Projeto_Redes_Neurais_LSTM.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2VnJfNuIWJB0"
   },
   "source": [
    "## IF702 - REDES NEURAIS\n",
    "\n",
    "**Grupo:**\n",
    "- Adriano Felipe Cabral Filho (afcf)\n",
    "- Felipe Rodrigues de Souza (frs3)\n",
    "- Lucas Vinícius José da Silva (lvjs)\n",
    "- Pedro José Carneiro de Souza (pjcs) \n",
    "- Vitor Sousa Silva (vss2)\n",
    "\n",
    "\n",
    "Obs: por alguma razão os gráficos do optuna não apareceram no arquivo, mas são tecidos comentários sobre os mesmos na última seção do documento."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "aA8EhILHB2dP",
    "outputId": "f31da54b-3dd8-46ca-dc59-14e2b64a8fac"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fatal: destination path 'predicting-bitcoin-prices-using-LSTM' already exists and is not an empty directory.\n",
      "WARNING:tensorflow:From /home/vitor/.local/lib/python3.8/site-packages/tensorflow/python/compat/v2_compat.py:101: disable_resource_variables (from tensorflow.python.ops.variable_scope) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "non-resource variables are not supported in the long term\n",
      "Tamanho do X treino (1018, 7, 1)\n",
      "Tamanho do y treino (1018, 1)\n",
      "Tamanho do X teste  (248, 7, 1)\n",
      "Tamanho do y teste  (248, 1)\n",
      "[]\n"
     ]
    }
   ],
   "source": [
    "!git clone https://github.com/brynmwangy/predicting-bitcoin-prices-using-LSTM/\n",
    "!pip3 install tensorflow --quiet\n",
    "!pip3 install numpy --quiet\n",
    "!pip3 install scikit-learn --quiet\n",
    "!pip3 install pandas --quiet\n",
    "!pip3 install matplotlib --quiet\n",
    "!pip3 install optuna --quiet\n",
    "!pip3 install plotly --quiet\n",
    "!pip3 install nbformat --quiet\n",
    "\n",
    "import tensorflow as tf\n",
    "import tensorflow.compat.v1 as tf\n",
    "tf.disable_v2_behavior()\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "scaler = StandardScaler()\n",
    "\n",
    "dados = pd.read_csv('./predicting-bitcoin-prices-using-LSTM/btc.csv')\n",
    "dados = dados[::-1]\n",
    "dados.reset_index()\n",
    "\n",
    "dados = scaler.fit_transform(dados['Close'].to_numpy().reshape(-1,1))\n",
    "\n",
    "def window_data(data, window_size):\n",
    "    X = []\n",
    "    y = []\n",
    "    \n",
    "    i = 0\n",
    "    while (i + window_size) <= len(data) - 1:\n",
    "        X.append(data[i:i+window_size])\n",
    "        y.append(data[i+window_size])\n",
    "        \n",
    "        i += 1\n",
    "    assert len(X) ==  len(y)\n",
    "    return X, y\n",
    "X, y = window_data(dados, 7)\n",
    "\n",
    "X_treino = np.array(X[:1018])\n",
    "y_treino = np.array(y[:1018])\n",
    "X_teste  = np.array(X[1018:])\n",
    "y_teste  = np.array(y[1018:])\n",
    "\n",
    "print(\"Tamanho do X treino {}\".format(X_treino.shape))\n",
    "print(\"Tamanho do y treino {}\".format(y_treino.shape))\n",
    "print(\"Tamanho do X teste  {}\".format(X_teste.shape))\n",
    "print(\"Tamanho do y teste  {}\".format(y_teste.shape))\n",
    "\n",
    "gpus = tf.config.list_physical_devices('GPU')\n",
    "if gpus:\n",
    "  try:\n",
    "    for gpu in gpus:\n",
    "      tf.config.experimental.set_memory_growth(gpu, True)\n",
    "    logical_gpus = tf.config.list_logical_devices('GPU')\n",
    "    print(len(gpus), \"Physical GPUs,\", len(logical_gpus), \"Logical GPUs\")\n",
    "  except RuntimeError as e:\n",
    "    print(e)\n",
    "print(gpus)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "l5kVXTk_Ou_x"
   },
   "outputs": [],
   "source": [
    "configs = {\n",
    "  \"batch_size\"   : 7,\n",
    "  \"window_size\"  : 7,\n",
    "  \"hidden_layer\" : 32,\n",
    "  \"clip_margin\"  : 20,\n",
    "  \"learning_rate\": 0.001,\n",
    "  \"epochs\"       : 60,\n",
    "  \"stddev\"       : 0.05,\n",
    "  \"epochs_print\" : 10\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "6lxBlbfCWJB6"
   },
   "source": [
    "### Parâmetros\n",
    "Tamanho do batch/janela: 7\n",
    "Hidden layer: 32\n",
    "Margem de clip: 20\n",
    "Taxa de aprendizado: 0.001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "E7TtNMjuYmIb"
   },
   "outputs": [],
   "source": [
    "# Pesos da via de entrada, inputs, objetivos, camada oculta e viés\n",
    "\n",
    "inputs = tf.placeholder( tf.float32, [ configs[\"batch_size\"], configs[\"window_size\"], 1 ] )\n",
    "targets = tf.placeholder( tf.float32, [ configs[\"batch_size\"], 1 ] )\n",
    "\n",
    "weights_input_gate = tf.Variable( tf.truncated_normal([ 1, configs[\"hidden_layer\"] ], stddev=0.05) )\n",
    "weights_input_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev=0.05) )\n",
    "bias_input = tf.Variable( tf.zeros( [ configs[\"hidden_layer\"] ] ) )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "MuqhkS_mamaN"
   },
   "outputs": [],
   "source": [
    "# Pesos da via do esquecimento, camada oculta e viés\n",
    "\n",
    "weights_forget_gate = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], stddev=0.05) )\n",
    "weights_forget_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev = configs[\"stddev\"] ) )\n",
    "bias_forget = tf.Variable( tf.zeros( configs[\"hidden_layer\"] ) )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "45-vNPNIbaKi"
   },
   "outputs": [],
   "source": [
    "# Pesos da via de saída, camada oculta e viés\n",
    "\n",
    "weights_output_gate = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], configs[\"hidden_layer\"] ) )\n",
    "weights_output_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev = configs[\"stddev\"] ) )\n",
    "bias_output = tf.Variable( tf.zeros( configs[\"hidden_layer\"] ) )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "PBWFgXd052yx"
   },
   "outputs": [],
   "source": [
    "# Pesos da célula de memória, camada oculta e viés\n",
    "\n",
    "weights_memory_cell = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], stddev=0.05 ) )\n",
    "weights_memory_cell_hidden = tf.Variable( tf.truncated_normal( [configs[\"hidden_layer\"], configs[\"hidden_layer\"]], stddev = configs[\"stddev\"] ) )\n",
    "bias_memory_cell = tf.Variable( tf.zeros( [configs[\"hidden_layer\"] ] ) )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "n3DeL7Za6QCd"
   },
   "outputs": [],
   "source": [
    "# Pesos da camada de saída (e viés)\n",
    "\n",
    "weights_output = tf.Variable( tf.truncated_normal( [configs[\"hidden_layer\"], 1], stddev = configs[\"stddev\"] ) )\n",
    "bias_output_layer = tf.Variable( tf.zeros([1]) )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "CL8S_BWr6gsg"
   },
   "outputs": [],
   "source": [
    "# Montando nossa célula LSTM\n",
    "\n",
    "def LSTM_cell(input, output, state):\n",
    "    # Via de entrada\n",
    "    input_gate = tf.sigmoid(tf.matmul(input, weights_input_gate) + tf.matmul(output, weights_input_hidden) + bias_input)\n",
    "    \n",
    "    # Via do esquecimento\n",
    "    forget_gate = tf.sigmoid(tf.matmul(input, weights_forget_gate) + tf.matmul(output, weights_forget_hidden) + bias_forget)\n",
    "        \n",
    "    # Célula de memória\n",
    "    memory_cell = tf.tanh(tf.matmul(input, weights_memory_cell) + tf.matmul(output, weights_memory_cell_hidden) + bias_memory_cell)\n",
    "    \n",
    "    # Estado (combinação dos anteriores)\n",
    "    state = state * forget_gate + input_gate * memory_cell\n",
    "\n",
    "    # Via de saída\n",
    "    output_gate = tf.sigmoid(tf.matmul(input, weights_output_gate) + tf.matmul(output, weights_output_hidden) + bias_output)\n",
    "    \n",
    "    # Saída (aplicação de função tangente à via de saída)\n",
    "    output = output_gate * tf.tanh(state)\n",
    "    return state, output\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "Ax6uW6pz7l_E"
   },
   "outputs": [],
   "source": [
    "# Percorrendo janela de dados e salvando resultados\n",
    "\n",
    "outputs = []\n",
    "for bs in range(configs[\"batch_size\"]):\n",
    "    batch_state  = np.zeros([1, configs[\"hidden_layer\"]], dtype=np.float32)\n",
    "    batch_output = np.zeros([1, configs[\"hidden_layer\"]], dtype=np.float32)\n",
    "\n",
    "    for ws in range(configs[\"window_size\"]):\n",
    "        batch_state, batch_output = LSTM_cell( tf.reshape( inputs[bs][ws], (-1, 1) ), batch_state, batch_output )\n",
    "\n",
    "    outputs.append( tf.matmul( batch_output, weights_output ) + bias_output_layer )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "xvEBRO9b8owK",
    "outputId": "b4bf3c21-6e7f-4e56-b7ee-7d4950bd89f9"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0/60  LOSS: 0.2114369124174118\n",
      "Epoch 10/60  LOSS: 0.012839341536164284\n",
      "Epoch 20/60  LOSS: 0.012306488119065762\n",
      "Epoch 30/60  LOSS: 0.009708047844469547\n",
      "Epoch 40/60  LOSS: 0.0033579072915017605\n",
      "Epoch 50/60  LOSS: 0.0004409407847560942\n"
     ]
    }
   ],
   "source": [
    "# Definindo perda\n",
    "\n",
    "losses = []\n",
    "\n",
    "for o in range(len(outputs)):\n",
    "    losses.append( tf.losses.mean_squared_error( tf.reshape( targets[o], (-1, 1) ), outputs[o] ) )\n",
    "\n",
    "loss = tf.reduce_mean(losses)\n",
    "\n",
    "# Definindo o otimizador com o gradiente de clipagem (?)\n",
    "gradients   = tf.gradients( loss, tf.trainable_variables() )\n",
    "clipped, _  = tf.clip_by_global_norm( gradients, configs[\"clip_margin\"] )\n",
    "optimizer   = tf.train.AdamOptimizer( configs[\"learning_rate\"] )\n",
    "trained_opt = optimizer.apply_gradients( zip( gradients, tf.trainable_variables() ) )\n",
    "\n",
    "session = tf.Session()\n",
    "session.run( tf.global_variables_initializer() )\n",
    "\n",
    "for e in range(configs[\"epochs\"]):\n",
    "    trained_scores = []\n",
    "    ee = 0\n",
    "    epoch_loss = []\n",
    "    while(ee + configs[\"batch_size\"]) <= len(X_treino): \n",
    "        X_batch = X_treino[ee:ee+configs[\"batch_size\"]]\n",
    "        y_batch = y_treino[ee:ee+configs[\"batch_size\"]]\n",
    "\n",
    "        o, c, _ = session.run( [outputs, loss, trained_opt], feed_dict={inputs:X_batch, targets: y_batch} )\n",
    "\n",
    "        epoch_loss.append(c)\n",
    "        trained_scores.append(o)\n",
    "        ee += configs[\"batch_size\"]\n",
    "\n",
    "    if(e % configs[\"epochs_print\"]) == 0:\n",
    "        print( 'Epoch {}/{}'.format(e, configs[\"epochs\"]), ' LOSS: {}'.format( np.mean( epoch_loss ) ) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "HuRwrn5dBOKj"
   },
   "outputs": [],
   "source": [
    "\n",
    "sup = []\n",
    "for ts in range(len(trained_scores)):\n",
    "    for tsr in range(len( trained_scores[ts] )):\n",
    "        sup.append( trained_scores[ts][tsr][0] )\n",
    "\n",
    "tests = []\n",
    "t = 0\n",
    "while( t + configs[\"batch_size\"] <= len(X_teste) ):\n",
    "    o = session.run( [outputs], feed_dict = { inputs: X_teste[t:t+configs[\"batch_size\"]] } )\n",
    "    t += configs[\"batch_size\"]\n",
    "    tests.append(o)\n",
    "\n",
    "novos_tests = []\n",
    "for ts in range(len(tests)):\n",
    "    for tsr in range(len( tests[ts][0] )):\n",
    "        novos_tests.append( tests[ts][0][tsr] )\n",
    "\n",
    "resultados_tests = []\n",
    "for r in range(1264):\n",
    "    if r >= 1019:\n",
    "        resultados_tests.append(novos_tests[r-1019])\n",
    "    else:\n",
    "        resultados_tests.append(None)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "9n1nqCQPHcNJ",
    "outputId": "8a00f514-51a2-4808-c9cb-148056850a8c"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/vitor/.local/lib/python3.8/site-packages/numpy/core/_asarray.py:136: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return array(a, dtype, copy=False, order=order, subok=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize=(16, 7))\n",
    "plt.title('Bitcoin prices from December 2014 to May 2018')\n",
    "plt.xlabel('Days')\n",
    "plt.ylabel('Scaled Price of Bitcoin')\n",
    "plt.plot(dados, label='Original data')\n",
    "plt.plot(sup, label='Training data')\n",
    "plt.plot(resultados_tests, label='Testing data')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8zXWTWuOWJB_"
   },
   "source": [
    "### Parâmetros do Estudo 1\n",
    "Trials: 16 <br>\n",
    "Tamanho do batch/janela: 7 <br>\n",
    "Hidden layer: 32 à 128 <br>\n",
    "Margem de clip: 4 à 64 <br>\n",
    "Taxa de aprendizado: 0.05 à 0.15 <br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "DolrC2_Do7OH",
    "outputId": "4db90357-3386-457d-e960-7f76f106291a"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:34:43,781]\u001b[0m A new study created in memory with name: no-name-59a039c6-5c2c-4259-b3a1-4db85caf2e61\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 71, 'clip_margin': 14, 'learning_rate': 0.0633319318479532, 'epochs': 60, 'stddev': 0.12933382482918607, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.24634768068790436\n",
      "Epoch 10/60  LOSS: 0.42830589413642883\n",
      "Epoch 20/60  LOSS: 0.006934761069715023\n",
      "Epoch 30/60  LOSS: 0.014057870022952557\n",
      "Epoch 40/60  LOSS: 0.006995856296271086\n",
