Expe 2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "#如果第二层多几个节点，会完全拟合数据吗？\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "sampleNo = 40;\n",
    "mu = 0\n",
    "sigma = 3\n",
    "x = np.random.normal(mu, sigma, sampleNo)\n",
    "mu = 0\n",
    "sigma = 3\n",
    "y = np.random.normal(mu, sigma, sampleNo)\n",
    "# group\n",
    "s = np.vstack((x,y)).T\n",
    "if s.shape[1] == 2:\n",
    "    s = np.insert(s, 2, values=0, axis=1) # 增加一列标记值\n",
    "s0 = np.array([])\n",
    "s1 = np.array([])\n",
    "for i in range(sampleNo):\n",
    "    if(abs(s[i,0])+abs(s[i,1])<=5):\n",
    "        s0 = np.append(s0, np.array([s[i,0],s[i,1]]), axis = 0)\n",
    "    else:\n",
    "        s1 = np.append(s1, np.array([s[i,0],s[i,1]]), axis = 0)\n",
    "        s[i,2] = 1\n",
    "     \n",
    "s0 = s0.reshape(int(s0.size/2), 2)\n",
    "s1 = s1.reshape(int(s1.size/2), 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAD8CAYAAAB3u9PLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAW7klEQVR4nO3df4zkd33f8ecL26TUpWcIhzG+M+cSGxSoccLKCKEgwNgYgzFUxjF/RE5LdbjFCiGqlENIxnIVFdLQEw4E6wgWtGqAayuHC2dsjmtUFymA96h9PQP+gWvq41x8DmBAOMCRd/+Y75r1emZvd2e+M9+ZeT6k0cz3+/3s7Pur0c57P79TVUiS9JRJByBJ6gYTgiQJMCFIkhomBEkSYEKQJDVMCJIkYEQJIcmNSR5OcmjZuWcm2Zfk3ub5GQN+9sqmzL1JrhxFPJKk9RtVDeETwEUrzu0A9lfVWcD+5vgJkjwTeB/wMuA84H2DEockqV0jSQhVdRvwvRWnLwU+2bz+JPDmPj/6OmBfVX2vqr4P7OPJiUWSNAYntvjep1bVQwBV9VCSZ/cpczrw4LLjw825J0myHdgOcPLJJ7/0hS984YjDlaTZduDAgUeqavOg620mhLVIn3N919Koql3ALoCFhYVaXFxsMy5JmjlJvr3a9TZHGX03yWlNEKcBD/cpcxjYuux4C3CkxZgkSQO0mRD2AEujhq4EPtunzK3AhUme0XQmX9ickySN2aiGnX4K+BvgBUkOJ3k78H7ggiT3Ahc0xyRZSPLnAFX1PeDfArc3j+uac5KkMcs0Ln9tH4IkrV+SA1W1MOi6M5UlSYAJQZLUMCFIkgATgiSpYUKQJAEmBElSw4QgSQJMCJKkhglBkgSYECRJDROCJAkwIUiSGiYESRJgQpAkNUwIkiTAhCBJapgQJEmACUGS1Gg1ISR5QZI7lj1+mOT3V5R5VZJHl5W5ps2YJEn9ndjmm1fV3cC5AElOAL4D3NSn6P+sqje2GYskaXXjbDI6H/hWVX17jL9TkrRG40wIVwCfGnDt5UnuTPL5JC8aY0ySpMZYEkKSpwJvAv5Ln8tfA55XVS8B/hT4ywHvsT3JYpLFo0ePthesJM2pcdUQXg98raq+u/JCVf2wqn7cvL4ZOCnJs/qU21VVC1W1sHnz5vYjlqQ5M66E8DYGNBcleU6SNK/Pa2L62zHFJUlqtDrKCCDJPwQuAN6x7NxVAFV1A3AZ8K+SHAMeA66oqmo7LknSE7WeEKrqJ8Cvrjh3w7LXHwY+3HYckqTVOVNZkgSYEKTRObgbdr4Yrj2l93xw96Qjktal9SYjaS4c3A1/9Xvw88d6x48+2DsGOOfyycUlrYM1BGkU9l/3y2Sw5OeP9c5LU8KEII3Co4fXd17qIBOCNAqbtqzvvNRBJgRpFM6/Bk562hPPnfS03nlpSpgQpFE453K45HrYtBVI7/mS6+1Q1lRxlJE0KudcbgLQVLOGIEkCTAiSpIYJQZIEmBAkSQ0TgiQJMCFIkhomBEkSYEKQJDVMCJIkwISgSXNTGakzWl+6IskDwI+AXwDHqmphxfUAHwIuBn4C/G5Vfa3tuNQBbiojdcq4agivrqpzVyaDxuuBs5rHduCjY4pJk+amMpNl7UwrdKHJ6FLgP1bPl4FTkpw26aA0BjO2qczOffdMOoS1W6qdPfogUL+snZkU5to4EkIBX0hyIMn2PtdPBx5cdny4OfcESbYnWUyyePTo0ZZC1VjN2KYyH9p/76RDWDtrZ+pjHAnhFVX1m/Saht6Z5JUrrqfPz9STTlTtqqqFqlrYvHlzG3Fq3NxUZjIO7m5qBn1Mae1Mo9F6p3JVHWmeH05yE3AecNuyIoeBrcuOtwBH2o5LHbDUcbz/ut4X0aYtvWQwRR3KO/fd84SawbYdewF41/ln8e4Lzp5UWIMtNRUNMqW1M41GqwkhycnAU6rqR83rC4GVddI9wNVJPg28DHi0qh5qMy51yJRvKvPuC85+/It/2469PPD+N0w4ouPo11S0xNrZ3Gu7yehU4EtJ7gS+CuytqluSXJXkqqbMzcD9wH3Ax4B/3XJM0lh1qrN5tSYht/yce60mhKq6v6pe0jxeVFV/1Jy/oapuaF5XVb2zqp5fVf+0qhbbjEkataUv/Hedf1bf653qbB7Ykb/VZKBODDuVptrSF/6a+gwmPfbfjnytovVOZWkkDu6eqs7nfp3Nb3rKl/jgP7iRk/7+73onJzEzewY68tWeVD1phGfnLSws1OKiLUtzY+USF9D7r3aCbd4rv/CX9Btd9Hhn884X9x/uuWkrvPtQW6FKj0tyYMCKEYA1BE2D1SZRTSghbGh00YzNzNbssQ9B3bfeL9JJt9Ov8Hhn84zNzNbsMSGo+9bzRTqBNXoGjS5a8ngTkh266jgTgrpvPV+kE1ijZ80zks+5vNfvsWkrkN6zY//VIfYhqPvWMzKm6+30Uz4zW7PNhKDpsNYv0k1bBozksZ1eOh6bjDRbbKeXNsyEoNliO720YTYZafbYTi9tiDUEzY+OzU+QusYagubDyuUvJrGOkNRx1hA0H9xDWDouE4LmQ9fnJ0gdYELQfJjCdYQ6tdOa5oIJQfNhCucndGqnNc2F1hJCkq1J/jrJN5LcleRdfcq8KsmjSe5oHt3969R0c36CdFytbZCT5DTgtKr6WpKnAweAN1fV15eVeRXwb6rqjet5bzfImTMd3i1t57571r643Rrfb60b70jrNbENcqrqIeCh5vWPknwDOB34+qo/KC3X8eGiH9p/70i/qDe08Y40ImPpQ0iyDfgN4Ct9Lr88yZ1JPp/kRau8x/Yki0kWjx492lKk6hyHi0pj0/rEtCT/CPhvwO9X1Q9XXP4a8Lyq+nGSi4G/BPruNlJVu4Bd0GsyajFkdUkHh4uubNbZtmMvMPpmneNtvCONWmt9CABJTgI+B9xaVf9hDeUfABaq6pHVytmHMEc6vjG9zTqaJsfrQ2hzlFGAjwPfGJQMkjynKUeS85p4/ratmDSFpnC46HLOJdA0abMP4RXA7wCvWTas9OIkVyW5qilzGXAoyZ3A9cAV1WaVRdOn48NFj9es41wCTZM2Rxl9CchxynwY+HBbMWhGdHg5a4eCapa42qk0YuPqdJZGrdVO5bbYqaxpYaezumRincqSpOliQpBa5FwCTRMTgtQi+ww0TUwI0og450DTzoQgjYhzDjTtTAhSS5ZqDNYcNC2chyANYbU5B0tLY496iWypLSYEaQir7V8wiSakUW/Yo/liQpBGaFCNYVyzla2NaBgmBGlElr7sV9YYnK2saWFC0Gh0eN/jcZnUf+aunaRRMSFoeB3f93jJuNvXl2Yptz1b2X2YNSoOO9XwpmTf43F38i59SftfuqaFCUHD6+C+x/PKtZM0DJuMNLxNWwbse7xl/LGsMG/t67N4Txofawga3hj2Pd7obN93X3A2D7z/DY+3qy+9nqYvTmc6a1xaTwhJLkpyd5L7kuzoc/1Xknymuf6VJNvajkkjNoZ9j+d5naB5vneNV6tNRklOAD4CXAAcBm5Psqeqvr6s2NuB71fVryW5AvgA8NttxqUWdHjf4yW2r0ura3ULzSQvB66tqtc1x+8BqKp/t6zMrU2Zv0lyIvD/gM21SmBuoTkfVrb/L5nV9v/l5vne1Z7jbaHZdkK4DLioqv5lc/w7wMuq6uplZQ41ZQ43x99qyjyy4r22A9sBzjjjjJd++9vfbi1udU/XxtePc05D1+5d02vSeyqnz7mVGWgtZaiqXVW1UFULmzdvHklw0kbZrq9Z1HZCOAxsXXa8BTgyqEzTZLQJ+F7LcWnKjLr9v6sjd/rFZd+HxqXthHA7cFaSM5M8FbgC2LOizB7gyub1ZcB/X63/QPNp1M0zG/kPf+e+e9i2Y+8TVjDdtmPvSJNLv7jsM9C4tDrKqKqOJbkauBU4Abixqu5Kch2wWFV7gI8D/ynJffRqBle0GZO0Ua4ZpFnX+kzlqroZuHnFuWuWvf474K1txyF1ddZyV+PS/Gl1lFFbHHaqYSx94Q7zH35bo4yseahNkx5lJM0k/3PXLHJxO82FfhO9tu3Y27lmGUcUaZJsMtLcsVlG88omI0nSmpgQNHdslpH6MyFo7nSpz0DqEhOCJAkwIUiSGiYEjc7B3bDzxXDtKb3ng7snHZGkdXAegkbj4G74q9+Dnz/WO370wd4xdH4nNUk91hA0Gvuv+2UyWPLzx3rnJU0FE4JG49HD6zuv6WFT4NwwIWg0Nm1Z33lNh6WmwEcfBOqXTYEmhZlkQtBonH8NnPS0J5476Wm985peNgXOFROCRuOcy+GS62HTViC950uut0N52tkUOFccZaTROedyE8Cs2bSlaS7qc14zxxqCpMFsCpwrrdQQkvx74BLgZ8C3gH9eVT/oU+4B4EfAL4Bjqy3LKmkClmp8+6/rNRNt2tJLBtYEZ1JbTUb7gPdU1bEkHwDeA/zhgLKvrqpHWopD0rBsCpwbrTQZVdUXqupYc/hlwAZHSeq4cfQh/Avg8wOuFfCFJAeSbF/tTZJsT7KYZPHo0aMjD1Id5+QoqXUbbjJK8kXgOX0uvbeqPtuUeS9wDPjPA97mFVV1JMmzgX1JvllVt/UrWFW7gF3Q20Jzo3FrCrlOkjQWG04IVfXa1a4nuRJ4I3B+Ddi4uaqONM8PJ7kJOA/omxA0Yw7uXntH5WqTo0wI0si00mSU5CJ6nchvqqqfDChzcpKnL70GLgQOtRGPOma9yyEMnBzVZ3y8pA1rqw/hw8DT6TUD3ZHkBoAkz01yc1PmVOBLSe4EvgrsrapbWopHXbLe5RAGToKKfQnSCLU1yujXqmprVZ3bPK5qzh+pqoub1/dX1Uuax4uq6o/aiEUdtN7lEM6/BkifC9XJNXV27rtn0iFIG+JMZY3feldGPedyegPS+ujgmjof2n/v+H6Zo680QiYEjd9GlkPYtHXA+Tme4uLS1BoxE4LGbyMro3Z8TZ2d++5h2469bNuxF+Dx1602H7k0tUYsA0aEdtrCwkItLi5OOgyN23qGqk7Qth17eeD9b2j/F117Cv2b0gLXPmnpMIkkB1ZbM87lrzU9XFPniVyaWiNmk5E0Yu86/6zx/KKON6Np+pgQpBF79wVnj+cXuUudRswmI2ma2YymEbKGIPXj+H7NIWsI0kqurqo5ZQ1BWsnx/ZpTJgRppfWutSTNCBOCtNJ611qSZoQJQVrJ8f2aUyYEaSXH92tOOcpI6sfx/ZpD1hAkSYAJQZLUaC0hJLk2yXeaPZXvSHLxgHIXJbk7yX1JdrQVjwQ4A1laRdt9CDur6k8GXUxyAvAR4ALgMHB7kj1V9fWW49K0GmZPBGcgS6uadJPRecB9VXV/Vf0M+DRw6YRjUlcNu2WkM5ClVbWdEK5OcjDJjUme0ef66cDyHT4ON+eeJMn2JItJFo8ePdpGrOq6Yb/QnYEsrWqohJDki0kO9XlcCnwUeD5wLvAQ8MF+b9HnXN89PatqV1UtVNXC5s2bhwlb02rYL3RnIEurGqoPoapeu5ZyST4GfK7PpcPA1mXHW4Ajw8SkGTbslpHnX/PEPgRwBrK0TJujjE5bdvgW4FCfYrcDZyU5M8lTgSuAPW3FpCk37JISzkCWVtXmKKM/TnIuvSagB4B3ACR5LvDnVXVxVR1LcjVwK3ACcGNV3dViTJpmS1/cGx1ltPQeJgCpr1T1bbLvtIWFhVpcXJx0GJI0VZIcqKqFQdcnPexUktQRJgRJEmBCkCQ1TAiSJMCEIElqmBAkSYAJQZLUMCFIkgATgiSpYUKQJAEmBElSw4QgSQJMCNLsOLgbdr4Yrj2l97zWrUWlRpvLX0sal6X9ppc2/1nabxpc7ltrZg1BmgXD7jctYUKQZsOw+01LmBCk2TBoX+m17jctYUKQZsOw+01LtNSpnOQzwAuaw1OAH1TVuX3KPQD8CPgFcGy1rd0krWIU+01r7rWSEKrqt5deJ/kg8OgqxV9dVY+0EYc0V8653ASgobQ67DRJgMuB17T5eyRJw2u7D+G3gO9W1b0DrhfwhSQHkmxvORZJ0io2XENI8kXgOX0uvbeqPtu8fhvwqVXe5hVVdSTJs4F9Sb5ZVbcN+H3bge0AZ5xxxkbDliQNkKpq542TE4HvAC+tquMOhk5yLfDjqvqT45VdWFioxcXF4YOUpDmS5MBqg3fabDJ6LfDNQckgyclJnr70GrgQONRiPJKkVbSZEK5gRXNRkucmubk5PBX4UpI7ga8Ce6vqlhbjkSStorVRRlX1u33OHQEubl7fD7ykrd8vSVofZypLkgATgtQN7mWgDnA/BGnS3MtAHWENQZo09zJQR5gQpElzLwN1hAlBmjT3MlBHmBCkcVit09i9DNQRdipLbTtep7F7GagjTAhS21brNF760ncvA3WATUZS2+w01pQwIUhts9NYU8KEILXNTmNNCROC1LZzLodLrodNW4H0ni+53j4DdY6dytI42GmsKWANQZIEmBAkSQ0TgiQJMCFIkhomBEkSMGRCSPLWJHcl+fskCyuuvSfJfUnuTvK6AT9/ZpKvJLk3yWeSPHWYeCRJGzdsDeEQ8M+A25afTPLrwBXAi4CLgD9LckKfn/8AsLOqzgK+D7x9yHgkSRs0VEKoqm9U1d19Ll0KfLqqflpV/we4DzhveYEkAV4D/Nfm1CeBNw8TjyRp49qamHY68OVlx4ebc8v9KvCDqjq2SpnHJdkObG8Of5xkeSJ6FvDIUBF306zeF8zuvXlf02dW763ffT1vtR84bkJI8kXgOX0uvbeqPjvox/qcqw2U+eWFql3ArgExLlbVQr9r02xW7wtm9968r+kzq/e2kfs6bkKoqtduIJbDwNZlx1uAIyvKPAKckuTEppbQr4wkaUzaGna6B7giya8kORM4C/jq8gJVVcBfA5c1p64EBtU4JEktG3bY6VuSHAZeDuxNcitAVd0F7Aa+DtwCvLOqftH8zM1Jntu8xR8Cf5DkPnp9Ch/fYCh9m5JmwKzeF8zuvXlf02dW723d95XeP+qSpHnnTGVJEmBCkCQ1pjohDFo6I8m2JI8luaN53DDJONdr2CVBpkGSa5N8Z9lndPGkYxpWkouaz+W+JDsmHc+oJHkgyf9uPqfFScczjCQ3Jnk4yaFl556ZZF+zhM6+JM+YZIwbMeC+1v03NtUJgQFLZzS+VVXnNo+rxhzXsIZdEmRa7Fz2Gd086WCG0XwOHwFeD/w68Lbm85oVr24+p2kfr/8Jen87y+0A9jdL6OxvjqfNJ3jyfcE6/8amOiGssnTGVBtmSRBNzHnAfVV1f1X9DPg0vc9LHVJVtwHfW3H6UnpL58CULqEz4L7WbaoTwnGcmeR/JfkfSX5r0sGMyOnAg8uOV13uYwpcneRgU92dumr6CrP22SxXwBeSHGiWkJk1p1bVQwDN87MnHM8oretvrPMJIckXkxzq81jtv6+HgDOq6jeAPwD+Isk/Hk/Ea7PB+1rXch+Tdpx7/CjwfOBcep/XByca7PCm6rNZp1dU1W/Saw57Z5JXTjogrcm6/8baWtxuZDaydEZV/RT4afP6QJJvAWcDnekQa3FJkM5Y6z0m+RjwuZbDadtUfTbrUVVHmueHk9xEr3msX7/dtPpuktOq6qEkpwEPTzqgUaiq7y69XuvfWOdrCBuRZPNSZ2uSf0Jv6Yz7JxvVSBx3SZBp0fzhLXkLvY70aXY7cFaz6dNT6XX+75lwTENLcnKSpy+9Bi5k+j+rlfbQWzoHZmgJnY38jXW+hrCaJG8B/hTYTG/pjDuq6nXAK4HrkhwDfgFcVVVDd7iMy6D7qqq7kiwtCXKMZUuCTKE/TnIuvWaVB4B3TDac4VTVsSRXA7cCJwA3Nku4TLtTgZuSQO/74i+q6pbJhrRxST4FvAp4VrPszvuA9wO7k7wd+L/AWycX4cYMuK9XrfdvzKUrJEnAjDYZSZLWz4QgSQJMCJKkhglBkgSYECRJDROCJAkwIUiSGv8fz2gYM/8/k3wAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(s0[:,0],s0[:,1],'+')\n",
    "plt.plot(s1[:,0],s1[:,1],'o')\n",
    "plt.axis(\"equal\")\n",
    "plt.axis([-10,10,-10,10])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def sigmoid(z):\n",
    "    return 1 / (1 + np.exp(-z))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\r\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\r\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\r\n",
       "<!-- Generated by graphviz version 2.38.0 (20140413.2041)\r\n",
       " -->\r\n",
       "<!-- Title: %3 Pages: 1 -->\r\n",
