circuit_autoencoder_completo_200-2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "dd9ba1be",
   "metadata": {},
   "source": [
    "# Circuito cuántico para autoencoder\n",
    "\n",
    "Propuesta para el hackathon es diseñar un circuito cuántico que pueda disminuir el número de variables de imagenes sin perder información y poder trabajar con menos qubits para una clasificación usando el conjunto de datos MNIST.\n",
    "\n",
    "Se indican a continuación las dependencias necesarias para poder trabajar en este problema."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "8ed9401c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "# Bibliotecas necesarias de Qiskit \n",
    "from qiskit import QuantumCircuit, transpile, Aer, IBMQ, execute, QuantumRegister, ClassicalRegister\n",
    "from qiskit.tools.jupyter import *\n",
    "from qiskit.visualization import *\n",
    "from qiskit.circuit import Parameter, ParameterVector\n",
    "\n",
    "#Bliblioteca para la adquisición y preprocesamiento del conjunto MNIST.\n",
    "import tensorflow as tf\n",
    "\n",
    "#Bibliotecas para graficar \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set_theme()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "513a9fcb",
   "metadata": {},
   "source": [
    "## Preprocesamiento\n",
    "\n",
    "Inicializamos las carácteristicas de nuestro circuito cuántico que es el conjunto de datos MNIST, para esto nos apoyamos de tensorflow que ya tiene el conjunto de datos.\n",
    "\n",
    "Cada conjunto tiene 10 clases : **[0,1,2,3,4,5,6,7,8,9]**,\n",
    "y van de 0 a 255, por ello nosotros pasamos un proceso de **normalización que va de 0.0 a 1.0**, donde negro representa el fondo y el blanco y escala de gris a los números.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "b37cbb4c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Imagenes del conjunto de entrenamiento: 60000\n",
      "Imagenes del conjunto de entrenamiento: 10000\n"
     ]
    }
   ],
   "source": [
    "(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()\n",
    "\n",
    "# Rescale the images from [0,255] to the [0.0,1.0] range.\n",
    "x_train, x_test = x_train[..., np.newaxis]/255.0, x_test[..., np.newaxis]/255.0\n",
    "\n",
    "print(\"Imagenes del conjunto de entrenamiento:\", len(x_train))\n",
    "print(\"Imagenes del conjunto de entrenamiento:\", len(x_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a14c1d20",
   "metadata": {},
   "outputs": [],
   "source": [
    "def filter_01(x, y):\n",
    "    keep = (y == 0) | (y == 1)\n",
    "    x, y = x[keep], y[keep]\n",
    "    return x,y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "541c3f72",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Imagenes del conjunto de entrenamiento: 12665\n",
      "Imagenes del conjunto de entrenamiento: 2115\n"
     ]
    }
   ],
   "source": [
    "x_train, y_train = filter_01(x_train, y_train)\n",
    "x_test, y_test = filter_01(x_test, y_test)\n",
    "\n",
    "print(\"Imagenes del conjunto de entrenamiento:\", len(x_train))\n",
    "print(\"Imagenes del conjunto de entrenamiento:\", len(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae624263",
   "metadata": {},
   "source": [
    "Se representan las imagenes que tienen un tamaño original de 28x28"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6d6c1f06",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x1db303d3040>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(x_train[0, :, :, 0])\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "502cfac4",
   "metadata": {},
   "source": [
    "### Reducción de la imagen\n",
    "\n",
    "Tensorflow tiene el método *tf.image.resize* que disminuye las imagenes a partir de los siguientes posibles criteros:\n",
    "\n",
    "<ul>\n",
    "<li><b>bilinear</b>: Bilinear interpolation. If antialias is true, becomes a hat/tent filter function with radius 1 when downsampling.</li>\n",
    "<li><b>lanczos3</b>: Lanczos kernel with radius 3. High-quality practical filter but may have some ringing, especially on synthetic images.</li>\n",
    "<li><b>lanczos5</b>: Lanczos kernel with radius 5. Very-high-quality filter but may have stronger ringing.</li>\n",
    "<li><b>bicubic</b>: Cubic interpolant of Keys. Equivalent to Catmull-Rom kernel. Reasonably good quality and faster than Lanczos3Kernel, particularly when upsampling.</li>\n",
    "<li><b>gaussian</b>: Gaussian kernel with radius 3, sigma = 1.5 / 3.0.</li>\n",
    "<li><b>nearest</b>: Nearest neighbor interpolation. antialias has no effect when used with nearest neighbor interpolation.</li>\n",
    "<li><b>area</b>: Anti-aliased resampling with area interpolation. antialias has no effect when used with area interpolation; it always anti-aliases.</li>\n",
    "<li><b>mitchellcubic</b>: Mitchell-Netravali Cubic non-interpolating filter. For synthetic images (especially those lacking proper prefiltering), less ringing than Keys cubic kernel but less sharp.</li>\n",
    "    \n",
    "</ul>\n",
    "\n",
    "\n",
    "Ejemplo de dicho preprocesamiento de reducción de datos se emplea a continuación con el método nearest y con el tamaño de imagen 8x8."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cdc4efcd",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train_small = tf.image.resize(x_train, (8,8), method='nearest', preserve_aspect_ratio=True).numpy()\n",
    "x_test_small = tf.image.resize(x_test, (8,8), method='nearest', preserve_aspect_ratio=True).numpy()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d0e50e61",
   "metadata": {},
   "source": [
    "En este punto se tienen imagenes de tamaño 8x8 y se tienen que pasar como un estado de amplitud ya que $8x8 = 64$  y esto nos dará un vector de $2^6$ que recordando el valor 6 es el número de qubits a usar "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "29c71f0e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x1db305a4b80>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(x_train_small[0,:,:,0], vmin=0, vmax=1)\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "80bebd2f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((12665, 64), (2115, 64))"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train = x_train_small.reshape(len(x_train_small), 64)\n",
    "x_test = x_test_small.reshape(len(x_test_small), 64)\n",
    "x_train.shape,x_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "990434ab",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ]],\n",
       "\n",
       "       [[0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.93333333],\n",
       "        [0.92941176],\n",
       "        [0.        ],\n",
       "        [0.        ]],\n",
       "\n",
       "       [[0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.98823529],\n",
       "        [0.99215686],\n",
       "        [0.74117647],\n",
       "        [0.        ],\n",
       "        [0.        ]],\n",
       "\n",
       "       [[0.        ],\n",
       "        [0.        ],\n",
       "        [0.69803922],\n",
       "        [0.0745098 ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.76470588],\n",
       "        [0.        ]],\n",
       "\n",
       "       [[0.        ],\n",
       "        [0.        ],\n",
       "        [0.98823529],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.58039216],\n",
       "        [0.        ]],\n",
       "\n",
       "       [[0.        ],\n",
       "        [0.        ],\n",
       "        [0.88235294],\n",
       "        [0.        ],\n",
       "        [0.44705882],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ]],\n",
       "\n",
       "       [[0.        ],\n",
       "        [0.        ],\n",
       "        [0.98823529],\n",
       "        [0.98823529],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ]],\n",
       "\n",
       "       [[0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ],\n",
       "        [0.        ]]])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_small[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7f1eac32",
   "metadata": {},
   "source": [
    "## De imagen a qubits\n",
    "\n",
    "Por las capacidades que actualmente se usa CPU para el diseño de circuitos cuánticos en Qiskit, no es posible ocupar todo el conjunto de imagenes, se diseñaron 8 experimentos por lo menos desde 10 iteraciones hasta 50, ocupando las siguientes entradas:\n",
    "\n",
    "<ul>\n",
    "    <li>Las 5 primeras imagenes del conjunto de entrenamiento.</li>\n",
    "    <li>Las 10 primeras imagenes del conjunto de entrenamiento.</li>\n",
    "    <li>Las 12 primeras imagenes del conjunto de entrenamiento.</li>\n",
    "    <li>Las 50 primeras imagenes del conjunto de entrenamiento.</li>\n",
    "    <li>Las 70 primeras imagenes del conjunto de entrenamiento.</li>\n",
    "    <li>Las 100 primeras imagenes del conjunto de entrenamiento.</li>\n",
    "    <li>Las 200 primeras imagenes del conjunto de entrenamiento.</li>\n",
    "    <li>Las 500 primeras imagenes del conjunto de entrenamiento.</li>\n",
    "\n",
    "</ul>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "619f1ddc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((12665, 64), (2115, 64))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train = (x_train)\n",
    "x_test = (x_test)\n",
    "\n",
    "x_train.shape,x_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45c2a60c",
   "metadata": {},
   "source": [
    "Se pasa de la matriz de tamaño 8x8 a un vector de tamaño 64x1 por cada imagen"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e1562960",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.        , 0.93333333, 0.92941176, 0.        ,\n",
       "       0.        , 0.        , 0.        , 0.        , 0.98823529,\n",
       "       0.99215686, 0.74117647, 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.69803922, 0.0745098 , 0.        , 0.        ,\n",
       "       0.76470588, 0.        , 0.        , 0.        , 0.98823529,\n",
       "       0.        , 0.        , 0.        , 0.58039216, 0.        ,\n",
       "       0.        , 0.        , 0.88235294, 0.        , 0.44705882,\n",
       "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "       0.98823529, 0.98823529, 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.        , 0.        , 0.        ])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "153e4958",
   "metadata": {},
   "source": [
    "Se elimina las posibles imagenes que no tengan información, es decir, puros 0's para el conjutno de entrenamiento"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "966561a2",
   "metadata": {},
   "outputs": [],
   "source": [
    "k = 0\n",
    "\n",
    "while k < len(x_train):\n",
    "  a = x_train[k].copy()\n",
    "  if np.sum(a) == 0.:\n",
    "    print(k,x_train[k])\n",
    "    x_train = np.delete(x_train, k, axis=0)\n",
    "    y_train = np.delete(y_train, k, axis=0)\n",
    "    k -= 1\n",
    "  k+=1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a29eff1",
   "metadata": {},
   "source": [
    "Se elimina las posibles imagenes que no tengan información, es decir, puros 0's para el conjutno de testeo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "7e9f3b3f",
   "metadata": {},
   "outputs": [],
   "source": [
    "k = 0\n",
    "\n",
    "while k < len(x_test):\n",
    "  a = x_test[k].copy()\n",
    "  if np.sum(a) == 0.:\n",
    "    print(k,x_test[k])\n",
    "    x_test = np.delete(x_test, k, axis=0)\n",
    "    y_test = np.delete(y_test, k, axis=0)\n",
    "    k -= 1\n",
    "  k+=1\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a392ee4",
   "metadata": {},
   "source": [
    "Ahora se aplicara una renormalización para poder interpretar un vector de estado para pdoer aplicarse en nuestro modelo de circuito cuántico, siguiendo el critero:\n",
    "\n",
    "$ \\frac{vector-de-entrada}{\\sqrt{\\sum_{i=0}^{n-1} (vector-de-entrada_i)^2}}$,\n",
    "\n",
    "donde vector-de-entrada  es  el vector de 64x1 que representa una imagen ya sea del conjunto de entrenamiento o de prueba con la finalidad de convertirse en un vector de estado $| \\psi \\rangle$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "358db9e9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "la suma de los estados de la imagen del conjunto de entrenamiento 0 1.0\n"
     ]
    }
   ],
   "source": [
    "import cmath\n",
    "def Normalize(row):\n",
    "  suma = np.sqrt(np.sum(row**2))\n",
    "  if suma == 0.:\n",
    "    return 0.0\n",
    "  row = row/suma\n",
    "  return row\n",
    "\n",
    "\n",
    "for i in range(len(x_train)):\n",
    "    x_train[i] = Normalize(x_train[i])\n",
    "    \n",
    "for i in range(len(x_test)):\n",
    "    x_test[i] = Normalize(x_test[i])\n",
    "    \n",
    "print(\"la suma de los estados de la imagen del conjunto de entrenamiento 0\",np.sum(x_train[0]**2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "afb03ff4",
   "metadata": {},
   "source": [
    "# Análisis y Diseño del Autoencoder\n",
    "\n",
    "Para este trabajo se diseñará los circuitos necesarios para indicar el autoencoder en su versión cuántica.\n",
    "\n",
    "Para este proceso se consideran 6 qubits que recordando equivalen al vector de estado 64x1, donde usaremos el mapeo por amplitud, y se ocupo por cuestiones de tiempo y recursos 1 sola capa por cada 1.\n",
    "\n",
    "Más adelante se indicaran el circuito que se uso para esta aplicación pero requiere de 10 paramétros."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "712a2f34",
   "metadata": {},
   "outputs": [],
   "source": [
    "n=6\n",
    "num_layers = 1\n",
    "params = np.random.random(10*(num_layers))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4fb58bbf",
   "metadata": {},
   "source": [
    "Se valida que el vector de entrada con indice 0 este normalizado a un vector de estado"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "df09cf0d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.        , 0.30173661, 0.30046881, 0.        ,\n",
       "       0.        , 0.        , 0.        , 0.        , 0.31948582,\n",
       "       0.32075362, 0.23961437, 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.22566856, 0.02408822, 0.        , 0.        ,\n",
       "       0.24722117, 0.        , 0.        , 0.        , 0.31948582,\n",
       "       0.        , 0.        , 0.        , 0.18763453, 0.        ,\n",
       "       0.        , 0.        , 0.2852552 , 0.        , 0.1445293 ,\n",
       "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "       0.31948582, 0.31948582, 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "       0.        , 0.        , 0.        , 0.        ])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98ebe795",
   "metadata": {},
   "source": [
    "Al ingresar el vector de estado como entrada a una función que genera un circuito cuántico de 6 qubits que usando el método \n",
    "`initialize(vectorstate,qubits)` que genera la representación en términos de qubits. \n",
    "\n",
    "Una observación de este método es que se puede variar el costo computacional por el tipo de método que se genere, y más si predominan estados con estados 0 de amplitud, generando que los equipos de computo de nuestro equipo fue limitado."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "ef1de1c2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 447.797x445.48 with 1 Axes>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def input_data(n,inputs):\n",
    "    circuit = QuantumCircuit(n,1)\n",
    "    circuit.initialize(inputs,range(0,n,1))\n",
    "    circuit.barrier()\n",
    "    return circuit\n",
    "input_data(n,x_train[0]).draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9460dabe",
   "metadata": {},
   "source": [
    "En este proceso generamos el circuito variacional cuántico que representa el autoencoder cuántico, consideramos de apoyo el circuito cuántico propuesto en [1], pero algunos problemas daban por el costo computacional por lo que consideramos [2],[3],[4],[5] para generar a partir de varias propuestas de redes tensoriales como peude ser MERA, para nuestro trabajo y con ciertas modificaciones para dejarlo en 6 qubits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "d2d3ad7c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def vqc(n, num_layers,params):\n",
    "    #Set the number of layers and qubits\n",
    "    #ParameterVectors are initialized with a string identifier and an integer specifying the vector length\n",
    "    \n",
    "    \n",
    "    parameters = ParameterVector('θ', 10*(num_layers))\n",
    "    len_p = len(parameters)\n",
    "    circuit = QuantumCircuit(n, 1)\n",
    "    \n",
    "\n",
    "    \n",
    "    for layer in range(num_layers):\n",
    "        for i in range(n):\n",
    "            circuit.ry(parameters[(layer)+i], i)\n",
    "        circuit.barrier()\n",
    "\n",
    "        circuit.cx(2,0)\n",
    "        circuit.cx(3,1)\n",
    "        circuit.cx(5,4)\n",
    "        circuit.barrier()\n",
    "        \n",
    "        circuit.ry(parameters[6+(layer)],0)\n",
    "        circuit.ry(parameters[7+(layer)],1)\n",
    "        circuit.ry(parameters[8+(layer)],4)\n",
