RNN-LSTM-Timeseries.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "5aa71f20",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import yfinance as yf\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "fae1fe4c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[*********************100%***********************]  1 of 1 completed\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>42.049999</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>38.230000</td>\n",
       "      <td>38.230000</td>\n",
       "      <td>573576400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>36.529999</td>\n",
       "      <td>36.660000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>34.029999</td>\n",
       "      <td>34.029999</td>\n",
       "      <td>168192700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>32.610001</td>\n",
       "      <td>33.590000</td>\n",
       "      <td>30.940001</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>101786600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>31.370001</td>\n",
       "      <td>32.500000</td>\n",
       "      <td>31.360001</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>73600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>32.950001</td>\n",
       "      <td>33.209999</td>\n",
       "      <td>31.770000</td>\n",
       "      <td>33.029999</td>\n",
       "      <td>33.029999</td>\n",
       "      <td>50237200</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Open       High        Low      Close  Adj Close     Volume\n",
       "0  42.049999  45.000000  38.000000  38.230000  38.230000  573576400\n",
       "1  36.529999  36.660000  33.000000  34.029999  34.029999  168192700\n",
       "2  32.610001  33.590000  30.940001  31.000000  31.000000  101786600\n",
       "3  31.370001  32.500000  31.360001  32.000000  32.000000   73600000\n",
       "4  32.950001  33.209999  31.770000  33.029999  33.029999   50237200"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stock_symbol = 'FB' # Stock Symbol Google / Alphabet\n",
    "# last 5 years data with interval 1 day\n",
    "data = yf.download(tickers=stock_symbol, period='10y', interval='1d')\n",
    "s = pd.Series(list(range(0, len(data))))\n",
    "data.set_index(s, inplace=True)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "6122fc53",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>42.049999</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>38.230000</td>\n",
       "      <td>573576400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>36.529999</td>\n",
       "      <td>36.660000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>34.029999</td>\n",
       "      <td>168192700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>32.610001</td>\n",
       "      <td>33.590000</td>\n",
       "      <td>30.940001</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>101786600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>31.370001</td>\n",
       "      <td>32.500000</td>\n",
       "      <td>31.360001</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>73600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>32.950001</td>\n",
       "      <td>33.209999</td>\n",
       "      <td>31.770000</td>\n",
       "      <td>33.029999</td>\n",
       "      <td>50237200</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Open       High        Low      Close     Volume\n",
       "0  42.049999  45.000000  38.000000  38.230000  573576400\n",
       "1  36.529999  36.660000  33.000000  34.029999  168192700\n",
       "2  32.610001  33.590000  30.940001  31.000000  101786600\n",
       "3  31.370001  32.500000  31.360001  32.000000   73600000\n",
       "4  32.950001  33.209999  31.770000  33.029999   50237200"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_training_size = int(len(data)*0.7)\n",
    "data_test_size = len(data) - data_training_size\n",
    "data_training, data_test = data.iloc[0:data_training_size,:], data.iloc[data_training_size:len(data),:]\n",
    "data_training = data_training.drop(['Adj Close'], axis=1)\n",
    "data_test = data_test.drop(['Adj Close'], axis=1)\n",
    "data_training.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "bb511d4a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.12128111, 0.13341652, 0.10395486, 0.10261801, 1.        ],\n",
       "       [0.09335154, 0.09178937, 0.07853803, 0.08159383, 0.28587298],\n",
       "       [0.07351751, 0.07646618, 0.06806629, 0.06642639, 0.16889149],\n",
       "       ...,\n",
       "       [0.63180528, 0.63104567, 0.6289142 , 0.63322819, 0.02870152],\n",
       "       [0.63246307, 0.6303469 , 0.62672831, 0.63307803, 0.01798936],\n",
       "       [0.63281721, 0.63433992, 0.63552259, 0.63107575, 0.01232227]])"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sc = MinMaxScaler()\n",
    "training_data = sc.fit_transform(data_training)\n",
    "training_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "8fa00278",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((1613, 60, 5), (1613,))"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "window = 60# check that window is less than shape[0]\n",