       "<svg width=\"310pt\" height=\"226pt\"\r\n",
       " viewBox=\"0.00 0.00 309.78 226.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\r\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 222)\">\r\n",
       "<title>%3</title>\r\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-222 305.778,-222 305.778,4 -4,4\"/>\r\n",
       "<!-- x1 -->\r\n",
       "<g id=\"node1\" class=\"node\"><title>x1</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"79.7782\" cy=\"-200\" rx=\"27\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"79.7782\" y=\"-196.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">x1</text>\r\n",
       "</g>\r\n",
       "<!-- h1 -->\r\n",
       "<g id=\"node2\" class=\"node\"><title>h1</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"185.778\" cy=\"-113\" rx=\"27\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"185.778\" y=\"-109.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">h1</text>\r\n",
       "</g>\r\n",
       "<!-- x1&#45;&gt;h1 -->\r\n",
       "<g id=\"edge1\" class=\"edge\"><title>x1&#45;&gt;h1</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M92.3078,-183.985C101.498,-173.487 114.62,-159.519 127.778,-149 136.55,-141.988 146.847,-135.338 156.311,-129.755\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"158.112,-132.757 165.053,-124.752 154.635,-126.681 158.112,-132.757\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"139.778\" y=\"-152.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w11</text>\r\n",
       "</g>\r\n",
       "<!-- h2 -->\r\n",
       "<g id=\"node3\" class=\"node\"><title>h2</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"113.778\" cy=\"-113\" rx=\"27\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"113.778\" y=\"-109.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">h2</text>\r\n",
       "</g>\r\n",
       "<!-- x1&#45;&gt;h2 -->\r\n",
       "<g id=\"edge2\" class=\"edge\"><title>x1&#45;&gt;h2</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M58.5457,-188.508C42.1146,-178.917 23.7191,-163.983 33.7782,-149 45.5551,-131.458 58.4409,-139.514 77.7782,-131 79.5811,-130.206 81.4287,-129.378 83.2888,-128.534\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"84.9354,-131.629 92.5433,-124.255 81.9977,-125.275 84.9354,-131.629\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"45.7782\" y=\"-152.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w21</text>\r\n",
       "</g>\r\n",
       "<!-- h3 -->\r\n",
       "<g id=\"node4\" class=\"node\"><title>h3</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"41.7782\" cy=\"-113\" rx=\"27\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"41.7782\" y=\"-109.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">h3</text>\r\n",
       "</g>\r\n",
       "<!-- x1&#45;&gt;h3 -->\r\n",
       "<g id=\"edge3\" class=\"edge\"><title>x1&#45;&gt;h3</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M53.9093,-194.696C36.2078,-190.258 13.9875,-181.408 2.77823,-164 -4.36781,-152.902 3.73473,-141.232 14.3699,-131.96\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"16.6732,-134.602 22.3798,-125.675 12.352,-129.095 16.6732,-134.602\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"14.7782\" y=\"-152.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w13</text>\r\n",
       "</g>\r\n",
       "<!-- h4 -->\r\n",
       "<g id=\"node5\" class=\"node\"><title>h4</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"257.778\" cy=\"-113\" rx=\"27\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"257.778\" y=\"-109.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">h4</text>\r\n",
       "</g>\r\n",
       "<!-- x1&#45;&gt;h4 -->\r\n",
       "<g id=\"edge4\" class=\"edge\"><title>x1&#45;&gt;h4</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M105.577,-194.639C130.519,-189.696 168.83,-180.149 198.778,-164 202.842,-161.809 219.567,-147.541 234.178,-134.805\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"236.763,-137.193 241.983,-127.974 232.153,-131.926 236.763,-137.193\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"227.778\" y=\"-152.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w14</text>\r\n",
       "</g>\r\n",
       "<!-- y -->\r\n",
       "<g id=\"node7\" class=\"node\"><title>y</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"148.778\" cy=\"-22\" rx=\"18\" ry=\"18\"/>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"148.778\" cy=\"-22\" rx=\"22\" ry=\"22\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"148.778\" y=\"-18.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">y</text>\r\n",
       "</g>\r\n",
       "<!-- h1&#45;&gt;y -->\r\n",
