Ch04.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 04 분류하는 뉴런을 만듭니다"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "이 노트북을 주피터 노트북 뷰어(nbviewer.jupyter.org)로 보거나 구글 코랩(colab.research.google.com)에서 실행할 수 있습니다.\n",
    "\n",
    "<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
    "  <td>\n",
    "    <a target=\"_blank\" href=\"https://nbviewer.jupyter.org/github/rickiepark/do-it-dl/blob/master/Ch04.ipynb\"><img src=\"https://jupyter.org/assets/main-logo.svg\" width=\"28\" />주피터 노트북 뷰어로 보기</a>\n",
    "  </td>\n",
    "  <td>\n",
    "    <a target=\"_blank\" href=\"https://colab.research.google.com/github/rickiepark/do-it-dl/blob/master/Ch04.ipynb\"><img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" />구글 코랩(Colab)에서 실행하기</a>\n",
    "  </td>\n",
    "</table>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 04-2 시그모이드 함수를 알아봅니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEGCAYAAACKB4k+AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy86wFpkAAAACXBIWXMAAAsTAAALEwEAmpwYAAAbKUlEQVR4nO3de5hcdZ3n8fe3qvqadNK5dC7mQiOESwRBbBmUWRVBnwy6xN1FHryMkWE2c8EZR+fZldnRR5+Z2X10V8fLzixOHHCisgKLjMTxNhkGBRxBwj0kICFALjRJNyTp7nR3Xb/7xzlVXWk7SSfUOaeq6/N6nnrqXOt8T3dSnz6/37mYuyMiIgKQSroAERGpHwoFERGpUCiIiEiFQkFERCoUCiIiUpFJuoBXY+HChd7b25t0GSIiDeWhhx4adPeeqeY1dCj09vayZcuWpMsQEWkoZvbC0eap+UhERCoUCiIiUqFQEBGRishCwcxuMrP9Zra1atp8M9tsZs+E7/PC6WZmXzWzHWb2uJldEFVdIiJydFEeKfwDsGbStOuBu9x9FXBXOA7wW8Cq8LUeuCHCukRE5CgiCwV3vwd4ZdLktcDGcHgj8N6q6d/0wP1At5ktjao2ERGZWtx9CovdvT8cfglYHA4vA3ZXLbcnnPZrzGy9mW0xsy0DAwPRVSoi0oQS62j24J7dJ3zfbnff4O597t7X0zPltRciIjPWaK7AF37yNI/uPhjJ58cdCvvKzULh+/5w+l5gRdVyy8NpIiJSZWiswN/cvYNtLw5F8vlxh8ImYF04vA64s2r6h8OzkC4CDlU1M4mISChbKALQlonm6zuy21yY2XeAtwMLzWwP8Bngc8BtZnYt8AJwVbj4D4HLgR3AKHBNVHWJiDSyXKEEQFtLg4WCu7//KLMunWJZB66LqhYRkZkiWw6FTDqSz9cVzSIiDSTq5iOFgohIA8nmy0cKCgURkaZXbj5qVSiIiMhE85H6FEREml424rOPFAoiIg1EfQoiIlKh5iMREalQ85GIiFRMXLymUBARaXqVU1LTCgURkaaXLRRpy6Qws0g+X6EgItJAsvlSZBeugUJBRKShZAulyM48AoWCiEhDKTcfRUWhICLSQLKFUmSno4JCQUSkoeTUfCQiImVBn4KOFEREBMjm1acgIiKhoE9BzUciIkIQClFdzQwKBRGRhpItFHX2kYiIBLJ5dTSLiEhIVzSLiEiFrmgWEZGKnK5oFhERAHdX85GIiARyxWifugYKBRGRhhH1ozhBoSAi0jCyeYWCiIiEsoUigPoURESkqvlopp19ZGYfN7MnzWyrmX3HzNrN7FQze8DMdpjZrWbWmkRtIiL1akY2H5nZMuCPgT53PwdIA1cDnwe+5O6nAweAa+OuTUSkns3k5qMM0GFmGaAT6AfeAdwezt8IvDeZ0kRE6lNuJp595O57gS8AuwjC4BDwEHDQ3QvhYnuAZVOtb2brzWyLmW0ZGBiIo2QRkbowI/sUzGwesBY4FXgNMAtYM9313X2Du/e5e19PT09EVYqI1J+J6xRmVvPRZcBz7j7g7nngDuBioDtsTgJYDuxNoDYRkbpV7lNonUnNRwTNRheZWaeZGXApsA24G7gyXGYdcGcCtYmI1K0ZefaRuz9A0KH8MPBEWMMG4JPAJ8xsB7AAuDHu2kRE6lkczUeZ4y9Se+7+GeAzkybvBC5MoBwRkYYwcUrqDDpSEBGRkzMjzz4SEZGTU+5TaE0rFEREml6uWCSTMjIKBRERyeZLkfYngEJBRKRhZAsl2lqiO/MIFAoiIg0jWyhG2p8ACgURkYYRHCkoFEREBPUpiIhIlWyhGOnVzKBQEBFpGNmCjhRERCSUU5+CiIiUBUcKaj4SERHKfQo6UhAREYIjhSgfsAMKBRGRhqFTUkVEpEKnpIqISIVOSRURkQrd5kJERAAoFEsUS67mIxERgVwxfBSnmo9ERKT8KE6FgoiIkC2EoaCH7IiISLZQBNBDdkREpPpIQaEgItL0JvoU1HwkItL0ys1H6mgWEZGJ5iOFgoiIVI4UdPaRiIjkdKQgIiJlaj4SEZGK8tlHM/IhO2bWbWa3m9lTZrbdzN5sZvPNbLOZPRO+z0uiNhGRejRx9tHM7FP4CvBjdz8LOA/YDlwP3OXuq4C7wnEREWEGX7xmZnOBtwI3Arh7zt0PAmuBjeFiG4H3xl2biEi9msl9CqcCA8A3zOwRM/t7M5sFLHb3/nCZl4DFU61sZuvNbIuZbRkYGIipZBGRZGXzM/feRxngAuAGd38DcJhJTUXu7oBPtbK7b3D3Pnfv6+npibxYEZF6UH4Up5lFup0kQmEPsMfdHwjHbycIiX1mthQgfN+fQG0iInUpjuczQwKh4O4vAbvN7Mxw0qXANmATsC6ctg64M+7aRETqVfB85mjPPIKgKScJfwTcbGatwE7gGoKAus3MrgVeAK5KqDYRkbqTLRRjOVJIJBTc/VGgb4pZl8ZciohIQ8gWSpFfuAa6ollEpCFk86XIL1wDhYKISEOom+YjM1sOXA38O+A1wBiwFfgB8CN3L0VaoYiI1MfZR2b2DeAmIAd8Hng/8IfAvwBrgPvM7K1RFyki0uzq5eyjL7r71immbwXuCM8eWln7skREpFo2X6Stqy3y7RwzFKoDIQyAswiuNH46vGdRDtgRbYkiIpKLqfloWqekmtm7ga8BzwIGnGpmv+fuP4qyOBERCQR9Csk3H5V9EbjE3XcAmNlphB3NURUmIiIT6u06heFyIIR2AsMR1CMiIlOom1NSQ1vM7IfAbQR9Cu8DHjSz/wjg7ndEVJ+IiFA++6h+QqEd2Ae8LRwfADqAf08QEgoFEZGIuHvY0VwnfQrufk3UhYiIyNTieuoanMRtLszs4SgKERGRqdV1KBCckioiIjHJFoJHccZxRfPJhMIPal6FiIgcVa6ejxTc/VNRFCIiIlOr9+YjAMzsiVoWIiIiUxsaywMwuy3656Idcwvl6xCmmgUsqX05IiIy2eBIDoCepG+IB9wK3ExwLcJk7bUvR0REJhsYzgL1EQqPA1+Y6vbZZnZZNCWJiEi1cigsmBV9KByvT+FPgKGjzPsPtS1FRESmMjiSpbuzJZYb4h3veQr3HmPeltqXIyIikw0MZ+mZHf1RAhz/cZyfMrP5x5j/DjN7T+3LEhGRsoGRLAtjCoXj9Sk8AXzfzMaBhwluhNcOrALOJ3hW8/+IskARkWY3OJLlvOXdsWzreM1HdwJ3mtkq4GJgKUEfw7eB9e4+Fn2JIiLNbWA4G8uZR3D86xT+DPixuz8CPBNLRSIiUnE4W2A0V6yb5qOdwMfM7DzgMYLHb/6zux+IvDIREWFwJL5rFOD4zUe3ElzAhpm9AVgDfNfMMgT9CT92919GXqWISJMqX6OwcHZrLNub1kmvZtYOXAJcCBwENgPPAb8bWWUiIlJfRwpVvgkMA18Nxz8AvN7d3xdJVSIiAsR7iwuYfiic4+6rq8bvNrNtURQkIiITBoazmMH8zjpqPgIeNrOLyiNm9hvAq7qi2czSZvaImf1TOH6qmT1gZjvM7FYzi+cnICJSxwZGciyY1UomHf0tLmD6ofBG4N/M7Hkzex74BfAmM3vCzB4/yW1/DNheNf554EvufjpwALj2JD9XRGTGGBiO72pmmH7z0ZpabtTMlgPvBv478AkzM+AdBH0VABuBzwI31HK7IiKNZnAkvgvXYJqh4O4v1Hi7Xwb+K9AVji8ADrp7IRzfAyybakUzWw+sB1i5cmWNyxIRqS8Dw1leu3BWbNuLp5GqSngDvf3u/tDJrO/uG9y9z937enp6alydiEj9cHcG6vFIocYuBq4ws8sJbq43B/gK0G1mmfBoYTmwN4HaRETqxnC2QK5QirVPIfYjBXf/M3df7u69wNXAv7r7B4G7gSvDxdYBd8Zdm4hIPYn7GgVIIBSO4ZMEnc47CPoYbky4HhGRRCURCkk0H1W4+0+Bn4bDOwluoyEiIkzc4mJGNx+JiMj0NHvzkYiIVBkcyZJJGd0dLbFtU6EgIlKnBoazLJjdSiplsW1ToSAiUqfifAxnmUJBRKRODY7k6ImxkxkUCiIidSvum+GBQkFEpC6VSh77zfBAoSAiUpcOjeUplFyhICIiMJDAhWugUBARqUt7D44BsGRue6zbVSiIiNSh7f1DAJyxuOs4S9aWQkFEpA5t7x9mWXcHc2O8mhkUCiIidWl7/xBnL50T+3YVCiIidWY8X2TnwAirX6NQEBFpek+/NEzJYfXSePsTQKEgIlJ3yp3Maj4SERG29Q8xuy3DinmdsW9boSAiUme29w9x1pKuWG+ZXaZQEBGpI6WSs71/OJGmI1AoiIjUlT0HxhjJFhQKIiIS9CcAiZyOCgoFEZG6sq1/iJTBmTHf3qJMoSAiUke29w/Ru3AWHa3pRLavUBARqSPb+4dYnVB/AigURETqxqGxPHsOjCXWyQwKBRGRuvFUuZNZoSAiIvftGCRlcN6K7sRqUCiIiNSJzdv20XfKfObPak2sBoWCiEgd2P3KKE+9NMw7Vy9OtA6FgohIHdi8bR+AQkFERIJQWLVoNr0LZyVaR+yhYGYrzOxuM9tmZk+a2cfC6fPNbLOZPRO+z4u7NhGRJBwczfHL519J/CgBkjlSKAB/6u6rgYuA68xsNXA9cJe7rwLuCsdFRGa8u5/eT7HkzRkK7t7v7g+Hw8PAdmAZsBbYGC62EXhv3LWJiCRh87Z9LOpq47zl3UmXkmyfgpn1Am8AHgAWu3t/OOslYMrINLP1ZrbFzLYMDAzEU6iISESyhSI/e3qAS89enMhDdSZLLBTMbDbwXeBP3H2oep67O+BTrefuG9y9z937enp6YqhURCQ6/7bjZQ7niryrDpqOIKFQMLMWgkC42d3vCCfvM7Ol4fylwP4kahMRidPGXzzPwtmtvPm0BUmXAiRz9pEBNwLb3f2vq2ZtAtaFw+uAO+OuTUQkTtv7h/jp0wN85C29tLckc6vsyTIJbPNi4LeBJ8zs0XDafwM+B9xmZtcCLwBXJVCbiEhsvn7PTjpb03zoolOSLqUi9lBw9/uAo/WmXBpnLSIiSdl7cIxNj73Iurf00t2Z3L2OJtMVzSIiCbjx3ucA+J3fPDXhSo6kUBARidnB0Ry3PLiLK857Dcu6O5Iu5wgKBRGRmH3tZzsZzRVZ/7bXJl3Kr1EoiIjEaOveQ3z93p1c+cblnLUkuSesHY1CQUQkJoViiU9+93Hmdbby6XevTrqcKSVxSqqISFP6+r3P8eSLQ9zwwQuY29mSdDlT0pGCiEgMnhs8zJf/5Vesed0SfuvcpUmXc1QKBRGRiA2N5/n9bz1EWybFX6x9XdLlHJOaj0REIpQvlrju5od5dmCEjb9zIYvmtCdd0jEpFEREIuLufPp7W7n3mUH+55Wv5+LTFyZd0nGp+UhEJALuzpc2/4pbHtzNRy85nav6ViRd0rToSEFEpMaKJeezm57kW/e/wPveuJxPvPOMpEuaNoWCiEgNjeeLfPzWR/nR1pf4vbe9luvXnEXwxIDGoFAQEamRXS+P8se3PMKjuw/y6fes5to6u9nddCgURERq4M5H9/Ln/7gVM7jhgxfU9bUIx6JQEBF5FfYNjfNXP9jO9x97kb5T5vHlq89n+bzOpMs6aQoFEZGTkCuUuOnnz/G/73qGfNH5+GVncN0lp5FJN/ZJnQoFEZETkC+W+MeH9/I3d+9g1yujXHb2Ij79ntWcsmBW0qXVhEJBRGQaDmcL3PHIXv7uZ8+y58AY5y6byzeueROXnLko6dJqSqEgInIMO/YP8+37d/Hdh/YwnC1w3opu/nLtObz9zJ6GOtV0uhQKIiKT7B8a5/uP9/O9R/byxN5DtKSNy89dyofffAoXrJw3I8OgTKEgIk3P3dk5eJjN2/axeds+Ht51AHc4Z9kcPvXus1l7/jJ6utqSLjMWCgURaUr7h8f55XOvcN8zg9z7zCB7D44BQRB87NJVvOf1Szl9UVfCVcZPoSAiM16hWOJX+0Z4bM9BHtl1gAefP8Bzg4cB6GrP8JbTFvD7bz+Nd5y1iGXdHQlXmyyFgojMKMPjeZ7ZP8LTLw3z5IuHePLFIZ7qH2YsXwRgXmcLbzxlPu+/cAVv6p3PucvmNvy1BbWkUBCRhlMsOf2Hxtj18ig7Bw/z7MAIzw4c5tn9I5VmIAiOAlYvncPVF67g/BXdnL+im5XzO2d0R/GrpVAQkbqTL5bYP5yl/+AYew+O8eLBcfYcGGX3gTH2vDLKngNj5IqlyvIdLWlOWzSLvt55fGDxSs5c3MWZS7pYPq9DAXCCFAoiEgt3ZzRX5OWRHIOHswwMZxkcyTI4nGPf8Dj7h7LsHx7npUPjDIxkcT9y/e7OFlbM6+TMJV28c/ViTlkwi94FnfQunMWSOe2kUvryrwWFgoicsGyhyNBYgUNjeQ6N5Rkay3NwLMfB0Xz4ynFgNM+B0VzwOpzn5cNZxvOlKT9v/qxWFnW1sWhOO2ct6WLJ3A6Wzm1nydx2lnd3sLS7g9lt+rqKg37KIk3A3ckWSozmiozmCozmihzOFjicLXI4V2A0V2AkW55WYCRbYGQ8eB8eLzA8nmd4vMDQeIGh8Ty5wtRf7mVz2jPMm9VKd0cLPbPbOGNxFwtmtbJgdlv43krP7HZ6utpYMLuVFnX01g2FgkgCiiUnVyiRK5TIFopkCyWyVePj+Ynp4/lwfj6YPp4vMl6YGB7LF8nmS4zli4zlgvHxfDEMgCJjuQJj+SIlP35dAGYwuzVDV3uG2e0ZZrdl6O5sZcX8TrraW5jTkWFOewtz2jPM6WhhbkcLczpamNcZhEBXe0Zn8zSwugoFM1sDfAVIA3/v7p9LuCSpY+5OoeQUS+F70cmXSpXxQrEUvjv5ynCJfNEplCbmFYol8uXli06uWKosly+VyBeC5XPh/HyxRL5YIleYGM4XS+SKTr4QLBfMD4bLX/7Vw4XpfkMfRTpltGdStLekaW9J09Gapr0lRUdLmq72DIu62uhsTdPRmqGjJU1na5rOtjSdLWk62zLMas0E01rTzGrLBK/WNLPbg+XVOdu86iYUzCwN/C3wTmAP8KCZbXL3bclWduLcHXcouVMK38vjTvBXolfNq54fzJtYt3rZYmnSsv7r80qlSfNKVIaLRwx7ZZ1iaaKOYBqUSsFnFEteGZ6YNlHrxGdMvBcq6wSfUyiVKuuU5xXCL+/Ky4Mv6GLVdsvjwRd/6cgACF9xMYOWVIrWTIqWtJFJp2hNHzncmgneO1rSzGnPhMtOTD/iPZOiLZOmrTKcoq0lTWs6RXvLxPz2lonlggAIxlsz+ktcolE3oQBcCOxw950AZnYLsBaoeSjc9uBuNty784gv2PJw8Kr6MufI8VKpPG1ivdLkdeL7ropdOmWkzUilCN+talrwnk5NeoXzMqnyMpBJpcikUrRlgmXK8zJV6wXLBNNb0hOflUmnjlwubbSkUkcslymvn554bzliOEU6Faw31byW8Au/PJzWmS3SJOopFJYBu6vG9wC/MXkhM1sPrAdYuXLlSW2ou7OFMxd3YQYpsyPeDSNVPW7BuFnwJVg+rE6nytON1BTLlIfLp8mlU4Yx8bnB+hPbTk1ax8rrhPOrt1f+Uq4Mh9tPpY4cTtvk9Y10amJ76dTEF3kqnF75vBRV69gR+yIiM1c9hcK0uPsGYANAX1/fSf1N/q7XLeFdr1tS07pERGaCemqY3AusqBpfHk4TEZGY1FMoPAisMrNTzawVuBrYlHBNIiJNpW6aj9y9YGYfBX5CcErqTe7+ZMJliYg0lboJBQB3/yHww6TrEBFpVvXUfCQiIglTKIiISIVCQUREKhQKIiJSYd7A92QwswHghZNcfSEwWMNyGkUz7ncz7jM053434z7Die/3Ke7eM9WMhg6FV8PMtrh7X9J1xK0Z97sZ9xmac7+bcZ+htvut5iMREalQKIiISEUzh8KGpAtISDPudzPuMzTnfjfjPkMN97tp+xREROTXNfORgoiITKJQEBGRihkfCma2xsyeNrMdZnb9FPPbzOzWcP4DZtabQJk1NY19/oSZbTOzx83sLjM7JYk6a+14+1213H8yMzezhj91cTr7bGZXhb/vJ83s/8ZdYxSm8W98pZndbWaPhP/OL0+izloys5vMbL+ZbT3KfDOzr4Y/k8fN7IKT2pCHD3KfiS+CW3A/C7wWaAUeA1ZPWuYPga+Fw1cDtyZddwz7fAnQGQ7/QaPv83T3O1yuC7gHuB/oS7ruGH7Xq4BHgHnh+KKk645pvzcAfxAOrwaeT7ruGuz3W4ELgK1HmX858CPAgIuAB05mOzP9SOFCYIe773T3HHALsHbSMmuBjeHw7cClVn4Qc2M67j67+93uPhqO3k/wlLtGN53fNcBfAp8HxuMsLiLT2ef/DPytux8AcPf9MdcYhenstwNzwuG5wIsx1hcJd78HeOUYi6wFvumB+4FuM1t6otuZ6aGwDNhdNb4nnDblMu5eAA4BC2KpLhrT2edq1xL8ddHojrvf4eH0Cnf/QZyFRWg6v+szgDPM7Odmdr+ZrYmtuuhMZ78/C3zIzPYQPKPlj+IpLVEn+n9/SnX1kB2Jl5l9COgD3pZ0LVEzsxTw18BHEi4lbhmCJqS3ExwR3mNm57r7wSSLisH7gX9w9y+a2ZuBb5nZOe5eSrqwejfTjxT2AiuqxpeH06ZcxswyBIeaL8dSXTSms8+Y2WXAnwNXuHs2ptqidLz97gLOAX5qZs8TtLluavDO5un8rvcAm9w97+7PAb8iCIlGNp39vha4DcDdfwG0E9w0biab1v/945npofAgsMrMTjWzVoKO5E2TltkErAuHrwT+1cNemwZ13H02szcAf0cQCDOhjRmOs9/ufsjdF7p7r7v3EvSlXOHuW5Iptyam8+/7ewRHCZjZQoLmpJ0x1hiF6ez3LuBSADM7myAUBmKtMn6bgA+HZyFdBBxy9/4T/ZAZ3Xzk7gUz+yjwE4IzFm5y9yfN7C+ALe6+CbiR4NByB0EnztXJVfzqTXOf/xcwG/h/YZ/6Lne/IrGia2Ca+z2jTHOffwK8y8y2AUXgv7h7Ix8JT3e//xT4upl9nKDT+SMN/sceZvYdgoBfGPaVfAZoAXD3rxH0nVwO7ABGgWtOajsN/nMSEZEamunNRyIicgIUCiIiUqFQEBGRCoWCiIhUKBRERKRCoSAiIhUKBRERqVAoiNSQmfWa2VNmdrOZbTez282sM+m6RKZLoSBSe2cC/8fdzwaGCJ7ZIdIQFAoitbfb3X8eDn8b+M0kixE5EQoFkdqbfO8Y3UtGGoZCQaT2Vob38Af4AHBfksWInAiFgkjtPQ1cZ2bbgXnADQnXIzJtM/rW2SIJKbj7h5IuQuRk6EhBREQq9DwFERGp0JGCiIhUKBRERKRCoSAiIhUKBRERqVAoiIhIxf8Hl5k08d52F1kAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "probs = np.arange(0, 1, 0.01)\n",
    "odds = [p/(1-p) for p in probs]\n",
    "plt.plot(probs, odds)\n",
    "plt.xlabel('p')\n",
    "plt.ylabel('p/(1-p)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "probs  = np.arange(0.001, 0.999, 0.001)\n",
    "logit = [np.log(p/(1-p)) for p in probs]\n",
    "plt.plot(probs, logit)\n",
    "plt.xlabel('p')\n",
    "plt.ylabel('log(p/(1-p))')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEGCAYAAABo25JHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy86wFpkAAAACXBIWXMAAAsTAAALEwEAmpwYAAAjbElEQVR4nO3deZRcdZ338fe39yydfU86GyQhYUlImgjIJlsCQjKjgGEURRgzKpnjHJcZfPRBHvTMM8roM3iGEaOisgyIOGDUYBM2USE7AdJZSJO1O+kla3fS6aW6vs8fdROKpjvpTrrq1vJ5nVOpqnt/VfXpW5X61v3d5WfujoiIZK+csAOIiEi4VAhERLKcCoGISJZTIRARyXIqBCIiWS4v7ADdNWTIEB8/fnzYMURE0sqaNWv2uvvQjualXSEYP348q1evDjuGiEhaMbMdnc1T15CISJZTIRARyXIqBCIiWU6FQEQky6kQiIhkuYQVAjN72MxqzWx9J/PNzH5oZhVm9paZzUxUFhER6Vwi1wh+Acw9wfzrgEnBZSHwowRmERGRTiTsOAJ3f9XMxp+gyXzgEY+dB3u5mQ0ws5HuvidRmUQkc7g7zZEoza1RmiJttESiRKJOWzRKa5vTFnUiUSfSdmy609oWDa6PzY8Sdccdoh57Tndw4qbhRB1wf68NH2wfuw/R4NT+x+YB+Ptyx92Om/P+6R0/4Kqpw5leMqCHluB7wjygbDSwK+5+ZTDtA4XAzBYSW2tg7NixSQknIonj7tQ3RahraKK2oZn9R1qoPxqhvqmV+qOtwXWEhqZW6psiHGmO0ByJ0tTaFlxiX/7ZMpyKWex6WL+ijCsEXebui4HFAKWlpVny1oukL3en6uBRdu5rZOf+Rnbsj13vPniUuoZm6hqaaY5EO3xsXo5RXJRHv1759CvKp7goj4G9e9OrIJeivByK8nMpyj92nXv8fn5uDvm5Rl5ODnk5Rl5u7Do3x8gLpufmGPm5wbTgfm6OkWNgGGYEl/em5RhgkGOG8f55lgNGMC9oA8FzHHtswOy9O3GTsU7aJFOYhaAKKIm7PyaYJiJppLUtSvnuetZXHWJTdT2b9jSwqbqBw82R423ycowxA3sxemAvLhg/iKHFhQwrLmRocBnSt5B+Rfn065VHr/zc0L4Qs1WYhWAJsMjMngQ+BBzS9gGR1NcSibJ25wFWbtvPym37WbvzAI0tbQAUF+UxdUQ/PjZzNFNGFDNhcB/GDu7NyP69yM3Rl3uqSlghMLMngCuAIWZWCXwLyAdw94eApcD1QAXQCHw2UVlE5PQ0