    "x_train = []\n",
    "y_train = []\n",
    "for i in range(window, training_data.shape[0]):\n",
    "    # take from current minus 60 first dimension, empty seconds dimension to grab all 5\n",
    "    x_train.append(training_data[i-window:i])\n",
    "    y_train.append(training_data[i, 0]) # zero second dimension only to grab which result should be\n",
    "x_train, y_train = np.array(x_train), np.array(y_train)\n",
    "x_train.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "f6efe67f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Layer lstm_12 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "WARNING:tensorflow:Layer lstm_13 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "WARNING:tensorflow:Layer lstm_14 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "WARNING:tensorflow:Layer lstm_15 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
      "Model: \"sequential_3\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_12 (LSTM)               (None, 60, 80)            27520     \n",
      "_________________________________________________________________\n",
      "dropout_12 (Dropout)         (None, 60, 80)            0         \n",
      "_________________________________________________________________\n",
      "lstm_13 (LSTM)               (None, 60, 80)            51520     \n",
      "_________________________________________________________________\n",
      "dropout_13 (Dropout)         (None, 60, 80)            0         \n",
      "_________________________________________________________________\n",
      "lstm_14 (LSTM)               (None, 60, 80)            51520     \n",
      "_________________________________________________________________\n",
      "dropout_14 (Dropout)         (None, 60, 80)            0         \n",
      "_________________________________________________________________\n",
      "lstm_15 (LSTM)               (None, 120)               96480     \n",
      "_________________________________________________________________\n",
      "dropout_15 (Dropout)         (None, 120)               0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 1)                 121       \n",
      "=================================================================\n",
      "Total params: 227,161\n",
      "Trainable params: 227,161\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = keras.models.Sequential([\n",
    "    # 50 nodes comming out, return sequences means returns the data as it enters\n",
    "    # input shape entering is window, result (60, 5) -> five data all the table\n",
    "    keras.layers.LSTM(units=80, activation='relu', return_sequences=True, input_shape=(x_train.shape[1], 5)),\n",
    "    # Drop 20% random -> regularization to avoid exploding gradient\n",
    "    keras.layers.Dropout(0.2),\n",
    "    keras.layers.LSTM(units=80, activation='relu', return_sequences=True),\n",
    "    keras.layers.Dropout(0.3),\n",
    "    keras.layers.LSTM(units=80, activation='relu', return_sequences=True),\n",
    "    keras.layers.Dropout(0.4),\n",
    "    # the last one never must have return sequences true, bc you dont want your input to be the output\n",
    "    # you want only a result\n",
    "    keras.layers.LSTM(units=120, activation='relu'),\n",
    "    keras.layers.Dropout(0.5),\n",
    "    # output layer, dense layer, return one result -> units\n",
    "    keras.layers.Dense(units=1)\n",
    "])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "7c709453",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-11-16 17:41:50.153218: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "51/51 [==============================] - 612s 12s/step - loss: 0.0051\n",
      "Epoch 2/10\n",
      "51/51 [==============================] - 610s 12s/step - loss: 0.0039\n",
      "Epoch 3/10\n",
      "51/51 [==============================] - 610s 12s/step - loss: 0.0036\n",
      "Epoch 4/10\n",
      "51/51 [==============================] - 614s 12s/step - loss: 0.0035\n",
      "Epoch 5/10\n",
      "51/51 [==============================] - 614s 12s/step - loss: 0.0035\n",
      "Epoch 6/10\n",
      "51/51 [==============================] - 615s 12s/step - loss: 0.0033\n",
      "Epoch 7/10\n",
      "51/51 [==============================] - 617s 12s/step - loss: 0.0034\n",
      "Epoch 8/10\n",
      "51/51 [==============================] - 615s 12s/step - loss: 0.0038\n",
      "Epoch 9/10\n",
      "51/51 [==============================] - 613s 12s/step - loss: 0.0031\n",
      "Epoch 10/10\n",
      "51/51 [==============================] - 613s 12s/step - loss: 0.0029\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x35ed43580>"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.compile(\n",
    "    optimizer='adam', #already does a batch aproach, re listen the podcast to remember later\n",
    "    # adam algorithm is stochastic gradient decent, using batch but different from SGD uses adam algorithm to optimize\n",