    "        circuit.barrier()\n",
    "        \n",
    "        circuit.cx(4,1)\n",
    "        circuit.barrier()\n",
    "\n",
    "        circuit.ry(parameters[9+(layer)], 1)\n",
    "        circuit.barrier()\n",
    "        \n",
    "\n",
    "    \n",
    "    params_dict = {}\n",
    "    i = 0\n",
    "    for p in parameters:\n",
    "        params_dict[p] = params[i]\n",
    "        i += 1\n",
    "    #Assign parameters using the assign_parameters method\n",
    "    circuit = circuit.assign_parameters(parameters = params_dict)\n",
    "    return circuit"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce042567",
   "metadata": {},
   "source": [
    "El circuito de nuestra red tensorial se ve afectada por 10 compuertas $Ry(\\theta)$ y 4 $C_{not}$, considerando como costo ligado al número de $C_{not}$ sería de 4."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "1adf27e3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1110x445.48 with 1 Axes>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vqc(n,num_layers,params).draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76029eab",
   "metadata": {},
   "source": [
    "Considerando [6] se tiene la oportunidad de usar el Swap-test para buscar el valor de y se identifico de [7] la forma de trabajo de la Swap-Test se comparan dos estados $| \\psi \\rangle$  y $| \\phi \\rangle$,  donde el primero es el vector de referencia $| 0 \\rangle$  y el segundo los qubits que se quieren eliminar para disminuir variables, estos son $| \\phi_0 \\phi_1 \\rangle$ donde al medirse el qubit que tiene la Hadamard del Swap-Test y este se acerque más al estado $|0 \\rangle$ significa que se disminuyó de manera correcta la informacion en $| \\phi_0 \\phi_1 \\rangle$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "5d9848ca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 387.597x325.08 with 1 Axes>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def swap_test(n):\n",
    "    qubits_values = 2*n+1\n",
    "    qc = QuantumCircuit(qubits_values)\n",
    "    qc.h(0)\n",
    "    for i in range(n):\n",
    "        qc.cswap(0,i+1,2*n-i)\n",
    "    qc.h(0)\n",
    "    qc.barrier()\n",
    "    return qc\n",
    "swap_test(2).draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad207188",
   "metadata": {},
   "source": [
    "El siguiente proceso indica el circuito cuántico variacional dle autoencoder para generar la disminución de dos qubits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "fb790b64",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABMAAAAPbCAYAAAC6/PdAAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAACt2klEQVR4nOzde3zP9f//8ft7BzvYZmZGRjkMzRxickhspBIVCSUp+ZU+TIrk8/mk8qm+pY/Ep/pEp8+ndPBRJqeSTmwUClFNy5zXnIY5jY0d3r8/Ftl57P1+v17v1/t2vVx2Ya/38/16PfZ09zSPvQ42u91uFwAAAAAAAGBRXkYXAAAAAAAAADgTDTAAAAAAAABYGg0wAAAAAAAAWBoNMAAAAAAAAFgaDTAAAAAAAABYGg0wAAAAAAAAWBoNMAAAAAAAAFgaDTAAAAAAAABYGg0wAAAAAAAAWBoNMAAAAAAAAFgaDTAAAAAAAABYGg0wAAAAAAAAWBoNMAAAAAAAAFgaDTAAAAAAAABYGg0wAAAAAAAAWBoNMAAAAAAAAFgaDTAAAAAAAABYGg0wAAAAAAAAWBoNMAAAAAAAAFgaDTAAAAAAAABYGg0wAAAAAAAAWBoNMAAAAAAAAFgaDTAAAAAAAABYGg0wAAAAAAAAWBoNMAAAAAAAAFgaDTAAAAAAAABYGg0wAAAAAAAAWJqP0QUA53yyQdp71JhjR9aWBnY05tgAAAAAAMC5aIDBNPYelXZkGl0FAAAAAACwGi6BBAAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAIClmboBVlhYqOnTp6t58+by9/dXu3btlJycrJYtW2rUqFFGlweDvXy3TYd/Tym1/Y2/hCvj1yTXFwQAAAAAQDXZ7VJWupT6lfTLUiltpZR9yOiqKnfmlLRrnfTLp1LKMmn/Fqkg3+iq/uRjdAEVGTlypBYuXKgnn3xSsbGxWrNmjYYOHapDhw5pwoQJRpcHAAAAAADgMGdPS5sXSif2S7L9sdEupW+U6kdLrW6UvEzYyUn/UdqWJNkLdb7uA79KaUnSVQOlWpcZWNwfTHsG2Ny5czVnzhwtWbJEEydOVM+ePTV58mR17dpV+fn5io2NlSQdPHhQN9xwgwIDA9WuXTtt2rTJ4MoBAAAAAAAuTmGhtGnBH80vSbL/8fGHA6lFZ4WZzf5fpbQVfzS/pGJ15+VKP86XTh8zqLgLmLBvWGTq1Knq06eP4uLiim2PioqSr6+v2rRpI0kaPXq0rrzySi1evFjvv/++Bg0apLS0NHl7extRNlxs/tPXSF7F+7hnc04YVA0AAAAAAJfm8E7p5MGKx+zfIjXpKgWGuqSkStnt0o7vKhogFZyVft8otbzOZWWVyZQNsIyMDKWkpGj8+PGlXktPT1dMTIz8/Px08uRJffbZZ9q7d68CAgI0atQoPf/881q3bp26detmQOVwtcFT1ii8Ueti2974S7hB1QAAAAAAcGn2b1HR5YP2iscdSJWadnVFRZU7vl/KPV75uH1bpBa9JJut8rHOYtoGmCTVr1+/2PacnBwlJyerb9++kqRt27apTp06Cg//s+HRpk0b/frrr5U2wPr166dly5Y5uHJUx+2TV6phdLwhx05OTtK4G3oacmwAAAAAAGaMWaU2TbpXOKagMF8z/jlbry0e56KqKtY15lY9M2JxpeMKzkp+vn7KKzjrsGO3bdtWP/30U5XHm/IeYOcaWmlpacW2T5s2Tfv371eHDh0kSadOnVJISEixMSEhIcrOznZNoQAAAAAAAA5wPPuQCgorfmyil81bx0+Z55GQx6v4eMqcM9kObX5Jf548VVWmPAOsadOmatu2rZ5//nmFhYUpMjJSiYmJ58/YOncD/Jo1a+rkyZPF3nvixAkFBQVVeozPPvvM8YWjWl79StqRacyx4+Lilfh/lZxnCgAAAACAk2SmST8vqXiMzWbTG/Of0fuhz7imqErY7dJ3b0m5ldyKu3nnINntjvs/d3x8/EW/x5RngHl5eWn+/PmKiYnR6NGjdd999yk8PFwJCQny8fFR27ZtJUnNmzfX4cOHdeTIkfPvTUlJUatWrYwqHQAAAAAA4KKFR0lBdVV0H7By1G9lnhvgS0X39Gpa0R2obJK3r3R5rMtKKpcpzwCTpBYtWmjlypXFtg0fPlzR0dEKCAiQJAUHB6tfv3569tln9cILL+iDDz6QzWZTly5djCgZLvbwB2V3jx98/bCLKwEAAAAAoHq8vKT2g6TNn5R4GuQfN8aPaCFF32BUdeVrECPl5UjbklXqBv6+flK726TA2oaUVoxpG2Bl2bBhQ6nm1uzZs3X33Xerdu3aat68uRYsWCBvb2+DKgQAAAAAALg0fjWlTndLWenSpvlF2xq1ly6LkULqGVtbRa7oKNW/Utr7i7Tzu6JtrfpI9VoWnQFmBm7TAMvOzlZaWprGjBlTbHu9evX01VdfGVQVAAAAAACA49hsUp0r/vy8ZS/jarkYfkFS065/NsAatDa2npLcpgEWFBSkgoICo8sAAAAAAACAmzHlTfABAAAAAAAAR6EBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS/MxugDgnMjannlsAAAAAADgXDTAYBoDOxpdAQAAAAAAsCIugQQAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAAAwyNEMac1/pP1bKh9rt0s7vpPW/lfKPen82qzE1A2wwsJCTZ8+Xc2bN5e/v7/atWun5ORktWzZUqNGjTK6PAAAAAAAgGrJPyOdPipt+Vzal1L+OLtd2pYs7VornT4mFRa4rERLMHUDbOTIkXr22Wf14IMP6vPPP9eQIUM0dOhQ7dy5U7GxsUaXBwAAAAAAUC11m0lR3Yt+/+vysptg55pf6Rskm5fU5lYpMNSlZbo90zbA5s6dqzlz5mjJkiWaOHGievbsqcmTJ6tr167Kz88/3wCbMmWKWrVqJS8vLyUmJhpcNQAAAAAAwMVp3Ln8JlhZza+IKGPqdGembYBNnTpVffr0UVxcXLHtUVFR8vX1VZs2bSRJzZs318svv6xOnToZUSYAAAAAAEC1lWyCnUPzyzFM2QDLyMhQSkqKBg8eXOq19PR0xcTEyM/PT5J099136/rrr5e/v7+rywQAAAAAAHCYC5tg59D8cgwfowsoS0ZGhiSpfv36xbbn5OQoOTlZffv2rfYx+vXrp2XLllV7PwAAAAAAAI50R8+/6v6+L0iS8gvy9cy7t2vto0sMrqpqvnrRLkmy2WxOPU5YWNhFjTflGWDh4eGSpLS0tGLbp02bpv3796tDhw5GlAUAAAAAAOB0tWqGn/+9l5eXggNqG1iNNZjyDLCmTZuqbdu2ev755xUWFqbIyEglJiaeP2PLEU+A/Oyzz6q9DwAAAAAAAEe58Ib353jZvPTYHe/qnXfeVYPWxtVWVV9PL/rVbrc77Rjx8fEX/R5TngHm5eWl+fPnKyYmRqNHj9Z9992n8PBwJSQkyMfHR23btjW6RAAAAAAAAIcp+bTHc8p7OiQujinPAJOkFi1aaOXKlcW2DR8+XNHR0QoICDi/LS8vTwUFBSosLFReXp5yc3Pl5+fn9GtNAQAAAAAAHKFk86vNrdLPi4pea9y56Nftq/98OqQ7nAlmNqY8A6w8GzZsKHX54wMPPKCAgACtXr1ad911lwICArRnzx6DKgQAAAAAAKi6sppfJZ/2eOHTITkT7NK4TQMsOztbaWlppW6A/+6778putxf7aNy4sTFFAgAAAAAAXITMrRU3v84p2QQ7dcR1NVqBaS+BLCkoKEgFBQVGlwEAAAAAAOAwAaFSaKTUuJMU3qzisY07S96+0sE0yTeg4rEozm0aYAAAAAAAAFYTUl/qOLTq4xt1KPrAxXGbSyABAAAAAACAS0EDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAF+XwLuno71Ubay+U9m+RTmU5t6aKmLoBVlhYqOnTp6t58+by9/dXu3btlJycrJYtW2rUqFFGlwcAAAAAAOBx7Hbpl6XSxo+kfSmVjC2UtiyXtnwu7fjWNfWVxdQNsJEjR+rZZ5/Vgw8+qM8//1xDhgzR0KFDtXPnTsXGxhpdHgAAAAAAgMex2aQmXYp+/+vy8ptg55pfB36VvH2lyzu6rsaSTNsAmzt3rubMmaMlS5Zo4sSJ6tmzpyZPnqyuXbsqPz9fsbGxOnPmjEaMGKHIyEiFhoaqV69eSk1NNbp0AAAAAAAAS2vcSWrWvej3ZTXBSja/2t8uhTZwfZ3nmLYBNnXqVPXp00dxcXHFtkdFRcnX11dt2rRRfn6+oqKi9MMPP+jIkSPq3bu37rjjDoMqBgAAAAAA8BxNOpffBCvV/GpoTI3nmLIBlpGRoZSUFA0ePLjUa+np6YqJiZGfn59q1qypJ554QpGRkfL29tZDDz2kX375Rbm5uQZUDQAAAAAA4FlKNsHOMVPzS5J8jC6gLBkZGZKk+vXrF9uek5Oj5ORk9e3bt8z3rVmzRo0bN5a/v3+lx+jXr5+WLVtW/WIBAAAAAAAc7KsX7ZIkm81mcCVVc2fPv+n/9Z16/vOcM9n6+2t9tOWR75xyvLCwsIsab8ozwMLDwyVJaWlpxbZPmzZN+/fvV4cOHUq95+jRo0pISNBzzz3nkhoBAAAAAABQ5OOkadq+d9P5zz9Z/S9t2e2c5telsNntdrvRRZRUWFio9u3ba//+/Zo+fboiIyOVmJioZcuWKT09XevWrVPnzp3Pj8/JydENN9ygHj160AADAAAAAABu7+vpRb/2nmhsHVVx4Q3vL9Sqj9SgteOPFx8fL0lKSkqq8ntMeQaYl5eX5s+fr5iYGI0ePVr33XefwsPDlZCQIB8fH7Vt2/b82Pz8fA0ZMkTNmzen+QUAAAAAAOBCJZ/2eKGyng5pFFPeA0ySWrRooZUrVxbbNnz4cEVHRysgIOD8tvvvv1+FhYV68803XV0iAAAAAACAxyrZ/Gp/u7RhXtFrzbpLO1b/eWN8Z5wJdjFM2wAry4YNG9SlS5fzn+/Zs0dz5syRv7+/QkNDz2//9ddfdfnllxtQIQAAAAAAgPWV1fy68GmPTf64c5VZmmBu0wDLzs5WWlqaxowZc37bFVdcIRPewgwAAAAAAMDSfvu6/ObXOSWbYN6+Ur2Wrq3zHLdpgAUFBamgoMDoMgAAAAAAADya3S4dTKu4+XXOhU2wzG00wAAAAAAAAOAGbDap0zDJy0fyD658fJPOUngTyT/E+bWVhwYYAAAAAAAALkpg7YsbHxzhnDqqysvYwwMAAAAAAADORQMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAlkYDDAAAAAAAAJZGAwwAAAAAAACWRgMMAAAAAAAAluZjdAEAUF1ZWVnKz8+vcMyLL76oxx57rNzXfXx8FBYW5ujSgEtGrlGZyjJCPuBuWPdch7mGFZFrVMZmt9vtRhcBANWRmZmprKysCsdER0crNTW13NfDwsIUERHh6NKAS0auUZnKMkI+4G5Y91yHuYYVWTHXX08v+rX3RGPruFiuqDs+Pl6SlJSUVOX3cAkkAAAAAAAALI0GGAAAAAAAACyNBhgAj5CYmGh0CYDDkWtUhHzAisi16zDXsCJy7dlogAEAAAAAAMDSaIAB8AiDBg0yugTA4cg1KkI+YEXk2nWYa1gRufZsPkYXYCZbV0gnM42uAvAswRFSy15GV2FerEvORwYBc2Hdcz7WPQCAJ6IBdoGTmdKxDKOrAIA/sS4B8DSsewAAwBm4BBKAR0hISDC6BMDhyDUqQj5gReTadZhrWBG59mw0wAB4hLFjxxpdAuBw5BoVIR+wInLtOsw1rIhcezYaYAA8Qo8ePYwuAXA4d8n16TPS1v1S8m/SVylSUqqUuk/KzjW6Mmtzl3wAF4Ncuw5zDSsi156Ne4AB8AiHDh0yugTLeXR2vFL3rJW3t6+8vLxVv3YT3XXdZMW1G2x0aR7DzLm224uaXN+mFf1qL2dc83rStS2kNo0kL5tLS7Q8M+fDXbHuGY9cuw5zDSsi156NBhgA4JIN6/2khvV+QgUF+Vq85t+aOvcuRUW2V2R4lNGlwUDHT0sf/yBt2Vv52G0Hiz6aRUhDu0jhwc6vD6gO1j0AANwTl0AC8AitWrUyugRL8/b20U2dH1BBYb527NtsdDkew4y5/j1Lmrasas2vC+3IlF5cVtQMg2OYMR9WwrpnDHLtOsw1rIhcezYaYAA8woIFC4wuwdLy8s/q0zWzJUkNw1sYXI3nMFuuDx6XZn8jnTpzae8/ky+9uVLac9ixdXkqs+XDalj3jEGuXYe5hhWRa8ey24s+Lma8kWiAAfAITz31lNElWNLcb57TgCdDdfPjAXrniyc0YfDbatqgrSTpuQ+H6ruURefHTnl3gDZtX2FQpdZkplwXFEofrJFOn6143APxRR/lySuQ3l8jnc13ZHWeyUz5sBLWPWORa9dhrmFF5Npx7Hbphw+k9R9KZ09VPv7o79Lq2dL21c6vrTymboAVFhZq+vTpat68ufz9/dWuXTslJyerZcuWGjVqlMvq2L53k8a+0km3PF5TD864Sr/uWeeyYwNwjPnz5xtdgiXddd1kLXr2mBL/cVidruyrzRf8R29M/5f13pdTdDr3pFb/8omCA8PUPqqXgdVaj5lyvWpr0eWPlYmJLPqoyOGT0he/OKYuT2amfFgJ656xyLXrMNewInLtWDZJJw5IGz+uuAl29Hdp0wLp7GnJZuBDj0zdABs5cqSeffZZPfjgg/r88881ZMgQDR06VDt37lRsbKxLajibl6sn37lFN3a8TwufPabbrh2np98bqLz8Sn7EDQAeJDiwtiYMfls//LZMa1IWS5JqB0VoYPfxem3xOM395v806ubpBlcJZykslJJ/c+w+v9tWdEkkYFasewAAT2azSe0GSjXrSKeOlN8EO9f8KsyXLouRml7j+lrPMW0DbO7cuZozZ46WLFmiiRMnqmfPnpo8ebK6du2q/Pz88w2wYcOGqV69eqpVq5Y6deqktWvXOrSOzTtWymbz0i3XjJaPt6/6dBqpkMA6WvfrUoceBwDcXUhgmG7vPkH/Xf64CgsLJUk3Xj1Cew9v04Bu4xQSGGZwhXCWrQekY6cdu8/cPOmndMfuE3A01j0AgCfzqyl1GFJ+E6xk86vVjZLNwC6UaRtgU6dOVZ8+fRQXF1dse1RUlHx9fdWmTRtJ0uTJk/X777/r+PHjmjJlim6//XaH1vF75m+6IqL4kyIa1W2p3Qe3OPQ4AJwrOTnZ6BI8wm3dH1bWif36auN757c1qBOlyPAoA6uyLrPkememc/a765Bz9uspzJIPq2Pdcy1y7TrMNayIXDteWU2wc8zU/JIkH2MPX7aMjAylpKRo/PjxpV5LT09XTEyM/Pz8JP35GFO73S5fX18dOHBAubm58vf3d0gtuWdPya9GYLFtfr6BOpPn4B91A3CqLVu2KCIiwugyLOWl0UmlttX0D9Enz1ThRlBwCLPkOuOok/ZLlKrFLPmwEtY945Fr12GuYUXk2jnONcF+/LioCXaOmZpfkokbYJJUv379YttzcnKUnJysvn37Fts+bNgwLViwQGfOnFFCQkKVml/9+vXTsmXLim2b/peVatcsvtg2P99Anc3LKbbtTN5pBdQIquqXA6ACyclJunpoz0t+/4QJE9SjR49Kf5ozc+bMMpvq58TFxWnVqlWaMWPGJdfiDGWtS3Cs6mbQGdwt13f8Y53qR3Uutu2B+Ipvdv+vYcU/37JXeiupxLatu2Xr28QhNVpNVTJilnxcLNY952Pd82zMNazIqrn+6kW7JMlm5J3jL0JoUIRefWid6ocVff+WtHmepk4apkJ7oVOOFxZ2cbcaMEEPrrTw8HBJUlpaWrHt06ZN0/79+9WhQ4di2z/88EOdPHlSixYtUpcuXRxay+X1ovX7oa3Ftv1+aKsur9eqnHcAAC406c531brJtUaXASeyO+mbmsLCAqfsF3A21j0AgCdqVLelagX9eYZd4/qtFRJYx8CKirPZ7Xa70UWUVFhYqPbt22v//v2aPn26IiMjlZiYqGXLlik9PV3r1q1T586dy3xvu3btNH/+fLVo0eKij7thnnQso/i2s/lndM/Uprqz5990c9e/6OuN7+s/y/6m9x7fpYAaNS/lywNwgdCGUsc7q7ePzMxMZWVVfPlJdHS0UlNTy309LCzMlKdDl7UuwbEckUFncKdcv7ta2lzFG9afO/PrkQ8rH9ssQnro+kuvy+oqy4hZ8nGxWPecj3UPzDWsyIq5/vqPhwn3nmhsHVVx4Q3vI1pKpw4XXQ5Zs44UO0RydPskPj5ekpSUlFTl95jyDDAvLy/Nnz9fMTExGj16tO677z6Fh4crISFBPj4+atu2bbnvPXv2rHbv3u2wWmr4+On5//e5Vm7+nwZOCdOi717V0/ctofkFuJmnn37a6BIAhzNLrhs56UF3l5vnB4ZuySz5AByJXLsOcw0rItfOUfJpj236Vfx0SKOY8h5gktSiRQutXLmy2Lbhw4crOjpaAQEBkqQjR45oxYoV6tevn3x8fPTWW29p3759pS6RrK6mDdrq5bFrHLpPAK41ZMgQo0sAHM4suW55mbR0s3P2i0tnlnwAjkSuXYe5hhWRa8cr2fw6d8P7kjfG3/ixc84EuximPAOsPBs2bFBsbGyxba+++qouu+wy1atXT/PmzdNnn312/h5iAHBOdHS00SUADmeWXDcMk65w8D+94cFSi/qVj0P5zJIPwJHItesw17Aicu1Y5TW/zjnXBDPLmWBu0wDLzs5WWlpasbO76tSpo1WrVun48eM6evSoVq9erWuv5YajAAC42o2tHbu/G1pLXu7xwCMAAACPY7dLvywtv/l1TskmWFrFD+p0KtNeAllSUFCQCgp4GhQAXOjw8b16Y+mjCg4M0xX1YtS/W4Ikaf1vy7Vmy2L5+vjpqmY9dU3r/jqVc1zTPx6pwfGPqdUVRU/MPZV7QuNe7aIXH1yhsJD6GvdqVzWs21KSlND/ZW3avkLf/rJAXl7eGtj9ETW9rJ3+vegh+dcI1MGje/TXO9/Tlxvm6KedSapVM1zXx96rlo06SpKWff+WtmX8qIdvn12qno4tb9SM+fcrODBMNXz89cDN0/Rj2tf6auN7ssuuW7qOVkzja0rV8+uetcX206XVLaXqSdn1ban9wPlaRUpXN5XW76x43Ja9VdhXA+nqJo6pC9ZU3tr35YY5Wrnpf6odXF9XRfVUeK2G+nrj+5KkH7d9pX+O+lpHju+tdK2pGVBL639brs9/eFtP3ZOoE6eOaGbiKNUOrqc6IQ00tNfj8vLyKraGhgZFaOaCUQr0C1Ze/hmNGzhLkortp7CwsNSa9dOOpErXx0/XvlFsnW0e2YG1DwBgKJtNatC6qOnV9Jqym1/nnGuCbUuWIqJcV2NJbnMGGABUx7mnhFjNp+ve0IBrx2ncwFn6/rfPlF+QJ0lasuY1BQeGKTvnmKIi20uS5n7znPxqBJ5/b2Fhof6z7O9qUKeZJOnQsQzlnj0lX58aahTRUjUDaslms2niHe+of7ex+jZloc7knVbST/N0/NRhedm8VMPXX6t+nq+/Df1Ao2+ZqXkrpkqSft65Sr9nbpVd9jLryc45pn5dHlTCgFeUcThNZ/PPKHH1DE0Y/LbGD3pTH638Z5n1lNxPWfWU3I+VmS3Xt3es/Mb1byUVfZQnIkS6q2vRN1WoHrPlw5HKW/t+2blK4bUiJdkVfXkXdWh+nSbd+a6ubX2b7uz1d11RL7pKa83m7Su198h25ZzJllS0prVp2kPjBs5SoH+INu9YWWoN/Wlnki4La6rRt85UrZp1tWX3mlL7KWvNqsr6WHKd9eS1z8q5NhvmGlZErh0rqofU7NqKm1/n+NWUWveVIlo4v67y0AAD4BFmz55tdAlOcfTkAdUNbSRJCg6orVO5xyVJuw+k6K7rJmv49U/pneVPSJIeuHna+f+oSdL7Xz2tfl1GKTiwqGPh5xugvw39QOMHvakTp45oy+416tZ6gLbsXqMZ8x9Q+6hestvtembEYj12xzuqH9ZEP+9cpaHXPa7pH92n979+RnkFZ3Qga7e+S1mo/t3Gnj9WyXrCQuqrTdPumpk4Sjabl2yySXa7fH1qyM83QGfzc8usp+R+yqqn5H6szGy59veVRve69Ht3XV5Heqi3FOTv2Lo8ldny4UjlrX03Xj1S4wbO1qibp+utzyZJkgoK8rVk7Szd2nVM0ZursNZcFdVTAy5YwzpF91Xm0T16bdE4pR9M1bHsg6XW0AtrqhvaUEdO7Cu1n7LWrKqsjyXXWU9e+6yca7NhrmFF5Nqz0QAD4BFGjx5tdAlOERF6uQ4fy5AknTydpSD/0KLtta9QDR9/hQSWfTrOsexDSk1fp6VrZys1fa0+Tpqmg0f36Gj2QUlSSM06yi84qx+3faO2TXto1iMbNfeb57Qx7Utt/X29JCks5DKdzj2hQ8d+12N3vKs74ifJx8tX32z6UCdPH9U7yycrZddq/bp7bal6Dh/fpz0HUzV+0Jtq1yxeP/y2TDV8/JWXf1Zn8nJUw8e/zHpK7qesekrux8rMmOuAGtJfehWdDVajijda8PGS+raTHr5BCg5wbn2exIz5cJTy1r4tu7+Tt5ePAv1Dzp+BuubXJYprN0ReXkXf9lZlrSnpxKnD6tiyjxIGvPLHZZCRpdbQiNDLdfh4UU2HjmWoTkiDUvspa82qyvpYcp315LXPyrk2G+YaVkSuPZvb3AMMAKojKSnJ6BKc4qZO9+v1pRP05YZ3dW2bgZq9dLwevPkl9e82VtM+uldeNm8N6Tmp1PtCg+rqhQe+kCRNmzdCQ+InqYavv+atfEHrty5XQUGe2jaN0+c//EfT5t0rm82m6zrcrQ7Ne+vFj0boQNYuncnL0a1dx+j71M/0z3nDlV+Qp3v7PKsm9Yvuhn4ga7fmrXxBrRp3LVVPDV9/vb50vEICw3U697iuj71HQQG1NWP+/covzNOw3k/qsjpNS9VTcj8RtRqVqifAL7jYfqzMrLn2skndW0odm0jrd0kbd0l7j0r5hX+O8faSLguV2l8udW7GWV/OYNZ8OEJ5a1+toLp6af7/k91eqKE9/y5J+m3POvXueM/5997W/ZFK15qSQoMi9PXGv+q7lIUK9A9Ru2ZxatesaNy5NbR2cD19vfF9vbb4YUkq8x5cZa2hZ/NzK10ff96RXGyd9eS1z8q5NhvmGlZErj2bzW63240uwiw2zJP++GEiABcJbSh1vLN6+8jMzFRWVlaFY6Kjo5Wamlru62FhYYqIiKheIU7AuuR8jsigM1gt1wWF0uGT0tRPiz6ffqfk421sTe6usoy4Uz4uxLrnfKx7YK5hRVbM9dfTi37tPdHYOszo3P3cLqapySWQAADA6by9pHq1/vyc5hcAAABciQYYAI9Q0U96AHdFrlER8gErIteuw1zDisi1Z6MBBsAjfPzxx0aXADgcuUZFyAesiFy7DnMNKyLXno2b4APwCFOmTNGQIUOMLuOiBZvnFgSW5c5z7K65hmu4az7c+e+ku3DnOXbXXLsj5hpWRK49Gw0wADCxlr2MrgAAXIt1DwAAOAOXQAIAAAAAAMDSaIAB8AizZs0yugTA4cg1KkI+YEXk2nWYa1gRufZsNMAAeISYmBijSwAcjlyjIuQDVkSuXYe5hhWRa89GAwyAR4iLizO6BMDhyDUqQj5gReTadZhrWBG59mw0wAAAAAAAAGBpPAUSgNsLCgpSQEBApeMiIyPLfc3b29uRJQHVRq5RmapkhHzAnbDuuQ5zDSsi16gMDTAAbi83N1f5+fkVjunatatycnLKfd3Hx0eBgYGOLg24ZOQalaksI+QD7oZ1z3WYa1gRuUZlaIABcHv5+fnKysqqcMzatWsrHBMWFubosoBqIdeoTGUZIR9wN6x7rsNcw4rINSrDPcAAAAAAAABgaTTAAAAAAAAAYGk0wAB4hMTERKNLAByOXKMi5ANWRK5dh7mGFZFrz0YDDAAAAAAAAJZGAwyARxg0aJDRJQAOR65REfIBKyLXrsNcw4rItWfjKZAAYGJbV0gnM42uwtqCI6SWvYyuAsA5rHuuwdoHq2INcT7WD7grGmAAYGInM6VjGUZXAQCuw7oHoDpYQwCUh0sgAXiEhIQEo0sAHI5coyLkA1ZErl2HuYYVkWvPRgMMgEcYO3as0SUADkeuXaOwsOjD3ZAPWJG75dpulwoKpUK70ZVcPHeba6AqyLVn4xJIAB6hR48eWrVqldFlWMqjs+OVumetvL195eXlrfq1m+iu6yYrrt1go0vzGOTaOU7kSN/vkLZnShlZ0qkzRdtr+kkNw6Tm9aTOTaXgAGPrrAz5cDzWPeO5Q64zsqQfdkp7Dkv7jkl5BUXbw4OK1pDWDaWrLpd8vA0ts1LuMNfuhjXEeOTas9EAA+ARDh06ZHQJljSs95Ma1vsJFRTka/Gaf2vq3LsUFdlekeFRRpfmEci1Y50+Ky3dVPQf14Iyzvg6dUbaur/o4/Ofi5pgt7SXAmq4vtaqIB/OwbpnLDPnev8xaf4P0s5ySjycXfSxOV1atFHq01a6prnkZXNpmVVm5rl2Z6whxiLXno1LIAEA1ebt7aObOj+ggsJ87di32ehygIu2M1P656fS2u1lN79KKiiU1myX/vmZtIvvpT0S6x7OsdulpFRp+uflN79Kyj4jJa6XZn8jZec6tz6YE2sI4Ho0wAB4hFatWhldgqXl5Z/Vp2tmS5IahrcwuBrPQa4dY9sBadYK6XjOxb/32Glp1jfS9oOOr6u6yIdzse4Zw4y5/vxnadGPVWuel7TtoPTvr83ZBDPjXFsJa4gxyLVnowEGwCMsWLDA6BIsae43z2nAk6G6+fEAvfPFE5ow+G01bdBWkvTch0P1Xcqi82OnvDtAm7avMKhSayLX1Xf0lPSfVVJ+QfljHogv+ihPXoH0n+SiZpiZkA/nYN0zltlyvXG39GVKxWMqW0MOHJfmfGu+G+Wbba6tgjXEWGbN9ZlT0s41Uu6Jqo0/cVDatU4qzHduXVZj6gZYYWGhpk+frubNm8vf31/t2rVTcnKyWrZsqVGjRrm8nt/Sf9AdzzZw+XEBVN9TTz1ldAmWdNd1k7Xo2WNK/MdhdbqyrzZf8E3amP4v670vp+h07kmt/uUTBQeGqX1ULwOrtR5yXT12u/TR91JuXsXjYiKLPiqSkyd9/H3RPs2CfDgH656xzJTrkznSgvWVj6vKGrLtoLRmm2PqchQzzbWVsIYYy6y5PryjqAG24X/S6WMVjz36u7RxnrTjW+kkt2G4KKZugI0cOVLPPvusHnzwQX3++ecaMmSIhg4dqp07dyo2NtZlddjtdi3/4b/621s3KD//rMuOC8Bx5s+fb3QJlhYcWFsTBr+tH35bpjUpiyVJtYMiNLD7eL22eJzmfvN/GnXzdIOrtB5yXT3bM6Xf9jtuf7/uq/r9f1yBfDgX654xzJTrFalFD89wlM9/rvhsVFcz01xbEWuIMcya63pXSrUaSLknpR8/Kr8JdvR3afMnUkGedFkrKaS+S8t0e6ZtgM2dO1dz5szRkiVLNHHiRPXs2VOTJ09W165dlZ+fX6oB9tFHH8lmsykxMdHxtax4Xgu/fVl3XTfZ4fsGAKsICQzT7d0n6L/LH1dhYdGNUG68eoT2Ht6mAd3GKSQwzOAKgeK+TXOPfcK8WPc819l86fsdjt3nqTNFT4iE52ANwTk+NaT2t1fcBCvZ/GrVR7KZ9CmyZmXaBtjUqVPVp08fxcXFFdseFRUlX19ftWnT5vy206dP67nnnlNMTIxTaulz9Ui9Pn6zWja62in7BwCruK37w8o6sV9fbXzv/LYGdaJ4tDdMp6BQ+nWv4/e7JUMqvIQbYcN9se55pl2HHHv21zm/ZDh+nzA31hCc4+NXugl2TpnNL9N2c8zLx+gCypKRkaGUlBSNHz++1Gvp6emKiYmRn5/f+W1Tp07ViBEjtGTJEqfUUyfkMqfsF4DrJCcnG12C5bw0OqnUtpr+IfrkmSzXF+OhyPWlO3i86Ob1jna2QMo8KdWv5fh9Xyzy4Xise8YzS65/d9IfeYaJomSWubYS1hDjmT3X55pgmxZIx/f9uZ3ml2OYtgEmSfXrF7+gNScnR8nJyerbt+/5bbt379aSJUu0YcOGi2qA9evXT8uWLSu2bfpfVqpds/hLLxzARUtOTtLVQ3te8vsnTJigHj16VPqP2Y4dO9SsWbNyX4+Li9OqVas0Y8aMS67FGViXnK+6GXQGK+f64Q+K7hJvM/ic/WYdb9PNj3xSbNsD8RXfqPpfw0pv27JXeiup+LZuvfpr54/O+aHcOVXJiDvmQ2LdcxWzrX3utu71vv9txcT/v2LbqruG2O12Hcm2ycvbR/ZC590MzN3m+mKxhjif2dYPyXq5DvQL1vP3f66Yxt0kFTW/vtwwRy9NGqlCO6eanxMWdnGXCZuybxgeHi5JSksrfiONadOmaf/+/erQocP5bRMmTNCzzz4rX19fl9YIwL046wxRVG7Sne+qdZNrjS7Dksj1pfPydt7PAL28zfE9CfkwDuue85gl1874e37uBwOsIWANcR53yfXpMyf1wVfPyv7H46VzzmTrw6+fpflVTTa73UwP7C5SWFio9u3ba//+/Zo+fboiIyOVmJioZcuWKT09XevWrVPnzp21YsUKPf/88/r6668lSfHx8Ro7dqwGDRp0ScfdME86VsF19z/tSNIz7w3SgqcPX9L+AZQW2lDqeGf19pGZmamsrIpPHY+OjlZqamq5r4eFhSkiIqJ6hThBZesSqs8RGXQGq+b6kQ+Lfi3rTAhX+nWv9GZS1caeq/Vc7ZX5Sy/pShfcPaGyjLhjPiTWPVcx49rnTute4vqqP/TiYtYQby/pxTslLyefJOtOc32xWEOcz4zrh2StXF94zy/fACkvR/IPljrcIQWGGl2dOcTHx0uSkpKSqvweU54B5uXlpfnz5ysmJkajR4/Wfffdp/DwcCUkJMjHx0dt27aVJK1atUrr1q1TeHi4wsPD9d1332nkyJGaOnWqwV8BAAAws8jaTtx3qPP2DcAcnLWGNAh1fvMLgLmVvOH9NSMrfjokqs6U9wCTpBYtWmjlypXFtg0fPlzR0dEKCAiQVHT54/3333/+9cGDB+u+++7TnXc6px3drlk8Z38Bburpp582ugTA4cj1pasVKIUGSsdOO3a/tWtKwQGO3eelIh+wIrPk+oo6TtpvuHP2eynMMteAI5k91+U97fHCG+P/+BFngl0qU54BVp4NGzYoNjb2/OchISFq2LDh+Q8/Pz+FhYUpJCTEwCoBmNGQIUOMLgFwOHJdPZ2ausc+LxX5gBWZJdeXhUoNnXAWGGsI4FxmznV5zS/pz6dDciZY9bhNAyw7O1tpaWnFboBfUlJS0iXf/wuAtUVHRxtdAuBw5Lp6ukYV3W/HUXy8ivZpFuQDVmSWXNtsUveWjt3nFeHS5U46s+xSmGWuAUcya66zD5Xf/DqnrCZYXq4x9bor014CWVJQUJAKCpz3OGAAcFeHj+/VG0sfVXBgmK6oF6P+3RLOv3Yq94TGvdpFLz64QmEh9XUq57imfzxSg+MfU6sruujTtW/op51JqlUzXNfH3qvmkR3070UPyb9GoA4e3aO/3vmevtwwp9SYmQtGKdAvWHn5ZzRu4Cz9e9FDOp17UkdO7FXt4PqadMecUmPW/7Zca7Yslq+Pn65q1lNdWt1SakzJei4La6KZiaNUO7ie6oQ00NBej8vLy0vrf1uuz394W0/dkyip6LHQW39fr4du+7eSNn+kH377XJK0Me0LvTVxi0ICL+4RyfAMtWtKfdpIn/1U8bgte6u2v5vaFl1WCecrb91bv/ULrd2yWAWF+Ypp3E29Owwvtc58+8tCffvLAnl5eWtg90dUN7SRXl2YoKCA2so9e0qPDv6Pftm5Sl9tfE922XVL19FqWLdFqTFfrH+n2HrVslHHUmtsybVxwqC3NGP+/QoODFMNH389cPO0Umtj6ybXllr3Nm77qtjXdUPHeyVJy75/S9syftTDt8+WVHrNh3Nd3VT6foe081DF46qyhnjZpMFXO6YuVK6i750q+h7jdO5JzV46XkH+oTp+6pAm3TlHkpR57HdNmNVdHzy+u8wx73/5tHYd+EX+NYIU126wrry8s15e8BfVCWmg8FoNdUfPSaWOlXMmWy8v+IuCa9ZRTf9aGnHjM/pywxzt2LdZJ04d0XUd7lbHljdIkv7z+eOq6ReiO3v9rUo1v//l09qftUv+NWpqcPxEBQfU1hufTlStmuEqKMjXg7dMd+UfBySdzPyj+RUjtbqxdPPrnHNNsE2fSMf3SrknJF9/19bqztzmDDAAQNk+XfeGBlw7TuMGztL3v32m/II8SUVP1P3Psr+rQZ1m58fO/eY5+dX483/oq36er78N/UCjb5mpeSum6kzeaSX9NE/HTx2Wl81LNXz9S435aWeSLgtrqtG3zlStmnW1ZfcajR3wqsYPelMhgXX00G2vlTlmyZrXFBwYpuycY4qKbF/mmJLH+nnnKrVp2kPjBs5SoH+INu9Yqc3bV2rvke3KOZMtSUr+6WNlnTyggsJ8SVL8VXdo0p3vKiqyvR4Z9CbNL1SoVyupad2Kx7yVVPRRkWYRUk9z/lDZkspb91J2rdaOfT9p3+Htahjeosx1xmazaeId76h/t7H6NmWhjp48qLt6TdYjt7+uoIBQHcjapcTVMzRh8NsaP+hNfbTyn2WOKbleSaXX2JJrY3bOMfXr8qASBryijMNpOpt/ptTaWNa6V/LrkqSfd67S75lbZVfRA93LWvPhXF42aWhXKaBGxeOqsob0bSc15J8rlylvDanse4ysk/u17telOpadqZDAotP1cs+e1txvnlO92o3LHbNlzxqFBtWTt7ePmkfG6uTpoxrW+0klDHhFP+1MKvNYyT99rE7RfZXQ/2VlHk3X4eN79WPaV/r90FYdPLr7/N/1lZvnKfePeqta8+YdKzXpznf1/256Xu99MUWff/+2Av2CdfzUYbVsRCfWCPVbSd0eKPvMr5J8/KSOd0jXPigFG//ASrdCAwyARzj3mFwrOnrygOqGNpIkBQfU1qnc45Kk9796Wv26jFJw4J/XUzxw87Ri/zkaet3jmv7RfXr/62eUV3BGdrtdz4xYrMfueEf1w5ro552rSo258Hh1QxvqyIl9kqQv1r+jrjH9VdM/pMwxuw+k6K7rJmv49U/pneVPlDmm5LE6RfdV5tE9em3ROKUfTNWx7IO6KqqnBnQbe/5riGs3RPHt7ig2JydPH9XPO5PVtdUtjp5uU7Fyrl3F20u6P756N55uHC7dHyd5mey7Kivno7x1r0Pz3pr24Nd6cvh8/W/l1DLXmW6tB2jL7jWaMf8BtY/qpcb1Y9S0QVt9n7pMNpuXGkW0lOx2+frUkJ9vgM7m55Y5puR6JZVeY6Xia2NYSH21adpdMxNHyWbzkk22UmtjWeteya/rQNZufZeyUP0vWAvLWvOtyGy5rhssje5VeROsIte1KvowG7PNtSOVt4ZU9j1GgF+wpt7/hf469D0dy87U/qxdenvZXzXsuidUw8e/zDEHsnZraK+/66Hb/q2B3R/RnC+fUsO6zdWwbgv9Y85A1Qu9osxjZZ08oLq1imoMrxWpIyf26/qO9+rZEUs0ftBbmrviOW39fYP2H9mpa9sMrHLNB7J269Zrxmj6xyO1eM1rOpN3WvuO7FCbJj00YdBbWr7+v8o96+AnxJiIWXNts0kBtYp+rdJ4L8k/2Lk1WZHJvlUDAOeYPXu20SU4TUTo5Tp8LEOSdPJ0loL8Q3Us+5BS09dp6drZSk1fq4+TppX53kPHftdjd7yrO+InycfLVxvTvtTW39dLksJCLtPp3BOlxkSEXq7DxzP+eH+G6oQ0kCR9n/qp4toN+bOmEmMial+hGj7+53/6WNaYksc6ceqwOrbso4QBr6h2cD1F/PFNYmW+WP+Obr0mofKBbs7KuXalwBrSmOuk7i0u/r09Wha9tzr/+XUWK+ejrHVPkt778h/y8fJVzYBQFRTml7nO/LjtG7Vt2kOzHtmoud88J0n64KtntffwNiX0f1mSVMPHX3n5Z3UmL+f8f2pLjim5XpXnwrXx8PF92nMwVeMHval2zeL1w2/LSq2NZa17Jb+ubzZ9qJOnj+qd5ZOVsmu1ftm5ukprvhWYMdeX15Em9Kn8bNKS/H2lu7pKt7Sv+n96XcmMc+0o5a0hlVmxaa4OZO2SJIUGRSgn96QOZu3Wh9/8n/Zk/qrE5Bmlx5w5qdT072Wz2RQSWEd5+WeVcWibjmcf0j/u/UR+vgHavndzmTUe+mP9Onx8r+qENND7Xz0tb28fhQXXV86ZbH254V0dPLpbiatm6Lsti5RxKK3yms9mK+fsKU0c8l/1an+XggPDFBZcX4H+IbLZbAoOqC2bGQPpIFbONSrnNvcAA4DqGD16tGX/wbup0/16fekEfbnhXV3bZqBmLx2vB29+SS888IUkadq8ERoSP6nM9wb5h+qf84YrvyBP9/Z5VhG1GunFj0boQNYuncnL0a1dx+j71M+KjWlcL0Zfb3xfry1+WJIU0/ga5Zw9pRq+AfL28v5jW7dSY/p3G6tpH90rL5u3hvScVOZ+jmcfKnas0KAIfb3xr/ouZaEC/UPUukm3Ks3Jb79/r35dRlVrXt2BlXPtan4+0u1XS+2vkFakSlsy9MeFZaXZJLVuWHT5ZJOL/A+vK1k5H+WtezdefZ+en3uX/HwDdXv38WWuRcu+f1vT5t0rm82m6zrcra82vKevf3xfVzbqpBfm3q27r39Kt3V/RDPm36/8wjwN6/1kmWNKrp9lKbk21vD11+tLxyskMFync4/r+th7VFCYX2xtLGvdK/l1xba4XpJ0IGu35q18QW2adtcLTStf863ArLmuGyyN7S39sEtatVXad7T8sf6+RU977NXK3PcNNOtcO0J5a4ivT8U/zejRdpBeXZign3aslK+Pn5o2aKtnRy6VJP39rT4aFDdBB4/uKTamyWVttObXJXp5wWjlnj2lYb2fkLeXt2YvGa/awfVUUJivK+qVPgWwe9vbNTNxlLbs/k6R4VEKr9VAnaP76Z//u0d2e6GG9npcUZFXSZJ+2pGk1D3r1LBu6Z/klKy5Sf3WStn1raZ/PFI5Z7I1+tZ/ycvmpVlLHta6X5eqZaNO8vMNqP4km5SVc43K2ex2e3nf33mcDfOkP34QAMBFQhtKHe+s3j4yMzOVlZVV4Zjo6GilpqaW+3pYWJgiIsx3ET3rkvM5IoPOYNVcP/Jh0a//GmZsHZU5dlramSn9niWt/GOKe0ZLjcKkphHm+E9rZRlxx3xIrHuuYsa1zyrrnt0u7T0q7TlS9OuabUXb+7Ytus9Xs3pFTXcjWWWuy8Ia4nxmXD8ka+capZ27nDUpKanK7+EMMAAAgBJCA6UOjYs+zjXA+ncwsiIA7sJmK2p0nbup/bkG2A1tjKsJAMA9wAAAAAAAAGBxNMAAeISKTnUG3BW5RkXIB6yIXLsOcw0rIteejUsgAXiEjz/+WEOGDDG6jIsWzC0InM6d59hdcw3XcNd8uPPfSXfirvPsrrl2R+461+6abXfiznPsrrmGY9AAA+ARpkyZ4pb/2LXsZXQFMDN3zTVcw13zwbqHirhrrt2Ru841awgq4q65hmNwCSQAAAAAAAAsjQYYAAAAAAAALI0GGACPMGvWLKNLAByOXKMi5ANWRK5dh7mGFZFrz0YDDIBHiImJMboEwOHINSpCPmBF5Np1mGtYEbn2bDTAAHiEuLg4o0sAHI5coyLkA1ZErl2HuYYVkWvPxlMgAbi9oKAgBQQEVDouMjKy3Ne8vb0dWRJQbeQalalKRsgH3Anrnusw17Aico3K0AAD4PZyc3OVn59f4ZiuXbsqJyen3Nd9fHwUGBjo6NKAS0auUZnKMkI+4G5Y91yHuYYVkWtUhgYYALeXn5+vrKysCsesXbu2wjFhYWGOLguoFnKNylSWEfIBd8O65zrMNayIXKMy3AMMAAAAAAAAlkYDDAAAAAAAAJZGAwyAR0hMTDS6BMDhyDUqQj5gReTadZhrWBG59mw0wAAAAAAAAGBpNMAAeIRBgwYZXQLgcOQaFSEfsCJy7TrMNayIXHs2ngIJACa2dYV0MtPoKqwtOEJq2cvoKgAAAAA4Ew0wADCxk5nSsQyjqwAAAAAA98YlkAA8QkJCgtElAA5HrlER8gErIteuw1zDisi1Z6MBBsAjjB071ugSAIcj16gI+YAVkWvXYa5hReTas9EAA+ARevToYXQJgMORa1SEfMCKyLXrMNewInLt2bgHGACPcOjQIaNLsJxHZ8crdc9aeXv7ysvLW/VrN9Fd101WXLvBRpfmMcg1KkI+YEXk2nWYa1gRufZsNMAAAJdsWO8nNaz3EyooyNfiNf/W1Ll3KSqyvSLDo4wuDQAAAADO4xJIAB6hVatWRpdgad7ePrqp8wMqKMzXjn2bjS7HY5BrVIR8wIrItesw17Aicu3ZaIAB8AgLFiwwugRLy8s/q0/XzJYkNQxvYXA1noNcoyLkA1ZErl2HuYYVkWvPRgMMgEd46qmnjC7BkuZ+85wGPBmqmx8P0DtfPKEJg99W0wZtJUnPfThU36UsOj92yrsDtGn7CoMqtSZyjYqQD1gRuXYd5hpWRK49m6kbYIWFhZo+fbqaN28uf39/tWvXTsnJyWrZsqVGjRrlsjpSdn2rh17prP5P1tI9U5vp07VvuOzYABxj/vz5RpdgSXddN1mLnj2mxH8cVqcr+2rzBQ2uMf1f1ntfTtHp3JNa/csnCg4MU/uoXgZWaz3ulOv8AumX3//8/Gy+cbV4CnfKB1BV5Np1mGtYEbn2bKa+Cf7IkSO1cOFCPfnkk4qNjdWaNWs0dOhQHTp0SBMmTHBJDSdPH9VT79yqhAGvqudVQ7Vj32b99c3ealCnmTq06O2SGgDA7IIDa2vC4Ld17wvNtCZlsa5p3V+1gyI0sPt4vbZ4nHbu/0n/HPW10WXCAHa79G2a9PnP0umzf25/coF0XSupd2vJy2ZcfQAAAPAMpj0DbO7cuZozZ46WLFmiiRMnqmfPnpo8ebK6du2q/Px8xcbGSpLi4+MVEBCgoKAgBQUFacSIEQ6t4+DRPeoU3U/XdRgmLy8vNW/YQe2a9dSWPWscehwAcHchgWG6vfsE/Xf54yosLJQk3Xj1CO09vE0Duo1TSGCYwRXCCN/8Ki3YULz5JUln8qVlP0uLfzSmLgAAAHgW0zbApk6dqj59+iguLq7Y9qioKPn6+qpNmzbnt73//vvKzs5Wdna23n33XYfWERV5lf429P3zn588fVQpu1ar2WXtHHocAM6VnJxsdAke4bbuDyvrxH59tfG989sa1IlSZHiUgVVZl9lzfSJHWvZTxWOSf5MOHHdNPZ7G7PkALgW5dh3mGlZErj2bKRtgGRkZSklJ0eDBg0u9lp6erpiYGPn5+bm8rlM5x/XkO7eoecNYdWl1i8uPD+DSbdmyxegSLOel0Uka1vuJYttq+ofok2eydOPVI4wpysOYPdc/7JQK7ZWPW7vd+bV4IrPnA7gU5Np1mGtYEbn2bKa8B1hGRoYkqX79+sW25+TkKDk5WX379i22ffTo0RozZoyuvvpq/etf/1Lz5s0rPUa/fv20bNmyYtum/2Wl2jWLL3P8/qxdevK/N+uyOs30xN0fycvLlL1DwO0kJyfp6qE9L/n9EyZMUI8ePSr9ac7MmTM1fvz4cl+Pi4vTqlWrNGPGjEuuxRkqWpfgGNXNoDNYJdfXj/qvruw2XF7e5X+7UVhYoHfnLdfAjje7sLKL8/AHRV08m808NyurSkbMng/gQlZZ98pitjXEynMNz0WuPVNY2MXdYsWUXZzw8HBJUlpaWrHt06ZN0/79+9WhQ4di23bv3q3du3erQ4cOuvXWW5Wf79hHS23L+FHjXumsji1u1NP3LpKfb4BD9w8AVjbpznfVusm1RpcBA+SfzZEq/Q+fvWgcAAAA4EQ2u91ehYsTXKuwsFDt27fX/v37NX36dEVGRioxMVHLli1Tenq61q1bp86dO5d6n91uV7169ZSUlKRWrVpd9HE3zJOOZRTfdvTkQY16qY1uj3tUd/b866V+SQDKEdpQ6nhn9faRmZmprKysCsdER0crNTW13NfDwsIUERFRvUKcoKx1CY7liAw6gxVy/ete6c2kysfd1VXq1NTp5VyyRz4s+vVfw4yto6TKMmL2fAAlWWHdK4sZ1xCrzjU8G7n2LPHx8ZKkpKSkKr/HlGeAeXl5af78+YqJidHo0aN13333KTw8XAkJCfLx8VHbtm3LfJ/NZnP4qcWf//AfHTt1SB9+/axumRx0/uO/n0926HEAONfTTz9tdAmAw5k911c2kCJCyj8JzGaTgvylqy53bV2ewuz5AC4FuXYd5hpWRK49mynvASZJLVq00MqVK4ttGz58uKKjoxUQUHQJ4rFjx7R+/Xr16NFDdrtdzz//vEJDQ9WiRQuH1XHXdY/rrused9j+ABhjyJAhRpcAOJzZc+1lkx6Il/79lXS8jKscA3ylv/SUapj2uxH3ZvZ8AJeCXLsOcw0rIteezZRngJVnw4YNio2NPf95Xl6e/va3vyk8PFyRkZHasGGDlixZIh8fvpMGUFx0dLTRJQAO5w65rhssTeon3XxV0e/9fKSwmtKNbaS/3iw1vLh7l+IiuEM+gItFrl2HuYYVkWvP5jadouzsbKWlpWnMmDHnt9WtW1cbN240sCoAAFCZmn5S75iiDwAAAMAIbtMACwoKUkFBgdFlAICpHD6+V28sfVTBgWG6ol6M+ndLOP/a+t+W6/Mf3tZT9ySqsLBQ/170kPxrBOrg0T36653vKWXXt/pq43uyy65buo5Wo7otNTNxlGoH11OdkAYa2utxbUz7Umu2LJavj5+uatZTHVveqBnz71dwYJhq+PjrgZun6dtfFurbXxbIy8tbA7s/oqjI9hr3alc1rNtSkpTQ/2Vt2r6i2Ji6oY306sIEBQXUVu7ZU3p08H/k61NDp3JPaNyrXfTigysUFBBa6lhfbpijlZv+p9rB9XVVVE/17jC81Nf15meP6XTuSR05sVe1g+vrb0PfN+qPBwAAAIBJuNUlkABwqc49JcRqPl33hgZcO07jBs7S9799pvyCPEnS5u0rtffIduWcyZYknck7raSf5un4qcPysnmphq+/ElfP0ITBb2v8oDf10cp/6uedq9SmaQ+NGzhLgf4h2rxjpZaseU3BgWHKzjmmqMj2ys45pn5dHlTCgFeUcThNZ/PPyGazaeId76h/t7H6NmWhDh3LUO7ZU/L1qaFGES1VM6BWqTFHTx7UXb0m65HbX1dQQKgOZO1SYWGh/rPs72pQp5kklXmsX3auUnitSEl2RV/epcyva+yAVzV+0JsKCayjh257zag/Gpewaq7hGOQDVkSuXYe5hhWRa89GAwyAR5g9e7bRJTjF0ZMHVDe0kSQpOKC2TuUelyRdFdVTA7qNPT/ObrfrmRGL9dgd76h+WBP9vHOVZLfL16eG/HwDdDY/V52i+yrz6B69tmic0g+m6lj2Qe0+kKK7rpus4dc/pXeWP6GwkPpq07S7ZiaOks3mJZts6tZ6gLbsXqMZ8x9Q+6he8vMN0N+GfqDxg97UiVNHtGX3mlJjGtePUdMGbfV96jLZbF5qFNFS73/1tPp1GaXgwDqSVOaxbrx6pMYNnK1RN0/XW59NKvvrkvTF+nfUNaa/avqHuPhPxLWsmms4BvmAFZFr12GuYUXk2rPRAAPgEUaPHm10CU4REXq5Dh/LkCSdPJ2lIP/QMsdtTPtSW39fL0kKC7lMp3NPqIaPv/Lyz+pMXo5q+PjrxKnD6tiyjxIGvKLawfUUEXqFImpfoRo+/gr5oyl1+Pg+7TmYqvGD3lS7ZvH64bdl+nHbN2rbtIdmPbJRc795TgeP7tHR7IOSpJCadZRfcLbUGEn64KtntffwNiX0f1nHsg8pNX2dlq6drdT0tfo4aVqZx9qy+zt5e/ko0D9EdtnL/Lok6fvUTxXXzvpP+bFqruEY5ANWRK5dh7mGFZFrz+Y29wADgOpISkoyugSnuKnT/Xp96QR9ueFdXdtmoGYvHa8Hb35Jvj41io3r0Ly3XvxohA5k7dKZvBzd2nWMAvyCNWP+/covzNOw3k8qNChCX2/8q75LWahA/xC1btJN/buN1bSP7pWXzVtDek5SDV9/vb50vEICw3U697iuj71Hq35O1LR598pms+m6DnfrsjpNNW/lC1q/dbkKCvLUtmmcPv/hP8XGfLXhPX394/u6slEnvTD3bt19/VN64YEvJEnT5o3QkPhJ8vGpUepYJ3OO6qX5/092e6GG9vy7rqjXqtTXlXP2lGr4Bsjby9uIPxKXsmqu4RjkA1ZErl2HuYYVkWvPZrPb7XajizCLDfOkP06kAOAioQ2ljndWbx+ZmZnKysqqcEx0dLRSU1PLfT0sLEwRERHVK8QJWJeczxEZdAYr59rdPPJh0a//GmZsHSVVlhHyAXdj1XXPjGuIVecano1ce5Zz93O7mKYml0ACAAAAAADA0miAAfAIFf2kB3BX5BoVIR+wInLtOsw1rIhcezYaYAA8wscff2x0CYDDkWtUhHzAisi16zDXsCJy7dm4CT4AjzBlyhQNGeJ+TwUM5hYETufOc+yuuYZrkA9YEbl2HeYaVkSuPRsNMAAwsZa9jK4AAAAAANwfl0ACAAAAAADA0miAAfAIs2bNMroEwOHINSpCPmBF5Np1mGtYEbn2bDTAAHiEmJgYo0sAHI5coyLkA1ZErl2HuYYVkWvPRgMMgEeIi4szugTA4cg1KkI+YEXk2nWYa1gRufZs3AQfgNsLCgpSQEBApeMiIyPLfc3b29uRJQHVRq5RmapkhHzAnbDuuQ5zDSsi16gMDTAAbi83N1f5+fkVjunatatycnLKfd3Hx0eBgYGOLg24ZOQalaksI+QD7oZ1z3WYa1gRuUZlaIABcHv5+fnKysqqcMzatWsrHBMWFubosoBqIdeoTGUZIR9wN6x7rsNcw4rINSrDPcAAAAAAAABgaTTAAAAAAAAAYGk0wAB4hMTERKNLAByOXKMi5ANWRK5dh7mGFZFrz0YDDG7vi/Xv6tHZ8VUeP+Zfsdqw9UvnFQQAAAAAAEyFBhg8TteY/lqzZbHRZcDFBg0aZHQJgMORa1SEfMCKyLXrMNewInLt2XgKJDzONTH99eR/b9a4ga8ZXQpQqa0rpJOZRldhbcERUsteRlcBAAAAwJlogMFyFn37qr5LWagX/7JCHye9qF92rtKzI5dqzhdTtGv/z/rHiIXy9vJRWsZGtWgYa3S5QIVOZkrHMoyuAgAAAADcG5dAwnK279ukZpHtJUk79m0u8/ddYm7lMkgPk5CQYHQJgMORa1SEfMCKyLXrMNewInLt2WiAwXK2ZfyoqAZFja7tezepWYOrJP3RAPvj952v7KcfUj8zqEIYYezYsUaXADgcuUZFyAesiFy7DnMNKyLXno0GGCwlL/+s0jN/VVRke53Jy1HG4TRFRbbXydNHlXksXVF/nAGWeWyPIkIvN7hauFKPHj2MLgFwOHKNipAPWBG5dh3mGlZErj0b9wCDpRw8ulv5BXlqEB6lHfs2K6BGkC4La6K1vy5VvdpXKCK0kSRp7ZYlurbN7QZXC1c6dOiQ0SVYzqOz45W6Z628vX3l5eWt+rWb6K7rJiuu3WCjS/MY5BoVIR+wInLtOsw1rIhcezbOAIOl2GxFkd53eLt27C265PFsXq7mrZiqGzveJ0nKPXtam3esVJdWNxtZKmAJw3o/qaXPZeuTfxzRjVeP0NS5d2nv4e1GlwUAAAAAxdAAg6VEhkcpvt0dGj/rWn2cNE37j+zQiH821+UR0bqr9xOSpA1bv1BUZHvVqhlucLVwpVatWhldgqV5e/vops4PqKAwXzv2bTa6HI9BrlER8gErIteuw1zDisi1Z+MSSFjO5LvnKS1jo/7x7gDFtbtDt3V/+Pylj5K09tcluqZVfwMrhBEWLFhgdAmWlpd/Vp+umS1JahjewuBqPAe5RkXIB6yIXLsOcw0rIteejTPAYEnNGlylE6ePqHfs8GLNL0mqV/sK9eAeRR7nqaeeMroES5r7zXMa8GSobn48QO988YQmDH5bTRu0lSQ99+FQfZey6PzYKe8O0KbtKwyq1JrINSpCPmBF5Np1mGtYEbn2bKZugBUWFmr69Olq3ry5/P391a5dOyUnJ6tly5YaNWqUy+pI/uljjXwxWrdMDtL902OK/YcOxmvW4Crd0HFEsW2/Z/6mwsICXVGv9Cmu99zwD9WrfYWLqoNZzJ8/3+gSLOmu6yZr0bPHlPiPw+p0ZV9tvqDBNab/y3rvyyk6nXtSq3/5RMGBYWof1cvAaq2HXKMi5ANWRK5dh7mGFZFrz2bqSyBHjhyphQsX6sknn1RsbKzWrFmjoUOH6tChQ5owYYJLasg4lKbpH92nF0Z9pZjG1+jHtK/1xH/76X9P7uUeUiYRFXmVoiKvKratcf0YLXvhjDEFAR4oOLC2Jgx+W/e+0ExrUhbrmtb9VTsoQgO7j9dri8dp5/6f9M9RXxtdJlBldrv0615pddqf215fIXVvKbVqINlsxtUGAACAi2faM8Dmzp2rOXPmaMmSJZo4caJ69uypyZMnq2vXrsrPz1dsbKwkyW63a8qUKbrssssUEhKi3r17O7SOhnVb6OMpBxXT+BrlnMnWkZP7FeAXLB/vGg49DgC4u5DAMN3efYL+u/xxFRYWSpJuvHqE9h7epgHdxikkMMzgCoGqsdulxPXSW8nS1gN/bt96QHorSfpkQ9EYAAAAuA/TngE2depU9enTR3FxccW2R0VFydfXV23atJEkvfTSS9q0aZM2b96s8PBw/fTTTw6vJcAvSPuzdmnEC1Gyy65xt81WTf8Qhx8HgPMkJycbXYJHuK37w/pk9Ux9tfE93Xj1CElSgzpRigyPMrYwiyLXzrFmm/TdtqLfX9joOvf71WlSg9pSV5PHmnzAisi16zDXsCJy7dlM2QDLyMhQSkqKxo8fX+q19PR0xcTEyM/PTwUFBZo5c6Y2btyoevXqSZI6dOjglJoiajXSZ1Nz9cuu1XrqnVsVWbc597IB3MiWLVsUERFhdBmW8tLopFLbavqH6JNnslxfjIci145nt0srf5Nskso7ycsmaWWq1KWZuS+FJB+wInLtOsw1rIhcezbTNsAkqX79+sW25+TkKDk5WX379pVU1Ayz2Wx69dVX9eabb6pevXr6v//7Pw0YMKDSY/Tr10/Lli0rtm36X1aqXbP4Msd7exdNVfuoXure5natSVlEAwxwgOTkJF09tOclv3/ChAnq0aNHpT/NmTlzZplN9XPi4uK0atUqzZgx45JrcYaK1iU4RnUz6AxWz7WZ1YpoqhEzdlQ4xi4p84QUWq+JThza7ZK6SqpKRsgH3ImV172HPyhqp9tM0jG38lzDc5FrzxQWdnG3WDHlPcDCw4tuLp+WllZs+7Rp07R///7zZ3nt27dPe/fuVU5OjjIyMjR79mzde++92r17t8Nq+T51mSa9Ufy+YnkFZ1UzINRhxwAAK5t057tq3eRao8sAqsTb198pYwEAAGAsm91uvtu4FhYWqn379tq/f7+mT5+uyMhIJSYmatmyZUpPT9e6devUuXNnbdq0SR06dNCRI0fOd/5uvfVWDRw4UCNGjLjo426YJx3LKL4t68QBjXzxSiUMeFXXtR+m9VuX6/m5Q/XqQ9/r8ogrHfDVAp4ttKHU8c7q7SMzM1NZWRVfdhcdHa3U1NRyXw8LCzPl6dBlrUtwLEdk0BmsnGszy82TnkiU8gsrHufrLf3f7ZKfr2vqKktlGSEfcDdWXfce+bDo138NM7aOC1l1ruHZyLVniY+PlyQlJSVV+T2mPAPMy8tL8+fPV0xMjEaPHq377rtP4eHhSkhIkI+Pj9q2bStJat68uXx9nfudZ1hIfT1731It/PZl3fZUqOZ8+ZSevncRzS/AzTz99NNGlwA4HLl2PH9fqWOTovt8lccm6eomxja/qoJ8wIrItesw17Aicu3ZTHkPMElq0aKFVq5cWWzb8OHDFR0drYCAAElSUFCQ+vfvr6lTp+r555/Xxo0btWrVKr3yyisOraVN0+6a9fAGh+4TgGsNGTLE6BIAhyPXztGnrbRlr5SdW/pG+DZJwQHSDW2MqOzikA9YEbl2HeYaVkSuPZspzwArz4YNGxQbG1ts2+zZs5WamqratWvr3nvv1bvvvqvGjRsbUyAA04qOjja6BMDhyLVzhAZKj9woRdUr/Vrz+tIjNxSNMTvyASsi167DXMOKyLVnM+0ZYCVlZ2crLS1NY8aMKbY9PDxcn376qUFVAQAAK6oTJCX0lg6ekPYcLjrzq3G4VDfE6MoAAABwKdymARYUFKSCggKjywAAUzl8fK/eWPqoggPDdEW9GPXvlnD+tfW/LdfnP7ytp+5JlCSNe7WrGtZtKUlK6P+ytv6+Xl9tfE922XVL19FqVLelZiaOUu3geqoT0kBDez0uLy8vnco9oXGvdtGLD65QaFCE/r3oIfnXCNTBo3v01zvf0087krRmy2L5+vjpqmY95e8XpK83vi9J+nHbV/rnqK/1844kbc1Yr5wz2eodO1x+voGlxmQeS9faLYtVUJivmMbdFF6rYakxV9SLrvTr+jV9XbH93NDxXtf8YcCS6oUUfQAAAMC9uU0DDACq49xTQqzm03VvaMC14xTT+Bo9/p++6tdllHy8fbV5+0rtPbJdOWeyJUmHjmUo9+wp+frUUP2wJqoZUEuJq2fo6XsXqdBeoOc+uFM3dByhNk17aGD3h7Xw21e0ecdKXdWsp/6z7O9qUKeZJOlM3mkl/TRPXVvdKi+bl2r4+mvJmtfU5LK2Onx8r6Ii2yui9uXq0Pw6rUlZrBaNOuqKetHasW+zJg75r06cOqJXF43V5GH/KzVmxaYPtWPfT6rh46cbO96nVo27lhpTla8rZdfqYvuxMqvmGo5BPmBF5Np1mGtYEbn2bG51DzAAuFSzZ882ugSnOHrygOqGNpIkBQfU1qnc45Kkq6J6akC3sefH+fkG6G9DP9D4QW/qxKkj2rJ7jWS3y9enhvx8A3Q2P1edovsq8+gevbZonNIPpupY9kG9/9XT6tdllIID60iS7Ha7nhmxWI/d8Y7qhzXRzztXafeBFN113WQNv/4pvbP8CUlSQUG+lqydpVu7Fl223qv9UOWcydbrSx/V0J5/L3NMh+a9Ne3Br/Xk8Pn638qpZY6pytdV1n6syqq5hmOQD1gRuXYd5hpWRK49Gw0wAB5h9OjRRpfgFBGhl+vwsQxJ0snTWQryDy1z3MGje3Q0+6AkKaRmHeUXnFUNH3/l5Z/Vmbwc1fDx14lTh9WxZR8lDHjlj8sgI5Wavk5L185WavpafZw0TRvTvtTW39dLksJCLtPp3BOKqH2Favj4K+SPJpkkrfl1ieLaDZGXV9E/M+mZv+nlBX/R8BumqGmDtmWOee/Lf8jHy1c1A0JVUJhf5piqfF1l7ceqrJprOAb5gBWRa9dhrmFF5NqzcQkkAI+QlJRkdAlOcVOn+/X60gn6csO7urbNQM1eOl4P3vySfH1qFBt3WZ2mmrfyBa3fulwFBXlq2zROkk0z5t+v/MI8Dev9pEKDIvT1xr/qu5SFCvQPUbtmcWrXLE6SNG3eCA2JnyQ/3wC9+NEIHcjapTN5Obq16xidzc/VtI/ulZfNW0N6TpIk/bZnnXp3vEdS0VljT/73ZjW9rJ3mLH9Sjeu31p29/lZsjCTdePV9en7uXfLzDdTt3ceX2k9Zyvq6ytqPVVk113AM8gErIteuw1zDisi1Z7PZ7Xa70UWYxYZ50h8nUgBwkdCGUsc7q7ePzMxMZWVlVTgmOjpaqamp5b4eFhamiIiI6hXiBKxLzueIDDqDlXMNx6gsI+QD7saq694jHxb9+q9hxtZxIavONTwbufYs5+7ndjFNTS6BBAAAAAAAgKXRAAPgESr6SQ/grsg1KkI+YEXk2nWYa1gRufZsNMAAeISPP/7Y6BIAhyPXqAj5gBWRa9dhrmFF5NqzcRN8AB5hypQpGjJkiNFlXLRgbkHgdO48x+6aa7gG+YAVkWvXYa5hReTas9EAAwATa9nL6AoAAAAAwP1xCSQAAAAAAAAsjQYYAI8wa9Yso0sAHI5coyLkA1ZErl2HuYYVkWvPRgMMgEeIiYkxugTA4cg1KkI+YEXk2nWYa1gRufZsNMAAeIS4uDijSwAcjlyjIuQDVkSuXYe5hhWRa89GAwwAAAAAAACWxlMgAbi9oKAgBQQEVDouMjKy3Ne8vb0dWRJQbeQalalKRsgH3Anrnusw17Aico3K0AAD4PZyc3OVn59f4ZiuXbsqJyen3Nd9fHwUGBjo6NKAS0auUZnKMkI+4G5Y91yHuYYVkWtUhgYYALeXn5+vrKysCsesXbu2wjFhYWGOLguoFnKNylSWEfIBd8O65zrMNayIXKMy3AMMAAAAAAAAlkYDDAAAAAAAAJZGAwyAR0hMTDS6BMDhyDUqQj5gReTadZhrWBG59mw0wAAAAAAAAGBpNMAAeIRBgwYZXQLgcOQaFSEfsCJy7TrMNayIXHs2ngIJACa2dYV0MtPoKqwtOEJq2cvoKgCcw7rnGqx9AABPQwMMAEzsZKZ0LMPoKgDAdVj3AACAM3AJJACPkJCQYHQJgMORa1SEfMCKyLXrMNewInLt2WiAAfAIY8eONboEwOHINSpCPmBF5Np1mGtYEbn2bDTAAHiEHj16GF0C4HDukOuCQmnbAembLdKHa6R3VksffCd9lSL9tl/KLzC6Qutyh3wAF4tcuw5zDSsi156Ne4AB8AiHDh0yugTLeXR2vFL3rJW3t6+8vLxVv3YT3XXdZMW1G2x0aR7DzLk+my8l/Sat2SYdO13+uCB/qWszqWcrKbCG6+rzBGbOh7ti3TMeuXYd5hpWRK49Gw0wF/hi/buat3KqHrn9TbVrFqfZS8YrLWODoiI7KKH/y2W+x2xjzFiTmeredSBF/0ocJS8vbzWoE6WJQ/4rm81W6j2p6d/r9SXj5eXlrRYNO2r0rTMddixnfd0XY+/h7XrmvUHq0upm3dfn/6q9P5jfsN5PaljvJ1RQkK/Fa/6tqXPvUlRke0WGRxldGgy0M1Oau1Y6nF352Oxc6ast0g87pTs6S60inV8fUB2sewAAuCcugXSRwXGPqV2zOG3L+FG5Z09p5pjVys8/q62/ry811mxjzFiT2epuVLelXh67RjPHrJYkpWVsKPM99UKv0IsPrtDMMat1LDtTu/b/4pBjOXMOL0ZkeJTG9P9XtffjDK1atTK6BEvz9vbRTZ0fUEFhvnbs22x0OR7DjLnetEf699dVa35d6HiO9GaS9F2aU8rySGbMh5Ww7hmDXLsOcw0rIteejQaYi/26Z606NO8tSerQvLdS96wz/Rgz1mS2un28fc//3tfHT3VrNSrzPWEh9VXD11+S5O3lIy8vb4ccy1lft5UsWLDA6BIsLS//rD5dM1uS1DC8hcHVeA6z5TrtgPT+d1Kh/dL3MX+9tHmP42ryZGbLh9Ww7hmDXLsOcw0rIteejQaYi2XnHlOgX4gkqaZ/LZ3MOWr6MWasyYx1r9myRA9Mb61j2ZkKqVmnwvfs3Pezjp86rCvqlf4JxKUcy1lfk5U89dRTRpdgSXO/eU4DngzVzY8H6J0vntCEwW+raYO2kqTnPhyq71IWnR875d0B2rR9hUGVWpOZcp1ztuiyx8qaXw/EF31U5KMfis4IQ/WYKR9WwrpnLHLtOsw1rIhcezZTN8AKCws1ffp0NW/eXP7+/mrXrp2Sk5PVsmVLjRo1yuX1HD15UIP/