timeSeriesPredicter.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy of timeSeriesPredicter.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SQlwyobxZi1a"
      },
      "source": [
        "# Stock Cost Time-Series Predicter App\n",
        "This notebook gonna show  you how to predict a time-series using a LSTM tecnique and create a succesful model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZeiASA6jjpx0"
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib as mpl\n",
        "import pandas_datareader as dt\n",
        "import datetime as date\n",
        "\n",
        "from  sklearn.preprocessing import MinMaxScaler\n",
        "from tensorflow.keras.layers import Dense,Dropout, LSTM\n",
        "from tensorflow.keras.models import Sequential\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4GCebAiMgOiV"
      },
      "source": [
        "Adding local variable for Yahoo finance"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HkO8OIZlcnVJ"
      },
      "source": [
        "ticker='CNY=X'   ## USD-China Yuan\n",
        "start_date = date.datetime(2020,1,1)\n",
        "end_date=date.datetime.now()\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 532
        },
        "id": "KhoVoN6vfTA7",
        "outputId": "fb921cd8-237b-4bea-a3a2-752126b67ac4"
      },
      "source": [
        "data = dt.DataReader(name=ticker, data_source='yahoo', start=start_date,end=end_date )\n",
        "data.head(15)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>High</th>\n",
              "      <th>Low</th>\n",
              "      <th>Open</th>\n",
              "      <th>Close</th>\n",
              "      <th>Volume</th>\n",
              "      <th>Adj Close</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Date</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2020-01-01</th>\n",
              "      <td>6.9623</td>\n",
              "      <td>6.9622</td>\n",
              "      <td>6.9623</td>\n",
              "      <td>6.9623</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9623</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-02</th>\n",
              "      <td>6.9730</td>\n",
              "      <td>6.9598</td>\n",
              "      <td>6.9622</td>\n",
              "      <td>6.9622</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9622</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-03</th>\n",
              "      <td>6.9744</td>\n",
              "      <td>6.9591</td>\n",
              "      <td>6.9634</td>\n",
              "      <td>6.9634</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9634</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-06</th>\n",
              "      <td>6.9775</td>\n",
              "      <td>6.9645</td>\n",
              "      <td>6.9645</td>\n",
              "      <td>6.9645</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9645</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-07</th>\n",
              "      <td>6.9750</td>\n",
              "      <td>6.9308</td>\n",
              "      <td>6.9745</td>\n",
              "      <td>6.9745</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9745</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-08</th>\n",
              "      <td>6.9530</td>\n",
              "      <td>6.9315</td>\n",
              "      <td>6.9444</td>\n",
              "      <td>6.9444</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9444</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-09</th>\n",
              "      <td>6.9516</td>\n",
              "      <td>6.9451</td>\n",
              "      <td>6.9452</td>\n",
              "      <td>6.9452</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9452</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-10</th>\n",
              "      <td>6.9560</td>\n",
              "      <td>6.9093</td>\n",
              "      <td>6.9517</td>\n",
              "      <td>6.9517</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9517</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-13</th>\n",
              "      <td>6.9142</td>\n",
              "      <td>6.8837</td>\n",
              "      <td>6.9093</td>\n",
              "      <td>6.9093</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9093</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-14</th>\n",
              "      <td>6.9003</td>\n",
              "      <td>6.8646</td>\n",
              "      <td>6.8927</td>\n",
              "      <td>6.8927</td>\n",
              "      <td>0</td>\n",
              "      <td>6.8927</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-15</th>\n",
              "      <td>6.9004</td>\n",
              "      <td>6.8828</td>\n",
              "      <td>6.9004</td>\n",
              "      <td>6.9003</td>\n",
              "      <td>0</td>\n",
              "      <td>6.9003</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-16</th>\n",
              "      <td>6.8907</td>\n",
              "      <td>6.8701</td>\n",
              "      <td>6.8898</td>\n",
              "      <td>6.8898</td>\n",
              "      <td>0</td>\n",
              "      <td>6.8898</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-17</th>\n",
