stock_exchange_mlp_predictions.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Previsão básica de ações utilizando Multi Layer Perceptron**\n",
    "\n",
    "- Diogo Nunes Batista\n",
    "- Rafael de Oliveira Évora\n",
    "- Vinicius Gabriel \n",
    "\n",
    "### Carregamento dos dados"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ABERTURA</th>\n",
       "      <th>FECHAMENTO</th>\n",
       "      <th>VARIAÇÃO</th>\n",
       "      <th>MÍNIMO</th>\n",
       "      <th>MÁXIMO</th>\n",
       "      <th>VOLUME</th>\n",
       "      <th>SELIC</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DATA</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>02/01/2024</th>\n",
       "      <td>0.0696</td>\n",
       "      <td>0.0704</td>\n",
       "      <td>0.10168</td>\n",
       "      <td>0.0691</td>\n",
       "      <td>0.0710</td>\n",
       "      <td>0.05515</td>\n",
       "      <td>0.043739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>03/01/2024</th>\n",
       "      <td>0.0703</td>\n",
       "      <td>0.0713</td>\n",
       "      <td>0.10127</td>\n",
       "      <td>0.0696</td>\n",
       "      <td>0.0718</td>\n",
       "      <td>0.04610</td>\n",
       "      <td>0.043739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>04/01/2024</th>\n",
       "      <td>0.0711</td>\n",
       "      <td>0.0701</td>\n",
       "      <td>0.09837</td>\n",
       "      <td>0.0680</td>\n",
       "      <td>0.0711</td>\n",
       "      <td>0.06905</td>\n",
       "      <td>0.043739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>05/01/2024</th>\n",
       "      <td>0.0696</td>\n",
       "      <td>0.0701</td>\n",
       "      <td>0.10000</td>\n",
       "      <td>0.0687</td>\n",
       "      <td>0.0706</td>\n",
       "      <td>0.03739</td>\n",
       "      <td>0.043739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>08/01/2024</th>\n",
       "      <td>0.0695</td>\n",
       "      <td>0.0691</td>\n",
       "      <td>0.09860</td>\n",
       "      <td>0.0684</td>\n",
       "      <td>0.0703</td>\n",
       "      <td>0.04757</td>\n",
       "      <td>0.043739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>08/11/2024</th>\n",
       "      <td>0.0622</td>\n",
       "      <td>0.0585</td>\n",
       "      <td>0.09451</td>\n",
       "      <td>0.0584</td>\n",
       "      <td>0.0607</td>\n",
       "      <td>0.13190</td>\n",
       "      <td>0.041957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11/11/2024</th>\n",
       "      <td>0.0581</td>\n",
       "      <td>0.0577</td>\n",
       "      <td>0.09863</td>\n",
       "      <td>0.0570</td>\n",
       "      <td>0.0581</td>\n",
       "      <td>0.05002</td>\n",
       "      <td>0.041957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12/11/2024</th>\n",
       "      <td>0.0581</td>\n",
       "      <td>0.0555</td>\n",
       "      <td>0.09619</td>\n",
       "      <td>0.0554</td>\n",
       "      <td>0.0579</td>\n",
       "      <td>0.10470</td>\n",
       "      <td>0.041957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13/11/2024</th>\n",
       "      <td>0.0578</td>\n",
       "      <td>0.0523</td>\n",
       "      <td>0.09423</td>\n",
       "      <td>0.0521</td>\n",
       "      <td>0.0558</td>\n",
       "      <td>0.13273</td>\n",
       "      <td>0.041957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14/11/2024</th>\n",
       "      <td>0.0545</td>\n",
       "      <td>0.0528</td>\n",
       "      <td>0.10096</td>\n",
       "      <td>0.0520</td>\n",
       "      <td>0.0542</td>\n",
       "      <td>0.12821</td>\n",
       "      <td>0.041957</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>222 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            ABERTURA  FECHAMENTO  VARIAÇÃO  MÍNIMO  MÁXIMO   VOLUME     SELIC\n",
       "DATA                                                                         \n",
       "02/01/2024    0.0696      0.0704   0.10168  0.0691  0.0710  0.05515  0.043739\n",
       "03/01/2024    0.0703      0.0713   0.10127  0.0696  0.0718  0.04610  0.043739\n",
       "04/01/2024    0.0711      0.0701   0.09837  0.0680  0.0711  0.06905  0.043739\n",
       "05/01/2024    0.0696      0.0701   0.10000  0.0687  0.0706  0.03739  0.043739\n",
       "08/01/2024    0.0695      0.0691   0.09860  0.0684  0.0703  0.04757  0.043739\n",