EaF1v356yfsI8g/V6TMnJEmnzpxQUECo6ceYsSYz1n1NzK16a2KKwmtFat2vn5b7nhOns/TvRWP16OD/XHLNJY/lrK/JSubPn290CZZ013WTtejZY0r8x2F1urKvNl/wH70x/V/We19O0enck1r9yycKDgxT+6heBlZrPWbK9ZcpFd/s/pyYyKKPiuSclZZuckxdnsxM+bAS1j1juVOuz+ZL63f++fmhE8bVcincaa6BqiLXns3UDbCRI0fq2Wef1YMPPqjPP/9cQ4YM0dChQ7Vz507Fxsa6vJ6X5v8/nTh9pFr7aHVFV23a9o0kadO2rxV9eRcVFOTr6MmDph1jxprMUPeFzuafOf/7QL8Q+fkGlLvfF/53tx64+UWFhdSXJIccy5lzCFRFcGBtTRj8tn74bZnWpCyWJNUOitDA7uP12uJxmvvN/2nUzdMNrhLOcjZfWrfdsfvctEc6mevYfQKOxLqHiny3TXrqE+nDtX9ue26p9HaydPqscXUBgCczbQNs7ty5mjNnjpYsWaKJEyeqZ8+emjx5srp27ar8/HzFxsYqPT1dQUFB5z9q1qwpm82mjRs3OryepWtfl3+NmqobWvp+SxejecMO8vX11/hZ3WWzeenKyzvpwNHdemf5E6YdY8aazFB31okD+vCb5yRJG35brgmz4zRhdpyOZh9UbIsbynxP8s/zlfb7er392V/16Ox4/bp7rUOO5aw5vPC4Vfkcni0kMEy3d5+g/y5/XIWFhZKkG68eob2Ht2lAt3EKCQwzuEI4S0qGlJPn2H0WFEo/7nbsPgFHY91DWdZul+b/IOWWsS6mZEhvrJDyC1xfFwB4Oh+jCyjP1KlT1adPH8XFxRXbHhUVJV9fX7Vp00Z+fn7Kzv7zMVNLlizRQw89pA4dOji0loxD25S46iW9+tD3GvOvi993gF+Q5q18QZHhzdWuWZwS+r9c7PWd+35Sz/ZDi20z2xgz1mR03WEh9TXsusmSpGta99c1rftXeuxe7YeqV4n9rv55QbWP5aw5vPC4Vfl87+HtenvZ39Sj7WCZTXJystEleITbuj+sT1bP1Fcb39ONV4+QJDWoE6XI8ChjC7Mos+R6T/VOji5XupP26ynMkg+rY91zLbPnOq9AWvJjxWP2HJF+Spdim7impktl9rkGLgW59mymbIBlZGQoJSVF48ePL/Vaenq6YmJi5OfnV+q1Dz/8UHfddZdsNpvDaikoyNc//3e3xtz68iX/FK9H20Hq0XZQua93b3t7pfsw2xgz1uTquh3FEcdy9RyWJzI8Sq8+ZM6nSW7ZskURERFGl2EpL41OKrWtpn+IPnkmy/XFeCiz5Hqfk56h4az9egqz5MNKWPeMZ/Zc//J75WfE2mzSmu3mb4CZfa6BS0GuPZtpG2CSVL9+/WLbc3JylJycrL59+5Z6z8mTJ7V06VL98MMPVTpGv379tGzZsmLbpv9lpdo1iy+27YOvn1WzBlepc3TpYwKovuTkJF09tOclv3/ChAnq0aNHpT/NmTlzZplN9XPi4uK0atUqzZgx45JrcYay1iU4VnUz6Azulus7nv5e9Zt1KrbtgfiKb3b/r2HFP9+yV3orqfi2rTv2yGZr7IgSLacqGTFLPi4W655rmG3tc7d1rzyxN0/StXf+s8Ixdru06dc9st3Q2DVFlWCVuQYuRK49U1jYxZ2kZMoGWHh4uCQpLS2tWLNr2rRp2r9/f5mXOH7yySdq3ry5Wrdu7dBakn/6SEdO7lfSTx9Jkk6fOaHnPrxTw657Qnf2+ptDjwUAVjTpzneNLgFOVljg4BuAndtvPneKhnti3fNcebknKx1jLyzU2Rw3eyQkAFiAzW63240uoqTCwkK1b99e+/fv1/Tp0xUZGanExEQtW7ZM6enpWrdunTp37lzsPTfccIN69+6tSZMmXfJxN8yTjmVUPObu5xtr7IB/q0urmy/5OAD+FNpQ6nhn9faRmZmprKyKLz+Jjo5Wampqua+HhYWZ8nToqqxLqB5HZNAZ3CnXc9dKP+ys2thzZ3498mHlY1s1kEaZ5wQV06ksI2bJx8Vi3XMNM6597rTulefYaenpRUVneVXkprbSjW1cUlKZrDDXQEnk2rPEx8dLkpKSkqr8HlM+BdLLy0vz589XTEyMRo8erfvuu0/h4eFKSEiQj4+P2rZtW2z8gQMHlJSUpKFDS98AHQAk6emnnza6BMDhzJLrRk560F1DHqBXLWbJB+BIZs91aKAU27j8122S/H2lrm7wjASzzzVwKci1ZzPlJZCS1KJFC61cubLYtuHDhys6OloBAQHFts+bN0/XXHONGjVq5PS6Pnh8t9OPAcDxhgwZYnQJgMOZJdcxkdInGyRHn1Lexvn/rFuaWfIBOJI75HpIJ+lEjpR2oOiG9xeeDebnK42Kl0ICyn27abjDXAMXi1x7NlOeAVaeDRs2KDY2ttT2Dz/8UMOGDSvjHQBQJDo62ugSAIczS67DgqRWFdzw/lJcEe68M8s8hVnyATiSO+S6ho/0l57S/XFS9GVSREjRenbzVdLkW6WmbnJ1lTvMNXCxyLVnM+0ZYCVlZ2crLS1NY8aMKfXa+vXrDagIAACc07edlLpPKnTQaWA3X+WY/QCAEby8pNYNiz4AAObgNg2woKAgFRQUGF0GAJjK4eN79cbSRxUcGKYr6sWof7cESdKna9/QTzuTVKtmuK6PvVfNIzto5oJRCvQLVl7+GY0bOKvUmJaNOupUznFN/3ikBsc/plZXdNGXG+Zo5ab/qXZwfV0V1VM3dLy31Jgf077WVxvfk1123dJ1tJpHdtCM+fcrODBMNXz89cDN00od67KwJpqZOEq1g+upTkgDDe31uDZvX3HR+2ke2UH/XvSQ/GsE6uDRPfrrne+phq+/TuWe0LhXu+jFB1coLKS+wX9KniGydtENnT//ueJxW/ZWvq9rW0jN6zmmLlhTeWvft78s1Le/LJCXl7cGdn9EdUMb6dWFCQoKqK3cs6f06OD/6PvUzyod88X6d4qtNU3qty61HkkqttbUqhmu1xaPU6B/iE7nntCY/i/Lx9u32Bgfb99Sa5+Xl5fW/7Zcn//wtp66J1GFhYWl1rUvN8wpVk9Ug6tKHWvTtm+0Zsti+fr46apmPXVN6/5G/hEBAGA6bnUJJABcqnNPCbGaT9e9oQHXjtO4gbP0/W+fKb8gT5K06uf5+tvQDzT6lpmat2KqftqZpMvCmmr0rTNVq2Zdbdm9ptQYSZr7zXPyqxF4fv+/7Fyl8FqRkuyKvrxLmWMSV8/QhMFva/ygN/XRyn8qO+eY+nV5UAkDXlHG4TSdzT9T6lg/71ylNk17aNzAWQr0D9HmHSsvaT9n8k4r6ad5On7qsLxsXqrh66/CwkL9Z9nf1aBOM9f9QRjEbLm+vrXUsXHFY95KKvooT6sG0oAODizKg5ktH45U3tpns9k08Y531L/bWH2bslBHTx7UXb0m65HbX1dQQKgOZO2q0piSa01Z61HJtea7LYt0Nj9Xp3NP6PKIaPl4+5YaU9bat3n7Su09sl05Z7Ilqcx1rWQ9ZR1ryZrXFBwYpuycY4qKbG/MH4wLWDnXZsNcw4rItWejAQbAI8yePdvoEpzi6MkDqhtadKfw4IDaOpV7XJI09LrHNf2j+/T+188or+BMsXF1QxvqyIl9pcZI0gM3TyvWOLrx6pEaN3C2Rt08XW99NqnMMbLb5etTQ36+ATqbn6uwkPpq07S7ZiaOks3mJZtspY7VKbqvMo/u0WuLxin9YKqOZR+8pP3Y7XY9M2KxHrvjHdUPa6Kfd67S+189rX5dRik4sI4r/ggMZbZce9mku7pKPaOLnnR2sbo0k0b2kHy8HV6aRzJbPhypvLWvW+sB2rJ7jWbMf0Dto3qpcf0YNW3QVt+nLpPN5qVGES2rNKbkWlPWelRyrdl3ZIea1G+jcQNnKS1jg/Yd3lFqTFlr31VRPTWg29jzX1tZ61rJeso61u4DKbrruskafv1Temf5Ey7+E3EdK+fabJhrWBG59mw0wAB4hNGjRxtdglNEhF6uw8cyJEknT2cpyD9UknTo2O967I53dUf8JPl4+RaNO57xx2sZqhPSoNSYsmzZ/Z28vXwU6B8ieznP+Kvh46+8/LM6k5ejGj7+Onx8n/YcTNX4QW+qXbN4/fDbslLHOnHqsDq27KOEAa+odnA9RYRecUn72Zj2pbb+XnQfyLCQy3Tw6B6lpq/T0rWzlZq+Vh8nTXPwjJuLGXPt5SX17yA9dL3UoHbV3lM3uOipaHd2ofnlSGbMh6OUt/b9uO0btW3aQ7Me2ai53zwnSfrgq2e19/A2JfR/ucpjSq41JdejNVsWl1prwoLrK9A/RJIUUjNcJ3OySo0pa+0rqeS6djr3RKl6Sh7LLrsiahetoyEWb/5bOddmw1zDisi1Z3Obe4ABQHUkJSUZXYJT3NTpfr2+dIK+3PCurm0zULOXjteDN7+kIP9Q/XPecOUX5OnePs+qcb0Yfb3xfb22+GFJUkzja3Q8+1CxMWWpFVRXL83/f7LbCzW059/LHHNb90c0Y/79yi/M07DeT6qGr79eXzpeIYHhOp17XNfH3qOfdyQXO1ZoUIS+3vhXfZeyUIH+IWrdpJsKCvMvej8RtRrpxY9G6EDWLp3Jy9HDA2fr+tjhkqRp80ZoSPwk50y8SZg5100jpMduknYfljbuktKzpP3HpLw/bufZsLbUqI7U/oqi+33ZLuWUMVTIzPmorvLWvgNZuzRt3r2y2Wy6rsPd+mrDe/r6x/d1ZaNOemHu3br7+qeqNKbkGlpyPRp7278V126wpD/XmpoBtTRj/v3afSBFNXwD1LLR1XrhgS+KjQmpWafU2ldSh+a9i61rt3Ydo+9TPytWT4M6zYodKzI8Sv27jdW0j+6Vl81bQ3pad+2zcq7NhrmGFZFrz2az2+0Oel6T+9swT/rjh4kAXCS0odTxzurtIzMzU1lZWRWOiY6OVmpqarmvh4WFKSLCfM8lZ11yPkdk0BmsmutHPiz69V/DjK3DCirLiDvmQ2LdcxUzrn1WXffMiLmGFZFrz3Lufm4X09TkEkgAAAAAAABYGg0wAB6hop/0AO6KXKMi5ANWRK5dh7mGFZFrz8Y9wAB4hI8//lhDhgwxuoyLFswZ2E7nznPsrrmGa7hrPtz576Q7cdd5dtdcuyPmGlZErj0bDTAAHmHKlClu+Y9dy15GVwAzc9dcwzXcNR+se6iIu+baHTHXsCJy7dm4BBIAAAAAAACWRgMMAAAAAAAAlkYDDIBHmDVrltElAA5HrlER8gErIteuw1zDisi1Z6MBBsAjxMTEGF0C4HDkGhUhH7Aicu06zDWsiFx7NhpgADxCXFyc0SUADkeuURHyASsi167DXMOKyLVn4ymQANxeUFCQAgICKh0XGRlZ7mve3t6OLAmoNnKNylQlI+QD7oR1z3WYa1gRuUZlaIABcHu5ubnKz8+vcEzXrl2Vk5NT7us+Pj4KDAx0dGnAJSPXqExlGSEfcDese67DXMOKyDUqQwMMgNvLz89XVlZWhWPWrl1b4ZiwsDBHlwVUC7lGZSrLCPmAu2Hdcx3mGlZErlEZ7gEGAAAAAAAAS6MBBgAAAAAAAEujAQbAIyQmJhpdAuBw5BoVIR+wInLtOsw1rIhcezYaYAAAAAAAALA0GmAAPMKgQYOMLgFwOHKNipAPWBG5dh3mGlZErj0bT4EEABPbukI6mWl0FdYWHCG17GV0FQAAAACciQYYAJjYyUzpWIbRVQAAAACAe+MSSAAeISEhwegSAIcj16gI+YAVkWvXYa5hReTas9EAA+ARxo4da3QJgMORa1SEfMCKyLXrMNewInLt2WiAAfAIPXr0MLoEwOHINSpCPmBF5Np1mGtYEbn2bNwDDIBHOHTokNElWM6js+OVumetvL195eXlrfq1m+iu6yYrrt1go0vzGOQaFSEfsCJy7TrMNayIXHs2GmAAgEs2rPeTGtb7CRUU5Gvxmn9r6ty7FBXZXpHhUUaXBgAAAADncQkkAI/QqlUro0uwNG9vH93U+QEVFOZrx77NRpfjMcg1KkI+YEXk2nWYa1gRufZsNMAAeIQFCxYYXYKl5eWf1adrZkuSGoa3MLgaz0GuURHyASsi167DXMOKyLVnowEGwCM89dRTRpdgSXO/eU4DngzVzY8H6J0vntCEwW+raYO2kqTnPhyq71IWnR875d0B2rR9hUGVWhO5RkXIB6yIXLsOcw0rIteezdQNsMLCQk2fPl3NmzeXv7+/2rVrp+TkZLVs2VKjRo1yWR0fJ72om/5WQ7dMDjr/8cvO1S47PoDqmz9/vtElWNJd103WomePKfEfh9Xpyr7afEGDa0z/l/Xel1N0OvekVv/yiYIDw9Q+qpeB1VqPu+W60G50BZ7F3fIBVAW5dh3mGlZErj2bqW+CP3LkSC1cuFBPPvmkYmNjtWbNGg0dOlSHDh3ShAkTXFbHjn2bNbLP8xocP9FlxwQAdxIcWFsTBr+te19opjUpi3VN6/6qHRShgd3H67XF47Rz/0/656ivjS4TBjl8UlqZKq3f+ee2uWulntHSZaGGlQUAAAAPYtozwObOnas5c+ZoyZIlmjhxonr27KnJkyera9euys/PV2xsrCTpxx9/VJcuXRQSEqIWLVo45Zre7Xs3qVmDqxy+XwCwkpDAMN3efYL+u/xxFRYWSpJuvHqE9h7epgHdxikkMMzgCmGE9CPSi59La7ZJZwv+3L5+p/TS59LW/cbVBgAAAM9h2gbY1KlT1adPH8XFxRXbHhUVJV9fX7Vp00aSdM8992jgwIE6duyY3nrrLd177706fPiww+rIPXtaGYfTtPDblzXk6foa+WK0lv/wX4ftH4BrJCcnG12CR7it+8PKOrFfX2187/y2BnWiFBkeZWBV1mX2XOcXSG8nSWfzpZJXP9olFRRK/10l5Zw1oDgPYPZ8AJeCXLsOcw0rIteezZQNsIyMDKWkpGjw4MGlXktPT1dMTIz8/PwkSXv27NHQoUPl5eWluLg4RUREaNeuXQ6r5Wj2QcU07qabu47Wh5PTNf72N/X60gn64bfPHXYMAM63ZcsWo0uwnJdGJ2lY7yeKbavpH6JPnsnSjVePMKYoD2P2XP+SIZ3Ilezl3PvLLulMvrTBcf9s4wJmzwdwKci16zDXsCJy7dlMeQ+wjIwMSVL9+vWLbc/JyVFycrL69u17ftvYsWP1/vvva9KkSUpOTlZhYaFiYmIqPUa/fv20bNmyYtum/2Wl2jWLL7btsrAmmjH6zy5xm6bd1Tt2uNakLFKnK2+62C8NQAnJyUm6emjPS37/hAkT1KNHj0p/mjNz5kyNHz++3Nfj4uK0atUqzZgx45JrcYay1iU4VnUz6AxWyXWvkW8oJn6kvLzK/3ajsLBA/3x9iXr8a6ALK3N/VcmI2fMBXMgq6547YK5hReTaM4WFXdwtVkx5Blh4eLgkKS0trdj2adOmaf/+/erQocP5bTfeeKNef/11+fv7q2/fvpo5c6YCAwMdVsu2jB81b8ULxbadzcuVr6+/w44BAFY26c531brJtUaXAQN4efuUvvaxrHEVNMgAAAAAR7DZ7eVdmGCcwsJCtW/fXvv379f06dMVGRmpxMRELVu2TOnp6Vq3bp06d+6sI0eOqGnTpnr//ffVr18/bdmyRX379tXy5cvVunXriz7uhnnSsYzi2zIOpenBGe3016Hv69rWA7V5x0r9Y84AvfSXZDVv2KHsHQGostCGUsc7q7ePzMxMZWVlVTgmOjpaqamp5b4eFhamiIiI6hXiBGWtS3AsR2TQGayQ6+TfpIUbKx93Q2upbzvn12M1lWXE7PkASrLCuucumGtYEbn2LPHx8ZKkpKSkKr/HlGeAeXl5af78+YqJidHo0aN13333KTw8XAkJCfLx8VHbtm0lSTt37lRYWJhuvfVWeXt7q23bturSpYtWr17tsFoa1m2hJ4Z/rA++ekb9nwjWqwvHaOKQd2h+AW7m6aefNroEwOHMnuurm0o+3hWPsUnqyjMSnMLs+QAuBbl2HeYaVkSuPZtprzlo0aKFVq5cWWzb8OHDFR0drYCAgPNjjh07pmXLlummm27Sli1btGrVKj388MMOraVrq1vUtdUtDt0nANcaMmSI0SUADmf2XAfWkIZ0kuauLWp0XXjK+bnPb2kv1a5pTH1WZ/Z8AJeCXLsOcw0rIteezZRngJVnw4YNio2NPf95rVq1NG/ePD3++OMKCQlRv379NGnSJHXv3t3AKgGYUXR0tNElAA7nDrnu1FT6fz2kerWKbw8LkoZ1lXq1MqYuT+AO+QAuFrl2HeYaVkSuPZtpzwArKTs7W2lpaRozZkyx7TfeeKNuvPFGg6oCAACVadNIat1Q2ntUOpEj1fSTGtWRvGxGVwYAAABP4TYNsKCgIBUUFBhdBgCYyuHje/XG0kcVHBimK+rFqH+3hPOvnco9oXGvdtGLD65QWEh9SdKy79/Stowf9fDts7V+6xdau2WxCgrzFdO4m27oeK/GvdpVDeu2lCQl9H9ZNQNqFdtPaFCE/r3oIfnXCNTBo3v01zvfUw1ff63/bbk+/+FtPXVPok6cOqKZiaNUO7ie6oQ00NBej+uzdW9oa8Z65ZzJVu/Y4erU8ia9tnicAv1DdDr3hMb0f1krNs3Vyk3/U+3g+roqqqeuaz+s1JhN21cUqzm+3R2aMf9+BQeGqYaPvx64eZq+3DCn2H5u6HivIX82KM5mkxpe3JOqAQAAAIdxq0sgAeBSnXtKiNV8uu4NDbh2nMYNnKXvf/tM+QV5koqepvufZX9XgzrNzo/9eecq/Z65VfY/7sSUsmu1duz7SfsOb1fD8BY6dCxDuWdPydenhhpFtFTNgFql9nMm77SSfpqn46cOy8vmpRq+/tq8faX2HtmunDPZ54/TpmkPjRs4S4H+Idq8Y6VqBoRq4pD/6uGBs7Vi01x9t2WRzubn6nTuCV0eES0fb1/9snOVwmtFSrIr+vIuZY4pWXN2zjH16/KgEga8oozDaTqbf6bUfqzMqrmGY5APWBG5dh3mGlZErj0bDTAAHmH27NlGl+AUR08eUN3QRpKk4IDaOpV7XJL0/ldPq1+XUQoOrCNJOpC1W9+lLFT/bmPPv7dD896a9uDXenL4fP1v5VT5+Qbob0M/0PhBb+rEqSPasntNqf3Y7XY9M2KxHrvjHdUPa6Kfd67SVVE9NeCC/XaK7qvMo3v02qJxSj+YqmPZB9Wr/VDlnMnW60sf1dCef9e+IzvUpH4bjRs4S2kZG7Tv8A7dePVIjRs4W6Nunq63PptU5piSNYeF1Febpt01M3GUbDYv2WQrtR8rs2qu4RjkA1ZErl2HuYYVkWvPRgMMgEcYPXq00SU4RUTo5Tp8LEOSdPJ0loL8Q3Us+5BS09dp6drZSk1fq4+TpumbTR/q5Omjemf5ZKXsWq1fd6/Ve1/+Qz5evqoZEKqCwnwdPLpHR7MPSpJCatbR4eN7S+1nY9qX2vr7eklSWMhlOp17olRNJ04dVseWfZQw4BXVDq6niNArlJ75m15e8BcNv2GKmjZoq7Dg+gr0D/njWOGyy64tu7+Tt5ePAv1DZJe9zDElaz58fJ/2HEzV+EFvql2zeP3w27JS+7Eyq+YajkE+YEXk2nWYa1gRufZsbnMPMACojqSkJKNLcIqbOt2v15dO0Jcb3tW1bQZq9tLxevDml/TCA19IkqbNG6Eh8ZPO3wPsQNZuzVv5glo17qobr75Pz8+9S36+gbq9+3hdVqep5q18Qeu3LldBQZ7u7Pk3xbUbXGw/fr4BevGjETqQtUtn8nJ0a9cxpWoKDYrQ1xv/qu9SFirQP0Qxja/RiH82V9PL2mnO8ifVuH5r3db9Yc2Yf792H0hRDd8ARYZHqVZQXb00///Jbi/U0J5/V7PIq0qNKVlzDV9/vb50vEICw3U697iuj71HJ3OOFtuPlVk113AM8gErIteuw1zDisi1Z7PZ7XZr/3j8ImyYJ/1xIgUAFwltKHW8s3r7yMzMVFZWVoVjoqOjlZqaWu7rYWFhioiIqF4hTsC65HyOyKAzWDnXcIzKMkI+4G5Y91yHuYYVkWvPcu5+bhfT1OQSSAAAAAAAAFgaDTAAHqGin/QA7opcoyLkA1ZErl2HuYYVkWvPRgMMgEf4+OOPjS4BcDhyjYqQD1gRuXYd5hpWRK49G/cAu8DWFdLJTKOrADxLcITUslf19mHl6/1Zl5zPERl0BivnGo7BPcBgNax7rsNcw4rItWe5lHuA8RTIC5jxP0AAPBvrEgAAAABUH5dAAgAAAAAAwNJogAHwCLNmzTK6BMDhyDUqQj5gReTadZhrWBG59mw0wAB4hJiYGKNLAByOXKMi5ANWRK5dh7mGFZFrz0YDDIBHiIuLM7oEwOHINSpCPmBF5Np1mGtYEbn2bDTAAAAAAAAAYGk8BRKA2/Px8VFYWFil4yoa4+PDcghzIdeoTFUyQj7gTlj3XIe5hhWRa1SGP10Abq8q/9BNmTJFERERLqgGcAxyjcpUlhHyAXfDuuc6zDWsiFyjMja73W43uggAAAAAAACgKuLj4yVJSUlJVX4P9wADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApdEAAwAAAAAAgKXRAAMAAAAAAICl0QADAAAAAACApbl9A8xut2vGjBlq2bKlAgMDFRMTo/fff9/osgAAAAAAAGASbt8AmzRpkiZNmqTbb79dn3zyia6//nrdc889WrRokdGlAQAAAAAAwARsdrvdbnQRl+rYsWOKiIjQo48+qqlTp57fftttt6lRo0Z65ZVXyn3vO/M/19adv7uiTAAAAAAAADjI3DdeUmCAn37b8nOV3+PjxHqcbt26dcrLy9PAgQOLbV+4cKFBFQEAAAAAAMBs3LoBduTIEUlSvXr1Lvq99w2+ydHlAAAAAAAAwMnWfT73ot/j1g2wWrVqSZIOHTqkyy+//Pz2LVu26Pjx47rmmmvKfS+XQAIAAAAAALifoIgmqh9R56Le49YNsE6dOsnHx0dLly5VbGzs+e2PPfaYsrOztWrVKgOrAwAAAAAAgKO1ju2qlk0bXdR73LoBFhERoYSEBE2dOlVeXl7q3Lmzli9fruXLl+uzzz6r8L1cAgkAAAAAAOAZ3PopkJJUWFioqVOn6q233tLBgwd15ZVXasqUKRowYIDRpQEAAAAAAMAE3L4BBgAAAAAAAFTEy+gCAAAAAAAAAGeiAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYAAAAAAABLowEGAAAAAAAAS6MBBgAAAAAAAEujAQYA+P/t3W+Qlndh//vPwlJYpFHXbcDZY5qJLLisLCPbmmBHFnr8w9nMGKtAixYNjMXfsuk5ExpzOkVCM6nQQ0lodSb0p09kUqnNLsFiQWc6I7s4JaniX4jM2U7SiKsomG006MZm2T0PfhPOrHeWJHrDde+V12vGB/le9839YcZH77muCwAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFITwAAAAAAoNQEMAAAAgFKrL3oAV9/w8HBGR0cnvf43f/M3+chHPjLp9fr6+jQ2Nl6JaQAAAABVVzc+Pj5e9AiurnPnzmV4eHjS662trTl9+vSk1xsbG3PttddeiWkAAAAAVecRSAAAAABKTQADAAAAoNQEMCr09fUVPQEAAACgagQwAAAAAEpNAKPC6tWri54AAAAAUDX1RQ+odQ+dSL7/X8X8dvOrk/f8TjG/DQAAAFAWAtgL+P5/JY+dK3oFAAAAAL8qj0BSoaenp+gJAAAAAFUjgFHhtttuK3oCAAAAQNUIYFRYvnx50RMAAAAAqkYAo8L58+eLngAAAABQNQIYAAAAAKUmgFFh0aJFRU8AAAAAqBoBjAoHDhwoegIAAABA1QhgVLjrrruKngAAAABQNTUdwMbGxrJ79+60tLRk1qxZWbJkSQYGBrJw4cJs2rSp6HnP6+/+uC4//t6pivP/+T+aMvSd/qs/6FfQ29tb9AQAAACAqqkvesDlbNy4MQcPHsy2bdvS0dGR48ePZ926dTl//ny2bNlS9DwAAAAApoCavQNs//792bdvXw4dOpQ77rgjK1euzNatW7Ns2bKMjo6mo6MjSfKjH/0o73jHOzJ79uwsWbIk3/jGNwpeDgAAAEAtqdk7wHbu3JlVq1als7Nzwvn8+fMzY8aMLF68OEnS3d2dN7zhDfnnf/7nPPDAA1m9enUGBwczffr0ImYnSXrvfksybWJb/O+Rnxa05qUbGBgoegIAAABA1dRkABsaGsqpU6dy++23V1w7c+ZM2traMnPmzDz99NM5fPhwvv/976ehoSGbNm3Kjh078sgjj+T3fu/3LvsbS5Ysybe//e0X3PLerUfzv7WueEn712w/nqbXvXHC2f/8H00v6c9IkoGB/vyf71j5kr93OVu2bMny5csvG7kee+yxvP71r5/0emdnZ44dO5b77ruvqtsAAAAAXoyurq4cPnz4RX++Jh+BHBoaSpLMmzdvwvnIyEgGBgYuPf74H//xH3nNa16Tpqb/Py4tXrw43/nOd170b1Dp0KFDRU8AAAAAqJqavAPsuaA1ODiYrq6uS+e7du3K2bNns3Tp0iTJz372s1xzzTUTvnvNNdfkwoULL/gbzz1C2d/ff9nPfeJfk8fOvZT11dPZuSJ9fzVe9T/33LlzWbhw4aTX9+zZc9l/ZbOxsTG33HJL7r333qpvAwAAAKi2mgxgN9xwQ9rb27Njx440Njamubk5fX19OXLkSJJcugPsFa94RZ5++ukJ3/3pT3+aOXPmXPXNAAAAANSmmnwEctq0aent7U1bW1u6u7uzYcOGNDU1paenJ/X19Wlvb0+StLS05Mc//nGefPLJS989depUFi1aVNT0Urj77ruLngAAAABQNTV5B1iSLFiwIEePHp1wtn79+rS2tqahoSFJ8pu/+Zu5+eabc8899+Sv//qv8w//8A+pq6vLTTfdVMTkJMn/9Q/P/8jih//+x1d5ya9u7dq1RU8AAAAAqJqavANsMidOnLj0+ONz9u7dm0cffTSvfvWr8/GPfzwHDhzI9OnTC1pYDq2trUVPAAAAAKiamr0D7JdduHAhg4OD2bx584TzuXPn5l//9V8LWgUAAABArZsyAWzOnDm5ePFi0TMAAAAAmGKm1COQXB0rVqwoegIAAABA1QhgVNi7d2/REwAAAACqRgCjQnd3d9ETAAAAAKpGAKNCf39/0RMAAAAAqkYAAwAAAKDUBDAAAAAASk0Ao8Lp06eLngAAAABQNfVFD6h1za9++f32gw8+mLVr1xbz4wAAAABVJoC9gPf8TtELrr7t27cLYAAAAEBpeAQSAAAAgFITwAAAAAAoNQGMCvfff3/REwAAAACqRgCjQltbW9ETAAAAAKpGAKNCZ2dn0RMAAAAAqsa/AvkyNGfOnDQ0NFz2M83NzZNemz59erUnAQAAAFwxAtjL0DPPPJPR0dFJry9btiwjIyOTXq+vr8/s2bOvxDQAAACAqhPAXoZGR0czPDw86fWHH374stcbGxuvxCwAAACAK8I7wAAAAAAoNQEMAAAAgFITwKjQ19dX9AQAAACAqhHAAAAAACg1AYwKq1evLnoCAAAAQNUIYAAAAACUmgAGAAAAQKkJYFTo6ekpegIAAABA1QhgVLjtttuKngAAAABQNQIYFZYvX170BAAAAICqEcCocP78+aInAAAAAFSNAAYAAABAqQlgVFi0aFHREwAAAACqRgCjwoEDB4qeAAAAAFA1AhgV7rrrrqInAAAAAFRNTQewsbGx7N69Oy0tLZk1a1aWLFmSgYGBLFy4MJs2bSp6Xmn19vYWPQEAAACgauqLHnA5GzduzMGDB7Nt27Z0dHTk+PHjWbduXc6fP58tW7YUPQ8AAACAKaBm7wDbv39/9u3bl0OHDuWOO+7IypUrs3Xr1ixbtiyjo6Pp6OhIkmzfvj2LFi3KtGnT0tfXV/BqAAAAAGpNzQawnTt3ZtWqVens7JxwPn/+/MyYMSOLFy9OkrS0tOTv/u7v8uY3v7mImaU0MDBQ9AQAAACAqqnJADY0NJRTp05lzZo1FdfOnDmTtra2zJw5M0nyx3/8x3n729+eWbNmXe2ZpfXoo48WPQEAAACgamryHWBDQ0NJknnz5k04HxkZycDAQLq6un7t3zh58mSGh4dTV1f3a/9ZU8mWLVuyfPnyy97ltWfPntx+++2TXu/s7MyxY8dy3333XYmJAAAAAJfV1dWVw4cPv+jP1+QdYE1NTUmSwcHBCee7du3K2bNns3Tp0iJmAQAAADAF1eQdYDfccEPa29uzY8eONDY2prm5OX19fTly5EiSXHoB/q/juXeI9ff3/9p/1lRz7ty5LFy4cNLre/bsyaZNmya93tjYmFtuuSX33nvvlZgHAAAAUFU1eQfYtGnT0tvbm7a2tnR3d2fDhg1pampKT09P6uvr097eXvTEUrv77ruLngAAAABQNTV5B1iSLFiwIEePHp1wtn79+rS2tqahoeHS2bPPPpuLFy9mbGwszz77bJ555pnMnDnzZfdur2pau3Zt0RMAAAAAqqYm7wCbzIkTJyoef/yTP/mTNDQ05Mtf/nLe9773paGhId/97ncLWlgOra2tRU8AAAAAqJopE8AuXLiQwcHBihfgf/rTn874+PiE/11//fXFjAQAAACg5tTsI5C/bM6cObl48WLRMwAAAACYYqbMHWBcPStWrCh6AgAAAEDVCGBU2Lt3b9ETAAAAAKpGAKNCd3d30RMAAAAAqkYAo0J/f3/REwAAAACqRgADAAAAoNQEMAAAAABKTQCjwunTp4ueAAAAAFA1AhgVHnzwwaInAAAAAFSNAEaF7du3Fz0BAAAAoGoEMAAAAABKTQADAAAAoNQEMCrcf//9RU8AAAAAqBoBjAptbW1FTwAAAACoGgGMCp2dnUVPAAAAAKgaAQwAAACAUqsvegBX35w5c9LQ0HDZzzQ3N096bfr06dWeBAAAAHDFCGAvQ88880xGR0cnvb5s2bKMjIxMer2+vj6zZ8++EtMAAAAAqk4AexkaHR3N8PDwpNcffvjhy15vbGy8ErMAAAAArgjvAAMAAACg1AQwAAAAAEpNAKNCX19f0RMAAAAAqkYAAwAAAKDUBDAqrF69uugJAAAAAFUjgAEAAABQagIYAAAAAKUmgFGhp6en6AkAAAAAVSOAUeG2224regIAAABA1QhgVFi+fHnREwAAAACqRgCjwvnz54ueAAAAAFA1AhgAAAAApSaAUWHRokVFTwAAAACoGgGMCgcOHCh6AgAAAEDV1HQAGxsby+7du9PS0pJZs2ZlyZIlGRgYyMKFC7Np06ai55XWXXfdVfQEAAAAgKqp6QC2cePG3HPPPfnwhz+cL3zhC1m7dm3WrVuXxx9/PB0dHUXPK63e3t6iJwAAAABUTc0GsP3792ffvn05dOhQ7rjjjqxcuTJbt27NsmXLMjo6mo6OjvziF7/Irbfemubm5rzqVa/K7//+7+f06dNFTwcAAACghtRsANu5c2dWrVqVzs7OCefz58/PjBkzsnjx4oyOjmb+/Pn5yle+kieffDJve9vb8od/+IcFLQYAAACgFtVkABsaGsqpU6eyZs2aimtnzpxJW1tbZs6cmVe84hX56Ec/mubm5kyfPj1/+qd/mpMnT+aZZ54pYHV5DAwMFD0BAAAAoGrqix7wfIaGhpIk8+bNm3A+MjKSgYGBdHV1Pe/3jh8/nuuvvz6zZs16wd84efJkhoeHU1dX9+sPnkK2bNmS5cuXXzZyPfbYY3n9618/6fXOzs4cO3Ys991335WYCAAAAHBZXV1dOXz48Iv+fE3eAdbU1JQkGRwcnHC+a9eunD17NkuXLq34zn/913+lp6cnH/vYx67KxjI7dOhQ0RMAAAAAqqZufHx8vOgRv2xsbCxvetObcvbs2ezevTvNzc3p6+vLkSNHcubMmTzyyCO58cYbL31+ZGQk73jHO7J8+fIXHcBWrFiRJOnv778Cf4Padu7cuQwPD096vbW19bL/mEBjY2OuvfbaKzENAAAAoOpq8g6wadOmpbe3N21tbenu7s6GDRvS1NSUnp6e1NfXp729/dJnR0dHs3bt2rS0tLj7CwAAAIAKNfkOsCRZsGBBjh49OuFs/fr1aW1tTUNDw6WzD33oQxkbG8snP/nJqz2xtO6+++6iJwAAAABUTc0GsOdz4sSJ3HTTTZf++7vf/W727duXWbNm5VWvetWl8+985zu57rrrClhYDmvXri16AgAAAEDV1OQjkM/nwoULGRwcnPAC/N/+7d/O+Ph4RkZGcuHChUv/E79+Pa2trUVPAAAAAKiaKXMH2Jw5c3Lx4sWiZwAAAAAwxUyZO8AAAAAA4FchgFFhxYoVRU8AAAAAqBoBjAp79+4tegIAAABA1QhgVOju7i56AgAAAEDVCGBU6O/vL3oCAAAAQNUIYAAAAACUmgAGAAAAQKkJYFQ4ffp00RMAAAAAqkYAo8KDDz5Y9AQAAACAqhHAqLB9+/aiJwAAAABUjQAGAAAAQKkJYAAAAACUmgBGhfvvv7/oCQAAAABVI4BRoa2tregJAAAAAFUjgFGhs7Oz6AkAAAAAVVNf9ACuvjlz5qShoeGyn2lubp702vTp06s9CQAAAOCKEcBehp555pmMjo5Oen3ZsmUZGRmZ9Hp9fX1mz559JaYBAAAAVJ0A9jI0Ojqa4eHhSa8//PDDl73e2Nh4JWYBAAAAXBHeAQYAAABAqQlgAAAAAJSaAEaFvr6+oicAAAAAVI0ABgAAAECpCWBUWL16ddETAAAAAKpGAAMAAACg1AQwAAAAAEpNAKNCT09P0RMAAAAAqkYAo8Jtt91W9AQAAACAqhHAqLB8+fKiJwAAAABUjQBGhfPnzxc9AQAAAKBqBDAAAAAASk0Ao8KiRYuKngAAAABQNQIYFQ4cOFD0BAAAAICqEcCocNdddxU9AQAAAKBqajqAjY2NZffu3WlpacmsWbOyZMmSDAwMZOHChdm0aVPR80qrt7e36AkAAAAAVVNf9IDL2bhxYw4ePJht27alo6Mjx48fz7p163L+/Pls2bKl6HkAAAAATAE1ewfY/v37s2/fvhw6dCh33HFHVq5cma1bt2bZsmUZHR1NR0dHkuT9739/5s6dm1e+8pV585vfnIcffrjg5QAAAADUkpoNYDt37syqVavS2dk54Xz+/PmZMWNGFi9enCTZunVrvve97+UnP/lJtm/fnve+971FzC2VgYGBoicAAAAAVE1NBrChoaGcOnUqa9asqbh25syZtLW1ZebMmUmSRYsW5Td+4zcyPj6eGTNm5Ic//GGeeeaZqz25VB599NGiJwAAAABUTU2+A2xoaChJMm/evAnnIyMjGRgYSFdX14Tz97///Tlw4EB+8YtfpKenJ7NmzXrB3zh58mSGh4dTV1dXveFTwJYtW7J8+fLL3uW1Z8+e3H777ZNe7+zszLFjx3LfffddiYkAAAAAl9XV1ZXDhw+/6M/X5B1gTU1NSZLBwcEJ57t27crZs2ezdOnSCeef+cxn8vTTT+dzn/tcbrrppqu2EwAAAIDaVzc+Pj5e9IhfNjY2lje96U05e/Zsdu/enebm5vT19eXIkSM5c+ZMHnnkkdx4443P+90lS5akt7c3CxYsuOxvrFixIknS399f5fW179y5cxkeHp70emtra06fPj3p9cbGxlx77bVXYhoAAABA1dXkHWDTpk1Lb29v2tra0t3dnQ0bNqSpqSk9PT2pr69Pe3v7pN/97//+7zzxxBNXb2wJ3X333UVPAAAAAKiamnwHWJIsWLAgR48enXC2fv36tLa2pqGhIUny5JNP5ktf+lJuvvnm1NfX51Of+lR+8IMfVDwiyUuzdu3aoicAAAAAVE1N3gE2mRMnTqSjo2PC2Sc+8Ym89rWvzdy5c/PZz342hw8fvvQOMX41ra2tRU8AAAAAqJqavQPsl124cCGDg4PZvHnzpbPXvOY1OXbsWIGrAAAAAKh1UyaAzZkzJxcvXix6BgAAAABTzJR6BJKr47l/IRMAAACgDAQwKuzdu7foCQAAAABVI4BRobu7u+gJAAAAAFUjgFGhv7+/6AkAAAAAVSOAAQAAAFBqAhgAAAAApSaAUeH06dNFTwAAAACoGgGMCg8++GDREwAAAACqRgCjwvbt24ueAAAAAFA1AhgAAAAApSaAAQAAAFBqAhgV7r///qInAAAAAFSNAEaFtra2oicAAAAAVI0ARoXOzs6iJwAAAABUjQAGAAAAQKnVFz2Aq2/OnDlpaGi47Geam5snvTZ9+vRqTwIAAAC4YgSwl6Fnnnkmo6Ojk15ftmxZRkZGJr1eX1+f2bNnX4lpAAAAAFUngL0MjY6OZnh4eNLrDz/88GWvNzY2XolZAAAAAFeEd4ABAAAAUGoCGAAAAAClJoBRoa+vr+gJAAAAAFUjgAEAAABQagIYFVavXl30BAAAAICqEcAAAAAAKDUBDAAAAIBSE8Co0NPTU/QEAAAAgKoRwKhw2223FT0BAAAAoGoEMCosX7686AkAAAAAVSOAUeH8+fNFTwAAAACoGgEMAAAAgFITwKiwaNGioicAAAAAVI0ARoUDBw4UPQEAAACgamo6gI2NjWX37t1paWnJrFmzsmTJkgwMDGThwoXZtGlT0fNK66677ip6AgAAAEDV1HQA27hxY+655558+MMfzhe+8IWsXbs269aty+OPP56Ojo6i55VWb29v0RMAAAAAqqZmA9j+/fuzb9++HDp0KHfccUdWrlyZrVu3ZtmyZRkdHa0IYP/0T/+Uurq69PX1FbQYAAAAgFpUswFs586dWbVqVTo7Oyecz58/PzNmzMjixYsvnf385z/Pxz72sbS1tV3tmQAAAADUuJoMYENDQzl16lTWrFlTce3MmTNpa2vLzJkzL53t3Lkzt956a5qamq7mzNIaGBgoegIAAABA1dQXPeD5DA0NJUnmzZs34XxkZCQDAwPp6uq6dPbEE0/k0KFDOXHiRA4dOvSif+PkyZMZHh5OXV1ddUZPEVu2bMny5csvG7kee+yxvP71r5/0emdnZ44dO5b77rvvSkwEAAAAuKyurq4cPnz4RX++Ju8Ae+5OrsHBwQnnu3btytmzZ7N06dJLZ1u2bMk999yTGTNmXNWNZfZSQiIAAABAravJO8BuuOGGtLe3Z8eOHWlsbExzc3P6+vpy5MiRJLn0AvwvfelL+elPf5p3vetdL/k3nnuHWH9/f9V2TxXnzp3LwoULJ72+Z8+ebNq0adLrjY2NueWWW3LvvfdeiXkAAAAAVVWTd4BNmzYtvb29aWtrS3d3dzZs2JCmpqb09PSkvr4+7e3tSZJjx47lkUceSVNTU5qamvJv//Zv2bhxY3bu3Fnw3wAAAACAWlGTd4AlyYIFC3L06NEJZ+vXr09ra2saGhqS/K/HHz/0oQ9dur5mzZps2LAhf/RHf3RVt5bN3XffXfQEAAAAgKqp2QD2fE6cOJGbbrrp0n9fc801ueaaay7998yZM9PY2DjhjJdu7dq1RU8AAAAAqJqafATy+Vy4cCGDg4MTXoD/y/r7+7N69eqruKqcWltbi54AAAAAUDVT5g6wOXPm5OLFi0XPAAAAAGCKmTJ3gAEAAADAr0IAo8KKFSuKngAAAABQNQIYFfbu3Vv0BAAAAICqEcCo0N3dXfQEAAAAgKoRwKjQ399f9AQAAACAqhHAAAAAACg1AQwAAACAUhPAqHD69OmiJwAAAABUjQBGhQcffLDoCQAAAABVI4BRYfv27UVPAAAAAKgaAQwAAACAUhPAAAAAACg1AYwK999/f9ETAAAAAKpGAKNCW1tb0RMAAAAAqkYAo0JnZ2fREwAAAACqpr7oAVx99fX1aWxsvOxnLne9vt7/bQAAAICpQ8l4GXqh+LV9+/Zce+21V2kNAAAAwJXlEUgq/OVf/mXREwAAAACqRgADAAAAoNQEMAAAAABKTQADAAAAoNQEMAAAAABKTQADAAAAoNQEMAAAAABKTQADAAAAoNQEMAAAAABKTQADAAAAoNQEMAAAAABKTQADAAAAoNQEMAAAAABKTQADAAAAoNQEMAAAAABKTQADAAAAoNRqOoCNjY1l9+7daWlpyaxZs7JkyZIMDAxk4cKF2bRpU9HzAAAAAJgC6osecDkbN27MwYMHs23btnR0dOT48eNZt25dzp8/ny1bthQ9DwAAAIApoGYD2P79+7Nv37709/ens7MzSbJy5cp8/etfz0MPPZSOjo4kyYoVK/Lv//7vmT59epJk9erV+fSnP13UbAAAAABqTM0GsJ07d2bVqlWX4tdz5s+fnxkzZmTx4sWXzh544IGsXr36ak8EAAAAYAqoyXeADQ0N5dSpU1mzZk3FtTNnzqStrS0zZ84sYBkAAAAAU01N3gE2NDSUJJk3b96E85GRkQwMDKSrq2vCeXd3dzZv3pzf/d3fzd/+7d+mpaXlBX/j5MmTGR4eTl1dXfWGAwAAAHDFdXV15fDhwy/68zV5B1hTU1OSZHBwcML5rl27cvbs2SxdunTC2RNPPJEnnngiS5cuzbve9a6Mjo5e1b0AAAAA1K668fHx8aJH/LKxsbG86U1vytmzZ7N79+40Nzenr68vR44cyZkzZ/LII4/kxhtvrPje+Ph45s6dm/7+/ixatOiyv7FixYokSX9//xX4GwAAAABQK2ryDrBp06alt7c3bW1t6e7uzoYNG9LU1JSenp7U19envb39eb9XV1fnkUYAAAAAJqjJd4AlyYIFC3L06NEJZ+vXr09ra2saGhqSJE899VS++tWvZvny5RkfH8+OHTvyqle9KgsWLChiMgAAAAA1qCbvAJvMiRMn0tHRcem/n3322fz5n/95mpqa0tzcnBMnTuTQoUOpr6/ZrgcAAADAVTZlStGFCxcyODiYzZs3Xzr7rd/6rXzta18rcBUAAAAAtW7KBLA5c+bk4sWLRc8AAAAAYIqZUo9AAgAAAMBLJYABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGoCGAAAAAClJoABAAAAUGo1HcDGxsaye/futLS0ZNasWVmyZEkGBgaycOHCbNq0qeh5AAAAAEwB9UUPuJyNGzfm4MGD2bZtWzo6OnL8+PGsW7cu58+fz5YtW4qeBwAAAMAUULN3gO3fvz/79u3LoUOHcscdd2TlypXZunVrli1bltHR0XR0dCRJxsfHs3379rz2ta/NNddck7e97W0FLwcAAACgltTsHWA7d+7MqlWr0tnZOeF8/vz5mTFjRhYvXpwkuffee/ONb3wj3/zmN9PU1JRvfetbRcwFAAAAoEbVZAAbGhrKqVOncvvtt1dcO3PmTNra2jJz5sxcvHgxe/bsyde+9rXMnTs3SbJ06dKrPRcAAACAGlazASxJ5s2bN+F8ZGQkAwMD6erqSvK/YlhdXV0+8YlP5JOf/GTmzp2bv/qrv8q73/3uF/yNkydPZnh4OHV1dVXfDwAAAMCV09XVlcOHD7/oz9fkO8CampqSJIODgxPOd+3albNnz166y+sHP/hBvv/972dkZCRDQ0PZu3dvPvjBD+aJJ5642pMBAAAAqFE1eQfYDTfckPb29uzYsSONjY1pbm5OX19fjhw5kiSXXoA/e/bsJMlHP/rRzJw5M29961vT2dmZ/v7+3HrrrZf9jefeIdbf33/F/h4AAAAAFK8m7wCbNm1aent709bWlu7u7mzYsCFNTU3p6elJfX192tvbkyQtLS2ZMWNGwWsBAAAAqGV14+Pj40WPeLHWr1+fb33rW/n2t7996WzNmjW5/vrrs2PHjnzta1/LqlWr8s1vfjPXX3/9Zf+sFStWJHEHGAAAAEDZ1eQjkJM5ceJEbrrppglne/fuza233ppXv/rVaW5uzqc//ekXjF8AAAAAvHxMmQB24cKFDA4OZvPmzRPOm5qa8i//8i8FrQIAAACg1k2ZADZnzpxcvHix6BkAAAAATDE1+RJ8AAAAAKgWAQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACg1AQwAAACAUhPAAAAAACi1mg5gY2Nj2b17d1paWjJr1qwsWbIkAwMDWbhwYTZt2lT0PAAAAACmgPqiB1zOxo0bc/DgwWzbti0dHR05fvx41q1bl/Pnz2fLli1FzwMAAABgCqjZALZ///7s27cv/f396ezsTJKsXLkyX//61/PQQw+lo6MjZ86cyaJFiy59Z3x8PD//+c9z4sSJdHR0FDUdAAAAgBpSs49A7ty5M6tWrboUv54zf/78zJgxI4sXL851112XCxcuXPrfP/7jP+a6667L0qVLC1oNAAAAQK2pyQA2NDSUU6dOZc2aNRXXzpw5k7a2tsycObPi2mc+85m8733vS11d3dWYCQAAAMAUUJOPQA4NDSVJ5s2bN+F8ZGQkAwMD6erqqvjO008/nc9//vP5yle+8qJ+4+TJkxkeHhbLAAAAAKaYrq6uHD58+EV/vibvAGtqakqSDA4OTjjftWtXzp49+7yPOD700ENpaWnJG9/4xquyEQAAAICpoSbvALvhhhvS3t6eHTt2pLGxMc3Nzenr68uRI0eS5HlfcP+Zz3wm73//+1/0byxevDhJ0t/fX5XNAAAAANSmmrwDbNq0aent7U1bW1u6u7uzYcOGNDU1paenJ/X19Wlvb5/w+R/+8Ifp7+/PunXrCloMAAAAQK2qyTvAkmTBggU5evTohLP169entbU1DQ0NE84/+9nP5i1veUte97rXXc2JAAAAAEwBNXkH2GROnDhRlccfAQAAAHj5qNk7wH7ZhQsXMjg4mM2bN1dc++pXv1rAIgAAAACmgikTwObMmZOLFy8WPQMAAACAKWZKPQIJAAAAAC+VAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqQlgAAAAAJSaAAYAAABAqdV0ABsbG8vu3bvT0tKSWbNmZcmSJRkYGMjChQuzadOmoucBAAAAMAXUFz3gcjZu3JiDBw9m27Zt6ejoyPHjx7Nu3bqcP38+W7ZsKXoeAAAAAFNAzd4Btn///uzbty+HDh3KHXfckZUrV2br1q1ZtmxZRkdH09HRkST5+te/nptuuinXXHNNFixYkAMHDhS8HAAAAIBaUrMBbOfOnVm1alU6OzsnnM+fPz8zZszI4sWLkyQf+MAH8p73vCdPPfVUPvWpT+WDH/xgfvzjHxcxGQAAAIAaVJMBbGhoKKdOncqaNWsqrp05cyZtbW2ZOXNmkuS73/1u1q1bl2nTpqWzszPXXntt/vM///NqTwYAAACgRtXkO8CGhoaSJPPmzZtwPjIykoGBgXR1dV06u+222/LAAw/kzjvvzMDAQMbGxtLW1vaCv3Hy5MkMDw+nrq6uuuMBAAAAuKK6urpy+PDhF/35mrwDrKmpKUkyODg44XzXrl05e/Zsli5deunsne98Z/7+7/8+s2bNSldXV/bs2ZPZs2df1b0AAAAA1K6avAPshhtuSHt7e3bs2JHGxsY0Nzenr68vR44cSZJLL8B/8sknc8stt+SBBx7IzTffnEcffTRdXV1paWnJG9/4xsv+xnPvEOvv77+ifxcAAAAAilWTd4BNmzYtvb29aWtrS3d3dzZs2JCmpqb09PSkvr4+7e3tSZLHH388jY2Nede73pXp06envb09N910U7785S8X/DcAAAAAoFbUZABLkgULFuTo0aP52c9+ljNnzuSee+7JyZMn09ramoaGhkufeeqpp3LkyJGMj4/n1KlTOXbs2Ave/QUAAADAy0fNBrDnc+LEiUuPPybJK1/5ynz2s5/NX/zFX+Saa67JzTffnDvvvDNvfetbC1wJAAAAQC2pyXeAPZ8LFy5kcHAwmzdvnnD+zne+M+985zsLWgUAAABArZsyAWzOnDm5ePFi0TMAAAAAmGKm1COQAAAAAPBSCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlNqUD2Dj4+O57777snDhwsyePTttbW154IEHip4FAAAAQI2Y8gHszjvvzJ133pn3vve9eeihh/L2t789H/jAB/K5z32u6GkAAAAA1IC68fHx8aJH/KqeeuqpXHvttfmzP/uz7Ny589L5H/zBH+R1r3tdPv7xj0/63Te0tefnI7/I+z78Z1djKgAAAABVsvCG12XDmv/jRX++/gpuueIeeeSRPPvss3nPe94z4fzgwYMFLQIAAACg1kzpAPbkk08mSebOnfuSvzvvtxqTJH/9f2+q6iYAAAAAasuUDmCvfOUrkyTnz5/Pddddd+n80UcfzU9+8pO85S1vmfS789/4O/nhuSfz5//PJ6/4TgAAAACq52X1COSb3/zm1NfX5/Of/3w6OjounX/kIx/JhQsXcuzYsUm/+3ud/3v+38e/dzVmAgAAAFCgKR3Arr322vT09GTnzp2ZNm1abrzxxnzxi1/MF7/4xRw+fPiy330plRAAAACAqWtK/yuQSTI2NpadO3fmU5/6VH70ox/lDW94Q7Zv3553v/vdRU8DAAAAoAZM+QAGAAAAAJczregBAAAAAHAlCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECpCWAAAAAAlJoABgAAAECp/X88c8UuEViYWwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1591.6x1288.28 with 1 Axes>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "size_reduce = 2\n",
    "circuit_init = input_data(n,x_train[0])\n",
    "circuit_vqc = vqc(n,num_layers,params)\n",
    "circuit_swap_test = swap_test(size_reduce)\n",
    "\n",
    "circuit_full = QuantumCircuit(n+size_reduce+1,1)\n",
    "\n",
    "circuit_full = circuit_full.compose(circuit_init,[i for i in range(size_reduce+1,n+size_reduce+1)])\n",
    "circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(size_reduce+1,n+size_reduce+1)])\n",
    "circuit_full = circuit_full.compose(circuit_swap_test,[i for i in range(2*size_reduce+1)])\n",
    "circuit_full.draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2856a6a4",
   "metadata": {},
   "source": [
    "Qiskit tiene la propeidad de generar de un circuito cuántico su inverso que es necesario para nosotros al momento de decodificar la disminuci´no de variables al tamaño original del vector de estado. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "f6f16fdd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1110x445.48 with 1 Axes>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vqc(n,num_layers,params).inverse().draw(output = \"mpl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1e79f93",
   "metadata": {},
   "source": [
    "## Comprimir datos\n",
    "\n",
    "En este trabajo al no encontrar una manera correecta de usar los optimziadores para un circuito cuántico que utiliza le mapeo por amplitud, se utilizo la libreria scikitlearn por su método de optimización que es el mismo que usan en Qiskit. y también se usa un shuffle para que en cada iteración logre ocupar algunas imagenes nuevas en cada iteración."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "7be5aa04",
   "metadata": {},
   "outputs": [],
   "source": [
    "from random import shuffle\n",
    "from scipy.optimize import minimize "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b7e2e85",
   "metadata": {},
   "source": [
    "Se uso para identificar el costo de cada entrada por el valor esperado en el eje z, es decir, $\\langle z \\rangle $, el cuál se define \n",
    "\n",
    "$\\langle Z \\rangle = \\langle q | Z | q\\rangle =\\langle q|0\\rangle\\langle 0|q\\rangle - \\langle q|1\\rangle\\langle 1|q\\rangle\n",
    "=|\\langle 0 |q\\rangle|^2 - |\\langle 1 | q\\rangle|^2 $\n",
    "\n",
    "\n",
    "pero al adecuarl oal criterio del Swap test debe quedar como:\n",
    "\n",
    "\n",
    "$1 -\\langle Z \\rangle = 1 - \\langle q | Z | q\\rangle = 1- [\\langle q|0\\rangle\\langle 0|q\\rangle - \\langle q|1\\rangle\\langle 1|q\\rangle] = 1 - [|\\langle 0 |q\\rangle|^2 - |\\langle 1 | q\\rangle|^2] = 1 - |\\langle 0 |q\\rangle|^2 + |\\langle 1 | q\\rangle|^2 $\n",
    "\n",
    "para mayor información se puede observar en https://qiskit.org/textbook/ch-labs/Lab02_QuantumMeasurement.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "9586bc24",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " cost:  0.2796875\n",
      "[ 0.62254674  0.26851417  0.18249782  1.72136515  2.0098116  -0.0834305\n",
      " -1.52525618  0.47399993 -0.21121215 -0.16145395]\n"
     ]
    }
   ],
   "source": [
    "def objective_function(params):\n",
    "    costo = 0\n",
    "    shuffle(x_train)\n",
    "    lenght= 5\n",
    "    for i in range(lenght):\n",
    "        circuit_init = input_data(n,x_train[i])\n",
    "        circuit_vqc = vqc(n,num_layers,params)\n",
    "        circuit_swap_test = swap_test(size_reduce)\n",
    "\n",
    "        circuit_full = QuantumCircuit(n+size_reduce+1,1)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(size_reduce+1,n+size_reduce+1)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(size_reduce+1,n+size_reduce+1)])\n",
    "        circuit_full = circuit_full.compose(circuit_swap_test,[i for i in range(2*size_reduce+1)])\n",
    "        circuit_full.measure(0, 0)\n",
    "        #qc.draw()\n",
    "        shots= 8192\n",
    "        job = execute( circuit_full, Aer.get_backend('qasm_simulator'),shots=shots )\n",
    "        counts = job.result().get_counts()\n",
    "        probs = {}\n",
    "        for output in ['0','1']:\n",
    "            if output in counts:\n",
    "                probs[output] = counts[output]/shots\n",
    "            else:\n",
    "                probs[output] = 0\n",
    "        costo += (1 +probs['1'] -  probs['0'])\n",
    "    \n",
    "    return costo/lenght\n",
    "\n",
    "for i in range(1):    \n",
    "    minimum = minimize(objective_function, params, method='COBYLA', tol=1e-6)\n",
    "    params = minimum.x\n",
    "    print(\" cost: \",objective_function(params))\n",
    "    print(params)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9dee656",
   "metadata": {},
   "source": [
    "Al finalizar las iteraciones necesarias, depende del número del conjunto de instancia seleccionadas, se pasa por vector de estado por el complejo conjugado de nuestro circuito cuántico, dónde [6] y [7] mencionaban que nos debe dar la información original. Esto se realiza a todo el conjunto de entrenamiento como de prueba."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "30a754be",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compress_result_test(params):\n",
    "    reduce = []\n",
    "    for i in range(len(x_test)):\n",
    "        circuit_init = input_data(n,x_test[i])\n",
    "        circuit_vqc = vqc(n,num_layers,params)\n",
    "\n",
    "        circuit_full = QuantumCircuit(n,n-size_reduce)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(n)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(n)])\n",
    "        len_cf = len(circuit_full)\n",
    "        for i in range(n-size_reduce):\n",
    "            circuit_full.measure(size_reduce+i, i)\n",
    "        job = execute( circuit_full, Aer.get_backend('qasm_simulator'),shots=8192 )\n",
    "        result = job.result().get_counts()\n",
    "        probs = {k: np.sqrt(v / 8192) for k, v in result.items()}\n",
    "        reduce.append(probs)\n",
    "        \n",
    "    return reduce\n",
    "        \n",
    "reduce_img =compress_result_test(params)\n",
    "test_reduce = []\n",
    "for i in reduce_img:\n",
    "    index_image = []\n",
    "    for j in range(16):\n",
    "        bin_index = bin(j)[2:]\n",
    "        while len(bin_index) <4:\n",
    "            bin_index = '0'+bin_index \n",
    "        try:   \n",
    "            index_image.append(i[bin_index])\n",
    "        except:\n",
    "            index_image.append(0)\n",
    "    \n",
    "    test_reduce.append(np.array(index_image))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "9cbc6e18",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compress_result_train(params):\n",
    "    reduce = []\n",
    "    for i in range(len(x_train)):\n",
    "        circuit_init = input_data(n,x_train[i])\n",
    "        circuit_vqc = vqc(n,num_layers,params)\n",
    "\n",
    "        circuit_full = QuantumCircuit(n,n-size_reduce)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(n)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(n)])\n",
    "        len_cf = len(circuit_full)\n",
    "        for i in range(n-size_reduce):\n",
    "            circuit_full.measure(size_reduce+i, i)\n",
    "        job = execute( circuit_full, Aer.get_backend('qasm_simulator'),shots=8192 )\n",
    "        result = job.result().get_counts()\n",
    "        probs = {k: np.sqrt(v / 8192) for k, v in result.items()}\n",
    "        reduce.append(probs)\n",
    "        \n",
    "    return reduce\n",
    "        \n",
    "reduce_img =compress_result_train(params)\n",
    "train_reduce = []\n",
    "for i in reduce_img:\n",
    "    index_image = []\n",
    "    for j in range(16):\n",
    "        bin_index = bin(j)[2:]\n",
    "        while len(bin_index) <4:\n",
    "            bin_index = '0'+bin_index \n",
    "        try:   \n",
    "            index_image.append(i[bin_index])\n",
    "        except:\n",
    "            index_image.append(0)\n",
    "    \n",
    "    train_reduce.append(np.array(index_image))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7472a411",
   "metadata": {},
   "source": [
    "En este punto  se muestra las primeras 5 imagenes del conjunto de prueba de tamaño 8x8 como se reducen a un tamaño de 4x4 cada una."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "4d1a3e7a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1db360c24c0>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "\n",
    "#subplot(r,c) provide the no. of rows and columns\n",
    "f, axarr = plt.subplots(5,1) \n",
    "\n",
    "# use the created array to output your multiple images. In this case I have stacked 4 images vertically\n",
    "axarr[0].imshow(x_test[0].reshape(8,8)*255)\n",
    "axarr[1].imshow(x_test[1].reshape(8,8)*255)\n",
    "axarr[2].imshow(x_test[2].reshape(8,8)*255)\n",
    "axarr[3].imshow(x_test[3].reshape(8,8)*255)\n",
    "axarr[4].imshow(x_test[4].reshape(8,8)*255)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "0a6f25e6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1db36449ee0>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "#subplot(r,c) provide the no. of rows and columns\n",
    "f, axarr = plt.subplots(5,1) \n",
    "\n",
    "# use the created array to output your multiple images. In this case I have stacked 4 images vertically\n",
    "axarr[0].imshow(test_reduce[0].reshape(4,4)*255)\n",
    "axarr[1].imshow(test_reduce[1].reshape(4,4)*255)\n",
    "axarr[2].imshow(test_reduce[2].reshape(4,4)*255)\n",
    "axarr[3].imshow(test_reduce[3].reshape(4,4)*255)\n",