    "    loss = 'mean_squared_error'\n",
    ")\n",
    "model.fit(x_train, y_train, epochs=10, batch_size=32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "1ca0a5bb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>155.399994</td>\n",
       "      <td>159.460007</td>\n",
       "      <td>155.009995</td>\n",
       "      <td>158.779999</td>\n",
       "      <td>19180100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>159.559998</td>\n",
       "      <td>160.490005</td>\n",
       "      <td>157.949997</td>\n",
       "      <td>159.419998</td>\n",
       "      <td>17592000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>158.509995</td>\n",
       "      <td>158.660004</td>\n",
       "      <td>153.279999</td>\n",
       "      <td>154.919998</td>\n",
       "      <td>21675100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>155.860001</td>\n",
       "      <td>157.350006</td>\n",
       "      <td>153.550003</td>\n",
       "      <td>154.050003</td>\n",
       "      <td>19761300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>154.759995</td>\n",
       "      <td>157.339996</td>\n",
       "      <td>154.460007</td>\n",
       "      <td>154.779999</td>\n",
       "      <td>15424700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>773</th>\n",
       "      <td>332.489990</td>\n",
       "      <td>333.190002</td>\n",
       "      <td>325.510010</td>\n",
       "      <td>327.640015</td>\n",
       "      <td>21872600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>774</th>\n",
       "      <td>329.820007</td>\n",
       "      <td>332.459991</td>\n",
       "      <td>327.000000</td>\n",
       "      <td>327.739990</td>\n",
       "      <td>12376600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>775</th>\n",
       "      <td>330.179993</td>\n",
       "      <td>341.859985</td>\n",
       "      <td>329.779999</td>\n",
       "      <td>340.890015</td>\n",
       "      <td>25387200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>776</th>\n",
       "      <td>344.339996</td>\n",
       "      <td>353.649994</td>\n",
       "      <td>343.200012</td>\n",
       "      <td>347.559998</td>\n",
       "      <td>25059000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>777</th>\n",
       "      <td>343.829987</td>\n",
       "      <td>346.649994</td>\n",
       "      <td>340.869995</td>\n",
       "      <td>342.709991</td>\n",
       "      <td>9420380</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>778 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Open        High         Low       Close    Volume\n",
       "0    155.399994  159.460007  155.009995  158.779999  19180100\n",
       "1    159.559998  160.490005  157.949997  159.419998  17592000\n",
       "2    158.509995  158.660004  153.279999  154.919998  21675100\n",
       "3    155.860001  157.350006  153.550003  154.050003  19761300\n",
       "4    154.759995  157.339996  154.460007  154.779999  15424700\n",
       "..          ...         ...         ...         ...       ...\n",
       "773  332.489990  333.190002  325.510010  327.640015  21872600\n",
       "774  329.820007  332.459991  327.000000  327.739990  12376600\n",
       "775  330.179993  341.859985  329.779999  340.890015  25387200\n",
       "776  344.339996  353.649994  343.200012  347.559998  25059000\n",
       "777  343.829987  346.649994  340.869995  342.709991   9420380\n",
       "\n",
       "[778 rows x 5 columns]"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "past_window_days = data_training.tail(window)\n",
    "# get window size of previous data to continue from the training data to the test data\n",
    "# that way we put a link between test and time line data\n",
    "\n",
    "df = past_window_days.append(data_test, ignore_index = True)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "bc4f05ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.69479859, 0.7047168 , 0.69875962, 0.70606197, 0.02337125],\n",
       "       [0.71584698, 0.70985779, 0.71370473, 0.70926565, 0.02057364],\n",
       "       [0.71053427, 0.70072377, 0.68996541, 0.68673974, 0.02776646],\n",
       "       ...,\n",
       "       [1.57913373, 1.6151235 , 1.5871797 , 1.61766038, 0.03430572],\n",
       "       [1.65077917, 1.67397056, 1.65539856, 1.65104869, 0.03372756],\n",
       "       [1.64819867, 1.63903171, 1.64355423, 1.62677074, 0.00617845]])"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inputs = sc.transform(df) # we transfor the data into an array and then gather the window sizes with the prediction\n",
    "\n",
    "x_test = []\n",
    "y_test = []\n",
    "for i in range(window, inputs.shape[0]):\n",
    "    # take from current minus 60 first dimension, empty seconds dimension to grab all 5\n",
    "    x_test.append(inputs[i-window:i])\n",
    "    y_test.append(inputs[i, 0]) # zero second dimension only to grab which result should be\n",
    "x_test, y_test = np.array(x_test), np.array(y_test)\n",