      "Epoch 50/60  LOSS: 0.007985973730683327\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:35:31,800]\u001b[0m Trial 0 finished with value: 0.015463595697359217 and parameters: {'hidden_layer': 71, 'clip_margin': 14, 'learning_rate': 0.0633319318479532, 'stddev': 0.12933382482918607}. Best is trial 0 with value: 0.015463595697359217.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 35, 'clip_margin': 48, 'learning_rate': 0.040600832440222806, 'epochs': 60, 'stddev': 0.06186367460250364, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.07793401926755905\n",
      "Epoch 10/60  LOSS: 0.06869269162416458\n",
      "Epoch 20/60  LOSS: 0.004831113386899233\n",
      "Epoch 30/60  LOSS: 0.025925491005182266\n",
      "Epoch 40/60  LOSS: 0.011901136487722397\n",
      "Epoch 50/60  LOSS: 0.011899720877408981\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:36:11,226]\u001b[0m Trial 1 finished with value: 0.10919421006841856 and parameters: {'hidden_layer': 35, 'clip_margin': 48, 'learning_rate': 0.040600832440222806, 'stddev': 0.06186367460250364}. Best is trial 0 with value: 0.015463595697359217.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 68, 'clip_margin': 33, 'learning_rate': 0.027684182198906518, 'epochs': 60, 'stddev': 0.10167581247726706, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.04525112360715866\n",
      "Epoch 10/60  LOSS: 0.009183439426124096\n",
      "Epoch 20/60  LOSS: 0.08996368199586868\n",
      "Epoch 30/60  LOSS: 0.006026847753673792\n",
      "Epoch 40/60  LOSS: 0.0035885737743228674\n",
      "Epoch 50/60  LOSS: 0.01802368275821209\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:37:05,005]\u001b[0m Trial 2 finished with value: 0.01911936511345755 and parameters: {'hidden_layer': 68, 'clip_margin': 33, 'learning_rate': 0.027684182198906518, 'stddev': 0.10167581247726706}. Best is trial 0 with value: 0.015463595697359217.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 81, 'clip_margin': 6, 'learning_rate': 0.034666614115877646, 'epochs': 60, 'stddev': 0.10058708121365512, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.09068140387535095\n",
      "Epoch 10/60  LOSS: 0.007022745907306671\n",
      "Epoch 20/60  LOSS: 0.005236579105257988\n",
      "Epoch 30/60  LOSS: 0.01195207517594099\n",
      "Epoch 40/60  LOSS: 0.01938570663332939\n",
      "Epoch 50/60  LOSS: 0.1386166512966156\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:37:59,172]\u001b[0m Trial 3 finished with value: 0.023053284938131236 and parameters: {'hidden_layer': 81, 'clip_margin': 6, 'learning_rate': 0.034666614115877646, 'stddev': 0.10058708121365512}. Best is trial 0 with value: 0.015463595697359217.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 124, 'clip_margin': 6, 'learning_rate': 0.036836180955451665, 'epochs': 60, 'stddev': 0.07123679781276535, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 1.0196027755737305\n",
      "Epoch 10/60  LOSS: 0.09389112144708633\n",
      "Epoch 20/60  LOSS: 0.01529747061431408\n",
      "Epoch 30/60  LOSS: 0.015114999376237392\n",
      "Epoch 40/60  LOSS: 0.013435245491564274\n",
      "Epoch 50/60  LOSS: 0.011418633162975311\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:38:55,237]\u001b[0m Trial 4 finished with value: 0.005761461484676852 and parameters: {'hidden_layer': 124, 'clip_margin': 6, 'learning_rate': 0.036836180955451665, 'stddev': 0.07123679781276535}. Best is trial 4 with value: 0.005761461484676852.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 50, 'clip_margin': 32, 'learning_rate': 0.05657534665892468, 'epochs': 60, 'stddev': 0.10152453377377922, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.027741944417357445\n",
      "Epoch 10/60  LOSS: 0.012591270729899406\n",
      "Epoch 20/60  LOSS: 0.00860643107444048\n",
      "Epoch 30/60  LOSS: 0.004774277564138174\n",
      "Epoch 40/60  LOSS: 0.02148338221013546\n",
      "Epoch 50/60  LOSS: 0.0023393852170556784\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:39:46,447]\u001b[0m Trial 5 finished with value: 0.005365704418303768 and parameters: {'hidden_layer': 50, 'clip_margin': 32, 'learning_rate': 0.05657534665892468, 'stddev': 0.10152453377377922}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 125, 'clip_margin': 50, 'learning_rate': 0.09895332669439717, 'epochs': 60, 'stddev': 0.09590280597657397, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 1.0036219358444214\n",
      "Epoch 10/60  LOSS: 0.01969262771308422\n",
      "Epoch 20/60  LOSS: 0.00839470699429512\n",
      "Epoch 30/60  LOSS: 0.018914133310317993\n",
      "Epoch 40/60  LOSS: 0.04784975200891495\n",
      "Epoch 50/60  LOSS: 0.035986728966236115\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:40:41,718]\u001b[0m Trial 6 finished with value: 0.03267629125666488 and parameters: {'hidden_layer': 125, 'clip_margin': 50, 'learning_rate': 0.09895332669439717, 'stddev': 0.09590280597657397}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 39, 'clip_margin': 45, 'learning_rate': 0.07764038042649608, 'epochs': 60, 'stddev': 0.07979700376619127, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.2186170518398285\n",
      "Epoch 10/60  LOSS: 0.02035936340689659\n",
      "Epoch 20/60  LOSS: 0.020047931000590324\n",
      "Epoch 30/60  LOSS: 0.04344770312309265\n",
      "Epoch 40/60  LOSS: 0.019691869616508484\n",
      "Epoch 50/60  LOSS: 0.02671739086508751\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:41:30,677]\u001b[0m Trial 7 finished with value: 0.023230003413726958 and parameters: {'hidden_layer': 39, 'clip_margin': 45, 'learning_rate': 0.07764038042649608, 'stddev': 0.07979700376619127}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 71, 'clip_margin': 56, 'learning_rate': 0.04094016257400482, 'epochs': 60, 'stddev': 0.059985094814200515, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.060604508966207504\n",
      "Epoch 10/60  LOSS: 0.013889957219362259\n",
      "Epoch 20/60  LOSS: 0.010826966725289822\n",
      "Epoch 30/60  LOSS: 0.009556318633258343\n",
      "Epoch 40/60  LOSS: 0.00945307593792677\n",
      "Epoch 50/60  LOSS: 0.0277418065816164\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:42:20,352]\u001b[0m Trial 8 finished with value: 0.017076246655018435 and parameters: {'hidden_layer': 71, 'clip_margin': 56, 'learning_rate': 0.04094016257400482, 'stddev': 0.059985094814200515}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 53, 'clip_margin': 63, 'learning_rate': 0.08822946005987356, 'epochs': 60, 'stddev': 0.11519463172441914, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.028217285871505737\n",
      "Epoch 10/60  LOSS: 0.19234664738178253\n",
      "Epoch 20/60  LOSS: 0.04247713461518288\n",
      "Epoch 30/60  LOSS: 0.013777139596641064\n",
      "Epoch 40/60  LOSS: 0.0024612327106297016\n",
      "Epoch 50/60  LOSS: 0.06594859808683395\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:43:01,555]\u001b[0m Trial 9 finished with value: 0.008818625053248949 and parameters: {'hidden_layer': 53, 'clip_margin': 63, 'learning_rate': 0.08822946005987356, 'stddev': 0.11519463172441914}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 97, 'clip_margin': 27, 'learning_rate': 0.011666642412384555, 'epochs': 60, 'stddev': 0.1386461365656535, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.2219933271408081\n",
      "Epoch 10/60  LOSS: 0.007945024408400059\n",
      "Epoch 20/60  LOSS: 0.003504543099552393\n",
      "Epoch 30/60  LOSS: 0.006697486154735088\n",
      "Epoch 40/60  LOSS: 0.003918299917131662\n",
      "Epoch 50/60  LOSS: 0.008770763874053955\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:43:57,264]\u001b[0m Trial 10 finished with value: 0.014287055238849556 and parameters: {'hidden_layer': 97, 'clip_margin': 27, 'learning_rate': 0.011666642412384555, 'stddev': 0.1386461365656535}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 125, 'clip_margin': 20, 'learning_rate': 0.05977132500452518, 'epochs': 60, 'stddev': 0.07806664466493624, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.37427809834480286\n",
      "Epoch 10/60  LOSS: 0.06734833121299744\n",
      "Epoch 20/60  LOSS: 0.005465580150485039\n",
      "Epoch 30/60  LOSS: 0.006255015265196562\n",
      "Epoch 40/60  LOSS: 0.005888625048100948\n",
      "Epoch 50/60  LOSS: 0.03961179405450821\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:44:50,831]\u001b[0m Trial 11 finished with value: 0.02200287514004795 and parameters: {'hidden_layer': 125, 'clip_margin': 20, 'learning_rate': 0.05977132500452518, 'stddev': 0.07806664466493624}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 105, 'clip_margin': 4, 'learning_rate': 0.014604565345756368, 'epochs': 60, 'stddev': 0.07968879257407203, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.1366111934185028\n",
      "Epoch 10/60  LOSS: 0.0026706361677497625\n",
      "Epoch 20/60  LOSS: 0.00672927126288414\n",
      "Epoch 30/60  LOSS: 0.006328681483864784\n",
      "Epoch 40/60  LOSS: 0.06800814718008041\n",
      "Epoch 50/60  LOSS: 0.14811848104000092\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:45:38,821]\u001b[0m Trial 12 finished with value: 0.006055218872586857 and parameters: {'hidden_layer': 105, 'clip_margin': 4, 'learning_rate': 0.014604565345756368, 'stddev': 0.07968879257407203}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 48, 'clip_margin': 38, 'learning_rate': 0.051191575402560564, 'epochs': 60, 'stddev': 0.11740063446692542, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.01984240487217903\n",
      "Epoch 10/60  LOSS: 0.018642157316207886\n",
      "Epoch 20/60  LOSS: 0.010722176171839237\n",
      "Epoch 30/60  LOSS: 0.017138483002781868\n",
      "Epoch 40/60  LOSS: 0.029165877029299736\n",
      "Epoch 50/60  LOSS: 0.014209786430001259\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:46:07,739]\u001b[0m Trial 13 finished with value: 0.011112098819351025 and parameters: {'hidden_layer': 48, 'clip_margin': 38, 'learning_rate': 0.051191575402560564, 'stddev': 0.11740063446692542}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 105, 'clip_margin': 15, 'learning_rate': 0.0723551271722748, 'epochs': 60, 'stddev': 0.05180155951874785, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.14075283706188202\n",
      "Epoch 10/60  LOSS: 0.01780426874756813\n",
      "Epoch 20/60  LOSS: 0.008228941820561886\n",
      "Epoch 30/60  LOSS: 0.0036683904472738504\n",
      "Epoch 40/60  LOSS: 0.060153473168611526\n",
      "Epoch 50/60  LOSS: 0.034152351319789886\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:47:03,776]\u001b[0m Trial 14 finished with value: 0.05931481265958829 and parameters: {'hidden_layer': 105, 'clip_margin': 15, 'learning_rate': 0.0723551271722748, 'stddev': 0.05180155951874785}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 92, 'clip_margin': 26, 'learning_rate': 0.050000640106874, 'epochs': 60, 'stddev': 0.08939693540768119, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.4528822898864746\n",
      "Epoch 10/60  LOSS: 0.03881917521357536\n",
      "Epoch 20/60  LOSS: 0.01317871268838644\n",
      "Epoch 30/60  LOSS: 0.038725025951862335\n",
      "Epoch 40/60  LOSS: 0.035278867930173874\n",
      "Epoch 50/60  LOSS: 0.0555773563683033\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:47:51,336]\u001b[0m Trial 15 finished with value: 0.05780296813058597 and parameters: {'hidden_layer': 92, 'clip_margin': 26, 'learning_rate': 0.050000640106874, 'stddev': 0.08939693540768119}. Best is trial 5 with value: 0.005365704418303768.\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "import optuna\n",
    "\n",
    "configs = {\n",
    "  \"batch_size\"   : 7,\n",
    "  \"window_size\"  : 7,\n",
    "  \"hidden_layer\" : 32,\n",
    "  \"clip_margin\"  : 32,\n",
    "  \"learning_rate\": 0.01,\n",
    "  \"epochs\"       : 60,\n",
    "  \"stddev\"       : 0.05,\n",
    "  \"epochs_print\" : 10\n",
    "}\n",
    "bestCase = {\n",
    "   \"batch_size\": 0,\n",
    "   \"outputs\": [],\n",
    "   \"loss\": 1000000\n",
    "}\n",
    "bestTrial = 0\n",
    "\n",
    "def perda1(trial):\n",
    "    \n",
    "    configs[\"batch_size\"] = configs[\"window_size\"] = 7 #trial.suggest_int(\"batch_size\", 3, 31)\n",
    "    configs[\"hidden_layer\"] = trial.suggest_int(\"hidden_layer\", 32, 128)\n",
    "    configs[\"clip_margin\"] = trial.suggest_int(\"clip_margin\", 4, 64)\n",
    "    configs[\"learning_rate\"] = trial.suggest_float(\"learning_rate\", 0.01, 0.1)\n",
    "    configs[\"stddev\"] = trial.suggest_float(\"stddev\", 0.05, 0.15)\n",
    "\n",
    "    print(\"\\nConfigs\", configs, \"\\n\")\n",
    "\n",
    "    # Pesos da via de entrada, inputs, objetivos, camada oculta e viés\n",
    "    inputs = tf.placeholder( tf.float32, [ configs[\"batch_size\"], configs[\"window_size\"], 1 ] )\n",
    "    targets = tf.placeholder( tf.float32, [ configs[\"batch_size\"], 1 ] )\n",
    "\n",
    "    weights_input_gate = tf.Variable( tf.truncated_normal([ 1, configs[\"hidden_layer\"] ], stddev=0.05) )\n",
    "    weights_input_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev=0.05) )\n",
    "    bias_input = tf.Variable( tf.zeros( [ configs[\"hidden_layer\"] ] ) )\n",
    "\n",
    "    # Pesos da via do esquecimento, camada oculta e viés\n",
    "    weights_forget_gate = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], stddev=0.05) )\n",
    "    weights_forget_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev = configs[\"stddev\"] ) )\n",
    "    bias_forget = tf.Variable( tf.zeros( configs[\"hidden_layer\"] ) )\n",
    "\n",
