TCN_report.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/breakwa/CS-Notes/blob/master/TCN_report.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gt8-lqLPJafi"
      },
      "source": [
        "# config"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 165,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "34ghg7-VDv09",
        "outputId": "e762adfc-9490-4982-acc2-52472cfd0f8f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting git+https://github.com/philipperemy/keras-tcn.git@master\n",
            "  Cloning https://github.com/philipperemy/keras-tcn.git (to revision master) to /tmp/pip-req-build-5t14iuot\n",
            "  Running command git clone -q https://github.com/philipperemy/keras-tcn.git /tmp/pip-req-build-5t14iuot\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (from keras-tcn==3.4.4) (1.21.6)\n",
            "Requirement already satisfied: tensorflow in /usr/local/lib/python3.7/dist-packages (from keras-tcn==3.4.4) (2.8.2+zzzcolab20220527125636)\n",
            "Requirement already satisfied: tensorflow_addons in /usr/local/lib/python3.7/dist-packages (from keras-tcn==3.4.4) (0.17.1)\n",
            "Requirement already satisfied: six>=1.12.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (1.15.0)\n",
            "Requirement already satisfied: keras-preprocessing>=1.1.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (1.1.2)\n",
            "Requirement already satisfied: protobuf<3.20,>=3.9.2 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (3.17.3)\n",
            "Requirement already satisfied: tensorboard<2.9,>=2.8 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (2.8.0)\n",
            "Requirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (0.26.0)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (57.4.0)\n",
            "Requirement already satisfied: absl-py>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (1.1.0)\n",
            "Requirement already satisfied: google-pasta>=0.1.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (0.2.0)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (1.1.0)\n",
            "Requirement already satisfied: gast>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (0.5.3)\n",
            "Requirement already satisfied: wrapt>=1.11.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (1.14.1)\n",
            "Requirement already satisfied: tensorflow-estimator<2.9,>=2.8 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (2.8.0)\n",
            "Requirement already satisfied: keras<2.9,>=2.8.0rc0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (2.8.0)\n",
            "Requirement already satisfied: astunparse>=1.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (1.6.3)\n",
            "Requirement already satisfied: grpcio<2.0,>=1.24.3 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (1.46.3)\n",
            "Requirement already satisfied: h5py>=2.9.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (3.1.0)\n",
            "Requirement already satisfied: flatbuffers>=1.12 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (2.0)\n",
            "Requirement already satisfied: libclang>=9.0.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (14.0.1)\n",
            "Requirement already satisfied: typing-extensions>=3.6.6 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (4.1.1)\n",
            "Requirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.7/dist-packages (from tensorflow->keras-tcn==3.4.4) (3.3.0)\n",
            "Requirement already satisfied: wheel<1.0,>=0.23.0 in /usr/local/lib/python3.7/dist-packages (from astunparse>=1.6.0->tensorflow->keras-tcn==3.4.4) (0.37.1)\n",
            "Requirement already satisfied: cached-property in /usr/local/lib/python3.7/dist-packages (from h5py>=2.9.0->tensorflow->keras-tcn==3.4.4) (1.5.2)\n",
            "Requirement already satisfied: google-auth<3,>=1.6.3 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (1.35.0)\n",
            "Requirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (0.6.1)\n",
            "Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (1.8.1)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (3.3.7)\n",
            "Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (2.23.0)\n",
            "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (1.0.1)\n",
            "Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (0.4.6)\n",
            "Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (0.2.8)\n",
            "Requirement already satisfied: rsa<5,>=3.1.4 in /usr/local/lib/python3.7/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (4.8)\n",
            "Requirement already satisfied: cachetools<5.0,>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (4.2.4)\n",
            "Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.7/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (1.3.1)\n",
            "Requirement already satisfied: importlib-metadata>=4.4 in /usr/local/lib/python3.7/dist-packages (from markdown>=2.6.8->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (4.11.4)\n",
            "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata>=4.4->markdown>=2.6.8->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (3.8.0)\n",
            "Requirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /usr/local/lib/python3.7/dist-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (0.4.8)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (1.24.3)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (3.0.4)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (2022.6.15)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (2.10)\n",
            "Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.7/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.9,>=2.8->tensorflow->keras-tcn==3.4.4) (3.2.0)\n",
            "Requirement already satisfied: typeguard>=2.7 in /usr/local/lib/python3.7/dist-packages (from tensorflow_addons->keras-tcn==3.4.4) (2.7.1)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.7/dist-packages (from tensorflow_addons->keras-tcn==3.4.4) (21.3)\n",
            "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging->tensorflow_addons->keras-tcn==3.4.4) (3.0.9)\n"
          ]
        }
      ],
      "source": [
        "\n",
        "!pip install git+https://github.com/philipperemy/keras-tcn.git@master\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 166,
      "metadata": {
        "id": "-hAmB8ooDo6m"
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from tensorflow.keras import Sequential\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow import keras\n",
        "import tensorflow as tf\n",
        "from tcn.tcn import TCN"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n",
        "os.environ['KAGGLE_CONFIG_DIR'] = '/content/drive/MyDrive/kaggle/' #注意kaggle文件夹包含json文件\n",
        " \n",
        "os.chdir('/content/drive/MyDrive/kaggle/') #切换到kaggle文件夹"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iN6W-EfSq-jR",
        "outputId": "7e1dd359-9a3c-482d-8489-d27ecbb728bb"
      },
      "execution_count": 167,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def normalization(data):\n",
        "    _range = np.max(data) - np.min(data)\n",
        "    return (data - np.min(data)) / _range\n",
        " \n",
        " \n",
        "def standardization(data):\n",
        "    mu = np.mean(data, axis=0)\n",
        "    sigma = np.std(data, axis=0)\n",
        "    return (data - mu) / sigma"
      ],
      "metadata": {
        "id": "jkH_7dnaEFbW"
      },
      "execution_count": 168,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "milk = pd.read_csv('/content/drive/MyDrive/Network Analytics.csv', index_col=0, parse_dates=True)\n",
        "milk = normalization(milk)\n",
        "print(milk.head())\n",
        "milk = milk.values  # just keep np array here for simplicity."
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NMhJyeSK6FEv",
        "outputId": "aeb9d3db-8ada-4600-a382-8366040619c3"
      },
      "execution_count": 169,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                     OutboundUtilzation (%)\n",
            "Timestamp                                  \n",
            "2017-10-10 07:01:00                0.962485\n",
            "2017-10-10 07:21:00                0.991176\n",
            "2017-10-10 07:26:00                0.972936\n",
            "2017-10-10 07:31:00                0.980330\n",
            "2017-10-10 07:36:00                0.995308\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from keras.callbacks import History \n",
        "history = History()"
      ],
      "metadata": {
        "id": "7HZTz0ea6F64"
      },
      "execution_count": 170,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V0zFTtXVJgGX"
      },
      "source": [
        "# 1 step pred"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 171,
      "metadata": {
        "id": "qChc5U-0DgUB",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "65b98254-3155-4a02-c947-0c1a03542d22"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(25567, 64, 1)\n",
            "(25567, 1)\n",
            "[[0.51176079]\n",
            " [0.48188243]\n",
            " [0.87708541]\n",
            " ...\n",
            " [0.05389427]\n",
            " [0.35506167]\n",
            " [0.5888253 ]]\n",
            "Model: \"sequential_35\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " tcn_12 (TCN)                (None, 64)                91136     \n",
            "                                                                 \n",
            " dense_14 (Dense)            (None, 1)                 65        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 91,201\n",
            "Trainable params: 91,201\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Train...\n",
            "Epoch 1/20\n",
            "799/799 - 11s - loss: 0.1870 - 11s/epoch - 14ms/step\n",
            "Epoch 2/20\n",
            "799/799 - 10s - loss: 0.0734 - 10s/epoch - 12ms/step\n",
            "Epoch 3/20\n",
            "799/799 - 5s - loss: 0.0677 - 5s/epoch - 6ms/step\n",
            "Epoch 4/20\n",
            "799/799 - 5s - loss: 0.0654 - 5s/epoch - 6ms/step\n",
            "Epoch 5/20\n",
            "799/799 - 6s - loss: 0.0633 - 6s/epoch - 7ms/step\n",
            "Epoch 6/20\n",
            "799/799 - 5s - loss: 0.0615 - 5s/epoch - 6ms/step\n",
            "Epoch 7/20\n",
            "799/799 - 5s - loss: 0.0603 - 5s/epoch - 6ms/step\n",
            "Epoch 8/20\n",
            "799/799 - 5s - loss: 0.0595 - 5s/epoch - 6ms/step\n",
            "Epoch 9/20\n",
            "799/799 - 5s - loss: 0.0588 - 5s/epoch - 6ms/step\n",
            "Epoch 10/20\n",
            "799/799 - 5s - loss: 0.0586 - 5s/epoch - 6ms/step\n",
            "Epoch 11/20\n",
            "799/799 - 5s - loss: 0.0585 - 5s/epoch - 6ms/step\n",
            "Epoch 12/20\n",
            "799/799 - 5s - loss: 0.0581 - 5s/epoch - 6ms/step\n",
            "Epoch 13/20\n",
            "799/799 - 5s - loss: 0.0586 - 5s/epoch - 6ms/step\n",
            "Epoch 14/20\n",
            "799/799 - 5s - loss: 0.0580 - 5s/epoch - 6ms/step\n",
            "Epoch 15/20\n",
            "799/799 - 5s - loss: 0.0579 - 5s/epoch - 6ms/step\n",
            "Epoch 16/20\n",
            "799/799 - 5s - loss: 0.0582 - 5s/epoch - 6ms/step\n",
            "Epoch 17/20\n",
            "799/799 - 5s - loss: 0.0579 - 5s/epoch - 6ms/step\n",
            "Epoch 18/20\n",
            "799/799 - 5s - loss: 0.0576 - 5s/epoch - 6ms/step\n",
            "Epoch 19/20\n",
            "799/799 - 5s - loss: 0.0581 - 5s/epoch - 6ms/step\n",
            "Epoch 20/20\n",
            "799/799 - 6s - loss: 0.0575 - 6s/epoch - 7ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# https://datamarket.com/data/set/22ox/monthly-milk-production-pounds-per-cow-jan-62-dec-75#!ds=22ox&display=line\n",
        "\n",
        "\n",
        "##\n",
        "# It's a very naive (toy) example to show how to do time series forecasting.\n",
        "# - There are no training-testing sets here. Everything is training set for simplicity.\n",
        "# - There is no input/output normalization.\n",
        "# - The model is simple.\n",
        "##\n",
        "\n",
        "\n",
        "\n",
        "lookback_window = 64  # months.\n",
        "\n",
        "\n",
        "\n",
        "x, y = [], []\n",
        "for i in range(lookback_window, len(milk)):\n",
        "    x.append(milk[i - lookback_window:i])\n",
        "    y.append(milk[i])\n",
        "x = np.array(x)\n",
        "y = np.array(y)\n",
        "\n",
        "print(x.shape)\n",
        "print(y.shape)\n",
        "print(y)\n",
        "# noinspection PyArgumentEqualDefault\n",
        "model = Sequential([\n",
        "    TCN(input_shape=(lookback_window, 1),\n",
        "        kernel_size=2,\n",
        "        use_skip_connections=False,\n",
        "        use_batch_norm=False,\n",
        "        use_weight_norm=False,\n",
        "        use_layer_norm=False\n",
        "        ),\n",
        "    Dense(1, activation='linear')\n",
        "])\n",
        "\n",
        "model.summary()\n",
        "model.compile('adam', 'mae',)\n",
        "\n",
        "print('Train...')\n",
        "model.fit(x, y, epochs=20, verbose=2, callbacks=[history])\n",
        "\n",
        "p = model.predict(x)\n",
        "\n",
        "p = p[:100]\n",
        "y = y[:100]\n",
        "plt.plot(p)\n",
        "plt.plot(y)\n",
        "plt.title('Monthly Milk Production (in pounds)')\n",
        "plt.legend(['predicted', 'actual'])\n",
        "plt.show()\n",
        "\n",
        "\n",
        "plt.plot(history.history['loss'][:10])\n",
        "plt.title('loss')\n",
        "plt.legend(['loss',])\n",
        "plt.show()\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 172,
      "metadata": {
        "id": "BVcdH4YTFowB",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e7d5ba29-a133-4d6e-b702-3cd16e4a3fc7"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((25567, 64, 1), (100, 1))"
            ]
          },
          "metadata": {},
          "execution_count": 172
        }
      ],
      "source": [
        "x.shape,y.shape"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# try merge curve"
      ],
      "metadata": {
        "id": "lEyleMJQrpnc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "milk = pd.read_csv('/content/drive/MyDrive/Network Analytics.csv', index_col=0, parse_dates=True)\n",
        "milk = normalization(milk)\n",
        "print(milk.head())\n",
        "milk = milk.values  # just keep np array here for simplicity."