    "x_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "23d91abc",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-11-16 19:24:04.017633: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0.5871101 ],\n",
       "       [0.5909943 ],\n",
       "       [0.5958951 ],\n",
       "       [0.60162157],\n",
       "       [0.6079391 ],\n",
       "       [0.61455154],\n",
       "       [0.6211817 ],\n",
       "       [0.6275158 ],\n",
       "       [0.6332419 ],\n",
       "       [0.63823915],\n",
       "       [0.6425197 ],\n",
       "       [0.6460747 ],\n",
       "       [0.6488051 ],\n",
       "       [0.6510615 ],\n",
       "       [0.6535679 ],\n",
       "       [0.6570111 ],\n",
       "       [0.66193676],\n",
       "       [0.6686669 ],\n",
       "       [0.6769757 ],\n",
       "       [0.6863483 ],\n",
       "       [0.69619536],\n",
       "       [0.70590466],\n",
       "       [0.71494496],\n",
       "       [0.7229098 ],\n",
       "       [0.729513  ],\n",
       "       [0.7346038 ],\n",
       "       [0.73822314],\n",
       "       [0.7404414 ],\n",
       "       [0.74141747],\n",
       "       [0.74144787],\n",
       "       [0.740903  ],\n",
       "       [0.74003834],\n",
       "       [0.73903227],\n",
       "       [0.73801833],\n",
       "       [0.7371396 ],\n",
       "       [0.7365517 ],\n",
       "       [0.73653454],\n",
       "       [0.7372127 ],\n",
       "       [0.7385455 ],\n",
       "       [0.7405146 ],\n",
       "       [0.74310285],\n",
       "       [0.7462558 ],\n",
       "       [0.7496841 ],\n",
       "       [0.7527633 ],\n",
       "       [0.75483936],\n",
       "       [0.75561076],\n",
       "       [0.7551538 ],\n",
       "       [0.75378436],\n",
       "       [0.7518696 ],\n",
       "       [0.7497371 ],\n",
       "       [0.74772865],\n",
       "       [0.7461081 ],\n",
       "       [0.7449809 ],\n",
       "       [0.74435586],\n",
       "       [0.74427754],\n",
       "       [0.7447765 ],\n",
       "       [0.7459019 ],\n",
       "       [0.7477885 ],\n",
       "       [0.75053996],\n",
       "       [0.75410575],\n",
       "       [0.75828344],\n",
       "       [0.76291186],\n",
       "       [0.767837  ],\n",
       "       [0.77288157],\n",
       "       [0.7778811 ],\n",
       "       [0.78268784],\n",
       "       [0.78719693],\n",
       "       [0.791273  ],\n",
       "       [0.7948572 ],\n",
       "       [0.79799163],\n",
       "       [0.80061084],\n",
       "       [0.80297273],\n",
       "       [0.8054954 ],\n",
       "       [0.80853945],\n",
       "       [0.81229275],\n",
       "       [0.81681997],\n",
       "       [0.82196623],\n",
       "       [0.8275565 ],\n",
       "       [0.83327484],\n",
       "       [0.83870536],\n",
       "       [0.8435047 ],\n",
       "       [0.8473869 ],\n",
       "       [0.8501976 ],\n",
       "       [0.8517049 ],\n",
       "       [0.8517578 ],\n",
       "       [0.85050565],\n",
       "       [0.8483339 ],\n",
       "       [0.8456509 ],\n",
       "       [0.84271306],\n",
       "       [0.83980805],\n",
       "       [0.83716905],\n",
       "       [0.8347898 ],\n",
       "       [0.8326556 ],\n",
       "       [0.8307335 ],\n",
       "       [0.8290116 ],\n",
       "       [0.82751185],\n",
       "       [0.8260587 ],\n",
       "       [0.8238038 ],\n",
       "       [0.82013184],\n",
       "       [0.81505257],\n",
       "       [0.80889785],\n",
       "       [0.8021751 ],\n",
       "       [0.79552084],\n",
       "       [0.7895685 ],\n",
       "       [0.784749  ],\n",
       "       [0.7813294 ],\n",
       "       [0.77948123],\n",
       "       [0.779316  ],\n",
       "       [0.78085893],\n",
       "       [0.7839604 ],\n",
       "       [0.78848284],\n",
       "       [0.79412013],\n",
       "       [0.80064034],\n",
       "       [0.8076711 ],\n",
       "       [0.81477886],\n",
       "       [0.82169944],\n",
       "       [0.82818395],\n",
       "       [0.8341455 ],\n",
       "       [0.83960074],\n",
       "       [0.8446271 ],\n",
       "       [0.8492747 ],\n",
       "       [0.85354716],\n",
       "       [0.8574451 ],\n",
       "       [0.86107016],\n",
       "       [0.8646013 ],\n",
       "       [0.86814123],\n",
       "       [0.87176   ],\n",
       "       [0.8755067 ],\n",
       "       [0.8793128 ],\n",
       "       [0.88297266],\n",
       "       [0.88626057],\n",
       "       [0.8890284 ],\n",
       "       [0.89126235],\n",
       "       [0.8927844 ],\n",
       "       [0.8935178 ],\n",
       "       [0.89350605],\n",
       "       [0.8928416 ],\n",
       "       [0.8916493 ],\n",
       "       [0.8900277 ],\n",
       "       [0.88800865],\n",
       "       [0.8855364 ],\n",
       "       [0.88229257],\n",
       "       [0.87817067],\n",
       "       [0.8732467 ],\n",
       "       [0.86784214],\n",
       "       [0.86239904],\n",
       "       [0.8572182 ],\n",
       "       [0.8525357 ],\n",
       "       [0.8482911 ],\n",
       "       [0.84438914],\n",