NLXy4sZalm2s4dXNdTQ0RzCDKcOLuXnWGC6YMIgZJQMYPaCXfs2noUTuNXTrSeY7cFeiXl9ETk9b1PlLxV5+s6aSsvJqmiNRhhYX8tHzRnLlWcP40ITB9O+dH3ZM6QFpsbFYRJLncHOEp1bt4uevbWPX/qP075XPLaUl/M35ozm/ZAA56uLJOCoEIgJAfVMrP3l1K7/463YamiOUjhvI3XOncvW0YRTm5YYdTxJIhUAkyzW1tvHY8h08+HIFBxpbuf7cEXzu0omcP3Zg2NEkSVQIRLLY6+/u43898zbb9h7h0klD+Oc5Z3HumP5hx5IkUyEQyUKHjrbyf5du5MlVuxg7qDeP3DGbyyZ3OJytZAEVApEss27XQe56fC3V9U38w+UT+aerJtOrQNsAspkKgUiWcHceeX0H3/nDBoYVF/GbL1zMjASct0bSjwqBSBZoiUT5l9+8xTNvVHHVWcP4/i3TGdC7IOxYkiJUCEQy3OHmCF94bA1/3rKXL18zmUUfOVPHAsj7qBCIZLC9h5v57M9XsWFPPfffdB43l5ac/EGSdVQIRDLUvsPNfOLHr1N18CiLb5vFVVOHhx1JUpQKgUgGamhq5TM/X0nlgaP88o7ZXDhxcNiRJIUlcsxiEQlBU2sbd/5yNZv2NPDQp2apCMhJaY1AJINEo84/PvEGq7bv5z8+MYOPnDUs7EiSBrRGIJJB/uPFLSzbUMM9N0xj/ozRYceRNKFCIJIhni+v5ocvbuHmWWO4/eLxYceRNKJCIJIBKmoP8+Wn3uS8Mf359t+co1HCpFtUCETSXFNrG59/bA2FeTk89KlZFOXrvEHSPdpYLJLmvvfHzVTUHubRO2czakCvsONIGtIagUgae/3dfTz81218+qJxXDpJp5GWU6NCIJKmGppa+eqv32TCkD7cfd1ZYceRNKauIZE09Z3fb2TPoaM8/YWL6V2g/8py6rRGIJKGVmzdx69W72LhZWcwU2MLy2lSIRBJM5G2KN9aUs7oAb340lWTwo4jGUCFQCTNPL5iJ5uqG/jmR6dqiEnpESoEImlk3+Fmvv/8Zi45cwhzzxkRdhzJECoEImnk/rLNNLa0ce+8aTp6WHqMCoFImthUXc+vVu/i9ovHc+aw4rDjSAZRIRBJE99//h36FuSx6Mozw44iGUaFQCQNrNt1kGUbavjcZRMZ0Lsg7DiSYVQIRNLA95/fzMDe+dxxyYSwo0gGSmghMLO5ZrbZzCrM7O4O5o81s5fN7A0ze8vMrk9kHpF0tGLrPv68ZS9fuOIM+hbqCGLpeQkrBGaWCzwIXAdMA241s2ntmn0TeMrdzwcWAP+VqDwi6cjd+ffnNzOsuJBPXzQ+7DiSoRK5RjAbqHD3re7eAjwJzG/XxoF+we3+wO4E5hFJO8u37mfV9gMsuvJMjTMgCZPIQjAa2BV3vzKYFu9e4FNmVgksBf6xoycys4VmttrMVtfV1SUiq0hK+vGr7zK4TwG3lJaEHUUyWNgbi28FfuHuY4DrgUfN7AOZ3H2xu5e6e+nQoTrnumSHTdX1vLK5jtsvHq+1AUmoRBaCKiD+Z8yYYFq8O4GnANz9daAIGJLATCJpY/GrW+mVn8ttF40LO4pkuEQWglXAJDObYGYFxDYGL2nXZidwFYCZTSVWCNT3I1lv98GjLFm3mwWzS3TcgCRcwgqBu0eARUAZsJHY3kHlZnafmc0Lmn0F+JyZvQk8Adzu7p6oTCLp4ud/3YYDd+q4AUmChO6U7O5LiW0Ejp92T9ztDcCHE5lBJN3UN7Xy3yt2csN5IxkzsHfYcSQLhL2xWETa+c2aSo60tPH3l0wMO4pkCRUCkRTi7jy2fAczSgZw7pj+YceRLKFCIJJCXt+6j3frjnDbhdpTSJJHhUAkhTy2fAcDeufz0fNGhh1FsogKgUiKqKlvoqy8hltKS3QAmSSVCoFIinhi5U7aos4nPzQ27CiSZVQIRFJAa1uUJ1bu5PLJQxk3uE/YcSTLqBCIpICXN9VSU9/Mp7SRWEKgQiCSAp5eU8mQvoV8ZIpOqijJp0IgErK9h5t5aVMtH5s5mrxc/ZeU5NOnTiRkv123m0jUuWnWmLCjSJZSIRAJkbvz69W7mD6mP5OHF4cdR7KUCoFIiMp317OpukFrAxIqFQKRED29ppKC3BzmTW8/iqtI8qgQiISkJRLlt+uquObs4fTvnR92HMliKgQiIXl5cy0HGlvVLSShUyEQCcmSdbsZ3KeAS8/UMN0SLhUCkRA0NLXywsYaPnreSB07IKHTJ1AkBMs21NAciTJv+qiwo4ioEIiEYcmbuxk9oBczxw4MO4qICoFIsu073Myft+zlxumjyMmxsOOIqBCIJNvS9dW0RV3dQpIyVAhEkmzJuiomDevL1JE6pYSkBhUCkSSqOniUVdsPMG/6KMzULSSpQYVAJImee3sPADeqW0hSiAqBSBKVlVdz1ohixg/RcJSSOlQIRJKkrqGZ1TsOcO3ZI8KOIvI+KgQiSfLCxhrcYc7Zw8OOIvI+KgQiSVJWXs2Ygb2YNrJf2FFE3keFQCQJGppaea1iH3POHqG9hSTldKkQmFmRmd1kZg+Y2a/N7BEz+2czO/skj5trZpvNrMLM7u6kzS1mtsHMys3sv0/ljxBJdS9vrqOlLcocbR+QFJR3sgZm9n+AG4BXgBVALVAETAb+zcyKgK+4+1vtHpcLPAhcA1QCq8xsibtviGszCfg68GF3P2Bmw3rkrxJJMWXl1QzuU8CscTq3kKSekxYCYKW7f6uTeT8IvrzHdjBvNlDh7lsBzOxJYD6wIa7N54AH3f0AgLvXdjm5SJpojrTxyqZabpw+ilydW0hS0Em7htz9DwBmdmnwK/84M5vp7rXuvrqDh44GdsXdrwymxZsMTDazv5rZcjOb21EGM1toZqvNbHVdXd3JIouklNcq9nGkpU3dQpKyurOxuAx4qV33zU9P8/XzgEnAFcCtwE/MbED7Ru6+2N1L3b106NChp/mSIslVVl5N38I8Lj5zcNhRRDrUnUKwGbgf+JOZXRxMO9F6bhVQEnd/TDAtXiWwxN1b3X0b8A6xwiCSEdqizrINNVwxZSiFebknf4BICLpTCNzdfw/MA/7TzBYBfoL2q4BJZjbBzAqABcCSdm2eJbY2gJkNIdZVtLUbmURS2podB9h3pEXdQpLSulMIDMDdtwCXBZfzOmvs7hFgEbEupY3AU+5ebmb3mdm8oFkZsM/MNgAvA19z933d/zNEUlNZeTUFuTlcMUVdmpK6urLXEADufn7c7cPALWbW0d5C8Y9ZCixtN+2euNsOfDm4iGQUd6esvJoPnzmY4qL8sOOIdOqUjiw2s7UA7r6zZ+OIZI4Ne+qpPHBU3UKS8k71FBPaGVrkJMrKa8gxuHqaTjInqe1UC8EfejSFSAZ6vrya0nGDGNK3MOwoIid0SoXA3b/Z00FEMsmOfUfYVN3AtTrltKSBLhcCM7s9gTlEMkpZeTWAtg9IWujq2UfvAa5OcBaRjFFWXsO0kf0oGdQ77CgiJ3XSQmBmi4GpwKcTH0ck/dU2NLF25wGtDUja6MpxBLcCH3L3aKLDiGSCZRuCISnP0fYBSQ9d6Rq6EXjKzM5IdBiRTFBWXsO4wb2ZMrw47CgiXdKV01C/Quw8QY8lPI1ImqtvauX1d/dqSEpJK13aWOzu64GPJziLSNp7eVMtrW3OHO02Kmmky7uPuvvuRAYRyQRl5dUMLS7k/BINSSnpozvHEUw2sxfNbH1w/zwz04FlIoGm1jZe2VzHNdOGk6MhKSWNdOfI4p8QG2i+FSAYrH5BIkKJpKO/bNlLo4aklDTUnULQ291XtpsW6ckwIumsrLya4qI8LpqoISklvXSnEOwNdiF1ADO7CdiTkFQiaSbSFuWFjTVcedYwCvJO9VyOIuHo8sA0wF3AYuAsM6sCtgGfTEgqkTSzavsBDjS2qltI0lJ3RijbClxtZn2AHHdvSFwskfRSVl5NQV4Ol0/WkJSSfk5lHfZiFQGR97g7yzbUcNmkIfQp7M5KtkhqOJVC8N0eTyGSxtZX1VN18CjXqltI0pS2aomcprLy6tiQlFN1NLGkpy6tx5rZz4ntLWTAWDN7+Ng8d78jQdlE0kJZeTUXjB/EoD4FYUcROSVd7dD8RdztS4Bf9nwUkfSzte4wW2oPc88N08KOInLKulQI3P1Px26bWUP8fZFsVlZeA6CxiSWtnco2gpYeTyGSpsrKqzlndD/GDNSQlJK+ul0I3P3CRAQRSTfVh5pYt+sgc6ZpbyFJb6e115CZ9e2pICLpZtmGagDmnKNCIOntdHcf3dAjKUTSUFl5DROG9GHSMP0ekvR20o3FZvblzmYB+h8gWelQYyvLt+7jzksnaEhKSXtdWSP4V2AgUNzu0reLjxfJOMs21hCJOnN1NLFkgK7sProWeNbd17SfYWZ/3/ORRFLfH9fvYVT/ImaUDAg7ishp68ov+s8COzqZV3qiB5rZXDPbbGYVZnb3Cdp93MzczE74fCKpoKGplVff2cvcc0aqW0gywkkLgbtvdve9ncyr6exxZpYLPAhcB0wDbjWzDxx+aWbFwJeAFV0NLRKmlzbV0tIW5bpz1S0kmeGkhcDMfmJm53Yyr4+Z3WFmHQ1QMxuocPet7t4CPAnM76Ddt4md0bSpG7lFQvPc29UMKy5k1tiBYUcR6RFd2UbwIPC/g2KwHqgDioBJQD/gYeDxDh43GtgVd78S+FB8AzObCZS4+x/M7GudBTCzhcBCgLFjx3YhskhiNLZEeOWdWm6eVUJOjrqFJDOctBC4+zrgluDgsVJgJHAU2Ojum0/1hc0sB/gBcHsXMiwmNkwmpaWlfqqvKXK6/rS5jqZWdQtJZunKcQSLgeeAF9z9lW48dxVQEnd/TDDtmGLgHOCVYIPbCGCJmc1z99XdeB2RpFm6vppBfQqYPX5Q2FFEekxX9hr6GTAdWGpmL5rZv5jZ9C48bhUwycwmmFkBsABYcmymux9y9yHuPt7dxwPLARUBSVlNrW28tLGGOWcPJy9Xh9BI5ujKXkMr3P1ed78UuAXYCXzFzN4ws4fN7JZOHhcBFgFlwEbgKXcvN7P7zGxeD/4NIknx5y17OdLSxtxzRoYdRaRHdWukbXffBzwBPGGx/pyvAWeeoP1SYGm7afd00vaK7mQRSbbn1u+hf698Lj5jcNhRRHpUtwpBPHd3M1vk7tqNRzJeSyTKsg01XDttBPnqFpIM05WNxW91NgvQsEySFV57dy8NTRGu195CkoG6skYwHJgDHGg33YDXejyRSAp67u1q+hbmccmkIWFHEelxXSkEvwf6BscTvI+ZvdLTgURSTXOkjT+WV3P11GEU5uWGHUekx3XlgLI7TzDv73o2jkjqefWdvRw62sr8GaPDjiKSENrqJXISS97czcDe+eoWkoylQiByAo0tEV7YUMP1547U3kKSsfTJFjmBZRtqONraxrzpo8KOIpIwKgQiJ7Bk3W5G9i/iAp1bSDKYCoFIJw42tvDqljpunD5Kp5yWjKZCINKJ59ZX09rm6haSjKdCINKJZ9+oYuKQPpw9ql/YUUQSSoVApAM79zWyYtt+PjZztAaol4ynQiDSgd+srcQMPjZzTNhRRBJOhUCknWjUeXpNJZecOYRRA3qFHUck4VQIRNpZvm0fVQePctMsrQ1IdlAhEGnn6dWVFBfmMedsnXJasoMKgUichqZWlq7fww3TR1GUrzONSnZQIRCJs/TtPTS1RtUtJFlFhUAkzlOrK5k4tA8zxw4IO4pI0qgQiAQ27qlnzY4DLLigRMcOSFZRIRAJPLZ8BwV5Odw8qyTsKCJJpUIgQmwj8bNvVHHjeaMY2Kcg7DgiSaVCIELsvEJHWtq47aJxYUcRSToVAsl67s6jy3dw7uj+TB/TP+w4IkmnQiBZb+W2/bxTc5jbLhynjcSSlVQIJOs9tmIn/YryuFHjDkiWUiGQrFZ18ChL397DzaUl9CrQkcSSnVQIJKs9/JdtANxxyYSQk4iER4VAstahxlaeWLmTedNHMVqnm5YsltBCYGZzzWyzmVWY2d0dzP+ymW0ws7fM7EUz0757kjSPrdhBY0sbCy+bGHYUkVAlrBCYWS7wIHAdMA241cymtWv2BlDq7ucBTwPfS1QekXhNrW38/K/buXzyUKaO1JjEkt0SuUYwG6hw963u3gI8CcyPb+DuL7t7Y3B3OaBTPkpSPPNGFXsPN/MPWhsQSWghGA3sirtfGUzrzJ3Acx3NMLOFZrbazFbX1dX1YETJRpG2KItf3cq5o/tz0RmDw44jErqU2FhsZp8CSoH7O5rv7ovdvdTdS4cOHZrccJJxnnmjim17j3DXR87QAWQiQF4Cn7sKiD+N45hg2vuY2dXAN4DL3b05gXlEaIlEeeDFLZw7ur+GohQJJHKNYBUwycwmmFkBsABYEt/AzM4HfgzMc/faBGYRAeBXq3dReeAoX7l2stYGRAIJKwTuHgEWAWXARuApdy83s/vMbF7Q7H6gL/BrM1tnZks6eTqR09bU2sZ/vrSFC8YP5PLJ6mIUOSaRXUO4+1Jgabtp98TdvjqRry8S79HXd1BT38wDC87X2oBInJTYWCySaIcaW/nRn97l0klDuHCi9hQSiadCIFnh/73wDgcbW/iXuWeFHUUk5agQSMbbuKeeR17fzt99aCznjNbAMyLtqRBIRnN3vrWknP698vnqtVPCjiOSklQIJKP97q09rNy2n6/NOYsBvTUovUhHVAgkY9U3tfKvf9jIOaP78YkLSk7+AJEsldDdR0XCdN/vNlB3uJmHbptFbo52FxXpjNYIJCMt21DD02sq+eIVZzCjZEDYcURSmgqBZJx9h5v5+v+8xdmj+vGPV04KO45IylPXkGQUd+cbz6yn/miEx/9+BgV5+q0jcjL6XyIZ5ZHXd/DH8mq+fO1kpowoDjuOSFpQIZCMsXLbfr79+w1cPXUYCy/VyGMiXaVCIBlhz6GjfPHxNYwd1JsffGIGOdpLSKTLtI1A0l5TaxtfeGwtR1vaeOJzF9KvKD/sSCJpRYVA0lprW5S7Hl/Lm5UH+dEnZzFpuLYLiHSXuoYkbUWjzld//SYvbqrlvvnnMPccDT0pcipUCCQtuTv3/q6c367bzdfmTOG2C8eFHUkkbalrSNJOW9T55rPreWLlTv7hsol88Yozwo4kktZUCCStNLW28aUn36CsvIa7PnIGX712ioadFDlNKgSSNg42trDw0TWs3Lafb904jc9+eELYkUQyggqBpIV1uw5y1+NrqW1o4oEFM5g/Y3TYkUQyhgqBpDR355HXd/CdP2xgWHERT3/+YqbrbKIiPUqFQFLWrv2NfOPZ9bz6Th1XnjWMH9wyXaOMiSSACoGknLao84vXtvPvZZsxg3tvnManLxqv00aIJIgKgaQMd+f5DTXcX7aZitrDfGTKUL7zt+cyekCvsKOJZDQVAgldNOr86Z06fvjSFt7YeZCJQ/vw0KdmMufsEdo1VCQJVAgkNI0tEZ59Yzc/+8tW3q07wsj+RXz34+fy8ZljyMvVQe8iyaJCIEkVjTrLt+3jf9ZW8dzbezjS0sY5o/vxwIIZXH/uSPJVAESSToVAEu5Ic4TX3t3HixtreGFjLXsPN9O3MI8bzhvFTaVjKB03UF1AIiFSIZAed7CxhVXbD7Bq+35WbNvP+qpDtEWd4sI8Lp8ylGvPHsE1U4fTqyA37KgiggqBnIbGlgg79zdSUXuYTXsa2FRdz8Y9DVQdPApAQW4OM0oG8PnLJ3LRxCHMnjBIg8mLpKCEFgIzmws8AOQCP3X3f2s3vxB4BJgF7AM+4e7bE5lJTs7dOdwcoa6hmdqGZuqCS21DMzX1Tezc38iOfY3sPdx8/DG5OcYZQ/swa9xAPnnhWGaNHcj0kgEU5etXv0iqS1ghMLNc4EHgGqASWGVmS9x9Q1yzO4ED7n6mmS0Avgt8IlGZ0o27E4k6bcElcvw6GrtuC+a5H7/f0halqbWNptY2miOx282tUZoiwXVrG02RNppaozQ0tdLQFKG