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "M2nA2JJLFc2W",
        "outputId": "ab1d533f-2e0b-4c8e-e083-21027a841da6"
      },
      "execution_count": 173,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                     OutboundUtilzation (%)\n",
            "Timestamp                                  \n",
            "2017-10-10 07:01:00                0.962485\n",
            "2017-10-10 07:21:00                0.991176\n",
            "2017-10-10 07:26:00                0.972936\n",
            "2017-10-10 07:31:00                0.980330\n",
            "2017-10-10 07:36:00                0.995308\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# https://datamarket.com/data/set/22ox/monthly-milk-production-pounds-per-cow-jan-62-dec-75#!ds=22ox&display=line\n",
        "\n",
        "\n",
        "##\n",
        "# It's a very naive (toy) example to show how to do time series forecasting.\n",
        "# - There are no training-testing sets here. Everything is training set for simplicity.\n",
        "# - There is no input/output normalization.\n",
        "# - The model is simple.\n",
        "##\n",
        "\n",
        "lookback_window = 64  # months.\n",
        "\n",
        "\n",
        "\n",
        "x, y = [], []\n",
        "for i in range(lookback_window, len(milk)):\n",
        "    x.append(milk[i - lookback_window:i])\n",
        "    y.append(milk[i])\n",
        "x = np.array(x)\n",
        "y = np.array(y)\n",
        "\n",
        "print(x.shape)\n",
        "print(y.shape)\n",
        "\n",
        "# noinspection PyArgumentEqualDefault\n",
        "model = Sequential([\n",
        "    TCN(input_shape=(lookback_window, 1),\n",
        "        kernel_size=4,\n",
        "        use_skip_connections=False,\n",
        "        use_batch_norm=False,\n",
        "        use_weight_norm=False,\n",
        "        use_layer_norm=False\n",
        "        ),\n",
        "    Dense(1, activation='linear')\n",
        "])\n",
        "\n",
        "model.summary()\n",
        "model.compile('adam', 'mae')\n",
        "\n",
        "print('Train...')\n",
        "model.fit(x, y, epochs=20, verbose=2, callbacks=[history])\n",
        "\n",
        "p = model.predict(x)\n",
        "\n",
        "p = p[:100]\n",
        "y = y[:100]\n",
        "plt.plot(p)\n",
        "plt.plot(y)\n",
        "plt.title('Monthly Milk Production (in pounds)')\n",
        "plt.legend(['predicted', 'actual'])\n",
        "plt.show()\n",
        "\n",
        "\n",
        "plt.plot(history.history['loss'][:10])\n",
        "plt.title('loss')\n",
        "plt.legend(['loss',])\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "sZ2xyi2hrwF9",
        "outputId": "9f6f733c-7fa5-4c4f-eea3-5713021bb6cf"
      },
      "execution_count": 174,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(25567, 64, 1)\n",
            "(25567, 1)\n",
            "Model: \"sequential_36\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " tcn_13 (TCN)                (None, 64)                181376    \n",
            "                                                                 \n",
            " dense_15 (Dense)            (None, 1)                 65        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 181,441\n",
            "Trainable params: 181,441\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Train...\n",
            "Epoch 1/20\n",
            "799/799 - 7s - loss: 0.1519 - 7s/epoch - 9ms/step\n",
            "Epoch 2/20\n",
            "799/799 - 5s - loss: 0.0749 - 5s/epoch - 7ms/step\n",
            "Epoch 3/20\n",
            "799/799 - 5s - loss: 0.0707 - 5s/epoch - 7ms/step\n",
            "Epoch 4/20\n",
            "799/799 - 5s - loss: 0.0631 - 5s/epoch - 7ms/step\n",
            "Epoch 5/20\n",
            "799/799 - 10s - loss: 0.0630 - 10s/epoch - 12ms/step\n",
            "Epoch 6/20\n",
            "799/799 - 5s - loss: 0.0624 - 5s/epoch - 7ms/step\n",
            "Epoch 7/20\n",
            "799/799 - 5s - loss: 0.0603 - 5s/epoch - 7ms/step\n",
            "Epoch 8/20\n",
            "799/799 - 5s - loss: 0.0597 - 5s/epoch - 7ms/step\n",
            "Epoch 9/20\n",
            "799/799 - 5s - loss: 0.0602 - 5s/epoch - 7ms/step\n",
            "Epoch 10/20\n",
            "799/799 - 5s - loss: 0.0592 - 5s/epoch - 7ms/step\n",
            "Epoch 11/20\n",
            "799/799 - 5s - loss: 0.0588 - 5s/epoch - 7ms/step\n",
            "Epoch 12/20\n",
            "799/799 - 5s - loss: 0.0587 - 5s/epoch - 7ms/step\n",