       "       [0.8408218 ],\n",
       "       [0.8377324 ],\n",
       "       [0.83517694],\n",
       "       [0.83317655],\n",
       "       [0.8316112 ],\n",
       "       [0.8301878 ],\n",
       "       [0.82873935],\n",
       "       [0.827228  ],\n",
       "       [0.8256766 ],\n",
       "       [0.82427055],\n",
       "       [0.82320917],\n",
       "       [0.82255924],\n",
       "       [0.8223794 ],\n",
       "       [0.8227924 ],\n",
       "       [0.8237962 ],\n",
       "       [0.825286  ],\n",
       "       [0.82705355],\n",
       "       [0.8289647 ],\n",
       "       [0.83095914],\n",
       "       [0.8329315 ],\n",
       "       [0.83478606],\n",
       "       [0.8364518 ],\n",
       "       [0.8379087 ],\n",
       "       [0.8392176 ],\n",
       "       [0.84039074],\n",
       "       [0.8413947 ],\n",
       "       [0.8420194 ],\n",
       "       [0.8419997 ],\n",
       "       [0.8411551 ],\n",
       "       [0.8394176 ],\n",
       "       [0.8368625 ],\n",
       "       [0.8335838 ],\n",
       "       [0.8296606 ],\n",
       "       [0.82529235],\n",
       "       [0.8208462 ],\n",
       "       [0.8166768 ],\n",
       "       [0.8130365 ],\n",
       "       [0.8100987 ],\n",
       "       [0.8079595 ],\n",
       "       [0.8067005 ],\n",
       "       [0.8063869 ],\n",
       "       [0.8070306 ],\n",
       "       [0.808662  ],\n",
       "       [0.81127894],\n",
       "       [0.81461865],\n",
       "       [0.81845695],\n",
       "       [0.8223632 ],\n",
       "       [0.82595825],\n",
       "       [0.8290115 ],\n",
       "       [0.831513  ],\n",
       "       [0.8335077 ],\n",
       "       [0.8351743 ],\n",
       "       [0.8364782 ],\n",
       "       [0.8374441 ],\n",
       "       [0.8382818 ],\n",
       "       [0.8393143 ],\n",
       "       [0.8408096 ],\n",
       "       [0.84280264],\n",
       "       [0.8451201 ],\n",
       "       [0.8475421 ],\n",
       "       [0.8498355 ],\n",
       "       [0.85184413],\n",
       "       [0.8536143 ],\n",
       "       [0.855198  ],\n",
       "       [0.8566734 ],\n",
       "       [0.85810643],\n",
       "       [0.85960364],\n",
       "       [0.86126536],\n",
       "       [0.8631359 ],\n",
       "       [0.86524194],\n",
       "       [0.8675604 ],\n",
       "       [0.8700338 ],\n",
       "       [0.8726371 ],\n",
       "       [0.8754057 ],\n",
       "       [0.8782343 ],\n",
       "       [0.88087493],\n",
       "       [0.883202  ],\n",
       "       [0.8851637 ],\n",
       "       [0.88677466],\n",
       "       [0.88810104],\n",
       "       [0.8892255 ],\n",
       "       [0.89023423],\n",
       "       [0.89095026],\n",
       "       [0.89112407],\n",
       "       [0.8907006 ],\n",
       "       [0.8898183 ],\n",
       "       [0.88868946],\n",
       "       [0.8876806 ],\n",
       "       [0.88704747],\n",
       "       [0.88708806],\n",
       "       [0.88798267],\n",
       "       [0.8897414 ],\n",
       "       [0.8922848 ],\n",
       "       [0.89535916],\n",
       "       [0.89863443],\n",
       "       [0.9019162 ],\n",
       "       [0.9050856 ],\n",
       "       [0.9080494 ],\n",
       "       [0.9108977 ],\n",
       "       [0.9137794 ],\n",
       "       [0.91691786],\n",
       "       [0.92047733],\n",
       "       [0.9245432 ],\n",
       "       [0.9290671 ],\n",
       "       [0.9339481 ],\n",
       "       [0.9390468 ],\n",
       "       [0.9441199 ],\n",
       "       [0.9489221 ],\n",
       "       [0.95330566],\n",
       "       [0.9571225 ],\n",
       "       [0.96022004],\n",
       "       [0.9623144 ],\n",
       "       [0.96337056],\n",
       "       [0.9634956 ],\n",
       "       [0.96228844],\n",
       "       [0.9595257 ],\n",
       "       [0.95533544],\n",
       "       [0.95012265],\n",
       "       [0.94452053],\n",
       "       [0.9391102 ],\n",
       "       [0.9343359 ],\n",
       "       [0.9305033 ],\n",
       "       [0.9275183 ],\n",
       "       [0.9253015 ],\n",
       "       [0.9238279 ],\n",
       "       [0.92317754],\n",
       "       [0.92338914],\n",
       "       [0.92451376],\n",
       "       [0.9264214 ],\n",
       "       [0.9287258 ],\n",
       "       [0.9305506 ],\n",
       "       [0.9312121 ],\n",
       "       [0.9303883 ],\n",
       "       [0.92779773],\n",
       "       [0.9231979 ],\n",
       "       [0.916886  ],\n",
       "       [0.90926063],\n",
       "       [0.90080804],\n",
       "       [0.8919434 ],\n",
       "       [0.88279694],\n",
       "       [0.8731417 ],\n",
       "       [0.8630083 ],\n",
       "       [0.8525273 ],\n",
       "       [0.8414058 ],\n",
       "       [0.8296799 ],\n",
       "       [0.8171574 ],\n",
       "       [0.803737  ],\n",
       "       [0.78935856],\n",
       "       [0.7743105 ],\n",
       "       [0.7591122 ],\n",
       "       [0.7442045 ],\n",
       "       [0.7302132 ],\n",
       "       [0.7176845 ],\n",
       "       [0.70713264],\n",