              "      <td>6.8781</td>\n",
              "      <td>6.8526</td>\n",
              "      <td>6.8781</td>\n",
              "      <td>6.8781</td>\n",
              "      <td>0</td>\n",
              "      <td>6.8781</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-20</th>\n",
              "      <td>6.8659</td>\n",
              "      <td>6.8398</td>\n",
              "      <td>6.8587</td>\n",
              "      <td>6.8587</td>\n",
              "      <td>0</td>\n",
              "      <td>6.8587</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-01-21</th>\n",
              "      <td>6.9080</td>\n",
              "      <td>6.8658</td>\n",
              "      <td>6.8658</td>\n",
              "      <td>6.8658</td>\n",
              "      <td>0</td>\n",
              "      <td>6.8658</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "              High     Low    Open   Close  Volume  Adj Close\n",
              "Date                                                         \n",
              "2020-01-01  6.9623  6.9622  6.9623  6.9623       0     6.9623\n",
              "2020-01-02  6.9730  6.9598  6.9622  6.9622       0     6.9622\n",
              "2020-01-03  6.9744  6.9591  6.9634  6.9634       0     6.9634\n",
              "2020-01-06  6.9775  6.9645  6.9645  6.9645       0     6.9645\n",
              "2020-01-07  6.9750  6.9308  6.9745  6.9745       0     6.9745\n",
              "2020-01-08  6.9530  6.9315  6.9444  6.9444       0     6.9444\n",
              "2020-01-09  6.9516  6.9451  6.9452  6.9452       0     6.9452\n",
              "2020-01-10  6.9560  6.9093  6.9517  6.9517       0     6.9517\n",
              "2020-01-13  6.9142  6.8837  6.9093  6.9093       0     6.9093\n",
              "2020-01-14  6.9003  6.8646  6.8927  6.8927       0     6.8927\n",
              "2020-01-15  6.9004  6.8828  6.9004  6.9003       0     6.9003\n",
              "2020-01-16  6.8907  6.8701  6.8898  6.8898       0     6.8898\n",
              "2020-01-17  6.8781  6.8526  6.8781  6.8781       0     6.8781\n",
              "2020-01-20  6.8659  6.8398  6.8587  6.8587       0     6.8587\n",
              "2020-01-21  6.9080  6.8658  6.8658  6.8658       0     6.8658"
            ]
          },
          "metadata": {},
          "execution_count": 149
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wI-D7ov3-ZgX"
      },
      "source": [
        "## Normalize the \"Close\" cost"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5-qZUCzJfVKr",
        "outputId": "6d8274ae-be41-4b27-a04f-b20f259c74fd"
      },
      "source": [
        "Close = data['Close']\n",
        "scaler = MinMaxScaler(feature_range=(0,1))\n",
        "scaledData = scaler.fit_transform(Close.values.reshape(-1,1))\n",
        "scaledData.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(461, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 150
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DcVIe_QL-h5G"
      },
      "source": [
        "## Split training and test data set\n",
        "we have 461 day  cost data, I'm gonna split the data set, first 400 days  as training set and rest of them as test datasets.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Sa32H02nlhcv",
        "outputId": "932385aa-a2b3-4438-b8da-b74efd80e38b"
      },
      "source": [
        "training_set = scaledData[:400]\n",
        "test_set = scaledData[400:]\n",
        "\n",
        "training_set.shape, test_set.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((400, 1), (61, 1))"
            ]
          },
          "metadata": {},
          "execution_count": 151
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aQKtJ2ghoeJz"
      },
      "source": [
        "## Create a model "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1Rgk60GjodcN",
        "outputId": "d9c17970-7ae6-40a3-c549-0d1861f3ab8f"
      },
      "source": [
        "period=60 \n",
        "X_train = []\n",
        "y_train = []\n",
        "for x in range(period, len(training_set)):\n",
        "    X_train.append(training_set[x-period:x, 0])\n",
        "    y_train.append(training_set[x, 0])\n",
        "X_train, y_train = np.array(X_train), np.array(y_train)\n",
        "X_train = np.reshape(X_train, (X_train.shape[0], X_train.shape[1], 1))\n",
        "X_train.shape, y_train.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((340, 60, 1), (340,))"
            ]
          },
          "metadata": {},
          "execution_count": 152
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C6X4Ti3KBKme",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "da27c279-beb2-4755-a778-e053541516c4"
      },
      "source": [
        "model = Sequential()\n",
        "#First\n",
        "model.add(LSTM(units = 50, return_sequences = True, input_shape = (X_train.shape[1], 1)))\n",
        "model.add(Dropout(0.2))\n",
        "# Second\n",
        "model.add(LSTM(units = 50, return_sequences = True))\n",
        "model.add(Dropout(0.2))\n",
        "# Third \n",
        "model.add(LSTM(units = 50, return_sequences = True))\n",
        "model.add(Dropout(0.2))\n",
        "# Fourth\n",