            "Epoch 13/20\n",
            "799/799 - 5s - loss: 0.0586 - 5s/epoch - 7ms/step\n",
            "Epoch 14/20\n",
            "799/799 - 5s - loss: 0.0586 - 5s/epoch - 7ms/step\n",
            "Epoch 15/20\n",
            "799/799 - 5s - loss: 0.0578 - 5s/epoch - 7ms/step\n",
            "Epoch 16/20\n",
            "799/799 - 5s - loss: 0.0577 - 5s/epoch - 7ms/step\n",
            "Epoch 17/20\n",
            "799/799 - 5s - loss: 0.0575 - 5s/epoch - 7ms/step\n",
            "Epoch 18/20\n",
            "799/799 - 6s - loss: 0.0576 - 6s/epoch - 7ms/step\n",
            "Epoch 19/20\n",
            "799/799 - 5s - loss: 0.0575 - 5s/epoch - 7ms/step\n",
            "Epoch 20/20\n",
            "799/799 - 5s - loss: 0.0573 - 5s/epoch - 7ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEICAYAAABPgw/pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9eXxbV533/z5abO3eHa+J0yZpnDRxmqQbKaW0UFpoS8uUpZRh+jCUGX7DAPOwFYZn2IaZwrAUfpRhOg8dKFspS6EshXTft6ylSZxms+N9tyVZu3SeP869kux4kW05ju3zfr3yknV1dXV0I330ud/zPd+vkFKi0Wg0msWPZaEHoNFoNJr8oAVdo9Folgha0DUajWaJoAVdo9Folgha0DUajWaJoAVdo9Folgha0JcBQggphFgzyWO3CCGenufXf1AI8TcTvd5UYzudzNc4st/7PBz7TUKI32TdDwohzpqP1zpdCCEeF0K8f5p9CoUQzUKIitM1rsWCFvTTiBCiRQgRE0KUj9u+1xCUhjy8xrRfiHxhjLlXCGHL2mY3tqUXOEgpr5ZS/nCOr9VgvF7Q+NcihLhtLsfMN0KIzwshfpy9LR/vfQq+DNye9VoeKeXxeXqtMwYpZRS4Gzij/v/PBLSgn35OADeZd4QQmwDXwg1nzgwBV2fdv9rYNl8USyk9qHP4L0KIq8bvkP0Ds1QRQpwPFEkpn1/osSwQPwX+RghRuNADOZPQgn76+RHw3qz7fwPck72DEKJICHGPEKJPCNEqhPisEMJiPHaLEOJpIcTXhBBDQogTQoirjce+DLwW+I7hYr+Tddg3CCGOCCGGhRB3CiHE+IEZ278+btsDQoh/msH7ee8E7yenqwYhxCVCiDYhxGXT7SulfA44AJwrhLhMCNEuhPiUEKIb+B/jsvwOIUSn8e+O7C+/EOITQogu47H3TTXeCcJEG4UQDwkhBoUQPUKIzxg/LJ8B3mmc+/3jjyWEsBj/l63GVcw9Qogi4zHzCuRvhBAnhRD9Qoh/nuIUXA08MW7c6bCREOIHxv/nH4QQASHEC0KIsyc57+Zrf8A4H11CiI9nPT7puRx/bmY6DiHEG4UKn4wYn1eR9dgaIcQTxmP9Qoifm49JKdtRxuGiKc7RskML+unnecAnhGgUQliBdwE/HrfP/w8UAWcBr0OJ5P/KevxC4DBQDnwV+L4QQkgp/xl4CviQcfn9oaznXAOcD2wG3gG8aYKx/RC4KevHoxx4A8oNTcZvgEuFEMVCiBLUD8pvpzkHp2AI4s+Av5JSPj7NvkIIsQPYCOw1NlcBpcAq4APAP6O+7FuAJuAC4LNZr/Vx4I3AWuM95jpOL/Aw8CegBlgDPCKl/BPwb8DPjXPfNMHTbzH+vR71f+sBvjNun0uAc4ArUFcgjZMMZRPqMzAV7wK+AJQAR1Ehmql4Pep8XAl8SghhnpdJz2WOTDgO4/P1a+NY5cAxYEfW874E7DSeV4f6XmRzyBiPxkAL+sJguto3oj6UHeYDWSL/aSllQErZAnwd+Ous57dKKf9bSplEiXA1sGKa17xdSjkspTwJPIb6co5BSvkiMIISE4xxPC6l7JniuBHgd8A7jX8PGNtmwtuB/wKuNsYwFf3AIPB/gduklI8Y21PA56SUUSllGLgZ+KKUsldK2YcSFPMcvgP4HynlK1LKUeDzMxjrNUC3lPLrUsqI8X/0Qo7PvRn4hpTyuJQyCHwaeJcYGyL6gpQyLKXcD+xncsEqBgLTvN79UsoXpZQJ4CdM8H8+ji9IKUellH8B/odMaHCqc5kLk43jzcABKeUvpZRx4A6gO+t5cdQPdI1xrsdP3gdQ50FjoAV9YfgR8G6UW7tn3GPlgB1ozdrWCtRm3U9/6KWUIeNPzzSvmf1FCU2x/w+B9xh/v8cY63Tcg/qBOiXckiMfBe6TUr6Sw77lUsoSKWWjlPLbWdv7pJTZPyQ1nHoOa7Ieaxv3WK7Uo5zkbJhoTDbG/hjn+v80BHineb1cj2Uy/pxkn6/JzmUuTDaOMf8PUlUKzB7DJ1EhmBeFEAfGh8ZQ7394BuNY8mhBXwCklK2oydE3oy45s+kn40xMVpLl4qc7/ByH92PgrUKIJqARFVKZjqfIXCXMJgXy7cD1QoiPzOK5JuPfdyennsNO4+8ulDBnP5bNKGMnqquy/m5DhUtyGcN4JhpTApjqCmgyXgbWzeJ5UzH+nJjna6pzOeZcCSGyz9V0jPl/MOZ10vellN1SylullDXA3wHfFWNTSxtRVzEaAy3oC8ffApcbl/xpjDDKfcCXhRBeIcQq4H9zapx9MnqYXHCmxZhsegnlzH9lhC+me44ErgWuk7Orx9yJCvN8RAjxwVk8fyJ+BnxWCFFhxGr/hcw5vA+4RQixQQjhAj437rn7gLcJIVyGgPxt1mO/B6qFEB81Jgu9QogLjcd6gAZzDmKSMf2TEGK1EMJDJuaemMX7+yNqfiWf/B/jPW9EzdmYk5BTncv9wEYhxBYhhIOZha/+YDz3bUbY6cNk/XgKId4uhKgz7g6hfjBTxmO1qDmT5ZrlMyFa0BcIKeUxKeWuSR7+R5TzOY5yvD9F5d3mwreAG4XKgPn2tHtPzA9Rk265hFsAkFIekFIemOXrYcT2rwBuE/nJo/9XYBfKyf4F2GNsQ0r5ICpe+yhqku7Rcc/9JhBDCfQPUXFfc5wB1NzHtahQwhHUZCLAL4zbASHEngnGdDfqnD6JukKLoP6vZ4yUcg8wkvVjkg+eQJ2PR4CvSSl3GtunOpevAl9ETRQfYQZXaFLKftTV2e3AAGpC9pmsXc4HXhBCBFFzMx/JyrN/N/BDIyddYyB0gwvNeIQQl6Ic2KpZOm7NaUAIcSXw/0kpr5/jcRpQPzD2WV4tnFaMlMn9wKVSyt6FHs+ZhBZ0zRiEEHbgXmC/lPKLCz0ezfyz2ARdMzk65KJJY+Q8D6MmOO9Y4OFoNJoZoh26RqPRLBG0Q9doNJolwoIVMSovL5cNDQ0L9fIajUazKNm9e3e/lHLC0sELJugNDQ3s2jVZ1p5Go9FoJkIIMenKZh1y0Wg0miWCFnSNRqNZImhB12g0miXCku/sotFoTi/xeJz29nYikZlWUdZk43A4qKurw2635/wcLegajSavtLe34/V6aWhoQJzaGEuTA1JKBgYGaG9vZ/Xq1Tk/T4dcNBpNXolEIpSVlWkxnwNCCMrKymZ8laMFXaPR5B0t5nNnNudwSQp6MiW576U2EsnUQg9Fo9FoThtLUtB3tw7xyV+9zPPHBxd6KBqNZhHz+OOPc8011wDwwAMPcPvtt0+67/DwMN/97ndn/Bqf//zn+drXvjbrMWazdARdSuhtBiAYjY+51Wg0mmySyeSMn3Pddddx2223Tfr4bAU9nywdQT/xBHz3Qug/Siim/rPMW41Gs3xoaWlh/fr13HzzzTQ2NnLjjTcSCoVoaGjgU5/6FFu3buUXv/gFO3fu5OKLL2br1q28/e1vJxgMAvCnP/2J9evXs3XrVn7960zL3x/84Ad86EMfAqCnp4cbbriBpqYmmpqaePbZZ7nttts4duwYW7Zs4ROf+AQA//Ef/8H555/P5s2b+dznMp0Ov/zlL7Nu3TouueQSDh8+nLf3vnTSFoN9xm03oZjqZxuOa0HXaBaSL/zuAAc7/Xk95oYaH5+7duOU+xw+fJjvf//77Nixg/e9731p51xWVsaePXvo7+/nbW97Gw8//DBut5uvfOUrfOMb3+CTn/wkt956K48++ihr1qzhne9854TH//CHP8zrXvc67r//fpLJJMFgkNtvv51XXnmFffv2AbBz506OHDnCiy++iJSS6667jieffBK32829997Lvn37SCQSbN26lW3btuXl3CwdQU8YvYxjo4QNZx7WDl2jWZbU19ezY8cOAN7znvfw7W+r9rqmQD///PMcPHgwvU8sFuPiiy+mubmZ1atXs3bt2vRz77rrrlOO/+ijj3LPPfcAYLVaKSoqYmhoaMw+O3fuZOfOnZx33nkABINBjhw5QiAQ4IYbbsDlcgEqlJMvlpCgG71io4G0M9chF41mYZnOSc8X41P+zPtutxtQC3fe+MY38rOf/WzMfqa7zgdSSj796U/zd3/3d2O233HH/DUDWzox9Ljp0IM6hq7RLHNOnjzJc889B8BPf/pTLrnkkjGPX3TRRTzzzDMcPXoUgNHRUV599VXWr19PS0sLx44dAzhF8E2uuOIK/vM//xNQE6wjIyN4vV4CgUB6nze96U3cfffd6dh8R0cHvb29XHrppfzmN78hHA4TCAT43e9+l7f3vXQEPWGsqIoGCcdUn1vzVqPRLC/OOecc7rzzThobGxkaGuKDH/zgmMcrKir4wQ9+wE033cTmzZvT4RaHw8Fdd93FW97yFrZu3UplZeWEx//Wt77FY489xqZNm9i2bRsHDx6krKyMHTt2cO655/KJT3yCK6+8kne/+91cfPHFbNq0iRtvvJFAIMDWrVt55zvfSVNTE1dffTXnn39+3t73gvUU3b59u8xrg4uHvwBPfwMu+zT/PPQWfvLCSd6xvY6v3tiUv9fQaDTTcujQIRobGxfs9VtaWrjmmmt45ZVXFmwM+WKicymE2C2l3D7R/kvQoQfSk6E65KLRaJYTS0fQ45ksl5DOctFoli0NDQ1Lwp3PhqUj6KZDjwUJ6SwXjUazDFk6gm469OxJUb2wSKPRLCOWjqBnOXRTyHXIRaPRLCeWjqCnHXogk4ce12mLGo1m+bB0BD3t0PXSf41GkxuPP/44zz777JyO4fF48jSauZOToAshrhJCHBZCHBVCnFI/UgixUgjxmBBirxDiZSHEm/M/1GnQK0U1Gs0MyYegn0lMK+hCCCtwJ3A1sAG4SQixYdxunwXuk1KeB7wLOP1FgcesFDUcejzJQi2c0mg0C8f111/Ptm3b2LhxY7q41p/+9Ce2bt1KU1MTV1xxBS0tLXzve9/jm9/8Jlu2bOGpp57illtu4Ze//GX6OKb7DgaDXHHFFWzdupVNmzbx29/+dkHe13TkUpzrAuColPI4gBDiXuCtwMGsfSTgM/4uAjrzOciciCtBl7EgsWQSp91GOJ4kmkjhsFtP+3A0Gg3w4G3Q/Zf8HrNqE1w9eecggLvvvpvS0lLC4TDnn38+b33rW7n11lt58sknWb16NYODg5SWlvL3f//3eDwePv7xjwPw/e9/f8LjORwO7r//fnw+H/39/Vx00UVcd911Z1zv1FwEvRZoy7rfDlw4bp/PAzuFEP8IuIE35GV0M8EonyuQuIhS4nbRMRwmFEtqQddolhnf/va3uf/++wFoa2vjrrvu4tJLL2X16tUAlJaWzuh4Uko+85nP8OSTT2KxWOjo6KCnp4eqqqq8j30u5Kt87k3AD6SUXxdCXAz8SAhxrpRyTJdmIcQHgA8ArFy5Mk8vbRCPgM0BiQhuIpR7CgxBT1DqLsjva2k0mtyYxknPB48//jgPP/wwzz33HC6Xi8suu4wtW7bQ3Nw87XNtNhuplJKtVCpFLBYD4Cc/+Ql9fX3s3r0bu91OQ0MDkUhkXt/HbMhlUrQDqM+6X2dsy+ZvgfsApJTPAQ6gfPyBpJR3SSm3Sym3V1RUzG7Ek5EIg1sd0y3ClHkKAZ3potEsN0ZGRigpKcHlctHc3Mzzzz9PJBLhySef5MSJEwAMDqoG8uNL3jY0NLB7925ANYWOx+PpY1ZWVmK323nsscdobW09ze8qN3IR9JeAtUKI1UKIAtSk5wPj9jkJXAEghGhECXpfPgc6JckEpBLgVr8hbiKUGa5crxbVaJYXV111FYlEgsbGRm677TYuuugiKioquOuuu3jb295GU1NTunPRtddey/3335+eFL311lt54oknaGpq4rnnnks3xLj55pvZtWsXmzZt4p577mH9+vUL+RYnZdqQi5QyIYT4EPBnwArcLaU8IIT4IrBLSvkA8DHgv4UQ/4SaIL1Fns70ErP9nEsJuocIpR4l6Dp1UaNZXhQWFvLggw9O+NjVV1895v66det4+eWXx2x7/vnn039/5StfAaC8vDzdMGM8ZgOLM4GcYuhSyj8Cfxy37V+y/j4I7Mjv0GaAkeGSHXIpd+uQi0ajWV4sjZWipkN3Zxx6mXboGo1mmbE0BD3t0JWgu0QkPSka0m3oNJrTjl7QN3dmcw6XhqCnHboRciGcnhSN6ElRjea04nA4GBgY0KI+B6SUDAwM4HA4ZvS8fOWhLyymQ8+eFHXrkItGsxDU1dXR3t5OX9/pS3RbijgcDurq6mb0nKUh6KZDL3ARtzhwCy3oGs1CYbfb0ysyNaeXpRFyMR26zUnM6sQrIjjsVgptFp2HrtFolg1LQ9BNh253ELW48FmjALgKrDptUaPRLBuWhqCnHbqDsHDhFeq+q8CmQy4ajWbZsDQEPe3QnYSFE69FCbrDbiGs29BpNJplwtIQ9CyHHsKBB+3QNRrN8mNpCHqWQx/FgQt131lg1YKu0WiWDUtD0LMcelA6cKYdulUvLNJoNMuGpSHoibBqbiEEAenAmVIO3aUdukajWUYsDUE3uxUB/qSDQhkGKXHYddqiRqNZPiwNQU+Ewe4EYCRViAUJsVHDoessF41GszxYGoKe5dBHEkb/0FhQZ7loNJplxdIQ9EQE7E6klAwlVdlcokGcdivRRIpUSld902g0S5+lIehxNSkaTaQISqPcZCyIs8AK6L6iGo1mebA0BN1w6KFYkiAqlq5CLkrQddhFo9EsB5aGoBsOPRxPMmo6dCPkArqvqEajWR4sDUE3HHo4lmCUTMjFVaDKvYd0PReNRrMMWBqCbjj0UCzboQfSIRft0DUazXJgaQh6IgJ2Q9DTMfRRHDrkotFolhFLQ9DjYbA5CceS40IuelJUo9EsH5aGoGc59BQWUjbnmJBLSKctajSaZcDiF3QplaDbnOll/tLuHpOHHtEOXaPRLAMWv6AnjNK5dkdmAVGhB6JZWS66notGo1kGLH5BjxvNLYwYOgCFXoiNpvPQdchFo9EsBxa/oGc5dHPy01LogVgQh92CEDrLRaPRLA8Wv6BnO/R4Uol4oReiAYQQOO26yYVGo1keLH5BH+PQEypuXqAmRUF1LdLFuTQazXJg8Qt6up+oKs7ltFuhQE2KArPqWvSHFw/xy+ePkUim8j1ajUajmTcWv6AnjJCL3UE4llS554XeMQ59RlkuA8e4+I9XEv79J7nyjif50yvdSKnrqWs0mjOfxS/o4x16geHQY6OQSuGcSdei0QHkT26kFD/b3b0I4O9/vJvbH2yet+FrNBpNvlj8gp6dh26GXAo9gIR4CFeuIZd4GH72Lhjp5FBqJXWWQf780Uu5cHUpTx7pn9e3oNFoNPlg0Ql6jz/CQwd7MhsSGYcejhshlwKP2mbUc8lpUvSBD0P7S7S9/g4eTzXhjvRgE7C+ykv7YEiHXTQazRlPToIuhLhKCHFYCHFUCHHbJPu8QwhxUAhxQAjx0/wOM8Ov9rRz6z27GBqNqQ3xTAw9k+ViCHo0iKMgB4eeSsGBX8P293G0/HI6ZRkWGYfRPupLXQSiCUbC8fl6SxqNRpMXphV0IYQVuBO4GtgA3CSE2DBun7XAp4EdUsqNwEfnYawAbFtZAsDetiG1IduhmzH0QtOhB3DlkoceGoBUAirW0+uP0iXL1HZ/O3UlLgDaBsP5fisajUaTV3Jx6BcAR6WUx6WUMeBe4K3j9rkVuFNKOQQgpezN7zAzbK4rxmYR7G41BD3boY8PuRgVF6fNcgl0qVtvFb2BKJ2moI+0U1+q6qu3DYXy/E40Go0mv+Qi6LVAW9b9dmNbNuuAdUKIZ4QQzwshrproQEKIDwghdgkhdvX19c1qwM4CKxtqfOxpHVYbEhNkuZQ0qG0DR3EW2KaPoQeNmLy3ir5AlGBhlbo/0kF9qenQtaBrNJozm3xNitqAtcBlwE3AfwshisfvJKW8S0q5XUq5vaKiYtYvtnVlCfvahtXCn3gYLDaSwkoskcJlt0HxSnAUQ9d+XAVW4klJfKpFQmMceoRCbxnYnDDSjs9hp9hl56QWdI1Gc4aTi6B3APVZ9+uMbdm0Aw9IKeNSyhPAqyiBnxe2riohHE/S3B04pRa6s8ACQkB1E3TtT1dcnNKlB7rVrWcFvYEolUUOKKoDfzsA9SUu2oZ0DF2j0ZzZ5CLoLwFrhRCrhRAFwLuAB8bt8xuUO0cIUY4KwRzP4zjHsG2Vmhjd3TqkHLqRgw7gNGqgU90EPQdw2ZQznzLTJdAFzlKwFdLrj1LpdUBRLYyo3636Uift2qFrNJoznGkFXUqZAD4E/Bk4BNwnpTwghPiiEOI6Y7