SecondTest-simple.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Second Test - simple\n",
    "\n",
    "Just using `empymod.bipole`.\n",
    "\n",
    "Very simple 1D model to try the integration of empymod into SimPEG.\n",
    "\n",
    "**Limitations at this point:**\n",
    "\n",
    "- Frequency domain (for the time-domain we will need some scripts to convert the data into the frequency domain, and then run the frequency domain inversion)\n",
    "- Inversion only for resistivity ($\\rho_h$), hence fixed values for anisotropy ($\\lambda$), electric permeabilities ($\\epsilon_h$, $\\epsilon_v$) and magnetic permeabilities ($\\mu_h$, $\\mu_v$)\n",
    "- Fixed depth structure (only inverting for layer thicknesses)\n",
    "- No `numexpr`\n",
    "- Numerical derivatives.\n",
    "\n",
    "**Things to discuss**\n",
    "- Allow to invert for either or jointly several of $\\rho_h, \\rho_v, \\varepsilon_h, \\varepsilon_v, \\mu_h, \\mu_v; z$\n",
    "- Weighting\n",
    "- Penalty\n",
    "- Fixed Layers\n",
    "- Boundaries\n",
    "- Starting model: use best fitting halfspace\n",
    "- Invert using Real/Imaginary or Amplitude/Phase\n",
    "\n",
    "**Other considerations**\n",
    "\n",
    "- SimPEG supports, I think, Python 2.7 and 3+, however, empymod only supports Python 3.4 upwards."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# I was playing around with this on Azure, so needed the install step\n",
    "# !pip install SimPEG\n",
    "# !pip install empymod"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from scipy import optimize\n",
    "from copy import deepcopy as dc\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.constants import mu_0, epsilon_0\n",
    "\n",
    "import SimPEG\n",
    "import empymod"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Forward model, create data\n",
    "\n",
    "\n",
    "### 1.1 Define model and create data\n",
    "\n",
    "Define all data as required by `empymod.bipole`, except the time-domain arguments (`freq` instead `freqtime`; `signal`, `ft`, `ftarg`)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Case 1: 2 sources, 4 receivers (8 offsets), 3 frequencies; target at 600 m depth, 300 m thick\n",
    "\n",
    "# depth = np.array([ -np.infty, 0, 1000, 1600, 1900]) # Layer boundaries\n",
    "# res = np.array([2e14, .3, 1, 100, 1])    # Air, sea, overburden, target, underburden\n",
    "# freq = np.array([0.01, 0.1, 1.])         # 0.01, 0.1 & 1 Hz\n",
    "# sx = np.array([0, 500])                  # Source x-points\n",
    "# src = [sx, sx*0, 900, 0, 0]              # Src coordinates\n",
    "# x = np.arange(2, 6)*1000                 # Receiver x-points\n",
    "# rec = [x, x*0, 1000, 0, 0]               # Rec coordinates\n",
    "\n",
    "\n",
    "# # Case 2: 3 offsets, 3 frequencies; target at 1 km depth, 100 m thick\n",
    "\n",
    "depth = np.array([ -np.infty, 0, 1000, 2000, 2100]) # Layer boundaries\n",
    "res = np.array([2e14, .3, 1, 100, 1])    # Air, sea, overburden, target, underburden\n",
    "freq = np.array([0.01, 0.1, 1.])         # 0.01, 0.1 & 1 Hz\n",
    "sx = np.array([0])                       # Source x-points\n",
    "src = [sx, sx*0, 900, 0, 0]              # Src coordinates\n",
    "x = np.array([1000, 4000, 7000])         # Receiver x-points\n",
    "rec = [x, x*0, 1000, 0, 0]               # Rec coordinates"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.2 Create data and add noise\n",
    "\n",
    "Add 5 % random noise and a noise level of 1e-15 V/m."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      ":: empymod END; runtime = 0:00:00.010511 :: 1 kernel call(s)\n",
      "\n",
      "\n",
      ":: empymod END; runtime = 0:00:00.144878 :: 1 kernel call(s)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Data\n",
    "data = empymod.bipole(src, rec, depth, res, freq)\n",
    "\n",
    "# Generate noise\n",
    "mask = 1e-1*(0.5 - np.random.random_sample(data.shape))\n",
    "maskn = 1e-15*(0.5 - np.random.random_sample(data.shape))\n",
    "\n",
    "# Apply noise\n",
    "data = data*(1 + mask) + maskn\n",
    "\n",
    "# Lets also calculate a denser model\n",
    "xf = np.arange(16, 150)*50       # Offsets\n",
    "full = empymod.bipole([sx[0], sx[0], src[2], src[3], src[4]], [xf, xf*0, rec[2], rec[3], rec[4]], depth, res, freq)\n",
    "\n",
    "# Ensure dimensionality, because empymod.dipole squeezes the output\n",
    "if len(freq) == 1:\n",
    "    data = data[None, :]\n",
    "    full = full[None, :]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.3 Define plot function and plot model and data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABDAAAAF9CAYAAADyapgDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzs3Xd4lFXax/HvTRIMLfQeIEAACR0CigVRUYoUQSlRqQosRnRdVCz7KquyqMtasSGIgDSFFxEQEAsLFlR4pQYi0iTSNColirTz/jGTbAgJJGQmM0l+n+uai8xTzrmf0TznyT2nmHMOEREREREREZFgViTQAYiIiIiIiIiInI8SGCIiIiIiIiIS9JTAEBEREREREZGgpwSGiIiIiIiIiAQ9JTBEREREREREJOgpgSEiIiIiIiIiQU8JDBEREQkqZjbIzJyZDQp0LCIiIhI8lMAQEREJIt4/3F2g48guM9tlZrsCHUdmzGyM9/NsH+hYREREJPeUwBARERERERGRoKcEhoiIiIiIiIgEPSUwREREgpyZRXmHQrzl/Xm2mf1sZsfMbI2Zdc3knLR5JMzsBjP7wsxSzOxXM5trZvUyOWdFVsNXMs5LYWbtvcfWAmqlDn1JjTOb1xVtZu96Y0rxxnjDOY6/2swmmlmCmR02sz/MbJOZPWZm4RmO3QU85n37afr40h1T38ye8n6GP5nZn2a221tHZHauQURERPJOaKADEBERkWyrBXwN7ACmA+WAvsACM+vgnPs0k3N6AZ2B+cAKoDlwE3C1mV3mnEu8wFh2Af8A/up9/3y6fevOd7I3gfIlUB5Y4j0nGnjP+z4zo4GLgS+AxUA4cDkwBmjv/QxOpYvnRuAqYKo33ox6AX8BPvWWeRxoBNwBdDOzWOfcj+e7FhEREckbSmCIiIjkH+2BMc65f6RuMLOZwFLgfjx/iGfUDejmnFuU7px78PyB/wpw7YUE4pzbBYxJ7ZHhnBuTwyJexpO8+Ktz7oV0sfXAk8TIzJ3ATufcGb1EzOwJ4O/AzcAcbzzPm1kZPAmMt5xzKzIpbzrwnHPuzwzlXY8nifJ3YEQOr0tERET8RENIRERE8o/dwJPpNzjnlgE/AG2yOOeT9MkLrwnAduAaM6vl8yjPwzs84zpgpzeWNM65BcB/MjvPObcjY/LCK7X3R8ecxOGc+zFj8sK7/UNgc07LExEREf9SAkNERCT/WJduiER6e4CyWZxzVjLAW8Zn3rctfBRbTqTW+VkW17Mis5PMrISZPWxm35jZITM77Z3T4mfvIdVzEoR53GZmH3nnwDiZbp6MJjktT0RERPxLQ0hERETyj9+y2H6SrL+UOJDF9v3ef0vnKqILk1rn+WJLY2ZhwCd4eppswjNU5CfghPeQx4CLchjHs3jm8NgHLAN+BP7w7huEZ84RERERCRJKYIiIiBRslbPYXsX776F0204DmFmoc+5khuPL+DCm1DrPF1t6PfAkL6Y65wal32FmVfnviiPZYmaVgLvxJEMuc84dybA/LifliYiIiP9pCImIiEjBdlXGDWYWAlzhffttul2/ev+tkUk5sVmUfwoIyWFMqXVe4Y0lo/aZbIv2/jsvk31nXaNX6vCUzOqog+c56MNMkheR3v0iIiISRJTAEBERKdiuMbOuGbbdBdQFPnXO7U63/Wvvv0PTH2xm1wJZ9UhIBiqaWbHsBuScSwKWA7W9saSvqweZJyR2ef9tn+H4OsDT54gNoOY5yjsjiWJmJYE3UC9VERGRoKPGWUREpGBbCMw3s/nA90AzoAvwC55lSdObgmc51ofMrBmQANQHOgPzgZsyKf9joDWw1MxWAn8C651zC88TVzzwJfC8d9nS9Xh6WfT0xtwtk+v4HvibmTXB04ujJtAVWEzmSYpP8QyLGWdmjfH2MHHOPemc229ms4F+wDoz+xDP3BzXAceAdUDz81yDiIiI5CH1wBARESnY/hdPUqAGcA9wuXdbW+fc1vQHOucO4un9sARoB4zgv3/UZ1yKNdWTwGt4enQ8BDxB5omOMzjntgGX4hkScrk3thrAjd74Mh6fAlwDzAQa4Zm/oqm3vtuyqGMLMBDPpKB3eo99It0htwP/BIrhSah09F7nZZw5N4iIiIgEAct8OXURERHJz8xsEJ4eFYOdc28FNhoRERGR3FMPDBEREREREREJekpgiIiIiIiIiEjQUwJDRERERERERIKe5sAQERERERERkaCnHhgiIiIiIiIiEvSUwBARERERERGRoKcEhoiIiIiIiIgEPSUwRERERERERCToKYEhIiIiIiIiIkFPCQwRERERERERCXpKYIiIiIiIiIhI0FMCQ0RERERERESCnhIYIiIiIiIiIhL0lMAQERERERERkaCnBIaIiEgQMrO2Zvalmf3HzGaZWVigYxIREclI7ZXkJSUwREREgtNu4Brn3FXADqBHgOMRERHJjNoryTNKYIiISLaZ2RgzG5NHde0ysw7p3o8zs79m89yvzaxRhm3njT0ndfibc26vc+4P79uTwGnI/NpERCRv26jzKUxtVlbtFajNEt9TAiMImNlmM2t/AefdamYf+rOOvOa9Wb+dzWNXmNkd/o5JRLLmTTL8YWZHzWy/mb1lZiX9UE9FYADwehZ1p74meHePBx6/0DrMbJmZnXW+mfXwXmdoum3VzCwpw3ErzOxXM7soJzFkEVdtoDOwyLspx9cmIlIYZUyE52G9Z7RZ52mvII/arAC0V6A2S3xMCYwc8NeDunOukXNuxXnqjjIzl/6h2Tk3wzl3fU7ryEmSQEQkG7o550oCzYEWwEN+qGMQ8EG6b3jOqDvd6y7v9veBq82s6gXW8RbQ38wswzH9gRnOuZPptnUBlqa+MbMo4ErAAd1zUP9ZzCwCmAr0d84d926+kGsTEZG8M4iz26ys2ivIuzYrr9srUJslPqYERs7lxYO6iEi+45zbDyzDc38E0r7tmWdmP5nZTjO7O92+B81su5kdMbMEM+t5juI7A//JQSzHgLVAtpK8mdTxHlAOz4Ndarxlga7AtAzndQE+SPd+ALAazwPlwPQHehPh95vZBjNLMbPJZlbZzJZ4P4ePvPXgTVjPAsY45xJzeW0iIoVaTu6/3uOzbKPMrKWZfevd966ZzTGzJ9NVF6xtVrbbK28Z2frMsmqvcnFtIllSAuMCZfGgfpGZjTezH8zsgJm9ZmbFvPsqmNkiM/vNzH4xs1VmVsS7L617m5m1MbM1ZnbYW8az3uJXev/9zdsDpK2ZDTKzz7znvWZm49PHaGYLzOxv6esws07Aw0Bfbznrzay3ma3NcO4oM3svs2v3djV70sy+8Jax0MzKm9kMb9zfeDO6qcdf5t12yPvvZen21TbPjMVHzGw5UCFDXZd66/nNG2v77Pz3EZG8Z2aReB6ovve+LwIsBNYD1YFrgb+aWUfvKdvxPGyVBv4BvH2Ob2iaAIlZ7MvKFqBZDo5Pq8P7jdY7eB7uUvUBtjrn1qduMM9M6+2A5emOGwDM8L46mlnlDPXcBFwH1Ae6AUvw3Jcr4GmXU5M8ccAlwKPe+27fXFybiIhk//4LWbRRZlYUmI/nj/5yeP5wz5iAD7o26wLbK8jeZ3au9upCrk0kS0pgXKCMD+peT+P55W4ORON5YH/Uu28UkARUBCrj+cV3mRT9AvCCcy4CqIvnZgSeGw5AGW+3sy8znDcTT1LCvPGVxZPpnJ3+IOfcUuCfwBxvOc3wdO2qbWYN0x16GzD9HB9BPzzd0qp74/wSmILnRr4FeMwbRzlgMfAiUB54FlhsZuXTxb0Wz03wCdJlf82suvfcJ73l3gfMM8+YPxEJHu+Z2RFgD3AQ7+8/0Bqo6Jx73Dl33Dm3A3gDz/0D59y73om/Tjvn5gDbgDZZ1FEGOJJF3b+lew1Nt++I97zsyljHVKB3aiIaz4Pe1AzntAPWO+eOAJjZFUAt4B3n3Fo8D8C3ZDjnJefcAefcj8Aq4Cvn3LfOuT/xPBS3AHDOTXfOVXDOtfe+5uTi2kREJJv3XzhnG3UpEAq86Jw74Zz7X+DrDPVk1madq70C/7dZF9JeQTY+s/O0VxdybSJZUgIj5zJ9UPcmDoYC9zrnfvHeHP6J90EdOAFUBWp5b3arnHOZJTBOANFmVsE5d9Q5tzqbca3CkxBJ7Tp2M/Clc27v+U703oDm4ElaYJ6ZgqM4cwKejKY457Y75w7hycRud8595B1j9y7/bQBuALZ5b2wnnXOzgK1ANzOriecPnP9xzv3pnFuJ59vaVLfhGdv3gbfxWA6swdP9TUSCx43OuVJAe+Bi/tuTqhZQLf0DG57kbWUAMxtgZuvS7WtMhl5Y6fwKlMqi7jLpXm+k21cK+C0H13FGHc65z4CfgB5mVgfP/WpmhnMydscdCHzonPvZ+34mZ3fLPZDu5z8yeZ+duZVyem0iIpKD++852qhqwI8ZnuP3ZKgnszbrXO0V+L/NupD2CtRmSZBRAiPnsnpQrwgUB9amu9Et9W4H+Bee3hofmtkOM3swi/Jvx9OLY6t3uEXX7ATlvYnOxtOFCzwZ1Bk5uK6pwC3eREx/PNnYP89xfHZvZtXwrA2d3m48PTeqAb8651Iy7EtVC08mOf0fP1fgSQSJSJBxzv0HT5fa1OFse4CdGR7YSjnnuphZLTy9Me4CyjvnygCbgIwTkKXagOfemBMN8Qxfya7M6piG51us/nge9A5k2N8FT08xvN969QGuMs9Ez/uBe4FmZubrrrM5vTYREcmm87RR+4Dqqb2evWpkKCIY26xAtVegNkt8SAmMC5TJg/rPeP5wb5TuQb20d8JPnHNHnHOjnHN18Iwf+5uZXZtJuducc3FAJTxDUuaaWQkyH26S0SzgZu9N9xJgXlbhZ1LvauA4nh4ct3Du4SM5sRdPIiK9msCPeBqAst7rS78v1R5geoY/fko4557yUWwi4nvPA9eZWXM8XWoPm9loMytmZiFm1tjMWgOp97WfAMxsMJ5vt7LyAXBVdoMwz3JwrThzrO/5ZFbHNKADnh52ZwwfMc9ycRc557Z6N90InAJi8AwlbI7noW0VZ45LzpULvDYREcm+c7VRX+K5199lZqFm1oOzhz8GVZsVqPbKW7faLPEpJTByJ+1B3Tl3Gk+m9jkzqwSeORxSJ6szs65mFu3N1h7Gc9M4lbFAM7vNzCp6y0vtanUKzw30NFAnq2Ccc996j5sELHPOZdVV6wAQ5Z1gL71pwATgpLcbmi98ANQ3s1u8N/m+eG6Wi5xzu/EMCfmHmRX1jsXrlu7ct/EMNeno/cMn3Mzae+cfEZEg5Jz7Cc+95H+cc6fw/E43B3biSfROAko75xKAf+N5EDyAZzKyz89R9DSgS7qxvakWmmcy4dTXfO/27sCK7AyjO1cdzrldwBd4Hmbfz3D8DZzdHXeKc+4H59z+1Bee++qtlm4Z7Fy6kGsTEZFsOlcb5TxLhPbC02v6NzxDnhcB6XsuZ9ZmZdVegf/brEC1V6A2S3zMl/9zFjrOuZ/MbBrwP3hm6B2NZ9LO1WZWAU8vg1fxrFZSD89NoSKeMWuvOOdWZFJsJ+BZMyuOZzhFP+dZfggzGwt8bp5ZhDtlEdYs4HE83cKy8i6em22yme10zrX0bp+OZyLNJ7Jx+dninEv2DoN5Ac9n8T3QNd14u1vwZIh/wdNITMM7yY9zbo83q/2M97pO4flGd4Sv4hOR3HHORWWybUS6n/fy36FtGY97BHgkO2U753723m+H40keZ1p3OvfhebjMtszq8G5vn8UpXfDc11OPy/S+7Jx7h/9OyByVYd9tGd5PwpPkOZccX5uISGGUoR2JyrDvnPffc7VRzrk1nLkS4Vekm8ctY3tynvYK/N9mXUh7lePPLAtqs8SnzGU6j6QURt4M7kGgpXNuW6DjEZHgY2ZjAJxzYwIbSc75OnYzewDP7Ox/+KI8ERHJnbxqo8zsKjxLmP4M3Aq8BtRxzu3zYR1jwDfXovZKChL1wJD0RgDfKHkhIuewItAB5MIKXxbmnHvGl+WJiEiurcijehrg6alQEs/yozf7MnnhtcJXBam9koLErz0wzKwTnqEDIcCkjJMveid1mYZnYpdkoK937JbkMTPbhWdm5Ru9c2mIiIiIiIiIBA2/JTDMLAT4DrgOSAK+AeK8k+KkHnMn0NQ59xcz6wf0dM719UtAIiIiIiIiIpJv+XMVkjbA9865Hd7ZemcDPTIc04P/LvEzF7g2w5rKIiIiIiIiIiJ+nQOjOrAn3fsk4JKsjnHOnTSzQ0B5PBPipDGzYcAwgBIlSrS6+OKLM61wx08pANSpWCL30Yucx9q1a392zlUMdBy+UKFCBRcVFRXoMERExA8KSnultkpEpODKblvlzwRGZj0pMo5Xyc4xOOcmAhMBYmNj3Zo1azKtsO/rXwIwZ3jbnMQpckHMbHegY/CVqKgosvq9EhGR/C2/t1dm1g3oFh0drbZKRKSAym5b5c8hJElAjXTvI4G9WR1jZqFAaeAXP8YkIiIiIvmIc26hc25Y6dKlAx2KiIgEmD8TGN8A9cystpkVBfoB72c45n1goPfnm4FPnD+XRRERERERERGRfMlvCQzn3EngLmAZsAV4xzm32cweN7Pu3sMmA+XN7Hvgb8CD/opHRERERPIfM+tmZhMPHToU6FBERCTA/DkHBs65D4APMmx7NN3Px4De/oxBcufEiRMkJSVx7NixQIcSMOHh4URGRhIWFhboUERECiW1RdlTUNsr59xCYGFsbOzQQMciIpIVtVXZk9u2yq8JDMn/kpKSKFWqFFFRURTGFW6dcyQnJ5OUlETt2rUDHY7PpZ8YTUQkWBX2tig7Cnp7JSIS7NRWnZ8v2ip/zoEhBcCxY8coX758of0lNDPKly9fYDOpmhhNRPKDwt4WZUdBb69ERIKd2qrz80VbpQSGnFdh/yUs7NcvIhIMdC8+v4L6GWkODBHJLwrqfdiXcvsZKYEhIiIiIkFLvQVFRCSVEhgS9MyM/v37p70/efIkFStWpGvXrjkqJyoqip9//jnXx4iIiIiIiEjeUwJDgl6JEiXYtGkTf/zxBwDLly+nevXqAY5KMnLOBToEEREREREpwJTAkHyhc+fOLF68GIBZs2YRFxeXtu+XX37hxhtvpGnTplx66aVs2LABgOTkZK6//npatGjB8OHDz/gD++2336ZNmzY0b96c4cOHc+rUqby9oALEOcfzzz/Pc889pySGiBR4o0aNIiYmhpEjR+bovKVLl9KgQQOio6N56qmncnzckCFDqFSpEo0bN77g2Au73bt3895773Hy5MlAhyIi4lcFua3SMqqSbf9YuJmEvYd9WmZMtQge69bovMf169ePxx9/nK5du7JhwwaGDBnCqlWrAHjsscdo0aIF7733Hp988gkDBgxg3bp1/OMf/+CKK67g0UcfZfHixUycOBGALVu2MGfOHD7//HPCwsK48847mTFjBgMGDPDptRUmJ06cYM6cOVSoUEGfo4gUWDt27ODzzz8nISEhR+edOnWK+Ph4li9fTmRkJK1bt6Z79+7ExMRk+7hBgwZx1113Fcp7rK+W/A4JCeHgwYPs2rULLR8uIgVVQW+r1AND8oWmTZuya9cuZs2aRZcuXc7Y99lnn6XNkXHNNdeQnJzMoUOHWLlyJbfddhsAN9xwA2XLlgXg448/Zu3atbRu3ZrmzZvz8ccfs2PHjry9IB8yszpmNtnM5qbbVsLMpprZG2Z2q5/rp2/fvrRv354XX3yRRYsW+bM6EZGASExM5KqrrmL37t20aNGClJSUbJ/79ddfEx0dTZ06dShatCj9+vVjwYIFOTquXbt2lCtXzmfXk5/4ahLP6tWrU6pUKbZs2eKjyEREgkthaKvUA0OyLTs9Jfype/fu3HfffaxYsYLk5OS07ZkNW0hdniezZXqccwwcOJBx48b5L9hsMrM3ga7AQedc43TbOwEvACHAJOdcln24nHM7gNvTJzCAXsBc59xCM5sDzPDLBXjqZ/r06TjniI2N5YknnqBs2bJcfvnl/qpSRAqxf//73yQmJvq0zAYNGjBq1KjzHjNw4ECioqK444470rZfeeWVHDly5Kzjx48fT4cOHQD48ccfqVGjRtq+yMhIvvrqq7POye5xcmHMjIsvvphvvvmG3377jTJlygQ6JBEpoNRW+Y96YEi+MWTIEB599FGaNGlyxvZ27doxY4bn7/MVK1ZQoUIFIiIizti+ZMkSfv31VwCuvfZa5s6dy8GDBwHPHBq7d+/Owys5w1tAp/QbzCwEeBnoDMQAcWYWY2ZNzGxRhlelLMqNBPZ4f/brBB9mxu23387JkyeJjIwkOjqa0aNHs2nTJn9WKyKS5zZu3EizZs3O2LZq1SrWrVt31iv1gRDOnWhPL7vHyYVr0KABZsbWrVsDHYqIiF8U9LZKPTAk34iMjOSee+45a/uYMWMYPHgwTZs2pXjx4kydOhXwzI0RFxdHy5Ytueqqq6hZsyYAMTExPPnkk1x//fWcPn2asLAwXn75ZWrVqpWn1wPgnFtpZlEZNrcBvvf2rMDMZgM9nHPj8PTWyI4kPEmMdWSRqDSzYcAwIO2zuVBVqlRh8ODBvPrqq1x11VUsXryYe+65h8mTJxMVFZWrskVE0jvft0/+tHnzZho1OrM3Yna+1YqMjGTPnj1p+5KSkqhWrdpZ52T3OLlwxYsXp1atWnz33Xe0bt2akJCQQIckIgWQ2ir/UQJDgt7Ro0fP2ta+fXvat28PQLly5TIdn1W+fHk+/PDDtPfPPfdc2s99+/alb9++Z52za9eu3Aece9X5b+8J8CQjLsnqYDMrD4wFWpjZQ95Ex/8CE8zsBmBhZuc55yYCEwFiY2NzvXxI8+bN6dKlCx988AGDBw/mlVdeIT4+ntdff53IyMjcFi8iElBHjhwhLCyM4sWLn7E9dULpc2ndujXbtm1j586dVK9endmzZzNz5swLPq6w8dUknqkaNmzIrl272LVrF3Xr1vVJmSIiwaAwtFUaQiISfDLrg5VlgsE5l+yc+4tzrq43eYFzLsU5N9g5N8I5l+X8F2bWzcwmHjp0yAdhQ7du3WjcuDEffvghDz30EMeOHWP48OEkJSX5pHwRkUDZtGnTBS8LFxoayoQJE+jYsSMNGzakT58+Z3w71qVLF/bu3XvO4+Li4mjbti2JiYlERkYyefJkn1xXfuCrSTz379/P1KlTqVy5MiVLltQwEhEpcApDW2WZjWEJZrGxsW7NmjWZ7uv7+pcAzBneNi9DKtC2bNlCw4YNAx1GwGX2OZjZWudcbG7L9g4hWZQ6iaeZtQXGOOc6et8/BJCanPCHc/1e5VRKSgrjxo3j+PHj9O3blwceeIDw8HDi4uKYPXs2Bw4coHLlysTHx9O5c2ef1CkiBZvaouzzZ3sVaLltq1avXs2UKVOIjY2lZcuWrF27ln79+hEREeHDKEWksFJblX25aavUA0POK78luXwtANf/DVDPzGqbWVGgH/C+PyrydQ8MgBIlSjBixAj++OMPPv74Y1566SUOHz7M888/z/79+3HOsX//fsaOHcuSJUt8Vq+IiMi5XHrppfTq1Ys1a9awdetWTeYpIpIPKYEh5xQeHk5ycnKhTWI450hOTiY8PNwv5ZvZLOBLoIGZJZnZ7c65k8BdwDJgC/COc26zP+r3VbfcjKpXr86gQYPYvn07X375JSVKlDjrmGPHjvHyyy/7tF4REZFzuf766+nQoQMrV67koosuIjExkdOnTwc6LBERySZN4innFBkZSVJSEj/99FOgQwmY8PBwv01C6ZyLy2L7B8AHfqk0HV9PjJZeq1at6NGjBwsWLCAsLCzTYw4cOODzekVERLJiZtx0000cPXqUjRs3Uq9ePXbv3k3t2rUDHZqIiGSDEhhyTmFhYWrUCzDn3EJgYWxs7FB/lN+5c2cOHjzIl19+yZ9//skvv/xyxv7KlSv7o1oREZEsFSlShAEDBvDKK69w/PhxvvrqKz3riIjkExpCIiJ+Y2bcdtttVKhQgVq1ahEREUH9+vUJDfXkTtu1axfgCEVEJNj5Y76mkJAQhg0bxsmTJzl06BDffvutz8oWERH/UQJDpBDzx0NhRqGhoTz88MNERERQt25dSpYsSZ06dahRowbvvvsus2fP9lvdIiKS//lrvqaLLrqIW2+9FYDFixezbds2n5YvIiK+pwSGSCHmr4fCjEaPHs3Ro0cpUqQIZkapUqWoVKkSLVu2ZPz48bz44ouaRE1ERPJcxYoVqVGjBuXLl2fChAns3Lkz0CGJiMg5KIEhIn43duxY2rRpkzZ0xMxo1aoV48aN4+abb2batGmMHj2alJSUAEcqIiKFTbNmzQgJCaFy5cq88MIL/PDDDxw6dIjx48fjzx6KIiKSc0pgiBRieTGEBKB06dKEh4dz6tQpQkJCcM6xc+dOSpcuzejRoxk1ahQrV65k8ODBJCUl+TUWEZHcGDVqFDExMYwcOTJH5y1dupQGDRoQHR3NU089leVxQ4YMoVKlSjRu3Di3oUo2VatWjTJlylCvXj2KFSvG888/z5w5c/j+++9ZtGhRoMMTEcmxgtxWKYEhUojl1RASgMOHD9OuXTseeugh6tevzy+//MLUqVNxzhEXF8dLL71EcnIyAwYM4KuvvvJ7PCIiObVjxw4+//xzEhISeOmll7J93qlTp4iPj2fJkiUkJCQwa9YsEhISMj120KBBLF261FchSzaYGY0aNeLXX3/lxIkTpKSksHbtWpxzrFy5kuHDh3PXXXcFOkwRkWwp6G2VEhgikidGjBjBLbfcQo0aNRg1ahTdu3fnq6++4n//938BaNOmDVOnTqVixYqMHDmSGTNm4JwLcNQiIh6JiYlcddVV7N69mxYtWuRoyNvXX39NdHQ0derUoWjRovTr148FCxZkemy7du0oV66cr8KWbKpXrx5hYWFcf/31NGnSJG17aGgobdq0YezYsQGMTkQkewpDWxUakFpFJCiYWTegW3R0dJ7X3aVLFw4fPszy5cuJiIjg+uuvJzIykilTpvDYY4/x3HPPsWnTJh555BFKliyZ5/GJSHCaM2eOz4eaRUZG0rdv33Me06BBAwbiM7pCAAAgAElEQVQOHEhUVBR33HFH2vYrr7ySI0eOnHX8+PHj6dChAwA//vgjNWrUOKM+9TQLLkWLFqVevXokJiZSpkwZzAznHCdPnuTUqVPkRU9FESk41Fb5jxIYIoWYc24hsDA2NnZoXtdtZvTt25cjR44wb948ihcvzhVXXEHx4sV5+umnmTZtGq+++ioJCQmMHTtW48FFJOA2btxIjx49zti2atWq856XWW8yM/NZXOIbjRo1IiEhgaNHj9KuXTsaNWrEpEmTWLduHXv27DnjwV5EJFgV9LZKCQwRCZgiRYowePBgjh07xttvv03RokVp06YNRYoUYdCgQbRs2ZJHHnmE22+/nfj4eG677TaKFNHIN5HC7HzfPvnT5s2badSo0RnbsvOtVmRkJHv27Enbl5SURLVq1fwbrORY2bJlqVatGocPH6Zfv34UKVKE//mf/+HZZ5/l2Wef5d5776VmzZqBDlNE8gG1Vf6jBIaIBFRYWBh/+ctfeOmll5gyZQphYWG0aNECgKZNmzJz5kyeeOIJXnzxRdasWcNjjz1G+fLlAxy1iBQ2R44cISwsjOLFi5+xPTvfarVu3Zpt27axc+dOqlevzuzZs5k5c6a/QpVc+Pnnnzl+/DiTJk1K21anTh1OnTrFc889x913303t2rUDGKGISNYKQ1ulrzJFJOCKFi1KfHw8UVFRvPHGG2zatCltX6lSpXj66ad58MEHWbNmDX379mX58uUBjFZECqNNmzZd8FC20NBQJkyYQMeOHWnYsCF9+vQ549uxLl26sHfvXgDi4uJo27YtiYmJREZGMnnyZJ/En5/l1ZLfAMePH890e0hICMWLF+e5554jMTHR73GIiFyIwtBWWX6b5T82NtatWbMm0319X/8SgDnD2+ZlSFJImdla51xsoOPIjXSTeA7dtm1boMPh999/57nnnmPfvn2MHDmSBg0anLF/x44djBkzhoSEBK699loefPBBypYtG6BoRSSvbNmyhYYNGwY6jHwhs8+qILRXcO5nQF+ZOHFilvv69OnDCy+8wMGDBxk6dCjNmzf3aywikr+orcq+3LRV6oEhUog55xY654YFy+zqxYsX55577qFixYq8/PLLfP/992fsr1OnDm+++Sbx8fGsXLmSPn368NFHHwUoWhERKUzKlCnDqFGjiIyM5PXXX2f16tWBDklEpNBRAkNEgkrJkiX561//SpkyZXjxxRfJ2DMkNDSUwYMHM336dKpUqcKDDz7IAw88wIEDBwIUsYiIFBYlS5bk3nvvpV69ekyZMoVPPvkk0CGJiBQqSmCISNApXbo0o0aNokyZMrz00kt89913Zx0THR3NlClTuPPOO/n888/p3bs3b7/9NidPngxAxCIiUliEh4czcuRImjVrxpw5c1i4cGGmyw+KiIjv+TWBYWadzCzRzL43swcz2T/IzH4ys3Xe1x3+jEdE8o/UJEbZsmWzTGKEhoYyZMgQ3nnnHVq2bMnzzz/Pbbfdxrp16wIQsYiI5HfFihXL1vawsDCGDx9O27ZtWbRoEW+//TanTp3KixBFRAo1vyUwzCwEeBnoDMQAcWYWk8mhc5xzzb2vSZnsF5FCKjWJUb58eV566aUsZ36vXr06zz33HOPHj+fIkSPccccdjBkzhoMHD+ZxxCIikp/179+fYcOGMWzYMDp37gxA+/bt6d+//1nHhoSEMHDgQDp37sxnn33Ga6+9luUqJiIi4hv+7IHRBvjeObfDOXccmA308GN9IpJDebk03YWKiIjgb3/7W1oSY8uWLZkeZ2a0b9+euXPnMnDgQJYtW0avXr14/fXX+f333/M4ahERye8iIyMpW7YsGzduzHKIiJlx4403EhcXx8aNG3n22Wc5cuRIHkcqIlJ4+DOBUR3Yk+59kndbRjeZ2QYzm2tmNfwYj4hkEGyrkGQlNYlRqVIlJkyYwPr167M8tlixYowcOZK5c+dy5ZVX8sYbb9CzZ0/ee+89de8VEZFsMzOaNGlCcnIy+/btO+ex7du3Z/jw4SQlJfHMM8/w008/5VGUIiKFiz8TGJbJtozp64VAlHOuKfARMDXTgsyGmdkaM1ujBkGkcIqIiEhbvu61117j66+/Pufx1atXZ9y4cUyZMoXq1avz5JNPEhcXxyeffKLJ1kREJFuio6MJDw9nw4YN5z22RYsW3HvvvaSkpPD000+zffv2PIhQRKRw8WcCIwlI36MiEtib/gDnXLJz7k/v2zeAVpkV5Jyb6JyLdc7FVqxY0S/BikjwK1GiBPfeey/R0dG8+eab/Oc//znvOU2aNGHy5Mk89dRTnDp1igceeIDbbruNVatWKZEhIjk2atQoYmJiGDlyZI7OGzJkCJUqVaJx48bnPG7p0qU0aNCA6OhonnrqqdyEKj4QGhpKo0aN+OGHH/jtt9/Oe3zdunV54IEHKFasGM8++yxfffVVHkQpInKmgtxW+TOB8Q1Qz8xqm1lRoB/wfvoDzKxqurfdgcwHt4uIeKUuX9e4cWNmzpzJ0qVLz3uOmdGhQwfmzJnDmDFjOHr0KPfeey+DBg3iiy++UCJDRLJlx44dfP755yQkJPDSSy/l6NxBgwad93516tQp4uPjWbJkCQkJCcyaNYuEhITchCw+EBMTQ0hICBs3bszW8VWqVGH06NHUqVOHN998kwULFnD69Gk/Ryki4lHQ2yq/JTCccyeBu4BleBIT7zjnNpvZ42bW3XvY3Wa22czWA3cDg/wVj4gUHEWLFmXEiBG0bt2a+fPnM3/+/GwlIUJDQ+natSvz5s3j73//O8nJydx9990MGDCAjz76SHNkiEiWEhMTueqqq9i9ezctWrQgJSUlR+e3a9eOcuXKnfOYr7/+mujoaOrUqUPRokXp168fCxYsyE3Y4gPFihWjXr16fPfddxw7dixb55QsWZJ77rmHyy+/nA8++IA33nhDK5SIiN8VhrYq1J+FO+c+AD7IsO3RdD8/BDzkzxhEpGAKCQlhyJAhhIeHs3TpUo4dO0bfvn0pUuT8ednQ0FBuvPFGbrjhBhYvXsy0adN48MEHqVmzJgMGDKBLly4ULVo0D65CRHLqiy++IDk52adlli9fnssuu+ycxzRo0ICBAwcSFRXFHXfckbb9yiuvzHTVifHjx9OhQ4ccxfHjjz9So8Z/R99GRkZqCEKQaNKkCVu3bmXz5s20apXpiOezhIaG0r9/f6pWrcq8efNITk7mzjvvpEyZMn6OVkQCTW2V//g1gSEi4k9FihTh1ltvpVixYnz44YccO3aMAQMGEBISkq3zw8LCuPHGG+nWrRuffvopb731Fk8++SSvvfYaffr0oVevXnrQFJE0GzdupEePM1eEX7Vqlc/Kz6wnmVlmc6LnP2ZWB3gEKO2cuzmrbcGqbNmy1KxZk02bNtG0aVPCwsKydZ6Zcd1111GpUiUmT57MuHHjiI+Pp2bNmn6OWEQKq4LeVimBIZLP5feHwtwyM3r16kWxYsVYsGABR48eZejQoYSHh2e7jJCQEDp06MC1117L119/zdSpU3nllVeYPHkynTp1Ii4ujujoaD9ehYhk1/m+ffKnzZs306hRozO2+fJbrcjISPbs+e8K9ElJSVSrVu3CgvUhM3sT6AocdM41Tre9E/ACEAJMcs5lOZObc24HcLuZzT3XtmDWvHlz3n//fbZu3UqTJk1ydG6zZs24//77efnll/nXv/7F4MGDadmypZ8iFZFAU1vlP0pgiASQHgp9w8zo0qULJUuWZObMmTz77LPcddddRERE5LicSy65hEsuuYTt27czZ84cFi9ezIIFC4iNjeXmm2/mqquuyvY3byJScBw5coSwsDCKFy9+xnZffqvVunVrtm3bxs6dO6levTqzZ89m5syZPis/F94CJgDTUjeYWQjwMnAdnpXnvjGz9/G0W+MynD/EOXcwb0L1nypVqlC1alU2bNiQNrFnTtSoUYOHHnqIV199lddff51OnTrRo0ePbA19FBHJjsLQVumOKRJYbwGd0m9I91DYGYgB4swsxsyamNmiDK9KeR9y8GrXrh0jRoxg7969PPPMMxw8eOHPy3Xr1uXhhx/mgw8+YOTIkSQlJfHggw/SpUsXXnjhBXbv3u3DyEUk2G3atOm8y8qdS1xcHG3btiUxMZHIyEgmT56ctq9Lly7s3buX0NBQJkyYQMeOHWnYsCF9+vQ561u0QHDOrQR+ybC5DfC9c26Hc+44MBvo4Zzb6JzrmuF1wTdjMxtmZmvMbM1PP/2Ui6vwjebNm5OSksK2bdsu6PzSpUszatQo2rVrx9KlS3nxxRc5evSoj6MUkcKqMLRVlt+WD4yNjXVr1qzJdF/f178EYM7wtnkZkhRSZrbWORfrg3KigEWpPTDMrC0wxjnX0fv+IQDnXMZvtDKWMzfjcJHMtqXbNwwYBlCzZs1WBekP8p07dzJhwgTMjPj4eGrXrp3rMk+dOsXq1auZP38+q1at4tSpU7Rs2ZJevXpx9dVXc9FFF/kgchHJzJYtW2jYsGGgw8gXMvusfNFeZdJW3Qx0cs7d4X3fH7jEOXdXFueXB8bi6bExyTk3LrNt54rhXM+AecU5x/z58zlx4gS9e/fOVe+Jzz77jFmzZhEREcFf/vIXatWq5cNIRSSvqa3Kvty0VeqBIRJ8qgN70r1P8m7LlJmVN7PXgBapyY7MtmXknJvonIt1zsVWrFjRh+EHXu3atXnggQe46KKLePbZZ9m4cWOuywwJCeHyyy9n/PjxLF68mPj4eA4cOMDf//53unTpwr///W+2bNmSreVcRUQKgMxmbMvyBuicS3bO/cU5Vzc1UZHZtmBnZjRv3pxDhw6xc+fOXJV1xRVXcP/99+Oc45lnnuHzzz/3UZQiIgWXEhgiwSfPHgrNrJuZTTx06JAPwg4ulStXZvTo0VSpUoVXXnmFzz77zGdlV6hQgcGDBzN//nxeeeUV2rRpw9y5c+nfvz8333wzkyZNIikpyWf1iYgEoSSgRrr3kcBef1QUbG1V7dq1KV26NOvWrct10joqKopHHnmEunXrMm3aNGbMmMHJkyd9FKmISMGjBIZI8Mmzh0Ln3ELn3LDSpUv7o/iAi4iIYNSoUTRs2JDp06ezcOFCn/aQKFKkCG3atGHcuHEsW7aMRx55hPLly/Paa69x4403MmTIEN555x1+/fVXn9UpUlipd9P55fFn9A1Qz8xqm1lRoB/wvj8qCra2KrUXRnJy8hkz8V+oUqVKcc8993D99dezcuVKxo8fzy+/ZJxyRETyA7VV55fbz0gJDJHgk2cPhcH2rZY/hIeHEx8fz2WXXcaiRYt48803OXHihM/riYiIoGfPnkycOJFFixZx1113kZKSwjPPPEOnTp24++67WbBgAb/99pvP6xYp6MLDw0lOTtaD4Tk450hOTs7REtLZZWazgC+BBmaWZGa3O+dOAncBy4AtwDvOuc0+rzxI1atXj5IlS7Ju3TqflBcSEsJNN93EsGHD2Lt3L08++SQbNmzwSdkikjfUVp2fL9oqTeIpcoF8NCnaLKA9UAE4ADzmnJtsZl2A5/EsR/emc25sbuM9F19MjHb69OmgXgrOOceSJUtYsGABdevWZcSIEZQqVcrv9W7bto0lS5awfPly9u3bR0hICK1ateKaa67h6quvpnz58n6PQSS/O3HiBElJSRw7dizQoQS18PBwIiMjz1rq2VeTTgeKmXUDukVHRw+90NU//GHTpk188cUXdOvWjapVq/qs3AMHDvDGG2+wZ88eOnToQM+ePQkNDfVZ+SLiH2qrsie3bZUSGCIXKL8/EILvHgqXLFkCQOfOnX0Umf+sXbuWKVOmULp0aeLj46lWrVqe1OucY+vWrXzyySd8/PHH/PDDD5gZLVq04JprrqF9+/ZUqVIlT2IRkcKlILRXEByrkKR38uRJZs6cSYUKFejSpYtPyz5x4gRz585lxYoVREVFMXToUCpUqODTOkREgolWIRGR8/LVuOKiRYvmm6ERrVq14r777uP48eM8/fTTbN6cNz2ezYyGDRsSHx/PvHnzmD17NnfccQeHDh1i/PjxdO3alVtuuYVXX32VTZs2cfr06TyJS0RELkxoaChNmzYlKSmJAwcO+LTssLAw4uLiGD58OAcOHODJJ5/k//7v/3xah4hIfqQEhojkWpkyZThy5Ei+mTk9KiqKhx56iAoVKjBhwgRWrFiRp/WbGdHR0QwfPpw5c+Ywd+5c7rnnHkqWLMmUKVMYNGgQnTt35vHHH2fFihX8/vvveRqfiEgwCeb5mho1akR4eDhr1671S/ktW7bkkUceoXLlyrz++uvMmjXLL/M4iYjkFxpQJ1KIpRtCkqtyypQpA8Dhw4cpV66cDyLzv3LlynH//fczadIkZs2axf79++nduzchISF5HktUVBRRUVH079+fQ4cO8cUXX7By5Uo++eQT3n//fYoWLUrz5s1p27Ytbdu2pW7duphlttquiEjB45xbCCyMjY0dGuhYMgoLC6Np06Z8/fXXHDhwgMqVK/u8jooVK3L//fczf/58PvroI7Zv387tt9/u03k3RETyC/XAECnEfDWEJPX8/DKMJFV4eDh33nknHTp04NNPP2XChAmkpKQENKbSpUvTuXNnxo0bx0cffcRrr71G7969+fnnn3nhhRfo168fnTt3ZsyYMSxdulRLtIqIBJi/e2GAZ7hK7969iY+P59dff2Xs2LF8+umnWu1ARAod9cAQkVxL7YGR3xIYAEWKFKF3795UrVqVmTNn8tRTTzFixIg8m9zzXEJDQ4mNjSU2NpZ7772XAwcOsHr1alavXs3KlStZtGgRZsbFF1/MpZdeStu2bWnSpMlZszqLiIhHSkoKJUqU8GmZYWFhNGvWjK+++spvvTBSNW3alEcffZRp06Yxe/ZsNm7cyMCBA8ntFxEiIvmFemCIFGK+GlccGhpKyZIl82UCI9UVV1zB3/72N44dO8bTTz/N+vXrAx3SWSpXrkyPHj0YN24cy5cv56233mL48OEULVqUadOmMWzYMK6++mpGjhzJW2+9xaZNm/LNvCQiIlnxVVu1bNkyevXqxZdffumjyP4rJibG770wUpUuXZq77rqLuLg4vvvuO/7xj3/w7bff+r1eEZFgoGVURS5QQVmWDnyzNN0HH3zAsWPH6NWrl4+iCoxff/2VV199ld27d9O9e3c6d+5MkSLBn+s9cuQI33zzDd988w1r165lx44dAJQoUYKWLVvSqlUrYmNjqV+/fr64HhHxnYLSXuW2rdq2bRuPPPIIO3bs4NZbbyU+Pp6iRYv6LL7169fz1Vdf0b179zxbFnvfvn28+eab/PDDD1x++eX06dOH8PDwPKlbRMSXsttWaQiJiPhEmTJlSExMxDmXryeYLFu2LPfddx9vv/0277//PklJSQwcODDoHwhLlSrFNddcwzXXXAPAzz//zNq1a1mzZg1r165l1apVAERERNCqVStatWpF8+bNqVevXkAmLhURyWv16tVj2rRpvPDCC8yYMYM1a9YwduxYoqKifFJ+TEwM69ev5//+7//o0qWLT8o8n6pVqzJ69GgWLlzIsmXLSExMZMiQIdStWzdP6hcRyWtKYIgUYr5ahQQ8CYwTJ07w+++/+3x8cV4rWrQogwcPpkaNGsybN48DBw4wYsQIKlasGOjQsq1ChQp07NiRjh07AnDgwAHWrFmTltD49NNPAU8PjaZNm9KsWTNatGiRNhmdiEhBFB4ezujRo7nkkkt44oknuO2227jvvvvo0aNHrpPv6efC2L9/f571wggNDaVnz540btyYKVOm8K9//YsOHTrQvXt3n/YwEREJBhpCInKBCkqXXPDNEJIff/yRxYsXc8MNN1C9enUfRRZ4CQkJvPHGG5gZd9xxBzExMYEOySf279/PunXr+Pbbb1m3bh3bt28HPA/CDRs2pHnz5jRv3pxmzZqlTdIqIvlTQWmvfNFWpXfw4EEee+wxvvnmGy677DIeeughqlatypIlS3j55ZfTJuSMj4+nc+fO2SrzxIkTzJo1iwoVKuRZL4z0/vjjD+bOnctnn31GlSpVGDhwIHXq1MnzOEREciq7bZUSGCIXqKA8EIJvHgpTUlKYMWMGl19+OY0aNfJRZMHhp59+4pVXXmHfvn10796dTp06Fbh5JA4dOsSGDRvSkhoJCQlpE4DWqVOHZs2a0axZM5o2bUqNGjXy9TAhkcImv7dX6XoLDt22bZtPyz59+jRz5szhlVdeAeDqq6/mk08+4dixY2nHhIeH88gjj2Q7iRGIuTAySkhIYNq0afz2229cd911dO/eXStUiUhQUwJDxM/y+wNher5IYDjneOutt6hfvz6XX365jyILHn/++Sdvv/02X3/9NU2bNmXw4MEUL1480GH5zbFjx0hISGDdunWsW7eO9evXk5KSAnhmwG/cuDFNmjShSZMmNGrUiJIlSwY4YhHJSkFpr3zdAyO9vXv3Mm7cuCxXKKlSpQqLFi3KVlmpvTDKly/PDTfc4MswcyR9b4yqVasycOBAateuHbB4RETORZN4ikieMjPKlCmT7aVUnXMcPHiQnTt3snfvXk6ePMmpU6cICwsjIiKCiIgIKlSoQJUqVYLij+OLLrqIIUOGULt2bd59913++c9/Mnz4cGrUqBHo0PwiPDycli1b0rJlSwBOnTrFzp072bhxIxs3bmTTpk18/vnngOe/fe3atdMSGo0bN6ZOnToFrpeKiBRc1apV48UXX6R169aZ7j9w4EC2ywrUXBgZFStWjP79+9OyZUumT5/O008/zfXXX0+3bt3UG0NE8i0lMEQKMV9O4gmeb+b37dt31vbp06fzxx9/ZHpOkSJFqFKlChERERQpUoTjx49z6NAh9uzZw6lTpwDPChs1atSgVq1aVK1aldDQwNy6zIxrrrmGmjVr8sYbb/D0009z66230rZtwe/1FRISQnR0NNHR0fTs2RPwLN26efNmNm3axMaNG1mxYgULFiwAPJODNmrUKO3VsGFDKlWqpKEnIhK0zIwqVaqwf//+s/ZVrlw5R2XFxMSwYcMG1q5dG9BeGACNGjXiscce491332XZsmWsW7eO/v37U69evYDGJSJyIZTAECnEnHMLgYWxsbFDfVFemTJl+P777zlx4sQZ3+5klbwAGDBgQKazpJ8+fZpffvmF/fv38+OPP/Ldd9+RkJBAaGgo1atXp1atWkRFRQVkxYzo6GgefvhhJk2axFtvvcXOnTvp3bt3oftGq1SpUlx66aVceumlgKdXzZ49e9J6aGzYsIFp06alJaLKly9PTEwMMTExNGzYkJiYGMqVKxfISxAROUN8fDxjx449Yw4M8CTbN27cSJMmTbJVTmovjNWrVwe0F0aqYsWKMWDAAFq1asWMGTMYP3487dq1o1evXhQrViygsYmI5IQSGCLiM6mrVRw6dIgKFSpk65yslngrUqQIFSpUoEKFCjRu3JiTJ0+yb98+du/ezQ8//MDu3btZtWoVkZGR1K1bl6ioqDxdLq506dL89a9/Zf78+Sxfvpzdu3czbNgwypcvn2cxBBszo2bNmtSsWTPtG8djx46xbds2EhISSEhIYMuWLXz22Wekzr9UpUoVGjZsmNZLIyYmhlKlSgXyMkQkyPi6t+C5pE7UmX4Vkssuu4yVK1cyePBgrrvuOoYPH05UVNR5y2rYsCHr168Pil4YqVJ7Y7z//vt8/PHHrF+/nri4OFq0aBHo0EREskWTeIpcoIIyKRr4bmK05ORk5s2bR4cOHc5Ytm3ixIlZnjNs2LAc1+OcIzk5me+//57t27eTkpJCSEgINWvWpH79+tSoUSNP519Yu3Yt06ZNo0iRIgwaNIhmzZrlWd35UUpKComJiWckNfbs2ZO2v0aNGmm9NC6++GLq169PREREACMWyd8KSnvlz0k8z+f3339n6tSpzJw5kz///JOuXbsydOhQqlates7zNmzYwOrVqwO6IklWdu3axfTp00lKSqJ58+bExcVp2WwRCRhN4ikiee6iiy4CPCt2+JOZpfXOuOSSSzhw4ADbt29n+/bt7Ny5k+LFi1O/fn3q16+fJw9jrVq1okaNGkycOJFXXnmF6667jp49exISEuL3uvOjEiVKnDFBKMDhw4fZsmVLWlJj/fr1LFu2LG1/tWrVaNCgAfXr16dBgwY0aNAgyzk1lixZcsa3p/Hx8dle/lBEJDPFixdnxIgR9O3blylTpjB37lyWLFlCz549GTJkSJa9DmNiYoKuF0aqqKgoHn74YZYvX86iRYt47LHHuOmmm7jiiis0CbOIBC0lMETEZ1KHcBw/fjzP6kyddK1KlSq0bduWH374ga1bt7J+/XrWrVtHlSpViImJoXbt2n5NKFSqVInRo0fz7rvvsnz5cr7//nuGDh1aqIeU5ERERASXXHIJl1xySdq2X375he+++46tW7eSmJhIYmIiK1asSBt+UqZMmbRkRuorISGBf/7zn2nj1/fv38/YsWMBlMQQkVwrV64co0aN4tZbb2Xy5MnMmzeP9957j+7du9O/f3+qV69+xvGhoaFpc2Hs27fvvD028lpISAidOnWiZcuWvP3228yYMYMvvviCW2+9tcCusiUi+ZuGkIhcoILQJTfduOKh27Zty3V5zjkmTZpE8+bNz1iKLqtVSFKXePOHlJQUtm3bxtatWzl8+DDFihVLG5ZQvHhxv9SZas2aNUyfPl1DSvwg9b9rakIjMTGR7du3c/LkScCT0MqsXatSpQqLFi3K63BFgkJBaK8gsENIsrJnzx6mTZvGokWLOH36NNdffz0DBw4k/XwdJ0+eZNasWZQtW5auXbsGMNpzc86xevVq5s2bx9GjR7n66qvp3r27JvkUkTyR3bZKCQyRC1RQHgjBtw+FU6dOJTo6mssvvzxt2969e1m0aBHt27enfv36Pqknu5xzJIScEK4AACAASURBVCUlsWnTJvbs2UORIkWoU6cOTZo0oWLFin6r9+DBg0ycOJE9e/bQoUMHevbsGbDlXwu6EydOsGPHDhITE3n88cezPK5jx47Uq1cvbTnYypUra1lXKRQKSnsVjAmMVAcPHmTmzJnMmzePP/74gyuuuIK4uDjatGmDmbFx40a+/PJLunXrFnS9MDJKSUlhwYIFrFy5koiICHr37k1sbKzulyLiV0pgiPhZQXkgBN8+FM6aNYvKlStzzTXXpG376KOP+PHHH7n11lsD+kf8oUOH2Lx5M4mJiZw4cYLIyEiaN29O1apV/fJgduLECebOncuKFSuoXbu2hpTkga5du7J///6ztl900UWULVv2jH2lSpVKS2akJjbq1q1LiRIl8jJkEb8rKO1VMCcwUv3222+88847vPvuu/z666/UqVOHPn360LFjR9577z3KlClDt27dAh1mtuzatYuZM2eye/duGjRowC233BJ0E5GKSMGhBIaInxWUB0Lw7UPhvHnzKFGiBJ06dQLg9OnTvPnmm1x88f+zd+dRUd95/u+fH3ZFFtnXYt/3VXBBgktEjcbELJ3cpLvTSU9Pp2d+M+d3+55e5nTunT6Z7jPdc2e6O+mZnk46uUlMG2PcdxFQVBABQRBk3xEUFXeQ5XP/wKoBFSWxiirg8zinDlB861ufUsvvt17f9+f9CWfx4sV6eY4ndffuXaqrq6msrOTOnTu4u7sTHx+PRqMxSJChVimZOvv37+fdd9/V9cAAsLGx4ec//znZ2dncuHGDxsZG6uvraWhooL6+XreSjZa3t/e4UCM4OBhfX1/VlFWZtmbK8Wo6BBhaAwMDHD58mM2bN3P+/Hns7Ox4/vnnsbKyYu3atXh5eRl7iJMyMjJCQUEB27dv5+7du6xcuZLVq1dP6bLliqLMDmoVEkVRjMLKympcE8/+/n5GRkaYP3++EUc1npWVFfHx8URHR1NbW6tb8cLZ2ZmUlBR8fX31GmQkJSWh0Wh0q5RkZWXx3HPPYWlpqbfnUEZpG3VOtAqJnZ0d8fHxxMfH6x4jpeTChQvU19frgo2GhgYKCgoYGRkBRis4goKCCAoKIiQkRPe9s7OzKqtWFAMb06/J2EOZNGtra9auXcuaNWuoqKhg8+bNfP7557zwwgv89a9/JTIykqysLN3qXabKzMyMpUuXkpCQwFdffcX+/fspLi7mpZdeUmG8oihGoSowFOUbmilXtEC/V7UOHjzIjRs32LhxIwBXr17lyy+/JCsrC1M9+RwZGaG+vp6ysjJu3LiBp6cnKSkpei+VHRwcZNu2beTm5uLj48P3vve9aXMVbjbq7++nublZV6mh/Xr16lXdNvb29gQGBupuQUFBBAYG4uTkpIINxWTMlOPVdKrAeJje3l527NjByMgIu3fv5vbt26xevZo1a9YQFhY2Lf7PqKur469//StdXV1ER0fz4osv4u7ubuxhKYoyAxh9CokQ4i/AWuCilDL6Ib8XwO+A1cBt4DtSyrLH7VcFGIqpmCknhKDfk8K8vDwuXLjAK6+8AowuY7lr1y5Wr16Nj4+PXp7DUIaHhzl//jxlZWXcuXMHjUZDamoqTk5Oen2eyspKPv74YwYGBnjxxRdZsmTJtDhxVUb19vbS1NREU1MTjY2Nuu9v3Lih28bBwUEXZoy96fvfkqJMxkw5Xk33AANGVyTZvHkzQgiqqqrIy8tjaGiIgIAAsrOzWbVqlckH28PDw+Tm5rJnzx4GBwfJyspizZo1arUSRVGeiClMIfkYeA/4ZILfZwMh924LgP+891VRlGns/ikkAwMDACZfJgtgbm5OVFQUoaGhVFVVUVFRwVdffUV4eDgpKSnY2Njo5XliYmL4xS9+wUcffcSmTZs4d+4cr732GvPmzdPL/hXDcnFxwcXFhdTUVN19Ukp6e3t1gUZzczONjY0cOHCAmzdv6rZzdHQcV6mh/d7R0dEYL0VRlClmYWFBbGwsRUVF/MM//AM/+clPyMnJYf/+/fzxj3/kj3/8I/Hx8WRnZ7N8+XIcHByMPeQHmJubs2LFChYsWMCOHTvIycmhqKiIZ599loULF2JmZmbsISqKMoMZdAqJEMIf2DNBBcafgHwp5V/v/VwLZEopLzxqn4+rwKi+cJ1IT/snHfrXtj7em1cWaKb8eRXjMZUrWkKIQODngIOUcuO9+54F1gBuwPtSykOP2oc+r2qVlJRQVlbGW2+9hRCCuro68vPzefnll7G3n/r35pPo7++nrKyMc+fOYWlpSXJyMpGRkXo7ORsZGeHIkSNs374dOzs73njjDcLCwvSyb8U0SCm5dOnSA9UaTU1N4xqHOjk54e/vT0BAwLivaqlXRR9M5Xj1pGZCBQaMTif8/PPPcXd31zW8htElxw8cOMC+fftoaWnBwsKChQsXkpWVRUZGhskeQ1tbW/niiy9obGxEo9Hw0ksvmeyUUUVRTJcpVGA8jjfQPubnjnv3PRBgCCG+D3wfQKOZOCRYH++t3xFOUvWF6wAqwFC+tommWgkhVjE6xcoc+EBK+euJ9iGlbAK+J4TYOua+HcAOIcR84LfAIwMMfdJ2Jh8cHMTKykq3GsR0qMC4n42NDQsXLiQ8PJzCwkJOnjxJTU0N6enpepkOY2ZmxooVKwgLC+ODDz7g3//933n66adZt26dWvFihhBC4ObmhpubG2lpabr7pZRcvHhxXLDR0tLC4cOHuX79um67OXPm4OfnR0BAgO6rv78/Go1GNYFVlGnK0tKSmJgYSkpKuHz5sm55bS8vL9544w2++93vUltby759+8jJyeHYsWOYm5uTmppKVlYWmZmZJtUY28/Pjx//+MeUlJTw1Vdf8Zvf/IaUlBSee+45NW1OURS9M2YFxl7gV1LK4/d+PgL8X1LK0kft0xTTd9V7Y3bSxxUtIUQGcBP4RPs+EUKYA3XACkaDvdPAtxgNM3513y7ekFJevPe4rdoKjDH7/zdg0+P6y+jzfXX+/HmOHTvGK6+8wrx58zh9+jTl5eW8+eab0/pKspSS1tZWCgsLuXHjBsHBwaSnp+ttzm9/fz9btmzhxIkT+Pv78+abb+Lq6qqXfSvTh5SSK1eu0NLSQnNzM62trTQ3N9PS0kJ3d7duO3Nzc7y9vfH393+gckNNRVLupyowTM/AwACff/45Go2GZcuWTbjdyMgI1dXVHDlyhNzcXDo7OzEzMyM+Pl4XZui74fSTGBgY4ODBgxw6NHrdZNWqVaxcuVItu6ooymNNhwqMDsB3zM8+QJeRxqIoRiGlPHYv6BsrFWi4V1mBEGIzsF5K+StGqzUe616T3F8D+yfTHFeftCcpAwMDzJs3j4GBAaysrKZ1eAGjV9L9/f3x8fGhvLyc8vJy2tvbSUtLIzQ09Ilfn42NDa+//jpRUVF89tln/PKXv+Rb3/oWaWlp0/7PTpk8IQTOzs44OzuTlJQ07ne3b9+mra1NF2hov548eZKhoSHdds7OzrpAY+zNzc1NzU1XFBNhbW1NZGQkZ8+eJTk5ecJeF2ZmZkRHRxMdHc3f//3fU19fT25uLrm5ufz2t7/lt7/9LSEhISxevJjFixcTHR1t1Ao+a2tr1q1bx6JFi/jqq6/YvXs3x48f59lnnyU1NVX9H6QoyhMzZoCxC/jRvQ9nC4Brj+t/oSizxMOmV03Y4FYI4Qy8CyQIIX56L+j4O2A54CCECJZS/tdDHjepqVlflzbA0DbyHBgY0FvzS1NgYWFBcnIyQUFBHDt2jKNHj1JfX8+SJUv00mwtKSmJgIAAPvzwQz7++GPOnj3Lq6++qq6qK8ydO5fw8HDCw8PH3T80NERXVxfNzc3jqjb2798/rs+GtbU1fn5+aDQaNBoNfn5+up9NdW69osxkMTExVFVVUV5eztKlSx+7vRCC0NBQQkND+cEPfkBLSwsFBQUUFBTwySef8NFHH+Hg4MDChQtZsmQJ6enp2NnZTcEreZCzszPf//73qaurY+vWrXz00Ufk5OTw/PPPExERYZQxKYoyMxgswBBC/BXIBFyEEB3AO4AlwL0PU/sYXUK1gdFlVL9rqLEoyjTzsMvtE871klJeBn5w332/B37/qCeRUv63EOIC8IyVlVXSo7b9Oh4WYEzH/hePM3/+fNatW0dNTQ3FxcVs3bqVlJQUYmJinrhiwsnJif/9v/83hw4dYteuXTQ0NPDtb3+b6OgHZuMpChYWFrpQYuyHICklly9f1oUabW1ttLa2UltbS15eHsPDw7ptHR0ddWHG2GDDx8dnRr5/FcUUzJ07l7CwMM6fP09SUtLXDqq11VWvvfYa169fp6ioiIKCAk6ePMn+/ft1K2stWLCAtLQ0oqKisLCY2muXoaGh/OQnP6GkpIQdO3bwH//xH0RHR/P888+b/HKxiqKYJoP2wDAEU5z/qHpgzE76mlN8f68YIUQ68H9LKZ++9/NPAe5VVhiEPt9XfX19bNmyhaeeeoqQkBC2b9+OjY0N2dnZetm/Kbp9+zYFBQW0trbi6enJ0qVL9XZFu729nb/85S90dXWRkZHBxo0b1QdK5YkNDg7S2dmpCza04UZrayuXL1/WbSeEwMvLa1zVhvaru7u7KgefJqZ7DwwhxDPAM8HBwW/V19cbezh6dePGDTZv3kxUVBQLFy7Uyz6Hh4epqqri+PHjnDp1ipqaGqSU2NrakpyczIIFC1iwYAEajWZKpygODg6Sm5vL/v376e/vZ9GiRaxbt84kl4pVFGXqTYceGIqiPNxpIEQIEQB0Ai8DrxjiicacFOptn9oP19oKjP7+/hl/cjJ37lxWrlxJfX09J06cYOvWraSlpREREfHEJ4e+vr787Gc/Y+fOneTk5HD+/Hm++93vEhgYqKfRK7ORpaWl7urt/W7evDku1NB+raio4Pbt27rtrK2t8fX1HVe14evri6+vL/Pnz1e9WxS9kVLuBnYnJye/Zeyx6JudnR0hISHU1NSQkJCgl8bQ5ubmxMXFERcXx9tvv01fXx8lJSWcOnWKU6dOcfToUQA8PDx0YUZKSorBVzaxtLTk6aefZtGiRezbt4/8/HxOnz7NihUrWLFixYyabqooiuGoCgw9UBUYs5OeViHRTbUCeoB3pJQfCiFWA//B6Mojf5FSvvuk430Ufb6vhoeH+fDDD0lOTiYxMZGPP/6YkJAQFi1apJf9m7qbN29y9OhROjs78fHxYenSpdja2upl33V1dXz88cdcuXKF7Oxs1q5dq5ZbVaaMdkpKS0vLA1UbnZ2d46ak2Nra6sKM+29OTk4q3Jhi070CQ8sUzwH1QVu5GB8fT2pqqkGfS0pJR0eHLsw4ffo0N2/eBCAwMJCkpCSSkpJITEw0+BKoFy9eZPv27ZSVlWFvb8+6detYuHChOq4pyiw12WOVCjD0QAUYs9NMOCE0VFnuhx9+SFRUFKmpqXzwwQckJiaSnDyt/6i+FiklNTU1FBUVYW5uztKlSx96pfubuHPnDl988QWFhYVoNBreeOMNPD099bJvRfmmhoaG6OzspL29nY6ODtra2ujo6KC9vZ2urq4Hwg0fH5+HhhvOzs4q3DCAmXC8AtM8B9SXw4cP09nZyauvvoqlpeWUPe/Q0BA1NTWUlJRQVlZGeXk5d+7cASAgIEAXZiQmJuLi4mKQMTQ2NvLVV1/R2NiIu7s769evJyEhQU1RU5RZRgUYU0gFGLPTTDkhBP2/rz799FP8/PxITU3lk08+YeHChbOyAWVfXx+5ubn09vYSGRlJWlqa3hqonTlzhs8++4yBgQE2bNjAU089pU72FJM0NDTEhQsXaGtreyDguL9yY86cORNWbri4uKhw4xuaKccrUzwH1Jeenh527txp9OPl0NAQ58+fHxdoaKeO+fn5kZiYSFxcHPHx8Xh7e+vtPSmlpKKigh07dnDhwgU0Gg3PPvsskZGR6n2vKLOE6oGhKIrRWFlZcffuXQYGBgBmbdNJR0dH1q9fz+nTpzl79iwXLlxg2bJleinLTUhIIDAwkE8//ZQtW7Zw5swZXn/9ddzc3PQwckXRHwsLC10Icb+hoSG6u7tpb28fV7VRX19Pfn7+uHDDxsbmgVDDx8cHHx8f3NzcVICnTGvu7u64ublRVVVFZGSk