4mSZIkSZKkErNJJEmSJEmSJJtEkiRJkiRJskkkSZIkSZIkbBJJkiRJkiSJzK38NAWEEOYDPwJOBe6IMX6uxJEkTUHWIklJYC2SlATWIiWRTaKECyE8Dfwx0AP8OMb4e2N8qauBbcCcGGO6UPkmgxDCd4ANMcYvlTqLlFTWouKzFkkjsxYVn7VIGpm1qPisRaXjdLMECyFMA44AWoDTgNXjeLkjgJcsPpL2l7VIUhJYiyQlgbVIk50jiZLtRLJFI4TwLkYoQCGEdwO3AccCvwb+Isb482wX9o+BdAjhM8AfxhgfGnTsDOArwGVAHfACsDTG2B5C+CDwP4BFwHPAf44xrsketw64HfgT4Gjgh8AXge8AZwErgT+KMe4IIRwJrAU+CdwAlAFfjTHemn2t6cDNQMjGisC1Mca9IYRzge8DXwOuJdO1/2KM8Z8HHPu32WOnA/8K/Lds/mGPDSFcPejaPBpjvDiEcC3waWAOsAn4LzHGh/d1/aVJzFpkLZKSwFpkLZKSwFpkLZrUHEmUQCGEK0IIO4EngDOz338OuDmEsDOEsDjPMfOA+4H/BdQDfw/cH0KojzH+KfAD4JYYY83g4pP1VTKd8HcD84BrgN4QwrHA3cBngIOAnwE/DSFUDTj2Q8BSMoXvYuDfyBShFJmfsU8POtd7gGOA84HrQgjvy27/S+AM4BTgZKAJGDi88BCglkwhvBK4PYQwN/vczdnznwI0ZPe5fqRjY4x3DLo2F4cQjgM+BTTGGGcDFwDr8lwzaVKzFlmLpCSwFlmLpCSwFlmLpgpHEiVQtvP6zyGEZuDPgVbgXuDUfQxFfD/wmxjj97KP7w4hfJpMQfjOvs4XQigHPgGcEWPcmN388+xzHwbujzEuzz7+KvAXZArVY9l9vx5j3JJ9vhnYGmP8RfbxvwLvHXTKv44x7gZeCCH8M7AMeIhMt/jPY4xbs8f+NfAN4K+yx3UBN8YYu4GfhRDagONCCCuBq4AlMcbW7LH/HbgL+MK+jgWeynNJesh0uk8IIbwRY1y3r+snTVbWImuRlATWImuRlATWImvRVOFIooQJIczLdqJ38fYv+Stkfll2ZIfb5bMQeG3QttfIdGVHkgKqgd+O9Loxxl7g9UGvu2XA9+15HtcMes3XB2VcmO9cg54D2J4tIH32ZF/7IGAm8Gz22u0E/l92+0jHDhFjbCHTlb8B2BpC+GEIYWG+faXJylpkLZKSwFpkLZKSwFpkLZpKHEmUMNkua10I4XLgPTHGT2Y7vbcPMwSxzyYyC58NdDiZX8SRbAM6yMxXfT7P657U9yCEUAYcBmxk7A4DXh6QcdOAcx0BvJjnuX3ZRqbQvXNAl31/DOn8xxjvAu4KIcwh0ym/mcycXmlKsBZZi6QksBZZi6QksBZZi6YSm0TJNXCl/FOBZ0fY/2fA10MIHyGzmNiHgBOA+0Y6UYyxN4TwT8DfhxD+hEyXuSl7/khmTup7gcfJDGPcS3ao4xj9VQjhKmAxcAXw0ez2u4EvhRBWkSkK15NZzGw0+b8JfC2E8KkY49YQwiLgxBjjA6PIswU4qu9Bdr7rIjLzjTvIFDdH3WmqshZZi6QksBZZi6QksBZZiyY9L2pynQasDiHUAz0xxh372jnGuB34AJnF07aTWdTsAzHGbaM83+fJrJa/isz82puB8hjjK2QKxNfJdIMvBi6OMXbu/1vqt4LMLSMfJrNy/oPZ7V8BngF+mc2yOrttNK7NvuZTIYQ3ycyfPW6Ux36bzNzWnSGEn5CZ63oTmff7O+BgMou8SVORtchaJCWBtchaJCWBtchaNOmVpdPDrbElFVZ4+/aK0wbNPZWkCWMtkpQE1iJJSWAt0mCOJJIkSZIkSZJNIkmSJEmSJDndTJIkSZIkSTiSSJIkSZIkSdgkkiRJkiRJEjaJJEmSJEmShE0iSZIkSZIkYZNIkiRJkiRJwP8HyRROJfoeerQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1440x1008 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import scipy.stats as ss\n",
    "data_unit_indices = range(168)\n",
    "purpose_of_this_run = 'sweepPLSridgesitefit_%dsites' % len(data_unit_indices) #'3regswithridgesitefits_168sites'\n",
    "\n",
    "ni = 5120\n",
    "nf = 168\n",
    "nt = 46\n",
    "reg_methods = ['PLS', 'ridge']\n",
    "n_components_range = [5, 12, 25, 35, 50,80,100,128, 168]#np.arange(nf,1,-1)\n",
    "\n",
    "VEarraysite = np.zeros((len(n_components_range), len(reg_methods)))\n",
    "Narraysite = np.zeros_like(VEarraysite)\n",
    "Larraysite = np.zeros_like(VEarraysite)\n",
    "Rarraysite = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraypop = np.zeros_like(VEarraysite)\n",
    "Narraypop = np.zeros_like(VEarraysite)\n",
    "Larraypop = np.zeros_like(VEarraysite)\n",
    "Rarraypop = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraypopStd = np.zeros_like(VEarraysite)\n",
    "NarraypopStd = np.zeros_like(VEarraysite)\n",
    "LarraypopStd = np.zeros_like(VEarraysite)\n",
    "RarraypopStd = np.zeros_like(VEarraysite)\n",
    "\n",
    "VEarraysiteStd = np.zeros_like(VEarraysite)\n",
    "NarraysiteStd = np.zeros_like(VEarraysite)\n",
    "LarraysiteStd = np.zeros_like(VEarraysite)\n",
    "RarraysiteStd = np.zeros_like(VEarraysite)\n",
    "\n",
    "\n",
    "for inc, nc in enumerate(n_components_range):\n",
    "    trainfraci = 0.8  # image trainfrac\n",
    "    splitfract = 0.5  # trial splitfrac\n",
    "    nfoldi = 2\n",
    "    nfoldt = 2\n",
    "    \n",
    "    spearman_brown = False\n",
    "    \n",
    "    reg_params_list = [nc, [20, -10, 10]]\n",
    "    report_popfit = [True, True]  # [False, True, True]\n",
