sentiment_model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rows and columns: (21459, 2)\n",
      "                                                Text  Emotion\n",
      "0                            i didnt feel humiliated  sadness\n",
      "1  i can go from feeling so hopeless to so damned...  sadness\n",
      "2   im grabbing a minute to post i feel greedy wrong    other\n",
      "3  i am ever feeling nostalgic about the fireplac...    other\n",
      "4                               i am feeling grouchy    other\n",
      "Emotion\n",
      "other      8165\n",
      "happy      7029\n",
      "sadness    6265\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "df = pd.read_csv(\"Emotions in text.csv\") \n",
    "\n",
    "def convert_emotion(label):\n",
    "    if label in ['happy']:\n",
    "        return 'happy'\n",
    "    elif label in ['sadness']:\n",
    "        return 'sadness'\n",
    "    else:\n",
    "        return 'other'\n",
    "\n",
    "df['Emotion'] = df['Emotion'].apply(convert_emotion)\n",
    "\n",
    "# Check the dataset\n",
    "print('rows and columns:', df.shape)\n",
    "print(df.head())\n",
    "print(df['Emotion'].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiUAAAGJCAYAAABVW0PjAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABDK0lEQVR4nO3de1hUZfs+/HPYDQjO4CZmQBFIFMEU9zq5yQ2KipWFlWWKhpqGGVJqfDMkyizMvSiVBj6mpZWagYIIqaWoiOEGlbQgLByoFEZMAeF+/+hl/RxRBARnJefnONZxOPd9zT3Xmmel57NmrRmFEEKAiIiIyMTMTN0AEREREcBQQkRERDLBUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCdJ+Fh4dDoVDcl9caOHAgBg4cKD3eu3cvFAoFvv766/vy+hMnToSrq+t9ea26Ki4uxuTJk6HVaqFQKBAcHGzqlurFf+G9J7oVQwnRPYiNjYVCoZA2a2trODk5wdfXFytWrMCVK1fq5XXy8vIQHh6OjIyMelmvPsm5t5p4//33ERsbi+nTp2PDhg0YP378HWtdXV2N/ve+eRs+fPh97Ppf//X3nuhWFqZugOhBEBERATc3N5SVlUGv12Pv3r0IDg7GkiVLsGPHDnTu3FmqnTdvHt58881arZ+Xl4d33nkHrq6u6NKlS42ft3v37lq9Tl1U19unn36KioqKBu/hXqSkpKBPnz6YP39+jeq7dOmC119/vcq4k5NTfbd2V//1957oVgwlRPVgxIgR6NGjh/Q4NDQUKSkpGDVqFJ544gmcOXMGNjY2AAALCwtYWDTsf3r//PMPmjRpAisrqwZ9nbuxtLQ06evXREFBAby8vGpc36pVK7z44osN2FH9+C+890S34sc3RA1k8ODBePvtt/Hbb7/h888/l8Zvd01JUlIS+vXrB3t7e9jZ2cHDwwP/93//B+Df60B69uwJAJg0aZL0cUFsbCyAf68beeSRR5Ceno4BAwagSZMm0nNvvaakUnl5Of7v//4PWq0Wtra2eOKJJ3DhwgWjGldXV0ycOLHKc29e82693e66hqtXr+L111+Hs7MzlEolPDw88NFHH+HWHyxXKBSYMWMGtm/fjkceeQRKpRIdO3ZEQkLC7d/wWxQUFCAwMBAajQbW1tbw9vbG+vXrpfnK62uys7MRHx8v9Z6Tk1Oj9aszceJE2NnZITc3F6NGjYKdnR1atWqFqKgoAMDJkycxePBg2NrawsXFBZs2baqyxq+//opnnnkGzZs3R5MmTdCnTx/Ex8cb9W/q9/7KlSsIDg6Gq6srlEolHBwcMHToUBw7duxe30JqpBhKiBpQ5fUJ1X2MkpmZiVGjRqGkpAQRERFYvHgxnnjiCRw4cAAA4OnpiYiICADA1KlTsWHDBmzYsAEDBgyQ1vj7778xYsQIdOnSBcuWLcOgQYOq7WvBggWIj4/H3LlzMXPmTCQlJcHHxwfXrl2r1f7VpLebCSHwxBNPYOnSpRg+fDiWLFkCDw8PzJ49GyEhIVXqf/zxR7zyyisYO3YsIiMjcf36dfj7++Pvv/+utq9r165h4MCB2LBhA8aNG4dFixZBrVZj4sSJWL58udT7hg0b0LJlS3Tp0kXq/aGHHqp27bKyMvz1119Vtlvfu/LycowYMQLOzs6IjIyEq6srZsyYgdjYWAwfPhw9evTAhx9+iKZNm2LChAnIzs6Wnpufn49HH30UiYmJeOWVV7BgwQJcv34dTzzxBLZt2yab937atGlYs2YN/P39sXr1arzxxhuwsbHBmTNnqn0Pie5IEFGdxcTECAAiLS3tjjVqtVp07dpVejx//nxx8396S5cuFQDEn3/+ecc10tLSBAARExNTZe6xxx4TAER0dPRt5x577DHp8ffffy8AiFatWgmDwSCNb9myRQAQy5cvl8ZcXFxEQEDAXdesrreAgADh4uIiPd6+fbsAIN577z2jujFjxgiFQiHOnz8vjQEQVlZWRmPHjx8XAMTKlSurvNbNli1bJgCIzz//XBorLS0VOp1O2NnZGe27i4uL8PPzq3a9m2sB3HZbuHCh0X4DEO+//740dvnyZWFjYyMUCoX48ssvpfGzZ88KAGL+/PnSWHBwsAAgfvjhB2nsypUrws3NTbi6uory8nIhhOnfe7VaLYKCgmrwzhHVDM+UEDUwOzu7au/Csbe3BwB8++23db4wUalUYtKkSTWunzBhApo2bSo9HjNmDBwdHbFz5846vX5N7dy5E+bm5pg5c6bR+Ouvvw4hBHbt2mU07uPjg7Zt20qPO3fuDJVKhV9//fWur6PVavH8889LY5aWlpg5cyaKi4uxb9++Ou9D7969kZSUVGW7+bUqTZ48Wfqzvb09PDw8YGtri2effVYa9/DwgL29vdE+7dy5E7169UK/fv2kMTs7O0ydOhU5OTk4ffp0rftuiPfe3t4ehw8fRl5eXq37IbodhhKiBlZcXGwUAG713HPPoW/fvpg8eTI0Gg3Gjh2LLVu21CqgtGrVqlYXtbZr187osUKhgLu7e71cT1Gd3377DU5OTlXeD09PT2n+Zm3atKmyRrNmzXD58uW7vk67du1gZmb8V9ydXqc2WrZsCR8fnyqbi4uLUZ21tXWVj4LUajVat25d5ZoitVpttE+//fYbPDw8qrz2vfTfEO99ZGQkTp06BWdnZ/Tq1Qvh4eF3DYxE1WEoIWpAv//+O4qKiuDu7n7HGhsbG+zfvx979uzB+PHjceLECTz33HMYOnQoysvLa/Q6lXf21Kc7fcFbTXuqD+bm5rcdF7dcmClHd+r9v7JPNenz2Wefxa+//oqVK1fCyckJixYtQseOHaucdSGqKYYSoga0YcMGAICvr2+1dWZmZhgyZAiWLFmC06dPY8GCBUhJScH3338P4M4Boa7OnTtn9FgIgfPnzxvdrdGsWTMUFhZWee6t/4+6Nr25uLggLy+vysdZZ8+elebrg4uLC86dO1flbFN9v05DcXFxQVZWVpXxW/uXw3vv6OiIV155Bdu3b0d2djZatGiBBQsW1GktIoYSogaSkpKCd999F25ubhg3btwd6y5dulRlrPKLsEpKSgAAtra2AHDbkFAX//vf/4z+cfr6669x8eJFjBgxQhpr27YtDh06hNLSUmksLi6uyq3Dtelt5MiRKC8vx6pVq4zGly5dCoVCYfT692LkyJHQ6/XYvHmzNHbjxg2sXLkSdnZ2eOyxx+rldRrKyJEjceTIEaSmpkpjV69exSeffAJXV1fpe1VM+d6Xl5ejqKjIaMzBwQFOTk7ScUtUW/zyNKJ6sGvXLpw9exY3btxAfn4+UlJSkJSUBBcXF+zYsQPW1tZ3fG5ERAT2798PPz8/uLi4oKCgAKtXr0br1q2lCx3btm0Le3t7REdHo2nTprC1tUXv3r3h5uZWp36bN2+Ofv36YdKkScjPz8eyZcvg7u6OKVOmSDWTJ0/G119/jeHDh+PZZ5/FL7/8gs8//9zo4sfa9vb4449j0KBBeOutt5CTkwNvb2/s3r0b3377LYKDg6usXVdTp07Fxx9/jIkTJyI9PR2urq74+uuvceDAASxbtqzaa3zu5o8//jD63plKdnZ2GD169D10/f+8+eab+OKLLzBixAjMnDkTzZs3x/r165GdnY1vvvlGulbGlO/9lStX0Lp1a4wZMwbe3t6ws7PDnj17kJaWhsWLF9fL+0CNkClv/SH6r6u8Jbhys7KyElqtVgwdOlQsX77c6NbTSrfeEpycnCyefPJJ4eTkJKysrISTk5N4/vnnxc8//2z0vG+//VZ4eXkJCwsLo9tAH3vsMdGxY8fb9nenW4K/+OILERoaKhwcHISNjY3w8/MTv/32W5XnL168WLRq1UoolUrRt29fcfTo0SprVtfbrbelCvHvra2zZs0STk5OwtLSUrRr104sWrRIVFRUGNUBuO3tpne6VflW+fn5YtKkSaJly5bCyspKdOrU6ba3ztbXLcE372dAQICwtbWt8vw7/W91ux5++eUXMWbMGGFvby+sra1Fr169RFxcXJXnmuq9LykpEbNnzxbe3t6iadOmwtbWVnh7e4vVq1ff7q0jqhGFEDK7uoqIiIgaJV5TQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssAvT6uBiooK5OXloWnTpvX+dd9EREQPMiEErly5Aicnpyo/knkrhpIayMvLg7Ozs6nbICIi+s+6cOECWrduXW0NQ0kNVH4l9YULF6BSqUzcDRER0X+HwWCAs7NzjX7egaGkBio/slGpVAwlREREdVCTyx94oSsRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJ/+8bEXN+MN3UL1IByPvAzdQtERP8ZPFNCREREsmDSUFJeXo63334bbm5usLGxQdu2bfHuu+9CCCHVCCEQFhYGR0dH2NjYwMfHB+fOnTNa59KlSxg3bhxUKhXs7e0RGBiI4uJio5oTJ06gf//+sLa2hrOzMyIjI+/LPhIREVHNmDSUfPjhh1izZg1WrVqFM2fO4MMPP0RkZCRWrlwp1URGRmLFihWIjo7G4cOHYWtrC19fX1y/fl2qGTduHDIzM5GUlIS4uDjs378fU6dOleYNBgOGDRsGFxcXpKenY9GiRQgPD8cnn3xyX/eXiIiI7kwhbj4tcZ+NGjUKGo0G69atk8b8/f1hY2ODzz//HEIIODk54fXXX8cbb7wBACgqKoJGo0FsbCzGjh2LM2fOwMvLC2lpaejRowcAICEhASNHjsTvv/8OJycnrFmzBm+99Rb0ej2srKwAAG+++Sa2b9+Os2fP3rVPg8EAtVqNoqIiqFSqen0PeE3Jg43XlBBRY1ebf0NNeqbk0UcfRXJyMn7++WcAwPHjx/Hjjz9ixIgRAIDs7Gzo9Xr4+PhIz1Gr1ejduzdSU1MBAKmpqbC3t5cCCQD4+PjAzMwMhw8flmoGDBggBRIA8PX1RVZWFi5fvlylr5KSEhgMBqONiIiIGpZJ77558803YTAY0KFDB5ibm6O8vBwLFizAuHHjAAB6vR4AoNFojJ6n0WikOb1eDwcHB6N5CwsLNG/e3KjGzc2tyhqVc82aNTOaW7hwId5555162ksiIiKqCZOeKdmyZQs2btyITZs24dixY1i/fj0++ugjrF+/3pRtITQ0FEVFRdJ24cIFk/ZDRETUGJj0TMns2bPx5ptvYuzYsQCATp064bfffsPChQsREBAArVYLAMjPz4ejo6P0vPz8fHTp0gUAoNVqUVBQYLTujRs3cOnSJen5Wq0W+fn5RjWVjytrbqZUKqFUKutnJ4mIiKhGTHqm5J9//oGZmXEL5ubmqKioAAC4ublBq9UiOTlZmjcYDDh8+DB0Oh0AQKfTobCwEOnp6VJNSkoKKioq0Lt3b6lm//79KCsrk2qSkpLg4eFR5aMbIiIiMg2ThpLHH38cCxYsQHx8PHJycrBt2zYsWbIETz31FABAoVAgODgY7733Hnbs2IGTJ09iwoQJcHJywujRowEAnp6eGD58OKZMmYIjR47gwIEDmDFjBsaOHQsnJycAwAsvvAArKysEBgYiMzMTmzdvxvLlyxESEmKqXSciIqJbmPTjm5UrV+Ltt9/GK6+8goKCAjg5OeHll19GWFiYVDNnzhxcvXoVU6dORWFhIfr164eEhARYW1tLNRs3bsSMGTMwZMgQmJmZwd/fHytWrJDm1Wo1du/ejaCgIHTv3h0tW7ZEWFiY0XeZEBERkWmZ9HtK/iv4PSVUV/yeEiJq7P4z31NCREREVImhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZMGkocTV1RUKhaLKFhQUBAC4fv06goKC0KJFC9jZ2cHf3x/5+flGa+Tm5sLPzw9NmjSBg4MDZs+ejRs3bhjV7N27F926dYNSqYS7uztiY2Pv1y4SERFRDZk0lKSlpeHixYvSlpSUBAB45plnAACzZs3Cd999h6+++gr79u1DXl4enn76aen55eXl8PPzQ2lpKQ4ePIj169cjNjYWYWFhUk12djb8/PwwaNAgZGRkIDg4GJMnT0ZiYuL93VkiIiKqlkIIIUzdRKXg4GDExcXh3LlzMBgMeOihh7Bp0yaMGTMGAHD27Fl4enoiNTUVffr0wa5duzBq1Cjk5eVBo9EAAKKjozF37lz8+eefsLKywty5cxEfH49Tp05JrzN27FgUFhYiISGhRn0ZDAao1WoUFRVBpVLV6z67vhlfr+uRvOR84GfqFoiITKo2/4Za3Kee7qq0tBSff/45QkJCoFAokJ6ejrKyMvj4+Eg1HTp0QJs2baRQkpqaik6dOkmBBAB8fX0xffp0ZGZmomvXrkhNTTVao7ImODj4jr2UlJSgpKREemwwGOpvR4nuEwbeBxsDLz2IZHOh6/bt21FYWIiJEycCAPR6PaysrGBvb29Up9FooNfrpZqbA0nlfOVcdTUGgwHXrl27bS8LFy6EWq2WNmdn53vdPSIiIroL2YSSdevWYcSIEXBycjJ1KwgNDUVRUZG0XbhwwdQtERERPfBk8fHNb7/9hj179mDr1q3SmFarRWlpKQoLC43OluTn50Or1Uo1R44cMVqr8u6cm2tuvWMnPz8fKpUKNjY2t+1HqVRCqVTe834RERFRzcniTElMTAwcHBzg5/f/PiPt3r07LC0tkZycLI1lZWUhNzcXOp0OAKDT6XDy5EkUFBRINUlJSVCpVPDy8pJqbl6jsqZyDSIiIpIHk4eSiooKxMTEICAgABYW/+/EjVqtRmBgIEJCQvD9998jPT0dkyZNgk6nQ58+fQAAw4YNg5eXF8aPH4/jx48jMTER8+bNQ1BQkHSmY9q0afj1118xZ84cnD17FqtXr8aWLVswa9Ysk+wvERER3Z7JP77Zs2cPcnNz8dJLL1WZW7p0KczMzODv74+SkhL4+vpi9erV0ry5uTni4uIwffp06HQ62NraIiAgABEREVKNm5sb4uPjMWvWLCxfvhytW7fG2rVr4evre1/2j4iIiGpGVt9TIlf8nhKqK1Petslj68HGW4Lpv6I2/4aa/OMbIiIiIoChhIiIiGSCoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGSBoYSIiIhkgaGEiIiIZIGhhIiIiGTBwtQNEBHRf4frm/GmboEaWM4HfiZ7bZ4pISIiIllgKCEiIiJZYCghIiIiWWAoISIiIlkweSj5448/8OKLL6JFixawsbFBp06dcPToUWleCIGwsDA4OjrCxsYGPj4+OHfunNEaly5dwrhx46BSqWBvb4/AwEAUFxcb1Zw4cQL9+/eHtbU1nJ2dERkZeV/2j4iIiGrGpKHk8uXL6Nu3LywtLbFr1y6cPn0aixcvRrNmzaSayMhIrFixAtHR0Th8+DBsbW3h6+uL69evSzXjxo1DZmYmkpKSEBcXh/3792Pq1KnSvMFgwLBhw+Di4oL09HQsWrQI4eHh+OSTT+7r/hIREdGdmfSW4A8//BDOzs6IiYmRxtzc3KQ/CyGwbNkyzJs3D08++SQA4H//+x80Gg22b9+OsWPH4syZM0hISEBaWhp69OgBAFi5ciVGjhyJjz76CE5OTti4cSNKS0vx2WefwcrKCh07dkRGRgaWLFliFF6IiIjIdEx6pmTHjh3o0aMHnnnmGTg4OKBr16749NNPpfns7Gzo9Xr4+PhIY2q1Gr1790ZqaioAIDU1Ffb29lIgAQAfHx+YmZnh8OHDUs2AAQNgZWUl1fj6+iIrKwuXL1+u0ldJSQkMBoPRRkRERA3LpKHk119/xZo1a9CuXTskJiZi+vTpmDlzJtavXw8A0Ov1AACNRmP0PI1GI83p9Xo4ODgYzVtYWKB58+ZGNbdb4+bXuNnChQuhVqulzdnZuR72loiIiKpj0lBSUVGBbt264f3330fXrl0xdepUTJkyBdHR0aZsC6GhoSgqKpK2CxcumLQfIiKixsCkocTR0RFeXl5GY56ensjNzQUAaLVaAEB+fr5RTX5+vjSn1WpRUFBgNH/jxg1cunTJqOZ2a9z8GjdTKpVQqVRGGxERETUsk4aSvn37Iisry2js559/houLC4B/L3rVarVITk6W5g0GAw4fPgydTgcA0Ol0KCwsRHp6ulSTkpKCiooK9O7dW6rZv38/ysrKpJqkpCR4eHgY3elDREREpmPSUDJr1iwcOnQI77//Ps6fP49Nmzbhk08+QVBQEABAoVAgODgY7733Hnbs2IGTJ09iwoQJcHJywujRowH8e2Zl+PDhmDJlCo4cOYIDBw5gxowZGDt2LJycnAAAL7zwAqysrBAYGIjMzExs3rwZy5cvR0hIiKl2nYiIiG5h0luCe/bsiW3btiE0NBQRERFwc3PDsmXLMG7cOKlmzpw5uHr1KqZOnYrCwkL069cPCQkJsLa2lmo2btyIGTNmYMiQITAzM4O/vz9WrFghzavVauzevRtBQUHo3r07WrZsibCwMN4OTEREJCMKIYQwdRNyZzAYoFarUVRUVO/Xl/BnwB9spvwJcB5bDzZTHVs8rh589X1s1ebfUJN/zTwRERERwFBCREREMsFQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESyYNJQEh4eDoVCYbR16NBBmr9+/TqCgoLQokUL2NnZwd/fH/n5+UZr5Obmws/PD02aNIGDgwNmz56NGzduGNXs3bsX3bp1g1KphLu7O2JjY+/H7hEREVEtmPxMSceOHXHx4kVp+/HHH6W5WbNm4bvvvsNXX32Fffv2IS8vD08//bQ0X15eDj8/P5SWluLgwYNYv349YmNjERYWJtVkZ2fDz88PgwYNQkZGBoKDgzF58mQkJibe1/0kIiKi6lmYvAELC2i12irjRUVFWLduHTZt2oTBgwcDAGJiYuDp6YlDhw6hT58+2L17N06fPo09e/ZAo9GgS5cuePfddzF37lyEh4fDysoK0dHRcHNzw+LFiwEAnp6e+PHHH7F06VL4+vre130lIiKiOzP5mZJz587ByckJDz/8MMaNG4fc3FwAQHp6OsrKyuDj4yPVdujQAW3atEFqaioAIDU1FZ06dYJGo5FqfH19YTAYkJmZKdXcvEZlTeUat1NSUgKDwWC0ERERUcMyaSjp3bs3YmNjkZCQgDVr1iA7Oxv9+/fHlStXoNfrYWVlBXt7e6PnaDQa6PV6AIBerzcKJJXzlXPV1RgMBly7du22fS1cuBBqtVranJ2d62N3iYiIqBom/fhmxIgR0p87d+6M3r17w8XFBVu2bIGNjY3J+goNDUVISIj02GAwMJgQERE1MJN/fHMze3t7tG/fHufPn4dWq0VpaSkKCwuNavLz86VrULRabZW7cSof361GpVLdMfgolUqoVCqjjYiIiBqWrEJJcXExfvnlFzg6OqJ79+6wtLREcnKyNJ+VlYXc3FzodDoAgE6nw8mTJ1FQUCDVJCUlQaVSwcvLS6q5eY3Kmso1iIiISB5MGkreeOMN7Nu3Dzk5OTh48CCeeuopmJub4/nnn4darUZgYCBCQkLw/fffIz09HZMmTYJOp0OfPn0AAMOGDYOXlxfGjx+P48ePIzExEfPmzUNQUBCUSiUAYNq0afj1118xZ84cnD17FqtXr8aWLVswa9YsU+46ERER3cKk15T8/vvveP755/H333/joYceQr9+/XDo0CE89NBDAIClS5fCzMwM/v7+KCkpga+vL1avXi0939zcHHFxcZg+fTp0Oh1sbW0REBCAiIgIqcbNzQ3x8fGYNWsWli9fjtatW2Pt2rW8HZiIiEhmTBpKvvzyy2rnra2tERUVhaioqDvWuLi4YOfOndWuM3DgQPz000916pGIiIjuD1ldU0JERESNV51CycMPP4y///67ynhhYSEefvjhe26KiIiIGp86hZKcnByUl5dXGS8pKcEff/xxz00RERFR41Ora0p27Ngh/TkxMRFqtVp6XF5ejuTkZLi6utZbc0RERNR41CqUjB49GgCgUCgQEBBgNGdpaQlXV1fph++IiIiIaqNWoaSiogLAv7fZpqWloWXLlg3SFBERETU+dbolODs7u777ICIiokauzt9TkpycjOTkZBQUFEhnUCp99tln99wYERERNS51CiXvvPMOIiIi0KNHDzg6OkKhUNR3X0RERNTI1CmUREdHIzY2FuPHj6/vfoiIiKiRqtP3lJSWluLRRx+t716IiIioEatTKJk8eTI2bdpU370QERFRI1anj2+uX7+OTz75BHv27EHnzp1haWlpNL9kyZJ6aY6IiIgajzqFkhMnTqBLly4AgFOnThnN8aJXIiIiqos6hZLvv/++vvsgIiKiRq5O15QQERER1bc6nSkZNGhQtR/TpKSk1LkhIiIiapzqFEoqryepVFZWhoyMDJw6darKD/URERER1USdQsnSpUtvOx4eHo7i4uJ7aoiIiIgap3q9puTFF1/k794QERFRndRrKElNTYW1tXV9LklERESNRJ0+vnn66aeNHgshcPHiRRw9ehRvv/12vTRGREREjUudQolarTZ6bGZmBg8PD0RERGDYsGH10hgRERE1LnUKJTExMfXdBxERETVydQolldLT03HmzBkAQMeOHdG1a9d6aYqIiIganzpd6FpQUIDBgwejZ8+emDlzJmbOnInu3btjyJAh+PPPP+vUyAcffACFQoHg4GBp7Pr16wgKCkKLFi1gZ2cHf39/5OfnGz0vNzcXfn5+aNKkCRwcHDB79mzcuHHDqGbv3r3o1q0blEol3N3dERsbW6ceiYiIqOHUKZS8+uqruHLlCjIzM3Hp0iVcunQJp06dgsFgwMyZM2u9XlpaGj7++GN07tzZaHzWrFn47rvv8NVXX2Hfvn3Iy8szusi2vLwcfn5+KC0txcGDB7F+/XrExsYiLCxMqsnOzoafnx8GDRqEjIwMBAcHY/LkyUhMTKzLrhMREVEDqVMoSUhIwOrVq+Hp6SmNeXl5ISoqCrt27arVWsXFxRg3bhw+/fRTNGvWTBovKirCunXrsGTJEgwePBjdu3dHTEwMDh48iEOHDgEAdu/ejdOnT+Pzzz9Hly5dMGLECLz77ruIiopCaWkpACA6Ohpubm5YvHgxPD09MWPGDIwZM+aOXwBHREREplGnUFJRUQFLS8sq45aWlqioqKjVWkFBQfDz84OPj4/ReHp6OsrKyozGO3TogDZt2iA1NRXAv9+L0qlTJ2g0GqnG19cXBoMBmZmZUs2ta/v6+kpr3E5JSQkMBoPRRkRERA2rTqFk8ODBeO2115CXlyeN/fHHH5g1axaGDBlS43W+/PJLHDt2DAsXLqwyp9frYWVlBXt7e6NxjUYDvV4v1dwcSCrnK+eqqzEYDLh27dpt+1q4cCHUarW0OTs713ifiIiIqG7qFEpWrVoFg8EAV1dXtG3bFm3btoWbmxsMBgNWrlxZozUuXLiA1157DRs3bpTdt8CGhoaiqKhI2i5cuGDqloiIiB54dbol2NnZGceOHcOePXtw9uxZAICnp2eVj0mqk56ejoKCAnTr1k0aKy8vx/79+7Fq1SokJiaitLQUhYWFRmdL8vPzodVqAQBarRZHjhwxWrfy7pyba269Yyc/Px8qlQo2Nja37U2pVEKpVNZ4X4iIiOje1epMSUpKCry8vGAwGKBQKDB06FC8+uqrePXVV9GzZ0907NgRP/zwQ43WGjJkCE6ePImMjAxp69GjB8aNGyf92dLSEsnJydJzsrKykJubC51OBwDQ6XQ4efIkCgoKpJqkpCSoVCp4eXlJNTevUVlTuQYRERHJQ63OlCxbtgxTpkyBSqWqMqdWq/Hyyy9jyZIl6N+//13Xatq0KR555BGjMVtbW7Ro0UIaDwwMREhICJo3bw6VSoVXX30VOp0Offr0AQAMGzYMXl5eGD9+PCIjI6HX6zFv3jwEBQVJZzqmTZuGVatWYc