0f49W1hYEB0dTXR0NN/5zncYGhqitraWsrIySktLOXz4MNu3bwdGl0qNjY3V9dsIDw//xtUjQgji4+OJjY3l1KlT7N69m9///veEhoby7LPPEhQUpM+XqSjKNKYqMPRAVWDMTjPhipahppBs374da2trkpOT2bFjB6tWrUKj0eht/9NRe3s7+fn53L17l/T0dL00+ITRq1aFhYVs2bKFoaEhVY2hzBhDQ0P09PTQ3t7+QMDR2dnJ4OCgbltLS0u8vLzw9vbGx8fnga+zvTngTDhegWmeA+pTU1MTOTk5rFy5Um/TDvVtZGSEpqYmKioqdLfOzk5g9GJFZGQkcXFxxMbGEhsbi6Oj4zd6nsHBQY4fP86+ffu4fv06sbGxPPvss3h7e+vz5SiKYkLUFJIppAKM2WmmnBCC/t9Xe/fuZXBwkMTERA4cOMD69etxd3fX2/6nqzt37pCXl0dHRwfBwcEsWbJEb3Odr169ymeffUZVVRXBwcG8/vrr6s9cmbGGh4fp7u6ms7OTjo4O3XQU7fe3bt0at72Li8uE4cZsaCo6U45XpngOqE8jIyNs3rwZOzs7nnnmGWMPZ9J6e3vHBRrnz5/XVU/5+/sTGxtLdHQ0UVFRBAUFfa2plAMDAxw5coRDhw7R399PSkoK69atw9XV1VAvR1EUI1EBxhRSAcbsNFNOCEH/76ucnBwuX75MYmIieXl5vPjii9/4KsxMI6XkzJkzlJSU4OjoyIoVK/S2dJ2UkqKiIr744guGhoZ49tlnycrKUtUYyqwipeTatWsPhBqdnZ10dnZy8eJFxp77zJkz54FQQ/u9p6fnhCHj/v37ef/99+np6cHd3Z23336b7OzsqXqZX8tMOV6Z4jmgvp09e5aioiI2bNgwbT+k9/f3U11drQs0KisruXbtGjA6FSwiIoKoqChdqOHh4fHYEPHWrVscPHiQ3NxchoeHWbx4MdnZ2QZfKUVRlKmjAowppAKM2WkmnBAaagrJsWPHaG1tJSEhgZMnT/L666/P+hLu+3V2dpKbm8vg4CAZGRkEBwfrbd9Xr15l06ZNVFZWEhQUxLe//W1VjaEo9wwMDHDhwoUHKje0AYe2dw+AmZkZ7u7uD4QbLS0tfPTRR+O2tbGx4ec//7lJhhgz4XgFpnkOqG93795l06ZNBAQEkJmZaezh6IWUks7OTqqqqjh37hxVVVXU1tZy9+5dYLSXRlRU1LhQY968eQ/d17Vr19i7dy8nTpxASsnixYtZtWqVCjIUZQZQAcYUUgHG7DRTTghB/++roqIizp07R3x8PKWlpbz55puqCuAhbt26xZEjR+ju7iYiIoL09HS9rVIipeTUqVN88cUXDA4Osn79epYtW6b+HhTlEUZGRrh8+fJDqzc6Ojq4evXqIx9vb2/Pv/zLv+Dl5fXI6o2pNlOOV6Z4DmgIx48fp7a2lldffXXGhv+Dg4PU19ePCzVaW1t1v/f399c1Ew0PDyckJGRcQ/ArV66wf/9+Tpw4gRCCRYsWkZ2drbeKRkVRpp5ahURRFKOxsrJieHiY27dvY2VlpT40T8DW1pa1a9dSXFzM2bNnuXTpEsuXL8fe3v6J9y2EIC0tjYiICD777DO2bt1KaWkp3/72t/H09NTD6BVl5jEzM8PV1RVXV1cSEhIe+P2tW7fo7OzklVdeeejjr1+/zo9+9CNg9D3o5uaGt7e3rsHo2K8uLi7q/0bloSIjI6murqa2tpa4uDhjD8cgLC0tiYyMJDIyUnff9evXqa6u1oUaJ0+eZM+ePQCYm5sTFBREeHg4ERERRERE8Pzzz7Nq1Sr2799PQUEBJ06c0E0tUdNWFWXmUgGGosxiY6aQ6HW/VlZWANy4cWPWLqE6WWZmZqSlpeHh4UF+fj7btm0jKytLb6u2ODg48MMf/pDi4mI2b97ML3/5S7Kzs1m1apXJXB1WlOnC1taW0NBQPDw86O7ufuD3rq6uvPvuu3R2dtLV1aX7eurUKS5dujRuWysrKzw9PR8acHh7e2NnZzdVL0sxMU5OTnh6elJdXU1sbOyMbzKrZW9vT1paGmlpacBoJWF3dzc1NTXU1NRw/vx5jh07xq5du4DRUCMwMJDw8HDS0tLo6+vj2LFjHD9+nIyMDFatWoWDg4MxX5KiKAagppDogZpCMjvNlJJc0P/7qq6ujvz8fBwcHLC0tOS5557T275nsuvXr3P48GEuX75MSkoK8fHxej1xvX79Olu2bOH06dN4eHjw2muv6bX3hqLMFvv37+fdd9+lv79fd9/jemBoe2+MDTbGfr1x48a47e3s7PDy8uJv/uZvyMjIeKLxzpTjlSmeAxpKY2MjR44cUcuQ30dKSU9Pz7hQo6amRje9a86cOQQFBWFtbY2ZmRnR0dE8//zzeHh4GHnkiqI8jppCoiiK0YytwFDTFSbP3t6e9evXc/ToUU6fPk1vby+ZmZl6q5Swt7fnzTffJC0tjc8//5zf/OY3ZGRksGHDBubOnauX51CU2UAbUnydVUisra3x9/fH39//ob+/cePGQ4ONOXPmGOIlTDkhRCDwc8BBSrnx3n3PAmsAN+B9KeUhIw7RpPj7+zNnzhzOnTunAowxhBB4eHjg4eHBU089BfxPqKENM2pqamhoaGDOnDlUVFRQXl7O4OAgnp6eREZGEhoaSmhoqOqXoSjTlAowFEXRO+20kZGRETWF5GuysLAgKysLFxcXiouL2bFjB08//bRe+mJoRUdH84tf/ILdu3dz5MgRKioqePnll0lISJg1pcqK8qSys7P1uuKInZ0d4eHhhIeH622f+iKE+AuwFrgopYwec/8q4HeAOfCBlPLXE+1DStkEfE8IsXXMfTuAHUKI+cBvARVg3GNubk54eDhnzpzh5s2bE67KoYwPNbQrt0gpuXTpEqdPn+bkyZP09vZy+fJldu7cSXd3N/39/bi5uRESEkJoaChhYWGEhITg6+uretMoiolTAYaiKHqnrcAAZmwHdUMSQhAXF4ezszNHjhxh27ZtLFu2DF9fX709h42NDS+88AIpKSl89tln/OlPfyIuLo6XXnoJZ2dnvT2PoigzwsfAe8An2juEEObA+8AKoAM4LYTYxWiY8av7Hv+GlPLiI/b/T/f2pYwRFhbGmTNnqKurIzEx0djDmVa0TXTXrFnDmjVruHbtGocPH+bo0aM4Ozvj7OyMlJLm5maKiooYHh4GRqegaEONkJAQwsLCCA4OVucyimJCVIChKLOYoZt4AqoC4wn4+PiwYcMGDh06xIEDB0hJSSEuLk6vVRL+/v789Kc/JScnhz179vDOO++wevVqVqxYoZp8KooCgJTymBDC/767U4GGe5UVCCE2A+ullL9itFrjscTof2a/BvZLKcsm2Ob7wPeBWTeVwt7eHk9PT2pra1WF3BNycHBg48aNrFq1iry8PHJzc7l9+zZpaWn80z/9E+bm5tTX11NXV0ddXR0HDhxg69bRYiEzMzN8fX11lRraKSjOzs7q70RRjEAFGIoyi0kpdwO7k5OT39LnflWAoT9j+2IUFxfT29vL0qVL9RoumJub8/TTT5OSksKWLVvYuXMnRUVFvPzyy+OWuFMURRnDG2gf83MHsGCijYUQzsC7QIIQ4qf3go6/A5YDDkKIYCnlf93/OCnlfwP/DaNNPPU4/mkhLCyM/Px8Lly4gJeXl7GHM+3NmzePZ555hhUrVnD06FFycnL4wx/+QFBQECtXrmTt2rWYmZkhpeTChQvU1dVRW1tLfX09586d4/Dhw7p9zZ8/n+DgYIKDgwkKCtJ9nWzfmv3793+tPjqKooxSAYaiKHqnAgz9srS0ZNmyZbi4uHD69Gn6+vpYuXKlXvtiwOjSfT/4wQ+oqqpi8+bN/O53vyMpKYkXXnhBNTtTFOV+D7v0PGHAIKW8DPzgvvt+D/xez+OaUQIDAzlx4gR1dXUqwNAjGxsbnn76aZ566ilOnDjB4cOH+c///E/c3d1Zvnw5aWlpeHl54eXlpeurAaPNdsdWajQ0NLB9+3bdikRCCLy9vR8INnx9fbGw+J+PXfevZNTd3c27774LoEIMRXmMRwYYQggf4GVgCeAF3AGqgL2MlvuNGHyEiqJMO0IILC0tGRwcVAGGngghiI+Px8XFhSNHjrB9+3aWLVuGj4+P3p8rOjqad955h0OHDrF//36qqqpYs2YNy5YtG3cCpijKrNYBjG3M4wN0GeKJDDXdcTqwsLAgKCiIhoYGFi5cOO4CgfLkrKyseOqpp8jIyKCsrIxDhw6xadMmdu3aRWZmJpmZmeMaqNrZ2ZGYmDiuJ8nIyAidnZ00NDSMux07doyRkRHd8/j7++uCjU8//XTcMswA/f39vP/++yrAUJTHEFI+PCwXQnzEaHngHqAEuAjYAKHAU0AS8BMp5bGpGeooU1wD/KU/FQLwxd+kG3kkylSa7FrF04Eh3lebNm3i1q1brFu3Tq2/rmfXr1/n0KFDXL16ldTUVGJjYw02D7e3t5ctW7ZQUVGBh4cHGzduJDo6Ws37VZRpRB/Hq3s9MPZoVyERQlgAdcAyoBM4DbwipTz3ZKOdmCmeA06Fnp4edu7cSUZGhkmuUjOTSCmpq6vj8OHDVFZWYmlpyaJFi1i+fDmurq5fa1/9/f20tLTQ0NBAY2OjLti4dOnSIx+Xn5+vVp1RZqXJHqsedSnt36SUVQ+5vwrYJoSwAmZXNyVFUSbN2tqaW7duqQoMAxjbF+PUqVO6vhiGqI5wcXHhhz/8IWfPnuXLL7/kvffeIyIighdeeAFvb2+9P5+iKKZHCPFXIBNwEUJ0AO9IKT8U+vtp1QAAIABJREFUQvwIOMjoyiN/MWR4MZu5ubnh4OBAQ0ODCjAMTAhBWFgYYWFhdHV1kZOTw/Hjxzl69CgJCQmsXLmSgICASe3LxsbmoUsj9/X18eKLL3LlypWHPi4zMxMPDw8CAwPH3QICArC1tX3i16go092EZ7sThBdjf38XaND7iBRFmRG0Za4qwDCMsX0xiouLdX0x7OzsDPJ8sbGxREZGkp+fz969e/nlL3/JkiVLeOaZZ/Tei0NRFNMipfzWBPfvA/YZ+vln8xQSGP1QHRgYSHl5Obdv32bu3LnGHtKs4OXlxeuvv8769evJzc3l2LFjlJWVERgYSFZWFomJiZibm3/t/To6OvKP//iP43pgwOj50saNG3F0dKShoYHm5mZKSkq4e/eubht3d/dxgUZQUBABAQGqYkOZVR57uU4IsRb4JeB3b3sBSCmlOmNVlGnOkCeFKsAwPG1fDCcnJ3Jzc9m2bRvLly83WGWEhYWFrrnZ3r17yc/Pp7i4mNWrV5OVlaWWXVUUxSAMtWLWdBIUFMSZM2doamoiOjra2MOZVRwcHNiwYQPZ2dmcPHmSvLw8PvjgAxwdHcnIyCAjI+NrXzzQ9rl43Cokw8PDdHZ20tTURHNzM01NTTQ1NVFWVsbAwIBuOzc3twdCjcDAQINd1FAUY5qwB4ZuAyEagOeASvm4jaeAKc5/VD0wZifVA+PRcnNzaWlp4Y033tDrfpWHu3btGocOHaKvr4+0tLQp6VPR3d3N1q1bqaysxMXFhWeffZakpCTMzMwM+ryKonw9M+V4ZYrngFPpyy+/xMrKivXr1xt7KLPayMgI586dIzc3l+rqaiwsLEhJSSErKwuNZmpm1w8PD3PhwgVdoKG9NTc3jws2XF1dHwg1AgMDVeWkYpL00QNDqx2oMoXwQlGU6cPb21s1epxCDg4OrF+/nvz8fAoLC+nt7WXJkiUGXTXEw8ODH/3oR9TU1LB161Y++OADDhw4wLp16wzaWFRRlNlltk8h0QoKCqKkpISbN2+qKQNGZGZmRkxMDDExMVy4cIG8vDyKioooLCwkODiYrKws4uPjv9H0kskyNzfHx8cHHx8fMjIydPePjIyMCza0VRs7duzgzp07uu2cnZ0JCgrC39+fgIAA/P398ff3x8XFRR27FZM3mQqMFEankBwFdJGelPL/NezQHs4U03dVgTE7zZQrWmCa7yvlm5FScubMGUpKSnBxcWHlypVTcqI7MjJCSUkJu3fv5uLFiwQEBLB+/XoiIiIM/tyKojzaTDlezfZj1bVr1/jiiy9IS0sjNjbW2MNRxrh9+7Zueklvby/z588nIyODRYsW4eDgYOzhMTIyQk9PD42NjeOmojQ3N3P79m3ddra2tvj5+Y0LNfz9/fHx8VHTRBWDm+yxajIBxiHgJlAJjGjvl1L+P086yG/CFA9eKsCYnWbKCSGY5vtKeTKtra3k5uZibm7OihUr8PT0nJLnHR4eprCwkD179nD16lXCwsJYv349QUFBU/L8iqI8aKYcr9SxCrZt24YQgg0bNhh7KMpDjIyMUFlZSW5uLufPn8fMzIz4+HgyMjIICwt7YIrltWvX+POf/8xbb71llKBDSklvby/Nzc20tLTobq2trfT09Oi2Mzc3x9vb+4Fgw9/fX/XZUPRGn1NInKSUK/UwJkVRFGWK+Pn5sWHDBg4ePMiePXtIT08nKirK4KWh5ubmLF68mAULFlBQUMC+ffv413/9V2JiYli7di3+/v4GfX5FUZSZLCgoiFOnTqlpJCbKzMyMuLg44uLi6OnpoaCggJMnT1JWVoabmxtLlixh4cKFur+7PXv20NDQwJ49e3j11VenfLxCCFxdXXF1dSU1NXXc727dukVra+u4YKOlpYUTJ04wNDSk287Z2fmBUMPf3x93d3fVE0sxiMlUYPwayJVSHpqaIT2aKabvqgJjdpopV7TANN9Xin7cvXuX3Nxc2traCA0NZfHixQbti3G/gYEB8vLyOHjwILdv3yYiIoLs7GxCQ0PVPFtFmSLT/Xg1pgfGW/X19cYejlH19fWxZcsWFi9eTGRkpLGHo0zC4OAgpaWlFBQU0NDQgIWFBcPDwzzsM5ilpSXvvfeeEUY5eUNDQ1y4cOGBqo2WlhauX7+u287a2vqh01F8fX2xsbEx4itQTJU+p5DcAGwZ7X8xiJGXUTXFD1oqwJidTOWEUAgRCPwccJBSbrx3XwTwvwAX4IiU8j8ftQ9TfF8p+iOlpLS0lLKyMlxcXFi+fPmUdyDv7+/n6NGj5OTkcP36dYKCgsjOzp6S1VIUZbYzlePVk1LHqtH/z7/44gscHBweWHJTMX2dnZ0cO3aMwsLCcauFWFpakpCQwMaNG02iZ8Y3IaWkr6+PlpaWceFGa2srXV1d4wIbDw8PNBoNfn5+aDQa3feenp4GbX6qmDa9TSGRUj4wsUmos01F0QshxF+AtcBFKWX0mPtXAb8DzIEPpJS/nmgfUsom4HtCiK1j7qsBfiCEMAP+bKjxK9ODEILk5GRcXV3Jy8tj27ZtZGZmTul0DhsbG55++mmeeuopTp48ycGDB3nvvffw8vJi2bJlLFiwQDUIUxRFeQwhBBqNhpqaGoaGhqa0ok55ct7e3nzrW9/iueee47333qOurg4YrdJoamqivb0dOzu7aTn1QgjB/PnzmT9/PgkJCeN+19/fT3t7O83NzbS1tdHa2kpbWxv79+/n5s2buu0sLS3x8fEZF2poQw4nJyd1wUMBJhFgCCH+WUr5izE/mwGfAlM/UUtRZp6PgfeAT7R3CCHMgfeBFUAHcFoIsYvRMONX9z3+DSnlxYftWAixDvjJvf0rCn5+fjz//PPk5ORw6NAhYmNjSU1NndITJSsrKzIzM1myZAnFxcUcOXKETz/9lB07dpCRkUFmZqZan15RFOURNBoNVVVVdHZ24ufnZ+zhKN+AtbU1c+fOZenSpYSGhrJv3z66u7v5wx/+gKOjI2lpaSxcuBB3d3djD1UvbGxsCAkJISQkZNz92qqN1tZWXaihDThOnjzJ4OCgblvtCin3hxu+vr7Y2tpO9UtSjGgysa1GCPFTKeWvhBDWwJdAmYHHpSizgpTymBDC/767U4GGe5UVCCE2A+ullL9itFpjsvveBewSQuwFPtfPiJXpzs7OjnXr1lFYWMjZs2fp6elh2bJlU94MztzcnPT0dNLS0qirqyMnJ4d9+/Zx8OBBkpOTycjIIDAwUF1tURRFuY+npyeWlpa0tbWpAGMa+9u//Vvd98nJyQwNDXH27FlOnjzJoUOHOHDgAEFBQSxcuJDk5OQZ2TdibNVGfHz8uN8NDw/T3d09rmKjra2NiooKDh48OG5KiouLywMVGxqNBm9vb1XdOQNNJsD4LrBJCPFT4Clgv5Ty3w07LEWZ1byB9jE/dwALJtpYCOEMvAskjAkbM4HnAGtg3wSP+z7wfRi9mqPMHtqVQjw9PTl27Bjbtm0jKysLHx+fKR+LEIKwsDDCwsLo6ekhNzeXoqIiioqK8Pb2JiMjgwULFjBnzpwpH5uiKKZhTBNPYw/FJGiXtGxra0NKqYLeGcLCwoLExEQSExO5du0aRUVFnDx5kk8//ZQvvviCpKQkFixY8NDlWGci7b9zb29v0tPH9xns7++no6NjXMVGa2sreXl59PX1jduHl5eXrlJDo9Hg6+uLr68vHh4eqt/GNDVhE08hROKYHy2BPwEngA8BpJSPrMKYaG7/mN9nAjuB5nt3bZNS/vPjBmyKDZxUE8/ZSV9N0e5VYOzRvk+EEC8AT0sp37z382tAqpTy7570uR7y3Kqz+yzX19fH4cOHuXr1qu7EydgnRv39/Zw+fZqjR4/S3t6OtbU1ycnJpKWlERwcbPTxKcp0o5p4zjy1tbUcPXqU559/HmdnZ2MPRzEQKSXNzc2cPHmS06dP09/fj4ODAykpKSxYsABfX18VYN3n2rVrtLe309LSogs4tCHH2MapFhYWeHt76wINbcDh4+ODh4eH6i9jBPpo4vlv9/18FYi8d78Esh6z74+5b27/QxRIKSddEq8os0QH4DvmZx+gyxBPJKXcDexOTk5+yxD7V0yfo6MjGzZs4Pjx45SVldHZ2UlWVhZ2dg/0b54yNjY2LFmyhMWLF9Pa2sqxY8coKSnhxIkTuLi4sGDBAtLT03F1dTXaGBVFUYzJ13f0NKG1tVUFGDOYEILAwEACAwN58cUXqays5NSpU+Tl5ZGTk4OHhwepqamkpqaqY+I9Dg4OODg4EB09/vq5lJLe3l7a2tpob28fdyspKaG/v1+3rYWFBV5eXuMqNsZWbqhww7gm/NOXUj71JDueYG6/oiiPdxoIEUIEAJ3Ay8ArhngiVZarwOiBOjMzE29vb44fP87WrVtZsmQJxv53IYTQrRv/0ksvcebMGYqKiti3bx979+4lMDCQ5ORkEhMTmT9//rjHXrt2jT//+c+89dZb03ZJOkVRlInMnTsXFxcXurq6SExMfPwDlGnPysqKpKQkkpKSuHXrFqWlpRQXF7Nr1y527dpFYGAgqampJCcnG/UihKkSQuDq6oqrqytJSUnjfqcNN9rb22lra6Ojo0MXbpSWlnLnzh3dttqpLfcHGxqNRoUbU+RRU0j+D2CTnGADIUQQ4CmlPD7hzu8rjb/vd5nAV4xebe4C/k8p5bkJ9jN2rn5Sa2vrxK/ICNQUktlJHyW5Qoi/ApmAC9ADvCOl/FAIsRr4D0ZXHvmLlPLdJx3vo6iyXEXr+vXr5OXl0dPTQ3BwMIsXL8bKysrYwxrn6tWrnDp1itOnT9PR0QFAUFAQycnJJCQkMH/+fDZt2kRBQQFLlizh1VfVolnK7KamkMxMhYWFVFdX853vfEfN5Z/Frly5QnFxMcXFxXR2dmJmZkZERARJSUnEx8erFTqekJSSy5cvjws3xlZxPCzc0C4Fq/3q6+uLp6enCjceY7LHqkcFGP8LeAMovXe7BNgAwcBSoBf4iZRywonzjwkw7IERKeXNex/WfielDLl/u/uZ4sFLBRiz00w5IQTTfF8pxjMyMsKZM2coKytj3rx5ZGZm4unpaexhPVR3dzelpaWUlZXpwoyHsbS05L331IrCyuw0U45X6lg1XktLC4cOHWLdunV4eHgYeziKCejo6KC4uJjS0lJ6e3tVmGFg2nDj/lBDe7t9+7ZuW21DUR8fH3x8fHRBh/Y2E1eZ+bqeOMC4txNzRntdLAI8gTtADaMrkbRNYhD+TBBgPGTbFiBZStn7qO1M8eClAozZaSacEKomnsqjdHd3k5eXx40bN4iOjiY1NdWkrx50d3dTWFhIQUEBt27d0t3v5uZGdnY28fHxzJ0714gjVBTjmO7HK3Wserj+/n4++eQTUlNTH1iCUpndpJS0tbVRUlJCWVmZLswIDw/XhRlTvXz6bCOl5MqVKw+EGh0dHXR0dHDjxo1x27u4uEwYbjg6Os6KZq16CTD0MAh/Jq7A8AB6pJRSCJEKbAX8JpqyoqUCDMVUTPcTwrFM8X2lmIbBwUGKi4s5d+4c9vb2ZGZmmvyVPu30ETMzM4aHhzEzM2NkZETXUyMiIoKIiAgCAgLU+vDKrDBTjlfqWPWgL7/8knnz5pGdnW3soSgmShtmlJaWjqvM0IYZsbGx2NvbG3uYs861a9d0YYb21tnZSUdHBxcvXhy3ra2t7bhQY+z37u7uj724tH//ft5//316enpwd3fn7bffNsn/M/SxCsmTDkA3t18I0QG8w+hyrEgp/wvYCPytEGKI0cqOlx8XXiiKol+qiafyOJaWlixatIiAgACOHj3Krl27iImJISUlxWSrMa5fv05GRgZLliyhoKCAvr4+VqxYQU1NDTU1NRw4cIB9+/ZhZWVFSEgIYWFhBAcH4+fnZ7KvSVEU5WE8PT2pr69nZGRELTGtPJQQAj8/P/z8/NiwYYNu1Y3S0lI+/fRT3UoncXFxxMfH4+7ubuwhzwra1VKioqIe+N3AwABdXV0PBByNjY0UFBQwODio2/ZRU1O8vb3Jz8/n3Xff1a2y0t3dzbvvjrbWM8UQYzIMWoFhCKaYvqsKjNlpplzRAtN8XymmZ3BwkFOnTlFdXY2dnR2LFy/WLeU3ndy5c4e6ujqqq6upqamhp6cHGA1r/Pz8CA4OJjg4mMDAQDVfWJkRZsrxSh2rHtTQ0EBubi7PPfccLi4uxh6OMo1IKens7KS8vJzy8nLa29uB0VBMG2b4+fmpYMzEDA8Pc+nSpQeqNiaamqKtQr2fm5sbe/fuNampKSYxhcQQTPHgpQKM2WmmnBCCab6vFNPV1dXF8ePH6evrIygoiPT09GndW+LatWs0NjbS0NBAY2MjbW1tuoO9l5cXgYGB+Pn54e/vj7e3t+r2r0w7M+V4pY5VD7p16xabNm0iPT2dmJgYYw9HmcYuX75MRUUFFRUV1NXVMTIygoODA3FxccTFxREWFqamXU4D2qkp2mDjj3/844Tbaqem3F+14e3tjYeHx5T/festwBBCWAPPA/6MmXIipfznJxzjN2KKBy8VYMxOM+GEUDVGU76p4eFhKioqOHPmDObm5qSmphIREWFSSf43dffuXZqbm3WBRktLi64pqIWFBb6+vrpAw8/PDw8PD3WFSjFpM+F4BaZ5DmgKNm/ejJOTEytXrjT2UJQZ4tatW1RVVVFeXs65c+cYGBjAysqK8PBwYmJiiI6OxsnJydjDVCZh7dq1dHd3P3C/vb09q1ev1lVudHV1cffuXd3vzczMcHd31wUa998M0VhUnz0wdgLXGF1KdeBJB6YoiumQUu4GdicnJ79l7LEo04u5uTmJiYkEBgZy/Phxjh8/Tm1tLYsWLcLNzc3Yw3siVlZWhIWFERYWBoyW2fb29tLS0kJrayutra0UFhaSn58PgLW1NRqNRjfHWKPR4ObmpkINRVGmhIeHB62trUgpZ0SIrBifra0tCxYsYMGCBQwODlJbW0tlZSVVVVWcPXsWAG9vb6Kjo4mJiSEwMFBVJ5qot99+e1wPDAAbGxt+/OMfj+uBMTIyQm9vL52dnbqbtpLj+PHjXL58edx+586dOy7QiIqKmrIQdTIBho+UcpXBR6IoiqJMO46OjqxZs4aGhgZOnTrFjh07CAkJYcGCBdN6WslYQghcXV1xdXUlJSUFGD3Q9/T00NLSogs2jh49qmusZW1tjY+PDxqNBl9fXzQaDV5eXuoET1G+AdVw+tE8PT2pq6ujr6+P+fPnG3s4ygxjaWlJdHQ00dHRSCnp6emhsrKSyspKDh8+zMGDB5k7dy6RkZFER0cTGRmJg4ODsYet3KMNKR63ComZmRlubm64ubmRkJDwwH7u3LlDV1cXXV1d4wKOtrY2CgsLaWtrM6kA46QQIkZKWWnw0SiKoijTjhCCkJAQ/Pz8KC8v5+zZs7S0tJCQkEBMTMyM/NBuZmaGp6cnnp6epKePTh8cHh6mu7ubtrY23e3kyZMMDIwWL1pYWODt7Y1Go9HdvL291ZxiRXkMVS34aK6ursBoDwMVYCiGJITAw8MDDw8PVqxYwZ07d6ipqaGqqorKykq0U7x8fHyIiIggMjKS4OBgrKysjDzy2S07O/uJVxyZM2cOQUFBBAUFPfA7KeW4Cg9DmzDAEEJUAvLeNt8VQjQxOoVEAFJKGTs1Q1QUxVDUVS1Fn6ysrEhNTSUsLIzCwkKKi4upqakhJSWFoKCgGV/abG5uriul1IYaIyMjXLx4cVyoUVpaSkFBATAahHh5eemqNDQaDT4+PtjY2BjzpSiKMo04OjpiZmbGlStXjD0UZZaZM2cOiYmJJCYmMjIyQkdHB9XV1VRXV5OXl8fhw4exsLAgJCREF2h4e3urKZYzjBCCOXPmTN3zTdTEUwjh96gHSilbDTKixzDFBk6qiefsNFOaooFpvq+U6a+9vZ1Tp05x5coVXFxcWLBgAd7e3sYeltFJKbl8+fK4UKOtrU239JkQAnd393GVGr6+vjNmSo4y9WbK8Uodqyb21VdfMXfu3Ce+yvooN2/e5NKlS9y5c4fbt28jhMDKygpra2vmz5/P/PnzsbCYTHG3MhsMDAxQX1+vW7K8q6sLADs7O12YERERgaOjo5FHqpiKJ27iqQ0ohBCfSilfu2/nnwKvPfSBiqIoigL4+vri7e1NQ0MDJSUl7N27Fx8fH1JTU3FxcTH28IxGCIGLiwsuLi4kJiYCo6FGX1+fLsxob2+nvr6e4uJi3eNcXFzGhRoajQY7OztjvQxFUUyIk5OT7gOiPt25c4fKykpaW1u5evXqpMahXZJR9f2Z3aytrXW9MwD6+vqoqanRBRra45unpyfh4eGEhoYSGhrKvHnzjDlsZRqYTEwaNfYHIYQ5kGSY4SiKoigziZmZGaGhoQQGBlJdXc2ZM2fYtm0bwcHBpKSkqA/g9wghdFcw4+LidPdfv36d9vb2cZUaZWVlut/Pnz9/XKNQjUZjkKXNFEUxbU5OTtTX19Pf36+XKWiDg4NUVlZSUVHB0NAQXl5ehIWF4enpia2tre457t69S39/P1evXuXy5ct0d3dz7tw5KisrsbKyIiAggJCQEDw9PdX/S7Oco6Mj6enppKenMzIyQmdnJzU1NdTU1HDixAny8vIQQuDt7U1YWBihoaGEhIRga2tr7KErJuZRPTB+CvwMmCOEuM5o7wuAu8B/T8HYFEUxMNUDQ5kqFhYWxMbGEhYWRnl5OVVVVTQ1NREREUFCQoKaHjEBe3t7oqKiiIr6n2sJt27d0nX+1t7Onj2LdkqonZ2dLszQNtxSf76KMrM5OzsDcOXKFby8vJ5oXzdv3mT//v1cvXoVf39/UlNTJyzzt7GxwcbGBkdHRwICAgAYGhqiq6uLxsZGmpqaqK2txcHBgYiICMLDw1VDRwUzMzN8fX3x9fVl5cqVDA0N0dLSQm1tLbW1tRw7dowjR44ghMDX13dcoDGVvRYU0zRhDwzdBkL8Skr50ykaz2OZ4vxH1QNjdpopc4rBNN9Xysx28+ZNSktLqaurw8zMjMjISOLi4tQH7W+ov79/XKjR3t5OV1cXIyMjCCHw9PQkODiYoKAggoODcXZ2VldDZ5GZcrxSx6qJ3b59m88++4yFCxfqSva/ib6+Pvbu3cvdu3dZsWIFPj4+TzSuoaEhmpqaqKmpoaenB0tLS6KiooiJiVEfRJUJDQ4O0tzcTG1tLXV1dTQ1NTE0NIQQAj8/P12YERQUpCo0ZpAn7oExxs+EEM8BixldlaRASrnjSQeoKIqizF7z5s1j6dKlJCQkUFZWRlVVFdXV1URFRREfH69W4fiabGxsCA4OZmw1VX9/Py0tLTQ2NtLQ0EBxcTHHjh0DwMHBQRdmBAUFodFoVFd4RZnG5syZg42NzROtRNLX18fOnTsxMzPjmWee0UuvIgsLC11vg97eXs6cOUN5eTmVlZVEREQQGxureh4oD7C0tNT9u4HRqUraQKO2tpYjR45w6NAhALy8vAgJCdEdA52cnIw5dGUKTCbAeB8IBv567+cfCCFWSCnfNtywFEVRlNnA3t6ezMxM4uPjKSsr4+zZs9TU1BAVFUVsbKwKMp6AjY0N4eHhhIeHA+jmHDc2Nupu2n4ac+bMITQ0lPDwcMLCwvDy8lIVGooyjQghcHZ25vLly9/o8SMjI+Tn5wOwbt06HBwc9Di6US4uLqxYsYK+vj7Ky8s5d+4c1dXVhIeHk5iYqCrwlAlZWVkRFhZGWFgYMBpotLS0UF9fT0NDA6dOneLo0aPAaD8YbZgREhKCh4eHCuhnmMkEGEuBaHlvrokQ4v8DKg06KkVRFGVWcXR0JCsri4SEBEpLS3UntzExMcTExGBtbW3sIU57Y+ccZ2ZmAnD16lUaGho4f/48tbW1VFRUAKN9NLRhRkRExKxeNUYxPtWvaXKcnJyoqalhZGTka39gq6io4OLFi2RlZRkkvBjL0dGRzMxMkpKSKC8vp6amhvr6euLi4oiJicHS0tKgz69Mf1ZWVuMqNIaHh+ns7KShoYH6+nrOnz+vW+XE1tZWV3EYHByMn5+fWu53mpvM314toAFa7/3sC5w12IgURVGUWWv+/PksX76cK1euUFJSopteEhsbS3R0tGr+pmfz588nJSWFlJQUAC5fvsz58+d1gcbp06cBcHd3JyYmhujoaEJCQtTJnzKlpJS7gd3JyclvGXsspszJyYmhoSFu3LjxtUKIK1euUFpaSmBgIEFBQQYc4Xh2dnYsWbKEmJgYiouLKSkpobq6mqSkJMLCwtRVc2XSzM3Ndc2rs7KykFJy6dIlGhoadKHG2bOjH18tLS3x8/PT/XsPDAzE3t7eyK9A+TomcwbiDNQIIbSL0acAhUKIXQBSynWGGpyiKIoyOzk5ObFy5Up6e3spLS2lpKSEyspK4uLiiIqKUlfoDMTZ2ZlFixaxaNEipJT09PRQXV1NZWUl+fn55OTkYGNjQ0REBNHR0cTExBj8aq2iKJOjXYnk8uXLk35fSik5duwY1tbWLF682ChTxxwdHVm5ciXd3d2cOnWKgoICqqqqWLRo0ROvqKLMTkII3NzccHNzY+HChcDosuTaMKOpqWlcHw0XFxcCAwN1Nx8fH8zNzY35EpRHmEyA8QuDj0JRFKNQZbmKqXNxceHpp5/m0qVLlJSUUFxczNmzZ4mPjycyMlJVAhiQEAIPDw88PDzIyspiYGCA8+fPU1lZSVVVFWfOnAHAz8+PxMREkpKScHV1NfKoFWX2cnR0RAjBlStXCAwMnNRjLl26xMWLF1m0aJHRew55eHiwbt06WlpaKCoqYs+ePQQGBpKWlqYafSpPzN7ensTERBITE4HRPhptbW00NTXplvvVTjuxtLTE399/XKihqjRMx2PP/KSUR4UQfkCIlDJHCDEHsJBS3jD88BRFMSRVlqtMF66urmRnZ9PT00NJSQlFRUXZPg3wAAAgAElEQVRUVlaSkJBAWFiYulIyBaytrYmLiyMuLg4pJZ2dnVRWVlJeXs727dvZvn07Go1GF2a4ubkZe8iKMqtYWFjg4ODwtRp5VldXY2lpSUhIiAFHNnlCCAICAvD19aWiooLy8nLa2tpITEwkJiZG/V+v6I2VldW41buklFy5ckUXaDQ1NZGTk8Pw8DCgqjRMyWMDDCHEW8D3AScgCPAB/gtYZtihKYqiKMp47u7urFmzhq6uLk6fPs3x48epqKggMTGRkJAQNm3axJ07dx543Jw5c3jttdeMMOKZSQiBj48PPj4+ZGdn65ZHLC0tZceOHezYsQMfHx+SkpJISUlRlRmKMkUcHBy4fv36pLbt7++nsbGR0NBQk+svZGFhQVJSEqGhoRQWFlJcXExtbS3p6eloNBpjD0+ZgbQr+Tg7O+v6Qj2uSsPPz4+AgAD8/f3x9/fH2dlZreA1BSZTe/s2kAqcApBS1gsh1GUVRVEUxWi8vLxYt24dHR0dlJSUcPToUcrLyx8aXgAT3q/oh3Z5xBUrVnDlyhXKysooLS1l586d7Ny5k+DgYNLT00lKSmLOnDnGHq6izFi2trZ0d3dPatu6ujqGh4eJjIw