Bitcoin price prediction (Time series) using LSTM RNN.ipynb

{"cells":[{"cell_type":"markdown","metadata":{"id":"ohIvUi11epp3"},"source":["## Load libraries"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"0vNsMjuKepp8","outputId":"f75c7c42-5230-4f07-983e-c850039225d4"},"outputs":[{"name":"stderr","output_type":"stream","text":["Using TensorFlow backend.\n"]}],"source":["import numpy as np\n","import pandas as pd\n","import matplotlib.pyplot as plt\n","import math\n","\n","from sklearn.preprocessing import MinMaxScaler\n","from sklearn.metrics import mean_squared_error\n","\n","from keras.models import Sequential\n","from keras.layers import Dense\n","from keras.layers import LSTM"]},{"cell_type":"markdown","metadata":{"id":"4PaDPdpMepp-"},"source":["## Load the dataset"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"Q6MS19_Aepp-"},"outputs":[],"source":["url = \"https://raw.githubusercontent.com/LakshyaSatpal/bitcoin-prediction/master/used_data.csv\"\n","data = pd.read_csv(url)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"4UxQK6PVepp_","outputId":"85d02d57-699e-4d8c-8a56-2188bd6804bb"},"outputs":[{"data":{"text/html":["<div>\n","<style>\n","    .dataframe thead tr:only-child th {\n","        text-align: right;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: left;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>date_id</th>\n","      <th>datetime_id</th>\n","      <th>market</th>\n","      <th>rpt_key</th>\n","      <th>last</th>\n","      <th>diff_24h</th>\n","      <th>diff_per_24h</th>\n","      <th>bid</th>\n","      <th>ask</th>\n","      <th>low</th>\n","      <th>high</th>\n","      <th>volume</th>\n","      <th>created_at</th>\n","      <th>updated_at</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>2017-05-31</td>\n","      <td>2017-06-01 00:00:00</td>\n","      <td>bitstamp</td>\n","      <td>btc_eur</td>\n","      <td>1996.72</td>\n","      <td>2029.99</td>\n","      <td>-1.638924</td>\n","      <td>2005.50</td>\n","      <td>2005.56</td>\n","      <td>1950.00</td>\n","      <td>2063.73</td>\n","      <td>2314.500750</td>\n","      <td>2017-05-31 14:59:36</td>\n","      <td>2017-05-31 14:59:36</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>2017-05-31</td>\n","      <td>2017-06-01 00:00:00</td>\n","      <td>bitflyer</td>\n","      <td>btc_jpy</td>\n","      <td>267098.00</td>\n","      <td>269649.00</td>\n","      <td>-0.946045</td>\n","      <td>267124.00</td>\n","      <td>267267.00</td>\n","      <td>267124.00</td>\n","      <td>267267.00</td>\n","      <td>70922.880112</td>\n","      <td>2017-05-31 14:59:36</td>\n","      <td>2017-05-31 14:59:36</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>2017-05-31</td>\n","      <td>2017-06-01 00:00:00</td>\n","      <td>korbit</td>\n","      <td>btc_krw</td>\n","      <td>3003500.00</td>\n","      <td>3140000.00</td>\n","      <td>-4.347134</td>\n","      <td>3003500.00</td>\n","      <td>3004000.00</td>\n","      <td>3002000.00</td>\n","      <td>3209500.00</td>\n","      <td>6109.752872</td>\n","      <td>2017-05-31 14:59:36</td>\n","      <td>2017-05-31 14:59:36</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>2017-05-31</td>\n","      <td>2017-06-01 00:00:00</td>\n","      <td>bitstamp</td>\n","      <td>btc_usd</td>\n","      <td>2237.40</td>\n","      <td>2239.37</td>\n","      <td>-0.087971</td>\n","      <td>2233.09</td>\n","      <td>2237.40</td>\n","      <td>2154.28</td>\n","      <td>2293.46</td>\n","      <td>13681.282017</td>\n","      <td>2017-05-31 14:59:36</td>\n","      <td>2017-05-31 14:59:36</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>2017-05-31</td>\n","      <td>2017-06-01 00:00:00</td>\n","      <td>okcoin</td>\n","      <td>btc_usd</td>\n","      <td>2318.82</td>\n","      <td>2228.70</td>\n","      <td>4.043613</td>\n","      <td>2319.40</td>\n","      <td>2319.99</td>\n","      <td>2129.78</td>\n","      <td>2318.82</td>\n","      <td>4241.641516</td>\n","      <td>2017-05-31 14:59:36</td>\n","      <td>2017-05-31 14:59:36</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["      date_id          datetime_id    market  rpt_key        last    diff_24h  \\\n","0  2017-05-31  2017-06-01 00:00:00  bitstamp  btc_eur     1996.72     2029.99   \n","1  2017-05-31  2017-06-01 00:00:00  bitflyer  btc_jpy   267098.00   269649.00   \n","2  2017-05-31  2017-06-01 00:00:00    korbit  btc_krw  3003500.00  3140000.00   \n","3  2017-05-31  2017-06-01 00:00:00  bitstamp  btc_usd     2237.40     2239.37   \n","4  2017-05-31  2017-06-01 00:00:00    okcoin  btc_usd     2318.82     2228.70   \n","\n","   diff_per_24h         bid         ask         low        high        volume  \\\n","0     -1.638924     2005.50     2005.56     1950.00     2063.73   2314.500750   \n","1     -0.946045   267124.00   267267.00   267124.00   267267.00  70922.880112   \n","2     -4.347134  3003500.00  3004000.00  3002000.00  3209500.00   6109.752872   \n","3     -0.087971     2233.09     2237.40     2154.28     2293.46  13681.282017   \n","4      4.043613     2319.40     2319.99     2129.78     2318.82   4241.641516   \n","\n","            created_at           updated_at  \n","0  2017-05-31 14:59:36  2017-05-31 14:59:36  \n","1  2017-05-31 14:59:36  2017-05-31 14:59:36  \n","2  2017-05-31 14:59:36  2017-05-31 14:59:36  \n","3  2017-05-31 14:59:36  2017-05-31 14:59:36  \n","4  2017-05-31 14:59:36  2017-05-31 14:59:36  "]},"execution_count":3,"metadata":{},"output_type":"execute_result"}],"source":["data.head()"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"0oNfdKDPepp_","outputId":"51f68ede-0c5b-49d8-ebce-21520e1cf111"},"outputs":[{"data":{"text/plain":["btc_usd       125438\n","fx_btc_jpy     62719\n","eth_krw        62719\n","etc_krw        62719\n","ltc_usd        62719\n","btc_jpy        62719\n","eth_btc        62719\n","btc_eur        62719\n","btc_krw        62719\n","Name: rpt_key, dtype: int64"]},"execution_count":4,"metadata":{},"output_type":"execute_result"}],"source":["data['rpt_key'].value_counts()"]},{"cell_type":"markdown","metadata":{"id":"NCieoJCOepp_"},"source":["## Subset USD"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"S5qUcMqOepp_"},"outputs":[],"source":["df = data.loc[(data['rpt_key'] == 'btc_usd')]"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"ZdViEdTMepqA","outputId":"da555fcc-aa1d-494c-f67e-2a22637aeb43","scrolled":true},"outputs":[{"data":{"text/html":["<div>\n","<style>\n","    .dataframe thead tr:only-child th {\n","        text-align: right;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: left;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>date_id</th>\n","      <th>datetime_id</th>\n","      <th>market</th>\n","      <th>rpt_key</th>\n","      <th>last</th>\n","      <th>diff_24h</th>\n","      <th>diff_per_24h</th>\n","      <th>bid</th>\n","      <th>ask</th>\n","      <th>low</th>\n","      <th>high</th>\n","      <th>volume</th>\n","      <th>created_at</th>\n","      <th>updated_at</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>3</th>\n","      <td>2017-05-31</td>\n","      <td>2017-06-01 00:00:00</td>\n","      <td>bitstamp</td>\n","      <td>btc_usd</td>\n","      <td>2237.40</td>\n","      <td>2239.37</td>\n","      <td>-0.087971</td>\n","      <td>2233.09</td>\n","      <td>2237.40</td>\n","      <td>2154.28</td>\n","      <td>2293.46</td>\n","      <td>13681.282017</td>\n","      <td>2017-05-31 14:59:36</td>\n","      <td>2017-05-31 14:59:36</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>2017-05-31</td>\n","      <td>2017-06-01 00:00:00</td>\n","      <td>okcoin</td>\n","      <td>btc_usd</td>\n","      <td>2318.82</td>\n","      <td>2228.70</td>\n","      <td>4.043613</td>\n","      <td>2319.40</td>\n","      <td>2319.99</td>\n","      <td>2129.78</td>\n","      <td>2318.82</td>\n","      <td>4241.641516</td>\n","      <td>2017-05-31 14:59:36</td>\n","      <td>2017-05-31 14:59:36</td>\n","    </tr>\n","    <tr>\n","      <th>15</th>\n","      <td>2017-06-01</td>\n","      <td>2017-06-01 00:01:00</td>\n","      <td>bitstamp</td>\n","      <td>btc_usd</td>\n","      <td>2248.39</td>\n","      <td>2242.44</td>\n","      <td>0.265336</td>\n","      <td>2247.77</td>\n","      <td>2248.38</td>\n","      <td>2154.28</td>\n","      <td>2293.46</td>\n","      <td>13701.698603</td>\n","      <td>2017-05-31 15:00:36</td>\n","      <td>2017-05-31 15:00:36</td>\n","    </tr>\n","    <tr>\n","      <th>16</th>\n","      <td>2017-06-01</td>\n","      <td>2017-06-01 00:01:00</td>\n","      <td>okcoin</td>\n","      <td>btc_usd</td>\n","      <td>2320.42</td>\n","      <td>2228.40</td>\n","      <td>4.129420</td>\n","      <td>2320.99</td>\n","      <td>2321.49</td>\n","      <td>2129.78</td>\n","      <td>2322.00</td>\n","      <td>4260.261516</td>\n","      <td>2017-05-31 15:00:36</td>\n","      <td>2017-05-31 15:00:36</td>\n","    </tr>\n","    <tr>\n","      <th>23</th>\n","      <td>2017-06-01</td>\n","      <td>2017-06-01 00:02:00</td>\n","      <td>bitstamp</td>\n","      <td>btc_usd</td>\n","      <td>2248.35</td>\n","      <td>2238.58</td>\n","      <td>0.436437</td>\n","      <td>2248.35</td>\n","      <td>2248.69</td>\n","      <td>2154.28</td>\n","      <td>2293.46</td>\n","      <td>13742.110913</td>\n","      <td>2017-05-31 15:01:36</td>\n","      <td>2017-05-31 15:01:36</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["       date_id          datetime_id    market  rpt_key     last  diff_24h  \\\n","3   2017-05-31  2017-06-01 00:00:00  bitstamp  btc_usd  2237.40   2239.37   \n","4   2017-05-31  2017-06-01 00:00:00    okcoin  btc_usd  2318.82   2228.70   \n","15  2017-06-01  2017-06-01 00:01:00  bitstamp  btc_usd  2248.39   2242.44   \n","16  2017-06-01  2017-06-01 00:01:00    okcoin  btc_usd  2320.42   2228.40   \n","23  2017-06-01  2017-06-01 00:02:00  bitstamp  btc_usd  2248.35   2238.58   \n","\n","    diff_per_24h      bid      ask      low     high        volume  \\\n","3      -0.087971  2233.09  2237.40  2154.28  2293.46  13681.282017   \n","4       4.043613  2319.40  2319.99  2129.78  2318.82   4241.641516   \n","15      0.265336  2247.77  2248.38  2154.28  2293.46  13701.698603   \n","16      4.129420  2320.99  2321.49  2129.78  2322.00   4260.261516   \n","23      0.436437  2248.35  2248.69  2154.28  2293.46  13742.110913   \n","\n","             created_at           updated_at  \n","3   2017-05-31 14:59:36  2017-05-31 14:59:36  \n","4   2017-05-31 14:59:36  2017-05-31 14:59:36  \n","15  2017-05-31 15:00:36  2017-05-31 15:00:36  \n","16  2017-05-31 15:00:36  2017-05-31 15:00:36  \n","23  2017-05-31 15:01:36  2017-05-31 15:01:36  "]},"execution_count":6,"metadata":{},"output_type":"execute_result"}],"source":["df.head()"]},{"cell_type":"markdown","metadata":{"id":"3yJPI8VbepqA"},"source":["### Convert datetime_id to data type and filter dates greater than  2017-06-28 "]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"7dwDNc_pepqA"},"outputs":[],"source":["df = df.reset_index(drop=True)\n","df['datetime'] = pd.to_datetime(df['datetime_id'])\n","df = df.loc[df['datetime'] > pd.to_datetime('2017-06-28 00:00:00')]"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"jPkJzGD7epqA"},"outputs":[],"source":["df = df[['datetime', 'last', 'diff_24h', 'diff_per_24h', 'bid', 'ask', 'low', 'high', 'volume']]"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"SorYN3h3epqB","outputId":"31f2b3ba-ae1d-4cf9-e283-b56e0d01cd1c","scrolled":true},"outputs":[{"data":{"text/html":["<div>\n","<style>\n","    .dataframe thead tr:only-child th {\n","        text-align: right;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: left;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>datetime</th>\n","      <th>last</th>\n","      <th>diff_24h</th>\n","      <th>diff_per_24h</th>\n","      <th>bid</th>\n","      <th>ask</th>\n","      <th>low</th>\n","      <th>high</th>\n","      <th>volume</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>77762</th>\n","      <td>2017-06-28 00:01:00</td>\n","      <td>2344.00</td>\n","      <td>2491.98</td>\n","      <td>-5.938250</td>\n","      <td>2335.01</td>\n","      <td>2343.89</td>\n","      <td>2307.0</td>\n","      <td>2473.19</td>\n","      <td>20719.583592</td>\n","    </tr>\n","    <tr>\n","      <th>77763</th>\n","      <td>2017-06-28 00:01:00</td>\n","      <td>2499.39</td>\n","      <td>2682.25</td>\n","      <td>-6.817411</td>\n","      <td>2495.00</td>\n","      <td>2499.33</td>\n","      <td>2444.0</td>\n","      <td>2780.62</td>\n","      <td>2265.557866</td>\n","    </tr>\n","    <tr>\n","      <th>77764</th>\n","      <td>2017-06-28 00:02:00</td>\n","      <td>2337.18</td>\n","      <td>2491.98</td>\n","      <td>-6.211928</td>\n","      <td>2337.18</td>\n","      <td>2340.00</td>\n","      <td>2307.0</td>\n","      <td>2473.19</td>\n","      <td>20732.082581</td>\n","    </tr>\n","    <tr>\n","      <th>77765</th>\n","      <td>2017-06-28 00:02:00</td>\n","      <td>2492.76</td>\n","      <td>2682.25</td>\n","      <td>-7.064591</td>\n","      <td>2492.76</td>\n","      <td>2495.00</td>\n","      <td>2444.0</td>\n","      <td>2780.62</td>\n","      <td>2262.618866</td>\n","    </tr>\n","    <tr>\n","      <th>77766</th>\n","      <td>2017-06-28 00:03:00</td>\n","      <td>2335.02</td>\n","      <td>2491.98</td>\n","      <td>-6.298606</td>\n","      <td>2335.01</td>\n","      <td>2335.02</td>\n","      <td>2307.0</td>\n","      <td>2473.19</td>\n","      <td>20665.357191</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                 datetime     last  diff_24h  diff_per_24h      bid      ask  \\\n","77762 2017-06-28 00:01:00  2344.00   2491.98     -5.938250  2335.01  2343.89   \n","77763 2017-06-28 00:01:00  2499.39   2682.25     -6.817411  2495.00  2499.33   \n","77764 2017-06-28 00:02:00  2337.18   2491.98     -6.211928  2337.18  2340.00   \n","77765 2017-06-28 00:02:00  2492.76   2682.25     -7.064591  2492.76  2495.00   \n","77766 2017-06-28 00:03:00  2335.02   2491.98     -6.298606  2335.01  2335.02   \n","\n","          low     high        volume  \n","77762  2307.0  2473.19  20719.583592  \n","77763  2444.0  2780.62   2265.557866  \n","77764  2307.0  2473.19  20732.082581  \n","77765  2444.0  2780.62   2262.618866  \n","77766  2307.0  2473.19  20665.357191  "]},"execution_count":9,"metadata":{},"output_type":"execute_result"}],"source":["df.head()"]},{"cell_type":"markdown","metadata":{"id":"kG9JKDO7epqB"},"source":["### we require only the last value, so we subset that and convert it to numpy \n"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"F7Xgo9yMepqB"},"outputs":[],"source":["df = df[['last']]"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"EJ3FUEh2epqB"},"outputs":[],"source":["dataset = df.values\n","dataset = dataset.astype('float32')"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"_C9ICkfDepqB","outputId":"1316ac8c-35b7-457e-fad6-fd3e83494f51"},"outputs":[{"data":{"text/plain":["array([[ 2344.        ],\n","       [ 2499.38989258],\n","       [ 2337.17993164],\n","       ..., \n","       [ 2394.0300293 ],\n","       [ 2320.4699707 ],\n","       [ 2394.0300293 ]], dtype=float32)"]},"execution_count":12,"metadata":{},"output_type":"execute_result"}],"source":["dataset"]},{"cell_type":"markdown","metadata":{"id":"BwWVr2fuepqB"},"source":["Neural networks are sensitive to input data, especiallly when we are using activation functions like sigmoid or tanh activation functions are used. ISo we rescale our data to the range of 0-to-1, using MinMaxScaler"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"9dA81OYhepqC"},"outputs":[],"source":["scaler = MinMaxScaler(feature_range=(0, 1))\n","dataset = scaler.fit_transform(dataset)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"3aT26PODepqC","outputId":"31cdd2c4-82ad-457b-b886-990bf53f54d1"},"outputs":[{"data":{"text/plain":["array([[ 0.1997695 ],\n","       [ 0.49828053],\n","       [ 0.18666792],\n","       ..., \n","       [ 0.29587936],\n","       [ 0.15456724],\n","       [ 0.29587936]], dtype=float32)"]},"execution_count":14,"metadata":{},"output_type":"execute_result"}],"source":["dataset"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"PItMZGr6epqC","outputId":"d3b35554-d56d-41c3-b5fc-6fa1a9fbadba"},"outputs":[{"name":"stdout","output_type":"stream","text":["(31942, 15734)\n"]}],"source":["train_size = int(len(dataset) * 0.67)\n","test_size = len(dataset) - train_size\n","train, test = dataset[0:train_size, :], dataset[train_size:len(dataset), :]\n","print(len(train), len(test))"]},{"cell_type":"markdown","metadata":{"id":"KAv2XoHKepqC"},"source":["Now let us define the function called create_dataset, which take two inputs, \n","\n","1. Dataset - numpy array that we want to convert into a dataset\n","2. look_back - number of previous time steps to use as input variables to predict the next time period\n"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"au37elaVepqC"},"outputs":[],"source":["# convert an array of values into a dataset matrix\n","def create_dataset(dataset, look_back=1):\n","  dataX, dataY = [], []\n","  for i in range(len(dataset)-look_back-1):\n","    a = dataset[i:(i+look_back), 0]\n","    dataX.append(a)\n","    dataY.append(dataset[i + look_back, 0])\n","  return np.array(dataX), np.array(dataY)"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"wMSnHoDeepqD"},"outputs":[],"source":["look_back = 10\n","trainX, trainY = create_dataset(train, look_back=look_back)\n","testX, testY = create_dataset(test, look_back=look_back)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"Up_Y8YwnepqD","outputId":"e1d6eabd-a8fa-4213-de67-a5d5c15ac20e"},"outputs":[{"data":{"text/plain":["array([[ 0.1997695 ,  0.49828053,  0.18666792, ...,  0.49753141,\n","         0.19973087,  0.48600531],\n","       [ 0.49828053,  0.18666792,  0.4855442 , ...,  0.19973087,\n","         0.48600531,  0.18442059],\n","       [ 0.18666792,  0.4855442 ,  0.18251848, ...,  0.48600531,\n","         0.18442059,  0.48598576],\n","       ..., \n","       [ 0.53376245,  0.69436169,  0.53105354, ...,  0.70821238,\n","         0.52058411,  0.70815468],\n","       [ 0.69436169,  0.53105354,  0.70823145, ...,  0.52058411,\n","         0.70815468,  0.52665424],\n","       [ 0.53105354,  0.70823145,  0.53320551, ...,  0.70815468,\n","         0.52665424,  0.70815468]], dtype=float32)"]},"execution_count":18,"metadata":{},"output_type":"execute_result"}],"source":["trainX"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"eARnA_xjepqD","outputId":"68d7adab-0f0a-4356-9457-96be5da38e37"},"outputs":[{"data":{"text/plain":["array([ 0.18442059,  0.48598576,  0.19208527, ...,  0.52665424,\n","        0.70815468,  0.5206418 ], dtype=float32)"]},"execution_count":19,"metadata":{},"output_type":"execute_result"}],"source":["trainY"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"LcC7T1dfepqD"},"outputs":[],"source":["# reshape input to be [samples, time steps, features]\n","trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))\n","testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))"]},{"cell_type":"markdown","metadata":{"id":"1l8A6d08epqD"},"source":["## Build our Model"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"c7FkS6UAepqD","scrolled":true},"outputs":[],"source":["model = Sequential()\n","model.add(LSTM(4, input_shape=(1, look_back)))\n","model.add(Dense(1))\n","model.compile(loss='mean_squared_error', optimizer='adam')\n","model.fit(trainX, trainY, epochs=100, batch_size=256, verbose=2)"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"olbGASFRepqE"},"outputs":[],"source":["trainPredict = model.predict(trainX)\n","testPredict = model.predict(testX)"]},{"cell_type":"markdown","metadata":{"id":"_UZto9QFepqE"},"source":["We have to invert the predictions before calculating error to so that reports will be in same units as our original data\n","\n","\n"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"WJf31BgvepqE"},"outputs":[],"source":["trainPredict = scaler.inverse_transform(trainPredict)\n","trainY = scaler.inverse_transform([trainY])\n","testPredict = scaler.inverse_transform(testPredict)\n","testY = scaler.inverse_transform([testY])"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"Jxt4cZVMepqE","outputId":"524882a4-6c7a-447e-ae98-cae7d23f0ea9"},"outputs":[{"name":"stdout","output_type":"stream","text":["Train Score: 4.77 RMSE\n","Test Score: 5.57 RMSE\n"]}],"source":["\n","trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:, 0]))\n","print('Train Score: %.2f RMSE' % (trainScore))\n","testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:, 0]))\n","print('Test Score: %.2f RMSE' % (testScore))\n"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"lcD8bOwcepqE"},"outputs":[],"source":["# shift train predictions for plotting\n","trainPredictPlot = np.empty_like(dataset)\n","trainPredictPlot[:, :] = np.nan\n","trainPredictPlot[look_back:len(trainPredict) + look_back, :] = trainPredict"]},{"cell_type":"code","execution_count":null,"metadata":{"collapsed":true,"id":"BjiAecapepqE"},"outputs":[],"source":[" # shift test predictions for plotting\n","testPredictPlot = np.empty_like(dataset)\n","testPredictPlot[:, :] = np.nan\n","testPredictPlot[len(trainPredict) + (look_back * 2) + 1:len(dataset) - 1, :] = testPredict\n"," \n"]},{"cell_type":"code","execution_count":1,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":235},"executionInfo":{"elapsed":9,"status":"error","timestamp":1680632318997,"user":{"displayName":"Lakshya Satpal","userId":"09688867864267811029"},"user_tz":-330},"id":"pOPCqbToepqE","outputId":"23e727a8-1dfc-4559-f2af-2ea9e9a5db8c"},"outputs":[{"ename":"NameError","evalue":"ignored","output_type":"error","traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)","\u001b[0;32m<ipython-input-1-c4cecc8e10ba>\u001b[0m in \u001b[0;36m<cell line: 1>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'last'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'Actual'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDataFrame\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrainPredictPlot\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcolumns\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"close\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclose\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'Training'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDataFrame\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtestPredictPlot\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcolumns\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"close\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclose\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'Testing'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlegend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mloc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'best'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mNameError\u001b[0m: name 'plt' is not defined"]}],"source":["plt.plot(df['last'], label='Actual')\n","plt.plot(pd.DataFrame(trainPredictPlot, columns=[\"close\"], index=df.index).close, label='Training')\n","plt.plot(pd.DataFrame(testPredictPlot, columns=[\"close\"], index=df.index).close, label='Testing')\n","plt.legend(loc='best')\n","plt.show()"]}],"metadata":{"colab":{"provenance":[]},"kernelspec":{"display_name":"Python [conda root]","language":"python","name":"conda-root-py"},"language_info":{"codemirror_mode":{"name":"ipython","version":2},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython2","version":"2.7.11"}},"nbformat":4,"nbformat_minor":0}