6cOXjppZeQkpKCLVu2ID4+vja7TkRERA2sVmdKjh8/juHDh99xftiwYUhPT7/npiotXboUo0aNgr+/PwYMGACtVoutW7dK8+bm5oiLi4O5uTl0Oh1efPFFTJgwAREREVKNm5sb4uPjkZSUBG9vbyxevBhr166Fr69vvfVJRERE965WZ0ry8/NveyuwtJiFRZ2/0RX495tXb2ZtbY2oqChERUXd8TkuLi7YuXNntesOHDgQP/30U537IiIiooZXqzMlrVq1wqlTp+44f+LECTg6Ot5zU0RERNT41CqUjBw5Em+//TauX79eZe7atWuYP38+Ro0aVW/NERERUeNRq49v5s2bh61bt6J9+/aYMWMGPDw8AABnz55FVFQUysvL8dZbbzVIo0RERPRgq1Uo0Wg0OHjwIKZPn47Q0FAIIQAACoUCvr6+iIqKqvLtqUREREQ1UesvT6u8sPTy5cs4f/48hBBo166d0Y/pEREREdVWnb7RFQCaNWuGnj171mcvRERE1IjV6bdviIiIiOobQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREcmCSUPJmjVr0LlzZ6hUKqhUKuh0OuzatUuav379OoKCgtCiRQvY2dnB398f+fn5Rmvk5ubCz88PTZo0gYODA2bPno0bN24Y1ezduxfdunWDUqmEu7s7YmNj78fuERERUS2YNJS0bt0aH3zwAdLT03H06FEMHjwYTz75JDIzMwEAs2bNwnfffYevvvoK+/btQ15eHp5++mnp+eXl5fDz80NpaSkOHjyI9evXIzY2FmFhYVJNdnY2/Pz8MGjQIGRkZCA4OBiTJ09GYmLifd9fIiIiujOFEEKYuombNW/eHIsWLcKYMWPw0EMPYdOmTRgzZgwA4OzZs/D09ERqair69OmDXbt2YdSoUcjLy4NGowEAREdHY+7cufjzzz9hZWWFuXPnIj4+HqdOnZJeY+zYsSgsLERCQkKNejIYDFCr1SgqKoJKparX/XV9M75e1yN5yfnAz2SvzWPrwWaqY4vH1YOvvo+t2vwbKptrSsrLy/Hll1/i6tWr0Ol0SE9PR1lZGXx8fKSaDh06oE2bNkhNTQUApKamolOnTlIgAQBfX18YDAbpbEtqaqrRGpU1lWvcTklJCQwGg9FGREREDcvkoeTkyZOws7ODUqnEtGnTsG3bNnh5eUGv18PKygr29vZG9RqNBnq9HgCg1+uNAknlfOVcdTUGgwHXrl27bU8LFy6EWq2WNmdn5/rYVSIiIqqGyUOJh4cHMjIycPjwYUyfPh0BAQE4ffq0SXsKDQ1FUVGRtF24cMGk/RARETUGFqZuwMrKCu7u7gCA7t27Iy0tDcuXL8dzzz2H0tJSFBYWGp0tyc/Ph1arBQBotVocOXLEaL3Ku3Nurrn1jp38/HyoVCrY2NjctielUgmlUlkv+0dEREQ1Y/IzJbeqqKhASUkJunfvDktLSyQnJ0tzWVlZyM3NhU6nAwDodDqcPHkSBQUFUk1SUhJUKhW8vLykmpvXqKypXIOIiIjkwaRnSkJDQzFixAi0adMGV65cwaZNm7B3714kJiZCrVYjMDAQISEhaN68OVQqFV599VXodDr06dMHADBs2DB4eXlh/PjxiIyMhF6vx7x58xAUFCSd6Zg2bRpWrVqFOXPm4KWXXkJKSgq2bNmC+HheQU5ERCQnJg0lBQUFmDBhAi5evAi1Wo3OnTsjMTERQ4cOBQAsXboUZmZm8Pf3R0lJCXx9fbF69Wrp+ebm5oiLi8P06dOh0+lga2uLgIAARERESDVubm6Ij4/HrFmzsHz5crRu3Rpr166Fr6/vfd9fIiIiujOThpJ169ZVO29tbY2oqChERUXdscbFxQU7d+6sdp2BAwfip59+qlOPREREdH/I7poSIiIiapwYSoiIiEgWGEqIiIhIFhhKiIiISBYYSoiIiEgWGEqIiIhIFhhKiIiISBYYSoiIiEgWGEqIiIhIFhhKiIiISBYYSoiIiEgWGEqIiIhIFhhKiIiISBYYSoiIiEgWGEqIiIhIFhhKiIiISBYYSoiIiEgWGEqIiIhIFhhKiIiISBYYSoiIiEgWGEqIiIhIFhhKiIiISBYYSoiIiEgWGEqIiIhIFhhKiIiISBZMGkoWLlyInj17omnTpnBwcMDo0aORlZVlVHP9+nUEBQWhRYsWsLOzg7+/P/Lz841qcnNz4efnhyZNmsDBwQGzZ8/GjRs3jGr27t2Lbt26QalUwt3dHbGxsQ29e0RERFQLJg0l+/btQ1BQEA4dOoSkpCSUlZVh2LBhuHr1qlQza9YsfPfdd/jqq6+wb98+5OXl4emnn5bmy8vL4efnh9LSUhw8eBDr169HbGwswsLCpJrs7Gz4+flh0KBByMjIQHBwMCZPnozExMT7ur9ERER0ZxamfPGEhASjx7GxsXBwcEB6ejoGDBiAoqIirFu3Dps2bcLgwYMBADExMfD09MShQ4fQp08f7N69G6dPn8aePXug0WjQpUsXvPvuu5g7dy7Cw8NhZWWF6OhouLm5YfHixQAAT09P/Pjjj1i6dCl8fX3v+34TERFRVbK6pqSoqAgA0Lx5cwBAeno6ysrK4OPjI9V06NABbdq0QWpqKgAgNTUVnTp1gkajkWp8fX1hMBiQmZkp1dy8RmVN5Rq3KikpgcFgMNqIiIioYckmlFRUVCA4OBh9+/bFI488AgDQ6/WwsrKCvb29Ua1Go4Fer5dqbg4klfOVc9XVGAwGXLt2rUovCxcuhFqtljZnZ+d62UciIiK6M9mEkqCgIJw6dQpffvmlqVtBaGgoioqKpO3ChQumbomIiOiBZ9JrSirNmDEDcXFx2L9/P1q3bi2Na7ValJaWorCw0OhsSX5+PrRarVRz5MgRo/Uq7865uebWO3by8/OhUqlgY2NTpR+lUgmlUlkv+0ZEREQ1Y9IzJUIIzJgxA9u2bUNKSgrc3NyM5rt37w5LS0skJydLY1lZWcjNzYVOpwMA6HQ6nDx5EgUFBVJNUlISVCoVvLy8pJqb16isqVyDiIiITM+kZ0qCgoKwadMmfPvtt2jatKl0DYharYaNjQ3UajUCAwMREhKC5s2bQ6VS4dVXX4VOp0OfPn0AAMOGDYOXlxfGjx+PyMhI6PV6zJs3D0FBQdLZjmnTpmHVqlWYM2cOXnrpJaSkpGDLli2Ij4832b4TERGRMZOeKVmzZg2KioowcOBAODo6StvmzZulmqVLl2LUqFHw9/fHgAEDoNVqsXXrVmne3NwccXFxMDc3h06nw4svvogJEyYgIiJCqnFzc0N8fDySkpLg7e2NxYsXY+3atbwdmIiISEZMeqZECHHXGmtra0RFRSEqKuqONS4uLti5c2e16wwcOBA//fRTrXskIiKi+0M2d98QERFR48ZQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREssBQQkRERLLAUEJERESywFBCREREsmDSULJ//348/vjjcHJygkKhwPbt243mhRAICwuDo6MjbGxs4OPjg3PnzhnVXLp0CePGjYNKpYK9vT0CAwNRXFxsVHPixAn0798f1tbWcHZ2RmRkZEPvGhEREdWSSUPJ1atX4e3tjaioqNvOR0ZGYsWKFYiOjsbhw4dha2sLX19fXL9+XaoZN24cMjMzkZSUhLi4OOzfvx9Tp06V5g0GA4YNGwYXFxekp6dj0aJFCA8PxyeffNLg+0dEREQ1Z2HKFx8xYgRGjBhx2zkhBJYtW4Z58+bhySefBAD873//g0ajwfbt2zF27FicOXMGCQkJSEtLQ48ePQAAK1euxMiRI/HRRx/ByckJGzduRGlpKT777DNYWVmhY8eOyMjIwJIlS4zCCxEREZmWbK8pyc7Ohl6vh4+PjzSmVqvRu3dvpKamAgBSU1Nhb28vBRIA8PHxgZmZGQ4fPizVDBgwAFZWVlKNr68vsrKycPny5du+dklJCQwGg9FGREREDUu2oUSv1wMANBqN0bhGo5Hm9Ho9HBwcjOYtLCzQvHlzo5rbrXHza9xq4cKFUKvV0ubs7HzvO0RERETVkm0oMaXQ0FAUFRVJ24ULF0zdEhER0QNPtqFEq9UCAPLz843G8/PzpTmtVouCggKj+Rs3buDSpUtGNbdb4+bXuJVSqYRKpTLaiIiIqGHJNpS4ublBq9UiOTlZGjMYDDh8+DB0Oh0AQKfTobCwEOnp6VJNSkoKKioq0Lt3b6lm//79KCsrk2qSkpLg4eGBZs2a3ae9ISIiorsxaSgpLi5GRkYGMjIyAPx7cWtGRgZyc3OhUCgQHByM9957Dzt27MDJkycxYcIEODk5YfTo0QAAT09PDB8+HFOmTMGRI0dw4MABzJgxA2PHjoWTkxMA4IUXXoCVlRUCAwORmZmJzZs3Y/ny5QgJCTHRXhMREdHtmPSW4KNHj2LQoEHS48qgEBAQgNjYWMyZMwdXr17F1KlTUVhYiH79+iEhIQHW1tbSczZu3IgZM2ZgyJAhMDMzg7+/P1asWCHNq9Vq7N69G0FBQejevTtatmyJsLAw3g5MREQkMyYNJQMHDoQQ4o7zCoUCERERiIiIuGNN8+bNsWnTpmpfp3Pnzvjhhx/q3CcRERE1PNleU0JERESNC0MJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCREREckCQwkRERHJQqMKJVFRUXB1dYW1tTV69+6NI0eOmLolIiIi+v81mlCyefNmhISEYP78+Th27Bi8vb3h6+uLgoICU7dGREREaEShZMmSJZgyZQomTZoELy8vREdHo0mTJvjss89M3RoREREBsDB1A/dDaWkp0tPTERoaKo2ZmZnBx8cHqampVepLSkpQUlIiPS4qKgIAGAyGeu+touSfel+T5KMhjpma4rH1YDPVscXj6sFX38dW5XpCiLvWNopQ8tdff6G8vBwajcZoXKPR4OzZs1XqFy5ciHfeeafKuLOzc4P1SA8m9TJTd0APKh5b1FAa6ti6cuUK1Gp1tTWNIpTUVmhoKEJCQqTHFRUVuHTpElq0aAGFQmHCzv7bDAYDnJ2dceHCBahUKlO3Qw8QHlvUUHhs3TshBK5cuQInJ6e71jaKUNKyZUuYm5sjPz/faDw/Px9arbZKvVKphFKpNBqzt7dvyBYbFZVKxf+4qUHw2KKGwmPr3tztDEmlRnGhq5WVFbp3747k5GRprKKiAsnJydDpdCbsjIiIiCo1ijMlABASEoKAgAD06NEDvXr1wrJly3D16lVMmjTJ1K0RERERGlEoee655/Dnn38iLCwMer0eXbp0QUJCQpWLX6nhKJVKzJ8/v8pHY0T3iscWNRQeW/eXQtTkHh0iIiKiBtYorikhIiIi+WMoISIiIllgKCEiIiJZYCihBjVw4EAEBwebug2SGR4XJEcKhQLbt283dRuNGkMJ1Yu9e/dCoVCgsLDQ1K0QEdF/FEMJ/eeUlpaaugUiImoADCVUYyUlJZg5cyYcHBxgbW2Nfv36IS0tDTk5ORg0aBAAoFmzZlAoFJg4caL0vIqKCsyZMwfNmzeHVqtFeHi40bqFhYWYPHkyHnroIahUKgwePBjHjx+X5sPDw9GlSxesXbsWbm5usLa2vh+7Sw2suuNiyZIl6NSpE2xtbeHs7IxXXnkFxcXF0nxsbCzs7e2xfft2tGvXDtbW1vD19cWFCxekmsrj5uOPP4azszOaNGmCZ599VvrV7/3798PS0hJ6vd6or+DgYPTv379hd57qxddff41OnTrBxsYGLVq0gI+PD65evYq0tDQMHToULVu2hFqtxmOPPYZjx44ZPffcuXMYMGAArK2t4eXlhaSkJKP5nJwcKBQKbN26FYMGDUKTJk3g7e1d5Zflf/zxR/Tv3x82NjZwdnbGzJkzcfXqVWl+9erV0jGq0WgwZsyYu/bfqAmiGpo5c6ZwcnISO3fuFJmZmSIgIEA0a9ZM/PXXX+Kbb74RAERWVpa4ePGiKCwsFEII8dhjjwmVSiXCw8PFzz//LNavXy8UCoXYvXu3tK6Pj494/PHHRVpamvj555/F66+/Llq0aCH+/vtvIYQQ8+fPF7a2tmL48OHi2LFj4vjx4ybZf6o/dzsuli5dKlJSUkR2drZITk4WHh4eYvr06dLzY2JihKWlpejRo4c4ePCgOHr0qOjVq5d49NFHpZrK42bw4MHip59+Evv27RPu7u7ihRdekGrat28vIiMjpcelpaWiZcuW4rPPPrsP7wLdi7y8PGFhYSGWLFkisrOzxYkTJ0RUVJS4cuWKSE5OFhs2bBBnzpwRp0+fFoGBgUKj0QiDwSCEEKK8vFw88sgjYsiQISIjI0Ps27dPdO3aVQAQ27ZtE0IIkZ2dLQCIDh06iLi4OJGVlSXGjBkjXFxcRFlZmRBCiPPnzwtbW1uxdOlS8fPPP4sDBw6Irl27iokTJwohhEhLSxPm5uZi06ZNIicnRxw7dkwsX778rv03ZgwlVCPFxcXC0tJSbNy4URorLS0VTk5OIjIyUnz//fcCgLh8+bLR8x577DHRr18/o7GePXuKuXPnCiGE+OGHH4RKpRLXr183qmnbtq34+OOPhRD//uNiaWkpCgoKGmDPyBTudlzc6quvvhItWrSQHsfExAgA4tChQ9LYmTNnBABx+PBhIcS/x425ubn4/fffpZpdu3YJMzMzcfHiRSGEEB9++KHw9PSU5r/55hthZ2cniouL730nqUGlp6cLACInJ+euteXl5aJp06biu+++E0IIkZiYKCwsLMQff/wh1ezateu2oWTt2rVSTWZmpgAgzpw5I4QQIjAwUEydOtXotX744QdhZmYmrl27Jr755huhUqmkMFTX/hsTfnxDNfLLL7+grKwMffv2lcYsLS3Rq1cvnDlzptrndu7c2eixo6MjCgoKAADHjx9HcXExWrRoATs7O2nLzs7GL7/8Ij3HxcUFDz30UD3uEZladcfFnj17MGTIELRq1QpNmzbF+PHj8ffff+Off/6R6i0sLNCzZ0/pcYcOHWBvb290PLZp0watWrWSHut0OlRUVCArKwsAMHHiRJw/fx6HDh0C8O/HQs8++yxsbW3rf4epXnl7e2PIkCHo1KkTnnnmGXz66ae4fPkygH9/AX7KlClo164d1Go1VCoViouLkZubCwA4c+YMnJ2d4eTkJK13px9nvfk4dXR0BACjv79iY2ON/u7y9fVFRUUFsrOzMXToULi4uODhhx/G+PHjsXHjRukYrq7/xoyhhBqcpaWl0WOFQoGKigoAQHFxMRwdHZGRkWG0ZWVlYfbs2dJz+I/Eg+dOx0VOTg5GjRqFzp0745tvvkF6ejqioqIA1P9Fzg4ODnj88ccRExOD/Px87Nq1Cy+99FK9vgY1DHNzcyQlJWHXrl3w8vLCypUr4eHhgezsbAQEBCAjIwPLly/HwYMHkZGRgRYtWtTp+Ln5OFUoFABg9PfXyy+/bPR31/Hjx3Hu3Dm0bdsWTZs2xbFjx/DFF1/A0dERYWFh8Pb2RmFhYbX9N2YMJVQjbdu2hZWVFQ4cOCCNlZWVIS0tDV5eXrCysgIAlJeX12rdbt26Qa/Xw8LCAu7u7kZby5Yt63Uf6L8hPT0dFRUVWLx4Mfr06YP27dsjLy+vSt2NGzdw9OhR6XFWVhYKCwvh6ekpjeXm5ho999ChQzAzM4OHh4c0NnnyZGzevBmffPIJ2rZta3Q2kORNoVCgb9++eOedd/DTTz/BysoK27Ztw4EDBzBz5kyMHDkSHTt2hFKpxF9//SU9z9PTExcuXMDFixelscqzZbXRrVs3nD59usrfXe7u7tLfiRYWFvDx8UFkZCROnDiBnJwcpKSkVNt/Y9ZofiWY7o2trS2mT5+O2bNno3nz5mjTpg0iIyPxzz//IDAwEP/88w8UCgXi4uIwcuRI2NjYwM7O7q7r+vj4QKfTYfTo0YiMjJT+AYqPj8dTTz2FHj163Ie9Izlxd3dHWVkZVq5ciccffxwHDhxAdHR0lTpLS0u8+uqrWLFiBSwsLDBjxgz06dMHvXr1kmqsra0REBCAjz76CAaDATNnzsSzzz4LrVYr1fj6+kKlUuG9995DRETEfdlHuneHDx9GcnIyhg0bBgcHBxw+fBh//vknPD090a5dO2zYsAE9evSAwWDA7NmzYWNjIz3Xx8cH7du3R0BAABYtWgSDwYC33nqr1j3MnTsXffr0wYwZMzB58mTY2tri9OnTSEpKwqpVqxAXF4dff/0VAwYMQLNmzbBz505UVFTAw8Oj2v4bM54poRr74IMP4O/vj/Hjx6Nbt244f/48EhMT0axZM7Rq1QrvvPMO3nzzTWg0GsyYMaNGayoUCuzcuRMDBgzApEmT0L59e4wdOxa//fYbNBpNA+8RyZG3tzeWLFmCDz/8EI888gg2btyIhQsXVqlr0qQJ5s6dixdeeAF9+/aFnZ0dNm/ebFTj7u6Op59+GiNHjsSwYcPQuXNnrF692qjGzMwMEydORHl5OSZMmNCg+0b1R6VSYf/+/Rg5ciTat2+PefPmYfHixRgxYgTWrVuHy5cvo1u3bhg/frz0VQaVzMzMsG3bNly7dg29evXC5MmTsWDBglr30LlzZ+zbtw8///wz+vfvj65duyIsLEy6VsXe3h5bt27F4MGD4enpiejoaHzxxRfo2LFjtf03ZgohhDB1E0REtREbG4vg4OBqv0E4PDwc27dvR0ZGxl3XCwwMxJ9//okdO3bUX5NEVGv8+IaIGq2ioiKcPHkSmzZtYiAhkgGGEiJqtJ588kkcOXIE06ZNw9ChQ03dDlGjx49viIiISBZ4oSsRERHJAkMJERERyQJDCREREckCQwkRERHJAkMJERERyQJDCRE1CuHh4ejSpYup2yCiajCUEFGDmThxIhQKRZVt+PDhDfq6CoUC27dvNxp74403kJyc3KCvS0T3hl+eRkQNavjw4YiJiTEaUyqV970POzu7Gv1IJBGZDs+UEFGDUiqV0Gq1RluzZs0A/HtG4+OPP8aoUaPQpEkTeHp6IjU1FefPn8fAgQNha2uLRx99FL/88ovRmmvWrEHbtm1hZWUFDw8PbNiwQZpzdXUFADz11FNQKBTS41s/vqmoqEBERARat24NpVKJLl26ICEhQZrPycmBQqHA1q1bMWjQIDRp0gTe3t5ITU1tmDeKiBhKiMi03n33XUyYMAEZGRno0KEDXnjhBbz88ssIDQ3F0aNHIYQw+tXpbdu24bXXXsPrr7+OU6dO4eWXX8akSZPw/fffAwDS0tIAADExMbh48aL0+FbLly/H4sWL8dFHH+HEiRPw9fXFE088gXPnzhnVvfXWW3jjjTeQkZGB9u3b4/nnn8eNGzca6N0gauQEEVEDCQgIEObm5sLW1tZoW7BggRBCCABi3rx5Un1qaqoAINatWyeNffHFF8La2lp6/Oijj4opU6YYvc4zzzwjRo4cKT0GILZt22ZUM3/+fOHt7S09dnJykvqo1LNnT/HKK68IIYTIzs4WAMTatWul+czMTAFAnDlzppbvBBHVBM+UEFGDGjRoEDIyMoy2adOmSfOdO3eW/qzRaAAAnTp1Mhq7fv06DAYDAODMmTPo27ev0Wv07dsXZ86cqXFPBoMBeXl5NVrn5v4cHR0BAAUFBTV+LSKqOV7oSkQNytbWFu7u7nect7S0lP6sUCjuOFZRUdFAHVZPTr0QPeh4poSI/lM8PT1x4MABo7EDBw7Ay8tLemxpaYny8vI7rqFSqeDk5HTXdYjo/uKZEiJqUCUlJdDr9UZjFhYWaNmyZZ3Wmz17Np599ll07doVPj4++O6777B161bs2bNHqnF1dUVycjL69u0LpVIp3e1z6zrz589H27Zt0aVLF8TExCAjIwMbN26sU19EdO8YSoioQSUkJEjXYlTy8PDA2bNn67Te6NGjsXz5cnz00Ud47bXX4ObmhpiYGAwcOFCqWbx4MUJCQvDpp5+iVatWyMnJqbLOzJkzUVRUhNdffx0FBQXw8vLCjh070K5duzr1RUT3TiGEEKZugoiIiIjXlBAREZEsMJQQERGRLDCUEBERkSwwlBAREZEsMJQQERGRLDCUEBERkSwwlBAREZEsMJQQERGRLDCUEBERkSwwlBAREZEsMJQQERGRLPx/rEsVCfBwz3QAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset Description:\n",
      "This dataset consists of two columns, Text and Emotions. Emotions column has various categories.\n",
      "Each entry in the dataset represents a text message, along with an associated Emotion label.\n",
      "Emotion categories include 'happy', 'sadness', 'anger', 'fear', 'love', 'surprise'.\n",
      "happy : 33%\n",
      "sadness : 29%\n",
      "other('anger', 'fear', 'love', 'surprise') : 38%\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Count the occurrences of each unique value in the 'Emotion' column\n",
    "emotion_counts = df['Emotion'].value_counts()\n",
    "\n",
    "# Create a bar plot\n",
    "plt.figure(figsize=(6, 4))  # optional: Adjust the figure size as needed\n",
    "plt.bar(emotion_counts.index, emotion_counts.values)\n",
    "\n",
    "# Adding labels and title\n",
    "plt.xlabel('Emotion')\n",
    "plt.ylabel('Count')\n",
    "plt.title('Distribution of Emotions')\n",
    "\n",
    "# Display the plot\n",
    "plt.show()\n",
    "\n",
    "# Dataset Description\n",
    "print(\"Dataset Description:\")\n",
    "print(\"This dataset consists of two columns, Text and Emotions. Emotions column has various categories.\")\n",
    "print(\"Each entry in the dataset represents a text message, along with an associated Emotion label.\")\n",
    "print(\"Emotion categories include 'happy', 'sadness', 'anger', 'fear', 'love', 'surprise'.\")\n",
    "print(\"happy : 33%\")\n",
    "print(\"sadness : 29%\")\n",
    "print(\"other('anger', 'fear', 'love', 'surprise') : 38%\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train data size:  (17171, 2)\n",
      "test data size:  (4288, 2)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "np.random.seed(1234)\n",
    "\n",
    "# split df into train and test\n",
    "i = np.random.rand(len(df)) < 0.8\n",
    "train = df[i]\n",
    "test = df[~i]\n",
    "print(\"train data size: \", train.shape)\n",
    "print(\"test data size: \", test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "classes: ['happy', 'other', 'sadness']\n",
      "train shapes: (17171, 25000) (17171,)\n",
      "test shapes: (4288, 25000) (4288,)\n",
      "test first ten labels: [1 0 1 1 1 0 1 0 0 2]\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "# set up X and Y\n",
    "num_labels = 3      # Total number of emotion categories\n",
    "vocab_size = 25000  # Size of the vocabulary to consider\n",
    "batch_size = 100    # Number of samples per gradient update\n",
    "\n",
    "# fit the tokenizer on the training data\n",
    "tokenizer = Tokenizer(num_words=vocab_size)\n",
    "tokenizer.fit_on_texts(train.Text)\n",
    "\n",
    "# convert texts to TF-IDF vectors (Text)\n",
    "x_train = tokenizer.texts_to_matrix(train.Text, mode='tfidf')\n",
    "x_test = tokenizer.texts_to_matrix(test.Text, mode='tfidf')\n",
    "\n",
    "# fit the label encoder and transform emotion labels to integer labels (Emotion)\n",