    "    report_sitefit = [False, False]  # [True, True, True]\n",
    "    PCA_ncomponents_list = [ -1,  nc]\n",
    "\n",
    "    # file = open(resultdir+'unit_test_%s_ni%d_nf%d_nt%d_collilnearity%s_%s_SB%s_VariousSd%s_noisymap%s.pickle' % (\n",
    "    #     reg_methods, ni, nf, nt, Collinearity, noise_dist, spearman_brown, various_unit_stds, noisy_map), 'rb')\n",
    "    \n",
    "    # file = open(resultdir+'unit_test_%s_%s_%s, ni%d_nf%d_nt%d_collinearity%s_%s_SB%s_VariousSd%s_noisymap%s_statsfromHvM%s_%dcmp.pickle' % (\n",
    "    #     reg_methods, reg_params_list, PCA_ncomponents_list , ni, nf, nt, Collinearity, noise_dist, spearman_brown, various_unit_stds, noisy_map,stats_from_data, nc), 'rb')\n",
    "    \n",
    "    file = open(resultdir + 'MappingMtoD_%s_%s_%s, ni%d_nf%d_nt%d_SB%s_%dcmp_%s.pickle' % (\n",
    "              reg_methods, reg_params_list, PCA_ncomponents_list, ni, nf, nt,  spearman_brown, nc, purpose_of_this_run), 'rb')\n",
    "\n",
    "    \n",
    "    data_list = pickle.load(file)\n",
    "    file.close()\n",
    "    r12_reg,r11_reg,r22_reg,r12_reg_sitfit,r11_reg_sitfit,r22_reg_sitfit = data_list\n",
    "    \n",
    "    for ir, r in enumerate(reg_methods):\n",
    "        \n",
    "        VEpop = np.nanmedian(r12_reg[:,ir]/np.sqrt(r11_reg[:, ir]*r22_reg[:, ir]))\n",
    "        Npop = np.median(r12_reg[:, ir])\n",
    "        Lpop = np.median(r11_reg[:, ir])\n",
    "        Rpop = np.median(r22_reg[:, ir])\n",
    "        \n",
    "        VEarraypop[inc, ir] = VEpop\n",
    "        Narraypop[inc, ir] = Npop\n",
    "        Larraypop[inc, ir] = Lpop\n",
    "        Rarraypop[inc, ir] = Rpop\n",
    "        \n",
    "        VEarraypopStd[inc, ir] = ss.sem(r12_reg[:, ir]/np.sqrt(r11_reg[:, ir]*r22_reg[:, ir]), axis=None)\n",
    "        NarraypopStd[inc, ir] = ss.sem(r12_reg[:, ir], axis=None)\n",
    "        LarraypopStd[inc, ir] = ss.sem(r11_reg[:, ir], axis=None)\n",
    "        RarraypopStd[inc, ir] = ss.sem(r22_reg[:, ir], axis=None)\n",
    "        \n",
    "        VEsite = np.nanmedian(r12_reg_sitfit[:,ir]/np.sqrt(r11_reg_sitfit[:, ir]*r22_reg_sitfit[:, ir]))\n",
    "        Nsite = np.median(r12_reg_sitfit[:, ir])\n",
    "        Lsite = np.median(r11_reg_sitfit[:, ir])\n",
    "        Rsite = np.median(r22_reg_sitfit[:, ir])\n",
    "        \n",
    "        VEarraysite[inc, ir] = VEsite\n",
    "        Narraysite[inc, ir] = Nsite\n",
    "        Larraysite[inc, ir] = Lsite\n",
    "        Rarraysite[inc, ir] = Rsite\n",
    "        \n",
    "        VEarraysiteStd[inc, ir] = ss.sem(r12_reg_sitfit[:, ir]/np.sqrt(r11_reg_sitfit[:, ir]*r22_reg_sitfit[:, ir]), axis=None)\n",
    "        NarraysiteStd[inc, ir] = ss.sem(r12_reg_sitfit[:, ir], axis=None)\n",
    "        LarraysiteStd[inc, ir] = ss.sem(r11_reg_sitfit[:, ir], axis=None)\n",
    "        RarraysiteStd[inc, ir] = ss.sem(r22_reg_sitfit[:, ir], axis=None)\n",
    "\n",
    "\n",
    "# Plot figures\n",
    "\n",
    "\n",
    "reg_methods_labels = ['PLS','PCR']\n",
    "colors = ['gray','DarkRed']            \n",
    "n_regs = len(reg_methods)\n",
    "fig, axes = plt.subplots(nrows=3, ncols=4, figsize=[20, 14])\n",
    "\n",
    "for iax, ax in enumerate(np.ravel(axes).tolist()):\n",
    "    ax.set_ylim([-0.0, 1.02])\n",
    "\n",
    "\n",
    "for iax, ax in enumerate(axes[0,:len(reg_methods)].tolist()):\n",
    "    \n",
    "    ax.errorbar(n_components_range, VEarraypop[:,iax], yerr=VEarraypopStd[:,iax], marker='o', color='r', label='VE')\n",
    "    ax.errorbar(n_components_range, Narraypop[:,iax], yerr=NarraypopStd[:,iax], marker='o', color='salmon', label='N')\n",