c/AwNCiIPAY8AnpJQD8zXomiIHVT4He04OpR266cBdhiOnugmSMaqiLcA0fUUDPeCtRkpJXyBKhbdQOfSRjENvHwqTSulcdI1Gc+Ziy2UnKeUfgT+O2/YvWX9L4H8b/+YdIQTbVpUoh746PKYWutlHlOomACpHDwMN0zt0bxX+cIJYMkWltxASdWqyNBGjvtRFLJmiNxClqsgxz+9Oo9FoZseiWylqct7KYtqHwkQjo2DLCLrZR5TSs6HAQ4n/EADh+BSpi4Hu9IQoYDj0WkBCoDOd6aInRjUazZnMohV0M44eCAbBnmk/Z/YRxWKBqk0UDStBnzTkkkopJ27koANGDL1OPT7SQX2JkYuuBV2j0ZzBLFpB31hTRIHNQjgUJGV18MvdKlW+xGXP7FTdhGvwIBZSkwt6qB9kErzVaYde6SsEnyHo/g5qS5wIoRcXaTSaM5tFK+gFNguba4uIhEbZ0xnmN/s6+cfL17Cm0pPZqboJayLEaksXf2kfmfhAWTnofYZDz4RcgJE2Cm1WVngdc1r+PxyK8e1HjpDUE6sajWaeWLSCDirsIhIR+iIWvvPu8/jYlecghMjsYEyMvqNmkJ/vapt4cVHAWCXqqaLXH8Vht+AtNNrYOUvGpC7OxaH/+UA333joVV5uH571MTQajWYqFrWgX39eLUX2JBefU8s1m2tO3aH8HLA5eFNpD32BKI8c6jl1n3F1XCq9jsyPgq8O/Kagu2aeiz58Eh6/HVIpevzK/esFShqNZr5Y1ILeWO2jwpGi2OebeAerDSo3sDJ2hOoiBz954eSp+4xZJRpRKYsm43LRu/wRookZ9Cc9+AA8/u8wcIRuv4rP64lVjUYzXyxqQQdUCzq7c/LHq5uwdL3MO7fX8dSRfk4OjBPUYDe4ysBWQF8gqiZETYpqM4Je6kJK6ByO5D62iBG372umVwu6RqOZZxa/oCfCYJtisU91E0RHePc6iUXAz14a59ID3eCtBqA3EKXCkyXovlqIDEM0OLvURVPQe5vTDl3nsms0mvlicQt6MqEaU0zj0AEqg81cvn4Fv9jVRiyRNTlqrBKNxJMEIgkqfVk/DkVGTTJ/VhndmUyMZjl0M4auBV2j0cwXi1vQE8YEo61w8n0qGwEBvYe4+cKV9AdjPNqc1X/DXCXqz0pZNEmnLrazwufAbhUzS100BF32HqI/GMVmEXSNRKYu5avRaDSzZHELejzT3GJS7E4oPQt6D7JjTTlCwKEuv3oslYRgr0pZNBcVjZ8UBRhpx2oR1JW4TnHoUkru39vOvz946NR6MaZDHziKRSY5t7aIZErSNZM4vEaj0eTI4hP0aAAOP6j+TmTaz01JZSP0HqTAZmGF10HHsPG8UXOV6LhFRSbeakCkUxfPWeHloQM9/OvvDzIcijEQjPL3P97NP/18P//1xHFu/N6ztGcLftQPCEQqzirRwwWrSwEddtFoNPPD4hP0Z74F974bBo7l5tABVmyEweMQD1Nb4qTDzAVP56BXj63jYmK1g7cqnenypevP5frzavj+Mye49KuPceU3n+Sx5j4+8+b1/N/3bufkQIjrvvMMzx7rV8+PjBghH1gr2tlu1J/RJQQ0Gs18sPgE/fz3g8UGz31nZg5dpqD/VWqLnbQPG4Ka1Uu0NxDBahGUuQvGPrd8HRx/AmIhKryFfPXGJh78yGvZ3lBKfamLB/5xBx+49GzesGEFv/3QDkpcdt7/w10qXz0yAnXnA7BWdLClXjW41g5do9HMB4tP0L1V0HQT7P2JWokJ0zv0yg3qtucgdSVOuoYjqqZK9ipRf5RyTwEWixj73Nd9SrWie+Zb6U3rq3zcfcv5/OYfdrC+KrOo6awKDx++Yi2hWJKW3oAKufhqGC6sZr21nXJPIXUlTi3oGo1mXlh8gg7wmg9DMpYR2ekceulZYC2A3oPUljhJpKSaBM1aJdoXjI4Nt5g07ICNb4Nn7sj8gEyBWRyspdM4tqOITvsqzrF2YbGI2ZUQ0Gg0mhxYnIJevgYar4X2l9T96Ry61a7quvQeorZY7dsxFFaC7ioHq51ef3TshGg2V34JELDzs9MO7axyD0JAZ3dG0I9TzyrZAckE9aUu7dA1Gs28sDgFHeCSj2b+ns6hg5Hpcog6Y8VnuynoWatEKycT9KI6eO3H4OBv4cjDICcvgesssFJb7KS3z8h1dxRxMFFDAXEYamFlqYuhUJxAJJ7T29RoNJpcWbyCXrsNGl6r/p5q6b9JZSP426l1KCHtGA6nV4mOhOL0B6OsKnNP/vzX/CMUr4Kf/BX8ez187xJ46usT7rqm0sPQQJ+64yhiT3iF+ruvmfoSY8XpHGqrazQazUQsXkEHeMPn4azXq5or07FiIwDO4Vcpcxcoh260nmvuVguN1ld7J3++3QG3/B6uuh22vFv9iDzyRWh5+pRd11R4CI6o1MWwxcPL0Sr1QN8hVur+pBqNZp5Y3IJetx3e+5vcQy6QnhgdHugxBL2a5u4AAI1Vk5ThNSleCRd9EN78VXjvA6rWy4O3qRWnWayp9OBMBQHoizsI4WDUWQN9h9OCrqsuajSafLO4BX0mFNVDgQd6DlJb7OSy3h+pWPjGG2juDlDssrPCN0VNmPEUuOCNX4Sev8Cee8Y8tKbSgw8l2N0xdcxYyVrobabIZcfrsGmHrtFo8s7yEXQh0hOjjc4Rro/9AbnlJlixgeZuP+urvGPb1+XCxhtg5Wvg0S9BONNa7uwKDz4RQiLoDNvUxor10O8qf+cAACAASURBVP8qpJKsLD21JsxE3L+3PV1HXaPRaKZj+Qg6qAVGvQd5c//dAAxd8HFSKcnh7sCYBUI5IwRcfTuEBuGJr6Y3l7gLqLRHiFjc9ARiADjrNkIyCoPHWZlD6qI/Eueffr6fX+3pmPm4NBqA/T+HH9+40KPQnEaWn6CHB1nT9Tt+kLyStmQZbUMhQrEkjVNNiE5FdRNsersKu2SlM9Y4YgRw0e2P4C6w4li5XT3QsdtYXBQmlZo8/TEYSajbqE5v1EzAE18ds3p5Qo49CkcfgmDf6RmTZsFZZoKuJkaTBT6+m3grHcNhDnWpCdFZOXSTqk0QC6hKkAYV9ghDKSc9/ggrfA4VcinwQtuL1Je6iCVT6YJgExGMKkEfjc6gh6lm+XDwAbUuYiqMKqH0Hpj/8WjOCJaXoFdtApuT2I6PMYKH9qEQzd1+hIB1K2bp0CG9OCldGwYosYQZSrk41BVQgm6xQt02aH+Rs8pVvvuu1sFJD5kR9MTsx6VZukRHpnfeRpVQeg7O/3g0ZwTLS9BdpfCxQzgv/Qheh42OoTDNXQEaytw4C6yzP67XyDPPEnQvo/ilixP9o5nsmboLoOcAF9YWcFa5m+88enTSsIsp5KMxLeiaCYj4YbR38lXLUoK/U/2tHfqyYXkJOoCzBISgtthJx3A4neEyJ3w16tYs9gU4kgH8KCe+osjIk6+/AGQKW/c+PvKGtTR3B3jwle7xRwMygh7UIRfNeKRU4b1EBGLBifcJDahJeIDeQ6dvbJqpkRIO/V51SpsHlp+gG9SVODnSG6R1MDS3+DmAx1jan+XQrbEAIYsh6GYVx9pt6rbtRa7ZXMOaSg93PPyqKuU7DlPIQzrkohlPbFR12oLJhcEMtxTVQ28zpHQf2zOCkXb4+c3Tz3/MkmUr6LXFTloHQkg5zZL/XCj0QKEv49BTSUTUj82lOhRVmQ7dVQpla6H9JawWwUffsJYjvUF+/3LnKYfMOHQt6JpxRP2Zv0f7J97HnBBd+0aIjxLsOcrek0PzPzbN1HTuVbc1W+fl8MtW0OuMIlmQw5L/XPBWZRy68YVz+lQP0TErUOsvUGV/peTN51Zzzgov33rkiHLpyTj85h9g4FhayEPjG09rNJFsQZ/MoRuCvuaNADzx1BPc+L3nGBqNzfPgNFPStU91XDNqS+WbZSvotUYZXXeBNV1Sd054q8BvCHpkBABfcRmAynIxqTtfxTcHj2OxCD58xVqO942qPqSDJ2Dfj+H4YzrLRTM52Q59spCLvwMsdlhtVCTtPUgyJdnbpl36gtK5V6VP51J/ahYsX0E3Gl2cU+U9te3cbPDWZEIuhqCfv341X7r+3PRrAcqhQ7o5x7m16uqgLxCFiFE+IBrUWS6ayYmMDbnEkynVwzYbf4earC/0QkkDRYFXAdjTOoxmgZBSCXrNefP2EjkJuhDiKiHEYSHEUSHEbVPs91dCCCmE2J6/Ic4PpkNfX52HcAtkQi5SZjn0cv76olVja8RkLTAC8DrsAPjD8fTziAXTDj0ST5FI6gktTRbRkczfo7388/1/4dZ7do/dZ6RDNWYBqNxITfQEALtbtUNfMIZbITy0sIIuhLACdwJXAxuAm4QQGybYzwt8BHgh34OcD8rcBbz7wpXccF4OtdRzwVsNqbiq62IKs6Po1P2yFhgBeB2qeFcgksgU+IoGxoRaQnEdR9dkYTr0Ai+M9vFqT5BDXf6x+/jb030CwiXrWCm78NoS7G8f1gZhoejcp24X2KFfAByVUh6XUsaAe4G3TrDfl4CvAIuiPKAQgn+7YRPnN5Tm54DZi4umEnRILzAiGsRuteC0W/FH4uNCLhkR13F0zRjMEhNlZ0Gwj4HRKP3BKHFTqFMpNZ9jrI/odpyFTaT467OjhGJJDvcEJjmwZl7p3KvmNSpP8cN5IxdBrwXasu63G9vSCCG2AvVSyj9MdSAhxAeEELuEELv6+pZYwaD08v/ujIOaTNBrt4FMKVEHfE6bcuimoMcCY9IVdT0XzRiifkBASQOM9jIQjCElmdpAo33qatEIuRwTKwG4plqFW/ac1HH0BaFzr8pusc2g78IMmfOkqBDCAnwD+Nh0+0op75JSbpdSbq+oqJjrS59Z+LLquURGAKFy0yfCayxECqkcYp/Dbjh0w9lHVQy9wKb+e7RD14wh4lefLc8K5GhfOrW1e8S4OPYbi4qMkMuBSCVRaWMdJyn3FLJXx9FPP1KqkMs8hlsgN0HvAOqz7tcZ20y8wLnA40KIFuAi4IHFMDGaV7JXi0ZG1BfOMsnpdRphnrD6YnkdNvzhrBh6TGW5mPnrOtNFM4aoHxw+cFciIiMUoEos95jNUMwcdCPkcnwoyklLHbb+ZrauLGaPXmB0+hk8riazzwBBfwlYK4RYLYQoAN4FPGA+KKUckVKWSykbpJQNwPPAdVLKXfMy4jMVWyG4yjKCPlm4BVQ9GVATqIDPaScwgUM3SwbokItmDGmHrq5yS1Ehvq60QzcE3Qi5tPSP0uM8G3oOsnVVCS0DIQaCk5duno5dLYN6BfNM6TInRLfM68tMK+hSygTwIeDPwCHgPinlASHEF4UQ183r6BYb3mojhj6NoBd61WqxsBJ0r8OOPyuGLmMBw6Gbgq6/PJos0g5dCXq5UEYg7dD9HWBzKIMBtAyEiPhWQ6CT7bVqhfTemcbRD/0OogEi8STvuut5vvPo0fy8l+VC516wFkJF47y+TE4xdCnlH6WU66SUZ0spv2xs+xcp5QMT7HvZsnPnJmYu+nSCLoQKuxghF5/DNjYPPRokJaHCq0MumgkwQ3ruSgDKhZ8CmyUTQx8xFhUJwdBojJFwHLdXXRWeW2nHZhHsnknYJdANP38PPPZv+CNxEinJ00eXWFLDfNO5D6rOBVvBvL7Msl0pOi/k6tBBhV3GhFwSyKwYOqAdumZi0g69HFAOvbHKS3e2QzcmRFsGRgEo8qkJeoeMsrHGx56ZTIyaaZJ77iHsV5/ZA51+BnVdmNxIpZSgV89vuAW0oOcXbzUEe1QoZTpBd5WOmRSNJVNphy6SMewkMg5dx9A12aRj6Mqh19gCrCxzj3PoYwW9tNj4PMbDnLeyhJfbRzJ569MRNxqax4IU7LsHUEkbzx6bpNKjZizDLapFZXUToViC1/z7I/x23/w0f9eCnk+8VSq/3N+Rg0PPDrnYEaQyl9KAmzBehw2n3aoduiaDlBmHXuAmKhzU2oNUFzno9keQyYQK+xUpQT/RH8IioKzEFPQQ51R5CceTU/a0HUPc+KEo8FJ64H+woz6PzxzVgp4TZqvAolr2nhymcyRCkdM+Ly+lBT2fmIuLQH3hpiIr5OJ12PAQQSDTmQkeEcFbaMNdaGNUl9DVmMTDkEqkf/hHLMVUWVXf2lgixUhfh2p+YaQstg6MUlPsxF7oUc9PRHAZ7RbDuX6uTId+wfspDHXzFsvzVBc5eFoLem4Yxg1HCS8cH8AiYNuqknl5KS3o+cRc/g85hFxK0lkuPqedIqEujc1LZQ9h3IU23IXaoWuyMEvnGoZhgCLKLX6qjPmWoW5VhAtfJmVxdbkb7EbFz3gIV4GqH5S7oIfVbeN1BDxncavtD7xpwwraBsOcHAjN/T0tdcwV4M5iXjgxyLm1RemifPlGC3o+MXuLQm4hl0QE4mF8Dhs+DEE3HLrbFPQCGyGd5aIxMctKFKrPV2/KR6kcpqrImG/pa1WPF9UipeRE/yirylxZgh7GaTcceq5F30yHXuDhlVV/zUZLK28vbwHQLj0XDIcesfnY2zbMBfmqHzUBWtDzibsChHFKc8lyAQgN4nNkOXQj9ukRETyGQ9eLODRpshx6KiXpTHjwJYeoKlKCHRs4qR731TIciuOPJGgoG+vQnUbIJWejYDp0u5NXyq4iLq2sDb5EdZFDx9Fzwchee7kfYokUF6zWgr44sFgzJQByyXIBCA/ic9rxYbigIlVlQTl0K+5C2+JoQzfSDok8p7Ht+xn81+sgtQje/+nCXKtQ6GM4HKdfFuGMD1PpsSMESH8n2JzgLOGEkeEyVtAjaYceydmhZwR9JG6lkzLsgTZ2rCnnmWP9EzY512QRGYbCIl40motoQV9MmHH0XEIuAOEhvA4bPjEu5CIiuAvUpOgZ79CTcbjzInj2W/k7ZjwCj3xBLZkO9tI2GOI//txMarmLR5ZD7w9G6ZdFWEhhjw5T5i6kcmgPVK4HIWg1Bb3cDXajh248lJkUzVXQExlBD0YTdLICMdTCJWvKGQ7FOdjpn/r5y53wUDp+vr7KS7Fr/hYXaUHPN2amywxCLk67lRJhOnQl6KXWKBaLwF1gJXSm56EHe1SebcvT+Tvmvp9kmm77O/jvp45z52PH0nnVy5Z0DD0j6ACM9rHd00t9uBnOvRGAE32jWATUlzrHxtDTIZcZOnSbk9Fogh5rFQy18po1qrTAMzoffWrCw0hHMbtbh7hwHt05aEHPP7kKelbIRQhBhS1MCpF+folNhS/chbYzP8vF36luO/ZAKkXXSDj3DIqJSMbh6TvS4avUcBs7D/SolxgOz3W0i5sshz4QjDGAkR4b7OUa+TRJLLBJCfpzxwfYUOOj0GZVYRgYI+gzSlu0OcBiIRhNMGCvglA/lQUJKr2FHO8L5vMdLj3CQwQsXkKxJBesLpvXl9KCnm9WbABX+eS10E1Mh27MgJfZwoQtHrAVEhd2iqxq0Ye7wMZoLIGUZ3CowRT0qB/6X+X6O5/hS384OPvj7b8XRk7Clf8KQNfJo+ll7R1Dy1zQI0ZziwIvA8EofaZDD/ayI/woL7AJvFWMhOLsbh3i8nPUalKsNtUtJ5GV5TITh244/GA0wWCBkc01fJKqIgfd/tlXblwWRIbpi6vzN5/xc9CCnn+2/S/4yH41QToVdqdyTcbiohJLmFGLG4CIcFFkUQLmLrSRkqpZ9BmLKehA+MTz9Pij/H5/J7HELMacTMBTX1d1Lza9HewuOk8exWYRWAS0L3dBj/qNap0WBkZjDJkO/dADlMS6uC+2g0g8yZNH+khJuGx9Zea5dhfEw9itFmwWMbO0RSMGH4wmCDiNhmVDLazwOej1L4qukwtHeIiT4ULOqnCny3nMF1rQ843FCuaqvOnIqudSJEYJoJ4XEk68aUFXPwxn9MRooFNdkjuKiZxQPcL9kcTsUtoO3A9DJ+DST4AQyKI6Qn2tXHx2GdVFTh1yMeu4AP3BGFZXqSrF3Px7ElYnO1Pb6fFHeKy5lxKXnaa64sxz7c50TrmzwDqzGLotUygu5FLzPAy3ssJXmCkKpjkVKZHhYY4FbVw4z+EW0IK+sDhLMvVcRAi/VC5oFAceYQi6sarvjF5c5O9Usf/abdi6dqc3/+7lzimeNAknn1Xn5Zw3AxByVFEU7+XKjVXUljhpH1rmKxNNhw70B6OUeR0qxCdTDNa/iRAOOocjPP5qH69bV4HVIjLPtTvSE5xOu3UGaYuRjEOPJJDOMrC7YaiVKp+D4VA892MtN2KjiFSc3rhz3pb7Z6MFfSHJqufikUGGUsaXRjpxs4gcur9LlSyoOx/3yBHchHn9ORU8dKCHaGKGX/TQoFqgZbTva40XUy0GuHLDCuqKnTqGHhnJLPsPRinzFKQ7FyWM7JadB7sZHI3x+uxwC6RDLgCuGTn00JgYutdpVw2qjZALQK+Oo0+Msex/BM+8h1tAC/rC4ipN13NxpYIMptSXJpAqxCWVE3UXmg79DHZA/g7VJLvufCyk2GZv4b2vaSAQTfDkqzMMu4QHMxPGwD6/mwoxwgqXhboSJ93+SO5lX5ci0UA65DIwGqPcUwjeGvCswLvhDQD8ek8HFgGXrh3XiN3uTAu6w26dQQxdTYpKKQlGE8pklKyC4VaqipSg67DLJBhX4CPSjcf4Ls8nWtAXkqyQizMZYCDpIpFM4U85cErTSakPwRnr0KVU+eK+GqjdCsBrnWrRSbHLzh9mGnYJD6UXXbUPhdjn92BBvUZtiZOUzOpuvxwxS+cCA8EYZe5CeNO/wc2/wOty4im0MRKOc97KEkrc4xawjHPoM0pbtLuIxFOkJHgK7VC8CoZaWWG4Ti3ocLDTz+Vff3xs44+w6dDd+Bxa0Jc2Zk30RBR7KopfuhkOx/GnHBSmlEM3f9XP2MVFoQFIxlTIxVVKu7WWrdaj2K0WrtpYxUMHe2YWXw0NpR36o829dEljIsnfQV2JCkkt60wXY1I0Ek8SjCZUyKV8DVQ3AbDCpwT29edUnPpcm2PMpOhMHXogGgfAU2hVIZf4KFV2tdCrZzH/yLY+C20vzvkwjxzq