    "    # Pesos da via de saída, camada oculta e viés\n",
    "    weights_output_gate = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], configs[\"hidden_layer\"] ) )\n",
    "    weights_output_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev = configs[\"stddev\"] ) )\n",
    "    bias_output = tf.Variable( tf.zeros( configs[\"hidden_layer\"] ) )\n",
    "\n",
    "    # Pesos da célula de memória, camada oculta e viés\n",
    "    weights_memory_cell = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], stddev=0.05 ) )\n",
    "    weights_memory_cell_hidden = tf.Variable( tf.truncated_normal( [configs[\"hidden_layer\"], configs[\"hidden_layer\"]], stddev = configs[\"stddev\"] ) )\n",
    "    bias_memory_cell = tf.Variable( tf.zeros( [configs[\"hidden_layer\"] ] ) )\n",
    "\n",
    "    # Pesos da camada de saída (e viés)\n",
    "    weights_output = tf.Variable( tf.truncated_normal( [configs[\"hidden_layer\"], 1], stddev = configs[\"stddev\"] ) )\n",
    "    bias_output_layer = tf.Variable( tf.zeros([1]) )\n",
    "\n",
    "\n",
    "    # Montando nossa célula LSTM\n",
    "    def LSTM_cell(input, output, state):\n",
    "        # Via de entrada\n",
    "        input_gate = tf.sigmoid(tf.matmul(input, weights_input_gate) + tf.matmul(output, weights_input_hidden) + bias_input)\n",
    "        \n",
    "        # Via do esquecimento\n",
    "        forget_gate = tf.sigmoid(tf.matmul(input, weights_forget_gate) + tf.matmul(output, weights_forget_hidden) + bias_forget)\n",
    "            \n",
    "        # Célula de memória\n",
    "        memory_cell = tf.tanh(tf.matmul(input, weights_memory_cell) + tf.matmul(output, weights_memory_cell_hidden) + bias_memory_cell)\n",
    "        \n",
    "        # Estado (combinação dos anteriores)\n",
    "        state = state * forget_gate + input_gate * memory_cell\n",
    "\n",
    "        # Via de saída\n",
    "        output_gate = tf.sigmoid(tf.matmul(input, weights_output_gate) + tf.matmul(output, weights_output_hidden) + bias_output)\n",
    "        \n",
    "        # Saída (aplicação de função tangente à via de saída)\n",
    "        output = output_gate * tf.tanh(state)\n",
    "        return state, output\n",
    "\n",
    "    \n",
    "    outputs = []\n",
    "    for bs in range(configs[\"batch_size\"]):\n",
    "        batch_state  = np.zeros([1, configs[\"hidden_layer\"]], dtype=np.float32)\n",
    "        batch_output = np.zeros([1, configs[\"hidden_layer\"]], dtype=np.float32)\n",
    "\n",
    "        for ws in range(configs[\"window_size\"]):\n",
    "            batch_state, batch_output = LSTM_cell( tf.reshape( inputs[bs][ws], (-1, 1) ), batch_state, batch_output )\n",
    "\n",
    "        outputs.append( tf.matmul( batch_output, weights_output ) + bias_output_layer )\n",
    "\n",
    "    # Definindo perda\n",
    "\n",
    "    losses = []\n",
    "\n",
    "    for o in range(len(outputs)):\n",
    "        losses.append( tf.losses.mean_squared_error( tf.reshape( targets[o], (-1, 1) ), outputs[o] ) )\n",
    "\n",
    "    loss = tf.reduce_mean(losses)\n",
    "\n",
    "    # Definindo o otimizador com o gradiente de clipagem (?)\n",
    "    gradients   = tf.gradients( loss, tf.trainable_variables() )\n",
    "    clipped, _  = tf.clip_by_global_norm( gradients, configs[\"clip_margin\"] )\n",
    "    optimizer   = tf.train.AdamOptimizer( configs[\"learning_rate\"] )\n",
    "    trained_opt = optimizer.apply_gradients( zip( gradients, tf.trainable_variables() ) )\n",
    "\n",
    "    session = tf.Session()\n",
    "    session.run( tf.global_variables_initializer() )\n",
    "\n",
    "    best_epochs = [10.0]\n",
    "\n",
    "    for e in range(configs[\"epochs\"]):\n",
    "        trained_scores = []\n",
    "        ee = 0\n",
    "        epoch_loss = []\n",
    "        while(ee + configs[\"batch_size\"]) <= len(X_treino): \n",
    "            X_batch = X_treino[ee:ee+configs[\"batch_size\"]]\n",
    "            y_batch = y_treino[ee:ee+configs[\"batch_size\"]]\n",
    "\n",
    "            o, c, _ = session.run( [outputs, loss, trained_opt], feed_dict={inputs:X_batch, targets: y_batch} )\n",
    "\n",
    "            epoch_loss.append(c)\n",
    "            trained_scores.append(o)\n",
    "            ee += configs[\"batch_size\"]\n",
    "\n",
    "        if(e % configs[\"epochs_print\"]) == 0:\n",
    "            if e == 0:\n",
    "                print( 'Epoch  {}/{}'.format(e, configs[\"epochs\"]), ' LOSS: {}'.format( np.mean( epoch_loss ) ) )\n",
    "            else: \n",
    "                print( 'Epoch {}/{}'.format(e, configs[\"epochs\"]), ' LOSS: {}'.format( np.mean( epoch_loss ) ) )\n",
    "    \n",
    "    loss = sum(epoch_loss)/len(epoch_loss)\n",
    "    if loss < bestCase[\"loss\"]:\n",
    "        bestCase[\"loss\"] = loss\n",
    "        global sup\n",
    "        sup = []\n",
    "        for ts in range(len(trained_scores)):\n",
    "            for tsr in range(len( trained_scores[ts] )):\n",
    "                sup.append( trained_scores[ts][tsr][0] )\n",
    "\n",
    "        tests = []\n",
    "        t = 0\n",
    "        while( t + configs[\"batch_size\"] <= len(X_teste) ):\n",
    "            o = session.run( [outputs], feed_dict = { inputs: X_teste[t:t+configs[\"batch_size\"]] } )\n",
    "            t += configs[\"batch_size\"]\n",
    "            tests.append(o)\n",
    "\n",
    "        novos_tests = []\n",
    "        for ts in range(len(tests)):\n",
    "            for tsr in range(len( tests[ts][0] )):\n",
    "                novos_tests.append( tests[ts][0][tsr] )\n",
    "\n",
    "        global resultados_tests \n",
    "        resultados_tests = []\n",
    "        for r in range(1264):\n",
    "            if r >= 1019:\n",
    "                resultados_tests.append(novos_tests[r-1019])\n",
    "            else:\n",
    "                resultados_tests.append(None)\n",
    "\n",
    "    # Retornando a média de LOSS\n",
    "    return loss\n",
    "\n",
    "study_1 = optuna.create_study(direction='minimize')\n",
    "study_1.optimize(perda1, n_trials=16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 546
    },
    "id": "egql70IQ88bV",
    "outputId": "71f2729d-969d-4fe1-c20e-9a5e4d81640d"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/vitor/.local/lib/python3.8/site-packages/numpy/core/_asarray.py:136: VisibleDeprecationWarning:\n",
      "\n",
      "Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize=(16, 7))\n",
    "plt.title('Study 1: Last 20% data prediction using LSTM')\n",
    "plt.xlabel('Days')\n",
    "plt.ylabel('Scaled Price of Bitcoin')\n",
    "plt.plot(dados, label='Original data')\n",
    "plt.plot(sup, label=\"Training data\")\n",
    "plt.plot(resultados_tests, label='Testing data')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "BEAb8NPh_iJN",
    "outputId": "30494487-ade6-445c-cb27-f6e27442caa3"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "mode": "markers",
         "name": "Objective Value",
         "type": "scatter",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15
         ],
         "y": [
          0.015463595697359217,
          0.10919421006841856,
          0.01911936511345755,
          0.023053284938131236,
          0.005761461484676852,
          0.005365704418303768,
          0.03267629125666488,
          0.023230003413726958,
          0.017076246655018435,
          0.008818625053248949,
          0.014287055238849556,
          0.02200287514004795,
          0.006055218872586857,
          0.011112098819351025,
          0.05931481265958829,
          0.05780296813058597
         ]
        },
        {
         "name": "Best Value",
         "type": "scatter",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15
         ],
         "y": [
          0.015463595697359217,
          0.015463595697359217,
          0.015463595697359217,
          0.015463595697359217,
          0.005761461484676852,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768,
          0.005365704418303768
         ]
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Optimization History Plot"
        },
        "xaxis": {
         "title": {
          "text": "#Trials"
         }
        },
        "yaxis": {
         "title": {
          "text": "Objective Value"
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_optimization_history(study_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ejdI865M_kzs",
    "outputId": "a52ab5e1-f8e1-4570-9388-766c23c4399d"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": true
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          14,
          48,
          33,
          6,
          6,
          32,
          50,
          45,
          56,
          63,
          27,
          20,
          4,
          38,
          15,
          26
         ],
         "xaxis": "x",
         "y": [
          0.015463595697359217,
          0.10919421006841856,
          0.01911936511345755,
          0.023053284938131236,
          0.005761461484676852,
          0.005365704418303768,
          0.03267629125666488,
          0.023230003413726958,
          0.017076246655018435,
          0.008818625053248949,
          0.014287055238849556,
          0.02200287514004795,
          0.006055218872586857,
          0.011112098819351025,
          0.05931481265958829,
          0.05780296813058597
         ],
         "yaxis": "y"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          71,
          35,
          68,
          81,
          124,
          50,
          125,
          39,
          71,
          53,
          97,
          125,
          105,
          48,
          105,
          92
         ],
         "xaxis": "x2",
         "y": [
          0.015463595697359217,
          0.10919421006841856,
          0.01911936511345755,
          0.023053284938131236,
          0.005761461484676852,
          0.005365704418303768,
          0.03267629125666488,
          0.023230003413726958,
          0.017076246655018435,
          0.008818625053248949,
          0.014287055238849556,
          0.02200287514004795,
          0.006055218872586857,
          0.011112098819351025,
          0.05931481265958829,
          0.05780296813058597
         ],
         "yaxis": "y2"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          0.0633319318479532,
          0.040600832440222806,
          0.027684182198906518,
          0.034666614115877646,
          0.036836180955451665,
          0.05657534665892468,
          0.09895332669439717,
          0.07764038042649608,
          0.04094016257400482,
          0.08822946005987356,
          0.011666642412384555,
          0.05977132500452518,
          0.014604565345756368,
          0.051191575402560564,
          0.0723551271722748,
          0.050000640106874
         ],
         "xaxis": "x3",
         "y": [
          0.015463595697359217,
          0.10919421006841856,
          0.01911936511345755,
          0.023053284938131236,
          0.005761461484676852,
          0.005365704418303768,
          0.03267629125666488,
          0.023230003413726958,
          0.017076246655018435,
          0.008818625053248949,
          0.014287055238849556,
          0.02200287514004795,
          0.006055218872586857,
          0.011112098819351025,
          0.05931481265958829,
          0.05780296813058597
         ],
         "yaxis": "y3"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          0.12933382482918607,
          0.06186367460250364,
          0.10167581247726706,
          0.10058708121365512,
          0.07123679781276535,
          0.10152453377377922,
          0.09590280597657397,
          0.07979700376619127,
          0.059985094814200515,
          0.11519463172441914,
          0.1386461365656535,
          0.07806664466493624,
          0.07968879257407203,
          0.11740063446692542,
          0.05180155951874785,
          0.08939693540768119
         ],
         "xaxis": "x4",
         "y": [
          0.015463595697359217,
          0.10919421006841856,
          0.01911936511345755,
          0.023053284938131236,
          0.005761461484676852,
          0.005365704418303768,
          0.03267629125666488,
          0.023230003413726958,
          0.017076246655018435,
          0.008818625053248949,
          0.014287055238849556,
          0.02200287514004795,
          0.006055218872586857,
          0.011112098819351025,
          0.05931481265958829,
          0.05780296813058597
         ],
         "yaxis": "y4"
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Slice Plot"
        },
        "width": 1200,
        "xaxis": {
         "anchor": "y",
         "domain": [
          0,
          0.2125
         ],
         "title": {
          "text": "clip_margin"
         }
        },
        "xaxis2": {
         "anchor": "y2",
         "domain": [
          0.2625,
          0.475
         ],
         "title": {
          "text": "hidden_layer"
         }
        },
        "xaxis3": {
         "anchor": "y3",
         "domain": [
          0.525,
          0.7375
         ],
         "title": {
          "text": "learning_rate"
         }
        },
        "xaxis4": {
         "anchor": "y4",
         "domain": [
          0.7875,
          1
         ],
         "title": {
          "text": "stddev"
         }