        "model.add(LSTM(units = 50))\n",
        "model.add(Dropout(0.2))\n",
        "#  output layer\n",
        "model.add(Dense(units = 1))\n",
        "# Compiling\n",
        "model.compile(optimizer = 'adam', loss = 'mean_squared_error')\n",
        "\n",
        "# Fitting to the Training set\n",
        "model.fit(X_train, y_train, epochs = 100, batch_size = 32)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "11/11 [==============================] - 8s 124ms/step - loss: 0.0943\n",
            "Epoch 2/100\n",
            "11/11 [==============================] - 2s 138ms/step - loss: 0.0243\n",
            "Epoch 3/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0101\n",
            "Epoch 4/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0103\n",
            "Epoch 5/100\n",
            "11/11 [==============================] - 2s 136ms/step - loss: 0.0097\n",
            "Epoch 6/100\n",
            "11/11 [==============================] - 1s 136ms/step - loss: 0.0067\n",
            "Epoch 7/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0090\n",
            "Epoch 8/100\n",
            "11/11 [==============================] - 2s 139ms/step - loss: 0.0071\n",
            "Epoch 9/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0056\n",
            "Epoch 10/100\n",
            "11/11 [==============================] - 1s 134ms/step - loss: 0.0059\n",
            "Epoch 11/100\n",
            "11/11 [==============================] - 2s 138ms/step - loss: 0.0062\n",
            "Epoch 12/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0059\n",
            "Epoch 13/100\n",
            "11/11 [==============================] - 1s 130ms/step - loss: 0.0066\n",
            "Epoch 14/100\n",
            "11/11 [==============================] - 1s 128ms/step - loss: 0.0052\n",
            "Epoch 15/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0060\n",
            "Epoch 16/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0061\n",
            "Epoch 17/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0056\n",
            "Epoch 18/100\n",
            "11/11 [==============================] - 1s 126ms/step - loss: 0.0056\n",
            "Epoch 19/100\n",
            "11/11 [==============================] - 1s 130ms/step - loss: 0.0047\n",
            "Epoch 20/100\n",
            "11/11 [==============================] - 1s 128ms/step - loss: 0.0047\n",
            "Epoch 21/100\n",
            "11/11 [==============================] - 1s 127ms/step - loss: 0.0047\n",
            "Epoch 22/100\n",
            "11/11 [==============================] - 1s 127ms/step - loss: 0.0047\n",
            "Epoch 23/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0054\n",
            "Epoch 24/100\n",
            "11/11 [==============================] - 1s 128ms/step - loss: 0.0058\n",
            "Epoch 25/100\n",
            "11/11 [==============================] - 1s 128ms/step - loss: 0.0066\n",
            "Epoch 26/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0056\n",
            "Epoch 27/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0053\n",
            "Epoch 28/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0052\n",
            "Epoch 29/100\n",
            "11/11 [==============================] - 1s 125ms/step - loss: 0.0058\n",
            "Epoch 30/100\n",
            "11/11 [==============================] - 1s 126ms/step - loss: 0.0061\n",
            "Epoch 31/100\n",
            "11/11 [==============================] - 1s 130ms/step - loss: 0.0055\n",
            "Epoch 32/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0043\n",
            "Epoch 33/100\n",
            "11/11 [==============================] - 1s 131ms/step - loss: 0.0039\n",
            "Epoch 34/100\n",
            "11/11 [==============================] - 1s 128ms/step - loss: 0.0050\n",
            "Epoch 35/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0044\n",
            "Epoch 36/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0037\n",
            "Epoch 37/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0045\n",
            "Epoch 38/100\n",
            "11/11 [==============================] - 1s 135ms/step - loss: 0.0054\n",
            "Epoch 39/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0050\n",
            "Epoch 40/100\n",
            "11/11 [==============================] - 1s 127ms/step - loss: 0.0046\n",
            "Epoch 41/100\n",
            "11/11 [==============================] - 1s 125ms/step - loss: 0.0047\n",
            "Epoch 42/100\n",
            "11/11 [==============================] - 1s 128ms/step - loss: 0.0042\n",
            "Epoch 43/100\n",
            "11/11 [==============================] - 1s 124ms/step - loss: 0.0043\n",
            "Epoch 44/100\n",
            "11/11 [==============================] - 1s 127ms/step - loss: 0.0045\n",
            "Epoch 45/100\n",
            "11/11 [==============================] - 1s 131ms/step - loss: 0.0052\n",
            "Epoch 46/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0049\n",
            "Epoch 47/100\n",
            "11/11 [==============================] - 1s 126ms/step - loss: 0.0040\n",