    "    ax.errorbar(n_components_range, Larraypop[:,iax], yerr=LarraypopStd[:,iax], marker='o', color='b', label='L')\n",
    "    ax.errorbar(n_components_range, Rarraypop[:,iax], yerr=RarraypopStd[:,iax], marker='o', color='g', label='R')\n",
    "    ax.set_title(reg_methods_labels[iax]+ ' popfit')\n",
    "\n",
    "for iax, ax in enumerate(axes[1, :len(reg_methods)].tolist()):\n",
    "    \n",
    "    ax.errorbar(n_components_range, VEarraysite[:,iax], yerr=VEarraysiteStd[:,iax], marker='o', color='r', ls='--', label='VE')\n",
    "    ax.errorbar(n_components_range, Narraysite[:,iax], yerr=NarraysiteStd[:,iax], marker='o', color='salmon', ls='--', label='N')\n",
    "    ax.errorbar(n_components_range, Larraysite[:,iax], yerr=LarraysiteStd[:,iax], marker='o', color='b', ls='--', label='L')\n",
    "    ax.errorbar(n_components_range, Rarraysite[:,iax], yerr=RarraysiteStd[:,iax], marker='o', color='g', ls='--', label='R')\n",
    "    ax.set_title(reg_methods_labels[iax] + ' sitefit')\n",
    "    # ax.set_xlabel('ni/nf')\n",
    "\n",
    "axes[0,0].legend()\n",
    "\n",
    "#         \n",
    "axes[2,0].set_title('VE')\n",
    "\n",
    "#[axes[2,0].errorbar(n_components_range, VEarraysite[itrace, :], yerr=VEarraysiteStd[itrace, :], marker='o', ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [1]]\n",
    "if nc == 2:\n",
    "    axes[2,0].legend()\n",
    "[axes[2,0].errorbar(n_components_range, VEarraypop[:, itrace], yerr=VEarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "axes[2,1].set_title('N')\n",
    "\n",
    "#[axes[2,1].errorbar(n_components_range, Narraysite[itrace, :], yerr= NarraysiteStd[itrace, :], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [1]]\n",
    "\n",
    "[axes[2,1].errorbar(n_components_range, Narraypop[:, itrace], yerr=NarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "\n",
    "axes[2,2].set_title('L')\n",
    "\n",
    "#[axes[2,2].errorbar(n_components_range, Larraysite[itrace, :], yerr=LarraysiteStd[itrace, :], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [1]]\n",
    "\n",
    "[axes[2,2].errorbar(n_components_range, Larraypop[:, itrace], yerr=LarraypopStd[:, itrace], marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "axes[2,3].set_title('R')\n",
    "\n",
    "#[axes[2,3].errorbar(n_components_range, Rarraysite[itrace, :], yerr=RarraysiteStd[itrace, :], marker='o',ls='--', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in [1]]\n",
    "\n",
    "[axes[2,3].errorbar(n_components_range, Rarraypop[:, itrace], yerr=RarraypopStd[:, itrace],  marker='o', color=colors[itrace], label=reg_methods_labels[itrace]) for itrace in range(n_regs)]\n",
    "\n",
    "[ax.set_xlabel('# of components') for ax in axes[2, :]]\n",
    "[ax.legend() for ax in axes[2, :]]\n",
    "\n",
    "\n",
    "title = 'Model %s, Data %s'%('conv5', 'IT' )\n",
    "axes[0,0].text(1,1.2, title, fontsize=15)\n",
    "\n",
    "fig.savefig(resultdir+'MappingMtoD_SweepNumComponents_%s_%s_SB%s_%s.png'%(reg_methods, reg_params_list, spearman_brown, purpose_of_this_run), dpi=300)\n",
    "fig.savefig(resultdir+'MappingMtoD_SweepNumComponents_%s_%s_SB%s_%s.pdf'%(reg_methods, reg_params_list, spearman_brown, purpose_of_this_run), dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([nan, nan])"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "VEarraysite[1,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "ename": "NameError",