        },
        "yaxis": {
         "anchor": "x",
         "domain": [
          0,
          1
         ],
         "title": {
          "text": "Objective Value"
         }
        },
        "yaxis2": {
         "anchor": "x2",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        },
        "yaxis3": {
         "anchor": "x3",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        },
        "yaxis4": {
         "anchor": "x4",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_slice(study_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "_654btRqWJCD",
    "outputId": "58a7ad2f-6ff9-431e-ad21-28e5bdc04ecf"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "cliponaxis": false,
         "hovertemplate": [
          "clip_margin (IntUniformDistribution): 0.04751511331405367<extra></extra>",
          "learning_rate (UniformDistribution): 0.1189467887034442<extra></extra>",
          "stddev (UniformDistribution): 0.30973946449391426<extra></extra>",
          "hidden_layer (IntUniformDistribution): 0.5237986334885879<extra></extra>"
         ],
         "marker": {
          "color": [
           "rgb(8,81,156)",
           "rgb(8,48,107)",
           "rgb(8,48,107)",
           "rgb(8,81,156)"
          ]
         },
         "orientation": "h",
         "text": [
          "0.04751511331405367",
          "0.1189467887034442",
          "0.30973946449391426",
          "0.5237986334885879"
         ],
         "textposition": "outside",
         "texttemplate": "%{text:.2f}",
         "type": "bar",
         "x": [
          0.04751511331405367,
          0.1189467887034442,
          0.30973946449391426,
          0.5237986334885879
         ],
         "y": [
          "clip_margin",
          "learning_rate",
          "stddev",
          "hidden_layer"
         ]
        }
       ],
       "layout": {
        "showlegend": false,
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Hyperparameter Importances"
        },
        "xaxis": {
         "title": {
          "text": "Importance for Objective Value"
         }
        },
        "yaxis": {
         "title": {
          "text": "Hyperparameter"
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_param_importances(study_1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KP_uZBgEWJCD"
   },
   "source": [
    "### Parâmetros do Estudo 2\n",
    "Trials: 16 <br>\n",
    "Tamanho do batch/janela: 7 <br>\n",
    "Hidden layer: 64 à 256 <br>\n",
    "Margem de clip: 64 à 128 <br>\n",
    "Taxa de aprendizado: 0.1 à 0.5 <br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "6A8SExqubSx_",
    "outputId": "a2029dd2-a379-42e6-b71b-bba5b41b9b0d"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:39:29,770]\u001b[0m A new study created in memory with name: no-name-9c8c7dca-7a8c-495c-a2a7-29db1d73ff24\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 197, 'clip_margin': 87, 'learning_rate': 0.28731147414473723, 'epochs': 60, 'stddev': 0.10892521047779707, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 23.786651611328125\n",
      "Epoch 10/60  LOSS: 0.24432078003883362\n",
      "Epoch 20/60  LOSS: 0.41070613265037537\n",
      "Epoch 30/60  LOSS: 0.3337078094482422\n",
      "Epoch 40/60  LOSS: 0.3942481279373169\n",
      "Epoch 50/60  LOSS: 0.39073801040649414\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:40:37,928]\u001b[0m Trial 0 finished with value: 0.4070438464828103 and parameters: {'hidden_layer': 197, 'clip_margin': 87, 'learning_rate': 0.28731147414473723, 'stddev': 0.10892521047779707}. Best is trial 0 with value: 0.4070438464828103.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 170, 'clip_margin': 114, 'learning_rate': 0.4523912711864134, 'epochs': 60, 'stddev': 0.08878662748097808, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 49.38376998901367\n",
      "Epoch 10/60  LOSS: 1.1461553573608398\n",
      "Epoch 20/60  LOSS: 0.7815670371055603\n",
      "Epoch 30/60  LOSS: 1.511423945426941\n",
      "Epoch 40/60  LOSS: 1.233652949333191\n",
      "Epoch 50/60  LOSS: 0.8182525038719177\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:41:30,868]\u001b[0m Trial 1 finished with value: 0.8176512446790278 and parameters: {'hidden_layer': 170, 'clip_margin': 114, 'learning_rate': 0.4523912711864134, 'stddev': 0.08878662748097808}. Best is trial 0 with value: 0.4070438464828103.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 256, 'clip_margin': 64, 'learning_rate': 0.3996040075301285, 'epochs': 60, 'stddev': 0.14158656955436272, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 20.84975242614746\n",
      "Epoch 10/60  LOSS: 0.11732295900583267\n",
      "Epoch 20/60  LOSS: 0.721122145652771\n",
      "Epoch 30/60  LOSS: 0.13634297251701355\n",
      "Epoch 40/60  LOSS: 0.5880559086799622\n",
      "Epoch 50/60  LOSS: 0.5027512907981873\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:43:23,077]\u001b[0m Trial 2 finished with value: 3.2326829405936475 and parameters: {'hidden_layer': 256, 'clip_margin': 64, 'learning_rate': 0.3996040075301285, 'stddev': 0.14158656955436272}. Best is trial 0 with value: 0.4070438464828103.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 188, 'clip_margin': 101, 'learning_rate': 0.3315468413796342, 'epochs': 60, 'stddev': 0.0858313948787198, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 27.547929763793945\n",
      "Epoch 10/60  LOSS: 0.059783220291137695\n",
      "Epoch 20/60  LOSS: 0.2501884996891022\n",
      "Epoch 30/60  LOSS: 0.3942742645740509\n",
      "Epoch 40/60  LOSS: 0.39415571093559265\n",
      "Epoch 50/60  LOSS: 0.394089937210083\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:44:27,054]\u001b[0m Trial 3 finished with value: 0.394061185141298 and parameters: {'hidden_layer': 188, 'clip_margin': 101, 'learning_rate': 0.3315468413796342, 'stddev': 0.0858313948787198}. Best is trial 3 with value: 0.394061185141298.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 222, 'clip_margin': 128, 'learning_rate': 0.34958082828225423, 'epochs': 60, 'stddev': 0.08335425296033779, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 36.842529296875\n",
      "Epoch 10/60  LOSS: 0.3151966631412506\n",
      "Epoch 20/60  LOSS: 0.5250840187072754\n",
      "Epoch 30/60  LOSS: 0.4146926999092102\n",
      "Epoch 40/60  LOSS: 0.611014187335968\n",
      "Epoch 50/60  LOSS: 0.6533364057540894\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:45:59,566]\u001b[0m Trial 4 finished with value: 0.7215765356500422 and parameters: {'hidden_layer': 222, 'clip_margin': 128, 'learning_rate': 0.34958082828225423, 'stddev': 0.08335425296033779}. Best is trial 3 with value: 0.394061185141298.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 150, 'clip_margin': 89, 'learning_rate': 0.4352354648278759, 'epochs': 60, 'stddev': 0.05471615975105827, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 34.289833068847656\n",
      "Epoch 10/60  LOSS: 0.05729949474334717\n",
      "Epoch 20/60  LOSS: 0.4459030032157898\n",
      "Epoch 30/60  LOSS: 0.5124076008796692\n",
      "Epoch 40/60  LOSS: 0.5124531984329224\n",
      "Epoch 50/60  LOSS: 0.5288580060005188\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:46:47,212]\u001b[0m Trial 5 finished with value: 0.5112410216228329 and parameters: {'hidden_layer': 150, 'clip_margin': 89, 'learning_rate': 0.4352354648278759, 'stddev': 0.05471615975105827}. Best is trial 3 with value: 0.394061185141298.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 256, 'clip_margin': 116, 'learning_rate': 0.1403621396559157, 'epochs': 60, 'stddev': 0.13264440905635375, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 8.117172241210938\n",
      "Epoch 10/60  LOSS: 1.3416537046432495\n",
      "Epoch 20/60  LOSS: 0.021451493725180626\n",
      "Epoch 30/60  LOSS: 0.11527025699615479\n",
      "Epoch 40/60  LOSS: 0.5566946864128113\n",
      "Epoch 50/60  LOSS: 0.046272534877061844\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:48:27,664]\u001b[0m Trial 6 finished with value: 0.037891683366636494 and parameters: {'hidden_layer': 256, 'clip_margin': 116, 'learning_rate': 0.1403621396559157, 'stddev': 0.13264440905635375}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 145, 'clip_margin': 70, 'learning_rate': 0.17547804684358953, 'epochs': 60, 'stddev': 0.09628385490397316, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 5.875904083251953\n",
      "Epoch 10/60  LOSS: 0.0733150988817215\n",
      "Epoch 20/60  LOSS: 0.09903957694768906\n",
      "Epoch 30/60  LOSS: 0.12582091987133026\n",
      "Epoch 40/60  LOSS: 0.12872925400733948\n",
      "Epoch 50/60  LOSS: 0.1298445761203766\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:49:20,112]\u001b[0m Trial 7 finished with value: 0.1305868434950063 and parameters: {'hidden_layer': 145, 'clip_margin': 70, 'learning_rate': 0.17547804684358953, 'stddev': 0.09628385490397316}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 148, 'clip_margin': 79, 'learning_rate': 0.4852534232031165, 'epochs': 60, 'stddev': 0.05696840617526562, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 71.16302490234375\n",
      "Epoch 10/60  LOSS: 0.05358676239848137\n",
      "Epoch 20/60  LOSS: 0.1727895587682724\n",
      "Epoch 30/60  LOSS: 0.8098796010017395\n",
      "Epoch 40/60  LOSS: 0.7553268074989319\n",
      "Epoch 50/60  LOSS: 0.5882998704910278\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:50:03,662]\u001b[0m Trial 8 finished with value: 0.6990398244007745 and parameters: {'hidden_layer': 148, 'clip_margin': 79, 'learning_rate': 0.4852534232031165, 'stddev': 0.05696840617526562}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 137, 'clip_margin': 94, 'learning_rate': 0.17033627126963374, 'epochs': 60, 'stddev': 0.14136041009600375, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 1.870864748954773\n",
      "Epoch 10/60  LOSS: 0.7280149459838867\n",
      "Epoch 20/60  LOSS: 0.04417095333337784\n",
      "Epoch 30/60  LOSS: 0.040284786373376846\n",
      "Epoch 40/60  LOSS: 0.04415031149983406\n",
      "Epoch 50/60  LOSS: 0.04367367923259735\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:50:48,375]\u001b[0m Trial 9 finished with value: 0.044185904561928856 and parameters: {'hidden_layer': 137, 'clip_margin': 94, 'learning_rate': 0.17033627126963374, 'stddev': 0.14136041009600375}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 255, 'clip_margin': 128, 'learning_rate': 0.1037578631718944, 'epochs': 60, 'stddev': 0.12284902845629478, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 1.0462384223937988\n",
      "Epoch 10/60  LOSS: 0.11001550406217575\n",
      "Epoch 20/60  LOSS: 0.030465254560112953\n",
      "Epoch 30/60  LOSS: 0.42259588837623596\n",
      "Epoch 40/60  LOSS: 0.069508858025074\n",
      "Epoch 50/60  LOSS: 0.059178173542022705\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:52:44,870]\u001b[0m Trial 10 finished with value: 0.056302929960008505 and parameters: {'hidden_layer': 255, 'clip_margin': 128, 'learning_rate': 0.1037578631718944, 'stddev': 0.12284902845629478}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 228, 'clip_margin': 107, 'learning_rate': 0.18723507680238366, 'epochs': 60, 'stddev': 0.14822330642526704, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 2.756204843521118\n",
      "Epoch 10/60  LOSS: 0.11442955583333969\n",
      "Epoch 20/60  LOSS: 0.23447415232658386\n",
      "Epoch 30/60  LOSS: 0.2346375286579132\n",
      "Epoch 40/60  LOSS: 0.2365354597568512\n",
      "Epoch 50/60  LOSS: 0.2856990396976471\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:54:19,352]\u001b[0m Trial 11 finished with value: 0.05776479966226706 and parameters: {'hidden_layer': 228, 'clip_margin': 107, 'learning_rate': 0.18723507680238366, 'stddev': 0.14822330642526704}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 129, 'clip_margin': 113, 'learning_rate': 0.10347124612018005, 'epochs': 60, 'stddev': 0.13281001080714921, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 1.0992286205291748\n",
      "Epoch 10/60  LOSS: 0.06571561843156815\n",
      "Epoch 20/60  LOSS: 0.005645844619721174\n",
      "Epoch 30/60  LOSS: 0.008067562244832516\n",
      "Epoch 40/60  LOSS: 0.08349921554327011\n",
      "Epoch 50/60  LOSS: 0.049387332051992416\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:55:08,521]\u001b[0m Trial 12 finished with value: 0.060676627478727006 and parameters: {'hidden_layer': 129, 'clip_margin': 113, 'learning_rate': 0.10347124612018005, 'stddev': 0.13281001080714921}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 231, 'clip_margin': 96, 'learning_rate': 0.20224390082039828, 'epochs': 60, 'stddev': 0.11954527905283105, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 15.236525535583496\n",
      "Epoch 10/60  LOSS: 0.05388883128762245\n",
      "Epoch 20/60  LOSS: 0.20344148576259613\n",
      "Epoch 30/60  LOSS: 0.24556398391723633\n",
      "Epoch 40/60  LOSS: 0.24551323056221008\n",
      "Epoch 50/60  LOSS: 0.24550965428352356\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:56:43,754]\u001b[0m Trial 13 finished with value: 0.24551581033257866 and parameters: {'hidden_layer': 231, 'clip_margin': 96, 'learning_rate': 0.20224390082039828, 'stddev': 0.11954527905283105}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 200, 'clip_margin': 118, 'learning_rate': 0.25292799781418357, 'epochs': 60, 'stddev': 0.14813715053669255, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 21.37560272216797\n",