+qZX6oxEamlqpb4pQf7SVSNQ/kCc/1xhWXETJoF5cedZQxg3uQ8mg3kwc0odJw/tSmKcvfZF0lMg1gtlAhbtvBTCzJ4H5QHwhmA/cG9x+GvhPMzN3/+C30Gl6atUufvzquwB48M+xF3F3HDj2qo7j/t79E7Y5Pj+Yenz+e485Nj/+/rHX/0AbnGgUItEoHXwX94jcHKMoL4fionz69cqjuCifIX0LmDi0D8VFefQryqd/r3yG9StkaN+i4LqQ/r3ydXSvSAZKZCEYDeyKu18JfKizNu4eMbNDwGBgb3wjM1sILAQYO3bsKYUZ2KeAs0b0g+B7zGLPe+wuZu9NOzYfg2Mt3pvfbpodb/2+NrGpdnwa8c/dwfzj08zIzTHycmLXuWbk5h67n3N8el6OkRPXLi8nh9wcKMjLoSgvl8L8XIrycyjMi10X5edSlJ9LYV6OdtEUkfdJi43F7r4YWAxQWlp6Sr+Tr5k2nGumDe/RXCIimSCRPw2rgJK4+2OCaR22MbM8oD+xjcYiIpIkiSwEq4BJZjbBzAqABcCSdm2WAJ8Jbt8EvJSI7QMiItK5hHUNBX3+i4AyYruPPuzu5WZ2H7Da3ZcAPwMeNbMKYD+xYiEiIkmU0G0E7r4UWNpu2j1xt5uAmxOZQURETky7j4iIZDkVAhGRLKdCICKS5VQIRESynKXb3ppmVgfsOMWHD6HdUcspJFWzKVf3KFf3pWq2TMs1zt2HdjQj7QrB6TCz1e5eGnaOjqRqNuXqHuXqvlTNlk251DUkIpLlVAhERLJcthWCxWEHOIFUzaZc3aNc3Zeq2bImV1ZtIxARkQ/KtjUCERFpR4VARCTLZVwhMLObzazczKJmVtpu3tfNrMLMNpvZnE4eP8HMVgTtfhWcQrunM/7KzNYFl+1mtq6TdtvN7O2g3eqeztHJa95rZlVx+a7vpN3cYDlWmNndSch1v5ltMrO3zOwZMxvQSbukLLOT/f1mVhi8zxXB52l8orLEvWaJmb1sZhuC/wNf6qDNFWZ2KO79vaej50pQvhO+Nxbzw2CZvWVmM5OQaUrcslhnZvVm9k/t2iRlmZnZw2ZWa2br46YNMrNlZrYluB7YyWM/E7TZYmaf6ajNCbl7Rl2AqcAU4BWgNG76NOBNoBCYALwL5Hbw+KeABcHth4AvJDjv94F7Opm3HRiS5OV3L/DVk7TJDZbfRKAgWK7TEpzrWiAvuP1d4LthLbOu/P3AF4GHgtsLgF8l4b0bCcwMbhcD73SQ6wrg98n8THX1vQGuB54jNnrrhcCKJOfLBaqJHXiV9GUGXAbMBNbHTfsecHdw++6OPvfAIGBrcD0wuD2wO6+dcWsE7r7R3Td3MGs+8KS7N7v7NqACmB3fwGKDCl8JPB1M+iXwN4nKGrzeLcATiXqNBJkNVLj7VndvAZ4ktnwTxt2fd/dIcHc5sRHvwtKVv38+sc8PxD5PV9mxQasTxN33uPva4HYDsJHYuODpYj7wiMcsBwaY2cgkvv5VwLvufqpnLjgt7v4qsXFZ4sV/jjr7PpoDLHP3/e5+AFgGzO3Oa2dcITiB0cCuuPuVfPA/yWDgYNwXTkdtetKlQI27b+lkvgPPm9kaM1uYwBztLQpWzR/uZFW0K8syke4g9suxI8lYZl35+4+3CT5Ph4h9vpIi6Io6H1jRweyLzOxNM3vOzM5OViZO/t6E/blaQOc/ysJaZsPdfU9wuxroaOD1015uaTF4fXtm9gIwooNZ33D33yY7T0e6mPFWTrw2cIm7V5nZMGCZmW0KfjUkLBvwI+DbxP7TfptY19Udp/uap5vr2DIzs28AEeDxTp4mIcssnZhZX+A3wD+5e3272WuJdX0cDrb/PAtMSlK0lH1vgm2B84CvdzA7zGV2nLu7mSVkf/+0LATufvUpPKwKKIm7PyaYFm8fsdXRvOBXXEdteiSjmeUBHwNmneA5qoLrWjN7hliXxGn/x+nq8jOznwC/72BWV5Zlj+cys9uBG4CrPOgc7eA5ErLM2unK33+sTWXwXvcn9vlKKDPLJ1YEHnf3/2k/P74wuPtSM/svMxvi7gk/uVoX3puEfK666DpgrbvXtJ8R5jIDasxspLvvCbrJajtoU0VsO8YxY4htI+2ybOoaWgIsCPbmmECsoq+MbxB8ubwM3BRM+gyQqDWMq4FN7l7Z0Uwz62NmxcduE9tYur6jtj2pXZ/s33bymquASRbbw6qA2Cr1kgTnmgv8MzDP3Rs7aZOsZdaVv38Jsc8PxD5PL3VWvHpKsA3iZ8BGd/9BJ21GHNtWYWaziX0HJKNAdeW9WQJ8Oth76ELgUFy3SKJ1unYe1jILxH+OOvs+KgOuNbOBQVfutcG0rkv0lvBkX4h9eVUCzUANUBY37xvE9vbYDFwXN30pMCq4PZFYgagAfg0UJijnL4DPt5s2Clgal+PN4FJOrHskGcvvUeBt4K3gQziyfbbg/vXE9kp5NxnZgvdjF7AuuDzUPlcyl1lHfz9wH7FCBVAUfH4qgs/TxCQso0uIdem9Fbecrgc+f+yzBiwKls2bxDa6X5ykz1WH7027bAY8GCzTt4nb6y/B2foQ+2LvHzct6cuMWCHaA7QG32F3Etuu9CKwBXgBGBS0LQV+GvfYO4LPWgXw2e6+tk4xISKS5bKpa0hERDqgQiAikuVUCEREspwKgYhIllMhEBHJcioEIiJZToVARCTLqRCInCYz+3zcueq3mdnLYWcS6Q4dUCbSQ4Lz/LwEfM/dfxd2HpGu0hqBSM95gNg5hVQEJK2k5dlHRVJNcGbUccTOSyOSVtQ1JHKazGwWsdGjLvXYCFEiaUVdQyKnbxGx8WJfDjYY/zTsQCLdoTUCEZEspzUCEZEsp0IgIpLlVAhERLKcCoGISJZTIRARyXIqBCIiWU6FQEQky/1/3rKaLusSxw8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "zs = np.arange(-10., 10., 0.1)\n",
    "gs = [1/(1+np.exp(-z)) for z in zs]\n",
    "plt.plot(zs, gs)\n",
    "plt.xlabel('z')\n",
    "plt.ylabel('1/(1+e^-z)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 04-4 분류용 데이터셋을 준비합니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_breast_cancer\n",
    "cancer = load_breast_cancer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(569, 30) (569,)\n"
     ]
    }
   ],
   "source": [
    "print(cancer.data.shape, cancer.target.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.799e+01, 1.038e+01, 1.228e+02, 1.001e+03, 1.184e-01, 2.776e-01,\n",
       "        3.001e-01, 1.471e-01, 2.419e-01, 7.871e-02, 1.095e+00, 9.053e-01,\n",
       "        8.589e+00, 1.534e+02, 6.399e-03, 4.904e-02, 5.373e-02, 1.587e-02,\n",
       "        3.003e-02, 6.193e-03, 2.538e+01, 1.733e+01, 1.846e+02, 2.019e+03,\n",
       "        1.622e-01, 6.656e-01, 7.119e-01, 2.654e-01, 4.601e-01, 1.189e-01],\n",
       "       [2.057e+01, 1.777e+01, 1.329e+02, 1.326e+03, 8.474e-02, 7.864e-02,\n",