       "...              ...         ...       ...     ...     ...      ...       ...\n",
       "08/11/2024    0.0622      0.0585   0.09451  0.0584  0.0607  0.13190  0.041957\n",
       "11/11/2024    0.0581      0.0577   0.09863  0.0570  0.0581  0.05002  0.041957\n",
       "12/11/2024    0.0581      0.0555   0.09619  0.0554  0.0579  0.10470  0.041957\n",
       "13/11/2024    0.0578      0.0523   0.09423  0.0521  0.0558  0.13273  0.041957\n",
       "14/11/2024    0.0545      0.0528   0.10096  0.0520  0.0542  0.12821  0.041957\n",
       "\n",
       "[222 rows x 7 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas \n",
    "import numpy as np\n",
    "\n",
    "STOCK_NAME = 'CSNMineracao'\n",
    "\n",
    "df = pandas.read_csv(f'data/final/{STOCK_NAME}.csv', index_col='DATA', sep=';', decimal='.')\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Separação input e target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((208, 49), (208, 1), (7, 49), (7, 1))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# input = [(df[i:i+7].values, np.array([df.iat[i+7, 0]])) for i in range(0, df.index.size - 7)]\n",
    "N_DAYS = 7\n",
    "\n",
    "input = np.array([(df[i:i+N_DAYS].to_numpy().flatten()) for i in range(df.index.size - N_DAYS)])\n",
    "target = np.array([[df.iat[i+N_DAYS, 1]] for i in range(df.index.size - N_DAYS)])\n",
    "\n",
    "N_TEST = 7\n",
    "\n",
    "x_train = input[:-N_TEST]\n",
    "t_train = target[:-N_TEST]\n",
    "\n",
    "x_test = input[-N_TEST:]\n",
    "t_test = target[-N_TEST:]\n",
    "\n",
    "x_train.shape, t_train.shape, x_test.shape, t_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Treinamento"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test 0 finished with error: 0.15163748397498972 - epochs: 150000\n",
      "params: {'learning_rate': 0.001, 'max_epochs': 150000, 'tolerated_error': 0.001}\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mlp_model import MLP\n",
    "from concurrent.futures import ThreadPoolExecutor\n",
    "\n",
    "models = [\n",
    "    MLP(layers_sizes=[49, 200, 150,25,5, 1], activation='sigmoid'),\n",
    "    # MLP(layers_sizes=[49, 120, 120, 1], activation='sigmoid'),\n",
    "    # MLP(layers_sizes=[49, 100, 100, 1], activation='tanh'),\n",
    "    # MLP(layers_sizes=[49, 120, 120, 1], activation='tanh')\n",
    "]\n",
    "\n",
    "training_params = [\n",
    "    {'learning_rate': 0.001, 'max_epochs': 150000, 'tolerated_error': 0.001},\n",
    "    # {'learning_rate': 0.001, 'max_epochs': 10000, 'tolerated_error': 0.001},\n",
    "    # {'learning_rate': 0.001, 'max_epochs': 10000, 'tolerated_error': 0.001},\n",
    "    # {'learning_rate': 0.001, 'max_epochs': 10000, 'tolerated_error': 0.001}\n",
    "]\n",
    "\n",
    "def execute(i : int):\n",
    "    mlp = models[i]\n",
    "    # error, epochs = mlp.train(x_train, t_train, learning_rate=0.001, max_epochs=10000, tolerated_error=0.0001)\n",
    "    error, epochs = mlp.train(x_train, t_train, **training_params[i])\n",
    "\n",
    "    y_test = mlp.predict(x_test)\n",
    "\n",
    "    y = mlp.predict(x_train)\n",
    "\n",
    "    return y, y_test, error, epochs, training_params[i]\n",
    "\n",
    "with ThreadPoolExecutor(max_workers=4) as executor:\n",
    "    results = executor.map(execute, range(len(models)))\n",
    "\n",
    "    for i, result in enumerate(results):\n",
    "        y, y_test, error, epochs, params = result\n",
    "        print(f'Test {i} finished with error: {error} - epochs: {epochs}')\n",
    "        print(f'params: {params}')\n",
    "\n",
    "        df_test = df.iloc[-N_TEST:].copy()\n",
    "        df_test['PREVISTO'] = y_test\n",
    "\n",
    "        df_test.plot(y=['FECHAMENTO', 'PREVISTO'], xlabel='Data', ylabel='Valor de Fechamento', fontsize=8).get_figure().savefig(f'results/{STOCK_NAME}_{i}.png')\n",
    "\n",
    "        df_predict = df.iloc[N_DAYS:-N_TEST].copy()\n",
    "        df_predict['PREVISTO'] = y\n",
    "\n",
    "        df_predict.plot(y=['FECHAMENTO', 'PREVISTO'], xlabel='Data', ylabel='Valor de Fechamento', fontsize=8).get_figure().savefig(f'results/{STOCK_NAME}_test_{i}.png')\n",
    "        "
   ]
  }
 ],
 "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": "3.13.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}