       "<g id=\"edge9\" class=\"edge\"><title>h1&#45;&gt;y</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M178.819,-95.2598C173.784,-83.1497 166.851,-66.4729 160.914,-52.1915\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"164.069,-50.6629 156.998,-42.7727 157.605,-53.3501 164.069,-50.6629\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"181.778\" y=\"-65.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w31</text>\r\n",
       "</g>\r\n",
       "<!-- h2&#45;&gt;y -->\r\n",
       "<g id=\"edge10\" class=\"edge\"><title>h2&#45;&gt;y</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M120.003,-95.2912C123.661,-85.5549 128.411,-73.0549 132.778,-62 134.015,-58.869 135.324,-55.6039 136.639,-52.3541\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"140.001,-53.381 140.538,-42.7998 133.519,-50.7363 140.001,-53.381\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"144.778\" y=\"-65.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w32</text>\r\n",
       "</g>\r\n",
       "<!-- h3&#45;&gt;y -->\r\n",
       "<g id=\"edge11\" class=\"edge\"><title>h3&#45;&gt;y</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M57.5363,-97.742C68.8636,-87.6007 84.5796,-73.7485 98.7782,-62 106.71,-55.4369 115.519,-48.4646 123.506,-42.2577\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"125.684,-44.9978 131.46,-36.1159 121.406,-39.4572 125.684,-44.9978\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"110.778\" y=\"-65.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w33</text>\r\n",
       "</g>\r\n",
       "<!-- h4&#45;&gt;y -->\r\n",
       "<g id=\"edge12\" class=\"edge\"><title>h4&#45;&gt;y</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M241.562,-98.3189C229.535,-88.2133 212.708,-74.1593 197.778,-62 189.978,-55.6467 181.414,-48.7911 173.659,-42.6262\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"175.645,-39.7344 165.635,-36.2642 171.296,-45.2196 175.645,-39.7344\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"227.778\" y=\"-65.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w34</text>\r\n",
       "</g>\r\n",
       "<!-- x2 -->\r\n",
       "<g id=\"node6\" class=\"node\"><title>x2</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"220.778\" cy=\"-200\" rx=\"27\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"220.778\" y=\"-196.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">x2</text>\r\n",
       "</g>\r\n",
       "<!-- x2&#45;&gt;h1 -->\r\n",
       "<g id=\"edge5\" class=\"edge\"><title>x2&#45;&gt;h1</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M235.694,-184.763C244.434,-174.631 252.545,-160.782 245.778,-149 242.298,-142.941 229.676,-135.197 216.9,-128.533\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"218.206,-125.273 207.699,-123.912 215.065,-131.528 218.206,-125.273\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"259.778\" y=\"-152.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w12</text>\r\n",
       "</g>\r\n",
       "<!-- x2&#45;&gt;h2 -->\r\n",
       "<g id=\"edge6\" class=\"edge\"><title>x2&#45;&gt;h2</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M203.189,-185.852C194.57,-179.349 184.071,-171.34 174.778,-164 162.533,-154.328 149.103,-143.36 137.969,-134.166\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"139.893,-131.215 129.96,-127.531 135.428,-136.606 139.893,-131.215\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"186.778\" y=\"-152.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w22</text>\r\n",
       "</g>\r\n",
       "<!-- x2&#45;&gt;h3 -->\r\n",
       "<g id=\"edge7\" class=\"edge\"><title>x2&#45;&gt;h3</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M193.887,-197.018C164.787,-193.813 117.968,-185.489 83.7782,-164 73.4901,-157.534 64.4542,-147.717 57.4256,-138.489\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"60.2243,-136.385 51.5567,-130.292 54.5329,-140.461 60.2243,-136.385\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"95.7782\" y=\"-152.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w23</text>\r\n",
       "</g>\r\n",
       "<!-- x2&#45;&gt;h4 -->\r\n",
       "<g id=\"edge8\" class=\"edge\"><title>x2&#45;&gt;h4</title>\r\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M243.897,-190.627C255.698,-185.037 268.97,-176.402 275.778,-164 280.165,-156.009 278.484,-146.737 274.765,-138.364\"/>\r\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"277.755,-136.536 269.916,-129.409 271.6,-139.869 277.755,-136.536\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"289.778\" y=\"-152.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">w24</text>\r\n",
       "</g>\r\n",
       "</g>\r\n",
       "</svg>\r\n"
      ],
      "text/plain": [
       "<graphviz.dot.Digraph at 0x1f5427718d0>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from graphviz import Digraph\n",