08Ki+OTs7O1auXElHRwcnTpzgwIEDaDQaFi5cqEr6FYN7WJXG1atXaWxspKmpiebmZvLy8hgaGgJG/71qwwztTU1/0r/JBBgDUsq72jRJCGEBSIOOSlEURVEeQwiBr68vPj4+tLa2UlJSYuwhKYw2YF2+fLluNZmioiKKior49NNP2bx5MwkJCSxatIiwsDB1pUpR9MzW1paBgQGGhoYe2SNISklNTQ3u7u665p+mzMfHh40bN1JVVUVZWRlffvkl8fHxxMXFqV5IypQRQuDk5ISTk5OuSmNoaIjOzk5aWlp0t6qqKqQc/bjs6uqqCzO006NMreJpupnMO/6oEOJnwBwhxArgh8Buww5LURRFUSZHCIG/vz9+fn78+c9/NvZwlDGcnJxYvXo12dnZNDc3U1hYqGvG6u7uztKlS0lPT2fu3LnGHqqizAja99Lt27cfWaHQ1dXFtWvXSEhImKqhPTFzc3Pi4uIIDg6msLCQ0tJS6uvrWbhwoZpWohiNhYUFfn5++Pn5sXTpUmB0elZra6su0GhoaNAtS25mZoa3t/e4UMPT01MtG/w1TCbA+AnwPaAS+BtgH/CBIQelKIqiKF+XuppvuoQQusZnL774IqWlpeTn57Nlyxa2b99OamoqmZmZ6kOIojwhW1tbAG7duvXIAKO9vR0zM7NJr1ZiSmxtbVm+fDmdnZ26aSX+/v6kp6djZ2dn7OEpCjY2NoSFhREWFqa779q1a7S0tNDc3ExLSwulpaUUFBQAo02yNRqNLtDw9/fHyclJnddMYDKrkIwIIXYAO6SUl6ZgTIqifA1CiEDg54CDlHLjmPttgWPAO1LKPcYan6KYipKSEuLi4rC0tDT2UGY1S0tL0tLSSEtLo62tjaNHj1JcXMz/z969x1ddnnm//1xZScgBkkBCAiSEQEII4RgMEVEEQTwDFtG2Wrs79dGxHd229Xntaacz7TN9dse2e+rTg3asrR1rR0UHbUVUPIAIKEg4JOQEETmGYziDBEjCvf/ISgwhQUKy8ltZ+b5fr/Vyrd9pXTdIfivXuu7r/vDDDxk5ciQ33ngjOTk5+uAmchmaV2BczP79+0lKSurW0y9SU1O54447KCkpYf369bz88stMmDCBsWPHanUICTrx8fFNK3kBnDt3jurq6qaExvbt23n//fd59913gYZ+Go2VHY2PhIQEL4cQNNr8qWUNnxx+DDwEmH9TPfBb59xPuig+kZBmZn8CbgMOOOdGN9t+E/BrwAf80Tn3s7au4ZzbCtxnZgta7PpH4OXOj1okeEVHR7fasNPn87F+/Xo2bdrEpEmTyMzM1C/IQSA9PZ17772XO+64g+XLl7NkyRJ+9atfkZ6ezg033MCECRP0i4hIOzSvwGhLfX09Bw8eDOrmnZfK5/Mxfvz4pmklhYWFVFZWcvXVV5OWluZ1eCJtCgsLIyUlhZSUFCZNmgSc309j27Zt7Ny5k7KysqZ+GvHx8RckNS61me2xY8f4wx/+wP333098fHzAxtUVLpZ2/Q5wNTDRObcNmr7p/Q8z+65z7v90RYAiIe5Z4AngucYNZuYDngRmAlVAoZktpCGZ8ViL87/pnDvQ8qJmdj1QDkQFJmyR4HSxpVL379/Phx9+yNKlS9m0aROTJ0+mX79+XRidtCUmJoabbrqJGTNm8PHHH/POO+/wxz/+keTkZGbNmkV+fr7mB4tcgsjISHw+30UrMA4fPkx9fT3JyaGzqGDv3r2ZOXMmu3bt4qOPPuLNN99k2LBhTJo0SatASLfRWj+NM2fOUFVVxY4dO5oeJSUlTUmNvn37Np2TkZFBenp6q//PL1q0iC1btrBo0SLuueeeLh1XZ7tYAuPrwEzn3MHGDc65rWb2NeAdQAkMkQ5yzi03s4wWmwuALf7KCsxsPjDHOfcYDdUal+I6IBbIBWrM7E3n3LnOiVqke0pJSeH2229n06ZNFBYW8sorrzBmzBgmTJigjuBBIiIigmuuuYbJkydTVFTEG2+8wTPPPMPbb7/NnDlzGDNmjCpnRC7CzIiNjb1oBcb+/fuBhp+JoWbw4MHMmzeP4uJiNmzYwM6dO7niiisYPXq0qrmkW+rVqxeZmZlkZmY2bTt9+jQ7d+48L6lRVFTUtD8pKakpqfHaa69RX1/ftG/58uUsX76ciIgInnjiiS4dS2e5WAIjonnyopFzrtrMNIFYJHBSgV3NXlcBV7Z1sJklAj8F8szsB865x5xzP/Tv+wZwsLXkhZk9ADwAqHGe9BhhYWHk5uYydOhQ1qxZw8aNG9myZQtTpkxhyJAhXocnfmFhYUyYMIHx48ezdu1aFi5cyJNPPklmZia333472dnZXoco7dRavyYzGwk8AiQBS5xz/+FhiCHjixIYBw4cICYmpmm6Sajx+XxMmDCB4cOH89FHH/Hxxx+zefNmrrnmGgYNGuR1eCIdFhUVRXZ29nn3wlOnTl2Q1Fi3bt0F5/p8PoYPH96tqzAulsA4e5n7RKRjWvt60bV1sHPuEPBgG/uevch5TwNPA+Tn57d5fZFQFB0dzdSpU8nJyWH58uW8/fbbZGVlMXnyZKKiNPMqWISFhVFQUMAVV1zBhx9+yKJFi/jlL3/JhAkTuPPOOzUFqIsEql+Tc64CeNDMwgCtgdxJYmJiqK5uu+/+/v37SU5ODvlqpj59+nDjjTeyY8cOPvroIxYtWkRmZiaTJk0K2eSN9FwxMTHk5OSQk5PTtO3kyZPs3LmT1157je3btwMNPXA2bdrEj370IwYMGHBeP43Bgwd3i4rUiyUwxpnZ8Va2G5pXLxJIVcDgZq/TgD2BeCMzmwXMysrKCsTlRYJeSkoKc+fOZcOGDWzYsIHdu3dz9dVXd8ulBUOZz+fj2muvZdKkSbz77ru89dZblJSUcMstt3DDDTd065UUuolnCUC/Jv91ZgPf919fOkFsbCzbt2/HOXdBkqKmpoYTJ06ERAPPSzVkyBBSU1MpKiqiuLiYnTt3Mn78eMaMGaOfHRLSevfuTW5uLh988AFTp05lypQpLFmyhP379zNq1Ch27NhBeXk5q1evBhqmoA0aNIj09PSmnhppaWlBt3pbm/9qnXOaKCbijUJguJkNBXYDXwHuDsQbOedeB17Pz8+/PxDXF+kOfD4f+fn5DB06lA8++ID33nuPoUOHMmXKFFVjBJnIyEhuvfVWJk2axIIFC3jttddYu3Yt9957L0OHDvU6vJAVwH5NOOcWAgvN7A3ghc6JuGeLiYmhvr6es2fP0qtXr/P2hXL/i4sJDw8nPz+f7OxsVq9eTWFhIRUVFRQUFGhVKgl53/rWt5qef+Mb37hg/9GjR8+belJaWsqqVauAhkrI1NTUpiqN9PR0UlNTPU1qKO0o4iEzexGYBiSZWRXwY+fcM2b2EPA2Dd9k/ck5Vxag91cFhohfYmIit99+O8XFxaxbt44DBw4wffp0Bg4c6HVo0kJiYiJ///d/T3FxMS+88AI///nPufHGG5k1a5a+Ue06He7XZGbTgLlAL+DNNs5Tv6Z2ar6UassExoEDBzAzkpKSvAjNc3Fxcdxwww3s2bOHVatWsXTpUsrKyrjqqqtCalUWkfZISEggISGBcePGAeCc48iRI2zfvr0pqbF+/XpWrlwJNHzx05jUSE9PJyMjg0GDBnXZ/Vd3eREPOee+2sb2N2njw1wnv78qMESaCQsLIy8vj7S0NJYsWcKiRYuYMGECeXl5WsYzCI0bN47s7Gz++7//m8WLF1NWVsZ9992npFPX6HC/JufcMmDZxd5E/ZrarzGBcerUqQv6xOzfv5+kpKQen+gbNGgQX/rSl6isrKSwsJC//e1vZGVlUVBQoGVXpcczM/r160e/fv2YMGEC0JDUOHToUFNCY+fOnaxbt44VK1YAkJ2dzaOPPtol8fXsn14iPZwqMERa179/f+bOncvKlStZt24de/bs4brrrtMH2yAUHR3N17/+dcaOHctf/vIX/u3f/o17772XgoICr0MLderXFKRiYmIALliJxDlHdXW1VvHxCwsLIycnh2HDhlFUVERJSQnbtm1jzJgxjBs37oLqFZGerLFyKykpiSuuuAJo+Jly8OBBduzY0aVTSvR1kkgP5px73Tn3QHx8vNehiASdyMhIrrvuOqZNm0Z1dTWvvPIKVVVVXoclbRg/fjw/+tGPSE9P55lnnuHFF1+krq7O67Au25kzZ1i8eDHHjh3zOpS2NPVrMrNIGvo1LQzEG+le1T5tJTBqa2upq6sjLi7Oi7CCVmRkJAUFBdx1111kZGRQVFTE/PnzKSoq6tY/Q0QCzczo378/+fn5TdNPukLAEhhmNtjM3jezCjMrM7NHWjnGzOw3ZrbFzDaa2YRAxSMiFzKzWWb2dBB/QBbxlJmRnZ3N3LlziYmJ4a233mLjxo04p0r2YBQfH8/3vvc9Zs6cybJly/j1r399wS9xwa62tpalS5fyz//8z/z1r3+luLjY65Aa+zWtAkaYWZWZ3eecqwMa+zVVAC8Hql+TtE94eDi9evXi1KlT520/ffo0gCoL2tCnTx9mzJjB3LlzSU5OZs2aNcyfP5/y8nLOnTvndXgi4hfIKSR1wKPOufVm1gdYZ2bvOufKmx1zMzDc/7gS+A8u0gBKRDqXemCIXJqEhATmzJnDsmXLWL16NYcPH2bKlCn4fFqwK9j4fD7mzZvH4MGDee655/jFL37Bww8/3C2aFpaUlPDyyy9z4MABsrOzefDBB8nMzPQ6LM/7NWkKSfvFxsZekLw7c+YMgFZX+gJJSUncfPPN7N27lzVr1rBy5Uo2btxIfn6+ViwRCQIBq8Bwzu11zq33Pz9BQ3Y+tcVhc4DnXIPVQIKZqfOWiIgEncjISGbOnMmECROorKzkrbfeavqFQILPlVdeySOPPMLx48f5+c9/zp49AWnP0CkOHz7Mk08+yRNPPIGZ8fDDD/O9730vKJIXwUBTSNovJiZGFRgdNHDgQGbPns2NN95IeHg4S5cu5dVXX2X79u2qwhPxUJc08fSvHZ4HfNxiV2tLcKUCe7sirs5Uvvc4X/79Kq/DkE42Z3wqd18Zusu26VstkfYxM/Lz84mLi2P58uUsXLiQm2++Wc09g1R2djb/+I//yOOPP87jjz/Od7/7XVJTW36X4h3nHKtXr2b+/Pk455g7dy4zZszo8StESMfFxsZy+PDh87apAqP9zKxpqcgtW7awbt063nnnHRITE8nLy2Po0KGqyBDpYgFv4mlmvYFXgO8454633N3KKRekNM3sATNba2Zrq6urAxFmh8wZn0ruQDVECjXle4/zWtFur8MIKH2rJXJ5srOzufnmmzl58iQLFy4M5kaLPd6AAQN49NFH8fl8/PKXvwyaRqw1NTX8/ve/59lnnyUtLY1/+Zd/afqmV6SjYmJiqKmpOa93gyowLp+ZMXz4cO666y6mTZtGXV0d7733HgsWLGDLli3qkSHShQJ6lzSzCBqSF887515t5ZBLWoIr2NcAv/vK9JD+lr6nUkWNiFxMamoqs2bN4o033mDhwoXceuut9OvXz+uwpBUpKSk8+uijPP744/zmN7/h+9//vqd/V3v27OGpp56iurqauXPnMnPmTMLCtDBcW1Qt2H6xsbE456ipqSE2Nhb4vAJDCYzLFxYWRnZ2NllZWWzdupUNGzawdOlS1q1bR15eHllZWfq3LBJggVyFxIBngArn3ONtHLYQ+Lp/NZJJwDHnXLebPiIiIj1TUlISs2fPxsxYtGgRR44c8TokaUNycjIPP/wwZ86c4be//e0F/QG6SklJCT/72c+oqanhu9/9LjfeeKN+4fkCqhZsv9aWUj19+jSRkZH6/60ThIWFkZWVxbx587j++usJDw9n2bJlvPTSS5SVlWn5VZEACmQFxtXAvUCJmRX5t/0TkA7gnHuKhs7VtwBbgFPA3wUwHhFpQd9qiXRc3759mTVrFgsXLuSNN95g9uzZxMVpWmEwSk1N5Vvf+ha/+c1veOqpp3jkkUcuayWZv/zlL9TU1FywPTo6mnvvvbfN8z7++OOmKSPf/va36du3b7vfW+RSNFZdNE/UnTlzRtUXnczMGDZsGEOHDmXHjh0UFRXx4YcfsnbtWkaNGsWoUaOIjo72OkyRkBKwBIZzbiWt97hofowD/iFQMYjIxWkZVZHOER8fz6233srrr7/OokWLmD17thp70tCksra2ltraWs6ePdv037q6Os6dO9fqwzmHmTU9wsLCCAsLO+91eHg4ERER5/238eHz+S7aVC8nJ4d7772XZ599loULF/KlL32p3eNqLXlxse0Ay5Yt48UXX2TEiBF8+9vfViPFdlCyvf3aSmDo/7vAMDMyMjIYMmQI+/fvp7i4mPXr11NcXEx2djZjx45FFUQinUOdokRERDpBv379uOWWW1i0aBGLFy9m9uzZREZGeh1WwDjn+Oyzzzh+/HjT49SpU9TU1DT9t6ampsuXGzQzoqKiiIyMpFevXhc8IiMj6devH1dddRUrVqwgIyODMWPGBLR55qpVq3jxxRcZN24c999/PxEREQF7r1CkZHv7RUVFYWbnJTBOnz6tCowAMzMGDBjAgAEDOHr0KBs3bmTz5s1UVFSQkZHB2LFjSUlJ4b/+678uq4pLRJTAEBER6TT9+/dn5syZvPXWWyxdupQbbrghJOabnz59mkOHDnHw4EEOHjzIoUOHOHHiBPX19U3HmBkxMTFER0cTExNDUlIS0dHRTUmDyMhIIiIiiIyMJDw8vKmyonmFReNz51zTo3llhnOO+vp66urqqKuro7a2tul54+va2lrOnDnT9KipqeHo0aOcPXu2qYlho5ycHAoLCyksLCQiIqIp/uaPmJgYYmNjmx7t/QWwvLyc5557jpycHB544AGtMiJdIiwsjKioqAt6YKgKoOskJCRw7bXXkp+fT2lpKeXl5Wzfvp2kpKTLquISkQa6i4r0YCrLFel8aWlpXH311axcuZLVq1czefJkr0Nqt1OnTrF7926qqqrYu3cvJ0+ebNrXu3dvEhMTSU9PJy4urunRu3fvoE/WOOeaEhmnT59m165d/O1vfyM1NZURI0Y0VY4cOXKEPXv2XJDwAL4wAXHmzBkiIyMxM3bt2sXvf/97Bg4cyIMPPqjkhXSpmJgY9cAIAjExMRQUFJCXl0dlZSVlZWVehyTSrelO2gle+HgnrxXt9joM6WTle4+TOzC0G/GpLFckMHJzczl69CilpaUkJycT7ElC5xz79+9nx44dVFVVcejQIaBhucVBgwaRm5tLUlISSUlJ3XoOvZk1TSWJi4sjOTmZ6upq/vrXv3LttddekGyqr69vmhJz8uRJPvvsM06ePElJSUmb7/HnP/+Z8PBwevfuzd69e0lNTWXq1Kns37+fPn360KdPn5CeWiTBIzY2tqkC49y5c5w9e1YJDA9FREQwatQocnNz+cMf/uB1OCLdlhIYneC1ot094pfdniZ3YBxzxqd6HYaIdFOTJk3i4MGDrFixgv79+wdl6fbx48epqKjg008/5eTJk4SFhZGSksLEiRNJS0sjMTEx6KsqOmrGjBmsXLmSl19+mX/5l385b1USn89H79696d27N8nJyU3bt2zZ0mqpd2RkJBMmTODkyZNUVFTgnCMxMZHi4uLzjouKimpKZsTFxTU9j4+PJzY2NuT/zNtL1YKXJyYmhurqaoCmaqLunIAMFRdrMgywefNmMjMzVbEl0gb9y+gkuQPjeOnvr/I6DBERCRJhYWFMnz6dV155hSVLljBnzpzLWrKzsznn2LVrF6WlpVRVVWFmpKWlMXHiRIYMGdLjqgMiIiK48847+d3vfseyZcuYMWPGF57zRU32Nm7cyJo1a7jhhhuYO3cuZ86c4fjx45w4cYITJ040PT948CDbtm07r9FpWFhYUzIjLi6O4cOH079//w6PsztTteDliYmJoaamhnPnzimB0Y188MEHrF69mhEjRjBy5MigTH6LeEkJDBERkQDp3bs3U6dO5Z133qGwsJBJkyZ5Fotzjh07drB+/XoOHjxITEwMV1xxBTk5OU1LLvZUY8eOJTc3l9dff52CggL69Olz2df67LPPeO6550hLS2P27NlNq6JERUWdV8XR6Ny5c5w6dappJZdjx441Pd+zZw8pKSk9PoEhlycmJgZoaAx5+vRpAE0hCRLR0dFtrkIyffp0ysrKKCkpYePGjaSmpjJy5EgyMjJUnSWCEhgiIiIBlZGRwciRIykpKSErK4ukpKQuj+Hw4cN8+OGH7N27lz59+nDttdeSnZ2tD8N+Zsadd97JT37yE95//31mz5592ddavHgxJ0+e5JFHHrmk5VLDwsKapqkMGjTovH2NK6+IXI7GBMapU6eaKjCUwAgOX1TFlZqaymeffda0BOt7771HdHQ0OTk55OTkdCjJKtLdKYEhIiISYAUFBWzfvp0VK1YwZ86cLksc1NXVUVhYSGlpKZGRkVxzzTXk5OQocdGKQYMGMXr0aJYvX87NN998ScmHlo4ePcr7779PQUEBgwcP7nBMZvaF8+VF2tI8gdFYgaEpJN1HbGwsEyZMYPz48ezatYuKigo2bNjAhg0bSE9PZ+TIkQwePFg/z6XHUQJDpAdTYzSRrtGrVy8mT57MkiVLKC8vZ/To0QF/zyNHjrBkyRIOHz7MyJEjmThxon55+QLTp0/n17/+NWvXruWqq9rf1+rNN9+kvr6eWbNmBSA6kfZpnsCora0FlMDojsLCwhgyZAhDhgzhxIkTbNq0ic2bN/P2228TGxvLyJEjGTFiRI+fCig9h1J2Ij2Yc+5159wDahAlEnjDhg0jLS2NwsLCpqUNA2XLli28+uqrnDp1iptuuokpU6boF5dLMHLkSAYOHMjSpUvbPXWjurqaFStWcM0116hnRSczs1lm9vSxY8e8DqVbiY6OBj6vwDCzy6oskuDRp08fJk6cyN13383MmTNJSEhg7dq1vPDCC7zzzjvs2rVL084k5CmBISIi0gXMjGuuuYa6ujqKiooC9j4lJSUsXbqU/v37M2/ePNLT0wP2XqHGzJg+fTo7d+7k008/bde5b731Fj6fj1tvvTVA0fVcSrZfHp/PR1RUVFMPjF69emlKUogICwtj6NCh3HrrrXz5y19m7Nix7Nu3j7feeov58+ezYcMGTp065XWYIgGhBIaIiEgXiYuLY8SIEWzatKnTqzCcc6xZs4ZVq1aRkZHBLbfc0lRCLpfuyiuvJCYmhiVLllzyOWfOnGlaZSYhISGA0Ym0T0xMTFMFhqqwQlN8fDxXXnkl99xzDzNmzKBPnz4UFhby/PPP884771BVVaWqDAkp6oEhIiLShcaPH8/mzZspKiri6quv7rTrbty4kaKiInJycrjmmmvU2O0y9erVi0mTJrF8+XLOnj1LZGTkF55TXFzM2bNnKSgo6IIIRS5dYwIjIiJCK5CEOJ/PR2ZmJpmZmRw9epRNmzZRWVnJ9u3b6dOnDzk5OYwYMUKJben29OlGRESkC8XFxZGdnd2pVRjbtm3j448/ZtiwYUyZMkXJiw4aM2YMdXV1VFZWXtLxhYWF9O3bFzVElmDTmMBonEIiPUNCQgKTJk3innvuYfr06edVZbz77ruqypBuTRUYIiIiXSwvL4/KykqKi4uZPHlyh65VXV3N0qVLSU5OZtq0aZrj3gmysrKIiIigrKzsC1eMOXnyJKWlpcyYMUOJIwk6jQkM5xyJiYlehyNdzOfzkZWVRVZWVlNVxubNm9m2bRt9+vRh5MiRZGdnqypDuhUlMERERLpYXFwcWVlZbN68mYkTJ172ygB1dXUsXbqU6OhobrjhBsLDdVvvDJGRkQwfPpzy8vIvPHb9+vWcO3dO00ckKMXExOCc49SpU+qB0cM1VmVMnDiRbdu2UVFRwZo1a1i7di0ZGRnk5OSQmpqqJLgEPX1VINLNmdkwM3vGzBY02zbNzFaY2VNmNs3D8ESkDTk5OdTW1rJ9+/bLvkZRURHHjh3j2muv1TdonWzUqFHs27ePQ4cOXfS4wsJCUlJSGDx4cBdF1vNoGdXL1/zngqaQCHxelTFr1izuuusuRo0axe7du3nzzTd56aWXKCoqoqamxuswRdqkBIaIh8zsT2Z2wMxKW2y/ycw2m9kWM/v+xa7hnNvqnLuv5WbgJBAFVHVu1CLSGQYMGECfPn345JNPLuv8I0eOUFRURFZWFmlpaZ0cnYwaNQqAsrKyNo85cuQIn3zyCQUFBfrWMoC0jOrlUwJDLiYhIYGrrrqqqVdGbGwsa9as4fnnn+e9995j9+7d6pUhQUe1piLeehZ4AniucYOZ+YAngXFZd9MAACAASURBVJk0JB8KzWwh4AMea3H+N51zB1q57grn3AdmlgI8DtwTgNhFpAPMjOHDh7NhwwY+++wzYmNjL/lc5xwrVqwgIiKCq666KoBR9lwDBgygb9++lJeXc+2117Z6zKZNm3DOkZeX18XRiVya5j9XNIVE2hIeHn5er4yKigoqKyvZunUrcXFxTb0yoqOjvQ5VRAkMES8555abWUaLzQXAFufcVgAzmw/Mcc49Btx2idc95396BGj1KxczewB4ACA9Pb3dsYtIxw0fPpz169fzySefMH78+Es+b8eOHezbt49rr71WHygDxMwYNWoU69ato76+Hp/Pd8ExVVVVREREMHDgQA8iFPlizX8+qAJDLkVjVUbzXhkff/wxhYWFZGRkkJuby8CBA1V1Jp7RFBKR4JMK7Gr2usq/rVVmlmhmTwF5ZvYD/7a5ZvZ74C80VHhcwDn3tHMu3zmX379//86LXkQuWXx8PCkpKXzyySftKtMtKysjNjaW7OzsAEYnubm51NTUsG3btlb37969m4EDB2r1EQla4eHhTYkLVWBIe4SHhzN8+HBmz57NnXfeSW5uLrt372bRokW8/PLLFBcXq1eGeEIVGCLBp7WUdpu/2TjnDgEPttj2KvDqF76R2SxgVlZWVntjFJFOMnz4cFauXMnBgwe5lGTikSNH2L17NxMnTtQvzgGWk5MDwJYtW2jt5+SuXbswM44dO4b6M0iwiomJ4cyZM6rAkMvWt29fJk+eTEFBwQVVGUOHDmXkyJGqypAuo08+IsGnCmjezj4N2BOIN1JjNBHvDRs2DGiYFnIpysrK8Pl8Tb9cS+DExsYSGxvL4cOHL9h3/PhxTp48yYkTJ1i0aJEH0YlcmsZGnqrAkI5qXpUxb948cnNz2bVrV1NVxsaNGzl9+rTXYUqIUwWGSPApBIab2VBgN/AV4O5AvJEqMES8FxUVRb9+/ThwoLV+vOc7e/YslZWVZGZmqvdFF+nbt+8FCYyHHnqI2traptfLly9n+fLlRERE8MQTrc7aE/FMTEwMPp+P8HB97JfO069fv6aqjK1bt1JRUcHq1atZs2aNqjIkoPSTTMRDZvYiMA1IMrMq4MfOuWfM7CHgbRpWHvmTc67tdfw6wDn3OvB6fn7+/YG4vohcmpSUFLZs2YJz7qIf9jZv3kxdXV3TEp8SeH379uXIkSPnbfvpT3/Kb3/7W3btamhXFBkZyfjx45k3b54XIYpcVFZWVrtWORJpj/DwcLKzs8nOzubw4cNs2rSJyspKPv30U+Lj45tWMFEFkHQWJTBEPOSc+2ob298E3gz0+6sCQyQ4pKSkUFFRwZEjR+jXr1+bx33yySf079//knplSOfo168fW7duPW9bfHw8Z8+eBSAiIoLa2lqioqLUB0OC0uDBgxk8ePAXHyjSQW1VZRQWFpKZmcmoUaN0/5IOUwJDpAdTBYZIcEhJSQFg//79bSYwzp49y6FDh8jLy+vK0Hq8hIQEPvvsM86ePUtkZGTT9uPHj5OQkMBDDz3EihUrOHbsmIdRhjYl20W6l5ZVGeXl5VRWVlJZWUlycjK5ubkMGzZM05rksqiJp0gPZmazzOxpffAW8VZcXBxRUVHs37+/zWP279+Pc44BAwZ0YWTSmFBqPo2kvr6euro68vPzGTx4MHfffTff+ta3vAox5KnhtEj31a9fP6655hq+9rWvMXnyZM6cOcOyZct44YUXWLNmDSdOnPA6ROlmlPYS6cFUgSESHMyMlJSUiyYw9u3bh5mRnJzchZFJ3759ATh8+HBTpUx1dTW1tbWkpaV5GZqISLcRGRnJ6NGjGTVqFHv27KGsrIzi4mKKi4tJT09n1KhRpKamqumnfCElMERERIJASkoKO3bsoKamptUVRvbt20diYuJ50xgk8FqrwNi9ezcAqampnsQkItJdmRmpqamkpqZy8uRJysvL2bRpEzt27CA+Pp7c3Fyys7Pp1auX16FKkNIUEpEeTFNIRIJH49SQ1pZTra+v58CBA5o+4oGEhATg/ARGVVUVZsbAgQO9CktEpNvr3bs3BQUF3HPPPVx33XX06tWLVatW8fzzz7N8+XIOHTrkdYgShFSBIdKDaQqJSPBISkoiLCyM/fv3M2TIkPP2HTp0iPr6eiUwPBAREUGfPn04fPhw07bdu3czYMAAIiIiPIxMRCQ0+Hw+hg8fzvDhwzl48CBlZWV88sknbNq0iQEDBjB69GgyMjIIC9N376IEhoiISFAIDw8nMTGx1T4Y+/btA1ACwyN9+/bl6NGjTa+rqqrIyMjwLiARkRCVlJTE1KlTufLKK6msrKSsrIz33nuP3r17k5ubS05ODlFRUV6HKR5SAkNERCRIpKSkUFFRgXPuvEZm+/btIy4ujpiYGA+j67n69u1LdXU10LB86qFDh5g2bZq3QYmIhLCoqCjGjh3L6NGj2blzJ6WlpaxZs4Z169aRnZ3N6NGjm5osS8+iBIZID2Zms4BZWVlZXociIkBMTAz19fXU19cTHt5wi3bOsW/fPtLT0z2Orufq27cvlZWVAGzduhWAYcOGeRmSiEiPEBYWRkZGBhkZGRw6dIjS0lIqKyupqKggNTWVMWPGMHjwYK1e0oMEbCKRmQ02s/fNrMLMyszskVaOmWZmx8ysyP/4UaDiEZELOeded849EB8f73UoIgJNPRXOnj3btO3YsWOcPn1a00c81K9fP2pqajh9+jSffvopPp/vgj4lIiISWImJiUydOpV77rmHiRMncvToURYvXsxLL71EaWnpefdOCV2BrMCoAx51zq03sz7AOjN71zlX3uK4Fc652wIYh4iISLfQmMCora1t2nbw4EEAkpOTPYlJaCpTPnz4MFu3bmXw4MFq4Cki4pGoqCjy8vIYN24c27Zto7S0lI8++ojCwkJGjBjB6NGjiYuL8zpMCZCAJTCcc3uBvf7nJ8ysAkgFWiYwREREBIiMjATOT2CcOXMGgOjoaE9ikoYKDGhYDWbHjh1MmTLF44hERCQsLIzMzEwyMzM5cOAApaWllJeXU1paSnp6OmPGjGHQoEGaXhJiuqQHhpllAHnAx63svsrMioE9wP90zpW1cv4DwAOA5gCLiEjIam0KSeNzfePvncYKjJKSEmpra9X/QkQkyCQnJzN9+nQmTZpEeXk55eXlvPHGG/Tr148xY8aQlZWFz+fzOkzpBAFfTNfMegOvAN9xzh1vsXs9MMQ5Nw74LfC31q7hnHvaOZfvnMvv379/YAMW6UHMbJaZPX3s2DGvQxERWp9CUltbS1hYmD54eSghIQEzY/369QBkZmZ6HJGIiLQmJiaG/Px87r77bqZOnQrABx98wAsvvMD69es5ffq0xxFKRwU0gWFmETQkL553zr3acr9z7rhz7qT/+ZtAhJklBTImEfmcmniKBJfWppCcPXuWiIgIlcB6yOfzERcXx4kTJ4iPj9fSfZfBzIaZ2TNmtqDF9lgzW2dm6ocmIp0mPDycESNGcMcdd3DLLbeQmJjI2rVref7551m5ciX68q77CtgUEmv4pPUMUOGce7yNYwYA+51zzswKaEioHApUTCIiIsGsrSkkjYkN8U7fvn05duwYw4YN63HJJDP7E3AbcMA5N7rZ9puAXwM+4I/OuZ+1dQ3n3FbgvpYJDOAfgZc7P2oRETAz0tLSSEtL4/Dhw5SUlLBp0ybKy8sZMmQIY8eOZcCAAT3u53p3FsgeGFcD9wIlZlbk3/ZPQDqAc+4pYB7wLTOrA2qArzjnXABjEhERCVptTSFRAsN7/fr1Y/v27T21/8WzwBPAc40bzMwHPAnMBKqAQjNbSEMy47EW53/TOXeg5UXN7HoamrtHBSZsEZHP9evXj6lTpzJx4kTKy8spKytjx44d9O/fnzFjxjBs2DDCwgLeYUE6KJCrkKwELprKcs49QcMNUUREpMdrLYHROIVEvNU4baQnJjCcc8v9DdmbKwC2+CsrMLP5wBzn3GM0VGtciuuAWCAXqDGzN51z5zonahGR1jX2yRg/fjyVlZWUlJSwdOlS1qxZw+jRo8nJydEXB0GsS1YhERERkS9mZkRERJw3haS2tpaYmBgPoxJoaNy5ceNGhgwZ4nUowSIV2NXsdRVwZVsHm1ki8FMgz8x+4Jx7zDn3Q/++bwAHW0teaCU6EQmU8PBwcnNzGTlyJDt37mTjxo2sXr2adevWkZOTw5gxY+jdu7fXYUoLSmCI9GBmNguYlZWV5XUoIuIXERFxQQWGGu1674orruCKK67wOoxg0lqVbZvTgJ1zh4AH29j37EXOexp4GiA/P1/TjEWk05kZQ4YMYciQIVRXV1NSUkJpaSmlpaVkZmYybtw4EhMTvQ5T/DTJR6QH0yokIsGnZQJDPTAkSFUBg5u9TgP2BOKNtOS3iHSV/v37M336dL761a8yevRoduzYwSuvvMJbb73Fnj17ULtG7ymBISIiEkQiIyMvqMBQAkOCUCEw3MyGmlkk8BVgYSDeSMl2EelqvXv35qqrruLuu+8mPz+f6upqFi1axGuvvca2bduUyPCQppCIiIgEkeY9MOrr66mvr1cTT/GUmb0ITAOSzKwK+LFz7hkzewh4m4aVR/7knCvzMEwRkU7Xq1cvJkyYwNixY6msrKS4uJh3332X+Ph4xo0bx/Dhw/H5fF6H2aMogSEiIhJEIiIiOH36NPD5aiSqwBAvOee+2sb2N4E3A/3+6tckIl5rbPiZk5PDtm3bKC4uZvny5axdu5bRo0eTm5ure3UXUQJDREQkiDTvgdFYiaEPRdKTOedeB17Pz8+/3+tYRKRnCwsLIzMzk2HDhrF7926Ki4tZs2YNGzZsIDc3lzFjxmjlsABTAkNERCSIREZGNiUuGv+rKSQiIiLBw8xIS0sjLS2N6upqiouL2bhxIyUlJWRnZzN27FgSEhK8DjMkKYEhIiISRJpXYGgKiYiISHDr378/119/PcePH6e4uJjKyko2bdrE0KFDycvLIykpyesQQ4oSGCLdnJkNA34IxDvn5vm3hQH/G4gD1jrn/uxhiCLSDhEREZw7d476+npVYIigHhgi0j3ExcUxZcoUrrjiCkpLSykrK2Pbtm2kp6eTl5dHSkqK1yGGBC2jKuIhM/uTmR0ws9IW228ys81mtsXMvn+xazjntjrn7muxeQ6QCtQCVZ0btYgEUmO1RW1trSowRNAyqiLSvcTExFBQUNC0BOv+/ft57bXXWLRoEbt379YSrB2kCgwRbz0LPAE817jBzHzAk8BMGpIPhWa2kIZl6h5rcf43nXMHWrnuCGCVc+73ZrYAWBKA2EUkABqrLc6ePasmniIiIt1U4xKsY8aMoaKigo0bN/LGG2+QnJzMhAkTGDx4MGbmdZjdjhIYIh5yzi03s4wWmwuALc65rQBmNh+Y45x7DLjtEi9dBZz1P69v7QAzewB4ACA9Pb19gYtIwDQmMGprazWFREREpJuLiIhg7Nix5ObmUllZSVFREYsXLyYxMZG8vDwyMjIIC9PEiEulPymR4JMK7Gr2usq/rVVmlmhmTwF5ZvYD/+ZXgRvN7LfA8tbOc8497ZzLd87l9+/fv5NCF5GOajmFxMwID9f3DdJzmdksM3v62LFjXociInLZwsPDyc3N5Stf+QrTpk2jrq6O9957jwULFlBZWcm5c+e8DrFb0CcikeDTWi1Zm5PlnHOHgAdbbDsFtOyLceEbqTGaSNBpWYERGRmpElPp0ZxzrwOv5+fn3+91LCIiHRUWFkZ2djZZWVls27aNDRs2sGzZMtatW8e4ceMYMWIEPp/P6zCDliowRIJPFTC42es0YE8g3kiN0USCT/MeGLW1tZo+IiIiEoLCwsLIzMzkjjvu4MYbbyQ6OpqVK1fy4osvUlJSQl1dndchBiVVYIgEn0JguJkNBXYDXwHuDsQbqQJDJPg0n0LSWIEhIiIiocnMGDJkCOnp6ezZs4f169ezatUqioqKGD9+PCNHjtRU0mZUgSHiITN7EVgFjDCzKjO7zzlXBzwEvA1UAC8758oC8f6qwBAJPi2nkKgCQ3o69cAQkZ7AzEhNTWXWrFnMmjWLvn37smrVKlVktKBUjoiHnHNfbWP7m8CbgX5/VWCIBJ+WU0iioqI8jkjEW+qBISI9zcCBA7ntttvYu3cv69atU0VGM6rAEOnBVIEhEnzCwsLw+XyqwBAREenhGhMZqsj4XM9N3YiIKjBEglRkZKR6YIiIiAjweSJDPTKUwBDp0VSWKxKcIiIiqK2tpba2VgkMERERAWDQoEEMGjSoRycyQn+EIiIi3UxERARnzpyhrq5OU0hERETkPK0lMoqLi8nLyyMnJwefz+d1iAGjBIZID6YpJCLBKTIyks8++6zpuUhPpnuViEjrmicy1q5dy4cffsjGjRuZMGECw4cP5/nnn6empuaC86Kjo7n33ns9iLjj1MRTpAdTE0+R4BQREdGUwFAFhvR0uleJiFzcoEGDmDVrFjfffDO9evXigw8+YMGCBa0mL4A2t3cHqsAQEREJMo1TSEAVGCIiIvLFzIzBgweTlpbG9u3bWbt2rdchBYQqMERERIJM86oLJTBERETkUpkZQ4cO5Y477vA6lIBQAkOkBzOzWWb29LFjx7wORUSaaZ60UAJDRERE2issLDR/1Q/NUYnIJdG8YpHg1LwCQz0wREREpLO9/fbbHD582Osw2k0JDBERkSCjKSQiIiLSUdHR0a1uDw8PZ+/evSxYsIBly5Zx8uTJLo7s8qmJp4iISJBpnrRQBYb0dFpGVUTk8lxsqdTTp09TVFREWVkZn376KaNGjWL8+PFERUV1YYTtpwoMERGRIKMpJCKf03RHEZHOFxUVxaRJk/jyl79MZmYmGzduZP78+RQVFVFXV+d1eG1SAkOkB1MTT5Hg1Ji0iIiIwMw8jkZERERCVe/evZk2bRp33HEHKSkprFmzhpdeeonNmzdz7tw5r8O7gBIYIj2YvtUSCU6NU0jU/0JERES6QmJiIjfffDO33XYbMTExfPDBB7zyyivs3LkT55zX4TVRAkNERCTINFZgKIEhIiIiXWnQoEHcfvvtXH/99Zw7d47FixezaNEiqqurvQ4NCGACw8yizGyNmRWbWZmZ/Wsrx/Qys5fMbIuZfWxmGYGKR0REpLtQAkNERES8YmYMGzaMO++8k6uvvpojR47w17/+lffff9/zFUsCuQrJGWC6c+6kmUUAK83sLefc6mbH3Acccc5lmdlXgJ8DXw5gTCIiIkGvMXGhBp4iIiLilbCwMEaNGsXw4cMpKiqipKSErVu3MnbsWMaNG+fJFy0Bq8BwDRrTMxH+R8vJM3OAP/ufLwBmmLqViYhID6cKDBEREQkWkZGRFBQUcNdddzF06FA2bNjASy+9REVFRZc3+gxoDwwz85lZEXAAeNc593GLQ1KBXQDOuTrgGJAYyJhE5HNahUQkOPl8PsLCwlSBISIiIkGjT58+TJ8+ndtvv524uDhWrFjBK6+8wq5du7oshoAmMJxz9c658UAaUGBmo1sc0lq1xQUtTs3sATNba2Zrg6V5iEgo0CokIsErOTmZpKQkr8MQEREROU9ycjKzZ8/m+uuvp66ujtLS0i5770D2wGjinDtqZsuAm4Dmo6sCBgNVZhYOxAOHWzn/aeBpgPz8/OBZw0VERCRAZs+e7XUIIkHBzGYBs7KysrwORURE/BobfQ4ZMoQzZ8502fsGchWS/maW4H8eDVwPbGpx2ELg//I/nwcsdcG0yKyIiIiIeErVgiIiwcvn8xETE9Nl7xfICoyBwJ/NzEdDouRl59wiM/sJsNY5txB4BviLmW2hofLiKwGMR0RERERERES6qYAlMJxzG4G8Vrb/qNnz08CdgYpBREREREREREJDQJt4ioiIiIiIiIh0BiUwRERERERERCTodckqJCISOGY2DPghEO+cm+ffNgW4h4Z/47nOuckehigiIiIiItJhqsAQ8ZCZ/cnMDphZaYvtN5nZZjPbYmbfv9g1nHNbnXP3tdi2wjn3ILAI+HPnRy4iIiIiItK1VIEh4q1ngSeA5xo3+FfueRKYCVQBhWa2EPABj7U4/5vOuQMXuf7dwP/ozIBFRERERES8oASGiIecc8vNLKPF5gJgi3NuK4CZzQfmOOceA2671GubWTpwzDl3vI39DwAPAKSnp7c/eBERERERkS6kKSQiwScV2NXsdZV/W6vMLNHMngLyzOwHzXbdB/xnW+c55552zuU75/L79+/f0ZhFREREREQCShUYIsHHWtnm2jrYOXcIeLCV7T/+wjcymwXMysrKaleAIiIiIiIiXc2ca/P3oqBkZtXAjoscEg8ca+e+ltvb87rxeRJw8CJxXaqLxd+e41rbfynbWhtby+caa4MhzrkOly74p5Ascs6N9r++Cvhfzrkb/a9/AOCfQhIQl/DvqjN11v8/wSKUxhNKY4HQGk8ojQVCazzdYSydcr/yWhffq6B7/N1eqlAaC4TWeEJpLBBa4wmlsUDwj+fS7lXOuZB6AE+3d1/L7e153fgcWBvo+NtzXGv7L2Vba2Nr5bnG2okPIAMobfY6HNgKDAUigWJgVKDj6KpHV/yZajwaS6iNJ5TGEmrjCaWx6BG6f7ehNJZQG08ojSXUxhNKYwml8YRiD4zXL2Nfy+3teX2x97scl3q9Lzqutf2Xsq2tsXX2ONtzzVAYa6vM7EVgFTDCzKrM7D7nXB3wEPA2UAG87Jwr66qYREREREREglG3m0ISrMxsrXMu3+s4uoLGKh0Ran+moTSeUBoLhNZ4QmksEFrjCaWxyPlC6e82lMYCoTWeUBoLhNZ4QmksEDrjCcUKDK887XUAXUhjlY4ItT/TUBpPKI0FQms8oTQWCK3xhNJY5Hyh9HcbSmOB0BpPKI0FQms8oTQWCJHxqAJDRERERERERIKeKjBEREREREREJOgpgSEiIiIiIiIiQU8JDBHpEmb2JzM7YGalXsfSUWY22MzeN7MKMyszs0e8jqkjzCzKzNaYWbF/PP/qdUwdZWY+M9tgZou8jqWjzGy7mZWYWZGZrfU6no4wswQzW2Bmm/z/fq7yOqbLZWYj/H8njY/jZvYdr+OSjtP9KjiF4r0KQud+FUr3Kgid+1Uo3qvUAyNAzCwW+B1wFljmnHve45ACxsyGAT8E4p1z87yOJ5DM7HbgViAZeNI5947HIXUbZnYtcBJ4zjk32ut4OsLMBgIDnXPrzawPsA643TlX7nFol8XMDIh1zp00swhgJfCIc261x6FdNjP7HpAPxDnnbvM6no4ws+1AvnPuoNexdJSZ/RlY4Zz7o5lFAjHOuaNex9VRZuYDdgNXOud2eB2PdIzuV8EpFO9VEDr3q1C6V0Fo3q9C5V6lCox2aCsjb2Y3mdlmM9tiZt/3b54LLHDO3Q/M7vJgO6g9Y3XObXXO3edNpB3XzrH+zf93+g3gyx6E220555YDh72OozM45/Y659b7n58AKoBUb6O6fK7BSf/LCP+j22a3zSyNhkTjH72ORT5nZnHAtcAzAM65s939w2AzM4BPu/MHQvmc7lfBKdTuVaD7VbAK4ftVSNyrlMBon2eBm5pv8GeyngRuBnKBr5pZLpAG7PIfVt+FMXaWZ7n0sXZ3z9L+sf6zf7/0cGaWAeQBH3sbScf4S1iLgAPAu8657jyeXwH/D3DO60A6iQPeMbN1ZvaA18F0wDCgGvhPf7n0H/3ViqHgK8CLXgchcjGhcL8KsXsVhNb9KlTuVRC696uQuFcpgdEObWTkC4At/iqEs8B8YA5QRUMSA7rhn3M7x9qttWes1uDnwFuN32hIz2VmvYFXgO845457HU9HOOfqnXPjafi5VWBm3bJs2sxuAw4459Z5HUsnuto5N4GGhOo/+Mvbu6NwYALwH865POAz4PsXPyX4+UuLZwP/7XUsIm0JlftVqNyrICTvV6Fyr4IQvF+F0r2q2/1iHYRS+bzSAhoSF6nAq8AdZvYfwOteBBYArY7VzBLN7Ckgz8x+4E1ona6tv9eHgeuBeWb2oBeBSXDwz799BXjeOfeq1/F0Fn+J5DJaVCV1I1cDs/1zcecD083sv7wNqWOcc3v8/z0A/JWGBGt3VAVUNfvGdAENHxC7u5uB9c65/V4HItKaULxfhcC9CkLsfhVC9yoIzftVyNyrlMDoOGtlm3POfeac+zvn3LdCqIFnW2M95Jx70DmX6Zx7rMujCoy2xvob59wV/vE+1eVRSVDwNxJ7Bqhwzj3udTwdZWb9zSzB/zyahiTdJm+jujzOuR8459Kccxk0lEoudc59zeOwLpuZxfob7zU2h74B6JYrIzjn9gG7zGyEf9MMoNs1EmzFVwmBklwJTaF0vwqlexWE1v0qlO5VELL3q5C5V4V7HUAIqAIGN3udBuzxKJZA01jlspnZi8A0IMnMqoAfO+ee8Taqy3Y1cC9Q4p+LC/BPzrk3PYypIwYCf/b3fgkDXnbOdevl3EJICvDXht9BCAdecM4t9jakDnkYeN5fyroV+DuP4+kQM4sBZgJ/73Us0nl0vwpaulcFr1C7V0EI3a9C7V6lZVTbyd8AaVHjslpmFg5U0pCZ2w0UAnc758q8irGzaKyhOVYREREREZHuSFNI2sGfkV8FjDCzKjO7zzlXBzwEvE3D0lQvh8IvuRpraI5VRERERESku1IFhoiIiIiIiIgEPVVgiIiIiIiIiEjQUwJDRERERERERIKeEhgiIiIiIiIiEvSUwBAREc+ZWZqZvWZmn5jZp2b2a//SZY37XzSzjWb2XTPLMbMiM9tgZpntfJ9vmNmgi+z/lZld63++3cySLnM8t5nZv17OuSIiEpx0rxLxnhIYIiLiKWtYOP5V4G/OueFANtAb+Kl//wBgsnNurHPu/wC3A6855/Kcc5+28+2+AbT6odDM+gGTnHPLL28k53kDmO1fe11ERLo53atEVv+QQwAABxJJREFUgoMSGCIi4rXpwGnn3H8COOfqge8C3/R/qHoHSPZ/k/Vj4DvA/zCz980s1szeMLNiMys1sy8DmNkVZvaBma0zs7fNbKCZzQPygef914puEcc8YHHL4Mws2swWm9n9ZpZhZpvM7I/+93vezK43sw/938gV+MfggGXAbYH4AxMRkS6ne5VIEAj3OgAREenxRgHrmm9wzh03s51AFjAbWOScGw9N34KddM79u5ndAexxzt3q3xdvZhHAb4E5zrlq/wfFnzrnvmlmDwH/0zm3tpU4rgYWtNjWG5gPPOece87MMvwx3Qk8ABQCdwPX+OP8Jxq+dQNYC0wBXr7MPxcREQkeuleJBAElMERExGsGuHZsb64E+Hcz+zkNHxxXmNloYDTwbsPnR3zA3kuIYyBQ3WLba8AvnHPPN9u2zTlXAmBmZcAS55wzsxIgo9lxB2ijBFhERLod3atEgoCmkHQSM6v3l3mVmtnrZpZwmdf56CL7Eszs25d6fPP9rZ3bmfxlax+Ymc//OsPM3jSzzWZWaWY/uMzrRprZcjNTsk0kdJXRUC7bxMzigMHARecNO+cqgSto+HD4mJn9iIYPk2XOufH+xxjn3A2XEEcNENVi24fAzf5v0hqdafb8XLPX5zj/i4Eo/zVFRKT7071KJAgogdF5avw/fEYDh4F/uJyLOOcmX2R3AnBeEuILjm++/4JzO9k3gVedc/VmFga8AjzlnBsBjAHyzeyB9l7UOXcWWAJ8uVOjFZFgsgSIMbOvA/gTob8EnnXOnbrYidbQpf2Uc+6/gH8HJgCbgf5mdpX/mAgzG+U/5QTQp43LVdBQctvcj4BDwO/aPaqGBm+ll3GeiIgEH92rRIKAEhiBsQpIBTCzr5nZGn91xu+bVSi01czn5EX2/wzI9F/r/2s83sx+3ry6wsz+l5k92vx6Lc81s/9tZo80O+enZvZ/tzYYM/u2P4YdZvZwG2O+h4byNYAbge3OuYUAzrkzwEPA/2x2zUtqLuT3N//1RSQE+ZuIfQm408w+ASqB0zTM0f0iY4A1ZlYE/BD4f/2Jz3nAz82sGCgCGpO5zwJPWeuN0d4AprXyHt8BoszsF+0aGFznv6aIiHRzuleJBAdr+LcoHWVmJ51zvf0JivnAM8AO4BfAXOdcrZn9Dljtb65zB3CTc+5+//nxzrljza5zwX6gLw3z5kY3f18aGu/8yjk31b+t3H/uzmbXy2h+rv/1q865Cf6KiU+AAufcoRbjugP4GnAXkERD6dsA51xds2MigZ3OuQH+198DejvnftLiWvuAdOfcWf/7bwHyaCjJKwSKgftoaC70d8652/3n+YB9zrn+7ftbERFpHzNbCdzmnDvaweukAC8452Z0TmQiIiINdK+SnkwVGJ0n2p9VPQT0A94FZtAw363Qv28GMMx/fAlwvb96Ysr/3979hNqcxnEcf3+zIBYKMWZBWVrgJjsbC5KVxahZULZKFiYxZWVjoWRlIU2ZElMTs5hobkQ2knITFszCn5KFEVbKnfG1eB7Xmd8597r0u/XjvF+re8/T85zfqds99e37fL6Z+bpx3qfWJ2TmGGVs0/cRsRp4mZlPpnrYzHwEvIiIEWATMNYsXlR7gP2ZOZ6Zz4Bx+v9uFgG9/0D/o6QhT6h38uYC//a8/DAz72TmO0oR43Ktbv8vXKiOqXobEZO10klSW34ClrVwzrJ6liRJbfO7SkPLYMT2vMnMNbVT4k9KBkYCpzKzL8AyMx9ExFpgCyXMZ7S3Y2HQOvDrFO//O6UN7TtKB8h0nAR21j2/NBejjHdaVYOHiIilwD+15a1XM0zoKmV29f782OKzEbhVixUfTDdcCGA2pU1PkmZMZt5o6ZybbZwjSVKT31UaZnZgtKx2Suyh5D1cA36IiMUAEbEgIpbXnweF+UyYZH2qQJ+zwI+UIkZzNjST7D0PbAbWAX8N2LMSmB8RK+o1k8OUedXNz/wSmBURc+rvt4Ex4FD9LEuAo0zvjmCfiFgIPM/M8S/ZL0mSJEn6+tmBMQMyc6yG8awCDgKjtQAwTunMeEwJ8zkSEe/q67sax/StZ+aLGnJ5F7iYmft63vNevWLxtF71aD5T396aRXEFeFWvaTSNAKeBM8A8SmbGiUk+9iiwHrgUEQcoY6a21/N3A8uB4xGxtV5f+RwbgAufuUeSJEmS9A0xxHOI1aLKLWBbZv49YP0YcD0zf5vGWSPA3szcMQPPeQ74OTPvt322JEmSJOnr4BWSIRURKylTQC4PKl5UaygjnT6pBole+TAmti11wskfFi8kSZIkabjZgSFJkiRJkjrPDgxJkiRJktR5FjAkSZIkSVLnWcCQJEmSJEmdZwFDkiRJkiR1ngUMSZIkSZLUeRYwJEmSJElS51nAkCRJkiRJnWcBQ5IkSZIkdd57mAAyyQkgVNcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0a5aaf6940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Define offset; this has to be adjusted depending on the model!\n",
    "off = np.array([])\n",
    "for _, val in enumerate(sx):\n",
    "    off = np.r_[off, x-val]\n",
    "\n",
    "def plot_data(title, res, idata=None):\n",
    "    # Create figure\n",
    "    fig = plt.figure(figsize=(15, 5))\n",
    "    plt.suptitle(title, y=1.05, fontsize=20)\n",
    "\n",
    "    # Plot Resistivity Model\n",
    "    ax0 = plt.subplot(151)\n",
    "    plt.title('Resistivity model')\n",
    "    plt.ylabel('Depth (km)')\n",
    "    plt.xlabel(r'Resistivity $\\rho\\ (\\Omega\\,\\rm{m})$')\n",
    "    plt.xscale('log')\n",
    "    plt.ylim([1.5*depth[-1]/1000, 0])\n",
    "    \n",
    "    for key, value in res.items():\n",
    "        plt.plot(np.repeat(value[1][1:], 2),\n",
    "                 np.r_[np.repeat(value[0][1:], 2), 1.5*value[0][-1]][1:]/1000,\n",
    "                 label=key)\n",
    "    plt.legend()\n",
    "\n",
    "\n",
    "    # Define real subplot\n",
    "    ax1 = plt.subplot2grid((1, 5), (0, 1), colspan=2)\n",
    "    plt.title(r'|Real(E)| (V/Am$^2$)')\n",
    "    plt.xlabel('Offset (km)')\n",
    "\n",
    "    # Define imaginary subplot\n",
    "    ax2 = plt.subplot2grid((1, 5), (0, 3), colspan=2)\n",
    "    plt.title(r'|Imag(E)| (V/Am$^2$)')\n",
    "    plt.xlabel('Offset (km)')\n",
    "\n",
    "    # Loop over real, imaginary subplots\n",
    "    for t, ax in enumerate([ax1, ax2]):\n",
    "        # Get type\n",
    "        ri_type = (np.real, np.imag)[t]\n",
    "\n",
    "        # Activate required subplot\n",
    "        plt.sca(ax)\n",
    "\n",
    "        # Loop over frequencies\n",
    "        markers = ['o', '*', 's', 'd', 'p', '+', 'x']\n",
    "        for i, f in enumerate(freq):\n",
    "            # Plot True model\n",
    "            plt.semilogy(xf/1000, np.abs(ri_type(full[i, :])), str(0.2+i*0.2), label='$f=$'+str(f))\n",
    "\n",
    "            # Plot data\n",
    "            plt.semilogy(off/1000, np.abs(ri_type(data[i, :].ravel('F'))), ls='none', marker=markers[i],\n",
    "                         color=str(0.2+i*0.2))\n",
    "\n",
    "            # Plot inversion result\n",
    "            if np.any(idata):\n",
    "                plt.semilogy(off/1000, np.abs(ri_type(idata[i, :].ravel('F'))), 'C3', ls='none', marker=markers[i])\n",
    "\n",
    "        # Legend\n",
    "        plt.legend(loc='best')\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "    \n",
    "# Plot Model and input data\n",
    "plot_data('Input data', {'Model': (depth, res)})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. SimPEG wrapper class for an empymod problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# this is sketching out an idea for later... we can chat! \n",
    "\n",
    "# import properties\n",
    "\n",
    "# class EmpymodParameters(properties.HasProperties):\n",
    "    \n",
    "#     zsrc = properties.Array(\"z location of the source\", default=np.r_[10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from SimPEG.Problem import BaseProblem\n",
    "from SimPEG import Props\n",
    "\n",
    "class EmpymodProblem(BaseProblem):\n",
    "    \"\"\"\n",
    "    Problem class for a simulation conducted with empymod\n",
    "    \"\"\"\n",
    "    \n",
    "    rho, rho_map, rho_deriv = Props.Invertible(\n",
    "        \"Electrical Resistivity (Ohm m)\"\n",
    "    ) \n",
    "    \n",
    "    deleteTheseOnModelUpdate = ['_Japprox']\n",
    "    \n",
    "    # Stuff we don't touch at the moment and keep fixed\n",
    "    empymod_settings = {\n",
    "        'xdirect': False,\n",
    "        'ht': 'fht',\n",
    "        'htarg': None,\n",
    "        'opt': 'spline',\n",
    "        'loop': False,\n",
    "        'verb': 1\n",
    "    }\n",
    "        \n",
    "    def __init__(self, mesh, **kwargs):\n",
    "        assert mesh.dim == 1, \"only 1D modelling supported\"\n",
    "        super(EmpymodProblem, self).__init__(mesh, **kwargs)\n",
    "        \n",
    "        # set air at infinity\n",
    "        depth = self.mesh.gridN.copy()\n",
    "        depth[0] = -np.inf\n",
    "        \n",
    "        # Set the empymod_fixedparams\n",
    "        # Most of this was defined in point 1.2\n",
    "        self.fixed_params = {\n",
    "            'src': src,\n",
    "            'rec': rec,\n",
    "            'depth': depth,\n",
    "            'freqtime': freq,\n",
    "            'aniso': None,\n",
    "            'epermH': None,\n",
    "            'epermV': None,\n",
    "            'mpermH': None,\n",
    "            'mpermV': None,\n",
    "            'msrc': False,\n",
    "            'mrec': False,\n",
    "            'srcpts': 1,\n",
    "            'recpts': 1,\n",
    "            'strength': 0,\n",
    "        }\n",
    "\n",
    "    def _calc_fm(self, rho):\n",
    "        \"\"\"\n",
    "        compute data using empymod. The real and imaginary parts are separated so that \n",
    "        we are always working with real values\n",
    "        \"\"\"\n",
    "\n",
    "        # Calculate result\n",
    "        out = empymod.bipole(res=rho, **self.fixed_params, **self.empymod_settings)\n",
    "        # Ensure dimensionality, because empymod.dipole squeezes the output\n",
    "        if len(freq) == 1:\n",
    "            out = out[None, :]\n",
    "        out = np.ravel(out, order='F')\n",
    "        \n",
    "        return np.hstack([out.real, out.imag])\n",
    "    \n",
    "    def fields(self, m):\n",
    "        \"\"\"\n",
    "        Computes the fields\n",
    "        \"\"\"\n",
    "        \n",
    "        # set the model (this performs the mappings)\n",
    "        self.model = m\n",
    "        return self._calc_fm(self.rho)\n",
    "    \n",
    "    def Japprox(self, m, perturbation=0.1, min_perturbation=1e-3):\n",
    "        \"\"\"\n",
    "        Approximate sensitivity computed using a finite difference approach\n",
    "        \"\"\"\n",
    "        if getattr(self, '_Japprox', None) is None:\n",
    "            self.model = m \n",
    "            delta_m = min_perturbation # np.max([perturbation*m.mean(), min_perturbation])\n",
    "\n",
    "            J = []\n",
    "\n",
    "            for i, entry in enumerate(m):\n",
    "                mpos = m.copy()\n",
    "                mpos[i] = entry + delta_m\n",
    "                \n",
    "                mneg = m.copy()\n",
    "                mneg[i] = entry - delta_m\n",
    "                \n",
    "                pos = self._calc_fm(self.rho_map * mpos)\n",
    "                neg = self._calc_fm(self.rho_map * mneg)\n",
    "                J.append((pos - neg) / (2.*delta_m))\n",
    "\n",
    "            self._Japprox = np.vstack(J).T\n",
    "            \n",
    "        return self._Japprox\n",
    "    \n",
    "    def Jvec(self, m, v, f=None):\n",
    "        \"\"\"\n",
    "        Sensitivity times a vector\n",
    "        \"\"\"\n",
    "        self.model = m\n",
    "        return self.Japprox(m).dot(v)\n",
    "\n",
    "    def Jtvec(self, m, v, f=None):\n",
    "        \"\"\"\n",
    "        Adjoint Sensitivity times a vector\n",
    "        \"\"\"\n",
    "        self.model = m\n",
    "        return self.Japprox(m).T.dot(v)\n",
    "\n",
    "    \n",
    "class EmpymodSurvey(SimPEG.Survey.BaseSurvey):\n",
    "    \"\"\"\n",
    "    Survey class for a simulation conducted with empymod\n",
    "    \"\"\"\n",
    "    \n",
    "    @property \n",
    "    def nD(self):\n",
    "        # this will likely need to be generalized\n",
    "        nsrc = len(self.prob.fixed_params['src'][0])\n",
    "        nrec = len(self.prob.fixed_params['rec'][0])\n",
    "        nfreq = len(self.prob.fixed_params['freqtime'])\n",
    "        return nsrc * nrec * nfreq * 2\n",
    "        \n",
    "    def eval(self, f):\n",
    "        return f\n",
    "\n",
    "def re_field(inp):\n",
    "    inp = inp.reshape(2, -1)\n",
    "    inp = inp[0, :] + 1j*inp[1, :]\n",
    "    return inp.reshape((-1, len(x), len(sx)), order='F')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tests\n",
    "- **forward** make sure we can reproduce initial results\n",
    "- **Jvec** test sensitivity computation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# this is something we can work on cleaning up - simpeg needs a mesh for the forward problem,\n",
    "# working with infinity is tricky, so I make it finite here and swap it out for an inf in \n",
    "# the instantiation of the problem \n",
    "\n",
    "finite_air_z = 1000\n",
    "true_depth = depth[1:].copy()\n",
    "true_depth = np.hstack([np.r_[-finite_air_z], true_depth])\n",
    "meshfwd = SimPEG.Mesh.TensorMesh([np.diff(true_depth)], x0=[-finite_air_z])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0a5a5845f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the mesh\n",
    "meshfwd.plotGrid(nodes=True, centers=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we set up mappings - we will invert for log(resistivity) below the sea floor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for the inversion, we want to ignore air and probably the sea layers\n",
    "# and invert for log-resistivity \n",
    "\n",
    "exp_map_fwd = SimPEG.Maps.ExpMap(nP=len(res))\n",
    "inject_active_cells_fwd = SimPEG.Maps.InjectActiveCells(\n",
    "    meshfwd, \n",
    "    np.arange(2, len(res)),  # active indices (eg layers we want to invert for)\n",
    "    np.log(res[:2]),  # log-resistivity values of the inactive cells\n",
    "    nC=len(res)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the problem object\n",
    "probfwd = EmpymodProblem(meshfwd, rho_map=exp_map_fwd * inject_active_cells_fwd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# start with the true model and compute data \n",
    "\n",
    "m = np.log(res[2:])\n",
    "fields = probfwd.fields(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0a5a5e9e48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_data('SimPEG Forward Model', {'Model': (depth, res), 'SimPEG FM': (depth, probfwd.rho)}, (re_field(fields)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test the sensitivities\n",
    "\n",
    "Work with a 20 layer model, check that the sensitivities are second order"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from SimPEG import Tests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "finite_air_z = 1000\n",
    "seafloor = depth[2]\n",
    "nlayers = 20\n",
    "last_layer = 2500  # depth of last layer\n",
    "\n",
    "h = (last_layer-seafloor)/nlayers  # layer thicknesses\n",
    "\n",
    "# create a mesh\n",
    "inversion_depths = np.hstack([\n",
    "    np.r_[-finite_air_z, 0.], \n",
    "    np.arange(seafloor, last_layer+h, h)\n",
    "])\n",
    "\n",
    "simpeg_depths = np.diff(inversion_depths)\n",
    "mesh = SimPEG.Mesh.TensorMesh([simpeg_depths], x0=[-finite_air_z])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0a582868d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mesh.plotGrid(nodes=True, centers=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# mappings and problem \n",
    "\n",
    "exp_map = SimPEG.Maps.ExpMap(nP=nlayers+3)\n",
    "inject_active_cells = SimPEG.Maps.InjectActiveCells(\n",
    "    mesh, \n",
    "    2+np.arange(0, nlayers+1),  # active indices (eg layers we want to invert for)\n",
    "    np.log(res[:2]),  # log-resistivity values of the inactive cells\n",
    "    nC=nlayers+3\n",
    ")\n",
    "\n",
    "prob = EmpymodProblem(mesh, rho_map=exp_map * inject_active_cells)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# m0 = np.random.rand(nlayers+1)\n",
    "m0 = np.zeros(nlayers+1)\n",
    "def function_and_deriv(x):   \n",
    "    return prob.fields(x), lambda v: prob.Jvec(m0, v)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "==================== checkDerivative ====================\n",
      "iter    h         |ft-f0|   |ft-f0-h*J0*dx|  Order\n",
      "---------------------------------------------------------\n",
      " 0   1.00e-01    3.324e-12     2.494e-13      nan\n",
      " 1   1.00e-02    3.420e-13     2.527e-15      1.994\n",
      " 2   1.00e-03    3.431e-14     2.531e-17      1.999\n",
      " 3   1.00e-04    3.432e-15     2.533e-19      2.000\n",
      " 4   1.00e-05    3.432e-16     2.555e-21      1.996\n",
      " 5   1.00e-06    3.432e-17     2.712e-23      1.974\n",
      "========================= PASS! =========================\n",
      "You are awesome.\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEaCAYAAAA7YdFPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzt3Xd4FNX6wPHvS4BAaKH3ehFEEBADWEBAihS9YgEL+hNBvVxFEAuIBbkq2LBwxcZVLFQVEUE60qQXMVSlSe8gHVLf3x+zqaRskt1MNnk/z7NPsmdnZ98zs7vvnjkz54iqYowxxmRWPrcDMMYYE9gskRhjjMkSSyTGGGOyxBKJMcaYLLFEYowxJksskRhjjMkSSyTGJ0RkqIiM88N6VURq+3q9GXj9F0Tkcz+te5aIPOSPdedVIjJFRDq6HUdeY4nEeE1E7heRtSJyTkQOeb4IW7gdF4CI1PAknXOe2xER+VlE2mdlvao6XFUf8UF8lyVaVe2kql9ndd2ZiGWRiFzybKfjni/fismWGerZns2SlRcUkXdFZL/n+X+JyPuJHm8hIstF5LSInBSRZSLS1PNYTxGJSbSP4m6VPI/vFpGLInJWRE551tNHRDLyPfUmMCzzW8dkhiUS4xUReRr4ABgOlAeqAR8Dt7sZVwpCVbUo0AiYB/woIj0zsyIRye/LwHKYvp7tVAcIBRInAwEeBE4CyVtMg4EwoBlQDGgDrPc8rzjwM/AhUAqoDPwHiEj0/BWqWjTZ7WCix29T1WJAdZykMAj4wttKqepqoLiIhHn7HJN1lkhMukSkBPAq8ISqTlHV86oaparTVfW5RIsWFJFvPL8oNyf+MItIJRH5QUSOeX7F9kv0WJDnENJOz3PXiUjVFOJoISL7RKRNejGr6mFVHQkMBd6K+1WbThxDRWSyiIwTkTNAz8QtCRGZLSJ9k8UULiJ3ev4f6YnvjKcOLT3lHYEXgHs8v8DDPeWLROQREQn2/AJvkGi9ZT2/zst57t8qIr8n+qXeML1t4A1VPQn8ADRIVNwSqAT0B+4VkYKJHmsK/KiqB9WxW1W/8TxWx7POiaoao6oXVXWuqm7IRFynVXUacA/wUOJt44VFQJeMvqbJPEskxhvXA4WAH9NZ7p/AJJxfuNOAUQCeL/HpQDjOr9S2wFMicovneU8D9wGdgeJAL+BC4hV7lp0I3KWqCzMQ+xSgHFDXizjAaWFN9tRhfLJ1TfDEGRfTVTi/nGd4itYAjXF+jU8AvheRQqo6G6cl963nF3ijxCtV1QhPnPclKu4OLFbVoyLSBBgD/AsoDXwGTBOR4AxshxSJSBngLjytCo+HcLbTt577tyZ6bCXwtIg8LiJXe1ovcbYBMSLytYh0EpGSWY3P08LYj5PcUor/eRH5OVnxVpwWqckmlkiMN0oDx1U1Op3llqrqTFWNAcaS8GFuCpRV1VdVNVJVdwH/A+71PP4I8JKq/un5lRuuqicSrbcbMBro7PliyYi4wyalvIgDnEMvU1U1VlUvJlvXj0BjEanuud8DmOJJBKjqOFU9oarRqvouEAzU9TLOJEkKuN9TBvAo8JmqrvL80v8a53DRdV6uOyX/FZFTOEn1EE4yR0RCcLb3BFWNwkmqiQ9vvQG8hVP3tcAB8ZwwoKpngBaA4mzXYyIyTUTKJ3r+dZ5WVdxtpxexHsTZf5dR1TdV9dZkxWdxfgiYbGKJxHjjBFDGiz6Dw4n+vwAU8jynOlAp8RcIzqGeuC+YqkBaXyhPAd+p6sZMxF7Z8/ekF3EA7EttRap6Fqf1EZd47iVRq0VEnhGRrZ6O5lNACaCMl3EuAAqLSHNPompMQguwOvBMsrir4hx+SsJziDCuE/vTNF6vn6qGqmplVe2hqsc85XcA0cBMz/3xQCcRKevZBjGq+pGq3ojzZT0MGCMi9TyPb1XVnqpaBedwWSWcvrU4Kz2vG3f7hxfbpjLO/vNWMeBUBpY3WWSJxHhjBXAJ6JrJ5+8D/kr2BVJMVTsnejytL5RuQFcReSoTr30HcBT404s4wPk1nZaJwH0icj1QGFgI4OkPGYRzSKqkqoYCp4G4Qz9prldVY4HvcFol9wM/exIXnriHJYs7RFUnprCe4Yk6sfukU5eUPAQUBfaKyGHge6AASVtLca91UVU/Av4Grkrh8T+Ar0ja/5Ih4pzxVRlYmoGn1cNpaZlsYonEpEtVTwNDgI9EpKuIhIhIAc9x8Le9WMVq4IyIDBKRwp7O9QaeLwmAz4HXROQKcTQUkdKJnn8Qpz+jn4g87k3MIlLe0zH+CjDY80WdXhzemInTQngVp88j1lNeDOeX/DEgv4gMwenviXMEqCFpn8o6AadzuQcJh7XAOUzUx9NaEREpIiJdRKRYBuJOl4jE9RvditMiaoxzePItPIe3ROQpEWnt2X75PYe1igHrReRKT6usimfZqjgJaGUmYikuIrfi9LmNy2BrtBUwK6OvaTLPEonxiqq+h3Mc/SWcL8t9QF9gqhfPjQFuw/li+gs4jpM8SngWeQ/n1/hc4AzO6Z6Fk61jL86X3CARSeu6jlMich7YiNN5301Vx3gZR7oSdYy3I+mX/RycL69twB6cFlziw2Tfe/6eEJHfUln3KuA8zuGgWYnK1+L0k4zC+fW/A+jpbcwZ8CDwu+dMq8NxN+C/QEPPmVMXgXdxDmMeB57AOQFiF07fRHNglWcfrAQ2Ac8keo3r5fLrSBIn8ukichZn272I8954OLWAPYfyZiW63xQ4n4m+NJMFojaxlTEmlxCRH4AvVHVmugsbn7FEYowxJkvs0JYxxpgssURijDEmSyyRGGOMyRJLJMYYY7IkN49uGq9MmTJao0YNt8MwxpiAsm7duuOqWja95fJEIqlRowZr1651OwxjjAkoIrLHm+Xs0JYxxpgssURijDEmSyyRGGOMyRJLJMYYY7LEEokxxpgssURijDEmSyyRGGNMLhQbC6tXw/Ll/n+tPHEdiTHG5AUnT8LcuTBzJsyeDceOwc03wy+/+Pd1LZEYY0yAUoXwcCdxzJwJK1Y4LZHSpaFjR+jcGTp08H8clkiMMSaAnDkD8+c7iWPWLDh40Cm/9lp48UUneTRtCkFB2ReTJRJjjMnBVOGPPxJaHb/+ClFRULw43HKLkzg6doQKFdyL0RKJMcbkMBcuwMKFCclj926nvEEDGDDASR433AAFCrgaZjxLJMYYkwPs3JmQOBYuhIgICAmBdu3g+eehUyeoVs3tKFMWEIlERGoBLwIlVPVuT1lXoAtQDvhIVee6GKIxxmRIRAQsWZKQPLZtc8rr1IF//9tpddx0EwQHuxunN/yeSERkDHArcFRVGyQq7wiMBIKAz1X1zdTWoaq7gN4iMjlR2VRgqoiUBEYAlkiMMTna3r1OB/nMmc4puefPO4miTRvo29dpddSu7XaUGZcdLZKvgFHAN3EFIhIEfAS0B/YDa0RkGk5SeSPZ83up6tE01v+SZ13GGJOjREU5FwTGtTo2bXLKq1eHhx5yWh1t2jiHsAKZ3xOJqi4RkRrJipsBOzwtDURkEnC7qr6B03pJl4gI8CYwS1V/813ExhiTeYcPJ7Q65s51TtfNn985TDVihJM8rrwSRNyO1Hfc6iOpDOxLdH8/0Dy1hUWkNDAMuEZEBnsSzpNAO6CEiNRW1U+TPecx4DGAajm1h8oYE/BiYpyhSOJaHb95ftZWqgTduzuJo21b53Td3MqtRJJSLtbUFlbVE0CfZGX/Bf6bxnNGA6MBwsLCUl23McZk1IkTMGcOzJjh/D1xAvLlc07JHT7cSR4NG+auVkda3Eok+4Gqie5XAQ66FIsxxqQpNhZ+/z2h1bFypXOhYJkyTtLo0gXat4dSpdyO1B1uJZI1wBUiUhM4ANwL3O9SLMYYc5nTp2HevIShSA4fdsqbNoUhQ5wEEhbmtETyuuw4/Xci0BooIyL7gVdU9QsR6QvMwTlTa4yqbvZ3LMYYkxpV2Lw5odWxbBlER0NoaMJQJLfcAuXLux1pzpMdZ23dl0r5TGCmv1/fGGNSc+4cLFiQkDz2eU4BatQInnvOSR7XXeecdWVSZ5vHGJNnqML27QmJY/FiiIyEokWdPo4hQ5yLAitXdjvSwGKJxBiTq1265CSMGTOc5LFzp1Nerx48+aTT6mjRAgoWdDfOQGaJxBiT6+zf7ySNGTOcuTsuXIBChZzZAp9+2ml11KzpdpS5hyUSY0zAi4lxTsmdMcO5bdjglFevDj17OqfntmkDhQu7Gma2i4iOIDi//0d9tERijAlIJ08685LPmOH8PXnSmRWwRQt46y0neVx1Vd65KDCxsxFneX3J63y35Ts29NlAseBifn09SyTGmICg6rQ04vo64uYnL1sWbr3VSRwdOjin6+ZVsRrLuA3jGDR/EIfPHaZn455ExUb5/XUtkRhjcqzz553h1uOSx/79Tnnc/ORdujgXCNpFgXDq0ik6je/Eyv0raVa5GVPvmUrzKqkOYehTlkiMMTlK3EyBM2bAokXOBFBFizqtjf/8x+kor1jR7Shzjrh+kBLBJahdqjb/uvZf/F+j/yOfZF92tURijHFVZCQsXZrQUf7nn0553brw+ONOq6NlSzs9N7nImEhGrR7FO8vfYWXvlVQPrc7YO8a6EoslEmNMtoubs2PGDGfOjrNnnUTRurWTPDp3DsyZArPLnB1zeGrOU/xx/A861u6Ipj54erawRGKM8bvYWFi7NqHVsW6dU165Mtx7r9PqaNvWOYRlUhcTG8Pd39/N1D+mUrtUbabfN50uV3RBXD41zRKJMcYvTp92WhszZjitj6NHnU7x666DYcOc5JGX5uzIisiYSAoGFSQoXxC1QmvxVru36N+8f7ZcI+INSyTGGJ9Qha1bE86wWrrUGT23ZEno2NFJHB07QunSbkcaOFSVCRsnMGj+IKbcM4VmlZvx7i3vuh3WZSyRGGMy7eJF58yquENWu3c75Q0b2ui5WbXu4Dr6ze7H8n3LCasURoF8BdwOKVW2e40xGbJ3b0Kr45dfnGQSEuL0cTz/vJM8qlZNfz0mdU/PeZoPVn5A2SJl+eKfX9Czcc9sPZ03oyyRGGPSFB3tXEUe1+rYtMkpr1ULHnnESRytWzuDIprMi4qJIn++/IgIlYpVYsB1AxjSagglCpVwO7R0WSIxxlzm+PGEcazmzIG//3YOT7VsCSNGOP0ddetaR7mvzNs5j/6z+zO09VC61+/Oszc863ZIGWKJxBiDKvz+e0KrY9Uqp6x8ebj9didxtG8PJXL+j+OAsuvvXTwz9xmm/jGVWiVrEVooMAcKs0RiTB517pwzV0dcf8fBg05506bwyitO8mjSxMax8peRK0cyaP4g8ufLz/CbhzPg+gEUyh+YxwctkRiTh8RNMztjRsI0s8WLO+NYdenijGNVvrzbUeZeqkqMxpA/X34qFqtIt/rdeLPtm1QuHthz+4qqu5fWZ4ewsDBdu3at22EYk+0iI2HJkoRDVtu3O+X16jmJo0sXuPFGKJBzzyzNNdYfWk//2f3pckUXBrUY5HY4XhGRdaoalt5y1iIxJpc5dCih1TFvnnMIKzjYmSGwXz8nedg0s9nn+IXjvLTgJUavG02ZkDI82uRRt0PyOUskxgS42FhYsyah1fHbb055lSrQo4eTOG6+GYoUcTfOvGjylsk8Ov1RzkacpX/z/rzS+pWA7VBPiyUSYwLQqVNJx7E6dszpFL/+ehg+3EkeV19tp+e6JTo2mvz58lO1eFWaVW7G+7e8z1Vlr3I7LL+xRGJMAEg8jtWMGc44VjExUKpUwjhWt9xi41i5bfep3Twz9xnKhpTl01s/pXmV5sx5YI7bYfmdJRJjcqi0xrEaONBJHtddB0FBbkZpAC5EXeDNpW/yzvJ3yCf5eKnlS26HlK1yfCIRkVrAi0AJVb07UXkRYAnwiqr+7FZ8xvjS3r0JHeWJx7Fq1w4GD3aGI6lSxe0oTWLL9i7jvh/uY9+ZfdzX4D7ebv82VYrnrZ3k10QiImOAW4GjqtogUXlHYCQQBHyuqm+mtg5V3QX0FpHJyR4aBHzn+6iNyT6Jx7GaORM2bnTK48ax6tIFWrWycaxyopjYGILyBVGleBWqlqjK+DvH07J6S7fDcoW/WyRfAaOAb+IKRCQI+AhoD+wH1ojINJyk8kay5/dS1aPJVyoi7YAtgH28TMCxcawC24kLJ3h54cv8deovZt4/k+qh1VnWa5nbYbnKr4lEVZeISI1kxc2AHZ6WBiIyCbhdVd/Aab14ow1QBLgKuCgiM1U1NvECIvIY8BhAtWrVMl0HY7JKFcLDk45jFRsL5crZOFaBJDo2ms/WfsbLC1/mTMQZnmj6BFGxURQMKuh2aK5zo4+kMrAv0f39QPPUFhaR0sAw4BoRGayqb6jqi57HegLHkycRAFUdDYwG58p234VvTPrOnXP6OOIOWR044JSHhcHLLzvJ49prbRyrQLHtxDbu/u5uNh7dyM01b2Zkx5E0KNcg/SfmEW4kkpQa7Kl+0avqCaBPKo995aOYjMmynTsTWh2LFjnDkxQrlnQcqwoV3I7SZERcP0jFohUpWrAoP3T/gTuuvAOx445JuJFI9gOJ50+rAhx0IQ5jsiQy0rmeIy55/PmnU163LvTt6ySPFi2goB35CDgXoi7w9rK3+Xnbz6zovYJiwcVY3nu522HlWG4kkjXAFSJSEzgA3Avc70IcxmTY4cPOleQzZjhXlp896ySKNm3giSec03P/8Q+3ozSZpapM3jKZZ+c9y97Te7mn/j2cizxHycIl3Q4tR/P36b8TgdZAGRHZj3PNxxci0heYg3Om1hhV3ezPOIzJrNhYWLs2oa8jbhDpypXhvvucVkfbtjaOVW5w7Pwxuk/uzqLdi2hUvhFj7xjLTdVvcjusgODvs7buS6V8JjDTn69tTGadPp10HKujR51O8euug2HDnOTRsKGdnptbxGos+SQfpQqXIkiC+KTLJzza5FGC8tmQAd7K8Ve2G+NvqvDHH0nHsYqOhpIlE8ax6tjRxrHKbWJiYxi9bjQjV41kRe8VlCxcknkPzrOO9EywRGLypEuXko5j9ddfTvnVV8OzzyaMY5XfPiG50uLdi+k3ux8bjmygdY3WnI44TcnCJS2JZJJ9TEyesX9/QuL45Re4cAEKF3b6OAYOdDrK7drV3C0iOoKHpj7Et5u/pVqJanzf7XvuqneXJZAsskRicq2YGFi5MiF5bNjglNeoAQ8/7LQ6Wrd2konJ3eL6QYLzBwMwtNVQnrvxOUIKhLgcWe5gicTkKidPJoxjNXu2cz8oyLme4+23neRRr551lOcVqsrUP6Yy+JfBTL9vOleUvoKJd020FoiPWSIxAU3VGTE3rtWxYoVzym7ZsnDrrU7i6NABQnPf7KYmHZuPbqb/7P788tcvNCjXgLORZwEsifiBJRITcM6fhwULEq7t2OcZue3aa+Gll5zkERZm41jlVarKc/Oe44OVH1A8uDijOo3iX2H/In8++7rzF9uyJiD89VdCq2PhQoiIgKJFndbG0KHOOFYVK7odpXGTqiIiiAixGsujTR7ltZtfo0xIGbdDy/UskZgcKSoq6ThWf/zhlNepA48/7rQ6Wra0cayMY/m+5fSb1Y93O7xLqxqteLfDu3YIKxtZIjE5xpEjScexOnPGSRStWkGfPk7yqF3b7ShNTnLw7EEGzR/EuA3jqFysMheiLgDWD5LdLJEY18TGwm+/JbQ61q51Os8rVYLu3Z3E0a6dcwjLmORGrR7F8/OfJyo2ihdavMDgloMpWtDeLG6wRGKy1ZkzMG9ewjhWhw87p+I2bw6vvuokj8aN7fRck7q4vpCY2Bja1mrLex3e4x+lbMhlN4lq7p88MCwsTNfGDdtqspUqbNuW0Or49Ven/yM0FG65JWEcq7Jl3Y7U5HTbT2xnwJwBdLuqGw81fig+oRj/EZF1qhqW3nLWIjE+FxEBixcnJI+dO53y+vVhwAAnedxwg41jZbxzNuIsw34dxnsr3qNQ/kJ0vbIrYP0gOYl9lI1PHDjgXNMxYwbMn+9c61GoENx8Mzz9tJM8qld3O0oTaKb9OY0+P/fh0LlDPNz4YYa3HU6FojZfcU5jicRkSkwMrF6d0Or4/XenvFo1+L//cxJHmzYQYkMZmUxIfNiqaomq/HjPjzSv0tzlqExqrI/EeO3vv2HOnIRxrI4fd8axuuEGJ3F06eIcvrIjDiazjp4/you/vEiV4lV4pfUrqCqKkk9smAI3WB+JyTJV2Lw5odWxfLnTEilTxrmSPG4cq5I2nbXJoqiYKD5a8xFDFw3lfNR5Bt04CHD6QQT7ZZLTWSIxSVy44AxBEpc89u51yq+5BgYPdpJH06ZOS8QYX1i1fxUP//QwW49v5ZZ/3MIHHT/gyjJXuh2WyQBLJIY9exISx4IFzuyBRYpA+/bw8svOhE+VKrkdpcmtgvMHE6MxTLt3GrfWudXOxgpAlkjyoKgo5zBV3FlWmzc75bVrw7/+5bQ6broJgoPdjdPkTucjz/Pm0jc5cv4Io28bTeMKjdn6xFbrBwlglkjyiGPHEsaxmjMHTp+GAgWchNG7t5M86tRxO0qTm6kq323+jmfnPcv+M/vpcXUPYmJjCMoXZEkkwFkiyaVUYf36hENWq1c7ZRUqwF13JYxjVby425GavGDnyZ30mtaLJXuW0LhCYybeNZEW1Vq4HZbxEUskucjZs87FgHETPh065JyK27SpM2dHly5Op7lN+GSyW0iBEPae3sunXT7lkSaPEJTPztbITSyRBLjt2xNaHYsXO/0fxYsnjGPVqROUK+d2lCaviYmNYfS60czdNZcp3adQsVhFtj+53WYpzKVsrwaYyEhYsiQheWzf7pTXqwf9+zvJ48Ybnf4PY9ywZM8S+s3qR/iRcNrUaMPpiNOEFgq1JJKL5fg9KyK1gBeBEqp6t6csH/AaUBxYq6pfuxii3x06lHCG1bx5cO6cc0ZVmzbQr5+TPGrWdDtKk9eduHCCvrP6MmnTJKqVqMb33b7nrnp32em8eYBfE4mIjAFuBY6qaoNE5R2BkUAQ8LmqvpnaOlR1F9BbRCYnKr4dqAycBPb7I3Y3xcbCmjUJrY7ffnPKq1SBHj2cxHHzzc61HsbkFIULFGb9ofW80uoVBt44kJACNtBaXuHvFslXwCjgm7gCEQkCPgLa4ySBNSIyDSepvJHs+b1U9WgK660LrFDVzzwJ5hc/xJ6tTp1KOo7VsWNOp/j118Pw4U7yuPpqG8fK5Byqyk9//sRHaz5i+n3TCSkQwsZ/b6RAkB1XzWv8mkhUdYmI1EhW3AzY4WlpICKTgNtV9Q2c1os39gORnv9jUlpARB4DHgOoVq1axgLPBqqwZUtCq2PZMmccq1KlnImeunRxOsxLl3Y7UmMut/XYVvrP7s+8XfOoX7Y+B84c4B+l/mFJJI9yo4+kMrAv0f39QKrjQ4tIaWAYcI2IDPYknCnAhyLSEliS0vNUdTQwGpzRf30Ue5ZcvJh0HKs9e5zyhg1h4EAneVx3nY1jZXKuS9GXeOGXF/hw9YcULViU/3b8L/9u+m/rSM/j3Nj7KR2cSfWLXlVPAH2SlV0Aevs4Lr/YuzfpOFYXLzpzdLRr5wyC2LkzVK3qdpTGeKdgUEFWHVhFr8a9eP3m1ylbxOZINu4kkv1A4q/OKsBBF+Lwi+hoWLEiIXls2uSU16oFjzzitDpatXJmDzQmEKzYt4KXF77MhLsmUK5IORY+tJCCQQXdDsvkIG4kkjXAFSJSEzgA3Avc70IcPnP8uNNBHjeO1d9/O/ORt2wJI0Y4yaNuXesoN4Hl0NlDPP/L83wT/g0Vi1Zkx8kdlCtSzpKIuYy/T/+dCLQGyojIfuAVVf1CRPoCc3DO1Bqjqpv9GYevqUJ4eEKrY+VKp6xcObj9didxtG8PJUq4HakxGaeqjFg+gleXvEpkTCTP3/g8L7R8gWLBxdwOzeRQ/j5r675UymcCM/352r527hz88kvCOFYHDjjlYWEwZIiTPK691saxMoFPRFh3aB2ta7TmvQ7vcUXpK9wOyeRwdqpFGo4fh/HjE8axioyEYsWc6WXjxrGqUMHtKI3Juh0nd/Ds3GcZdvMw6perz9ddvyY4v01IY7yTbiLxXED4pqo+lw3x5CgnTsBTTzn9G337OsmjRQsoaIeITS5xLvIcw5YM472V7xEcFMzW41upX66+JRGTIekmElWNEZFrRURUNUdcj5Fd6tSBXbtsHCuTO03aNIln5j7DwbMH6dm4J2+0fYMKRa2JbTLO20Nb64GfROR74HxcoapO8UtUOYSIJRGTe4UfDqdyscpM6T6F5lVSvSbYmHSJN40MEfkyhWJV1V6+D8n3wsLCdO3atW6HYYyrjp0/xosLXuTOenfSsXZHIqIjKBBUwKa5NakSkXWqGpbecl61SFT14ayHZIxxQ1RMFJ+s/YQhC4dwPuo8V5a5ko61O1o/iPEZrxKJiFQBPgRuxBnOZCnQX1Vz3RDuxuQmi3cv5omZT7D52GY6/KMDH9zyAfXK1nM7LJPLeNtH8iUwAejmuf+Ap6y9P4IyxvjGthPbuBB1gan3TOWfdf9pk0wZv/C2j+R3VW2cXllOZX0kJq+4EHWBt5e9TdXiVendpDcxsTFExUZRKL8N7mYyzts+Em972Y6LyAMiEuS5PQCcyFqIxhhfUVUmb5lMvY/q8Z/F/2HdoXUABOULsiRi/M7bRNIL6A4cBg4Bd3vKjDEu23JsC22/aUu377tRslBJFvdczMddPnY7LJOHeHtl+12q+s9siMcYk0EHzhwg/Eg4n3T5hEebPEpQPpsZzWQvb69svx14PxviMcakIyY2hs9/+5zjF47z4k0v0v4f7dndf7eNzmtc4+2hrWUiMkpEWopIk7ibXyMzxlxm6d6lhP0vjD4z+rB4z2JiNRbAkohxlben/97g+ftqojIFbvZtOMaYlBw8e5Dn5j3HhI0TqFq8Kt/e/S3drupmp/OaHMGbPpJ8wCeq+l02xGOMScHpS6eZ/ud0htw0hEEtBhFSIMTtkIyJ500fSaxnRkNLJMZkE1Vl+rbpLNq9iPdueY96Zeux/+n9FA8u7nb8+b71AAAgAElEQVRoxlzG2z6SeSLyrIhUFZFScTe/RmZMHvXH8T/oNL4Tt0+6ndk7ZnMm4gyAJRGTY3nbRxJ3zcgTicoUqOXbcIzJu85EnOHVxa8yctVIihQowge3fMDjTR+nQFABt0MzJk3ejv5rs3IY42eRMZF89ftX9GzUk2Fth1GuSDm3QzLGK2ke2hKRgYn+75bsseH+CsqYvGL1gdU8Nv0xYjWWMiFl2P7kdv73z/9ZEjEBJb0+knsT/T842WMdfRyLMXnGkXNH6PVTL5p/3pzp26az6+9dAJQsXNLlyIzJuPQObUkq/6d03xiTjqiYKD5c/SH/WfwfLkZdZOANA3npppfsgkIT0NJLJJrK/yndN8akI1Zj+WTtJ7Ss1pL3bnmPOqXruB2SMVmWXiJpJCJncFofhT3/47lvY1Mb44WdJ3cy/Nfh/LfTfylSsAgre6+kdEhpt8MyxmfS7CNR1SBVLa6qxVQ1v+f/uPvZck6iiNQSkS9EZHKismoiMk1ExojI89kRhzEZdS7yHC/+8iJXfXwV327+Nn6OEEsiJrfx9oLETPF80R8VkU3JyjuKyJ8isiO9RKCqu1S1d7LiOsAMVe0FXOXjsI3JElVl4saJXDnqSoYvHc499e9h25PbuKn6TW6HZoxfeHtBYmZ9BYwCvokr8Mxv8hHOfO/7gTUiMg0IAt5I9vxeqno0hfWuB14UkXuAsX6I25gsGf3baMoXLc933b7jhqo3pP8EYwKYXxOJqi4RkRrJipsBO1R1F4CITAJuV9U3gFu9XPXDwCue9U8GvvRRyMZkyvELx/nPov8wqMUgqhSvwvfdvqdkoZI2yZTJE/x6aCsVlYF9ie7v95SlSERKi8inwDUiEncty2ygn6d8dyrPe0xE1orI2mPHjvkmcmOSiY6N5qPVH1Hnwzp8svYTFu1eBECZkDKWREye4e9DWylJ6fqTVE8lVtUTQJ9kZZtw5o1PlaqOBkYDhIWF2anKxucW7V5Ev1n92Hh0I21rtmVkx5HUL1ff7bCMyXZuJJL9QNVE96sAB12Iw5gs+Sb8G85GnuWH7j9wx5V32CRTJs9yI5GsAa4QkZrAAZxhWO53IQ5jMuRi1EXeWf4Ot9a5lSYVm/DeLe8RHBRM4QKF3Q7NGFf5NZGIyESgNVBGRPbjdJB/4Zkoaw7OmVpjVHWzP+MwJitUlSlbp/DM3GfYc3oPAE0qNiG0UKjLkRmTM/j7rK37UimfCcz052sb4wubj26m3+x+LPhrAVeXu5qFDy2kdY3WbodlTI7ixqEtYwLGD1t/YP2h9YzqNIp/hf2L/PnsI2NMcqKa+09oCgsL07Vr17odhgkAMbExjFk/hsrFK9P5is5cjLrI+ajzlAkp43ZoxmQ7EVmnqmHpLefGdSTG5EjL9i6j2efNeOznx5iwcQIAhQsUtiRiTDoskZg87+DZgzz444O0+LIFR84dYcKdExh7h428Y4y37ICvyfMW/LWA7zZ/xwstXmBwy8EULVjU7ZCMCSiWSEyeo6r8vO1nTl48yUONH6LH1T24qfpNVCtRze3QjAlIdmjL5Cl/Hv+TzhM6889J/+TjtR+jqoiIJRFjssASickTzkSc4bm5z9HgkwYs37ec9zq8x9KHl9qwJsb4gB3aMnnCxiMbeXfFu/S6phfDbh5G+aLl3Q7JmFzDEonJtVYfWM2KfSvof11/bqx2Izv67aBWyVpuh2VMrmOHtkyuc/jcYXr91IvmnzdnxIoRnI88D2BJxBg/sURico3ImEhGLB9BnQ/rMG7DOAbdOIgtj2+hSMEibodmTK5mh7ZMrnHgzAFeWvAS7Wq14/1b3ueK0le4HZIxeYIlEhPQdpzcwaRNk3jpppeoWbImmx7fRO1Std0Oy5g8xQ5tmYB0LvIcg+cPpv7H9Xlr2VvsPrUbwJKIMS6wRGICiqoybsM46o6qy5vL3uS+Bvexre82aoTWcDs0Y/IsO7RlAsrZyLM8M/cZqpeozpTuU2hepbnbIRmT51mLxOR4R88fZcjCIUTHRlM8uDhLH17KykdWWhIxJoewRGJyrKiYKD5Y+QF1PqzDG0vfYOX+lQBcUfoK8om9dY3JKezTaHKkeTvn0ejTRgyYM4DmVZqzoc8GWlRr4XZYxnhl3759tG3blnr16lG/fn0GDhxIbp6N1hKJyXFiYmMYMGcAETER/HTvT8zuMZt6Zeu5HZYxXsufPz9vvfUWW7duZf369axatYopU6a4HZbfWCIJAEuWLKFZs2YUKVIEEeHYsWM0atSIr7/+2rWYJk2aRHBwMFFRUWkup6rUrFkTEWHHjh2pLnc+8jyvL3md05dOE5QviGn3TWPz45v5Z91/IiI88cQT9O7d29fVuIyvtnV2xOvtPvCGqvqlnt7u/5wotW3izXavWLEiYWHOVOcFCxakYcOG7Nu3L/7x7Hh/bNmyhbZt2xISEkKlSpUYMmQIMTEx/nkxVc31t2uvvVYDVUREhJYpU0Z79OihS5Ys0VWrVumkSZO0atWqGhkZ6Vpczz//vDZp0iTd5ZYtW6aAioi+9tprlz0eGxurEzdO1CrvVVGGomPDx6a4nr/++kuDg4N1+/btWY49Nb7c1tkRr7f7wBv+qmd6+98tUVFRevr06TSXSW2bZHS7Hz9+XKtUqaJbtmyJL/P3++PkyZNasWJFbdu2rc6dO1c/+eQTDQkJ0RdffDFD6wHWqhffsa5/yWfHLZATyZw5cxTQw4cPx5fdcMMN+sILL/jsNRYuXKjObwrvdezYUXv37p3ucn379tVKlSppt27d9Kqrrkry2PpD67XlmJbKUPSaT6/RX/f8mua62rZtq08//XSG4swIX29rf8fr7T7whr/qmdb+d8vYsWO1b9++unv3bn3//fe1TZs2Ghoaqj169NDz58/HL5faNsnIdr906ZK2bt1aR4wYcdlj/nx/DB8+XENDQ5Mky7feeksLFy6cbgJNzBJJLkgkN954owJJbosXL1ZAf/vttyTLLlq0SAGdOXNmfNmuXbu0bNmy+uSTT6b5OplJJBUrVtQRI0bo4MGDtVKlSlq8eHF95JFHNCYmJn6Z6OhoLV++vPbv31+nTJmigIaHh8c/3mV8Fy3zdhl95rNnvIr9448/1vLlyyd5DV/xx7b2Z7yq3u0DVdUTJ05or169NDQ0VEuXLq3Dhw/X119/XevWrauqqtu3b/dLPdPb/xl5DV8ZOnSo3nHHHRobG6u9evXS1atXq6rqjz/+qIC+/PLLqpr6NlH1frtHR0frXXfdpQMGDEgxFn++P1q2bKn33HNPkrI9e/YooNOmTfN6PZZIckEiWb9+vbZv315btGihK1as0JUrV+qnn36qRYoUSfHN16ZNG73++utVVfXUqVN61VVXaZcuXTQ6OjrN18loIjl69KgCWr16dX3qqad07ty5+uqrryqgU6ZMiV9u7ty5Cujy5cv10qVLWrx4cW3/f+1158mdqqq67/Q+PXnhpNex//777wro77//7nWs3vLHtvZnvN7ug0uXLmmTJk20Tp06OnbsWJ0+fbo2bNhQa9Sood27d1dV9Vs9U9r/zz//fKZewxdmzZqlwcHBevjwYd2yZYvWr19fN27cGN8Kueqqq/Saa65R1dS3ibfbXVW1d+/e2rNnT42NjU0xnvTeH7GxsRoVFZXuLSVly5bVV1555bLykJAQffvtt9PcTonlmkQCdAX+B/wEdPCUFQG+9pT3SG8dgZpIVFUbNWqkAwcOjL//6KOPalhYWIrLLlmyRAGdM2eOdujQQRs1aqRnz569bLnkb9D58+cr4NUbVDXhC+Ldd99NUl6xYkV966234u8//PDDWq1aNY2NjdVfdv2ioc1ClVD0lQWvZCr2qKgoDQoK0tGjR6caW1b4elv7M15v98HQoUO1WLFieuTIkctiHzZsmKr6r56J97+q6oMPPqg1atS47IvV2/dtVjVu3FjvvfdeVVX9448/4luecX0XzZs313r16qlq6tvE2+2+dOlSBbRBgwbaqFEjbdSokY4cOTLJc9J7f3z55ZeXtZJTuqUkf/78+v77719WXrlyZR08eHBqm+gy3iYSvw6RIiJjgFuBo6raIFF5R2AkEAR8rqpvprYOVZ0KTBWRksAIYC5wJzBZVaeLyLfAeD9WwzXR0dFs3bqV5557Lr7s8OHDlClTJsXlW7ZsSbt27bjjjjsIDQ1l1apVFC1a9LLlvv76ax5++OHLygsUKJDkvvM+ulx4eDihoaE8+eSTSZY9depUfGwRERH8+OOPdH+gO90nd2fylsmUa1wOVkOH4A6Zij1//vyEhoZy+PDhy55/+vRpDh06lGK8iV155ZUplvtjW6cVb1Zj9mYfxMbGMmrUKAYMGEC5cuXil6tZsyYAjRo18ls94/b/I488gogAcO+99zJ27FhWrlzJ9ddfn6HXiJPZbbZp0yZ+//13Bg0aBEDdunWZPHkyIkK9es6p5fv27eO6665Lc5t4s90BbrzxxlQ/P3HSe3/cdtttrFmzJt26piZuuyemqimWZ5W/x9r6ChgFfBNXICJBwEdAe2A/sEZEpuEklTeSPb+Xqh71/P+S53kAVYCNnv/9dD6b+/744w8iIyNp2LBhfNmlS5cICQlJ9Tm1a9dm/vz5jBw5kipVqqS4TPI36Lp16+jTp4/Xb9rw8HBatWqVJPHs3LmTixcvUr9+fQBmzZrFqVOnOFX7FDO2zeDV1q/Sv2l/ak6uyaRJk7jhhhsyFXtwcDCXLl26rPz777/n0UcfTTf21D7c/trWqcWb1Zi92QcbNmzg+PHj3HbbbUmee+DAAYD4uvqjnnH7/5577okva9++PaVKlWLixIlJEom3rwGZ32arVq0CEpIowF133RX///bt2zl48GB8vKltE2+2e0ak9f4oVaoUJUqUyPA6AUqWLMmpU6cuKz99+jShoaGZWmeavGm2ZOUG1AA2Jbp/PTAn0f3BwOA0ni/AW0C7RGUPArd6/p+UyvMeA9YCa6tVq+Z1Uy4nGT9+vBYoUCDJ6Yf33HOPtmrVKsXlP/vsMy1YsKA2atRImzdv7vXrZLSPpEGDBvGdknEmT56s+fLl07Nnz+r3m7/Xtre21Vq1aumJCyd0z6k98cv17t1by5cvf9nxb29jDwkJSfEMmKzy17b2V7xp7YNz586pqurMmTMV0D179iRZ7u2339aSJUvG3/dHPbt37661atW6bLmU9n9m37cZ8c477yQ5jJXc448/rvXr14/vE0ltm3iz3TMirfdHVg5ttWzZMv4wXpy9e/cGbmd7ConkbpzDWYmTwqg0nt8PWAd8CvTRhD6SL4FPyMV9JAMHDtSrr746SdmQIUM0pcQ4d+5cLVCggH799de6YsWKy86ESUtGEklERITmz59fJ0+efHlcNatpm6/aKIPRoOCgFI/Fzps3TwGdN29ehmOP6+jMyAfBW/7Y1v6KN619ULt27fj7q1evVkB/+eWX+LJTp05ppUqVtHXr1kme58t6nj17VkNCQrza/1l532ZE3FljP/7442WPzZkzR0uUKKEbNmyIL0tpm3i73b2V3vvj+PHjumbNmnRvKRk+fLiWLFlSz5w5E1/2zjvvBO7pvykkkm4pJJIP/RlDoCaSjh07ao8ePZKUxV3rcPTo0fiyTZs2aYkSJZJcbNSuXTtt2rSpV6+TkUTy22+/KaDbtm2LLztx4YTWur6WUg8t9VYp7flqz/jO3B9//DHJbfLkyRocHKy9evXKcOwzZ85UEdHjx497FWtG+GNb+yvelPaBqmrXrl31zjvvjL8fERGhVatW1fr16+u0adN0ypQpGhYWpiVKlNCnnnoqfjlf13PcuHFe7f+svm8z4sKFC1q3bl1t0qSJnjhxQlVVY2Ji9PPPP9fatWtf9oWc0jbxdrt7y5/v55MnT2qFChW0Xbt2Om/ePP3ss8+0SJEigXtBYlYPbfniFqiJpFKlSknOBFF1vhxKlSql33zzjaqqHjlyRGvUqKF33313krNh4q6B+Pnnn9N9nYwkki+//FJDQkKSnBb5/or3lZJo0x5N9cSFE9qlS5d0m+OhoaF6+PDhDMXer1+/JL+kfckf29pf8aa0D1RVa9WqpUOHDk1StnLlSm3cuLEWKlRIr732Wp06daqWKlVKf/jhh/hlfF1Pb/Y/kOX3bUbt27dP77//fq1Tp47ecsst2rlzZ33ttdeS/GqPk3ybqGZsu3vDn+9nVdXNmzdrmzZttFChQlqhQgV96aWXMnxKdU5OJPmBXUBNoCAQDtT3ZwyBmkhS069fP+3cubOrMSzZvURnbnMOQURER+jGIxv9+nrR0dFauXJlHTs25SFU/CWz29qteNMzcuRILVu2rF68eDFJeW6rpy/483MWKNstRyQSYCJwCIjCOUOrt6e8M7AN2Am86M8YNBcmkn379mlISIj++eef2f7ae0/t1Xsn36sMRa/7/LpUL7bytYkTJ2rt2rXTvL7FHzK7rd2KN7GFCxfqsGHDdM6cOfrzzz/r448/rgUKFNBvv/32smUDuZ7+4s/PWaBsN28TiV9P/1XV+1IpnwnM9Odr52ZVqlThiy++4NChQ9SpUydbXvNS9CXeXf4uw5cOJ1ZjGXLTEAa1GOSXc9JToqp88cUX5M+fvbNDZ3ZbuxVvYhcuXGDcuHG89tpr5MuXj7CwMH7++Wc6dLj8Op5Arqe/+PNzltu2mzhJJ3cLCwvTtWvXuh1GQJv25zRun3Q7d9W7ixEdRlAjtIbbIRlj/ExE1qlqWHrL5Y50aPxi67GthB8J594G93JbndtY/chqmlZu6nZYxpgcxia2Mpc5dekUA2YPoOGnDXl27rNEREcgIpZEjDEpskRi4sXExvD5b59T58M6jFw1kt7X9Gb9v9YTnD/Y7dCMMTmYHdoy8TYf28xj0x/jhqo3MPuB2TSp2MTtkIwxAcASSR6359Qe5uycw2PXPkbD8g1Z+chKmlZqmm1nYxljAp8lkjzqfOR53l72Nm8vf5sgCaLrlV0pV6QczSo3czs0Y0yAsT6SPEZVmbhxIld+dCWvLnmVrld2ZcsTWyhXpFz6TzbGmBRYiySPOXL+CI9Mf4S6pesy4c4JtKze0u2QjDEBzhJJHnDk3BG+Cf+GZ294lgpFK7Cs1zKuLnc1QfmC3A7NGJML2KGtXCwiOoJ3lr3DFR9ewYsLXmTzsc0ANK7Q2JJIHnTx4kVatWpFTIwzqej8+fN58MEHkywTGRnJTTfdRHR0tBshmgBliSQXUlWm/zmdBp80YOD8gbSq0YpNj2+iQbkGbodmssGYMWNYsGABr7/+Ol26dGHMmDHx5XfeeSdBQc6PiPDwcK655pokzy1YsCBt27bl22+/zfa4TeCyQ1u50KXoS/x7xr8pFlyMWT1m0bF2R7dDylOeegp+/92362zcGD74IP3lBg4cSMuWLZk1axZPPPEEd911F9deey2dOnVi/PjxTJgwIX7Z8PBwKlSoQMuWLdm+fTvjxo2jXbt2dO3alcGDB9OjRw/fVsLkWtYiySX+vvg3/1n0HyJjIilcoDDz/28+G/pssCSSh2zevJkFCxZw2223cf78eQYMGEC9evUoW7Yse/fuZdeuXdSoUSN++fDwcMqUKcOvv/7Kxx9/zPjx4wFo0KABa9ascakWJhBZiyTARcdG8791/+PlhS/z96W/ub7q9XT4RweuLHOl26HlWd60HPxhypQp3HzzzQB8/PHHgNPncfToUapUqUJoaGj8slFRUZw8eZJnn30WgOjo6PjHg4KCKFiwIGfPnqVYsWLZXAsTiKxFEsAW/rWQa0dfy+MzH+fq8lfz22O/0eEfl881YfKGAwcOUKpUqSRlY8eOpVu3boSEhHDp0qX48i1bttCoUSPy5XO+AjZs2ECDBgl9aBERERQqVCh7AjcBz1okAUpVeW7ec5yJOMPkbpO5s96dNqxJHnf11VczZ86c+PvLly/nyy+/ZNq0aZQsWZKYmBguXbpEoUKFCA8Pp1GjRvHLbtiwgdtvvx2AEydOULZsWQoUKJDtdTCByRJJADkXeY4Ry0fQt1lfyoSUYXL3yZQvUp7CBQq7HZrJAR599FHWrFlD586dqVatGkWKFGH69OmULFkSgA4dOrB06VLatWtHeHg4zZolDIezadOm+BbJwoUL6dy5syt1MIHJZkgMALEay/gN4xk0fxCHzh3iy9u/pGfjnm6HZQLM+vXree+99xg7dmyay91555288cYb1K1bN5siMzmVzZCYS6zav4r+s/uz6sAqwiqFMbn7ZG6oeoPbYZkAdM0119CmTRtiYmLiryVJLjIykq5du1oSMRliLZIc7u7v7mbZvmW82fZNHmz0IPnEzo8wxmQPa5EEqEvRl3h/xfvcUe8OrixzJR91/oiQAiEUC7bTMI0xOZMlkhxCVZn6x1SemfsMf536C0V5oeULlC9a3u3QjDEmTZZIcoCNRzby1JynWPDXAuqXrc+8B+fRrlY7t8MyxhivWCLJAb4J/4b1h9YzqtMo/hX2L/Lns91ijAkc1tnugqiYKD5d+ykNyjWgTc02nI04S2RMJKVDSrsdmjHGxPO2sz0gTgESka4i8j8R+UlEOqRWFgjm7ZxH488a0292P77f8j0AxYKLWRIxxgQsvycSERkjIkdFZFOy8o4i8qeI7BCR59Nah6pOVdVHgZ7APamV5WQ7Tu7g9km302FcBy5FX2LqPVP5qPNHbodljDFZlh0H478CRgHfxBWISBDwEdAe2A+sEZFpQBDwRrLn91LVo57/X/I8L7GUynKcuTvnsuCvBbzZ9k2euu4pgvMHux2SMSmaOnUqM2bM4OjRozzxxBN06BAwDX7jFlX1+w2oAWxKdP96YE6i+4OBwWk8X4C3gHZplSV7zmPAWmBttWrVNLvFxMbol+u/1AkbJqiqalRMlB48czDb4zAmzoULF/Smm27S6OhoVVWdN2+ePvDAA6kuf/LkSe3Vq5eqqkZERGjLli01KioqW2I1OQOwVr34jnerj6QysC/R/f2estQ8CbQD7haRPmmUxVPV0aoapqphZcuW9VHY3lm+bznNP2/Owz89zKTNkwDIny8/FYtVzNY4TN6Ulal2E3v99dd54oknAJuC16TNrUSS0njnqZ4+pqr/VdVrVbWPqn6aWpnb9p/ZT48pPbhxzI0cPHuQcXeMY+o9U90Oy+QhAwcOpGzZssyaNYsHHniAESNG0LdvXw4dOsT48ePjh4oHJ5EcPnyYli1bUqFCBebPnw84RykGDRpEp06daNKkSfzyXbt2jZ9F0ZjE3LpgYT9QNdH9KsBBl2LxmY1HNvLDlh94qeVLDGoxiKIFi7odknFJ669aX1bWvX53Hm/6OBeiLtB5/OXDtPds3JOejXty/MJx7v7u7iSPLeq5KN3XjJtq9+2332bWrFkMGDCAH3/8Mc2pdhs0aMCvv/7KlClTGD9+PO3atePDDz9k/vz5nD59mh07dtCnj9Pgtyl4TWrcSiRrgCtEpCZwALgXuN+lWDJNVflh6w/sP7Ofp657ik5XdGL3U7upULSC26GZPMhXU+3269ePfv36XbZ+m4LXpMbviUREJgKtgTIish94RVW/EJG+wBycM7XGqOpmf8fiS+GHw+k/uz+L9yymWeVmPNnsSYLyBVkSMUDaLYiQAiFpPl4mpIxXLZDkDhw4kKTFAZmfajc1NgWvSYnfE4mq3pdK+Uxgpr9f39eOXzjOSwte4n+//Y+ShUryaZdPeaTJIwTlS3l+B2Oyi6+m2k2NTcFrUmODOmXQ4XOH+Tr8a/o168eQVkMoWbik2yEZA/huqt3U2BS8JjU21pYXZu+YzeLdi3mjnXOt5LHzxyhbJHtPKTYmq7ydajc1NgVv3pOrxtpy0wu/vECn8Z34YesPnIk4A2BJxASkxFPtZpRNwWvSYi2SdKw9uJbFuxfzZPMnKRhU0MeRGWNMzmVT7fpIWKUwwiqlux2NMSbPskNbxhhjssQSiTHGmCyxRGKMMSZLLJEYY4zJEkskxhhjssQSiTHGmCyxRGKMMSZLLJEYY4zJkjxxZbuIHAP2ACWA04keSut+4v/LAMd9EEry18vssqk9llJ5Zursq/qmFlNmlvNVnVN7LK/U2R/v69RiysxygVJnX32W03o8vc9u8jJ/1Lm6qqY/JpQ3E7vnlhsw2tv7yf5f64/Xz+yyqT2WUnlm6uyr+makzukt56s6p/ZYXqmzP97XebHOvvosZ6TO6W2D7NjPqd3y2qGt6Rm4n/wxf7x+ZpdN7bGUygOlzukt56s6p7c9fCEn19kf9c3IenNLnX31WU7rcW/eq9m9n1OUJw5tZZWIrFUvBi7LLfJafcHqnFdYnf0jr7VIMmu02wFks7xWX7A65xVWZz+wFokxxpgssRaJMcaYLLFEYowxJksskRhjjMkSSyRZICL5RGSYiHwoIg+5HU92EJHWIvKriHwqIq3djie7iEgREVknIre6HUt2EJF6nn08WUT+7XY82UFEuorI/0TkJxHp4HY82UFEaonIFyIyOSvrybOJRETGiMhREdmUrLyjiPwpIjtE5Pl0VnM7UBmIAvb7K1Zf8VGdFTgHFCLv1BlgEPCdf6L0LV/UWVW3qmofoDuQ40+X9VGdp6rqo0BP4B4/husTPqrzLlXtneVY8upZWyJyE84X4jeq2sBTFgRsA9rjfEmuAe4DgoA3kq2il+f2t6p+JiKTVfXu7Io/M3xU5+OqGisi5YH3VLVHdsWfGT6qc0OcYSYK4dT/5+yJPnN8UWdVPSoi/wSeB0ap6oTsij8zfFVnz/PeBcar6m/ZFH6m+LjOWfr+yp/ZJwY6VV0iIjWSFTcDdqjqLgARmQTcrqpvAJcd0hCR/UCk526M/6L1DV/UOZG/gWB/xOlLPtrPbYAiwFXARRGZqaqxfg08C3y1n1V1GjBNRGYAOTqR+Gg/C/AmMIwX+9oAAAH6SURBVCunJxHw+ec5S/JsIklFZWBfovv7geZpLD8F+FBEWgJL/BmYH2WoziJyJ3ALEAqM8m9ofpOhOqvqiwAi0hNPi8yv0flHRvdza+BOnB8LM/0amf9k9PP8JNAOKCEitVX1U38G5ycZ3c+lgWHANSIy2JNwMswSSVKSQlmqx/5U9QKQ5eOLLstonafgJNBAlqE6xy+g+pXvQ8k2Gd3Pi4BF/gomm2S0zv8F/uu/cLJFRut8AuiT1RfNs53tqdgPVE10vwpw0KVYsovV2eqcW1mds6nOlkiSWgNcISI1RaQgcC8wzeWY/M3qbHXOrazO2VTnPJtIRGQisAKoKyL7RaS3qkYDfYE5wFbgO1Xd7GacvmR1tjpjdbY6+yOWvHr6rzHGGN/Isy0SY4wxvmGJxBhjTJZYIjHGGJMllkiMMcZkiSUSY4wxWWKJxBhjTJZYIjHGBSJSI/nw38YEKkskxhhjssQSiTHuCfLMyLdZROaKSGG3AzImMyyRGOOeK4CPVLU+cAq4y+V4jMkUSyTGuOcvVf3d8/86oIaLsRiTaZZIjHFPRKL/Y7D5gUyAskRijDEmSyyRGGOMyRIbRt4YY0yWWIvEGGNMllgiMcYYkyWWSIwxxmSJJRJjjDFZYonEGGNMllgiMcYYkyWWSIwxxmSJJRJjjDFZ8v/ZcHf09djr5QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0a5a5e9710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Tests.checkDerivative(function_and_deriv, x0=m0, num=6, expectedOrder=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.49350416758e-11 1.49350416758e-11 3.23117426779e-27\n"
     ]
    }
   ],
   "source": [
    "# Test Adjoint\n",
    "u = prob.fields(m0)\n",
    "\n",
    "v = np.random.rand(len(u))\n",
    "w = np.random.rand(len(m0))\n",
    "\n",
    "vJw = v.dot(prob.Jvec(m0, w))\n",
    "wJtv = w.dot(prob.Jtvec(m0, v))\n",
    "\n",
    "print(vJw, wJtv, vJw - wJtv)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Invert \n",
    "\n",
    "- Use a 20 layer model, starting model and reference model of log(1 Ohm m) \n",
    "- we will invert the clean data for now\n",
    "- assign a 5% error and a 5e-15 floor so that we don't over-fit the zero-crossings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "survey = EmpymodSurvey()\n",
    "survey.pair(prob)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "rdata = np.ravel(data, order='F')\n",
    "survey.dobs = np.hstack([rdata.real, rdata.imag])  # observed data\n",
    "survey.std = np.abs(survey.dobs)*.05\n",
    "survey.eps = 5e-15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# regularization\n",
    "reg = SimPEG.Regularization.Simple(\n",
    "    SimPEG.Mesh.TensorMesh([len(m0)]), alpha_s=5e-5, alpha_x=1, mref=np.zeros(nlayers+1)\n",
    ")\n",
    "\n",
    "# data misfit\n",
    "dmis = SimPEG.DataMisfit.l2_DataMisfit(survey)\n",
    "\n",
    "# optimization\n",
    "opt = SimPEG.Optimization.InexactGaussNewton(maxIter=40)\n",
    "\n",
    "# inverse problem\n",
    "invProb = SimPEG.InvProblem.BaseInvProblem(dmis, reg, opt, beta=5)\n",
    "\n",
    "# directives \n",
    "target = SimPEG.Directives.TargetMisfit()\n",
    "directives = [\n",
    "#     SimPEG.Directives.BetaEstimate_ByEig(beta0_ratio=5e-2),\n",
    "    target\n",
    "]\n",
    "\n",
    "# inversion\n",
    "inv = SimPEG.Inversion.BaseInversion(invProb, directiveList=directives)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "m0 = np.zeros(nlayers+1) # starting model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "    ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  5.00e+00  9.28e+02  0.00e+00  9.28e+02    2.84e+02      0              \n",
      "   1  5.00e+00  2.96e+02  6.45e-01  2.99e+02    1.44e+03      1              \n",
      "   2  5.00e+00  7.44e+01  7.56e-01  7.82e+01    5.46e+01      0              \n",
      "   3  5.00e+00  6.37e+01  6.65e-01  6.71e+01    1.80e+02      1              \n",
      "   4  5.00e+00  5.84e+01  8.92e-01  6.29e+01    2.05e+02      0              \n",
      "   5  5.00e+00  5.52e+01  9.89e-01  6.01e+01    1.14e+02      0              \n",
      "   6  5.00e+00  5.32e+01  1.15e+00  5.90e+01    9.08e+01      0   Skip BFGS  \n",
      "   7  5.00e+00  5.31e+01  1.10e+00  5.86e+01    1.01e+02      0              \n",
      "   8  5.00e+00  5.29e+01  1.11e+00  5.84e+01    6.75e+01      0              \n",
      "   9  5.00e+00  5.24e+01  1.18e+00  5.83e+01    7.54e+01      0              \n",
      "  10  5.00e+00  5.26e+01  1.11e+00  5.82e+01    6.80e+01      0              \n",
      "  11  5.00e+00  5.21e+01  1.09e+00  5.75e+01    7.56e+01      0   Skip BFGS  \n",
      "  12  5.00e+00  5.16e+01  1.13e+00  5.73e+01    6.04e+01      0              \n",
      "  13  5.00e+00  5.09e+01  9.85e-01  5.59e+01    1.39e+02      0              \n",
      "  14  5.00e+00  4.96e+01  1.07e+00  5.50e+01    1.01e+02      0              \n",
      "  15  5.00e+00  4.77e+01  1.33e+00  5.43e+01    2.51e+02      0   Skip BFGS  \n",
      "  16  5.00e+00  4.69e+01  1.14e+00  5.25e+01    1.52e+02      0              \n",
      "  17  5.00e+00  4.51e+01  9.49e-01  4.99e+01    2.10e+02      1              \n",
      "  18  5.00e+00  4.40e+01  1.05e+00  4.93e+01    1.93e+02      1              \n",
      "  19  5.00e+00  4.15e+01  1.03e+00  4.67e+01    2.32e+02      1              \n",
      "  20  5.00e+00  4.02e+01  1.13e+00  4.58e+01    2.46e+02      1              \n",
      "  21  5.00e+00  3.21e+01  1.48e+00  3.95e+01    2.74e+02      1              \n",
      "  22  5.00e+00  2.70e+01  1.74e+00  3.56e+01    2.86e+02      1   Skip BFGS  \n",
      "  23  5.00e+00  2.62e+01  1.72e+00  3.48e+01    2.73e+02      1              \n",
      "  24  5.00e+00  2.17e+01  1.66e+00  3.00e+01    1.47e+02      0              \n",
      "  25  5.00e+00  1.26e+01  2.58e+00  2.55e+01    1.77e+02      0   Skip BFGS  \n",
      "  26  5.00e+00  1.21e+01  2.61e+00  2.52e+01    2.15e+02      0              \n",
      "  27  5.00e+00  1.11e+01  2.50e+00  2.36e+01    1.54e+02      0              \n",
      "  28  5.00e+00  1.00e+01  2.63e+00  2.32e+01    1.27e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 0.0000e+00 <= tolF*(1+|f0|) = 9.2850e+01\n",
      "0 : |xc-x_last| = 4.1888e-01 <= tolX*(1+|x0|) = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 1.2744e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 1.2744e+02 <= 1e3*eps       = 1.0000e-02\n",
      "0 : maxIter   =      40    <= iter          =     29\n",
      "------------------------- DONE! -------------------------\n"
     ]
    }
   ],
   "source": [
    "# run the inversion \n",
    "mrec = inv.run(m0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Target Resistance    : 9900 \n",
      "Inversion Resistance : 9587\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0a5a4a0be0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "thick = mesh.gridN[3:]-mesh.gridN[2:-1]\n",
    "TR = np.sum(thick*(prob.rho_map*mrec)[2:-1])-np.sum(thick*(prob.rho_map*m0)[2:-1])\n",
    "print('Target Resistance    :', 100*(100-1), '\\nInversion Resistance :', int(np.round(TR)))\n",
    "\n",
    "plot_data('SimPEG Inversion',\n",
    "          {'Model': (depth, res),\n",
    "           'Initial Model': (mesh.gridN, prob.rho_map*m0),\n",
    "           'Inversion': (mesh.gridN, prob.rho_map*mrec)},\n",
    "          (re_field(prob.fields(mrec))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test speed of standard `bipole` vs custom `bipole`\n",
    "\n",
    "**Conclusion**: We probably should just use the standard `bipole`; to be discussed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "from empymod.utils import (check_model, check_frequency, check_hankel, check_opt,\n",