      "Epoch 10/60  LOSS: 0.05967293307185173\n",
      "Epoch 20/60  LOSS: 0.1526317596435547\n",
      "Epoch 30/60  LOSS: 0.29312822222709656\n",
      "Epoch 40/60  LOSS: 0.29295840859413147\n",
      "Epoch 50/60  LOSS: 0.2928783595561981\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:57:59,679]\u001b[0m Trial 14 finished with value: 0.29284612049501324 and parameters: {'hidden_layer': 200, 'clip_margin': 118, 'learning_rate': 0.25292799781418357, 'stddev': 0.14813715053669255}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 172, 'clip_margin': 103, 'learning_rate': 0.1451839422241405, 'epochs': 60, 'stddev': 0.13346136615009427, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 1.589428424835205\n",
      "Epoch 10/60  LOSS: 0.17467264831066132\n",
      "Epoch 20/60  LOSS: 0.1492558717727661\n",
      "Epoch 30/60  LOSS: 0.22260628640651703\n",
      "Epoch 40/60  LOSS: 0.20385459065437317\n",
      "Epoch 50/60  LOSS: 0.2078532874584198\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 01:59:00,455]\u001b[0m Trial 15 finished with value: 0.3299177016489482 and parameters: {'hidden_layer': 172, 'clip_margin': 103, 'learning_rate': 0.1451839422241405, 'stddev': 0.13346136615009427}. Best is trial 6 with value: 0.037891683366636494.\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "configs = {\n",
    "  \"batch_size\"   : 7,\n",
    "  \"window_size\"  : 7,\n",
    "  \"hidden_layer\" : 32,\n",
    "  \"clip_margin\"  : 32,\n",
    "  \"learning_rate\": 0.01,\n",
    "  \"epochs\"       : 60,\n",
    "  \"stddev\"       : 0.05,\n",
    "  \"epochs_print\" : 10\n",
    "}\n",
    "bestCase = {\n",
    "   \"batch_size\": 0,\n",
    "   \"outputs\": [],\n",
    "   \"loss\": 1000000\n",
    "}\n",
    "bestTrial = 0\n",
    "\n",
    "def perda2(trial):\n",
    "    \n",
    "    configs[\"batch_size\"] = configs[\"window_size\"] = 7 #trial.suggest_int(\"batch_size\", 3, 31)\n",
    "    configs[\"hidden_layer\"] = trial.suggest_int(\"hidden_layer\", 128, 256)\n",
    "    configs[\"clip_margin\"] = trial.suggest_int(\"clip_margin\", 64, 128)\n",
    "    configs[\"learning_rate\"] = trial.suggest_float(\"learning_rate\", 0.1, 0.5)\n",
    "    configs[\"stddev\"] = trial.suggest_float(\"stddev\", 0.05, 0.15)\n",
    "\n",
    "    print(\"\\nConfigs\", configs, \"\\n\")\n",
    "\n",
    "    # Pesos da via de entrada, inputs, objetivos, camada oculta e viés\n",
    "    inputs = tf.placeholder( tf.float32, [ configs[\"batch_size\"], configs[\"window_size\"], 1 ] )\n",
    "    targets = tf.placeholder( tf.float32, [ configs[\"batch_size\"], 1 ] )\n",
    "\n",
    "    weights_input_gate = tf.Variable( tf.truncated_normal([ 1, configs[\"hidden_layer\"] ], stddev=0.05) )\n",
    "    weights_input_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev=0.05) )\n",
    "    bias_input = tf.Variable( tf.zeros( [ configs[\"hidden_layer\"] ] ) )\n",
    "\n",
    "    # Pesos da via do esquecimento, camada oculta e viés\n",
    "    weights_forget_gate = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], stddev=0.05) )\n",
    "    weights_forget_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev = configs[\"stddev\"] ) )\n",
    "    bias_forget = tf.Variable( tf.zeros( configs[\"hidden_layer\"] ) )\n",
    "\n",
    "    # Pesos da via de saída, camada oculta e viés\n",
    "    weights_output_gate = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], configs[\"hidden_layer\"] ) )\n",
    "    weights_output_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev = configs[\"stddev\"] ) )\n",
    "    bias_output = tf.Variable( tf.zeros( configs[\"hidden_layer\"] ) )\n",
    "\n",
    "    # Pesos da célula de memória, camada oculta e viés\n",
    "    weights_memory_cell = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], stddev=0.05 ) )\n",
    "    weights_memory_cell_hidden = tf.Variable( tf.truncated_normal( [configs[\"hidden_layer\"], configs[\"hidden_layer\"]], stddev = configs[\"stddev\"] ) )\n",
    "    bias_memory_cell = tf.Variable( tf.zeros( [configs[\"hidden_layer\"] ] ) )\n",
    "\n",
    "    # Pesos da camada de saída (e viés)\n",
    "    weights_output = tf.Variable( tf.truncated_normal( [configs[\"hidden_layer\"], 1], stddev = configs[\"stddev\"] ) )\n",
    "    bias_output_layer = tf.Variable( tf.zeros([1]) )\n",
    "\n",
    "\n",
    "    # Montando nossa célula LSTM\n",
    "    def LSTM_cell(input, output, state):\n",
    "        # Via de entrada\n",
    "        input_gate = tf.sigmoid(tf.matmul(input, weights_input_gate) + tf.matmul(output, weights_input_hidden) + bias_input)\n",
    "        \n",
    "        # Via do esquecimento\n",
    "        forget_gate = tf.sigmoid(tf.matmul(input, weights_forget_gate) + tf.matmul(output, weights_forget_hidden) + bias_forget)\n",
    "            \n",
    "        # Célula de memória\n",
    "        memory_cell = tf.tanh(tf.matmul(input, weights_memory_cell) + tf.matmul(output, weights_memory_cell_hidden) + bias_memory_cell)\n",
    "        \n",
    "        # Estado (combinação dos anteriores)\n",
    "        state = state * forget_gate + input_gate * memory_cell\n",
    "\n",
    "        # Via de saída\n",
    "        output_gate = tf.sigmoid(tf.matmul(input, weights_output_gate) + tf.matmul(output, weights_output_hidden) + bias_output)\n",
    "        \n",
    "        # Saída (aplicação de função tangente à via de saída)\n",
    "        output = output_gate * tf.tanh(state)\n",
    "        return state, output\n",
    "\n",
    "    \n",
    "    outputs = []\n",
    "    for bs in range(configs[\"batch_size\"]):\n",
    "        batch_state  = np.zeros([1, configs[\"hidden_layer\"]], dtype=np.float32)\n",
    "        batch_output = np.zeros([1, configs[\"hidden_layer\"]], dtype=np.float32)\n",
    "\n",
    "        for ws in range(configs[\"window_size\"]):\n",
    "            batch_state, batch_output = LSTM_cell( tf.reshape( inputs[bs][ws], (-1, 1) ), batch_state, batch_output )\n",
    "\n",
    "        outputs.append( tf.matmul( batch_output, weights_output ) + bias_output_layer )\n",
    "\n",
    "    # Definindo perda\n",
    "\n",
    "    losses = []\n",
    "\n",
    "    for o in range(len(outputs)):\n",
    "        losses.append( tf.losses.mean_squared_error( tf.reshape( targets[o], (-1, 1) ), outputs[o] ) )\n",
    "\n",
    "    loss = tf.reduce_mean(losses)\n",
    "\n",
    "    # Definindo o otimizador com o gradiente de clipagem (?)\n",
    "    gradients   = tf.gradients( loss, tf.trainable_variables() )\n",
    "    clipped, _  = tf.clip_by_global_norm( gradients, configs[\"clip_margin\"] )\n",
    "    optimizer   = tf.train.AdamOptimizer( configs[\"learning_rate\"] )\n",
    "    trained_opt = optimizer.apply_gradients( zip( gradients, tf.trainable_variables() ) )\n",
    "\n",
    "    session = tf.Session()\n",
    "    session.run( tf.global_variables_initializer() )\n",
    "\n",
    "    best_epochs = [10.0]\n",
    "\n",
    "    for e in range(configs[\"epochs\"]):\n",
    "        trained_scores = []\n",
    "        ee = 0\n",
    "        epoch_loss = []\n",
    "        while(ee + configs[\"batch_size\"]) <= len(X_treino): \n",
    "            X_batch = X_treino[ee:ee+configs[\"batch_size\"]]\n",
    "            y_batch = y_treino[ee:ee+configs[\"batch_size\"]]\n",
    "\n",
    "            o, c, _ = session.run( [outputs, loss, trained_opt], feed_dict={inputs:X_batch, targets: y_batch} )\n",
    "\n",
    "            epoch_loss.append(c)\n",
    "            trained_scores.append(o)\n",
    "            ee += configs[\"batch_size\"]\n",
    "\n",
    "        if(e % configs[\"epochs_print\"]) == 0:\n",
    "            if e == 0:\n",
    "                print( 'Epoch  {}/{}'.format(e, configs[\"epochs\"]), ' LOSS: {}'.format( np.mean( epoch_loss ) ) )\n",
    "            else:\n",
    "                print( 'Epoch {}/{}'.format(e, configs[\"epochs\"]), ' LOSS: {}'.format( np.mean( epoch_loss ) ) )\n",
    "    \n",
    "    loss = sum(epoch_loss)/len(epoch_loss)\n",
    "    if loss < bestCase[\"loss\"]:\n",
    "        bestCase[\"loss\"] = loss\n",
    "        global sup\n",
    "        sup = []\n",
    "        for ts in range(len(trained_scores)):\n",
    "            for tsr in range(len( trained_scores[ts] )):\n",
    "                sup.append( trained_scores[ts][tsr][0] )\n",
    "\n",
    "        tests = []\n",
    "        t = 0\n",
    "        while( t + configs[\"batch_size\"] <= len(X_teste) ):\n",
    "            o = session.run( [outputs], feed_dict = { inputs: X_teste[t:t+configs[\"batch_size\"]] } )\n",
    "            t += configs[\"batch_size\"]\n",
    "            tests.append(o)\n",
    "\n",
    "        novos_tests = []\n",
    "        for ts in range(len(tests)):\n",
    "            for tsr in range(len( tests[ts][0] )):\n",
    "                novos_tests.append( tests[ts][0][tsr] )\n",
    "\n",
    "        global resultados_tests \n",
    "        resultados_tests = []\n",
    "        for r in range(1264):\n",
    "            if r >= 1019:\n",
    "                resultados_tests.append(novos_tests[r-1019])\n",
    "            else:\n",
    "                resultados_tests.append(None)\n",
    "\n",
    "    # Retornando a média de LOSS\n",
    "    return loss\n",
    "\n",
    "study_2 = optuna.create_study(direction='minimize')\n",
    "study_2.optimize(perda2, n_trials=16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "_OX6U9iObj9w",
    "outputId": "f7160a48-211a-4b16-a6d2-ae8f8151e2bf"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/vitor/.local/lib/python3.8/site-packages/numpy/core/_asarray.py:136: VisibleDeprecationWarning:\n",
      "\n",
      "Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize=(16, 7))\n",
    "plt.title('Study 2: Last 20% data prediction using LSTM')\n",
    "plt.xlabel('Days')\n",
    "plt.ylabel('Scaled Price of Bitcoin')\n",
    "plt.plot(dados, label='Original data')\n",
    "plt.plot(dados[:int((len(dados)/5)*4)], label=\"Training data\")\n",
    "plt.plot(resultados_tests, label='Testing data')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ZpD5Cpz7bo8M",
    "outputId": "8367a932-ae34-432a-cda7-8128e908017e"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "mode": "markers",
         "name": "Objective Value",
         "type": "scatter",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15
         ],
         "y": [
          0.4070438464828103,
          0.8176512446790278,
          3.2326829405936475,
          0.394061185141298,
          0.7215765356500422,
          0.5112410216228329,
          0.037891683366636494,
          0.1305868434950063,
          0.6990398244007745,
          0.044185904561928856,
          0.056302929960008505,
          0.05776479966226706,
          0.060676627478727006,
          0.24551581033257866,
          0.29284612049501324,
          0.3299177016489482
         ]
        },
        {
         "name": "Best Value",
         "type": "scatter",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15
         ],
         "y": [
          0.4070438464828103,
          0.4070438464828103,
          0.4070438464828103,
          0.394061185141298,
          0.394061185141298,
          0.394061185141298,
          0.037891683366636494,
          0.037891683366636494,
          0.037891683366636494,
          0.037891683366636494,
          0.037891683366636494,
          0.037891683366636494,
          0.037891683366636494,
          0.037891683366636494,
          0.037891683366636494,
          0.037891683366636494
         ]
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Optimization History Plot"
        },
        "xaxis": {
         "title": {
          "text": "#Trials"
         }
        },
        "yaxis": {
         "title": {
          "text": "Objective Value"
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_optimization_history(study_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "Mn5lzi8pbrRr",
    "outputId": "ada2d96e-920b-4418-b967-e75829cd9c1b"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": true
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          87,
          114,
          64,
          101,
          128,
          89,
          116,
          70,
          79,
          94,
          128,
          107,
          113,
          96,
          118,
          103
         ],
         "xaxis": "x",
         "y": [
          0.4070438464828103,
          0.8176512446790278,
          3.2326829405936475,
          0.394061185141298,
          0.7215765356500422,
          0.5112410216228329,
          0.037891683366636494,
          0.1305868434950063,
          0.6990398244007745,
          0.044185904561928856,
          0.056302929960008505,
          0.05776479966226706,
          0.060676627478727006,
          0.24551581033257866,
          0.29284612049501324,
          0.3299177016489482
         ],
         "yaxis": "y"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          197,
          170,
          256,
          188,
          222,
          150,
          256,
          145,
          148,
          137,
          255,
          228,
          129,
          231,
          200,
          172
         ],
         "xaxis": "x2",
         "y": [