    "encoder = LabelEncoder()\n",
    "encoder.fit(train.Emotion)\n",
    "y_train = encoder.transform(train.Emotion)\n",
    "y_test = encoder.transform(test.Emotion)\n",
    "\n",
    "print(\"classes:\", list(encoder.classes_))   \n",
    "\n",
    "# check shape\n",
    "print(\"train shapes:\", x_train.shape, y_train.shape)\n",
    "print(\"test shapes:\", x_test.shape, y_test.shape)\n",
    "print(\"test first ten labels:\", y_test[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "\n",
    "# Convert texts to sequences of integers\n",
    "x_train_seq = tokenizer.texts_to_sequences(train.Text)\n",
    "x_test_seq = tokenizer.texts_to_sequences(test.Text)\n",
    "\n",
    "# Determine max sequence length for padding\n",
    "max_seq_length = max([len(seq) for seq in x_train_seq])\n",
    "\n",
    "# Pad sequences to ensure uniform length\n",
    "x_train_pad = pad_sequences(x_train_seq, maxlen=max_seq_length, padding='post')\n",
    "x_test_pad = pad_sequences(x_test_seq, maxlen=max_seq_length, padding='post')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras import layers, models\n",
    "\n",
    "max_features = 10000\n",
    "batch_size = 32\n",
    "\n",
    "model = models.Sequential()\n",
    "model.add(layers.Embedding(max_features, 500)) \n",
    "model.add(layers.Conv1D(32, 7, activation='relu')) \n",
    "model.add(layers.MaxPooling1D(5)) \n",
    "model.add(layers.Conv1D(32, 7, activation='relu')) \n",
    "model.add(layers.GlobalMaxPooling1D())\n",
    "model.add(layers.Dense(num_labels, activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "\n",
    "model.compile(optimizer=tf.keras.optimizers.RMSprop(learning_rate=1e-4),  # set learning rate\n",
    "              loss='categorical_crossentropy',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 64ms/step - accuracy: 0.3659 - loss: 1.0946 - val_accuracy: 0.4049 - val_loss: 1.0928\n",
      "Epoch 2/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 60ms/step - accuracy: 0.3763 - loss: 1.0880 - val_accuracy: 0.4049 - val_loss: 1.0915\n",
      "Epoch 3/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 62ms/step - accuracy: 0.3739 - loss: 1.0871 - val_accuracy: 0.4047 - val_loss: 1.0910\n",
      "Epoch 4/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 62ms/step - accuracy: 0.3815 - loss: 1.0850 - val_accuracy: 0.4084 - val_loss: 1.0900\n",
      "Epoch 5/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 62ms/step - accuracy: 0.4135 - loss: 1.0778 - val_accuracy: 0.4166 - val_loss: 1.0876\n",
      "Epoch 6/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 63ms/step - accuracy: 0.4352 - loss: 1.0698 - val_accuracy: 0.4323 - val_loss: 1.0842\n",
      "Epoch 7/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 61ms/step - accuracy: 0.4836 - loss: 1.0588 - val_accuracy: 0.4512 - val_loss: 1.0784\n",
      "Epoch 8/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 64ms/step - accuracy: 0.5357 - loss: 1.0393 - val_accuracy: 0.4710 - val_loss: 1.0658\n",
      "Epoch 9/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 68ms/step - accuracy: 0.5890 - loss: 1.0102 - val_accuracy: 0.4993 - val_loss: 1.0392\n",
      "Epoch 10/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 61ms/step - accuracy: 0.6263 - loss: 0.9573 - val_accuracy: 0.5394 - val_loss: 0.9899\n",
      "Epoch 11/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 62ms/step - accuracy: 0.6850 - loss: 0.8752 - val_accuracy: 0.5991 - val_loss: 0.9224\n",
      "Epoch 12/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 60ms/step - accuracy: 0.7663 - loss: 0.7516 - val_accuracy: 0.6591 - val_loss: 0.8266\n",
      "Epoch 13/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 62ms/step - accuracy: 0.8458 - loss: 0.6191 - val_accuracy: 0.7100 - val_loss: 0.7327\n",
      "Epoch 14/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 60ms/step - accuracy: 0.8987 - loss: 0.4775 - val_accuracy: 0.7287 - val_loss: 0.6764\n",
      "Epoch 15/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 61ms/step - accuracy: 0.9181 - loss: 0.3714 - val_accuracy: 0.7447 - val_loss: 0.6341\n",
      "Epoch 16/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 63ms/step - accuracy: 0.9385 - loss: 0.2819 - val_accuracy: 0.7636 - val_loss: 0.5647\n",
      "Epoch 17/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 67ms/step - accuracy: 0.9518 - loss: 0.2155 - val_accuracy: 0.7802 - val_loss: 0.5263\n",
