Pairs_Trading(LSTM).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Pairs Trading(LSTM).ipynb",
      "version": "0.3.2",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/tylee33/DLforFinance/blob/master/Pairs_Trading(LSTM).ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "Bby7mrOQSr2w",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import keras.callbacks\n",
        "from keras.models import Sequential\n",
        "from keras.layers.core import Dense, Activation, Dense, Dropout\n",
        "from keras.layers.recurrent import LSTM\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "import numpy.random as random"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "TvJqGlr6S7aS",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "LOOKBACK_WINDOW = 15\n",
        "PREDICT_WINDOW = 1\n",
        "BATCH_SIZE = 1     #e.g if 3, then trains with 3 samples/lookback windows, each with 15 timesteps and 2 features at once before updating gradients.\n",
        "HIDDEN = 32\n",
        "DROPOUT = 0.3\n",
        "OPTIMIZER = 'adam'\n",
        "EPOCHS = 3"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "mB4VsbGMS-Gx",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "BATCH_SIZE - 그라디언트 업데이트 당 샘플 수입니다. 가중치를 업데이트하기 전에 LSTM 네트워크를 학습하는 데 사용되는 N 개의 샘플 집합입니다. 일괄 처리로 인해 모델이 한 번 업데이트됩니다. 따라서 배치 크기가 작을수록 학습 과정이 빨라집니다. \n",
        "\n",
        "batch_input_shape - 형상을 3D 배열로 전달하는 경우 : (batch_size, timesteps, input_dim); feature를 2 차원 배열로 전달하는 경우 : (batch_size, output_dim). batch_input_shape는 신경망의 순차 분류가 정의 된 배치 크기의 입력 데이터를 받아 들일 수 있으며, 가변 길이 벡터의 생성을 제한합니다. \n",
        "\n",
        "EPOCH - 일반적으로 \"전체 데이터 세트에 대해 하나의 통과\"로 정의되는 임의 컷오프는 교육을 별도의 단계로 분리하는 데 사용되며 로깅 및 정기 평가에 유용합니다. \n",
        "\n",
        "HIDDEN - 숨겨진 레이어의 수입니다.\n",
        "\n",
        "LOOKBACK_WINDOW - 일괄 처리의 타임 스텝 수입니다. \n",
        "\n",
        "DROPOUT - 네트워크를 학습하는 동안 셀 상태에 대한 입력 및 순환 연결이 활성화 및 가중치 업데이트에서 확률 적으로 제외되는 정규화 방법입니다. 이렇게하면 모델 성능을 향상시키기 위해 지나친 조정이 줄어 듭니다."
      ]
    },
    {
      "metadata": {
        "id": "Y5R5cb8CTIQB",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def coint_path(N,delta,sigma,PO):\n",
        "   X = [PO]\n",
        "   Y = [PO]\n",
        "   for i in range(N):\n",
        "      dX = random.randn()+delta*(Y[-1] - X[-1])\n",
        "      X.append(X[-1]+dX)\n",
        "      dY = random.randn()+sigma*(X[-1] - Y[-1])\n",
        "      Y.append(Y[-1]+dY)\n",
        " \n",
        "   return X,Y\n",
        " \n",
        "#create artificial cointegrated series\n",
        "X,Y = coint_path(60,0.9,0.1,0)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "LB3XhXWWTLN4",
        "colab_type": "code",
        "outputId": "3e73d974-4e63-4fb0-eab9-41b7271edf13",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 347
        }
      },
      "cell_type": "code",
      "source": [
        "plt.plot(np.array(Y),label='Y')\n",
        "plt.plot(np.array(X),label='X')\n",
        "plt.legend()\n",
        "plt.show();"
      ],
      "execution_count": 103,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAd8AAAFKCAYAAABcq1WoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+AohhBAZJsFXCCGEyLD/H+d8Tj0R43vUAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "Q48xorRhTO5k",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "우리는 두 개의 시리즈를 함께 다차원 배열로 '쌓아야'합니다. 그런 다음 sklearn의 MinMaxScaler 함수를 사용하여 새로 생성 된 다차원 배열을 정규화합니다. 이 값은 -1과 1 사이의 값으로 변환됩니다."