    "u = Digraph()\n",
    "u.edge(\"x1\", \"h1\", label=\"w11\", color=\"blue\")\n",
    "u.edge(\"x1\", \"h2\", label=\"w21\", color=\"blue\")\n",
    "u.edge(\"x1\", \"h3\", label=\"w13\", color=\"blue\")\n",
    "u.edge(\"x1\", \"h4\", label=\"w14\", color=\"blue\")\n",
    "u.edge(\"x2\", \"h1\", label=\"w12\", color=\"blue\")\n",
    "u.edge(\"x2\", \"h2\", label=\"w22\", color=\"blue\")\n",
    "u.edge(\"x2\", \"h3\", label=\"w23\", color=\"blue\")\n",
    "u.edge(\"x2\", \"h4\", label=\"w24\", color=\"blue\")\n",
    "u.edge(\"h1\", \"y\", label=\"w31\", color=\"blue\")\n",
    "u.edge(\"h2\", \"y\", label=\"w32\", color=\"blue\")\n",
    "u.edge(\"h3\", \"y\", label=\"w33\", color=\"blue\")\n",
    "u.edge(\"h4\", \"y\", label=\"w34\", color=\"blue\")\n",
    "u.node(\"y\", shape=\"doublecircle\")\n",
    "u"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def twolayerNN2(s, sampleNo):\n",
    "    # 随机初始化各个参数\n",
    "    w11 = np.random.rand()\n",
    "    w12 = np.random.rand()\n",
    "    w21 = np.random.rand()\n",
    "    w22 = np.random.rand()\n",
    "    w13 = np.random.rand()\n",
    "    w14 = np.random.rand()\n",
    "    w24 = np.random.rand()\n",
    "    w23 = np.random.rand()\n",
    "    w31 = np.random.rand()\n",
    "    w32 = np.random.rand()\n",
    "    w33 = np.random.rand()\n",
    "    w34 = np.random.rand()\n",
    "    b1 = np.random.rand()\n",
    "    b2 = np.random.rand()\n",
    "    b3 = np.random.rand()\n",
    "    b4 = np.random.rand()\n",
    "    b5 = np.random.rand()\n",
    "    loss = 100000\n",
    "    epoch = 0\n",
    "    MaxEpoch = 1000\n",
    "    lr = 0.1\n",
    "\n",
    "    while True:\n",
    "        if(epoch%20==0):\n",
    "            print(\"epoch:\", epoch)\n",
    "            print(\"w11:%.2f, w12:%.2f, b1:%.2f\"%(w11,w12,b1))\n",
    "            print(\"w21:%.2f, w22:%.2f, b2:%.2f\"%(w21,w22,b2))\n",
    "            print(\"w13:%.2f, w23:%.2f, b3:%.2f\"%(w13,w23,b3))\n",
    "            print(\"w31:%.2f, w32:%.2f, w33:%.2f, b4:%.2f\"%(w31,w32,w33,b4))\n",
    "        lastloss = loss\n",
    "        loss = 0\n",
    "        partialw11 = 0\n",
    "        partialw12 = 0\n",
    "        partialb1 = 0\n",
    "        partialw21 = 0\n",
    "        partialw22 = 0\n",
    "        partialb2 = 0\n",
    "        partialw13 = 0\n",
    "        partialw23 = 0\n",
    "        partialb3 = 0\n",
    "        partialw14 = 0\n",
    "        partialw24 = 0\n",
    "        partialb4 = 0\n",
    "        partialw31 = 0\n",
    "        partialw32 = 0\n",
    "        partialw33 = 0\n",
    "        partialw34 = 0\n",
    "        partialb5 = 0\n",
    "        for k in range(sampleNo):\n",
    "            h1 = sigmoid(w11*s[k,0]+w12*s[k,1]+b1)\n",
    "            h2 = sigmoid(w21*s[k,0]+w22*s[k,1]+b2)\n",
    "            h3 = sigmoid(w13*s[k,0]+w23*s[k,1]+b3)\n",
    "            h4 = sigmoid(w14*s[k,0]+w24*s[k,1]+b4)\n",
    "            yhat = sigmoid(w31*h1+w32*h2+w33*h3+w34*h4+b5)\n",
    "            #print(\"%.1f,\\t%.1f,\\t%.1f,\\t%.1f,\\t%.1f,\\t%.1f\"%(s[k,0], s[k,1], s[k,2], h1, h2, yhat))\n",
    "            y = s[k,2]\n",
    "            loss += (yhat-y)*(yhat-y)\n",
    "\n",
    "            partialw31 += 2*(yhat-y)*yhat*(1-yhat)*h1\n",
    "            partialw32 += 2*(yhat-y)*yhat*(1-yhat)*h2\n",
    "            partialw33 += 2*(yhat-y)*yhat*(1-yhat)*h3\n",
    "            partialw34 += 2*(yhat-y)*yhat*(1-yhat)*h4\n",
    "            partialb5 += 2*(yhat-y)*yhat*(1-yhat)\n",
    "\n",
    "            partialh1 = 2*(yhat-y)*yhat*(1-yhat)*w31\n",
    "            partialh2 = 2*(yhat-y)*yhat*(1-yhat)*w32\n",
    "            partialh3 = 2*(yhat-y)*yhat*(1-yhat)*w33\n",
    "            partialh4 = 2*(yhat-y)*yhat*(1-yhat)*w34\n",
    "\n",
    "            partialw11 += partialh1*h1*(1-h1)*s[k,0]\n",
    "            partialw12 += partialh1*h1*(1-h1)*s[k,1]\n",
    "            partialb1 += partialh1*h1*(1-h1)\n",
    "\n",
    "            partialw21 += partialh2*h2*(1-h2)*s[k,0]\n",
    "            partialw22 += partialh2*h2*(1-h2)*s[k,1]\n",
    "            partialb2 += partialh2*h2*(1-h2)\n",
    "            \n",
    "            partialw13 += partialh3*h3*(1-h3)*s[k,0]\n",
    "            partialw23 += partialh3*h3*(1-h3)*s[k,1]\n",
    "            partialb3 += partialh3*h3*(1-h3)\n",
    "            \n",
    "            partialw14 += partialh4*h4*(1-h4)*s[k,0]\n",
    "            partialw24 += partialh4*h4*(1-h4)*s[k,1]\n",
    "            partialb4 += partialh4*h4*(1-h4)\n",
    "\n",
    "        if(epoch%20==0):\n",
    "            print(\"loss:%.2f\"%loss)\n",