          0.4070438464828103,
          0.8176512446790278,
          3.2326829405936475,
          0.394061185141298,
          0.7215765356500422,
          0.5112410216228329,
          0.037891683366636494,
          0.1305868434950063,
          0.6990398244007745,
          0.044185904561928856,
          0.056302929960008505,
          0.05776479966226706,
          0.060676627478727006,
          0.24551581033257866,
          0.29284612049501324,
          0.3299177016489482
         ],
         "yaxis": "y2"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          0.28731147414473723,
          0.4523912711864134,
          0.3996040075301285,
          0.3315468413796342,
          0.34958082828225423,
          0.4352354648278759,
          0.1403621396559157,
          0.17547804684358953,
          0.4852534232031165,
          0.17033627126963374,
          0.1037578631718944,
          0.18723507680238366,
          0.10347124612018005,
          0.20224390082039828,
          0.25292799781418357,
          0.1451839422241405
         ],
         "xaxis": "x3",
         "y": [
          0.4070438464828103,
          0.8176512446790278,
          3.2326829405936475,
          0.394061185141298,
          0.7215765356500422,
          0.5112410216228329,
          0.037891683366636494,
          0.1305868434950063,
          0.6990398244007745,
          0.044185904561928856,
          0.056302929960008505,
          0.05776479966226706,
          0.060676627478727006,
          0.24551581033257866,
          0.29284612049501324,
          0.3299177016489482
         ],
         "yaxis": "y3"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          0.10892521047779707,
          0.08878662748097808,
          0.14158656955436272,
          0.0858313948787198,
          0.08335425296033779,
          0.05471615975105827,
          0.13264440905635375,
          0.09628385490397316,
          0.05696840617526562,
          0.14136041009600375,
          0.12284902845629478,
          0.14822330642526704,
          0.13281001080714921,
          0.11954527905283105,
          0.14813715053669255,
          0.13346136615009427
         ],
         "xaxis": "x4",
         "y": [
          0.4070438464828103,
          0.8176512446790278,
          3.2326829405936475,
          0.394061185141298,
          0.7215765356500422,
          0.5112410216228329,
          0.037891683366636494,
          0.1305868434950063,
          0.6990398244007745,
          0.044185904561928856,
          0.056302929960008505,
          0.05776479966226706,
          0.060676627478727006,
          0.24551581033257866,
          0.29284612049501324,
          0.3299177016489482
         ],
         "yaxis": "y4"
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Slice Plot"
        },
        "width": 1200,
        "xaxis": {
         "anchor": "y",
         "domain": [
          0,
          0.2125
         ],
         "title": {
          "text": "clip_margin"
         }
        },
        "xaxis2": {
         "anchor": "y2",
         "domain": [
          0.2625,
          0.475
         ],
         "title": {
          "text": "hidden_layer"
         }
        },
        "xaxis3": {
         "anchor": "y3",
         "domain": [
          0.525,
          0.7375
         ],
         "title": {
          "text": "learning_rate"
         }
        },
        "xaxis4": {
         "anchor": "y4",
         "domain": [
          0.7875,
          1
         ],
         "title": {
          "text": "stddev"
         }
        },
        "yaxis": {
         "anchor": "x",
         "domain": [
          0,
          1
         ],
         "title": {
          "text": "Objective Value"
         }
        },
        "yaxis2": {
         "anchor": "x2",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        },
        "yaxis3": {
         "anchor": "x3",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        },
        "yaxis4": {
         "anchor": "x4",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_slice(study_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "SiW23Q0JWJCE",
    "outputId": "8b879a15-a486-4f61-c83f-d781452f1df2"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "cliponaxis": false,
         "hovertemplate": [
          "hidden_layer (IntUniformDistribution): 0.0703741123425194<extra></extra>",
          "stddev (UniformDistribution): 0.14505768628503404<extra></extra>",
          "learning_rate (UniformDistribution): 0.3335860555462762<extra></extra>",
          "clip_margin (IntUniformDistribution): 0.4509821458261703<extra></extra>"
         ],
         "marker": {
          "color": [
           "rgb(8,81,156)",
           "rgb(8,48,107)",
           "rgb(8,48,107)",
           "rgb(8,81,156)"
          ]
         },
         "orientation": "h",
         "text": [
          "0.0703741123425194",
          "0.14505768628503404",
          "0.3335860555462762",
          "0.4509821458261703"
         ],
         "textposition": "outside",
         "texttemplate": "%{text:.2f}",
         "type": "bar",
         "x": [
          0.0703741123425194,
          0.14505768628503404,
          0.3335860555462762,
          0.4509821458261703
         ],
         "y": [
          "hidden_layer",
          "stddev",
          "learning_rate",
          "clip_margin"
         ]
        }
       ],
       "layout": {
        "showlegend": false,
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Hyperparameter Importances"
        },
        "xaxis": {
         "title": {
          "text": "Importance for Objective Value"
         }
        },
        "yaxis": {
         "title": {
          "text": "Hyperparameter"
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_param_importances(study_2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tET7cz7bWJCE"
   },
   "source": [
    "### Parâmetros do Estudo 3\n",
    "Trials: 16 <br>\n",
    "Tamanho do batch/janela: 7 <br>\n",
    "Hidden layer: 256 à 512 <br>\n",
    "Margem de clip: 128 à 256 <br>\n",
    "Taxa de aprendizado: 0.3 à 0.7 <br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "b22mk8iLbtvG",
    "outputId": "b33593db-b3a2-43ec-8cb0-b05eb4010e5f"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:20:28,338]\u001b[0m A new study created in memory with name: no-name-beff6c9a-658c-42ac-870a-1735035dc870\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 30, 'clip_margin': 22, 'learning_rate': 0.0037347286149582125, 'epochs': 60, 'stddev': 0.10553049163883402, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.12742051482200623\n",
      "Epoch 10/60  LOSS: 0.012536275200545788\n",
      "Epoch 20/60  LOSS: 0.006031524855643511\n",
      "Epoch 30/60  LOSS: 0.0007955959299579263\n",
      "Epoch 40/60  LOSS: 0.000582517241127789\n",
      "Epoch 50/60  LOSS: 0.0005526210879907012\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:21:12,211]\u001b[0m Trial 0 finished with value: 0.00048017606419652393 and parameters: {'hidden_layer': 30, 'clip_margin': 22, 'learning_rate': 0.0037347286149582125, 'stddev': 0.10553049163883402}. Best is trial 0 with value: 0.00048017606419652393.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 20, 'clip_margin': 16, 'learning_rate': 0.002206639599309111, 'epochs': 60, 'stddev': 0.13337702038120913, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 3.8276615142822266\n",
      "Epoch 10/60  LOSS: 0.014626799151301384\n",
      "Epoch 20/60  LOSS: 0.011472244746983051\n",
      "Epoch 30/60  LOSS: 0.011270339600741863\n",
      "Epoch 40/60  LOSS: 0.010271896608173847\n",
      "Epoch 50/60  LOSS: 0.00484803831204772\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:21:55,501]\u001b[0m Trial 1 finished with value: 0.0013894868760589503 and parameters: {'hidden_layer': 20, 'clip_margin': 16, 'learning_rate': 0.002206639599309111, 'stddev': 0.13337702038120913}. Best is trial 0 with value: 0.00048017606419652393.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 11, 'clip_margin': 26, 'learning_rate': 0.004227672658599115, 'epochs': 60, 'stddev': 0.05890245612756484, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.1708863228559494\n",
      "Epoch 10/60  LOSS: 0.008308297954499722\n",
      "Epoch 20/60  LOSS: 0.004092466551810503\n",
      "Epoch 30/60  LOSS: 0.002523035043850541\n",
      "Epoch 40/60  LOSS: 0.0011005369015038013\n",
      "Epoch 50/60  LOSS: 0.0011642819736152887\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:22:40,207]\u001b[0m Trial 2 finished with value: 0.0011698623783621739 and parameters: {'hidden_layer': 11, 'clip_margin': 26, 'learning_rate': 0.004227672658599115, 'stddev': 0.05890245612756484}. Best is trial 0 with value: 0.00048017606419652393.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 14, 'clip_margin': 19, 'learning_rate': 0.003377067548621225, 'epochs': 60, 'stddev': 0.07679359438819774, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.2637307643890381\n",
      "Epoch 10/60  LOSS: 0.018171796575188637\n",
      "Epoch 20/60  LOSS: 0.009936193004250526\n",
      "Epoch 30/60  LOSS: 0.007227221038192511\n",
      "Epoch 40/60  LOSS: 0.0010843881173059344\n",
      "Epoch 50/60  LOSS: 0.0006544325733557343\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:23:20,744]\u001b[0m Trial 3 finished with value: 0.0006450670262454189 and parameters: {'hidden_layer': 14, 'clip_margin': 19, 'learning_rate': 0.003377067548621225, 'stddev': 0.07679359438819774}. Best is trial 0 with value: 0.00048017606419652393.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 31, 'clip_margin': 22, 'learning_rate': 0.0011244264980527455, 'epochs': 60, 'stddev': 0.07698484217075284, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.2072707861661911\n",
      "Epoch 10/60  LOSS: 0.013411223888397217\n",
      "Epoch 20/60  LOSS: 0.011885814368724823\n",
      "Epoch 30/60  LOSS: 0.006627291440963745\n",
      "Epoch 40/60  LOSS: 0.0006833273801021278\n",
      "Epoch 50/60  LOSS: 0.00045638190931640565\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:24:04,538]\u001b[0m Trial 4 finished with value: 0.0004387548195379936 and parameters: {'hidden_layer': 31, 'clip_margin': 22, 'learning_rate': 0.0011244264980527455, 'stddev': 0.07698484217075284}. Best is trial 4 with value: 0.0004387548195379936.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 18, 'clip_margin': 31, 'learning_rate': 0.003329377747437302, 'epochs': 60, 'stddev': 0.10821253658754612, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.808253288269043\n",
      "Epoch 10/60  LOSS: 0.01880737580358982\n",
      "Epoch 20/60  LOSS: 0.012180887162685394\n",
      "Epoch 30/60  LOSS: 0.008943503722548485\n",
      "Epoch 40/60  LOSS: 0.006223150063306093\n",
      "Epoch 50/60  LOSS: 0.0013978707138448954\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:24:47,960]\u001b[0m Trial 5 finished with value: 0.0011916806593286185 and parameters: {'hidden_layer': 18, 'clip_margin': 31, 'learning_rate': 0.003329377747437302, 'stddev': 0.10821253658754612}. Best is trial 4 with value: 0.0004387548195379936.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 23, 'clip_margin': 27, 'learning_rate': 0.0030071545557830153, 'epochs': 60, 'stddev': 0.054742582731521654, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 3.029115915298462\n",
      "Epoch 10/60  LOSS: 0.015111998654901981\n",
      "Epoch 20/60  LOSS: 0.0119004612788558\n",
      "Epoch 30/60  LOSS: 0.009546173736453056\n",
      "Epoch 40/60  LOSS: 0.008808310143649578\n",
      "Epoch 50/60  LOSS: 0.0065819378942251205\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:25:24,399]\u001b[0m Trial 6 finished with value: 0.0008531552140891955 and parameters: {'hidden_layer': 23, 'clip_margin': 27, 'learning_rate': 0.0030071545557830153, 'stddev': 0.054742582731521654}. Best is trial 4 with value: 0.0004387548195379936.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 8, 'clip_margin': 27, 'learning_rate': 0.001231174400482109, 'epochs': 60, 'stddev': 0.14094330501950722, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.3934575319290161\n",
      "Epoch 10/60  LOSS: 0.011271101422607899\n",
      "Epoch 20/60  LOSS: 0.0017917088698595762\n",
      "Epoch 30/60  LOSS: 0.0005708968383260071\n",
      "Epoch 40/60  LOSS: 0.00047031629947014153\n",
      "Epoch 50/60  LOSS: 0.00043338947580195963\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:26:06,656]\u001b[0m Trial 7 finished with value: 0.00043607260116098993 and parameters: {'hidden_layer': 8, 'clip_margin': 27, 'learning_rate': 0.001231174400482109, 'stddev': 0.14094330501950722}. Best is trial 7 with value: 0.00043607260116098993.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 24, 'clip_margin': 27, 'learning_rate': 0.0026469178069660902, 'epochs': 60, 'stddev': 0.06063117870609974, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 2.3672053813934326\n",
      "Epoch 10/60  LOSS: 0.018493549898266792\n",
      "Epoch 20/60  LOSS: 0.011574538424611092\n",
      "Epoch 30/60  LOSS: 0.010370494797825813\n",
      "Epoch 40/60  LOSS: 0.008938018232584\n",