      ]
    },
    {
      "metadata": {
        "id": "ELWHwL2nTNwP",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def prep_feature_data(X,Y):\n",
        "    data = np.column_stack([X,Y])\n",
        "    scaler = MinMaxScaler(feature_range=(-1,1)) #다차원 배열 정규화\n",
        "    data = scaler.fit_transform(data)\n",
        "    return data\n",
        " \n",
        "#format features for the model into a multidimensional array\n",
        "data = prep_feature_data(X,Y)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "TojOy_ebTU-o",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "이것은 다차원 배열과 같이 이제 형상 데이터가 어떻게 생겼는지를 나타냅니다."
      ]
    },
    {
      "metadata": {
        "id": "MIn62mQSTXW_",
        "colab_type": "code",
        "outputId": "b86f29a7-46fa-48c0-8ab9-c0313070d716",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "cell_type": "code",
      "source": [
        "print(data[0:3])\n",
        "print(data.shape)"
      ],
      "execution_count": 105,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[-0.22701844 -0.32011537]\n",
            " [-0.24623211 -0.11204855]\n",
            " [-0.16604053 -0.10926056]]\n",
            "(61, 2)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "WWW0vFoyTZdB",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "이제 우리는 nb_samples에 대해이 정수 값을 계산합니다. nb_samples는 전환 확인 기간을 설정할 다음 단계의 정수입니다. 얼마나 많은 총 샘플 / 관측치보다 예측 및 전환 확인 기간이 적어 지므로 이러한 창 생성을 중단해야하는지 알 수 있습니다."
      ]
    },
    {
      "metadata": {
        "id": "XA5qgEeTTb6t",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def window_stop(data,LOOKBACK_WINDOW,PREDICT_WINDOW):\n",
        "    examples = LOOKBACK_WINDOW\n",
        "    y_examples = PREDICT_WINDOW\n",
        "    nb_samples = len(data) - examples - y_examples #makes sure it can be split into lookback windows properly\n",
        "    return nb_samples\n",
        " \n",
        "#ensure it can be divided into the lookback window/batch size\n",
        "nb_samples = window_stop(data,LOOKBACK_WINDOW,PREDICT_WINDOW)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "gm8NuZP0TgV9",
        "colab_type": "code",
        "outputId": "136f9fc6-5e56-45b0-9ab1-3084205e7fb3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "print(nb_samples)"
      ],
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "45\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "8V8yy-BLTiV1",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "이제는 더 많은 형식 지정이 기능 데이터에 필요합니다. 여기서는 LSTM에 반영 할 전환 확인 기간을 만듭니다.\n",
        "\n",
        "np.expand_dims는 확장 배열 모양의 축 위치에 나타날 새 축을 삽입하여 배열의 모양을 확장합니다. 각 데이터 전환 확인 기간의 이전 모양은 (관측 개수, 기능 개수) 즉, 20 개의 전환 확인 기간이 있고 2 개의 기능이있는 경우 이전 모양은 (20,2)이고 새 모양은 (1, 20, 2)입니다.\n",
        "\n",
        "이렇게하면 모양이 (800,2)가 아닌 (40,2,2) 별도의 전환 확인 기간이 모두 연결되면됩니다. 즉 2 개의 기능이있는 20 개의 룩백 샘플로 구성된 40 개의 룩백 (lookback) 윈도우가 있습니다."
      ]
    },
    {
      "metadata": {
        "id": "lbtuQVSYTkdH",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def input_features(nb_samples,LOOKBACK_WINDOW):\n",
        "    input_list = [np.expand_dims(data[i:LOOKBACK_WINDOW+i,:], axis=0) for i in range(nb_samples)] #here nb_samples comes in handy\n",
        "    input_mat = np.concatenate(input_list, axis=0)\n",
        "    return input_mat\n",
        " \n",
        "#format the features into the batch size\n",
        "input_mat = input_features(nb_samples,LOOKBACK_WINDOW)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "YJaySeeCTntt",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "이것은 새로운 입력 행렬의 모양을 미리 보여줍니다. 전환 확인 기간은 항상 한 번에 한 단계 앞으로 이동하는 방법에 유의하십시오. 데이터를 20 개의 샘플 그룹으로 단순히 나누는 것과 같지 않습니다."