      "Epoch 18/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 60ms/step - accuracy: 0.9582 - loss: 0.1748 - val_accuracy: 0.7808 - val_loss: 0.5200\n",
      "Epoch 19/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 60ms/step - accuracy: 0.9672 - loss: 0.1340 - val_accuracy: 0.7898 - val_loss: 0.4851\n",
      "Epoch 20/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 60ms/step - accuracy: 0.9760 - loss: 0.1055 - val_accuracy: 0.7936 - val_loss: 0.4770\n",
      "Epoch 21/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 59ms/step - accuracy: 0.9780 - loss: 0.0912 - val_accuracy: 0.7948 - val_loss: 0.4672\n",
      "Epoch 22/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 59ms/step - accuracy: 0.9800 - loss: 0.0793 - val_accuracy: 0.7977 - val_loss: 0.4684\n",
      "Epoch 23/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 59ms/step - accuracy: 0.9869 - loss: 0.0656 - val_accuracy: 0.7980 - val_loss: 0.4739\n",
      "Epoch 24/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 61ms/step - accuracy: 0.9862 - loss: 0.0580 - val_accuracy: 0.8023 - val_loss: 0.4708\n",
      "Epoch 25/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 60ms/step - accuracy: 0.9881 - loss: 0.0516 - val_accuracy: 0.8041 - val_loss: 0.4718\n",
      "Epoch 26/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 58ms/step - accuracy: 0.9916 - loss: 0.0426 - val_accuracy: 0.8012 - val_loss: 0.4849\n",
      "Epoch 27/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 59ms/step - accuracy: 0.9905 - loss: 0.0403 - val_accuracy: 0.8093 - val_loss: 0.4806\n",
      "Epoch 28/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 62ms/step - accuracy: 0.9927 - loss: 0.0333 - val_accuracy: 0.8067 - val_loss: 0.4859\n",
      "Epoch 29/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 69ms/step - accuracy: 0.9927 - loss: 0.0320 - val_accuracy: 0.8049 - val_loss: 0.4921\n",
      "Epoch 30/30\n",
      "\u001b[1m108/108\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 61ms/step - accuracy: 0.9944 - loss: 0.0275 - val_accuracy: 0.8119 - val_loss: 0.4892\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.utils import to_categorical\n",
    "\n",
    "# convert labels to one-hot encoding format\n",
    "y_train_cat = to_categorical(y_train, num_classes=num_labels)\n",
    "y_test_cat = to_categorical(y_test, num_classes=num_labels)\n",
    "\n",
    "# Train the model using the training data with one-hot encoded labels\n",
    "history = model.fit(x_train_pad, y_train_cat,\n",
    "                    epochs=30,\n",
    "                    batch_size=128,\n",
    "                    verbose=1,           \n",
    "                    validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m134/134\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step - accuracy: 0.9432 - loss: 0.1602\n",
      "Accuracy:  0.9113805890083313\n"
     ]
    }
   ],
   "source": [
    "# Evaluate\n",
    "score = model.evaluate(x_test_pad, y_test_cat, batch_size=batch_size, verbose=1)\n",
    "print('Accuracy: ', score[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m134/134\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 6ms/step\n"
     ]
    }
   ],
   "source": [
    "# get predictions so we can calculate more metrics\n",
    "pred = model.predict(x_test_pad)\n",
    "\n",
    "# Obtain the index of the highest probability category for each sample, which represents the predicted label\n",
    "pred_labels = np.argmax(pred, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy score:  0.9113805970149254\n",
      "precision score:  0.9114450392948313\n",
      "recall score:  0.9113805970149254\n",
      "f1 score:  0.9113769270953002\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score\n",
    "\n",
    "print('accuracy score: ', accuracy_score(y_test, pred_labels))\n",
    "print('precision score: ', precision_score(y_test, pred_labels, average='weighted'))\n",
    "print('recall score: ', recall_score(y_test, pred_labels, average='weighted'))\n",
    "print('f1 score: ', f1_score(y_test, pred_labels, average='weighted'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save('sentiment.keras')  # save the model"
   ]
  }
 ],
 "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": "3.11.2"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}