       "        8.690e-02, 7.017e-02, 1.812e-01, 5.667e-02, 5.435e-01, 7.339e-01,\n",
       "        3.398e+00, 7.408e+01, 5.225e-03, 1.308e-02, 1.860e-02, 1.340e-02,\n",
       "        1.389e-02, 3.532e-03, 2.499e+01, 2.341e+01, 1.588e+02, 1.956e+03,\n",
       "        1.238e-01, 1.866e-01, 2.416e-01, 1.860e-01, 2.750e-01, 8.902e-02],\n",
       "       [1.969e+01, 2.125e+01, 1.300e+02, 1.203e+03, 1.096e-01, 1.599e-01,\n",
       "        1.974e-01, 1.279e-01, 2.069e-01, 5.999e-02, 7.456e-01, 7.869e-01,\n",
       "        4.585e+00, 9.403e+01, 6.150e-03, 4.006e-02, 3.832e-02, 2.058e-02,\n",
       "        2.250e-02, 4.571e-03, 2.357e+01, 2.553e+01, 1.525e+02, 1.709e+03,\n",
       "        1.444e-01, 4.245e-01, 4.504e-01, 2.430e-01, 3.613e-01, 8.758e-02]])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cancer.data[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.boxplot(cancer.data)\n",
    "plt.xlabel('feature')\n",
    "plt.ylabel('value')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['mean area', 'area error', 'worst area'], dtype='<U23')"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cancer.feature_names[[3,13,23]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1]), array([212, 357]))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(cancer.target, return_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = cancer.data\n",
    "y = cancer.target"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 04-5 로지스틱 회귀로 모델을 만들어봅니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(x, y, stratify=y, \n",
    "                                                    test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(455, 30) (114, 30)\n"
     ]
    }
   ],
   "source": [
    "print(x_train.shape, x_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1]), array([170, 285]))"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(y_train, return_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "class LogisticNeuron:\n",
    "    \n",
    "    def __init__(self):\n",
    "        self.w = None\n",
    "        self.b = None\n",
    "\n",
    "    def forpass(self, x):\n",
    "        z = np.sum(x * self.w) + self.b  # 직선 방정식을 계산합니다\n",
    "        return z\n",
    "\n",
    "    def backprop(self, x, err):\n",
    "        w_grad = x * err    # 가중치에 대한 그래디언트를 계산합니다\n",
    "        b_grad = 1 * err    # 절편에 대한 그래디언트를 계산합니다\n",
    "        return w_grad, b_grad\n",
    "\n",
    "    def activation(self, z):\n",
    "        z = np.clip(z, -100, None) # 안전한 np.exp() 계산을 위해\n",
    "        a = 1 / (1 + np.exp(-z))  # 시그모이드 계산\n",
    "        return a\n",
    "        \n",
    "    def fit(self, x, y, epochs=100):\n",
    "        self.w = np.ones(x.shape[1])      # 가중치를 초기화합니다.\n",
    "        self.b = 0                        # 절편을 초기화합니다.\n",
    "        for i in range(epochs):           # epochs만큼 반복합니다\n",
    "            for x_i, y_i in zip(x, y):    # 모든 샘플에 대해 반복합니다\n",
    "                z = self.forpass(x_i)     # 정방향 계산\n",
    "                a = self.activation(z)    # 활성화 함수 적용\n",
    "                err = -(y_i - a)          # 오차 계산\n",
    "                w_grad, b_grad = self.backprop(x_i, err) # 역방향 계산\n",
    "                self.w -= w_grad          # 가중치 업데이트\n",
    "                self.b -= b_grad          # 절편 업데이트\n",
    "    \n",
    "    def predict(self, x):\n",
    "        z = [self.forpass(x_i) for x_i in x]    # 정방향 계산\n",
    "        a = self.activation(np.array(z))        # 활성화 함수 적용\n",
    "        return a > 0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = np.array([1,2,3])\n",
    "b = np.array([3,4,5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4, 6, 8])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a + b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 3,  8, 15])"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a * b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "26"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sum(a * b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0.],\n",