    "        #if(lastloss<loss):\n",
    "        #    break\n",
    "        w11 = w11 - lr * partialw11\n",
    "        w12 = w12 - lr * partialw12\n",
    "        b1 = b1 - lr * partialb1\n",
    "        w21 = w21 - lr * partialw21\n",
    "        w22 = w22 - lr * partialw22\n",
    "        b2 = b2 - lr * partialb2\n",
    "        w13 = w13 - lr * partialw13\n",
    "        w23 = w23 - lr * partialw23\n",
    "        b3 = b3 - lr * partialb3\n",
    "        w14 = w14 - lr * partialw14\n",
    "        w24 = w24 - lr * partialw24\n",
    "        b4 = b4 - lr * partialb4\n",
    "        w31 = w31 - lr * partialw31\n",
    "        w32 = w32 - lr * partialw32\n",
    "        w33 = w33 - lr * partialw33\n",
    "        w34 = w34 - lr * partialw34\n",
    "        b5 = b5 - lr * partialb5\n",
    "        epoch = epoch+1\n",
    "        if(epoch>MaxEpoch):\n",
    "            break\n",
    "        if(loss<0.01):\n",
    "            break\n",
    "    print(\"epoch:\", epoch)\n",
    "    print(\"w11:%.2f, w12:%.2f, b1:%.2f\"%(w11,w12,b1))\n",
    "    print(\"w21:%.2f, w22:%.2f, b2:%.2f\"%(w21,w22,b2))\n",
    "    print(\"w13:%.2f, w23:%.2f, b3:%.2f\"%(w13,w23,b3))\n",
    "    print(\"w14:%.2f, w24:%.2f, b4:%.2f\"%(w14,w24,b4))\n",
    "    print(\"w31:%.2f, w32:%.2f, w33:%.2f,w34:%.2f, b5:%.2f\"%(w31,w32,w33,w34,b5))\n",
    "    print(\"loss:%.2f\"%loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch: 0\n",
      "w11:0.08, w12:0.83, b1:0.69\n",
      "w21:0.63, w22:0.26, b2:0.92\n",
      "w13:0.17, w23:0.07, b3:0.67\n",
      "w31:0.73, w32:0.48, w33:0.42, b4:0.52\n",
      "loss:14.90\n",
      "epoch: 20\n",
      "w11:-0.03, w12:0.91, b1:0.78\n",
      "w21:0.78, w22:0.32, b2:0.95\n",
      "w13:0.05, w23:-0.02, b3:0.64\n",
      "w31:-0.52, w32:-0.50, w33:0.37, b4:0.73\n",
      "loss:8.35\n",
      "epoch: 40\n",
      "w11:-0.21, w12:1.20, b1:1.28\n",
      "w21:0.91, w22:0.25, b2:1.12\n",
      "w13:0.10, w23:0.08, b3:0.66\n",
      "w31:-1.10, w32:-0.58, w33:0.64, b4:1.04\n",
      "loss:7.79\n",
      "epoch: 60\n",
      "w11:-0.23, w12:1.44, b1:1.87\n",
      "w21:0.95, w22:0.08, b2:1.34\n",
      "w13:0.12, w23:0.15, b3:0.63\n",
      "w31:-1.69, w32:-0.63, w33:0.94, b4:1.35\n",
      "loss:7.21\n",
      "epoch: 80\n",
      "w11:-0.22, w12:1.64, b1:2.37\n",
      "w21:0.80, w22:-0.36, b2:1.65\n",
      "w13:0.12, w23:0.19, b3:0.50\n",
      "w31:-2.23, w32:-0.83, w33:1.29, b4:1.64\n",
      "loss:6.54\n",
      "epoch: 100\n",
      "w11:-0.11, w12:1.93, b1:2.66\n",
      "w21:0.30, w22:-1.04, b2:1.97\n",
      "w13:0.06, w23:0.08, b3:0.46\n",
      "w31:-2.72, w32:-1.38, w33:1.66, b4:1.89\n",
      "loss:5.40\n",
      "epoch: 120\n",
      "w11:-0.10, w12:2.25, b1:2.82\n",
      "w21:0.36, w22:-1.26, b2:2.39\n",
      "w13:0.10, w23:0.01, b3:0.63\n",
      "w31:-3.10, w32:-2.07, w33:2.01, b4:2.19\n",
      "loss:4.62\n",
      "epoch: 140\n",
      "w11:-0.09, w12:2.48, b1:2.95\n",
      "w21:0.42, w22:-1.40, b2:2.78\n",
      "w13:0.12, w23:-0.02, b3:0.76\n",
      "w31:-3.41, w32:-2.57, w33:2.32, b4:2.45\n",
      "loss:4.12\n",
      "epoch: 160\n",
      "w11:-0.09, w12:2.65, b1:3.05\n",
      "w21:0.48, w22:-1.53, b2:3.09\n",
      "w13:0.15, w23:-0.04, b3:0.82\n",
      "w31:-3.67, w32:-2.94, w33:2.60, b4:2.68\n",
      "loss:3.78\n",
      "epoch: 180\n",
      "w11:-0.09, w12:2.78, b1:3.14\n",
      "w21:0.54, w22:-1.69, b2:3.35\n",
      "w13:0.21, w23:0.09, b3:0.91\n",
      "w31:-3.79, w32:-3.22, w33:2.81, b4:2.85\n",
      "loss:3.99\n",
      "epoch: 200\n",
      "w11:-0.08, w12:2.89, b1:3.22\n",
      "w21:0.58, w22:-1.78, b2:3.62\n",
      "w13:0.24, w23:0.02, b3:0.95\n",
      "w31:-3.97, w32:-3.49, w33:2.80, b4:3.02\n",
      "loss:3.55\n",
      "epoch: 220\n",
      "w11:-0.08, w12:3.00, b1:3.29\n",
      "w21:0.63, w22:-1.88, b2:3.84\n",
      "w13:0.30, w23:-0.00, b3:0.88\n",
      "w31:-4.10, w32:-3.68, w33:2.89, b4:3.17\n",
      "loss:3.36\n",
      "epoch: 240\n",
      "w11:-0.05, w12:3.09, b1:3.35\n",
      "w21:0.69, w22:-1.97, b2:4.03\n",
      "w13:0.41, w23:-0.06, b3:0.66\n",
      "w31:-4.15, w32:-3.86, w33:2.99, b4:3.28\n",
      "loss:3.13\n",
      "epoch: 260\n",
      "w11:-0.01, w12:3.18, b1:3.40\n",
      "w21:0.75, w22:-2.00, b2:4.21\n",
      "w13:0.63, w23:-0.17, b3:0.20\n",
      "w31:-4.09, w32:-4.02, w33:3.13, b4:3.34\n",
      "loss:2.73\n",
      "epoch: 280\n",
      "w11:-0.01, w12:3.24, b1:3.47\n",
      "w21:0.77, w22:-2.01, b2:4.37\n",
      "w13:0.87, w23:-0.14, b3:-0.31\n",
      "w31:-4.06, w32:-4.16, w33:3.29, b4:3.32\n",
      "loss:2.40\n",
      "epoch: 300\n",
      "w11:-0.03, w12:3.27, b1:3.54\n",
      "w21:0.79, w22:-2.03, b2:4.52\n",