    "axarr[4].imshow(test_reduce[4].reshape(4,4)*255)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "066d384e",
   "metadata": {},
   "source": [
    "###  Descomprimir datos\n",
    "\n",
    "Aquí recordando la aplicación del complejo conjugado de nuestra propuesta  de red tensorial  debemos acercarnos al valor original de entrada $|\\phi \\rangle$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "e9bbfd0b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1110x445.48 with 1 Axes>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vqc(n,num_layers,params).inverse().draw(output = \"mpl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "2f4173f1",
   "metadata": {},
   "outputs": [],
   "source": [
    "def decoder_result_test(params):\n",
    "    reduce = []\n",
    "    for i in range(len(test_reduce)):\n",
    "\n",
    "        circuit_init = input_data(6,np.concatenate((np.zeros(48), test_reduce[i]), axis=0))\n",
    "        circuit_vqc = vqc(n,num_layers,params).inverse()\n",
    "\n",
    "        circuit_full = QuantumCircuit(n,n)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(n)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(n)])\n",
    "        job = execute( circuit_full, Aer.get_backend('statevector_simulator') )\n",
    "        result = job.result().get_statevector()\n",
    "        reduce.append(result)\n",
    "    return reduce\n",
    "        \n",
    "decoder =decoder_result_test(params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "fe80a3ff",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1db334b5100>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "\n",
    "#subplot(r,c) provide the no. of rows and columns\n",
    "f, axarr = plt.subplots(5,1) \n",
    "\n",
    "# use the created array to output your multiple images. In this case I have stacked 4 images vertically\n",
    "axarr[0].imshow(decoder[0].real.reshape(8,8)*255)\n",
    "axarr[1].imshow(decoder[1].real.reshape(8,8)*255)\n",
    "axarr[2].imshow(decoder[2].real.reshape(8,8)*255)\n",
    "axarr[3].imshow(decoder[3].real.reshape(8,8)*255)\n",
    "axarr[4].imshow(decoder[4].real.reshape(8,8)*255)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "bd1ae5a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "def decoder_result_train(params):\n",
    "    reduce = []\n",
    "    for i in range(len(train_reduce)):\n",
    "        circuit_init = input_data(n,np.concatenate((np.zeros(48), train_reduce[i]), axis=0))\n",
    "        circuit_vqc = vqc(n,num_layers,params).inverse()\n",
    "\n",
    "        circuit_full = QuantumCircuit(n,n)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(n)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(n)])\n",
    "        job = execute( circuit_full, Aer.get_backend('statevector_simulator') )\n",
    "        result = job.result().get_statevector()\n",
    "        reduce.append(result)\n",
    "    return reduce\n",
    "        \n",
    "decoder_train =decoder_result_train(params)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96bac096",
   "metadata": {},
   "source": [
    "# métricas para comparar imagenes\n",
    "\n",
    "\n",
    "De cada una de las imágenes tanto de prueba como entrenamiento se realizará las siguientes métricas para validar la capacidad de nuestro autoencoder entre las imagenes de entrada y descomprimidas. \n",
    "\n",
    "- Error Cuadrático medio (o por sus siglas en inglés  MSE)\n",
    "\n",
    "$MSE=\\frac{1}{m n} \\sum_{i=0}^{m-1} \\sum_{j=0}^{n-1}[I(i, j)-K(i, j)]^{2},$ \n",
    "\n",
    "donde $m$ es el  alto de la imágen $I$, n el ancho de la imagen $K$ e $i$,$j$ las posiciones $x,y$ de los píxeles de las imágenes; entre más cercano a 0 sea su valor es mejor.\n",
    "\n",
    "- Proporción Máxima de Señal a Ruido (o por sus siglas en inglés  PSNR)\n",
    "\n",
    "$PSNR = 10×log_{10}(\\frac{(mxn)^2}{MSE},$\n",
    "\n",
    "donde $m$ el  alto de la imagen $I$, n el ancho de la imagen $K$ y $MSE$ el error cuadrático medio;entre más alto su valor es mejor.\n",
    "\n",
    "\n",
    "- Semejanza Estructural (o por sus siglas en inglés SSIM)\n",
    "\n",
    "$    \\operatorname{SSIM}(x, y)=\\frac{\\left(2 \\mu_{x} \\mu_{y}+c_{1}\\right)\\left(2 \\sigma_{x y}+c_{2}\\right)}{\\left(\\mu_{x}^{2}+\\mu_{y}^{2}+c_{1}\\right)\\left(\\sigma_{x}^{2}+\\sigma_{y}^{2}+c_{2}\\right)},$\n",
    "\n",
    "donde $\\mu$ es el promedio, $\\sigma$ es la varianza y $c$ es la covarianza \\cite{c1}}; peor caso -1, mejor caso 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "3d59c9e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def mse(imageA, imageB):\n",
    "    err = np.sum((imageA.astype(\"float\") - imageB.astype(\"float\")) ** 2)\n",
    "    err /= float(imageA.shape[0] * imageA.shape[1])\n",
    "    return err"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "1bdd3c6f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from skimage.metrics import structural_similarity as ssim"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79f45d65",
   "metadata": {},
   "source": [
    "# Comparar el conjunto de entrenamiento\n",
    "\n",
    "Se realiza los resultados de las tres métricas en el conjunto de entrenamiento"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "6fc065a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import math \n",
    "ssim_list = []\n",
    "mse_list = []\n",
    "psnr_list = []\n",
    "\n",
    "for i in range(len(x_train)):\n",
    "    test_img = x_train[i].reshape(8,8)*255\n",
    "    decoded_img = decoder_train[i].real.reshape(8,8)*255\n",
    "    Y = float(mse(decoded_img,test_img))\n",
    "    ssim_list.append(ssim(decoded_img.astype(\"float\"),test_img.astype(\"float\")))\n",
    "    mse_list.append(Y)\n",
    "    aux = (64**2)/Y\n",
    "    psnr_list.append(10*math.log10(aux))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "f61bda52",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "plt.plot(mse_list)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "ab3df8f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "plt.plot(psnr_list)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "47c3a26c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAD7CAYAAABnoJM0AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAABD5klEQVR4nO3deVzUdf4H8NfADAhyKc6AoamFaZJIioVkmGagHOqPbKu1aHNXs9o0LEtWY5daz3Wj7dgOU9tKAiuDSK7UTA3UIC8K8coLdRgOuY8Z5vv7A2aYGWbm+/0OczrvZ48eD2c+3+93Pt/hO9/35/4KGIZhQAghxGm52DoDhBBCbIsCASGEODkKBIQQ4uQoEBBCiJOjQEAIIU6OAgEhhDg5CgSEEOLkhLbOgCnq61ugVJo2/cHf3wu1tc1mzpH1UP5ti/JvW5R/07i4CDBo0ECD6Q4ZCJRKxuRAoNrfkVH+bYvyb1uUf/OjpiFCCHFyFAgIIcTJUSAghBAnxykQ5ObmIjY2FtHR0di+fXuf9IqKCiQmJiImJgarVq2CQqHQSv/tt99w1113qV83NjZi8eLFmD17NhYsWACZTNbP0yCEEGIq1kAglUqRnp6OjIwMZGdnIysrC2fPntXaZsWKFUhNTUVhYSEYhsGOHTvUaW1tbXjjjTcgl8vV77311lsIDw9Hfn4+HnnkEaxZs8aMp0QIIYQP1kBQXFyMiIgI+Pn5wdPTEzExMSgoKFCnV1VVob29HWFhYQCAxMRErfT169fjqaee0jrmvn37kJCQAACIj4/H/v37tQIFIYQQ62ENBNXV1RCLxerXEokEUqnUYLpYLFan79mzB+3t7Zg1a5bBYwqFQnh5eaGurq5/Z0KIDT2f/iMydp+2dTYIMQnrPAKlUgmBQKB+zTCM1mtD6TKZDO+//z4++eQT1kwwDAMXF+791v7+Xpy31Ucs9u7X/rZG+bctfflv6+jC7tIrWPb4JBvkiJ+b8ft3JPaYf9ZAEBgYiNLSUvVrmUwGiUSila7Z2VtTUwOJRIJ9+/bhxo0bWLBggTpt7ty52L59OyQSCWpqahAYGAiFQoGWlhb4+flxznRtbbPJkzLEYm/IZE0m7WsPKP+2xZZ/ez+3m/37t3e2yr+Li8BoAZq1GB4ZGYmSkhLU1dWhra0NRUVFiIqKUqcHBQXB3d0dZWVlAICcnBxERUXhkUcewe7du5GTk4OcnBx1mpeXF6ZNm4bs7GwAQF5eHsLDwyESifpznoQQQkzEGggCAgKQnJyMpKQkzJs3D/Hx8QgNDcWiRYtw8uRJAMCmTZuwbt06zJo1C62trUhKSjJ6zGXLluHYsWOIi4tDRkYGUlNTzXM2hBBCeBM44sPrqWmI8m8rhvK/cP1eAMDWlTOsnSVebtbv31E4bNMQIYSQmxsFAkIIcXIUCAghxMlRICCEECdHgYAQQpwcBQJCCHFyFAgIIcTJUSAghBAnR4GAEEKcHAUCQghxchQICCHEyVEgIIQQJ0eBgBBCnBwFAkIIcXIUCAghxMlRICCEECdHgYAQQpwcp0CQm5uL2NhYREdHY/v27X3SKyoqkJiYiJiYGKxatQoKhQIAUFpaisTERCQkJGDJkiVoaGgAAFy4cAFPPPEEEhIS8OSTT+L333834ykRQgjhgzUQSKVSpKenIyMjA9nZ2cjKysLZs2e1tlmxYgVSU1NRWFgIhmGwY8cOAEBKSgo2btyI3NxcBAcHY8uWLer3ExMTkZubi5deegkvvvii+c+MEEIcwA+/XMGV6mab5oE1EBQXFyMiIgJ+fn7w9PRETEwMCgoK1OlVVVVob29HWFgYACAxMVGdnpeXh+DgYMjlckilUvj4+ADorkHMmjULABAWFobq6mpcvnzZ3OdGCCF277Oi00jdesSmeWANBNXV1RCLxerXEokEUqnUYLpYLFani0QiVFZWYtq0aTh8+DDi4uIAAOPGjcOuXbsAACUlJbhx4wZkMpl5zogQQggvQrYNlEolBAKB+jXDMFqv2dLHjBmD4uJiZGZmIjk5GZmZmVi/fj3eeOMNfPbZZ4iKisLYsWMhEok4Z9rf34vztvqIxd792t/WKP+2ZSz/jnBujpBHY27W/NvyvFgDQWBgIEpLS9WvZTIZJBKJVrpmab6mpgYSiQQdHR04cOAAZs6cCQCYM2cONmzYAABQKBR477334ObmBrlcjqysLAwbNoxzpmtrm6FUMpy31yQWe0MmazJpX3tA+bcttvzb+7nd7N+/vTOWf0uel4uLwGgBmrVpKDIyEiUlJairq0NbWxuKiooQFRWlTg8KCoK7uzvKysoAADk5OYiKioJQKERaWhrKy8sBAPn5+Zg4cSIAID09HXv27AEAfPXVVxg/fjwGDRpk+lnamFLJIPen39HWobB1VgghhDfWGkFAQACSk5ORlJQEuVyO+fPnIzQ0FIsWLcLSpUsxfvx4bNq0CatXr0ZzczNCQkKQlJQEV1dXpKenIzU1FV1dXQgICMCaNWsAAC+//DJeffVVvPvuuwgICMC6dessfqKWVFpZjW8O/I4bzZ14MmaMrbNz0/n6x3PYVXIRH78yHS4uAvYdCCG8sAYCAEhISEBCQoLWe5s3b1b/e+zYsfjqq6/67BceHo6dO3f2eX/EiBHIzMzkm1e7JVcoAQDtnV02zsnNqfDIJQBAl5KhQECIBdDMYkIIcXIUCMzKtA5sYhxDXyshFkWBwAxUo2XpfmVZAmoVIsQiKBCYgQB0hyLE1k5fvoGF6/fi92uNts6Kw6FAYEB9UwdWbT6E2oZ27jtRlYAQmzl+tgYAUHGx3sY5cTwUCAw4ePIartW2Yt+xKvaNqWmIEOLAKBCYgaphiKFeTcLDxetNOHe1wdbZcHoNLZ1I++Rn1DXyqP1zUNfYjk/yT0HRpTTrcS2BAoE5UBcBMUHaJz9jzadlts7GTaeptZPX9sUnr+Hi9SbsLrti1nxsyz+F/cevOkRTFQUCA+jeToiD6fnRFh65bNfLvSi6lCj6+bJd1RQoELDYVXIR568aH4WgGjV0s7UMvfjOQby25bCts3HTfa/E8jrk1pnlf622BSs/LEGjnlqIocLk7tIryNxzBnt/4dD/aCUUCAzQHLP+z09LDW+Im3d8e2NLJ6pkLbbOBiHWYUKBI//wJVTXt+HYmRrDh9U5bmtPbaW9035qLRQIzMiSBdfy87U4eOKaBT/BfjlyoG3rUODTwkp09GMdqq9/PIfF/9pnvkw5gfqmDu4bW+r6cqDrlgKBOZm5DaOxtRNnr3SPKnlzx3Fszasw6/H74/TlG3jt48PotEIV3JGbhvIPX8K+o1X4vtT0R7HuKrloV+3J9kpzYucb/zNei9eHsVhRzv4vYAoEZiCwUJF17WdlWPu59qiSdjvpBNv+/WlU1bTgWm2rrbNi11RDio3dCqpvtGm93n/8Kg7/JjWwNTE3S/XxOVLfIQUCMzL292YYBleqm/u8X9PQhh8NTFqrrm/r856xsc6NrZ24Iuv7GZagurit0WxjrKRWeqoa//riqOUzYUEnzmq3L3+SfwoffvurjXJjms25v+LnU9U2zYPlSvSmcaQmTQoEBvAp5XPZcv/xq0jdegTlv9dqvb8x4yj+V1Bplo6j1z4+jNQtR/p9HG7s40f33+xyhxin7YiuVDdzrpmU/CrF+9nletPOXmlASfl1o/srupTIP2x6E1jlpXr8XGHZQNTQ0mnSOkb28UsxjgKBlVzqqQ1I67RL+U2tcgBADZ81jQxQHUufo2dk/NZNYqG6uC3VLKb3M3V+UXwnDhF+UrceMUvNZO3nZdj83W9Gtyk8cglf/nDO5CGVGzKO9vkNHakwb/Na6pbDJvU9OAJOgSA3NxexsbGIjo7G9u3b+6RXVFQgMTERMTExWLVqFRSK7tJtaWkpEhMTkZCQgCVLlqChobvjs6GhAYsWLcKcOXMwf/58VFTYTyeoikm3t36EfkuX5N/5+iTSPvlZ/Vpa34rWdsOBw5YWrt+LT/LZr4nrdQ7UP2GkoVhfE6CzUT3dz5zj/z/I+ZXTNaJeRp7l92usoGWUoePaUVWBNRBIpVKkp6cjIyMD2dnZyMrKwtmzZ7W2WbFiBVJTU1FYWAiGYbBjxw4AQEpKCjZu3Ijc3FwEBwdjy5YtAIBt27bhjjvuwLfffovnnnsOr7/+ugVOzfoYdLfvFpfb5zDP5rbeCznlw0P461sHTD+Yqo8A3VVmcweV/cft8zvki0uFydxLG9i7js4uq/Vl7TtaxXptmrtO29wmx4HjV3vXINO549tj1wFrICguLkZERAT8/Pzg6emJmJgYFBQUqNOrqqrQ3t6OsLAwAEBiYqI6PS8vD8HBwZDL5ZBKpfDx8QEAKJVKtLR0T1Rqa2vDgAEDzH1eBl2pbjJrEwmg/WPff/wqPv6Oew2HdweXma8iUxfaUudaACS/cxDL3/vJbHlyFG0dCixcvxc//OJcN3JN2/IqkPLRIV77vJ9TjtQtR/QOPVY9/9tcin6+jC27uP0ezdXZ/FHur9iWf8qhRtSxBoLq6mqIxWL1a4lEAqlUajBdLBar00UiESorKzFt2jQcPnwYcXFxAICFCxeipKQEU6dOxerVq7F06VKznRCbZzfsxYr3i9k3NOGGa3T1UTuqBmoytU1Wda6qr6lTbv5x7ifO1WLh+r0Gq+y2HpanmrRkTyX663WtZr+ZGnPgxDVIeTbRVV6+AQDoUvb9A35XfIF3HvSNxtN0o5mlL8nM/VyNPZ8nV3V82+lvX5OQbQOlUqnVIcgwjNZrtvQxY8aguLgYmZmZSE5ORmZmJt544w0sWLAASUlJOHr0KJKTk7Fr1y4MHDiQU6b9/b04bWeMWOxtNN1roLv2ax8PeLjr/7p8rzcBANzcetN1jz/AQwQA8PYeoJWm+V0ZypPW+wx73o2l66ZVN7SbdDxX1+4yxODBvX+zzbsq8Lc/3WP0WGz50/Tj8as6+3lBJHRVv65u6tRI43ZMoLsk7yZyhauLaTcA1We19/zOVd+FZtql640orZDC07P7Oho40J3X30Xfe2zn2Nwmx9/W78UDk4bhpT9O4vVZxvDZnss5qr72IUO84Dmg+3fh6enG+fN00981MFpJRSRyMXpML6/uv5GLqyua5UqMusWX8+d7DOj7uxb2XKOq68LH10NrH9W5eupcE3z/LubEGggCAwNRWtrbUy6TySCRSLTSZTKZ+nVNTQ0kEgk6Ojpw4MABzJw5EwAwZ84cbNiwAQCwZ88edb/A3XffDX9/f5w7dw6hoaGcMl1b2wylntIEHzJZk973L15vwoff/orwsRKt9z/f9Svm3X+b3n0ae5pXOjWGgOoev72nfb65uV0rTbMWYShPuu8b2k4zvbVdjte2HMHz/zcet93iY3Dfzk4Fp+Pp6uop7dTX95YGS05eYz2WWOxtdJvW9t7v8OhpmVaaTNYMkbD3pnvjRqtGmvHPVedbqcSijfsAAOkvTIXvwN4bkFyhxNWaFowINH4zU31WXV1386ZmiVSVtvw/+9HR2YXZEbcCAFpaOozmUV8a37+7qoZyrLJa77YNLZ0IHunP+bvi+rlct1WlqX66NTXN6sJVq8YIMGPH0Hf9dLIMvZbLlUaP2dzc/b0VHb6IosMX8eHLD2hdZ4bOAwDaevofmpp6f9cKRXeTl1LZ/RtpaGhTp4nF3mjpOdfWVu1rwlgeKy/Vo7lNgUljxAa3McbFRWC0AM3aNBQZGYmSkhLU1dWhra0NRUVFiIqKUqcHBQXB3d0dZWXdM2BzcnIQFRUFoVCItLQ0lJd3R+v8/HxMnDgRADB27Fjs3r0bAHDhwgVUV1dj1KhRJp2guX1z4Dyu17XiTE/1VYVL3DGpqYLvPhwKsfuPX8WZKw2ob+rAtz/9bnTbE+dqTRqGaalmmTNXbhj7VNb9G1o68VlRpcHx6F1dvcf4SGdo5PbvTyPtk5/N0oekav9WzS5tapWj5FfjY+n7y1jTZMWFOiS/cxAlJ6/2SWts7cTu0suO+2ClfmZb9yfV30Kmemi1GWcWb8g4ive+Odn/AxnAGggCAgKQnJyMpKQkzJs3D/Hx8QgNDcWiRYtw8mR3xjZt2oR169Zh1qxZaG1tRVJSElxdXZGeno7U1FTMnTsXhYWFWLNmDQBg/fr1+PrrrxEfH4/ly5djw4YN8Pa2XbVIRckwOHGuln1DdJceX/2gGCfO1Zit//bi9SZUXuqeHNWfCWaf5J/itf3feHb2AZrzCHjvajEZu08j+Z2D+OGXKqOrQarorlmvWm68heMIKEO/7w55V5+Z17vLrmBz7m9mfwqWPvrmdlyQdpc2Ky7Uo6G5AwvX78WX+7pH/23+9ldk7D6DXSUXLZ43Q4xdR0cqpDimM/u65kYbrxt2bUM73vryOPIO6TlHM1zDF6414t2dJ9GldMw1oVibhgAgISEBCQkJWu9t3rxZ/e+xY8fiq6++6rNfeHg4du7c2ef9kSNH4tNPP+WbV4trYOtU0lDf1A7ZjXZs//40HntwNOv2XC5Z1Tj/rStn4Lk393POS3+1m7IypoVKj6YeteZGG3aX2r7TVskwePbfP6pf635Ntlg8TlrXiqsay4lfren+d/6hS3jkgWA0t3UHxJ37zyM+cqTV82fMsbM1+CCnu+b27ov3A+heluWVD0oQGzEC8x+4nfUYAgGwNa8CFRfrceJcLSaPlUDs52F0H6WSwV82/oDZ996KR6YHs37GvmPdNS1etUk7qoDRzGIT6fsbcqlaW6IAve9oFTJ2n9abduJcLc5Vmf+5uMZmFl+rbcHWvApeJbaG5o5+NckcMWGdG7bcHT0tw8L1e7Fj79k+aa3tClTpGQvP5ZwvSZvwaWEl12ya5KeT19R9BikfHcJPLEs82BLDMNh/rG+TFQDc0FhOWtHz3Ta0dBfYKi7Wmfx5bFQBO//wJd7HV81bUP00dJs77agSrebUgaC1XY6X3vupXzdKQc9/gGkBnn8XQfdnad6gPi2sNFoaXvOZ8efi8i3cn7/aqJ4Nq++ifj/7Vxw8cY3XpKHkd3/iNqxXg7Gb6fGzNfhm/3lex9O1v2fUUsGRSyg9VY2F6/fi4vVGvLnjGNZ8VqouqWr6X4HxZrnMPWeRvuM49h1lH7b78n8Nz81gGAZvfXkcJ871bQJr61Bgy64KbMrUvxif5p9byTC4KOXXeWxux8/VolFn1m75+Vp8kKMzGqgn4y4C7bZ3LpevsZu/OW/MVbIW1DZqPwuh6GfTlyC3FqcNBIouJTZlHkN9UwfrjVIvzevKwJVUXd/K2hHbnxaWgiOXoDRwAD7LL/CdSKO1cB7Lr6issvsGWl3PMT88sqJq4tDnp/LryDVhTDrQ/WS2s1cacFyjvyj7YHen+7+3l6H8fJ3ByUI/nTRe8tZt69al+dyCukbjD1c5ca4Wb315Qv1adcNRXRP6mjqVSu1VcEs51KRqGtqw4r/F5puIqfE3bm6To6Wtb5/MmzuO4wjLInIW69vuR2R4d2dvhy6X9cMMBSiGYbBw/V7k6+vTsABOfQQ3o4+/+w0XrhsZrqUzakj3DqUOHgYumrYOBVZ+eAjubq54f/k00zPaJxfa+fjLhh/0bpelpzmDi055F9o6u7SGVRrD9ptRrV55UdoMySBPk/KkYsoP/5K0CbcG8BuI8OaO433eMxZ0zOmL3WdM3lcVCARG/io5+89pvebSP7T/+DXUNrbjp/JrmHOf+Ub3CQTA0v9wX+ZEd4CCuWYC8104UXajDR7uQnj1zA3SZO7Y9OW+c5gdMcLMR+3LaWsEbKUNNqp1e6rre0cvaN6onk/v7uxVP6LQTMUXhULZrxUQD5y42ncmqEbW/pV5FMnvHOR+QDsZNmQoF51yJZpaO3Ho1+s4dqYG6z7/RZ3WtzRmu96705dvQK7g3mlvUge/maVuOdznPbZZvpp0m1D00vOHVQc61Z+rn7+t7d/r719TqWts1+rkf/WDEqzezH+kHRuGYbD3lys2WQzSqWoE5noIve4DOI6ekRnY0rj6pg7eo0ikda0mrYmusi2vbxu25s/oXFX3sReu38vpeGxfmerYBYcvYbLOJD1rYMDgvW/KcbpPDQ+4JOW/8Jmx0jZXqs5OTeu3/4K7Rw/hfAxVQUNvbnhkkcswW0OuyPrWklK3HsH7L02Du8hVzx7aODU1aV6cDIP2ToV6dB3X27/u17Hyw0N4LznK4EoBul7+bzGmhg7Vek/dp9HPsoPm7uevNuLzotOouFiP5+bd1b8D8+RUgUA1TtwQY7+f74ov4rvii3hsRjAydZpdVO3CRquqeqLMSyYs1GbN9f8NOfRr71pTnRzXtelP8OoPhuke6ms2Fvz6j/bjpmwIW+mSrc/CFF1dDNC31UQP9ruo7iS8eo1aRH8qAjeaOzgHAgA4eEL/arhyE4YDG8q2qmlPX5+JpTlt05CuskoZajhM9vly3znWbW52mh3Rn5s4DPLC9caeUTjmGbFijr+LvkXQ+kvvBCYeNGsyus825kLzGRRccJ1MZxzL98hx/X9Au2+m7+bsz4O2tH49W1qnUGHLR31SIOjx3jcnseZTE0YPaTLTFWnOh3OYy9HTMr2PhNS9OTW2GH5ucv7hi3j41VwAvU0SuiXSHT+Y1sltDlyWDT5vgTkZxmQf6B0Cu/KDEr1NXMbIbvCrDR1nqSG0dSg43fyUDIOcg79rPQNDl7HBGvowjHbFmk+NQHMNK6B7fkL2gfNWW1aj/PdaXK1pQV7x7yg80jM3wY5W1XWqpiGb4vHX1ZyZai/e6RkWt3XlDKPbGXu04Zc/sJfajQ17ZdDdSb91VwVcOK4cyvZDP3WxHmNHDOJ0LHuwq+Rin+/o6GkZTpzntjRKf23Lq0BppQxBQ4yvFPzN/vPYVXIROQe117pqaZerB1DopnGhr9mE7Uq4KG3u0xf3r8xjAIDxt/nr3YdLC6yhodv6vJnVdyQaALuZXUyBwMpaO/ivIaQuQdiIsQte9wfD5UHy35deNrnU88b/fuZVyr1e12p0+41fHGUNbrak21x18nwtTp7Xnsfxzk6dxcjMcHPRXIdJ8xkBqpE+HSwjnPStW8QwDKd+MWO1iBfTtQtJDMPg1wvGrzljAzLYRgwZY2jotiOipiGejJUUyn9nn/L+9Y/8Z7uaOieAjx9+uWL4oSDG+sBN+Kwvdp9R3+AaWzrxA4dZtip8mzr+V8Deh/E6z3Z0azpzhX9TlDmaFt/+6kSf94zdoLl4d+dJTg8w0vfZQPe1oqu+icMQVCP0TfoSwH4eeqTL+Oq8pqMaAU+KLv5XiCmdfIbsYllW2lSfFZlaMjIcCgQCgcEg8uOx7ps/nyBgKReuN/V76eGbTd8JlcDu0isYNdSn78Y6DD0Lm+uoqLMG+mEKrFQzlncpIXK1fBn5el2rwQEK2wysIHz0dA1GD/Mze16oRmAFa1nmL/Bx5jJ704u5rfnMWP5Nu4F2mPBoS0t2JHMZMWaPzF1y/fi7CqMd0l02WD1VxRIdu/pqOZ9yqEWaw6HfpHhm074+71+UahdM2J4pYg4UCKygqU1uL31CJvn9Wu/ojr08HtTOMAzKTuufbGfKcsw/mPh8ZU5s3RZgR9Zv/8Vg2iUeM4fNQbPArNtJbqm/mK0X4dOdNZ59gALBTcEcs1Htxec6TUjGOr9tOS6ar5sxDpgyMMGemTIT3BSqlXVtyVprW6lQILAC3VE3Z67cMPkBJfZ2w2rrMNwxWVZp2tIbtmDpx0jebKy11pHuU+Q0WbJ4tTWvwoJHZ6dvGRLAfAvt6eLUWZybm4v3338fCoUCTz31FBYsWKCVXlFRgVWrVqGlpQXh4eFIS0uDUChEaWkp1q5dC7lcjqCgIGzYsAG+vr5ITExEV1f3hdTe3o7Lly9j//79GDKE+1orjuZHjQdvrPv8FyyZG2LD3FgHn6Wwbe3bny7YOgsO5d894/BtyZTlHbjq76KUjoa1RiCVSpGeno6MjAxkZ2cjKysLZ89qd9qtWLECqampKCwsBMMw2LFjBwAgJSUFGzduRG5uLoKDg7FlyxYAwM6dO5GTk4OcnBxMmDABS5cuvamDgD6GRkawab/JqvuEmKq6vg0HT+pfA4jwwxoIiouLERERAT8/P3h6eiImJgYFBQXq9KqqKrS3tyMsLAwAkJiYqE7Py8tDcHAw5HI5pFIpfHy0h56VlJTg1KlTWLRokRlPyTGY+nxdrou8EeIMrNGR6gxYA0F1dTXEYrH6tUQigVQqNZguFovV6SKRCJWVlZg2bRoOHz6MuLg4rWO//fbbSE5Ohqsr+5K1hBBCLIO1j0CpVGotfcwwjNZrtvQxY8aguLgYmZmZSE5ORmZmJgDgzJkzqK+vx/Tp03ln2t/fi/c+hBDi6Dw83CAW83viHhesgSAwMBClpb0TimQyGSQSiVa6TNY7OqSmpgYSiQQdHR04cOAAZs6cCQCYM2cONmzYoN5u9+7diI2NNSnTtbXNvGeCXqu17nAsQggxt7a2Tshk/Oc5uLgIjBagWZuGIiMjUVJSgrq6OrS1taGoqAhRUVHq9KCgILi7u6OsrHsJ55ycHERFRUEoFCItLQ3l5eUAgPz8fEycOFG937FjxxAeHs77hEzF9iBwQghxVqw1goCAACQnJyMpKQlyuRzz589HaGgoFi1ahKVLl2L8+PHYtGkTVq9ejebmZoSEhCApKQmurq5IT09Hamoqurq6EBAQgDVr1qiPe/nyZQQEBFj05AghhLATMNZ6MoMZmdI09Ovvdfh31jHLZIgQQqzgwUnDsOChO3jv1++mIUIIIfZhT5lpw87ZUCAghBAnR4GAEEKcnNMEgi4lzcglhBB9nCgQOFyfOCGEWIXTBAKHfjIMIYRYkNMEAooDhBCin9MEAkIIIfpRICCEECdHgYAQQpyc0wQCx1tIgxBCrMNpAgF1FxNCiH5OFAgIIYToQ4GAEEKcnNMEAuojIIQQ/ZwmEBBCCNGPAgEhhDg5ToEgNzcXsbGxiI6Oxvbt2/ukV1RUIDExETExMVi1ahUUCgUAoLS0FImJiUhISMCSJUvQ0NAAAGhubsZLL72EefPmYd68efj111/NeEqEEEL4YA0EUqkU6enpyMjIQHZ2NrKysnD27FmtbVasWIHU1FQUFhaCYRjs2LEDAJCSkoKNGzciNzcXwcHB2LJlCwBg3bp1GDp0KLKzs7F8+XL84x//MP+Z6aAuAkII0Y81EBQXFyMiIgJ+fn7w9PRETEwMCgoK1OlVVVVob29HWFgYACAxMVGdnpeXh+DgYMjlckilUvj4+IBhGBQVFWHx4sUAgKioKKxdu9YCp0YIIYQL1kBQXV0NsVisfi2RSCCVSg2mi8VidbpIJEJlZSWmTZuGw4cPIy4uDrW1tXBzc0NGRgYeffRRJCUloaury5znpBdDw4YIIUQvIdsGSqUSAoFA/ZphGK3XbOljxoxBcXExMjMzkZycjP/85z+oqamBt7c3srKy8NNPP+H555/Hnj17OGfa39+L87YqPlcaee9DCCH2Riz2NvsxWQNBYGAgSktL1a9lMhkkEolWukwmU7+uqamBRCJBR0cHDhw4gJkzZwIA5syZgw0bNmDQoEEQCoWIj48HANx3331obW1FbW0t/P39OWW6trYZSp5PHGtsauO1PSGE2COZrIn3Pi4uAqMFaNamocjISJSUlKCurg5tbW0oKipCVFSUOj0oKAju7u4oKysDAOTk5CAqKgpCoRBpaWkoLy8HAOTn52PixIlwc3NDZGQkdu3aBQA4duwYPDw8MGjQIN4nRwghpP9YawQBAQFITk5GUlIS5HI55s+fj9DQUCxatAhLly7F+PHjsWnTJqxevRrNzc0ICQlBUlISXF1dkZ6ejtTUVHR1dSEgIABr1qwBAKxZswapqanIyMiAUChEeno6XFwsO6WBuggIIUQ/AeOAvaimNA0d+u06Pvr2NwvliBBCrGPryhm89+l309BNw+HCHSGEWIfzBAJCCCF6OU0goAoBIYTo5zSB4Iqs2dZZIIQQu+Q0gUAuV9o6C4QQYpecJhAQQgjRjwIBIYQ4OQoEhBDi5CgQEEKIk6NAQAghTs55AoGAfRNCCHFGThMIhvh62DoLhBBil5wmEPh5udk6C4QQYpecJhAQQgjRjwIBIYQ4OacJBJrPUSaEENLLeQKBrTNACCF2ymkCASGEEP04BYLc3FzExsYiOjoa27dv75NeUVGBxMRExMTEYNWqVVAoFACA0tJSJCYmIiEhAUuWLEFDQwMA4MiRI7j33nsxd+5czJ07FykpKWY8Jf1CRg22+GcQQogjYg0EUqkU6enpyMjIQHZ2NrKysnD27FmtbVasWIHU1FQUFhaCYRjs2LEDAJCSkoKNGzciNzcXwcHB2LJlCwCgvLwcCxcuRE5ODnJycrBu3ToLnJo2D3ehxT+DEEIcEWsgKC4uRkREBPz8/ODp6YmYmBgUFBSo06uqqtDe3o6wsDAAQGJiojo9Ly8PwcHBkMvlkEql8PHxAQCcPHkSBw8eVNcUrl27ZoFTI4QQwgVrMbm6uhpisVj9WiKR4MSJEwbTxWIxpFIpAEAkEqGyshJPP/00hEIhli9fDgDw9vbG7NmzER0djS+++ALJycnIzMzknGl/fy/O2xJCyM1ELPY2+zFZA4FSqdQaeskwjNZrtvQxY8aguLgYmZmZ6hv+66+/rk5//PHH8e9//xtNTU3w9uZ2grW1zVAq6SnEhBDnI5M18d7HxUVgtADN2jQUGBgImUymkQkZJBKJwfSamhpIJBJ0dHRg9+7d6vfnzJmDyspKKJVKvP/+++jq6tL6HFdXV25nRAghxKxYA0FkZCRKSkpQV1eHtrY2FBUVISoqSp0eFBQEd3d3lJWVAQBycnIQFRUFoVCItLQ0lJeXAwDy8/MxceJEuLi44Pvvv0dhYSEAIDs7GxMmTICnp6clzo8QQggLAcMwrG0subm5+PDDDyGXyzF//nwsWrQIixYtwtKlSzF+/HicOnUKq1evRnNzM0JCQrBu3Tq4ubmhtLQUa9euRVdXFwICAvD6668jMDAQZ86cwWuvvYampiYMHjwYGzduxNChQzln2tSmoYXr9/LehxBC7MnWlTN478PWNMQpENgbCgSEEGdliUBAM4sJIcTJUSAghBAnR4GAEEKcHAUCQghxchQICCHEyVEgIIQQJ0eBgBBCnBwFAkIIcXIUCAixMjcR/eyIfaErkhArC73N39ZZIEQLBQJCrCxgMC2wSOwLBQJCrOz+CbfYOgt9hN5OtRRn5lSBYPkfJmi9Fro61ekTOyFg38SqZt1zK/7v/ttsnQ1iQ051JxQJtU932fxQG+WEmMPaxRG2zoJJ7C0QAMCIQPM//pDw5zvQzSaf61SBQFfIqMFwEzr1V+DQAqmt3SSjh/naOgvEgIBBHlj3jPULOE51F9T35AXdWsLNitqAiYrEz8PWWSCGCAQIGGT9Ao5z3AUdxFB/y10AmkHwD9ODtdJcBPbYWHETc6Cve4jvAF7ba46Imjre+FMHX34sDB+teMCUbN10oicPt+nncwoEubm5iI2NRXR0NLZv394nvaKiAomJiYiJicGqVaugUCgAAKWlpUhMTERCQgKWLFmChoYGrf2uX7+Oe+65B1euXDHDqZjGTeRq8r7mLmXbqqljsI+7TT5XnxkTg2ydBbVbAww/0cmavDxEZj2ebsU4IiTA4LZjRwzifFzfgW54feFk9euFcXciYpzhY48bOdimAzZ8PM37vbK5ZchAg2kTgocAYC8jCF0tU4pg/StIpVKkp6cjIyMD2dnZyMrKwtmzZ7W2WbFiBVJTU1FYWAiGYbBjxw4AQEpKCjZu3Ijc3FwEBwdjy5Yt6n2USiVWrVoFuVxu5lPi5+XHwnhtr/mHcKCCncNIjLrd1llQe/7/xts6CwCA2ffearFjb105A7cGGO4ojo0YwflYU0ICIRJqF6wWzwnRem1Ps6qnhVm+0HH36CF6/20IW+Xc34dfDY0r1r9KcXExIiIi4OfnB09PT8TExKCgoECdXlVVhfb2doSFhQEAEhMT1el5eXkIDg6GXC6HVCqFj4+Per+PP/4YkZGRGDSIe4nDEob6G47S+og12lddXAR49Y93mztLrMLHiK3+mf3h4e7K+TmrLvZzn7CbJjNzP1T87tHcrx9zfwMxk/sGtfCxEjN/in66I6Os8bD2+zSax/7v/tvwwMRh+je08aPjWX921dXVEIt7LxyJRAKpVGowXSwWq9NFIhEqKysxbdo0HD58GHFxcQCA8vJyHDp0CE8//bRJmfb394JY7M37fz8/7aYX1ft8LH2098bv5ibE1EnmK625uQk5bTf5LuNtr/qI3HpLagN1hqi5WLh6LhAIOH/Pw4MGIeWpyZAMYu/Q5Pu30xQbOZJ1m2G3WGZ0jf9gfk1OMZGj9L7v58WtSc9H5+89a2rvnAHN34c+gwcPxMa/3s/pczw83bSOo++4D2gEAlX6PSGBnI7fXy8tmKT1eqCn5Ydq+vj0XscBAT4YYqCj3te3+97k5iY0el2r0vn+b+zB9QDAeudRKpUQaJSMGIbRes2WPmbMGBQXFyMzMxPJycnYtm0b0tLS8J///AcuJhb/amuboVTyj6D1N1q1XstkTVqvVyVNwppPy4weQ97e25TV0aHocwxjvD1FaGo13BTmLuRW/mpu7sCooT74/Voj58/u7FD07t/SoZWm7FJyPo4pGIbh/D3JZE0YPdQbAweIALSxbquPh7sQbRrnK3QVQNGlfb3cMpg90LS1dGCwjzvqGru/ryG+A1DT0M66H5u6uhZe2ys69F8z99wpQdHPl1n3/+PM0fgg51f1a83vje3vUlfXwnlJjLbWTtZj19e39klvau7os50ltDS1Y+ytfjh16Ub369ZOzvsOl3jhcnUz789sbOy9ho191/UN3d+LXN5ldDsBuP+WNLm4CIwGA9Y7cWBgIGQymfq1TCaDRCIxmF5TUwOJRIKOjg7s3r1b/f6cOXNQWVmJ0tJS1NbW4tlnn8XcuXNRXV2NxYsX4/z587xPzly2vDodW16djts5lAC9jUz4WLlgotF9g4x0FgHAcEl3SSCKwxIEdtJqYXOL54zDgzrV7X89G6n1+t/P39dnP1cXF07NVUMt1IF/q4R7rcBQE9XkOyW4P5S9dnjPnQH4+JXpnD/PkvSdijNdyvcaqv2wlGsn9zSfhQWz9zOYgjUQREZGoqSkBHV1dWhra0NRURGioqLU6UFBQXB3d0dZWXdJOicnB1FRURAKhUhLS0N5eTkAID8/HxMnTsT999+PvXv3IicnBzk5OZBIJPjoo49w2222m+IuEAi0ajHG+A50Q9KsMXrT7hjuZ3C/d17kVr0GLDMywNh1Zs5RKabkfPrdpnfaRYwLxILoOzDnvpHq9zwHaFd0vfvRBDDlLu0f7rsr+n9DFQiA5D9MwNOzx7JuGxY8xOBcFxeBAE9E678W+2zrov2XWfXkJPyJw+fboxFGOrft3diRg42mG7oNJc0agzf/eh/mTNXfTNhfrIEgICAAycnJSEpKwrx58xAfH4/Q0FAsWrQIJ0+eBABs2rQJ69atw6xZs9Da2oqkpCS4uroiPT0dqampmDt3LgoLC7FmzRqLnARXAwdwa4NnY8o0cE9345/94CTtUu0kIx3CAvSvb2nC7dqliqVmWmpj68oZCDQyF+LFR/R/jj/Pseoeer5LPqNbgN4fHNsojMi7hmLpw735HhHoY2Rr7ny93HH/hFuwKH6c0e2MfZ8M0z0hcpiYW+3i7tFDkBR7JwDg9iBfTjVPe7QqaRI+eGkar33G3uqHIX6WGXHTX3cM9wNjpKgWJB6IgQNE8PNyt9gABk53xoSEBCQkJGi9t3nzZvW/x44di6+++qrPfuHh4di5c6fRY+/du5dLFszC2DA5fV58JBRvfXnCQrnRtuChO7CnrHc+xeKEcXim8keLfNYtQwZi68oZWLi++7vn2uloqpmTuifLhN6uv1r7UPgwfLXvHKdj/SX+TowZPghXa/m1s+tSjY9PW3gP2jsVePm/xQa3vUXMb2SZOen72buJXNAp59+v88LDoRCLvU1qY/7T7LHI/el31Db2rz3fHPcxoasLwHP6T/IfwuDq4gIvDrXDof6eWJQwDq9/UmpiDrkLGTkILz12N8rP1wLQ//fmM8rLVHY0WM++vPvi/X1uXA9P026+YowUy9/4y70WyRfQ3eZriiej7zBYKucqYlwAXDWaGV54WHusvb6x9/8XZbzZT3fsuTHBw/zg7zsA4/vxcJcRAd7w6bkheA4QYrCFxmabg74rzFKBe9RQw7WdqAm34F/P9e1r4WvoYOsF1cdmaM6g7/4m/zRrDMb0NOEGDPLAm3/te05rFkX0DFawDK2JqD2R0b1nVJ/u3/btZfdj3v2WaQ7SRIHAAM+eC+GFxN4bW9yUkQAAgZGW8A9ffgAfvzIdQUMG9qn282vO6f0MzeWz3USufdrA9enT3MMwmD5xmMFSOVeL54RoXci6fSvGZk+ag6F1crj28ZhC98gvPRqG1D+FW+zzNIl6hvY+FN67BAFbM6MmPrPflz0S2q/+KS7rdrm76Qn6FvrThYzq2x7vOUCEV/54N/725CRE3hVosLl4kLflasmaf0vVqQcH+eLPcXdiQfQdWtt6eYisMp+FAoEet93SWzK6+46+1TJRz+xITz2lBpHQRd0xp9vRqMIW4TXjhYtAoHVz5XpNWLJDTevCZLqXYng61rYdj+ZePFBzhJdu/A4ZNRgjefQV6OvT4Eu3NgoYvhZ05w1w5ePphkljuE3u0neTjZvCr5/GXHy9+J2vQCBAcJAvBAIBREJXvL2s70AOQ0tf8A0Q+gp/mvkd3rOMiUAgwH3jh2IAx7lE5kaBQEf6C1PxyuPGZwuPGzEIjz84Gk/oRG+u7hjmp/736wvvwT+enmx4Y2iXdpNmd3f2+WusD3SrxAvvsoxKMue8Rd3z/sfT9+D+0N6OR1ut6PrYg6ORyNIMxeb1hfd0/8OMhTC+a1Lp297Ymli6yzZoDmZ4dEaw7uZmcYueGfn9WbeLq8/TZvV5b9WTk7A4wXinu+6oKVOJzDD5cpjYCxN7CpjmOJ452Ecu7IjvQDfWC1ogEOChycPNUtIbJvHi3Int5+WGOT1r8TwdeycWxt4J34FumD/9dr21E0vx7UcbNd+RW3zGTUdPHo54DjOG+ervLcTY/vpGi2iOgOJSA3xu3l1aw2ef7BlS+tiDo3kvocKVPa0ZNMTXAxF6xucHDPbEuJGD8Mrjd8PVgmuX8B2xBvS2CoRzrIFZmv38NZ0Il9I52w3Aw12IqaFDkf7CVNw1yjLPGoi8KxDpL0w1vhFLPsfe6qf1+uPV0Uj/632sk7kGenQHjLn9HDd9n4HmOV1xU0YgaoLG5CwzVqGG6UweY+vP4Lsi52CfAZin8ajJ4GG+WLs4Ag+FG1jXxgziI0di3tRRWs2ohuhr1tJkqRZwoasLXn7sbl4rqP6TbZCHnszOf6C7cGZsHpGuYWIvbF05o8+1ofJE9B0YacWnxlEguEm599Rq/HkuMa05csTLQ9TvR+c9M/curddeHiJONYpF8ePw+MzR/V4K+s8s4/RVHp52O/40+069P3S+MeHecQFaTTKz7r0Vq5PC1c+bYAvyAoFpATAhcqS6RhQ42NOiHejuIlfMmToKq5P6dpo/OiMYf5w5Wv1aNcjC3g0cIDQ42GHUUOM35beX3Y+XHg0zW15mTByG1D8ZbzI2J9v0TDg5a0yp9xwgxL+ejURVTTPe+vIE5xFLA/SN6tDD2Lo7Hu5CyBWd8PNyg7eJs5a9Pd20RlcA4D2JyFae6Vl6OWtv93LtLgIBbrvFB8l/mIBfKmUmD/98KHw4xt7qh9YOBbbsqtBaCRdgH6bLhbEh0VzF3GO+hRiDg3zh7SnC0TM1ALRrU2+9MBVdJqw5xtcbf7kXLgJg//FrBn+7+mbnC10FDrMUDAWCm5i/7wBcr29l35DFumcikPLhIa33bhky0GAgSHliIk6cq+1zI+8va3RG6sP1t7xkbggURhbwG+LrgWgDN0k3oQs6FcYniT2uUcq+j+XpXzeDkYHeeGZOCPIOXVS/p3ljNXV0FF9BQwbimmoCo0DAafHBjc9OgbvIFWeuNBjdzl5QILCS/lTTVePGNUfmGMNWjeXL2DNU9Z1VwCBPPBTuoA+W51HAHDhAiJb27lVOoycP7zPR78FJw/r0kehSLXEx9/5ROFJRjYvXu2f9TgkJQP6hi5hipSWabclQHxffphEXgQBKE2o0A9xcIXR1waMzRhvcRlWLmzlpGKaGDkVXlxJ/fesAphh4utsQ3+7amo0fM8AZBQIrMvWacHdzxUcrHtCa0WtMyhO9667bc830hYfHQ85SCrYZDl/c+iVTsO7zX3C1pgXj9CwmtuAh9uHFY24dhNeeCseIQG8cqahWvy8Z5IkPXn6AT47NQrPAEj5W0qf5yRL8fQdg68oZSN1yGEFiLxz+TaqVPnmsBD8crWI9zpsv3IcX3z4IgO8ijy6sz072cBdqD3AQcX/YkiOgQGADfCoHqm35jCTR3Fa1IFl/Vvi0FGusoWJJAweIMNjHHVdr+rfukaqD3h6CtmYfwZI5IWYbf8/F63/uHrGjGwg0R/0Yy42PxjpCllwi4mZEo4ZsRN9aOVNCAnDniEGI0xiX3N+qpc9AN2xdOcOkxwEaC1hDelYM5bLcha1wrUH1h6rphu1ZE8RM7Lz3Neae4YgYZ9paYLZkv79iCzO2wJY1LJsf2qc903OACCt6ZjV3KW3fZGIsCD06IxjjRg7GaI1Z0vbk8QdH486R+sePc3mIuCbVkhr6JhBOCQm8edvx7fCea43g3h/G+hnsmdMGAraniWkKHuZrdBhkxLgAHNKpzuoj9h2A05eBAW5CuLgI4GLkl2aNTqZJd4hRdlrGvqEeIqGrepq8PXposv4RS28vu593LcbfdwAemxFs8YesDxN74cL1JrPMWL9ZebgLwf+BkaYZJvbCjEn216RqCU57xfFZD+dvGp2v+vwlfpzBp5ZpeiJ6DMJGizGCx4xBS9aEl8wLgULhIMMazMTUp7EZGvZpTk9E34H7xgci0EKPx+TCGkuV3DVqMO/Z01wN8nY324KLr//5HrMcxxE4bSAwJxcXAadVA93dXI0+eUwfS9YMXF1c4KozFNscE4rs0at/vBsd8i7W7VTLJAcaGTJrKW4iV4y5lftyCJbwh+m3Y1/PCB2uZZCBA4S8Hvq03IQZuA/cHYRfONRe9T2fmrDjFAhyc3Px/vvvQ6FQ4KmnnsKCBQu00isqKrBq1Sq0tLQgPDwcaWlpEAqFKC0txdq1ayGXyxEUFIQNGzbA19cXZ8+exerVq9Ha2gpfX1+sX78eQUHOUQWzpD/H3YnhPB6K7ky43mDFfh548ZEJGD3M18I5sk+mLIP8zotR7Bv1U1LMGCTFcHs+sz3xcLfNJEi+WOtnUqkU6enpyMjIQHZ2NrKysnD27FmtbVasWIHU1FQUFhaCYRjs2LEDAJCSkoKNGzciNzcXwcHB2LJlCwAgLS0Nzz33HL799lvExsbizTfftMCp3Rz4NA3dN34o78dxkr5Cb/endnpiFnfyWPDOllgDQXFxMSIiIuDn5wdPT0/ExMSgoKBAnV5VVYX29naEhYUBABITE9XpeXl5CA4Ohlwuh1QqhY9P90idbdu2ISoqCkqlElevXlW/T+yXm42eMUBsx5KL1jkLR/kOWYs91dXVEIt727UlEglOnDhhMF0sFkMq7R5BIxKJUFlZiaeffhpCoRDLly/v/lChEI2NjYiNjUV7ezs+++wzXpn29+9/84dYbN2SM9/PU/YsphV336g++1oy7yJR7yXh6emm/qzNqx5CQ3OHWT7b2DFMTbMme8mHqbjm3xJ/Cz77GdrWUb9/Vb7tMf+sgUCpVGpFNYZhtF6zpY8ZMwbFxcXIzMxEcnIyMjMzAQA+Pj44ePAg9u/fj2effRZ79uyBqyu39rTa2mb1jdJUMllTv/a3xudtfuUBuAgEWvuKxd4WzbtcrlD/u7W1U+uzvEQu/f5stvybmmYtlv7+LY1P/s35t/jXs5FoauvktN8fZ46GoovRu60jf/8yWZPN8u/iIjBagGat7wcGBkIm6+2tl8lkkEgkBtNramogkUjQ0dGB3bt3q9+fM2cOKisrAXQ3GalGp0RFRaG9vR0NDY6xSp81ubq4OEzVsr/ipozg9JAT4pj8fQdwfs7zzPDhmHWv5Yfrkl6sgSAyMhIlJSWoq6tDW1sbioqKEBXVO0ogKCgI7u7uKCsrAwDk5OQgKioKQqEQaWlpKC8vBwDk5+dj4sTuSVxbt27F999/DwA4dOgQBg0ahMGD+y7aRZzHw9Nu1/uQE0KI5bE2DQUEBCA5ORlJSUmQy+WYP38+QkNDsWjRIixduhTjx4/Hpk2bsHr1ajQ3NyMkJARJSUlwdXVFeno6UlNT0dXVhYCAAKxZswYAsH79erz22mt477334O3tjbffftviJ6qy/vmpqK/v3yJhfDzywO34ct85q33ezcjLQ4QHJ1nusYuEWNLfrfikMVNxGiOXkJCAhIQErfc2b96s/vfYsWPx1Vdf9dkvPDwcO3fu7PN+cHAwvvjiC755NYuQ2/whk1nngRYAMDtiBGab8HBr0uvtZdyXFCbE3vBZScBWaLA00aI5sXgmlcKJjvkP3I5alqdzEcdDgYDo9crjd2OIFR5KQuyPm8gFnXL9q9/GUu32pkSBgBCi5e9/mozfLtTbOhvEiigQEEK0DPUfiKH+9KAdZ0LrBhAtN+fao4QQYygQEL2cZB4bIQQUCAghxOlRICBaxt/WPcN7sM8AG+eEEGIt1FlMtMRGjMDU0FvgO9B6k+4IIbZFgYAAAISuArgIBBAIBBQECHEyFAgIAOC95Gm2zgIhxEYoEBAAgIieQEaI06JfPyGEODkKBIQQ4uQoEBBCiJOjQEAIIU6OAgEhhDg5CgSEEOLkHHL4qItL/1ZE6+/+tkb5ty3Kv21R/s3/mQKGYWjlYUIIcWLUNEQIIU6OAgEhhDg5CgSEEOLkKBAQQoiTo0BACCFOjgIBIYQ4OQoEhBDi5CgQEEKIk6NAQAghTs5pAkFubi5iY2MRHR2N7du32zo7Wt59913ExcUhLi4OGzduBAAUFxcjISEB0dHRSE9PV29bUVGBxMRExMTEYNWqVVAoFACAq1evYsGCBZg1axaeffZZtLS0WP08NmzYgJUrVzpc/vfu3YvExETMnj0b//znPx0u/zk5OerrZ8OGDQ6T/+bmZsTHx+PKlStmzXNjYyMWL16M2bNnY8GCBZDJZFbJf1ZWFuLj45GQkICUlBR0dnbadf61ME7g+vXrzPTp05n6+nqmpaWFSUhIYM6cOWPrbDEMwzA//fQT8+ijjzIdHR1MZ2cnk5SUxOTm5jLTpk1jLl26xMjlcmbhwoXMvn37GIZhmLi4OObo0aMMwzBMSkoKs337doZhGGbx4sXMd999xzAMw7z77rvMxo0brXoexcXFzL333su8+uqrTFtbm8Pk/9KlS8zUqVOZa9euMZ2dnczjjz/O7Nu3z2Hy39raykyePJmpra1l5HI5M3/+fGbPnj12n/9jx44x8fHxTEhICHP58mWzXjNpaWnMhx9+yDAMw3zzzTfMsmXLLJ7/8+fPMw899BDT1NTEKJVK5pVXXmG2bdtmt/nX5RQ1guLiYkRERMDPzw+enp6IiYlBQUGBrbMFABCLxVi5ciXc3NwgEolw++2348KFCxgxYgSGDx8OoVCIhIQEFBQUoKqqCu3t7QgLCwMAJCYmoqCgAHK5HD///DNiYmK03reWGzduID09HUuWLAEAnDhxwmHy//333yM2NhaBgYEQiURIT0+Hh4eHw+S/q6sLSqUSbW1tUCgUUCgU8PLysvv879ixA3//+98hkUgAmPea2bdvHxISEgAA8fHx2L9/P+RyuUXz7+bmhr///e/w8vKCQCDAHXfcgatXr9pt/nU55OqjfFVXV0MsFqtfSyQSnDhxwoY56jV69Gj1vy9cuID8/Hw88cQTffIrlUr7nIdYLIZUKkV9fT28vLwgFAq13reW1NRUJCcn49q1awD0f9/2mv+LFy9CJBJhyZIluHbtGh544AGMHj3aYfLv5eWFZcuWYfbs2fDw8MDkyZMd4vtfs2aN1mtz5llzH6FQCC8vL9TV1SEgIMBi+Q8KCkJQUBAAoK6uDtu3b8e6devsNv+6nKJGoFQqIRD0LsPKMIzWa3tw5swZLFy4EK+88gqGDx+uN7+GzkPf+Vjr/L788ksMHToUU6ZMUb9nKJ/2mP+uri6UlJRg7dq1yMrKwokTJ3D58mWHyf+pU6fw9ddf44cffsCBAwfg4uKCCxcuOEz+VSx5zTAMAxcX69zqpFIpnnrqKTz88MO49957HSb/TlEjCAwMRGlpqfq1TCZTV+nsQVlZGZYuXYq//e1viIuLw5EjR7Q6iFT5DQwM1Hq/pqYGEokEgwcPRlNTE7q6uuDq6mrV88vLy4NMJsPcuXPR0NCA1tZWVFVVwdXV1SHyP2TIEEyZMgWDBw8GAMycORMFBQUOk/+DBw9iypQp8Pf3B9DdxLBlyxaHyb+Kbt76k2eJRIKamhoEBgZCoVCgpaUFfn5+Fj+Hc+fO4S9/+QuefPJJLFy4UO952Wv+naJGEBkZiZKSEtTV1aGtrQ1FRUWIioqydbYAANeuXcPzzz+PTZs2IS4uDgAwYcIE/P7777h48SK6urrw3XffISoqCkFBQXB3d0dZWRmA7tEiUVFREIlECA8PR15eHgAgOzvbaue3bds2fPfdd8jJycHSpUsxY8YMfPzxxw6T/+nTp+PgwYNobGxEV1cXDhw4gFmzZjlM/seOHYvi4mK0traCYRjs3bvXoa4fFXPmedq0acjOzgbQXVAJDw+HSCSyaP6bm5vx5z//GcuWLVMHAQAOk3+nGDXEMAzz7bffMnFxcUx0dDTz0Ucf2To7am+88QYTFhbGzJkzR/1/RkYGU1xczCQkJDDR0dHMmjVrGKVSyTAMw1RUVDAPP/wwExMTwyxfvpzp6OhgGIZhrly5wjzxxBPM7NmzmYULFzI3btyw+rl8/fXXzKuvvsowDONQ+f/yyy/V10ZaWhrT1dXlUPn/8MMPmZiYGCY+Pp5JSUlh2tvbHSb/06dPZy5fvswwjPmumfr6euaZZ55hYmNjmUcffVR9fEvmf9u2bUxISIjW7/itt96y+/yr0BPKCCHEyTlF0xAhhBDDKBAQQoiTo0BACCFOjgIBIYQ4OQoEhBDi5CgQEEKIk6NAQAghTo4CASGEOLn/B0MSjwSwNaTsAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "plt.plot(ssim_list)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72eca357",
   "metadata": {},
   "source": [
    "# Comparar el conjunto de prueba (Test)\n",
    "\n",
    "Se realiza los resultados de las tres métricas en el conjunto de prueba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "1ac73141",
   "metadata": {},
   "outputs": [],
   "source": [
    "ssim_list = []\n",
    "mse_list = []\n",
    "psnr_list = []\n",
    "\n",
    "for i in range(len(x_test)):\n",
    "    test_img = x_test[i].reshape(8,8)*255\n",
    "    decoded_img = decoder[i].real.reshape(8,8)*255\n",
    "    Y = float(mse(decoded_img,test_img))\n",
    "    ssim_list.append(ssim(decoded_img.astype(\"float\"),test_img.astype(\"float\")))\n",
    "    mse_list.append(Y)\n",
    "    aux = (64**2)/Y\n",
    "    psnr_list.append(10*math.log10(aux))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "0fa19686",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "plt.plot(mse_list)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "41d04e61",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "plt.plot(psnr_list)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "c049c41d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "plt.plot(ssim_list)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f90c4ed4",
   "metadata": {},
   "source": [
    "Se repite el mis proceso pero ya con eentradas no aleatorias para guardar la información en archivos csv que se ocuapran para realizar una clasificación con las imagenes reducidas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "f9f683e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "(x_train_c, y_train_c), (x_test_c, y_test_c) = tf.keras.datasets.mnist.load_data()\n",
    "\n",
    "# Rescale the images from [0,255] to the [0.0,1.0] range.\n",
    "x_train_c, x_test_c = x_train_c[..., np.newaxis]/255.0, x_test_c[..., np.newaxis]/255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "7d4873e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train_c, y_train_c = filter_01(x_train_c, y_train_c)\n",
    "x_test_c, y_test_c = filter_01(x_test_c, y_test_c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "1a5ee904",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train_c = tf.image.resize(x_train_c, (8,8), method='nearest', preserve_aspect_ratio=True).numpy()\n",
    "x_test_c = tf.image.resize(x_test_c, (8,8), method='nearest', preserve_aspect_ratio=True).numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "de674932",
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(x_train_c)):\n",
    "    x_train_c[i] = Normalize(x_train_c[i])\n",
    "    \n",
    "for i in range(len(x_test)):\n",
    "    x_test_c[i] = Normalize(x_test_c[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "a002113d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(12665, 64)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_c = x_train_c.reshape(len(x_train_small), 64)\n",
    "x_test_c = x_test_c.reshape(len(x_test_small), 64)\n",
    "\n",
    "x_train_c.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "578d0056",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compress_result_train(params):\n",
    "    reduce = []\n",
    "    for i in range(len(x_train_c)):\n",
    "        circuit_init = input_data(n,x_train_c[i])\n",
    "        circuit_vqc = vqc(n,num_layers,params)\n",
    "\n",
    "        circuit_full = QuantumCircuit(n,n-size_reduce)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(n)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(n)])\n",
    "        len_cf = len(circuit_full)\n",
    "        for i in range(n-size_reduce):\n",
    "            circuit_full.measure(size_reduce+i, i)\n",
    "        job = execute( circuit_full, Aer.get_backend('qasm_simulator'),shots=8192 )\n",
    "        result = job.result().get_counts()\n",
    "        probs = {k: np.sqrt(v / 8192) for k, v in result.items()}\n",
    "        reduce.append(probs)\n",
    "        \n",
    "    return reduce\n",
    "reduce_train_c = compress_result_train(params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "870ea7d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compress_result_test(params):\n",
    "    reduce = []\n",
    "    for i in range(len(x_test_c)):\n",
    "        circuit_init = input_data(n,x_test_c[i])\n",
    "        circuit_vqc = vqc(n,num_layers,params)\n",
    "\n",
    "        circuit_full = QuantumCircuit(n,n-size_reduce)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(n)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(n)])\n",
    "        len_cf = len(circuit_full)\n",
    "        for i in range(n-size_reduce):\n",
    "            circuit_full.measure(size_reduce+i, i)\n",
    "        job = execute( circuit_full, Aer.get_backend('qasm_simulator'),shots=8192 )\n",
    "        result = job.result().get_counts()\n",
    "        probs = {k: np.sqrt(v / 8192) for k, v in result.items()}\n",
    "        reduce.append(probs)\n",
    "        \n",
    "    return reduce\n",