       "       [0., 0., 0.]])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.zeros((2, 3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[7, 7, 7],\n",
       "       [7, 7, 7]])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.full((2,3), 7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., 1., 1., 1.],\n",
       "       [0., 0., 0., 1., 1., 1.]])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.c_[np.zeros((2,3)), np.ones((2,3))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "neuron = LogisticNeuron()\n",
    "neuron.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8245614035087719"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(neuron.predict(x_test) == y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 04-6 단일층 신경망을 만들어 봅니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "class SingleLayer:\n",
    "    \n",
    "    def __init__(self):\n",
    "        self.w = None\n",
    "        self.b = None\n",
    "        self.losses = []\n",
    "\n",
    "    def forpass(self, x):\n",
    "        z = np.sum(x * self.w) + self.b  # 직선 방정식을 계산합니다\n",
    "        return z\n",
    "\n",
    "    def backprop(self, x, err):\n",
    "        w_grad = x * err    # 가중치에 대한 그래디언트를 계산합니다\n",
    "        b_grad = 1 * err    # 절편에 대한 그래디언트를 계산합니다\n",
    "        return w_grad, b_grad\n",
    "\n",
    "    def activation(self, z):\n",
    "        z = np.clip(z, -100, None) # 안전한 np.exp() 계산을 위해\n",
    "        a = 1 / (1 + np.exp(-z))  # 시그모이드 계산\n",
    "        return a\n",
    "        \n",
    "    def fit(self, x, y, epochs=100):\n",
    "        self.w = np.ones(x.shape[1])               # 가중치를 초기화합니다.\n",
    "        self.b = 0                                 # 절편을 초기화합니다.\n",
    "        for i in range(epochs):                    # epochs만큼 반복합니다\n",
    "            loss = 0\n",
    "            # 인덱스를 섞습니다\n",
    "            indexes = np.random.permutation(np.arange(len(x)))\n",
    "            for i in indexes:                      # 모든 샘플에 대해 반복합니다\n",
    "                z = self.forpass(x[i])             # 정방향 계산\n",
    "                a = self.activation(z)             # 활성화 함수 적용\n",
    "                err = -(y[i] - a)                  # 오차 계산\n",
    "                w_grad, b_grad = self.backprop(x[i], err) # 역방향 계산\n",
    "                self.w -= w_grad                   # 가중치 업데이트\n",
    "                self.b -= b_grad                   # 절편 업데이트\n",
    "                # 안전한 로그 계산을 위해 클리핑한 후 손실을 누적합니다\n",
    "                a = np.clip(a, 1e-10, 1-1e-10)\n",
    "                loss += -(y[i]*np.log(a)+(1-y[i])*np.log(1-a))\n",
    "            # 에포크마다 평균 손실을 저장합니다\n",
    "            self.losses.append(loss/len(y))\n",
    "    \n",
    "    def predict(self, x):\n",
    "        z = [self.forpass(x_i) for x_i in x]     # 정방향 계산\n",
    "        return np.array(z) > 0                   # 스텝 함수 적용\n",
    "    \n",
    "    def score(self, x, y):\n",
    "        return np.mean(self.predict(x) == y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9298245614035088"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer = SingleLayer()\n",
    "layer.fit(x_train, y_train)\n",
    "layer.score(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(layer.losses)\n",
    "plt.xlabel('epoch')\n",
    "plt.ylabel('loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 04-7 사이킷런의 경사 하강법을 사용해 봅니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import SGDClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8333333333333334"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sgd = SGDClassifier(loss='log', max_iter=100, tol=1e-3, random_state=42)\n",
    "sgd.fit(x_train, y_train)\n",
    "sgd.score(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 0, 0, 0, 0, 1, 0, 0, 0])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sgd.predict(x_test[0:10])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}