      "w13:1.11, w23:-0.08, b3:-0.77\n",
      "w31:-4.12, w32:-4.29, w33:3.50, b4:3.27\n",
      "loss:2.11\n",
      "epoch: 320\n",
      "w11:-0.03, w12:3.31, b1:3.62\n",
      "w21:0.81, w22:-2.05, b2:4.65\n",
      "w13:1.36, w23:-0.03, b3:-1.09\n",
      "w31:-4.20, w32:-4.46, w33:3.73, b4:3.23\n",
      "loss:1.90\n",
      "epoch: 340\n",
      "w11:-0.02, w12:3.34, b1:3.69\n",
      "w21:0.84, w22:-2.07, b2:4.78\n",
      "w13:1.59, w23:0.01, b3:-1.32\n",
      "w31:-4.30, w32:-4.62, w33:3.94, b4:3.21\n",
      "loss:1.75\n",
      "epoch: 360\n",
      "w11:-0.01, w12:3.38, b1:3.75\n",
      "w21:0.86, w22:-2.10, b2:4.90\n",
      "w13:1.79, w23:0.04, b3:-1.48\n",
      "w31:-4.41, w32:-4.77, w33:4.14, b4:3.21\n",
      "loss:1.63\n",
      "epoch: 380\n",
      "w11:-0.00, w12:3.41, b1:3.81\n",
      "w21:0.88, w22:-2.12, b2:5.00\n",
      "w13:1.96, w23:0.06, b3:-1.60\n",
      "w31:-4.52, w32:-4.90, w33:4.32, b4:3.22\n",
      "loss:1.53\n",
      "epoch: 400\n",
      "w11:0.00, w12:3.45, b1:3.87\n",
      "w21:0.90, w22:-2.15, b2:5.10\n",
      "w13:2.10, w23:0.08, b3:-1.69\n",
      "w31:-4.63, w32:-5.02, w33:4.48, b4:3.24\n",
      "loss:1.46\n",
      "epoch: 420\n",
      "w11:0.01, w12:3.48, b1:3.92\n",
      "w21:0.91, w22:-2.18, b2:5.19\n",
      "w13:2.22, w23:0.10, b3:-1.77\n",
      "w31:-4.74, w32:-5.13, w33:4.62, b4:3.27\n",
      "loss:1.40\n",
      "epoch: 440\n",
      "w11:0.01, w12:3.51, b1:3.96\n",
      "w21:0.92, w22:-2.20, b2:5.27\n",
      "w13:2.31, w23:0.12, b3:-1.83\n",
      "w31:-4.83, w32:-5.23, w33:4.75, b4:3.30\n",
      "loss:1.35\n",
      "epoch: 460\n",
      "w11:0.02, w12:3.54, b1:4.01\n",
      "w21:0.94, w22:-2.23, b2:5.35\n",
      "w13:2.39, w23:0.13, b3:-1.89\n",
      "w31:-4.93, w32:-5.33, w33:4.86, b4:3.33\n",
      "loss:1.30\n",
      "epoch: 480\n",
      "w11:0.02, w12:3.57, b1:4.05\n",
      "w21:0.95, w22:-2.25, b2:5.42\n",
      "w13:2.46, w23:0.14, b3:-1.94\n",
      "w31:-5.01, w32:-5.42, w33:4.97, b4:3.36\n",
      "loss:1.26\n",
      "epoch: 500\n",
      "w11:0.02, w12:3.59, b1:4.09\n",
      "w21:0.96, w22:-2.27, b2:5.49\n",
      "w13:2.52, w23:0.15, b3:-1.98\n",
      "w31:-5.09, w32:-5.50, w33:5.06, b4:3.39\n",
      "loss:1.23\n",
      "epoch: 520\n",
      "w11:0.02, w12:3.62, b1:4.12\n",
      "w21:0.97, w22:-2.30, b2:5.55\n",
      "w13:2.57, w23:0.17, b3:-2.02\n",
      "w31:-5.17, w32:-5.58, w33:5.15, b4:3.42\n",
      "loss:1.20\n",
      "epoch: 540\n",
      "w11:0.02, w12:3.64, b1:4.15\n",
      "w21:0.98, w22:-2.32, b2:5.60\n",
      "w13:2.61, w23:0.18, b3:-2.06\n",
      "w31:-5.24, w32:-5.66, w33:5.24, b4:3.45\n",
      "loss:1.18\n",
      "epoch: 560\n",
      "w11:0.02, w12:3.67, b1:4.18\n",
      "w21:0.99, w22:-2.34, b2:5.66\n",
      "w13:2.64, w23:0.20, b3:-2.09\n",
      "w31:-5.30, w32:-5.73, w33:5.32, b4:3.47\n",
      "loss:1.15\n",
      "epoch: 580\n",
      "w11:0.02, w12:3.70, b1:4.21\n",
      "w21:0.99, w22:-2.37, b2:5.70\n",
      "w13:2.67, w23:0.23, b3:-2.13\n",
      "w31:-5.34, w32:-5.81, w33:5.40, b4:3.50\n",
      "loss:1.15\n",
      "epoch: 600\n",
      "w11:0.02, w12:3.72, b1:4.24\n",
      "w21:1.00, w22:-2.38, b2:5.76\n",
      "w13:2.72, w23:0.23, b3:-2.15\n",
      "w31:-5.39, w32:-5.89, w33:5.47, b4:3.53\n",
      "loss:1.13\n",
      "epoch: 620\n",
      "w11:0.02, w12:3.74, b1:4.27\n",
      "w21:1.00, w22:-2.40, b2:5.81\n",
      "w13:2.75, w23:0.24, b3:-2.18\n",
      "w31:-5.44, w32:-5.96, w33:5.53, b4:3.57\n",
      "loss:1.11\n",
      "epoch: 640\n",
      "w11:0.02, w12:3.76, b1:4.29\n",
      "w21:1.01, w22:-2.41, b2:5.85\n",
      "w13:2.77, w23:0.25, b3:-2.21\n",
      "w31:-5.48, w32:-6.03, w33:5.59, b4:3.60\n",
      "loss:1.10\n",
      "epoch: 660\n",
      "w11:0.02, w12:3.79, b1:4.32\n",
      "w21:1.02, w22:-2.43, b2:5.89\n",
      "w13:2.79, w23:0.25, b3:-2.24\n",
      "w31:-5.53, w32:-6.09, w33:5.65, b4:3.62\n",
      "loss:1.08\n",
      "epoch: 680\n",
      "w11:0.02, w12:3.81, b1:4.34\n",
      "w21:1.02, w22:-2.44, b2:5.93\n",
      "w13:2.81, w23:0.26, b3:-2.27\n",
      "w31:-5.56, w32:-6.16, w33:5.71, b4:3.65\n",
      "loss:1.07\n",
      "epoch: 700\n",
      "w11:0.02, w12:3.83, b1:4.36\n",
      "w21:1.03, w22:-2.46, b2:5.97\n",
      "w13:2.82, w23:0.27, b3:-2.30\n",
      "w31:-5.60, w32:-6.22, w33:5.76, b4:3.68\n",
      "loss:1.06\n",
      "epoch: 720\n",
      "w11:0.02, w12:3.85, b1:4.38\n",
      "w21:1.03, w22:-2.47, b2:6.00\n",
      "w13:2.83, w23:0.28, b3:-2.34\n",
      "w31:-5.63, w32:-6.28, w33:5.81, b4:3.71\n",
      "loss:1.05\n",
      "epoch: 740\n",
      "w11:0.02, w12:3.87, b1:4.40\n",
      "w21:1.04, w22:-2.48, b2:6.04\n",
      "w13:2.84, w23:0.29, b3:-2.37\n",