    "reduce_test_c = compress_result_test(params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "2ceccdd8",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_reduce = []\n",
    "for i in reduce_test_c:\n",
    "    index_image = []\n",
    "    for j in range(16):\n",
    "        bin_index = bin(j)[2:]\n",
    "        while len(bin_index) <4:\n",
    "            bin_index = '0'+bin_index \n",
    "        try:   \n",
    "            index_image.append(i[bin_index])\n",
    "        except:\n",
    "            index_image.append(0)\n",
    "    \n",
    "    test_reduce.append(np.array(index_image))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "79e605ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_reduce = []\n",
    "for i in reduce_train_c:\n",
    "    index_image = []\n",
    "    for j in range(16):\n",
    "        bin_index = bin(j)[2:]\n",
    "        while len(bin_index) <4:\n",
    "            bin_index = '0'+bin_index \n",
    "        try:   \n",
    "            index_image.append(i[bin_index])\n",
    "        except:\n",
    "            index_image.append(0)\n",
    "    \n",
    "    train_reduce.append(np.array(index_image))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "940dc3aa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12665"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def decoder_result_train_c(params):\n",
    "    reduce = []\n",
    "    for i in range(len(train_reduce)):\n",
    "        circuit_init = input_data(n,np.concatenate((np.zeros(48), train_reduce[i]), axis=0))\n",
    "        circuit_vqc = vqc(n,num_layers,params).inverse()\n",
    "\n",
    "        circuit_full = QuantumCircuit(n,n)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(n)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(n)])\n",
    "        job = execute( circuit_full, Aer.get_backend('statevector_simulator') )\n",
    "        result = job.result().get_statevector()\n",
    "        reduce.append(result)\n",
    "    return reduce\n",
    "        \n",
    "decoder_train_c =decoder_result_train_c(params)\n",
    "len(decoder_train_c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "61d3f477",
   "metadata": {},
   "outputs": [],
   "source": [
    "def decoder_result_test_c(params):\n",
    "    reduce = []\n",
    "    for i in range(len(test_reduce)):\n",
    "\n",
    "        circuit_init = input_data(6,np.concatenate((np.zeros(48), test_reduce[i]), axis=0))\n",
    "        circuit_vqc = vqc(n,num_layers,params).inverse()\n",
    "\n",
    "        circuit_full = QuantumCircuit(n,n)\n",
    "\n",
    "        circuit_full = circuit_full.compose(circuit_init,[i for i in range(n)])\n",
    "        circuit_full = circuit_full.compose(circuit_vqc,[i for i in range(n)])\n",
    "        job = execute( circuit_full, Aer.get_backend('statevector_simulator') )\n",
    "        result = job.result().get_statevector()\n",
    "        reduce.append(result)\n",
    "    return reduce\n",
    "        \n",
    "decoder_c =decoder_result_test_c(params)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "652d34ba",
   "metadata": {},
   "source": [
    "### Guardar los resultados\n",
    "\n",
    "Se guardaron dos archivos train.csv y test.csv de las imágenes comprimidas obtenidas de nuestro autoencoder la primera para el  conjunto de entrenamiento y la segunda para el conjunto de prueba\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "5031c796",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd \n",
    "\n",
    "\n",
    "\n",
    "df = pd.DataFrame(train_reduce)\n",
    "df[16] = y_train\n",
    "df.to_csv(\"train_1.csv\",index=False)\n",
    "\n",
    "\n",
    "\n",
    "df = pd.DataFrame(test_reduce)\n",
    "df[16] = y_test\n",
    "df.to_csv(\"test_1.csv\",index=False)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49708844",
   "metadata": {},
   "source": [
    "# Resultados del autoencoder cuántico\n",
    "\n",
    "Siguiendo los resultados definimos en un histograma por métrica los mejores casos y nos dieron las siguientes gráficas\n",
    "\n",
    "\n",
    "## MSE\n",
    "\n",
    "Los resultados más cercanos al 0 son los mejores resultados, viendo de manera visual el mejor caso es con 200 imágenes.\n",
    "\n",
    "<img src=\"mse.png\"> \n",
    "\n",
    "\n",
    "## PSNR\n",
    "\n",
    "Los resultados con un valor mayor en el eje de las ordenadas son los mejores resultados, observando de manera visual que el mejor caso es con 200 imágenes.\n",
    "<img src=\"psnr.png\">\n",
    "\n",
    "\n",
    "## SSIM\n",
    "Los resultados más cercanos a 1 son los mejores resultados, viendo de manera visual el mejor caso es con 200 imágenes.\n",
    "\n",
    "<img src=\"ssim.png\">\n",
    "\n",
    "\n",
    "\n",
    "Por lo tanto consideraremos los resultados de 200 imagenes para realizar un clasificador binario."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82d69e25",
   "metadata": {},
   "source": [
    "# Parte del clasificador binario\n",
    "\n",
    "Se importan las bibliotecas necesarias para esto usando qiskit meachine learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "374ffcb5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scikit Imports\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n",
    "\n",
    "# Qiskit Imports\n",
    "from qiskit import Aer, execute\n",
    "from qiskit.circuit import QuantumCircuit, Parameter, ParameterVector\n",
    "from qiskit.circuit.library import PauliFeatureMap, ZFeatureMap, ZZFeatureMap\n",
    "from qiskit.circuit.library import TwoLocal, NLocal, RealAmplitudes, EfficientSU2\n",
    "from qiskit.circuit.library import HGate, RXGate, RYGate, RZGate, CXGate, CRXGate, CRZGate\n",
    "from qiskit_machine_learning.kernels import QuantumKernel\n",
    "\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import csv"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e10ab228",
   "metadata": {},
   "source": [
    "Se obtiene los datos de los csv tanto para poder procesar los datos en el clasificador cuántico variacional. pasando un lado el vector de entrada de tamaño 16x1 y el otro la etiqueta"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "aeee9d86",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((12665, 16), (12665,))"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_train = []\n",
    "label_train = []\n",
    "with open('train.csv', newline='') as csvfile:\n",
    "    reader = csv.reader(csvfile, delimiter=',')\n",
    "    next(reader, None)\n",
    "    for row in reader:\n",
    "        sample_train.append(row[0:-1])\n",
    "        label_train.append(row[-1])\n",
    "sample_train = np.array(sample_train).astype(np.float)\n",
    "label_train = np.array(label_train).astype(np.float)\n",
    "sample_train.shape, label_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "34619f78",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((2115, 16), (2115,))"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_test = []\n",
    "label_test = []\n",
    "with open('test.csv', newline='') as csvfile:\n",
    "    reader = csv.reader(csvfile, delimiter=',')\n",
    "    next(reader, None)\n",
    "    for row in reader:\n",
    "        sample_test.append(row[0:-1])\n",
    "        label_test.append(row[-1])\n",
    "sample_test = np.array(sample_test).astype(np.float)\n",
    "label_test = np.array(label_test).astype(np.float)\n",
    "sample_test.shape, label_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5d18fda",
   "metadata": {},
   "source": [
    "Se genera el vector de entrada para el conjunto de entrenamiento y prueba de tamañ0 16x1 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a58efee4",
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_train = sample_train.reshape(len(sample_train), 16)\n",
    "sample_test = sample_test.reshape(len(sample_test), 16)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9ec5347",
   "metadata": {},
   "source": [
    "Vamos a realizar un clasificador usando 4 qubits por lo cual hay disminuir el númeor de muestras para que nuestros dispositivos puedan correr los ejemplos, usando el método Hold-out 70-30, es decir, 70% entrenamiento y 30% de prueba."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "110140d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_size = 700\n",
    "sample_train = sample_train[:train_size]\n",
    "label_train = label_train[:train_size]\n",
    "\n",
    "test_size = 300\n",
    "sample_test = sample_test[:test_size]\n",
    "label_test = label_test[:test_size]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "620e6624",
   "metadata": {},
   "source": [
    "Lo siguiente es mapear el vector clasico a un estado cuántico para ello usaremos la funcion ZZFeatureMap (\"Se puede ocupar otro como PauliFeatureMap o ZFeatureMap\").\n",
    "\n",
    "Despues inicializaremos un Kernel cuántico del cual podemos calcular cada elemento de esta mátriz  en una computadora cuántica calculando la amplitud de transición. Esto nos proporciona una estimación de la matriz cuántica del kernel, que luego podemos usar en un algoritmo de aprendizaje automático del kernel, es este caso se usara en una maquina de soporte vectorial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a95849ae",
   "metadata": {},
   "outputs": [],
   "source": [
    "zz_map = ZZFeatureMap(feature_dimension=16, reps=1, entanglement='linear', insert_barriers=True)\n",
    "zz_kernel = QuantumKernel(feature_map=zz_map, quantum_instance=Aer.get_backend('statevector_simulator'))\n",
    "zz_map.draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e687dbe",
   "metadata": {},
   "source": [
    "Construimos las matrices de entrenamiento y prueba del kernel cuántico.\n",
    "Para cada par de puntos de datos en el conjunto de datos de entrenamiento"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5079e709",
   "metadata": {},
   "outputs": [],
   "source": [
    "matrix_train = zz_kernel.evaluate(x_vec=sample_train)\n",
    "matrix_test = zz_kernel.evaluate(x_vec=sample_test, y_vec=sample_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7a52c6c",
   "metadata": {},
   "source": [
    "Utilizamos las matrices de entrenamiento y prueba del kernel cuántico en un algoritmo de clasificación de máquina de vectores de soporte clásico."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c28f62cb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precomputed kernel classification test score: 0.97\n"
     ]
    }
   ],
   "source": [
    "zzpc_svc = SVC(kernel='precomputed')\n",
    "zzpc_svc.fit(matrix_train, label_train)\n",
    "zzpc_score = zzpc_svc.score(matrix_test, label_test)\n",
    "\n",
    "print(f'Precomputed kernel classification test score: {zzpc_score}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49d85e42",
   "metadata": {},
   "source": [
    "Probamos el algoritmo viendo que tal hace la clasificacion del set de prueba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7593b9f9",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = zzpc_svc.predict(matrix_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7bdb89af",
   "metadata": {},
   "source": [
    "Como se puede observar de 300 muestaras solo 6 no se clasificaron de manera correcta"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "78367dd8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0 0.0\n",
      "1.0 0.0\n",
      "1.0 0.0\n",
      "1.0 0.0\n",
      "0.0 1.0\n",
      "0.0 1.0\n",
      "0.0 1.0\n",
      "1.0 0.0\n",
      "0.0 1.0\n"
     ]
    }
   ],
   "source": [
    "for prediction,label in zip(predictions,label_test):\n",
    "    if(prediction != label):\n",
    "        print(prediction, label)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "089bdace",
   "metadata": {},
   "source": [
    "## Validar para el conjunto de 200 imagenes\n",
    "\n",
    "Se repite el mismo proceso pero considerando que el método de validación hold-out  se consideré válido se debe hacer la prueba con diferentes conjuntos de forma aleatoria que se realizará con el conjunto comprimido de 200 imágenes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "92898ea9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((12665, 16), (12665,), (2115, 16), (2115,))"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_train = []\n",
    "label_train = []\n",
    "with open('train_200.csv', newline='') as csvfile:\n",
    "    reader = csv.reader(csvfile, delimiter=',')\n",
    "    next(reader, None)\n",
    "    for row in reader:\n",
    "        sample_train.append(row[0:-1])\n",
    "        label_train.append(row[-1])\n",
    "sample_train = np.array(sample_train).astype(np.float)\n",
    "label_train = np.array(label_train).astype(np.float)\n",
    "\n",
    "sample_test = []\n",
    "label_test = []\n",
    "with open('test_200.csv', newline='') as csvfile:\n",
    "    reader = csv.reader(csvfile, delimiter=',')\n",
    "    next(reader, None)\n",
    "    for row in reader:\n",
    "        sample_test.append(row[0:-1])\n",
    "        label_test.append(row[-1])\n",
    "sample_test = np.array(sample_test).astype(np.float)\n",
    "label_test = np.array(label_test).astype(np.float)\n",
    "sample_train.shape, label_train.shape, sample_test.shape, label_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b13041a",
   "metadata": {},
   "source": [
    "Se genera 7 iteraciones con diferentes imagenes de rangos fijos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a6ac49e4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precomputed kernel classification test score: 0.97\n",
      "Precomputed kernel classification test score: 0.9366666666666666\n",
      "Precomputed kernel classification test score: 0.96\n",
      "Precomputed kernel classification test score: 0.9633333333333334\n",
      "Precomputed kernel classification test score: 0.9833333333333333\n",
      "Precomputed kernel classification test score: 0.9633333333333334\n",
      "Precomputed kernel classification test score: 0.9466666666666667\n"
     ]
    }
   ],
   "source": [
    "score = []\n",
    "for i in range(7):\n",
    "    train_size = 700\n",
    "    \n",
    "    sample_train_1 = sample_train[i*train_size:(i+1)*train_size]\n",
    "    label_train_1 = label_train[i*train_size:(i+1)*train_size]\n",
    "\n",
    "    test_size = 300\n",
    "    sample_test_1 = sample_test[i*test_size:(i+1)*test_size]\n",
    "    label_test_1 = label_test[i*test_size:(i+1)*test_size]\n",
    "    \n",
    "    \n",
    "    zz_map = ZZFeatureMap(feature_dimension=16, reps=1, entanglement='linear', insert_barriers=True)\n",
    "    zz_kernel = QuantumKernel(feature_map=zz_map, quantum_instance=Aer.get_backend('statevector_simulator'))\n",
    "    matrix_train = zz_kernel.evaluate(x_vec=sample_train_1)\n",
    "    matrix_test = zz_kernel.evaluate(x_vec=sample_test_1, y_vec=sample_train_1)\n",
    "    \n",
    "    zzpc_svc = SVC(kernel='precomputed')\n",
    "    zzpc_svc.fit(matrix_train, label_train_1)\n",
    "    zzpc_score = zzpc_svc.score(matrix_test, label_test_1)\n",
    "\n",
    "    print(f'Precomputed kernel classification test score: {zzpc_score}')\n",
    "    score.append(zzpc_score)\n",
    "    del matrix_train, matrix_test"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e33ae6b",
   "metadata": {},
   "source": [
    "El valor promedio para el conjunto que se dio del auto encoder usando Hold-out 70-30 se obtuvo  un valor de desempeño de "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "dca7ed51",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9604761904761905"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(score)/len(score)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e696fc3",
   "metadata": {},
   "source": [
    "## Autores\n",
    "\n",
    "- Martínez Vázquez María Fernanda (undergraduate)\n",
    "- Navarro Ambriz Ronaldo  (undergraduate)\n",
    "- Martinez Hernandez Luis Eduardo (undergraduate)\n",
    "- Galindo Reyes Agustin (undergraduate)\n",
    "- Alberto Maldonado Romo (master)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "75262f05",
   "metadata": {},
   "source": [
    "# Referencias\n",
    "\n",
    "\n",
    "[1] Bravo-Prieto, Carlos. (2020). Quantum autoencoders with enhanced data encoding. \n",
    "\n",
    "[2] Biamonte, Jacob. (2019). Lectures on Quantum Tensor Networks. \n",
    "\n",
    "[3] Kardashin, Andrey & Uvarov, Aleksey & Biamonte, Jacob. (2021). Quantum Machine Learning Tensor Network States. Frontiers in Physics. 8. 586374. 10.3389/fphy.2020.586374. \n",
    "\n",
    "[4] Stoudenmire, E. & Schwab, David. (2016). Supervised Learning with Quantum-Inspired Tensor Networks. \n",
    "\n",
    "[5] Liu, Ding & Yao, Zekun & Zhang, Quan. (2020). Quantum-Classical Machine learning by Hybrid Tensor Networks\n",
    "\n",
    "[6] Romero, Jonathan & Olson, Jonathan & Aspuru-Guzik, Alán. (2016). Quantum autoencoders for efficient compression of quantum data. Quantum Science and Technology. 2. 10.1088/2058-9565/aa8072. \n",
    "\n",
    "[7] Foulds, Steph & Kendon, Viv & Spiller, Tim. (2020). The controlled SWAP test for determining quantum entanglement. "
   ]
  }
 ],
 "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}