      ]
    },
    {
      "metadata": {
        "id": "oIJJMGoKTp-V",
        "colab_type": "code",
        "outputId": "d70a9f9d-4095-4f5d-b9c8-cd6e4ad0c955",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 561
        }
      },
      "cell_type": "code",
      "source": [
        "print(input_mat[0:2]) #the second sample of features will become the first sample in the next window\n",
        "print(input_mat.shape)"
      ],
      "execution_count": 108,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[[-0.22701844 -0.32011537]\n",
            "  [-0.24623211 -0.11204855]\n",
            "  [-0.16604053 -0.10926056]\n",
            "  [-0.19285018 -0.33406838]\n",
            "  [-0.3727985  -0.38184457]\n",
            "  [-0.48811632 -0.20575102]\n",
            "  [ 0.10386427 -0.0486566 ]\n",
            "  [-0.09296896  0.0356976 ]\n",
            "  [ 0.16624383  0.0088588 ]\n",
            "  [-0.12385044 -0.06313628]\n",
            "  [ 0.27848516 -0.07963483]\n",
            "  [-0.00122742 -0.43767753]\n",
            "  [-0.59782819 -0.76652298]\n",
            "  [-0.58583586 -0.84362597]\n",
            "  [-0.69777841 -0.82392044]]\n",
            "\n",
            " [[-0.24623211 -0.11204855]\n",
            "  [-0.16604053 -0.10926056]\n",
            "  [-0.19285018 -0.33406838]\n",
            "  [-0.3727985  -0.38184457]\n",
            "  [-0.48811632 -0.20575102]\n",
            "  [ 0.10386427 -0.0486566 ]\n",
            "  [-0.09296896  0.0356976 ]\n",
            "  [ 0.16624383  0.0088588 ]\n",
            "  [-0.12385044 -0.06313628]\n",
            "  [ 0.27848516 -0.07963483]\n",
            "  [-0.00122742 -0.43767753]\n",
            "  [-0.59782819 -0.76652298]\n",
            "  [-0.58583586 -0.84362597]\n",
            "  [-0.69777841 -0.82392044]\n",
            "  [-0.45662849 -0.57832139]]]\n",
            "(45, 15, 2)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "2Kkrh9o1TtNO",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "마지막으로 우리의 Feature Data Set가 준비되었습니다!\n",
        "\n",
        "이제 목표 값을 만듭니다. 이 예에서 우리는 LSTM 모델을 훈련하여 두 피쳐 간의 기울기 (베타)를 기반으로 선형 회귀선의 절편을 예측하려고합니다."
      ]
    },
    {
      "metadata": {
        "id": "JeIO5EtcTyQ8",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def target_values(input_mat):\n",
        "    targets = []\n",
        "    for i in range(len(input_mat)):\n",
        "        X1 = input_mat[i].T[0]\n",
        "        X2 = input_mat[i].T[1]\n",
        "        beta = np.polyfit(X1,X2,1)[0] #the slope is calculated from each lookback window \n",
        "        #targets.append(X1[-1] - beta*X2[-1])\n",
        "        #targets = np.array(X1[-1] - beta*X2[-1]) \n",
        "        targets = np.append(targets, X1[-1]-beta*X2[-1])\n",
        "        targets = np.array(targets)\n",
        "  \n",
        "    return targets"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "rWBC3JCJT1cr",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#format the targets into the batch size\n",
        "targets = target_values(input_mat)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "kIk4ulKoUoCF",
        "colab_type": "code",
        "outputId": "1b926cdf-29d6-4734-fb13-e2b90dc57b63",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "cell_type": "code",
      "source": [
        "print(targets)\n",
        "print(targets.shape)"
      ],
      "execution_count": 111,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[-0.02538364  0.02415468  0.43438918  0.26149216 -0.20152613 -0.46927685\n",
            " -0.29747156 -0.18410668 -0.43013045  0.25496894  0.11556141 -0.25550088\n",
            " -0.11287485 -0.10629178 -0.15871493 -0.04512547  0.03619763 -0.24549307\n",
            "  0.34055387 -0.0075504  -0.23907352 -0.36737035 -0.19625962 -0.17308257\n",
            "  0.00883774  0.27432831  0.12731167 -0.14250111  0.46878125  0.33475358\n",
            "  0.30177112  0.37631112  0.17771629  0.23805338  0.57681483  0.72649717\n",
            "  0.04582921  0.24343401  0.16745949  0.25907657 -0.25920669  0.25757496\n",
            " -0.05678766  0.09562598  0.27573018]\n",
            "(45,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "7UbGH79Vg3tX",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class LossHistory(keras.callbacks.Callback):\n",
        "  def on_train_begin(self, logs={}):\n",
        "    self.losses = [] \n",
        "    \n",
        "  def on_batch_end(self, batch, logs={}):\n",