            "Epoch 48/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0039\n",
            "Epoch 49/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0049\n",
            "Epoch 50/100\n",
            "11/11 [==============================] - 2s 141ms/step - loss: 0.0044\n",
            "Epoch 51/100\n",
            "11/11 [==============================] - 1s 134ms/step - loss: 0.0043\n",
            "Epoch 52/100\n",
            "11/11 [==============================] - 2s 137ms/step - loss: 0.0045\n",
            "Epoch 53/100\n",
            "11/11 [==============================] - 1s 130ms/step - loss: 0.0056\n",
            "Epoch 54/100\n",
            "11/11 [==============================] - 2s 139ms/step - loss: 0.0043\n",
            "Epoch 55/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0041\n",
            "Epoch 56/100\n",
            "11/11 [==============================] - 2s 138ms/step - loss: 0.0047\n",
            "Epoch 57/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0044\n",
            "Epoch 58/100\n",
            "11/11 [==============================] - 1s 134ms/step - loss: 0.0038\n",
            "Epoch 59/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0042\n",
            "Epoch 60/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0034\n",
            "Epoch 61/100\n",
            "11/11 [==============================] - 1s 134ms/step - loss: 0.0031\n",
            "Epoch 62/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0042\n",
            "Epoch 63/100\n",
            "11/11 [==============================] - 1s 131ms/step - loss: 0.0036\n",
            "Epoch 64/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0036\n",
            "Epoch 65/100\n",
            "11/11 [==============================] - 2s 139ms/step - loss: 0.0033\n",
            "Epoch 66/100\n",
            "11/11 [==============================] - 2s 142ms/step - loss: 0.0031\n",
            "Epoch 67/100\n",
            "11/11 [==============================] - 2s 138ms/step - loss: 0.0036\n",
            "Epoch 68/100\n",
            "11/11 [==============================] - 1s 129ms/step - loss: 0.0044\n",
            "Epoch 69/100\n",
            "11/11 [==============================] - 2s 137ms/step - loss: 0.0041\n",
            "Epoch 70/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0037\n",
            "Epoch 71/100\n",
            "11/11 [==============================] - 2s 137ms/step - loss: 0.0034\n",
            "Epoch 72/100\n",
            "11/11 [==============================] - 1s 136ms/step - loss: 0.0033\n",
            "Epoch 73/100\n",
            "11/11 [==============================] - 2s 137ms/step - loss: 0.0035\n",
            "Epoch 74/100\n",
            "11/11 [==============================] - 1s 130ms/step - loss: 0.0034\n",
            "Epoch 75/100\n",
            "11/11 [==============================] - 1s 127ms/step - loss: 0.0032\n",
            "Epoch 76/100\n",
            "11/11 [==============================] - 1s 124ms/step - loss: 0.0036\n",
            "Epoch 77/100\n",
            "11/11 [==============================] - 1s 124ms/step - loss: 0.0038\n",
            "Epoch 78/100\n",
            "11/11 [==============================] - 1s 130ms/step - loss: 0.0029\n",
            "Epoch 79/100\n",
            "11/11 [==============================] - 2s 136ms/step - loss: 0.0041\n",
            "Epoch 80/100\n",
            "11/11 [==============================] - 1s 131ms/step - loss: 0.0048\n",
            "Epoch 81/100\n",
            "11/11 [==============================] - 1s 127ms/step - loss: 0.0029\n",
            "Epoch 82/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0033\n",
            "Epoch 83/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0034\n",
            "Epoch 84/100\n",
            "11/11 [==============================] - 1s 132ms/step - loss: 0.0042\n",
            "Epoch 85/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0036\n",
            "Epoch 86/100\n",
            "11/11 [==============================] - 1s 128ms/step - loss: 0.0038\n",
            "Epoch 87/100\n",
            "11/11 [==============================] - 2s 138ms/step - loss: 0.0035\n",
            "Epoch 88/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0036\n",
            "Epoch 89/100\n",
            "11/11 [==============================] - 2s 138ms/step - loss: 0.0037\n",
            "Epoch 90/100\n",
            "11/11 [==============================] - 1s 134ms/step - loss: 0.0033\n",
            "Epoch 91/100\n",
            "11/11 [==============================] - 1s 135ms/step - loss: 0.0032\n",
            "Epoch 92/100\n",
            "11/11 [==============================] - 1s 131ms/step - loss: 0.0030\n",
            "Epoch 93/100\n",
            "11/11 [==============================] - 1s 134ms/step - loss: 0.0031\n",
            "Epoch 94/100\n",
            "11/11 [==============================] - 2s 139ms/step - loss: 0.0034\n",
            "Epoch 95/100\n",
            "11/11 [==============================] - 1s 131ms/step - loss: 0.0033\n",