      "Epoch 50/60  LOSS: 0.00445427093654871\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:26:47,408]\u001b[0m Trial 8 finished with value: 0.002078600580459571 and parameters: {'hidden_layer': 24, 'clip_margin': 27, 'learning_rate': 0.0026469178069660902, 'stddev': 0.06063117870609974}. Best is trial 7 with value: 0.00043607260116098993.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 23, 'clip_margin': 21, 'learning_rate': 0.003122987425426211, 'epochs': 60, 'stddev': 0.06153782807391946, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 3.0468831062316895\n",
      "Epoch 10/60  LOSS: 0.018257133662700653\n",
      "Epoch 20/60  LOSS: 0.012274855747818947\n",
      "Epoch 30/60  LOSS: 0.010673454031348228\n",
      "Epoch 40/60  LOSS: 0.009268088266253471\n",
      "Epoch 50/60  LOSS: 0.001639710390008986\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:27:18,532]\u001b[0m Trial 9 finished with value: 0.0016898407357718403 and parameters: {'hidden_layer': 23, 'clip_margin': 21, 'learning_rate': 0.003122987425426211, 'stddev': 0.06153782807391946}. Best is trial 7 with value: 0.00043607260116098993.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 8, 'clip_margin': 31, 'learning_rate': 0.0012914884095612668, 'epochs': 60, 'stddev': 0.1484905195053996, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.4017440378665924\n",
      "Epoch 10/60  LOSS: 0.015868326649069786\n",
      "Epoch 20/60  LOSS: 0.00542991841211915\n",
      "Epoch 30/60  LOSS: 0.0015131945256143808\n",
      "Epoch 40/60  LOSS: 0.0007766035269014537\n",
      "Epoch 50/60  LOSS: 0.0009121187031269073\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:27:55,680]\u001b[0m Trial 10 finished with value: 0.001072110742205919 and parameters: {'hidden_layer': 8, 'clip_margin': 31, 'learning_rate': 0.0012914884095612668, 'stddev': 0.1484905195053996}. Best is trial 7 with value: 0.00043607260116098993.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 32, 'clip_margin': 24, 'learning_rate': 0.0010899252712851255, 'epochs': 60, 'stddev': 0.0853726522904534, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.20904411375522614\n",
      "Epoch 10/60  LOSS: 0.012366831302642822\n",
      "Epoch 20/60  LOSS: 0.01079853530973196\n",
      "Epoch 30/60  LOSS: 0.00799987930804491\n",
      "Epoch 40/60  LOSS: 0.0005992865771986544\n",
      "Epoch 50/60  LOSS: 0.00042068568291142583\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:28:39,780]\u001b[0m Trial 11 finished with value: 0.000456215770326441 and parameters: {'hidden_layer': 32, 'clip_margin': 24, 'learning_rate': 0.0010899252712851255, 'stddev': 0.0853726522904534}. Best is trial 7 with value: 0.00043607260116098993.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 28, 'clip_margin': 29, 'learning_rate': 0.0017920075740621118, 'epochs': 60, 'stddev': 0.12802238957680404, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.2187896966934204\n",
      "Epoch 10/60  LOSS: 0.012097591534256935\n",
      "Epoch 20/60  LOSS: 0.005702266003936529\n",
      "Epoch 30/60  LOSS: 0.0004958926583640277\n",
      "Epoch 40/60  LOSS: 0.000359265977749601\n",
      "Epoch 50/60  LOSS: 0.0002908901951741427\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:29:25,913]\u001b[0m Trial 12 finished with value: 0.0003548882499517531 and parameters: {'hidden_layer': 28, 'clip_margin': 29, 'learning_rate': 0.0017920075740621118, 'stddev': 0.12802238957680404}. Best is trial 12 with value: 0.0003548882499517531.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 26, 'clip_margin': 29, 'learning_rate': 0.0018417951021639603, 'epochs': 60, 'stddev': 0.12806204320697395, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.3499254882335663\n",
      "Epoch 10/60  LOSS: 0.010308372788131237\n",
      "Epoch 20/60  LOSS: 0.0035767140798270702\n",
      "Epoch 30/60  LOSS: 0.006851503625512123\n",
      "Epoch 40/60  LOSS: 0.0030439298134297132\n",
      "Epoch 50/60  LOSS: 0.002161358715966344\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:30:05,012]\u001b[0m Trial 13 finished with value: 0.0023967705187182033 and parameters: {'hidden_layer': 26, 'clip_margin': 29, 'learning_rate': 0.0018417951021639603, 'stddev': 0.12806204320697395}. Best is trial 12 with value: 0.0003548882499517531.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 27, 'clip_margin': 29, 'learning_rate': 0.0018183981182289845, 'epochs': 60, 'stddev': 0.14930431145435258, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 0.29849743843078613\n",
      "Epoch 10/60  LOSS: 0.011748317629098892\n",
      "Epoch 20/60  LOSS: 0.004437736701220274\n",
      "Epoch 30/60  LOSS: 0.0026547126471996307\n",
      "Epoch 40/60  LOSS: 0.00284851947799325\n",
      "Epoch 50/60  LOSS: 0.001782021252438426\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:30:51,679]\u001b[0m Trial 14 finished with value: 0.0023361934949426887 and parameters: {'hidden_layer': 27, 'clip_margin': 29, 'learning_rate': 0.0018183981182289845, 'stddev': 0.14930431145435258}. Best is trial 12 with value: 0.0003548882499517531.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Configs {'batch_size': 7, 'window_size': 7, 'hidden_layer': 14, 'clip_margin': 32, 'learning_rate': 0.0016600287562503746, 'epochs': 60, 'stddev': 0.1261890471167614, 'epochs_print': 10} \n",
      "\n",
      "Epoch  0/60  LOSS: 1.624199390411377\n",
      "Epoch 10/60  LOSS: 0.0257050059735775\n",
      "Epoch 20/60  LOSS: 0.013584212400019169\n",
      "Epoch 30/60  LOSS: 0.00990284699946642\n",
      "Epoch 40/60  LOSS: 0.00888871680945158\n",
      "Epoch 50/60  LOSS: 0.007261344697326422\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2021-11-16 02:31:36,613]\u001b[0m Trial 15 finished with value: 0.0008607498633644919 and parameters: {'hidden_layer': 14, 'clip_margin': 32, 'learning_rate': 0.0016600287562503746, 'stddev': 0.1261890471167614}. Best is trial 12 with value: 0.0003548882499517531.\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "configs = {\n",
    "  \"batch_size\"   : 7,\n",
    "  \"window_size\"  : 7,\n",
    "  \"hidden_layer\" : 32,\n",
    "  \"clip_margin\"  : 32,\n",
    "  \"learning_rate\": 0.01,\n",
    "  \"epochs\"       : 60,\n",
    "  \"stddev\"       : 0.05,\n",
    "  \"epochs_print\" : 10\n",
    "}\n",
    "bestCase = {\n",
    "   \"batch_size\": 0,\n",
    "   \"outputs\": [],\n",
    "   \"loss\": 1000000\n",
    "}\n",
    "bestTrial = 0\n",
    "\n",
    "def perda3(trial):\n",
    "    \n",
    "    configs[\"batch_size\"] = configs[\"window_size\"] = 7 #trial.suggest_int(\"batch_size\", 3, 31)\n",
    "    configs[\"hidden_layer\"] = trial.suggest_int(\"hidden_layer\", 8, 32)\n",
    "    configs[\"clip_margin\"] = trial.suggest_int(\"clip_margin\", 16, 32)\n",
    "    configs[\"learning_rate\"] = trial.suggest_float(\"learning_rate\", 0.001, 0.005)\n",
    "    configs[\"stddev\"] = trial.suggest_float(\"stddev\", 0.05, 0.15)\n",
    "\n",
    "    print(\"\\nConfigs\", configs, \"\\n\")\n",
    "\n",
    "    # Pesos da via de entrada, inputs, objetivos, camada oculta e viés\n",
    "    inputs = tf.placeholder( tf.float32, [ configs[\"batch_size\"], configs[\"window_size\"], 1 ] )\n",
    "    targets = tf.placeholder( tf.float32, [ configs[\"batch_size\"], 1 ] )\n",
    "\n",
    "    weights_input_gate = tf.Variable( tf.truncated_normal([ 1, configs[\"hidden_layer\"] ], stddev=0.05) )\n",
    "    weights_input_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev=0.05) )\n",
    "    bias_input = tf.Variable( tf.zeros( [ configs[\"hidden_layer\"] ] ) )\n",
    "\n",
    "    # Pesos da via do esquecimento, camada oculta e viés\n",
    "    weights_forget_gate = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], stddev=0.05) )\n",
    "    weights_forget_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev = configs[\"stddev\"] ) )\n",
    "    bias_forget = tf.Variable( tf.zeros( configs[\"hidden_layer\"] ) )\n",
    "\n",
    "    # Pesos da via de saída, camada oculta e viés\n",
    "    weights_output_gate = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], configs[\"hidden_layer\"] ) )\n",
    "    weights_output_hidden = tf.Variable( tf.truncated_normal( [ configs[\"hidden_layer\"], configs[\"hidden_layer\"] ], stddev = configs[\"stddev\"] ) )\n",
    "    bias_output = tf.Variable( tf.zeros( configs[\"hidden_layer\"] ) )\n",
    "\n",
    "    # Pesos da célula de memória, camada oculta e viés\n",
    "    weights_memory_cell = tf.Variable( tf.truncated_normal( [1, configs[\"hidden_layer\"]], stddev=0.05 ) )\n",
    "    weights_memory_cell_hidden = tf.Variable( tf.truncated_normal( [configs[\"hidden_layer\"], configs[\"hidden_layer\"]], stddev = configs[\"stddev\"] ) )\n",
    "    bias_memory_cell = tf.Variable( tf.zeros( [configs[\"hidden_layer\"] ] ) )\n",
    "\n",
    "    # Pesos da camada de saída (e viés)\n",
    "    weights_output = tf.Variable( tf.truncated_normal( [configs[\"hidden_layer\"], 1], stddev = configs[\"stddev\"] ) )\n",
    "    bias_output_layer = tf.Variable( tf.zeros([1]) )\n",
    "\n",
    "\n",
    "    # Montando nossa célula LSTM\n",
    "    def LSTM_cell(input, output, state):\n",
    "        # Via de entrada\n",
    "        input_gate = tf.sigmoid(tf.matmul(input, weights_input_gate) + tf.matmul(output, weights_input_hidden) + bias_input)\n",
    "        \n",
    "        # Via do esquecimento\n",
    "        forget_gate = tf.sigmoid(tf.matmul(input, weights_forget_gate) + tf.matmul(output, weights_forget_hidden) + bias_forget)\n",
    "            \n",
    "        # Célula de memória\n",
    "        memory_cell = tf.tanh(tf.matmul(input, weights_memory_cell) + tf.matmul(output, weights_memory_cell_hidden) + bias_memory_cell)\n",
    "        \n",
    "        # Estado (combinação dos anteriores)\n",
    "        state = state * forget_gate + input_gate * memory_cell\n",
    "\n",
    "        # Via de saída\n",
    "        output_gate = tf.sigmoid(tf.matmul(input, weights_output_gate) + tf.matmul(output, weights_output_hidden) + bias_output)\n",
    "        \n",
    "        # Saída (aplicação de função tangente à via de saída)\n",
    "        output = output_gate * tf.tanh(state)\n",
    "        return state, output\n",
    "\n",
    "    \n",
    "    outputs = []\n",
    "    for bs in range(configs[\"batch_size\"]):\n",
    "        batch_state  = np.zeros([1, configs[\"hidden_layer\"]], dtype=np.float32)\n",
    "        batch_output = np.zeros([1, configs[\"hidden_layer\"]], dtype=np.float32)\n",
    "\n",
    "        for ws in range(configs[\"window_size\"]):\n",
    "            batch_state, batch_output = LSTM_cell( tf.reshape( inputs[bs][ws], (-1, 1) ), batch_state, batch_output )\n",
    "\n",
    "        outputs.append( tf.matmul( batch_output, weights_output ) + bias_output_layer )\n",
    "\n",
    "    # Definindo perda\n",
    "\n",
    "    losses = []\n",
    "\n",
    "    for o in range(len(outputs)):\n",
    "        losses.append( tf.losses.mean_squared_error( tf.reshape( targets[o], (-1, 1) ), outputs[o] ) )\n",
    "\n",
    "    loss = tf.reduce_mean(losses)\n",
    "\n",
    "    # Definindo o otimizador com o gradiente de clipagem (?)\n",
    "    gradients   = tf.gradients( loss, tf.trainable_variables() )\n",
    "    clipped, _  = tf.clip_by_global_norm( gradients, configs[\"clip_margin\"] )\n",
    "    optimizer   = tf.train.AdamOptimizer( configs[\"learning_rate\"] )\n",
    "    trained_opt = optimizer.apply_gradients( zip( gradients, tf.trainable_variables() ) )\n",
    "\n",
    "    session = tf.Session()\n",
    "    session.run( tf.global_variables_initializer() )\n",
    "\n",
    "    best_epochs = [10.0]\n",
    "\n",
    "    for e in range(configs[\"epochs\"]):\n",
    "        trained_scores = []\n",
    "        ee = 0\n",
    "        epoch_loss = []\n",
    "        while(ee + configs[\"batch_size\"]) <= len(X_treino): \n",
    "            X_batch = X_treino[ee:ee+configs[\"batch_size\"]]\n",
    "            y_batch = y_treino[ee:ee+configs[\"batch_size\"]]\n",
    "\n",
    "            o, c, _ = session.run( [outputs, loss, trained_opt], feed_dict={inputs:X_batch, targets: y_batch} )\n",
    "\n",
    "            epoch_loss.append(c)\n",
    "            trained_scores.append(o)\n",
    "            ee += configs[\"batch_size\"]\n",
    "\n",
    "        if(e % configs[\"epochs_print\"]) == 0:\n",
    "            if e == 0:\n",
    "                print( 'Epoch  {}/{}'.format(e, configs[\"epochs\"]), ' LOSS: {}'.format( np.mean( epoch_loss ) ) )\n",
    "            else:\n",
    "                print( 'Epoch {}/{}'.format(e, configs[\"epochs\"]), ' LOSS: {}'.format( np.mean( epoch_loss ) ) )\n",
    "    \n",
    "    loss = sum(epoch_loss)/len(epoch_loss)\n",
    "    if loss < bestCase[\"loss\"]:\n",
    "        bestCase[\"loss\"] = loss\n",
    "        global sup\n",
    "        sup = []\n",
    "        for ts in range(len(trained_scores)):\n",
    "            for tsr in range(len( trained_scores[ts] )):\n",
    "                sup.append( trained_scores[ts][tsr][0] )\n",
    "\n",
    "        tests = []\n",
    "        t = 0\n",
    "        while( t + configs[\"batch_size\"] <= len(X_teste) ):\n",
    "            o = session.run( [outputs], feed_dict = { inputs: X_teste[t:t+configs[\"batch_size\"]] } )\n",
    "            t += configs[\"batch_size\"]\n",
    "            tests.append(o)\n",
    "\n",
    "        novos_tests = []\n",
    "        for ts in range(len(tests)):\n",
    "            for tsr in range(len( tests[ts][0] )):\n",
    "                novos_tests.append( tests[ts][0][tsr] )\n",
    "\n",