    "                           check_bipole, get_abs, get_geo_fact, get_azm_dip,\n",
    "                           get_off_ang, get_layer_nr, printstartfinish, conv_warning)\n",
    "from empymod.model import bipole, dipole, fem\n",
    "\n",
    "\n",
    "def _bipole(src, rec, depth, freq, etaH, etaV, zetaH, zetaV, nsrcz, srcdipole,\n",
    "            nrecz, recdipole, isrz, irec, isrc, msrc, nrec, nsrc, srcpts, mrec,\n",
    "            recpts, strength, xdirect, ht, htarg, isfullspace, use_spline,\n",
    "            loop_off, loop_freq, use_ne_eval, verb):\n",
    "    \"\"\"Return the electromagnetic field due to an electromagnetic source.\"\"\"\n",
    "\n",
    "    # Pre-allocate output EM array\n",
    "    EM = np.zeros((freq.size, nrec*nsrc), dtype=complex)\n",
    "\n",
    "    # Initialize kernel count, conv (only for QWE)\n",
    "    # (how many times the wavenumber-domain kernel was calld)\n",
    "    kcount = 0\n",
    "    conv = True\n",
    "\n",
    "    # The kernel handles only 1 ab with one srcz-recz combination at once.\n",
    "    # Hence we have to loop over every different depth of src or rec, and\n",
    "    # over all required ab's.\n",
    "    for isz in range(nsrcz):  # Loop over source depths\n",
    "\n",
    "        # Get this source\n",
    "        srcazmdip = get_azm_dip(src, isz, nsrcz, srcpts, srcdipole, strength,\n",
    "                                'src', verb)\n",
    "        tsrc, srcazm, srcdip, srcg_w, srcpts, src_w = srcazmdip\n",
    "\n",
    "        for irz in range(nrecz):  # Loop over receiver depths\n",
    "\n",
    "            # Get this receiver\n",
    "            recazmdip = get_azm_dip(rec, irz, nrecz, recpts, recdipole,\n",
    "                                    strength, 'rec', verb)\n",
    "            trec, recazm, recdip, recg_w, recpts, rec_w = recazmdip\n",
    "\n",
    "            # Get required ab's\n",
    "            ab_calc = get_abs(msrc, mrec, srcazm, srcdip, recazm, recdip, verb)\n",
    "\n",
    "            # Pre-allocate temporary source-EM array for integration loop\n",
    "            sEM = np.zeros((freq.size, isrz), dtype=complex)\n",
    "\n",
    "            for isg in range(srcpts):  # Loop over src integration points\n",
    "\n",
    "                # This integration source\n",
    "                tisrc = [tsrc[0][isg::srcpts], tsrc[1][isg::srcpts],\n",
    "                         tsrc[2][isg]]\n",
    "\n",
    "                # Get layer number in which src resides\n",
    "                lsrc, zsrc = get_layer_nr(tisrc, depth)\n",
    "\n",
    "                # Pre-allocate temporary receiver EM arrays for integr. loop\n",
    "                rEM = np.zeros((freq.size, isrz), dtype=complex)\n",
    "\n",
    "                for irg in range(recpts):  # Loop over rec integration pts\n",
    "                    # Note, if source or receiver is a bipole, but horizontal\n",
    "                    # (dip=0), then calculation could be sped up by not looping\n",
    "                    # over the bipole elements, but calculate it all in one go.\n",
    "\n",
    "                    # This integration receiver\n",
    "                    tirec = [trec[0][irg::recpts], trec[1][irg::recpts],\n",
    "                             trec[2][irg]]\n",
    "\n",
    "                    # Get src-rec offsets and angles\n",
    "                    off, angle = get_off_ang(tisrc, tirec, isrc, irec, verb)\n",
    "\n",
    "                    # Get layer number in which rec resides\n",
    "                    lrec, zrec = get_layer_nr(tirec, depth)\n",
    "\n",
    "                    # Gather variables\n",
    "                    finp = (off, angle, zsrc, zrec, lsrc, lrec, depth, freq,\n",
    "                            etaH, etaV, zetaH, zetaV, xdirect, isfullspace, ht,\n",
    "                            htarg, use_spline, use_ne_eval, msrc, mrec,\n",
    "                            loop_freq, loop_off, conv)\n",
    "\n",
    "                    # Pre-allocate temporary EM array for ab-loop\n",
    "                    abEM = np.zeros((freq.size, isrz), dtype=complex)\n",
    "\n",
    "                    for iab in ab_calc:  # Loop over required ab's\n",
    "\n",
    "                        # Carry-out the frequency-domain calculation\n",
    "                        out = fem(iab, *finp)\n",
    "\n",
    "                        # Get geometrical scaling factor\n",
    "                        tfact = get_geo_fact(iab, srcazm, srcdip, recazm,\n",
    "                                             recdip, msrc, mrec)\n",
    "\n",
    "                        # Add field to EM with geometrical factor\n",
    "                        abEM += out[0]*np.squeeze(tfact)\n",
    "\n",
    "                        # Update kernel count\n",
    "                        kcount += out[1]\n",
    "\n",
    "                        # Update conv (QWE convergence)\n",
    "                        conv *= out[2]\n",
    "\n",
    "                    # Add this receiver element, with weight from integration\n",
    "                    rEM += abEM*recg_w[irg]\n",
    "\n",
    "                # Add this source element, with weight from integration\n",
    "                sEM += rEM*srcg_w[isg]\n",
    "\n",
    "            # Get required [s]tart- and [e]nd-[i]ndices, and [st]ep\n",
    "            if nrec == nrecz:\n",
    "                if nsrc == nsrcz:  # Case 1: Looped over each src and each rec\n",
    "                    si = isz*nrec + irz\n",
    "                    ei = si + 1\n",
    "                    st = 1\n",
    "                else:              # Case 2: Looped over each rec\n",
    "                    si = irz\n",
    "                    ei = nsrc*nrec\n",
    "                    st = nrec\n",
    "            else:\n",
    "                if nsrc == nsrcz:  # Case 3: Looped over each src\n",
    "                    si = isz*nrec\n",
    "                    ei = si + nrec\n",
    "                    st = 1\n",
    "                else:              # Case 4: All in one go\n",
    "                    si = 0\n",
    "                    ei = nsrc*nrec\n",
    "                    st = 1\n",
    "\n",
    "            # Get required scaling from src-strength and src/rec-length\n",
    "            src_rec_w = 1\n",
    "            if strength > 0:\n",
    "                src_rec_w *= np.repeat(src_w, irec)\n",
    "                src_rec_w *= np.tile(rec_w, isrc)\n",
    "\n",
    "            # Add this src-rec signal\n",
    "            EM[:, si:ei:st] = sEM*src_rec_w\n",
    "\n",
    "    # In case of QWE/QUAD, print Warning if not converged\n",
    "    conv_warning(conv, htarg, 'Hankel', verb)\n",
    "\n",
    "    # Reshape for number of sources\n",
    "    EM = np.squeeze(EM.reshape((-1, nrec, nsrc), order='F'))\n",
    "\n",
    "    return EM, conv, kcount\n",
    "\n",
    "\n",
    "def custom_bipole(n, src, rec, depth, res, freqtime, aniso=None, epermH=None,\n",
    "                  epermV=None, mpermH=None, mpermV=None, msrc=False,\n",
    "                  srcpts=1, mrec=False, recpts=1, strength=0, xdirect=True,\n",
    "                  ht='fht', htarg=None, opt=None, loop=None, verb=2):\n",
    "\n",
    "    # === 1.  LET'S START ============\n",
    "    t0 = printstartfinish(verb)\n",
    "\n",
    "    # === 2.  CHECK INPUT ============\n",
    "\n",
    "    # Check layer parameters\n",
    "    model = check_model(depth, res, aniso, epermH, epermV, mpermH, mpermV,\n",
    "                        xdirect, verb)\n",
    "    depth, res, aniso, epermH, epermV, mpermH, mpermV, isfullspace = model\n",
    "\n",
    "    # Check frequency => get etaH, etaV, zetaH, and zetaV\n",
    "    frequency = check_frequency(freqtime, res, aniso, epermH, epermV, mpermH,\n",
    "                                mpermV, verb)\n",
    "    freq, etaH, etaV, zetaH, zetaV = frequency\n",
    "\n",
    "    # Check Hankel transform parameters\n",
    "    ht, htarg = check_hankel(ht, htarg, verb)\n",
    "\n",
    "    # Check optimization\n",
    "    optimization = check_opt(opt, loop, ht, htarg, verb)\n",
    "    use_spline, use_ne_eval, loop_freq, loop_off = optimization\n",
    "\n",
    "    # Check src and rec, get flags if dipole or not\n",
    "    # nsrcz/nrecz are number of unique src/rec-pole depths\n",
    "    src, nsrc, nsrcz, srcdipole = check_bipole(src, 'src')\n",
    "    rec, nrec, nrecz, recdipole = check_bipole(rec, 'rec')\n",
    "\n",
    "    # === 3. EM-FIELD CALCULATION ============\n",
    "\n",
    "    # Define some indeces\n",
    "    isrc = int(nsrc/nsrcz)  # this is either 1 or nsrc\n",
    "    irec = int(nrec/nrecz)  # this is either 1 or nrec\n",
    "    isrz = int(isrc*irec)   # this is either 1, nsrc, nrec, or nsrc*nrec\n",
    "\n",
    "    for i in range(n):\n",
    "        EM, conv, kcount = _bipole(src, rec, depth, freq, etaH, etaV, zetaH, zetaV,\n",
    "                                   nsrcz, srcdipole, nrecz, recdipole, isrz, irec,\n",
    "                                   isrc, msrc, nrec, nsrc, srcpts, mrec, recpts,\n",
    "                                   strength, xdirect, ht, htarg, isfullspace,\n",
    "                                   use_spline, loop_off, loop_freq, use_ne_eval, verb)\n",
    "\n",
    "    # === 4.  FINISHED ============\n",
    "    printstartfinish(verb, t0, kcount)\n",
    "\n",
    "    return EM, kcount\n",
    "\n",
    "\n",
    "def multi_bipole(n, inp):\n",
    "\n",
    "    for i in range(n):\n",
    "        EM = bipole(**inp)\n",
    "        \n",
    "    return EM\n",
    "\n",
    "\n",
    "def multi_dipole(n, inp):\n",
    "\n",
    "    for i in range(n):\n",
    "        EM = dipole(**inp)\n",
    "\n",
    "    return EM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Simple survey (only one kernel call required)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Runs:  1 ; kernel calls per run:  1\n",
      "======================================\n",
      "\n",
      "=> Custom\n",
      "2.93 ms ± 39.1 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n",
      "\n",
      "=> Standard\n",
      "2.89 ms ± 46.5 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n",
      "\n",
      "=> Standard dipole\n",
      "2.53 ms ± 17.2 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n",
      "\n",
      "\n",
      "Runs:  10 ; kernel calls per run:  1\n",
      "======================================\n",
      "\n",
      "=> Custom\n",
      "26.5 ms ± 157 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n",
      "\n",
      "=> Standard\n",
      "28.7 ms ± 106 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n",
      "\n",
      "=> Standard dipole\n",
      "25.5 ms ± 236 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n",
      "\n",
      "\n",
      "Runs:  100 ; kernel calls per run:  1\n",
      "======================================\n",
      "\n",
      "=> Custom\n",
      "260 ms ± 1.94 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n",
      "=> Standard\n",
      "290 ms ± 3.47 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n",
      "=> Standard dipole\n",
      "253 ms ± 2.04 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n",
      "\n",
      "Runs:  1000 ; kernel calls per run:  1\n",
      "======================================\n",
      "\n",
      "=> Custom\n",
      "2.62 s ± 4.64 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n",
      "=> Standard\n",
      "2.89 s ± 15.4 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n",
      "=> Standard dipole\n",
      "2.68 s ± 68.5 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "inp = {\n",
    "    'src': [0, 0, 950, 0, 0],\n",
    "    'rec': [5000, 0, 1000, 0, 0],\n",
    "    'depth': [0, 1000, 2000, 2100],\n",
    "    'res': [2e14, 0.3, 1, 100, 1],\n",
    "    'freqtime': [0.01, 0.056, 0.1, 0.5, 1.],\n",
    "    'verb': 1}\n",
    "\n",
    "dinp = {\n",
    "    'src': [0, 0, 950],\n",
    "    'rec': [5000, 0, 1000],\n",
    "    'depth': [0, 1000, 2000, 2100],\n",
    "    'res': [2e14, 0.3, 1, 100, 1],\n",
    "    'freqtime': [0.01, 0.056, 0.1, 0.5, 1.],\n",
    "    'verb': 1}\n",
    "\n",
    "for n in [1, 10, 100, 1000]:\n",
    "    \n",
    "    _, kcount = custom_bipole(n, **inp)\n",
    "\n",
    "    print('\\nRuns: ', n, '; kernel calls per run: ', kcount)\n",
    "    print('======================================\\n')\n",
    "\n",
    "    print('=> Custom')\n",
    "    %timeit custom_bipole(n, **inp)\n",
    "\n",
    "    print('\\n=> Standard')\n",
    "    %timeit multi_bipole(n, inp)\n",
    "\n",
    "    print('\\n=> Standard dipole')\n",
    "    %timeit multi_dipole(n, dinp)\n",
    "    \n",
    "    print('')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### More complicated survey "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Runs:  1 ; kernel calls per run:  30\n",
      "======================================\n",
      "\n",
      "=> Custom\n",
      "790 ms ± 10.4 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n",
      "=> Standard\n",
      "801 ms ± 18.6 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n",
      "\n",
      "Runs:  10 ; kernel calls per run:  30\n",
      "======================================\n",
      "\n",
      "=> Custom\n",
      "7.91 s ± 106 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n",
      "=> Standard\n",
      "7.86 s ± 34.3 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "inp2 = {\n",
    "    'src': [[-50, 0], [50, 100], [-100, -100], [-150, -150], 900, 950],\n",
    "    'rec': [np.arange(1,11)*500, np.arange(1,11)*500+50, np.ones(10)*100, np.ones(10)*150, 1000, 1000],\n",
    "    'depth': [0, 1000, 2000, 2100],\n",
    "    'res': [2e14, 0.3, 1, 100, 1],\n",
    "    'freqtime': [0.01, 0.056, 0.1, 0.5, 1.],\n",
    "    'srcpts': 5,\n",
    "    'recpts': 1,\n",
    "    'strength': 100,\n",
    "    'verb': 1}\n",
    "\n",
    "for n in [1, 10]:  #, 100, 1000]:\n",
    "\n",
    "    _, kcount = custom_bipole(n, **inp2)\n",
    "\n",
    "    print('\\nRuns: ', n, '; kernel calls per run: ', kcount)\n",
    "    print('======================================\\n')\n",
    "\n",
    "    print('=> Custom')\n",
    "    %timeit custom_bipole(n, **inp2)\n",
    "\n",
    "    print('\\n=> Standard')\n",
    "    %timeit multi_bipole(n, inp2)\n",
    "    \n",
    "    print('')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<h3>Tue Jan 23 17:24:48 2018 CST</h3><table><tr style='border: 2px solid #fff; text-align: left;'><td style='background-color: #ccc;width: 10%; border: 2px solid #fff;'>1.5.0</td><td style='border: 2px solid #fff; text-align: left;'>empymod</td><td style='background-color: #ccc;width: 10%; border: 2px solid #fff;'>1.13.3</td><td style='border: 2px solid #fff; text-align: left;'>numpy</td><td style='background-color: #ccc;width: 10%; border: 2px solid #fff;'>Linux</td><td style='border: 2px solid #fff; text-align: left;'>OS</td><td style='background-color: #ccc;width: 10%; border: 2px solid #fff;'>6.2.1</td><td style='border: 2px solid #fff; text-align: left;'>IPython</td></tr><tr style='border: 2px solid #fff; text-align: left;'><td style='background-color: #ccc;width: 10%; border: 2px solid #fff;'>0.6.5</td><td style='border: 2px solid #fff; text-align: left;'>SimPEG</td><td style='background-color: #ccc;width: 10%; border: 2px solid #fff;'>1.0.0</td><td style='border: 2px solid #fff; text-align: left;'>scipy</td><td style='background-color: #ccc;width: 10%; border: 2px solid #fff;'>8</td><td style='border: 2px solid #fff; text-align: left;'>CPU(s)</td><td style='background-color: #ccc;width: 10%; border: 2px solid #fff;'>2.1.1</td><td style='border: 2px solid #fff; text-align: left;'>matplotlib</td></tr><tr style='border: 2px solid #fff; text-align: left;'><td style='border: 2px solid #fff; text-align: left;' colspan='8'>3.6.4 |Anaconda custom (64-bit)| (default, Dec 21 2017, 21:42:08) \n",
       "[GCC 7.2.0]</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import printinfo\n",
    "printinfo.info()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}