     "evalue": "name 'reg_method' is not defined",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-66-d80e3924fa54>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mMapping\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mMapping\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      6\u001b[0m \u001b[0mget_consistency\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mMapping\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_consistency\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 7\u001b[0;31m \u001b[0mr_sites_fixed_1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mr_sites_flexible_1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist_acceptable_neurons_1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_consistency\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreg_method\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreg_params\u001b[0m\u001b[0;34m,\u001b[0m  \u001b[0mmodel1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0minds1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minds2\u001b[0m\u001b[0;34m,\u001b[0m  \u001b[0mNeu_trial\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_inds\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtest_inds\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtime_interval_fixed\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mflexible_times\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtimes\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      8\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      9\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'reg_method' is not defined"
     ],
     "output_type": "error"
    }
   ],
   "source": [
    "   \n",
    "\n",
    "importlib.reload(MappingV36)\n",
    "from MappingV36 import MappingV36 as Mapping\n",
    "Mapping = Mapping()\n",
    "get_consistency = Mapping.get_consistency\n",
    "r_sites_fixed_1, r_sites_flexible_1, list_acceptable_neurons_1 = get_consistency(reg_method, reg_params,  model1,inds1, inds2,  Neu_trial, train_inds, test_inds, time_interval_fixed, flexible_times, times)\n",
    "\n",
    "\n",
    "datalist = [r_sites_fixed_1, r_sites_flexible_1, list_acceptable_neurons_1, flexible_times, time_interval_fixed, reg_method, reg_params]\n",
    "filename = resultdir+'model%s_IT_fixedflexible.pickle'%model_layer1\n",
    "pickle.dump(datalist, open(filename, 'wb'))\n",
    "print('Model1 done')\n",
    "\n",
    "# r_sites_fixed_2, r_sites_flexible_2, list_acceptable_neurons_2 = get_consistency(reg_method, reg_params, model2, inds1, inds2,  Neu_trial, train_inds, test_inds, time_interval_fixed, flexible_times, times)\n",
    "# print('Model2 done')\n",
    "# \n",
    "# \n",
    "# datalist = [r_sites_fixed_2, r_sites_flexible_2, list_acceptable_neurons_2, flexible_times, time_interval_fixed, reg_method, reg_params]\n",
    "# filename = resultdir+'model%s_IT_fixedflexible.pickle'%model_layer2\n",
    "# pickle.dump(datalist, open(filename, 'wb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'r2AX' is not defined",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-42-8f4f1c9fffe8>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mr2XAB\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mr2AX\u001b[0m \u001b[0;34m+\u001b[0m\u001b[0mr2BX\u001b[0m \u001b[0;34m-\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msqrt\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mr2AX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msqrt\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mr2BX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mrAB\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m/\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msquare\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrAB\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'r2AX' is not defined"
     ],
     "output_type": "error"
    }
   ],
   "source": [
    "r2XAB = (r2AX +r2BX -2*np.sqrt(r2AX)*np.sqrt(r2BX)*rAB)/(1-np.square(rAB))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}