        "    self.losses.append(logs.get('loss'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "2dNd5GE-UtBk",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def basic_LSTM_model(BATCH_SIZE,HIDDEN,LOOKBACK_WINDOW,DROPOUT,OPTIMIZER,EPOCHS,input_mat,targets):\n",
        "    num_features = input_mat.shape[2] #this is the same as input dimension, used in describing batch_input_shape\n",
        "    model = Sequential()\n",
        "    \n",
        "    model.add(LSTM(HIDDEN,batch_input_shape=(BATCH_SIZE,LOOKBACK_WINDOW,num_features)))\n",
        "    model.add(Dropout(DROPOUT))\n",
        "    model.add(Dense(1)) #the output is one dimensional\n",
        "    \n",
        "    model.compile(loss='mse',optimizer=OPTIMIZER)\n",
        "    history = LossHistory()\n",
        "    predictions,actuals,scores = [],[],[]\n",
        "    \n",
        "    assert(len(input_mat)==len(targets)) #checks each sample has a corresponding target value \n",
        "    \n",
        "    for i in range(1,len(input_mat)):\n",
        "        trainX = input_mat[0:i]\n",
        "        trainY = targets[0:i]\n",
        "        testX = input_mat[i].reshape(1,input_mat[0].shape[0],input_mat[0].shape[1]) #convert the input into 3 dimensional array\n",
        "        testY = np.array([targets[i]]).reshape(1,1) #convert the target value into a 2 dimensional array\n",
        "  \n",
        " #walk forward training, to predict the next timestep \n",
        " #you can adjust the verbose parameter to 1 or 2 to watch the LSTM's progress \n",
        "    model.fit(trainX,trainY,nb_epoch=EPOCHS,batch_size=BATCH_SIZE,callbacks=[history],verbose=0)\n",
        "    prediction = model.predict(testX,batch_size=BATCH_SIZE) \n",
        "    score = model.evaluate(testX,testY,batch_size=BATCH_SIZE,verbose=0)\n",
        "    \n",
        "    predictions.append(prediction[0][0])\n",
        "    actuals.append(targets[i])\n",
        "    scores.append(score)\n",
        "    \n",
        "    return model,history,predictions,actuals,scores"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "X9kpg8nSUvVt",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def evaluate_walk_forward_LSTM(model,history,predictions,actuals,scores):\n",
        "   print(np.mean(scores[int(len(scores)*0.75):])) #since the latter predictions have been trained on more data, we take the average of the testing loss scores for the last quarter of predictions\n",
        "  \n",
        "   plt.figure(1)\n",
        "   plt.plot(history.losses)\n",
        "   plt.title('Loss History')\n",
        "   plt.figure(2)\n",
        "   plt.plot(scores)\n",
        "   plt.title('Testing Loss')\n",
        "   plt.figure(3)\n",
        "   plt.plot(actuals,'b-',label='actual')\n",
        "   plt.plot(predictions,'g-',label='prediction')\n",
        "   plt.title('Basic LSTM')\n",
        "   plt.legend()\n",
        "   plt.grid('on')\n",
        "   plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "cAGTdvXPUzL1",
        "colab_type": "code",
        "outputId": "9ebcdec3-0b26-48ff-dcee-fe1753cfabcd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1137
        }
      },
      "cell_type": "code",
      "source": [
        "LOOKBACK_WINDOW = 15\n",
        "PREDICT_WINDOW = 1\n",
        "BATCH_SIZE = 1 #e.g if 3, then trains with 3 samples/lookback windows, each with 15 timesteps and 2 features at once.\n",
        "HIDDEN = 32\n",
        "DROPOUT = 0.3\n",
        "OPTIMIZER = 'adam'\n",
        "EPOCHS = 3\n",
        " \n",
        "model,history,predictions,actuals,scores = basic_LSTM_model(BATCH_SIZE,HIDDEN,LOOKBACK_WINDOW,DROPOUT,OPTIMIZER,EPOCHS,input_mat,targets)\n",
        "evaluate_walk_forward_LSTM(model,history,predictions,actuals,scores)"
      ],
      "execution_count": 115,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:23: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "0.04802286624908447\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/matplotlib/cbook/__init__.py:424: MatplotlibDeprecationWarning: \n",
            "Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
            "  warn_deprecated(\"2.2\", \"Passing one of 'on', 'true', 'off', 'false' as a \"\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeEAAAFZCAYAAACv05cWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBBXCFAqFQqEUCSqEKRQKhUIpEv8fXvXLkl5oBFwAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFZCAYAAAC173eYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ntAEAMATRBgDAEEQbAABDEG0AAAzxf9YYim6byCpKAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAecAAAFZCAYAAACizedRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}