       "       [0.69883674],\n",
       "       [0.6930356 ],\n",
       "       [0.6898825 ],\n",
       "       [0.6891402 ],\n",
       "       [0.6902733 ],\n",
       "       [0.6924986 ],\n",
       "       [0.6953335 ],\n",
       "       [0.6986925 ],\n",
       "       [0.7027347 ],\n",
       "       [0.70764965],\n",
       "       [0.71345705],\n",
       "       [0.7202146 ],\n",
       "       [0.72775143],\n",
       "       [0.7357103 ],\n",
       "       [0.74382323],\n",
       "       [0.7518639 ],\n",
       "       [0.7593059 ],\n",
       "       [0.76595473],\n",
       "       [0.772012  ],\n",
       "       [0.77768475],\n",
       "       [0.7832674 ],\n",
       "       [0.788848  ],\n",
       "       [0.7944477 ],\n",
       "       [0.8004697 ],\n",
       "       [0.8072534 ],\n",
       "       [0.8150143 ],\n",
       "       [0.823874  ],\n",
       "       [0.83375686],\n",
       "       [0.84452087],\n",
       "       [0.8559373 ],\n",
       "       [0.86764586],\n",
       "       [0.8791811 ],\n",
       "       [0.88979506],\n",
       "       [0.8988566 ],\n",
       "       [0.90610486],\n",
       "       [0.9117137 ],\n",
       "       [0.9160024 ],\n",
       "       [0.9195699 ],\n",
       "       [0.9231748 ],\n",
       "       [0.92744774],\n",
       "       [0.9327904 ],\n",
       "       [0.93893653],\n",
       "       [0.94543463],\n",
       "       [0.9518227 ],\n",
       "       [0.95787996],\n",
       "       [0.9635382 ],\n",
       "       [0.9688266 ],\n",
       "       [0.9736238 ],\n",
       "       [0.9778302 ],\n",
       "       [0.98144376],\n",
       "       [0.9845816 ],\n",
       "       [0.98756903],\n",
       "       [0.99016285],\n",
       "       [0.99216425],\n",
       "       [0.99357545],\n",
       "       [0.99470073],\n",
       "       [0.9957618 ],\n",
       "       [0.99691445],\n",
       "       [0.9982303 ],\n",
       "       [0.99985296],\n",
       "       [1.0019227 ],\n",
       "       [1.0043198 ],\n",
       "       [1.0067681 ],\n",
       "       [1.0081083 ],\n",
       "       [1.0073681 ],\n",
       "       [1.0045753 ],\n",
       "       [1.0004046 ],\n",
       "       [0.9958374 ],\n",
       "       [0.9917745 ],\n",
       "       [0.98895556],\n",
       "       [0.9877993 ],\n",
       "       [0.98857635],\n",
       "       [0.9912013 ],\n",
       "       [0.9953334 ],\n",
       "       [1.0002421 ],\n",
       "       [1.0054218 ],\n",
       "       [1.0104141 ],\n",
       "       [1.0149615 ],\n",
       "       [1.0189108 ],\n",
       "       [1.0222895 ],\n",
       "       [1.0250946 ],\n",
       "       [1.0271345 ],\n",
       "       [1.0280993 ],\n",
       "       [1.0279567 ],\n",
       "       [1.0268164 ],\n",
       "       [1.0248789 ],\n",
       "       [1.0223247 ],\n",
       "       [1.0197908 ],\n",
       "       [1.018101  ],\n",
       "       [1.0177858 ],\n",
       "       [1.0191047 ],\n",
       "       [1.0219793 ],\n",
       "       [1.026309  ],\n",
       "       [1.0321592 ],\n",
       "       [1.0391166 ],\n",
       "       [1.0467201 ],\n",
       "       [1.0545884 ],\n",
       "       [1.0623299 ],\n",
       "       [1.069623  ],\n",
       "       [1.0761693 ],\n",
       "       [1.0818121 ],\n",
       "       [1.0865892 ],\n",
       "       [1.09072   ],\n",
       "       [1.0944724 ],\n",
       "       [1.0980171 ],\n",
       "       [1.1018271 ],\n",
       "       [1.10663   ],\n",
       "       [1.11292   ],\n",
       "       [1.120698  ],\n",
       "       [1.1296798 ],\n",
       "       [1.13939   ],\n",
       "       [1.1488247 ],\n",
       "       [1.1568716 ],\n",
       "       [1.1625667 ],\n",
       "       [1.1655282 ],\n",
       "       [1.1657665 ],\n",
       "       [1.1635039 ],\n",
       "       [1.1592045 ],\n",
       "       [1.153574  ],\n",
       "       [1.1470556 ],\n",
       "       [1.13969   ],\n",
       "       [1.1315458 ],\n",
       "       [1.1226456 ],\n",
       "       [1.113229  ],\n",
       "       [1.1037586 ],\n",
       "       [1.094593  ],\n",
       "       [1.0861158 ],\n",
       "       [1.0788416 ],\n",
       "       [1.0732334 ],\n",
       "       [1.0696023 ],\n",
       "       [1.0681288 ],\n",
       "       [1.0686163 ],\n",
       "       [1.0707442 ],\n",
       "       [1.0739652 ],\n",
       "       [1.0775601 ],\n",
       "       [1.0811402 ],\n",
       "       [1.0845754 ],\n",
       "       [1.087997  ],\n",
       "       [1.0917737 ],\n",
       "       [1.0960891 ],\n",
       "       [1.1006025 ],\n",
       "       [1.1048943 ],\n",
       "       [1.1085315 ],\n",
       "       [1.1112512 ],\n",
       "       [1.1133616 ],\n",
       "       [1.1153569 ],\n",
       "       [1.1176558 ],\n",
       "       [1.1203697 ],\n",
       "       [1.1235081 ],\n",
       "       [1.1266214 ],\n",
       "       [1.1293776 ],\n",
       "       [1.1311078 ],\n",
       "       [1.1313775 ],\n",
       "       [1.1302248 ],\n",
       "       [1.1283848 ],\n",
       "       [1.1270043 ],\n",
       "       [1.1269872 ],\n",
       "       [1.1284273 ],\n",
       "       [1.1304957 ],\n",
       "       [1.1326267 ],\n",
       "       [1.1345625 ],\n",
       "       [1.1361804 ],\n",
       "       [1.1374424 ],\n",
       "       [1.1382978 ],\n",
       "       [1.1387026 ],\n",
       "       [1.1385759 ],\n",
       "       [1.1378347 ],\n",
       "       [1.136382  ],\n",
       "       [1.1344041 ],\n",
       "       [1.1323342 ],\n",
       "       [1.130624  ],\n",
       "       [1.1294193 ],\n",
       "       [1.1289697 ],\n",
       "       [1.1295632 ],\n",
       "       [1.1312515 ],\n",
       "       [1.1337897 ],\n",
       "       [1.1368738 ],\n",
       "       [1.1402698 ],\n",
       "       [1.1434715 ],\n",
       "       [1.145964  ],\n",
       "       [1.147427  ],\n",
       "       [1.1477618 ],\n",
       "       [1.1469494 ],\n",
       "       [1.1452515 ],\n",
       "       [1.1430508 ],\n",
       "       [1.1405678 ],\n",
       "       [1.137949  ],\n",
       "       [1.1351982 ],\n",
       "       [1.1323465 ],\n",
       "       [1.1295236 ],\n",
       "       [1.1268402 ],\n",
       "       [1.1246579 ],\n",
       "       [1.1232269 ],\n",
       "       [1.1225519 ],\n",
       "       [1.1224155 ],\n",
       "       [1.1226025 ],\n",
       "       [1.1225665 ],\n",
       "       [1.1221057 ],\n",
       "       [1.1211767 ],\n",
       "       [1.119468  ],\n",
       "       [1.1166822 ],\n",
       "       [1.1127286 ],\n",
       "       [1.1075766 ],\n",
       "       [1.101445  ],\n",
       "       [1.0949153 ],\n",
       "       [1.0888968 ],\n",
       "       [1.0842695 ],\n",
       "       [1.0816633 ],\n",
       "       [1.0814168 ],\n",
       "       [1.0836698 ],\n",
       "       [1.0879041 ],\n",
       "       [1.0932751 ],\n",
       "       [1.0985861 ],\n",
       "       [1.1029679 ],\n",
       "       [1.1061891 ],\n",
       "       [1.108353  ],\n",
       "       [1.1096436 ],\n",
       "       [1.1103379 ],\n",
       "       [1.1106743 ],\n",
       "       [1.1108649 ],\n",
       "       [1.1111412 ],\n",
       "       [1.1116657 ],\n",
       "       [1.1124941 ],\n",
       "       [1.1136351 ],\n",
       "       [1.1150376 ],\n",
       "       [1.1164789 ],\n",
       "       [1.1175159 ],\n",
       "       [1.1177275 ],\n",
       "       [1.1169233 ],\n",
       "       [1.1152234 ],\n",
       "       [1.1127405 ],\n",
       "       [1.1095768 ],\n",
       "       [1.1060477 ],\n",
       "       [1.1025574 ],\n",
       "       [1.0992795 ],\n",
       "       [1.0961744 ],\n",
       "       [1.0933355 ],\n",
       "       [1.0909114 ],\n",
       "       [1.0890999 ],\n",
       "       [1.0881752 ],\n",
       "       [1.0884165 ],\n",
       "       [1.089762  ],\n",
       "       [1.0921528 ],\n",
       "       [1.0956097 ],\n",
       "       [1.1001328 ],\n",
       "       [1.1055727 ],\n",
       "       [1.1116575 ],\n",
       "       [1.1184039 ],\n",
       "       [1.1258003 ],\n",
       "       [1.1334248 ],\n",
       "       [1.1405234 ],\n",
       "       [1.1465354 ],\n",
       "       [1.1514084 ],\n",
       "       [1.155316  ],\n",
       "       [1.1585441 ],\n",
       "       [1.1615833 ],\n",
       "       [1.164919  ],\n",
       "       [1.1690114 ],\n",
       "       [1.1740321 ],\n",
       "       [1.1799939 ],\n",
       "       [1.1867028 ],\n",
       "       [1.19376   ],\n",
       "       [1.2007027 ],\n",
       "       [1.2068595 ],\n",
       "       [1.2118304 ],\n",
       "       [1.215518  ],\n",
       "       [1.217843  ],\n",
       "       [1.2187805 ],\n",
       "       [1.218413  ],\n",
       "       [1.2168857 ],\n",
       "       [1.2144396 ],\n",
       "       [1.2114959 ],\n",
       "       [1.2084858 ],\n",
       "       [1.2057139 ],\n",
       "       [1.203814  ],\n",
       "       [1.2035705 ],\n",
       "       [1.2054341 ],\n",
       "       [1.2090333 ],\n",
       "       [1.2137425 ],\n",
       "       [1.2188851 ],\n",
       "       [1.2240278 ],\n",
       "       [1.228447  ],\n",
       "       [1.2314804 ],\n",
       "       [1.2327111 ],\n",
       "       [1.2321837 ],\n",
       "       [1.2304473 ],\n",
       "       [1.2281842 ],\n",
       "       [1.2259212 ],\n",
       "       [1.22385   ],\n",
       "       [1.2222267 ],\n",
       "       [1.2213697 ],\n",
       "       [1.2215269 ],\n",
       "       [1.2229589 ],\n",
       "       [1.2257912 ],\n",
       "       [1.229858  ],\n",
       "       [1.2348685 ],\n",
       "       [1.240398  ],\n",
       "       [1.2459986 ],\n",
       "       [1.251174  ],\n",
       "       [1.2555944 ],\n",
       "       [1.2592071 ],\n",
       "       [1.2621739 ],\n",
       "       [1.2646244 ],\n",
       "       [1.266577  ],\n",
       "       [1.2680328 ],\n",
       "       [1.269088  ],\n",
       "       [1.2699342 ],\n",
       "       [1.2706172 ],\n",