    "        global resultados_tests \n",
    "        resultados_tests = []\n",
    "        for r in range(1264):\n",
    "            if r >= 1019:\n",
    "                resultados_tests.append(novos_tests[r-1019])\n",
    "            else:\n",
    "                resultados_tests.append(None)\n",
    "\n",
    "    # Retornando a média de LOSS\n",
    "    return loss\n",
    "\n",
    "study_3 = optuna.create_study(direction='minimize')\n",
    "study_3.optimize(perda3, n_trials=16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "wQL6JaEPb0E2",
    "outputId": "a2adda29-1a1b-476c-90f9-2b38716d2c71"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/vitor/.local/lib/python3.8/site-packages/numpy/core/_asarray.py:136: VisibleDeprecationWarning:\n",
      "\n",
      "Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize=(16, 7))\n",
    "plt.title('Study 3: Last 20% data prediction using LSTM')\n",
    "plt.xlabel('Days')\n",
    "plt.ylabel('Scaled Price of Bitcoin')\n",
    "plt.plot(dados, label='Original data')\n",
    "plt.plot(dados[:int((len(dados)/5)*4)], label=\"Training data\")\n",
    "plt.plot(resultados_tests, label='Testing data')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "8xpdrq0cb1z0",
    "outputId": "c3411b7a-94c3-489b-b5c9-6fbf15aed896"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "mode": "markers",
         "name": "Objective Value",
         "type": "scatter",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15
         ],
         "y": [
          0.00048017606419652393,
          0.0013894868760589503,
          0.0011698623783621739,
          0.0006450670262454189,
          0.0004387548195379936,
          0.0011916806593286185,
          0.0008531552140891955,
          0.00043607260116098993,
          0.002078600580459571,
          0.0016898407357718403,
          0.001072110742205919,
          0.000456215770326441,
          0.0003548882499517531,
          0.0023967705187182033,
          0.0023361934949426887,
          0.0008607498633644919
         ]
        },
        {
         "name": "Best Value",
         "type": "scatter",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15
         ],
         "y": [
          0.00048017606419652393,
          0.00048017606419652393,
          0.00048017606419652393,
          0.00048017606419652393,
          0.0004387548195379936,
          0.0004387548195379936,
          0.0004387548195379936,
          0.00043607260116098993,
          0.00043607260116098993,
          0.00043607260116098993,
          0.00043607260116098993,
          0.00043607260116098993,
          0.0003548882499517531,
          0.0003548882499517531,
          0.0003548882499517531,
          0.0003548882499517531
         ]
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Optimization History Plot"
        },
        "xaxis": {
         "title": {
          "text": "#Trials"
         }
        },
        "yaxis": {
         "title": {
          "text": "Objective Value"
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_optimization_history(study_3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "acLJgVhtb3X_",
    "outputId": "9409dcb8-3d2d-4cdf-982e-4774c244dfdc"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": true
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          22,
          16,
          26,
          19,
          22,
          31,
          27,
          27,
          27,
          21,
          31,
          24,
          29,
          29,
          29,
          32
         ],
         "xaxis": "x",
         "y": [
          0.00048017606419652393,
          0.0013894868760589503,
          0.0011698623783621739,
          0.0006450670262454189,
          0.0004387548195379936,
          0.0011916806593286185,
          0.0008531552140891955,
          0.00043607260116098993,
          0.002078600580459571,
          0.0016898407357718403,
          0.001072110742205919,
          0.000456215770326441,
          0.0003548882499517531,
          0.0023967705187182033,
          0.0023361934949426887,
          0.0008607498633644919
         ],
         "yaxis": "y"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          30,
          20,
          11,
          14,
          31,
          18,
          23,
          8,
          24,
          23,
          8,
          32,
          28,
          26,
          27,
          14
         ],
         "xaxis": "x2",
         "y": [
          0.00048017606419652393,
          0.0013894868760589503,
          0.0011698623783621739,
          0.0006450670262454189,
          0.0004387548195379936,
          0.0011916806593286185,
          0.0008531552140891955,
          0.00043607260116098993,
          0.002078600580459571,
          0.0016898407357718403,
          0.001072110742205919,
          0.000456215770326441,
          0.0003548882499517531,
          0.0023967705187182033,
          0.0023361934949426887,
          0.0008607498633644919
         ],
         "yaxis": "y2"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          0.0037347286149582125,
          0.002206639599309111,
          0.004227672658599115,
          0.003377067548621225,
          0.0011244264980527455,
          0.003329377747437302,
          0.0030071545557830153,
          0.001231174400482109,
          0.0026469178069660902,
          0.003122987425426211,
          0.0012914884095612668,
          0.0010899252712851255,
          0.0017920075740621118,
          0.0018417951021639603,
          0.0018183981182289845,
          0.0016600287562503746
         ],
         "xaxis": "x3",
         "y": [
          0.00048017606419652393,
          0.0013894868760589503,
          0.0011698623783621739,
          0.0006450670262454189,
          0.0004387548195379936,
          0.0011916806593286185,
          0.0008531552140891955,
          0.00043607260116098993,
          0.002078600580459571,
          0.0016898407357718403,
          0.001072110742205919,
          0.000456215770326441,
          0.0003548882499517531,
          0.0023967705187182033,
          0.0023361934949426887,
          0.0008607498633644919
         ],
         "yaxis": "y3"
        },
        {
         "marker": {
          "color": [
           0,
           1,
           2,
           3,
           4,
           5,
           6,
           7,
           8,
           9,
           10,
           11,
           12,
           13,
           14,
           15
          ],
          "colorbar": {
           "title": {
            "text": "#Trials"
           },
           "x": 1,
           "xpad": 40
          },
          "colorscale": [
           [
            0,
            "rgb(247,251,255)"
           ],
           [
            0.125,
            "rgb(222,235,247)"
           ],
           [
            0.25,
            "rgb(198,219,239)"
           ],
           [
            0.375,
            "rgb(158,202,225)"
           ],
           [
            0.5,
            "rgb(107,174,214)"
           ],
           [
            0.625,
            "rgb(66,146,198)"
           ],
           [
            0.75,
            "rgb(33,113,181)"
           ],
           [
            0.875,
            "rgb(8,81,156)"
           ],
           [
            1,
            "rgb(8,48,107)"
           ]
          ],
          "line": {
           "color": "Grey",
           "width": 0.5
          },
          "showscale": false
         },
         "mode": "markers",
         "showlegend": false,
         "type": "scatter",
         "x": [
          0.10553049163883402,
          0.13337702038120913,
          0.05890245612756484,
          0.07679359438819774,
          0.07698484217075284,
          0.10821253658754612,
          0.054742582731521654,
          0.14094330501950722,
          0.06063117870609974,
          0.06153782807391946,
          0.1484905195053996,
          0.0853726522904534,
          0.12802238957680404,
          0.12806204320697395,
          0.14930431145435258,
          0.1261890471167614
         ],
         "xaxis": "x4",
         "y": [
          0.00048017606419652393,
          0.0013894868760589503,
          0.0011698623783621739,
          0.0006450670262454189,
          0.0004387548195379936,
          0.0011916806593286185,
          0.0008531552140891955,
          0.00043607260116098993,
          0.002078600580459571,
          0.0016898407357718403,
          0.001072110742205919,
          0.000456215770326441,
          0.0003548882499517531,
          0.0023967705187182033,
          0.0023361934949426887,
          0.0008607498633644919
         ],
         "yaxis": "y4"
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Slice Plot"
        },
        "width": 1200,
        "xaxis": {
         "anchor": "y",
         "domain": [
          0,
          0.2125
         ],
         "title": {
          "text": "clip_margin"
         }
        },
        "xaxis2": {
         "anchor": "y2",
         "domain": [
          0.2625,
          0.475
         ],
         "title": {
          "text": "hidden_layer"
         }
        },
        "xaxis3": {
         "anchor": "y3",
         "domain": [
          0.525,
          0.7375
         ],
         "title": {
          "text": "learning_rate"
         }
        },
        "xaxis4": {
         "anchor": "y4",
         "domain": [
          0.7875,
          1
         ],
         "title": {
          "text": "stddev"
         }
        },
        "yaxis": {
         "anchor": "x",
         "domain": [
          0,
          1
         ],
         "title": {
          "text": "Objective Value"
         }
        },
        "yaxis2": {
         "anchor": "x2",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        },
        "yaxis3": {
         "anchor": "x3",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        },
        "yaxis4": {
         "anchor": "x4",
         "domain": [
          0,
          1
         ],
         "matches": "y",
         "showticklabels": false
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_slice(study_3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "MzAEfamSWJCF",
    "outputId": "0bf90f44-95b9-4ce3-c82b-3310182a0b7b"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "cliponaxis": false,
         "hovertemplate": [
          "clip_margin (IntUniformDistribution): 0.11294769660968174<extra></extra>",
          "stddev (UniformDistribution): 0.2500839642667676<extra></extra>",
          "learning_rate (UniformDistribution): 0.26709902059431845<extra></extra>",
          "hidden_layer (IntUniformDistribution): 0.3698693185292321<extra></extra>"
         ],
         "marker": {
          "color": [
           "rgb(8,81,156)",
           "rgb(8,48,107)",
           "rgb(8,48,107)",
           "rgb(8,81,156)"
          ]
         },
         "orientation": "h",
         "text": [
          "0.11294769660968174",
          "0.2500839642667676",
          "0.26709902059431845",
          "0.3698693185292321"
         ],
         "textposition": "outside",
         "texttemplate": "%{text:.2f}",
         "type": "bar",
         "x": [
          0.11294769660968174,
          0.2500839642667676,
          0.26709902059431845,
          0.3698693185292321
         ],
         "y": [
          "clip_margin",
          "stddev",
          "learning_rate",
          "hidden_layer"
         ]
        }
       ],
       "layout": {
        "showlegend": false,
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "title": {
         "text": "Hyperparameter Importances"
        },
        "xaxis": {
         "title": {
          "text": "Importance for Objective Value"
         }
        },
        "yaxis": {
         "title": {
          "text": "Hyperparameter"
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "optuna.visualization.plot_param_importances(study_3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "40l_RX8GWJCF"
   },
   "source": [
    "# Resultados\n",
    "\n",
    "Utilizamos para todos os testes o mesmo tamanho de batch size e window size por acreditarmos ser uma boa configuração padrão.\n",
    "Isto posto, alteramos apenas as quantidades de hidden layer, taxa de aprendizagem, desvio padrão e margem de clip.\n",
    "Podemos ver a partir dos gráficos que o primeiro estudo foi o que obteve melhor resultado.\n",
    "No segundo estudo, ao tentarmos otimizar os parâmetros acabamos por exagerar e a predição saiu muito errada para cima, bastante distante do objetivo.\n",
    "No terceiro estudo, tentamos reduzir o problema do segundo estudo, e acabamos por encontrar a situação inversa, nossa predição errou os valores para baixo, ainda assim, foi melhor que a do segundo estudo.\n",
    "Quanto a importância dos parâmetros, observados o do primeiro, vemos que o tamanho da camada oculta é a que mais impactou no desempenho de nossa rede LSTM; em segundo lugar, temos o desvio padrão como fator importante ao primeiro estudo e a taxa de aprendizagem em terceiro. De acordo com o primeiro estudo, a melhor configuração seria algo em torno de 100 neurônios, uma taxa de clipagem em torno de 20, com taxa de aprendizado de 5% e desvio padrão indiferente."
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "machine_shape": "hm",
   "name": "Projeto_Redes_Neurais_LSTM.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}