            "Epoch 96/100\n",
            "11/11 [==============================] - 2s 136ms/step - loss: 0.0039\n",
            "Epoch 97/100\n",
            "11/11 [==============================] - 1s 133ms/step - loss: 0.0033\n",
            "Epoch 98/100\n",
            "11/11 [==============================] - 2s 137ms/step - loss: 0.0036\n",
            "Epoch 99/100\n",
            "11/11 [==============================] - 2s 143ms/step - loss: 0.0039\n",
            "Epoch 100/100\n",
            "11/11 [==============================] - 2s 139ms/step - loss: 0.0032\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fc28fa29390>"
            ]
          },
          "metadata": {},
          "execution_count": 153
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AM2jfhroFsai"
      },
      "source": [
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EX89vpfbDnfH",
        "outputId": "0cc4ee1f-ea1c-47e4-d660-11c840832827"
      },
      "source": [
        "dataset_train = data.iloc[:400,3:4]\n",
        "dataset_test = data.iloc[400:,3:4]\n",
        "\n",
        "dataset_total = pd.concat((dataset_train, dataset_test), axis = 0)\n",
        "inputs = dataset_total[len(dataset_total) - len(dataset_test) - period:].values\n",
        "inputs = inputs.reshape(-1,1)\n",
        "inputs = scaler.transform(inputs)\n",
        "X_test = []\n",
        "for i in range(period, len(inputs)):\n",
        "    X_test.append(inputs[i-period:i, 0])\n",
        "X_test = np.array(X_test)\n",
        "X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1], 1))\n",
        "print(X_test.shape)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(61, 60, 1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZRNVz4oQGzws"
      },
      "source": [
        "predicted_stock_price = model.predict(X_test)\n",
        "predicted_stock_price = scaler.inverse_transform(predicted_stock_price)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "b7cfNniqNeEd",
        "outputId": "26394bbb-c380-4127-ab2e-f490c694b25b"
      },
      "source": [
        "dataset_test.values"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[6.46750021],\n",
              "       [6.46829987],\n",
              "       [6.46010017],\n",
              "       [6.47819996],\n",
              "       [6.48939991],\n",
              "       [6.48390007],\n",
              "       [6.46700001],\n",
              "       [6.46969986],\n",
              "       [6.48040009],\n",
              "       [6.48169994],\n",
              "       [6.50969982],\n",
              "       [6.48939991],\n",
              "       [6.45529985],\n",
              "       [6.46049976],\n",
              "       [6.46099997],\n",
              "       [6.46920013],\n",
              "       [6.46519995],\n",
              "       [6.46029997],\n",
              "       [6.48250008],\n",
              "       [6.48519993],\n",
              "       [6.4849    ],\n",
              "       [6.47819996],\n",
              "       [6.47830009],\n",
              "       [6.4763999 ],\n",
              "       [6.47359991],\n",
              "       [6.48509979],\n",
              "       [6.48320007],\n",
              "       [6.49370003],\n",
              "       [6.49160004],\n",
              "       [6.48000002],\n",
              "       [6.46990013],\n",
              "       [6.47489977],\n",
              "       [6.48050022],\n",
              "       [6.47079992],\n",
              "       [6.46509981],\n",
              "       [6.45959997],\n",
              "       [6.45900011],\n",
              "       [6.45539999],\n",
              "       [6.45190001],\n",
              "       [6.45669985],\n",
              "       [6.46549988],\n",
              "       [6.46070004],\n",
              "       [6.45429993],\n",
              "       [6.44339991],\n",
              "       [6.45079994],\n",
              "       [6.43830013],\n",
              "       [6.43139982],\n",
              "       [6.45660019],\n",
              "       [6.46519995],\n",
              "       [6.46500015],\n",
              "       [6.46600008],\n",
              "       [6.46080017],\n",
              "       [6.45790005],\n",
              "       [6.46570015],\n",
              "       [6.45599985],\n",
              "       [6.45900011],\n",
              "       [6.4696002 ],\n",
              "       [6.4454999 ],\n",
              "       [6.44570017],\n",
              "       [6.44519997],\n",
              "       [6.44519997]])"
            ]
          },
          "metadata": {},
          "execution_count": 156
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "843Pw1eVG7hH",
        "outputId": "2b7450fa-26f8-4b6d-fc71-4f39819a11a5"
      },
      "source": [
        "\n",
        "mpl.pyplot.plot(dataset_test.values,label=\"Actual Data\")\n",
        "mpl.pyplot.plot(predicted_stock_price,label=\"Predicted data\")\n",
        "mpl.pyplot.title('USD-CHY Price Prediction')\n",
        "mpl.pyplot.xlabel('Time')\n",
        "mpl.pyplot.ylabel('USD-CHY Price')\n",
        "mpl.pyplot.legend()\n",
        "mpl.pyplot.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}