       "       [1.271109  ],\n",
       "       [1.2713299 ],\n",
       "       [1.2712659 ],\n",
       "       [1.2710818 ],\n",
       "       [1.2710822 ],\n",
       "       [1.27162   ],\n",
       "       [1.2728126 ],\n",
       "       [1.2746834 ],\n",
       "       [1.27736   ],\n",
       "       [1.2808099 ],\n",
       "       [1.2847905 ],\n",
       "       [1.2891523 ],\n",
       "       [1.2937102 ],\n",
       "       [1.2981691 ],\n",
       "       [1.3020759 ],\n",
       "       [1.3050919 ],\n",
       "       [1.3072553 ],\n",
       "       [1.3087368 ],\n",
       "       [1.3096278 ],\n",
       "       [1.3098873 ],\n",
       "       [1.3094217 ],\n",
       "       [1.3081055 ],\n",
       "       [1.3059242 ],\n",
       "       [1.3032018 ],\n",
       "       [1.3004253 ],\n",
       "       [1.2982975 ],\n",
       "       [1.2977831 ],\n",
       "       [1.2993591 ],\n",
       "       [1.3028952 ],\n",
       "       [1.3076184 ],\n",
       "       [1.3125658 ],\n",
       "       [1.3170428 ],\n",
       "       [1.3205221 ],\n",
       "       [1.3228085 ],\n",
       "       [1.3241283 ],\n",
       "       [1.324878  ],\n",
       "       [1.3253871 ],\n",
       "       [1.3258324 ],\n",
       "       [1.3262434 ],\n",
       "       [1.3266077 ],\n",
       "       [1.32697   ],\n",
       "       [1.327369  ],\n",
       "       [1.3277109 ],\n",
       "       [1.3278449 ],\n",
       "       [1.3276008 ],\n",
       "       [1.32693   ],\n",
       "       [1.3260512 ],\n",
       "       [1.3252923 ],\n",
       "       [1.3249893 ],\n",
       "       [1.3253162 ],\n",
       "       [1.3262951 ],\n",
       "       [1.3280237 ],\n",
       "       [1.330658  ],\n",
       "       [1.3342044 ],\n",
       "       [1.3383551 ],\n",
       "       [1.342675  ],\n",
       "       [1.3468384 ],\n",
       "       [1.350628  ],\n",
       "       [1.3539056 ],\n",
       "       [1.3566401 ],\n",
       "       [1.3588169 ],\n",
       "       [1.3604121 ],\n",
       "       [1.361361  ],\n",
       "       [1.3615817 ],\n",
       "       [1.3609712 ],\n",
       "       [1.3592608 ],\n",
       "       [1.3563738 ],\n",
       "       [1.3522238 ],\n",
       "       [1.3466454 ],\n",
       "       [1.3401219 ],\n",
       "       [1.3331435 ],\n",
       "       [1.3261127 ],\n",
       "       [1.3192006 ],\n",
       "       [1.3126087 ],\n",
       "       [1.3065212 ],\n",
       "       [1.30075   ],\n",
       "       [1.2949319 ],\n",
       "       [1.2891119 ],\n",
       "       [1.2835251 ],\n",
       "       [1.2784756 ],\n",
       "       [1.2741015 ],\n",
       "       [1.2701523 ],\n",
       "       [1.2665453 ],\n",
       "       [1.2634988 ],\n",
       "       [1.261056  ],\n",
       "       [1.2593112 ],\n",
       "       [1.2586334 ],\n",
       "       [1.259356  ],\n",
       "       [1.2614601 ],\n",
       "       [1.264058  ],\n",
       "       [1.2661377 ],\n",
       "       [1.2668004 ],\n",
       "       [1.2657273 ],\n",
       "       [1.2628989 ],\n",
       "       [1.2587681 ],\n",
       "       [1.2542499 ],\n",
       "       [1.2501758 ],\n",
       "       [1.2471278 ],\n",
       "       [1.2456052 ],\n",
       "       [1.2459803 ],\n",
       "       [1.2483546 ],\n",
       "       [1.2523637 ],\n",
       "       [1.2571111 ],\n",
       "       [1.2617874 ],\n",
       "       [1.2659397 ],\n",
       "       [1.2696916 ]], dtype=float32)"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = model.predict(x_test)\n",
    "\n",
    "# un scale the values to normal amount of google shares\n",
    "scale = (1/sc.scale_[0]) # scaling label -> amount that use scaler to scale\n",
    "# every single value of scale_ is the factor that was use for every single value in the scaler\n",
    "# sc.scale_[0] is the scaler factor\n",
    "# we do the 1/scale to do the reverse and have the original numbers\n",
    "sc.scale_, sc.scale_[0], scale\n",
    "\n",
    "y_pred_final = y_pred*scale\n",
    "y_test_final = y_test*scale\n",
    "# normal scales"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "176f89d8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(14,5))\n",
    "plt.plot(y_test, color='red', label='Real FB Stock Price')\n",
    "plt.plot(y_pred, color='blue', label='Predicted FB Stock Price')\n",
    "plt.title('FB Stock Price Prediction')\n",
    "plt.xlabel('Time')\n",
    "plt.ylabel('FB Stock Price')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "7d8df1fb",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-11-17 03:13:48.768596: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "23/23 [==============================] - 15s 588ms/step - loss: 0.0375\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.037505172193050385"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f6c32e54",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9 (tensorflow)",
   "language": "python",
   "name": "tensorflow"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}