      "w31:-5.66, w32:-6.34, w33:5.86, b4:3.74\n",
      "loss:1.03\n",
      "epoch: 760\n",
      "w11:0.02, w12:3.89, b1:4.42\n",
      "w21:1.04, w22:-2.50, b2:6.07\n",
      "w13:2.84, w23:0.30, b3:-2.41\n",
      "w31:-5.69, w32:-6.40, w33:5.92, b4:3.77\n",
      "loss:1.02\n",
      "epoch: 780\n",
      "w11:0.02, w12:3.91, b1:4.44\n",
      "w21:1.04, w22:-2.51, b2:6.10\n",
      "w13:2.85, w23:0.31, b3:-2.45\n",
      "w31:-5.71, w32:-6.46, w33:5.97, b4:3.81\n",
      "loss:1.01\n",
      "epoch: 800\n",
      "w11:0.02, w12:3.93, b1:4.46\n",
      "w21:1.05, w22:-2.52, b2:6.13\n",
      "w13:2.85, w23:0.32, b3:-2.49\n",
      "w31:-5.74, w32:-6.52, w33:6.02, b4:3.84\n",
      "loss:1.00\n",
      "epoch: 820\n",
      "w11:0.02, w12:3.95, b1:4.47\n",
      "w21:1.05, w22:-2.53, b2:6.16\n",
      "w13:2.84, w23:0.35, b3:-2.54\n",
      "w31:-5.76, w32:-6.57, w33:6.07, b4:3.88\n",
      "loss:1.03\n",
      "epoch: 840\n",
      "w11:0.02, w12:3.97, b1:4.49\n",
      "w21:1.06, w22:-2.54, b2:6.18\n",
      "w13:2.84, w23:0.37, b3:-2.60\n",
      "w31:-5.78, w32:-6.63, w33:6.12, b4:3.91\n",
      "loss:1.02\n",
      "epoch: 860\n",
      "w11:0.02, w12:3.99, b1:4.51\n",
      "w21:1.06, w22:-2.56, b2:6.21\n",
      "w13:2.83, w23:0.39, b3:-2.66\n",
      "w31:-5.81, w32:-6.68, w33:6.17, b4:3.95\n",
      "loss:1.02\n",
      "epoch: 880\n",
      "w11:0.02, w12:4.01, b1:4.52\n",
      "w21:1.07, w22:-2.57, b2:6.24\n",
      "w13:2.82, w23:0.41, b3:-2.74\n",
      "w31:-5.82, w32:-6.72, w33:6.23, b4:3.99\n",
      "loss:1.01\n",
      "epoch: 900\n",
      "w11:0.01, w12:4.03, b1:4.54\n",
      "w21:1.07, w22:-2.59, b2:6.26\n",
      "w13:2.80, w23:0.43, b3:-2.84\n",
      "w31:-5.84, w32:-6.77, w33:6.28, b4:4.04\n",
      "loss:1.00\n",
      "epoch: 920\n",
      "w11:0.01, w12:4.05, b1:4.55\n",
      "w21:1.08, w22:-2.60, b2:6.29\n",
      "w13:2.79, w23:0.45, b3:-2.95\n",
      "w31:-5.86, w32:-6.81, w33:6.33, b4:4.09\n",
      "loss:0.97\n",
      "epoch: 940\n",
      "w11:0.01, w12:4.06, b1:4.57\n",
      "w21:1.08, w22:-2.62, b2:6.31\n",
      "w13:2.80, w23:0.48, b3:-3.07\n",
      "w31:-5.89, w32:-6.85, w33:6.39, b4:4.14\n",
      "loss:0.96\n",
      "epoch: 960\n",
      "w11:0.01, w12:4.08, b1:4.58\n",
      "w21:1.08, w22:-2.63, b2:6.34\n",
      "w13:2.82, w23:0.50, b3:-3.18\n",
      "w31:-5.91, w32:-6.88, w33:6.44, b4:4.20\n",
      "loss:0.95\n",
      "epoch: 980\n",
      "w11:0.01, w12:4.09, b1:4.60\n",
      "w21:1.09, w22:-2.64, b2:6.37\n",
      "w13:2.86, w23:0.52, b3:-3.29\n",
      "w31:-5.94, w32:-6.91, w33:6.50, b4:4.26\n",
      "loss:0.94\n",
      "epoch: 1000\n",
      "w11:0.00, w12:4.11, b1:4.62\n",
      "w21:1.09, w22:-2.64, b2:6.40\n",
      "w13:2.90, w23:0.53, b3:-3.39\n",
      "w31:-5.97, w32:-6.94, w33:6.55, b4:4.32\n",
      "loss:0.92\n",
      "epoch: 1001\n",
      "w11:0.00, w12:4.11, b1:4.62\n",
      "w21:1.09, w22:-2.63, b2:6.40\n",
      "w13:2.90, w23:0.33, b3:-3.42\n",
      "w13:4.47, w23:1.53, b3:4.30\n",
      "w31:-6.00, w32:-6.93, w33:6.54,w34:-6.65, b5:9.73\n",
      "loss:0.92\n"
     ]
    }
   ],
   "source": [
    "twolayerNN2(s, sampleNo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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": [
    "from pylab import *\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "fig = figure()\n",
    "ax = Axes3D(fig)\n",
    "x1 = np.arange(-4, 4, 0.05)\n",
    "x2 = np.arange(-4, 4, 0.05)\n",
    "X, Y = np.meshgrid(x1, x2)\n",
    "Z = np.zeros(X.shape)\n",
    "width, height = X.shape\n",
    "\n",
    "w11 = 0.00\n",
    "w12 = 4.11\n",
    "b1 = 4.62\n",
    "w21 = 1.09\n",
    "w22 = -2.63\n",
    "b2 = 6.40\n",
    "w13 = 2.90\n",
    "w23 = -0.33\n",
    "b3 = -3.42\n",
    "w14 = 4.47\n",
    "w24 = 1.53\n",
    "b4 = 4.30\n",
    "w31 = -6.00\n",
    "w32 = -6.93\n",
    "w33 = 6.54\n",
    "w34 = -6.65\n",
    "b5 = 9.73\n",
    "for i in range(height):\n",
    "    for j in range(width):\n",
    "        x1 = X[i,j]\n",
    "        x2 = Y[i,j]\n",
    "        h1 = sigmoid(w11*x1+w12*x2+b1)\n",
    "        h2 = sigmoid(w21*x1+w22*x2+b2)\n",
    "        h3 = sigmoid(w13*x1+w23*x2+b3)\n",
    "        h4 = sigmoid(w14*x1+w24*x2+b4)\n",
    "        Z[i,j] = sigmoid(w31*h1+w32*h2+w33*h3+w34*h4+b5)\n",
    "\n",
    "ax.plot_surface(X, Y, Z, cmap='rainbow', rstride=1, cstride=1, alpha=0.9, lw=1)\n",
    "ax.set_xlabel(\"x1\")\n",
    "ax.set_ylabel(\"x2\")\n",
    "ax.set_zlabel(\"y\")\n",
    "#ax.set_zlim(0,40)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From what we can see the data is not fitted completely because our is still not equal to 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}