4XjfKCf6s/Lysx26FvQljqtU1bYeaQfUr3jXcIRRHBQkzZCLiqHP9+KiXn+E7zx6ZOZ1OcwO88aCqH2ps1kXPwxScs3mGkZjSR4/PIP+k+Gh9KKr432jjNiNOPBIB7XFKiS1rDNdDIc+YLjAcs9YF26GQC47p/KUp2J3qgqfgNM+gxpBhqCbnbM8DpsKuQDeUDtOuzXToHox8oePwc7Pzvkw+9qUG/eHs76rRgx9WHp0yGXJY8aKB1UNa7900TUSJoATezIEUqazXOalUfRffgl3bIZ4hG8+/Cpf2/kqL7fPsBu831ie76slHEvyfOxsvIlBGGnjorNKKXHZefhQT27HSkQhPgpOlWrX44+Q9BqLWEbaqC5WYrVsM13iEXU1VOij3+g2VOYeO9F2doWHuhInG2smWNhmd41x6DlfOSWUoAcj6jPoLrCpkAsgzvDUxVRKTr0oLxmH/iOqbd8ckFKy3/ju+CPxzANGyMUv3Onv8nyiBX0hMQt0DRmCjpuukQij0qG6F8VG05Oi8+LQn/9PGG5lpO0Vfr1HfaD3t81U0I0vgq+ajuEQe1Nr1P32l7BZLaxd4U0XiZoWs7OLcV66RiIUF5eoMgr+DgptVlb4Cpdvpkt62X8RA6NK0MvHZU7cdvV6fvsPOxBCjH/2mElRl906s/K5RmEuMBx68Ur12HArK3yOM9Khp1KSS77yKD96vnXynQaOQSqu5oHmUNWzYzhMf1BdNQUiWec1PEQSK6LAg8Uywf9JntGCvpCY9VwGjwPKoXeOhBnFqLsRC1Jgs2C3ivy3oes/Ch27AHjhhWeIJlK4C6zsbx+Z2XECXSBU2eD2oTCHZT0pa6Gq6wIzSzU0UjjN89Ljj1Dlc6ruO4aDqi12Lt8Y+pjCXEo8ysZNfLoKbJR5JkmPyxJ0Z66TolKmJ0XTgl5oU4vn3BUqF73IQc8ZmLbYG4jSORKZeoFb3yF1K5MQ7J31a+1vy3xvxjj0yDBhq3feOhSNRwv6QjIu5GLG0INShRaIqsmVeSnQ9fLPQViQFhtdR/Zw6boKdqwpn4VD71RFtCxWOobDJLCR8takywHUziTVMO3QS0imJL2BqOrEU1QL/nbjeK7lG0OPGqJhFOYC1KRorpiCLmXuk6JGzF3F0LMEHYxMlxaqfEaD6jOs3tDJQRVeOtg1RXnf3kOZv/2zWAhnsL99mAKbKqkw3qEHLd7TMiEKWtAXlvEhF+mma4xDDwAqZpnXtEUplaCvfh1+92pq46387SWraaov5nj/KCOh+JjdA5H4JAdCfQmM1nsdQ2FsFoHVW5V2O7XFM0g1NHLycZbSH4ySTEnVK7Mo49DrSpx0jYSXZ1OFLIc+EIziKrCm01pzwu5UTjQZx2m3Ek/K6X9o080tXATGC7qRi15pNKgeDk3xOVkATEFvGwyPdc3Z9B5SV5iQCR/Ogn1tw2ys8eF12MZ+X8LDBMTpmRAFLegLi+nQh1oA8ONSMXTGO/Q810Q/+TwMtyI3v5O9kWo22jq4dG05W+rVZOT+rInRfW3DbPniQ+w9OTTxsfydalERKp2wutiB8FaqGukohw45ZqZkOXTzB6CqyKlSIsODEAtRW+wknpT0Bs68mO28k106dzQ2M3cOEza5mHZi1OwnanNkHLrpNksaYKSdao+6f6ZNjJqCDtDcFZh4p75mqDtf/T1Lh55IpvhL+whNdcV4HfZTHPqIdOuQy7LAalOr/hIRsDmwF7ro8UcIykwMHVRcNK9ZLi/fC3YXLzl3sCu0gmrZi4iNsqmuCCHGTozev6edZEry4onBiY8VMJb9o0S7rtilQjCGQzdzx3OKo2fF0E1xSDt0AH8ndeYPxHKMo49rbjE+w2VajAJbxMM4ci2hGzdDLi6CkQQWQbr8LsWrIJWg3qZ+iPM2MRrogd0/nPNh2gZD3FzwJKtEN4cmCrskompStGEHWAvA38Ftv3qZD/10z4xe52hfkHA8yZb6YnxOG/7w2Bj60GnKQQct6AuP6dIdxXgdNuJJeYpD9+Qzhh6PqIqGjddy/4ERWq0q/Yy+w/gcds6u8KQdejIl+eMr3QAcmKjNWDSgXKORg94xFFaO3FOp4r3xMNVFZqphLg59UH2x7K60Q19RVJj+wcDfnhH05RhHz3LoncNh9WM3E0yHnl0TPVeHbmS5uAttmQwaIxd9RVJdjeVN0HfdDb/78JwmKQGG+zv5suV7fNjx4MRt8vqPqBBU5Qbw1TDY3cK9L7XNuEqoaYCa6ovxFp7q0AdSbrw65LJMMDNdHEX4jMuyqMVw6MYX2FVgTS/qmDOv/gkiI8jN7+TR5h5KGjar7cZsf1NdMfvaRpBS8sKJAfoCUTyFNg50TpD9kpWDHkuk6AlE1OIfs1F2sBeH3UqFt5CO4Rxyx81l/0LQ7Y9gswjK3cakKMBIBzXG4qJlmeliOPSEzcPJwRBnVbhn9vysNnSuXNvQZTWIDkYTY2PBRi56aVx9DrpH8pTp0ms0RxmdW2u78sG9AGyzHedQ9wSC3tesbisbkb4auttVttlQKH7KPNJU7Gsbweew0VDmMmLohqCnkhDxM5Bw4tUOfZlgOnSnulwDkAWqPKoZcvEU5jHk0voMFHg5UHgePf4omzY1qUtxY7Z/S30RjtF2Ij++iUf2HMZpt3LzhSs53j96at5yVg5610gYKY2YeZagg5oYzclRhzKFuXpGIqzwOVTurunQR9pVWp67YHkKetQPBR46/DHiSamaP8+EdAw9jCPXRtFph+5idLygF9WBsGIbPk6Zu4CefM1rmJkn4UnCfDkQjiU5O6p+GFbGW2jp7icxfgLYnBAtW0M35XgiPbx2bTkArYO5967d3zZMU30xQgh8TntmUjQyAkj6ky5V/+Y0oAV9oXFmHLo5cWItNL6o85G2ONQKJQ08cngAIeD1jdVQvi79JWqqL+Zm68M4jz1I9OCfuKKxkm2rSpASmrvHTSyZTZx9NemYdp0ZcgEYNQS9JMdc9PBw+oqlaySSrkuCrVDlPKdTF5dpXXSjjsvxfiU2Z81Y0E2HHsKVcwx9rEN3Zwu61a4mFF/dqRYX5aOeSzwMg8fU36GBWR+mfSjENsurpIQNC0nWJo9zon+cSPc1Q9nZJC0FPNFtp8oyxKfetA6AloHcPl/hWJLDPYF0QoHXYcNvOnSzdK6OoS8j0iGX4nQ1NrejAAq8mUnRQmv+FhYNt0LJKh5t7mFLfbEqv1rZmL78XL/Cy7XW5wDYHN/PNZur2Vir+laeEkfPquPSbjjw9KQopDNd6oqddA5HSE2XaphVOrfHH6G6yJl5zFc7ZnHR6YyhN3f7efZM6GwfVbXQT/QpYVo9a0GP5N4oelwM/ZTQwcYboPcATc6e/GS59L8K0nDScxD0tr5hNosTDJ11HQBbLEdPzUfvPQiVjfxqdzsHgx4KSHC2S72H1vHiPwmvdI6QTMl0Zyivw04wmlBptWZhLtw65LJsSE+KZhy621yJF1WO2FNgI5ZIzb0OuJQwfJKQu5b97SNcYTZAqGxU4hwepqBrF3Win4B0ssN6kMvOqaSmyEGR087BrDh6Ipli34FDJApLwK5WbwphFIdyG5X+ghmHHkum6A9OE2MNq0qLUkq6/ZF08wRAXd77M7noHUPh07aQ5fYHm/ngj/ecesl+ujEc+on+UXwOG6Xjy+NOR7ZDN2Po04VcshYWBSOJU+uRbFCCeXny2fxMimYv9JmDoIdad1Mo4tjPvRZZVM9W6/Gxgh4PqwV9FY38z7MtFJbWA+CMdFPlc+Ts0M3sry0rlaCbpiwYTaTruAxLj54UXTaYIZesGLq70AYFniyHnqcSuqN9EA/RHFavefl6w0lXNKrbvsPwyq+IiwL+M3EdtaIPR7ANIQQba3xjHPrTR/vp6zzB8aiPtsEQHUMq66LAZlGX4q6yTC66OZE5lauWMh1D90cShGJJtUrUZJxDjyZS6eXv883R3iAj4Ti7WyfJxT9dGJUWT/SPsrrCM3G9lqnImhQ10xYjOYdcjBj6eKfpq4GVF7Ml8AT9wdjcTUfPgXSmUzqNdRYUdr4EgHfNDkTtNrbZjo/NdOl/FZD4fWs41OVn7dpz1HZ/JyvLXJzMMYb+8KEemuqK1JUupBMbApF4VrcinYe+fJggy8VTaDSaTqctGm3o5joxOqSKFD0z4KamyEFjtTH5WmkIes9f4MD99FVfxs7UNrXtxJMAbKzx0dwdSH9hH9jXSa1liG5Ked8PXuLVnkBauNWgV4xx6DBN7ng8DMkouErTTq8qO+RSVKdWzoaHOKtCNdM+PD6mPw9E4sl0eOfR5rml0c2Z8BA4ijjRPzrz+DmMW1iUY6/arJBLYPykqMnGG6gIHeVs0UFvYI6ZLr2H1JyOu3xOgl42tI9OSzXCuwJqt1GV6qGrsz3rdVSIcU9YmZrGdRlBbyhz5eTQ+4NR9rUNZ4wRpEMr/nAiqxa658yKoQshrhJCHBZCHBVC3DbB4/9bCHFQCPGyEOIRIcSq/A91iTImD90IuRSMc+jml2+uE6PDStAf6nJweWNlxuEV1aumvy99H0Z7qbz43XzspmuR7sosQS8ilkhxvG+UcCzJnw90U28fZt2adZzoH+UvHSNp4QbUxOg4hz5l3Dtr2X96lWh2PuylqwAAIABJREFUyKXaSK88+QLnrSzGIuDFltl/4XPlRP8oUkKB1ZJ7GeD5IKnq5id8K+kYDs88fg6ZhUWJSFYeem5L/6WxUnRCQW+8DongLZYX5t5NqveQygt3lc0+5CIlq8Ov0OI6V92vVeakPnwos8K49yBY7Dzc68VTaKNxzdlgsYG/g1VlbvoC0TGJCH94uYtb79k1puTEY829SAlXNGZqz3uzHXpWt6IzZum/EMIK3AlcDWwAbhJCbBi3215gu5RyM/BL4Kv5HuiSxcwIcZePDbkUescsLII8tKEzSgwciZVyRZarwGKByvXqQ17gwbb+Kq7eXINYfSmceAqkTNfXPtA5wkOHepCxUbyJIarqzubf3rYJgFWlrqz3tSIt6F6HHZ/DNrVDN91Y1rJ/c1ESAPUXKUE6/hheh53Gah8vTbZ6NY8c61P/B9efV8OxvlFacpwsyzsjJyGVoK9A1c2ZlaBnOXSHXX31w7k4dJuDcEKSkozNcjHxVROqOp83W1+gdy5x9MiIymSqbJyToMvBE5TIYQZLz1MbarYghYUtlmMcMksA9DVD+VqePjbMhatLsdlsqjerv5OGMnVuW7Nc+n272njoYA8PHexOb3vkUC9VPseY2vOmQw9ElENPWB3EsJ+W9nOQm0O/ADgqpTwupYwB9wJvzd5BSvmYlNJ8988Ddfkd5hKmajPcdC+seWP6193rMB26+vCZE1hzXlw03ErQVkLS5uLis8vGPmbG0de/JRNrXf1aCHZD/xHOqvDgsFs40Onnwd1Hucf5DaSwQMMO3rG9np+8/0Ju2bE6czxPpQq5GBOX01ZJNOu4ZC37r/RlxdDtDlj1Gjj2GADnN5Syt22IWGKeJip7m6HnAMd6RxEC3v/aswB4ZKHCLgNq0UurVKtyZyXoVrvKu46HEULgzKVRdDwMNsfYWugTsfF61lvaCHcenPm4TMwJ0Tk69MCRpwFI1Bg1WgrcpMobaRLHVBx94BiceIrRsnNpGQjxmjUq9xyfEvRVZeqHz6zjn0zJ9PzJfz15HCkl0USSp4708dY1FsTuH6Q/5z6n+g77jRh6xKbE/kwKudQCbVn3241tk/G3wINzGdSyQgg452qw2jJpi+ksl0weOuSha9FQK11iBY1V3vSkWBozjn7ujZltqy9VtyeewGoRnFPlY/eRNm5p+SRb5UHEDXdBwyUA7FhTPjbrwrNCZUgYq11rp6uLHs5y6P4Ipe4CCm3jxnjW66H/MPg7uWB1KZF4auIVrPngwU/Cj27gZO8AtcVO1q3wsrbSw6PNCxR2MWrmH4oq8ZnxoiJQn7VxjaJzWilq1HGBzHzOeFxNN5CSgvKTc/jqmytE0w59dldg0RPP4ZdOfKs2pbdZ67ex1XacP+xtQf7q/WC18XjtBwB4jWlufDVGyMUQdKO416EuP65oH/9Q1cy+k4Psah3iheODuGL9fOTkR+H3H01f/Y516MOErT6sFpGpfzPP5HVSVAjxHmA78B+TPP4BIcQuIcSuvr4Z9JlcJpi/7uOzXPLVtUgOt3I0XsaGidqTbX4HXPE5OPvyzLaS1aq5RMtTALymPMRnhv4P28RhOi//Nmx+++QvNm61aF2Jyh2fNNUwq1tR90hk4jolZ12mbo8/zvYGNffw0nzF0Uf7IdjDWR2/5WxjEvbyxkpeOD44eSnW+WTwONjdHPA7qfQWzj4ma3ekBd2Rq0PP7ic6yYpH4auh2XYOFd1PzP6qqfeQ+twXr1TZX7EAJGaeyVTY9RJ7U2tZWebJbKzdhk8G+LvBryI698C13+aRDjtl7gLOWWEkBxgO3Vtoo9xTkHboL54Y5OO2+/jE8Bf5qeOr3PfoCzz3cjM/Lfx3XMEW9Vzjc54RdBVDH7Wo0rkzzkiaJbkIegdQn3W/ztj2/9o78/C2qjNxv0eLZcnyJst7vMeOs5DVIQkhBJJAgFACFFpamDJdpi2/oVMopUOndLr9pi0zXWY6pVDKUtqhhYHSEsJaAjQJEEISQhay2c7iNd53y5asM3+cK1m2ZVu2Zctx7vs8fqS7Sef4Xn33u986ACHEBuBbwLVSyqCubinlw1LKEillSXJykK7k5znZDhufKJnFJYVOZUP3uKDP3d8oeiLJRd4+aK2k3JPEvIz4odvtKbDma6oCpA8hlJZ+cge8dT93H7uVhaKcH9m+TuaaW0f+Pp9vwJdclKgSUwY00A1kkA09LT6IQE9dADYnlL1JSmw0uUk2dp+cpFBC7QZzXcczzHYq08/64lQ8XsmO4xFIMmoqA0c+Jxu7xmdu8THWrkWaht7eo25iMcNo6ADW4vUUeU7w29fHVq3Qz1mV6IMQ/dFfY03/d7US21bKXm+Rv9InAJklAHzMuIuXzZfTN3cz75Q1srIgqb81XFym+s11N5OTFMOpBqWh7y5vZK35MDgKKDEc577Tn+OmQ18ix1AHG3+ojtWuc4vJiMVk8NvQ25m6WugQmkB/HygUQuQJIaKAm4EtgTsIIZYAv0YJ8wjHdp27mI0G/v3GReQkxShNBaCn3X9BTEhDb6tCeD1UyBTmpQfR0Icj7xL1o3rrh3TkrGd9z09wXPjJ0TWOQdmi/bHow4SDdTcrc4BZ9adMDaahGwxKSy9/C6Rkea6DPaebRs9AHQ/dzbiT5pAp6lnv/hsAS7MTSLCZ2RYJs0tTOTjyVMjiWItyBRLQKNoWStcid9cADT12hJokeSuuxSgkB3e+4HcmD+YvH1TxwJulQzdI6c/cbOzo4T2fFBmrHb16PwLJKWvxQLNicjFE2emw53B3+6f4xbYT1La5+s0t4G/UQlsVOQ4bpxs7kVJSc+ojUmUDrLyd9s++xUkyyJRneXvZf8GCj6tjOvvFXmy02W9Db8U+ZVmiEIJAl1J6gDuAV4EjwP9KKQ8LIb4vhLhW2+0/ADvwjBBivxBiyzAfpxMqlv4CXVazESGguXMCiTSaja9SJvfHn4dC8SYo+RzctpX4zzzJ3Z/YwGdX545+3OACXYmjVEnUskR7PH00dvYOjHAJpOAy9eM5e5jleQ5autyUDiM8xo3bBZ5uqmZt4iNvDktOPw7ePkxGA+uKU3jxQA3P759Yl/gx0eeB5lO44nJp6uwNm4YebQ5VQ7fSEYKGTuYyvFGxXGI6xL88d3DIjba1y823nz/ET187NjSrtKNOKQ4p83ngzTJ+/rammY9ZoKsKi+2OCwauN5rg009j+9wWstOS+cUbJwBYXeDs38dfprmanKQYqltdfFTTxrye/Wp9/qU4suby7KJHWdX7AHPX3KCeGBEDSv3G+eq5dDfTIqcu7R9CtKFLKV+SUhZJKQuklP+mrftXKeUW7f0GKWWqlHKx9nftyJ+oMyoWn4begRCC1QVOnnzvDBVN4yxKpSUVkZgztrZl0XFwzc8hbw1CCG5YOiu046MTwGAeGos+okB3UKc1Gx621nf+Zeq1/E0uzFWP5cM23xgvWoZfTa+VBzybsbaVwxGlo/zL1XNZNCuBrz61n/+/9aOpKQfQWgFeD2fNSuDkOe2jHDACJqs/nX8sGnqHz4Y+knAymjHkXcJV1iO8d7KRp/dUDNj8yM5y2l0evBKe2zfohlh3GABvcjEvHayhWWpzHLNA30cVqSQmpQ3dlnsxBkcud11ehJSQER/td4ACAzT0XKda/9y+KlYbDuOJSYOk2QB8c9MCHrl9ozILGk0qCaqj/6ktNtpEd3cXuLto8tqmnclFJxIMKqH7448rjePuZz4MqZ/mG0fP8kzgD6rlNH0YSMqcHfahBsVg6A9dBBwxUUSbDcOHLnY1gTXBH7KYOpyGHp+pMgnL3iQnyUZyrCX8jlHNfl7RHcVO80XIpEJ49Vtw6m2cdgtP/sMKbluVwyM7T/Kl3+8N73cHQ4twOSWVkJq4hq6UAmsoGrrHFbxB9HAUXIa9u4rrsnv43guHVRRSdzMt9dU8tvMkV1+QxvLcRJ7ZWzHQQV5zAICD7gxq21y0Cc0sOMZIF1m1jw/68sgKzIkYxBXzUllT6OS6JZkDTYexaSqsU9PQAZ7fV8FFxo8wFqxVtn1UkMLS7MT+42JSoKM/yCPOakZ2qWuooc82ZWn/oAv06Yul34YOqpXbdz42j90nm3h0Z/mIh3b2ePj6Mwe47y+H/LWZextOUiMdFGc4JnXYAwjIFhVCkJlgHf4Jo7sZbA5qgiUVDSb/Mjj9DsLTw4W5jvAnGGkZfuUdUeSmxCFueFjFcP92E7zyL5i9PXxv8wK+tDafbUfrJmYKC4WAkEWDUM7zcTPIKTqWsMUB7eeGQ4uS+sEFdSRYo/jK796l75ErsD64jM19r3Hn+kJuWpZFeX0n+3x9anvaYdeDkLmMv5xwE2UycO0qleXZ1TIGl1xnA6K1gg+9+cxOGf4pRgjB7z+/gm9cWTxwg8GohLqW/g+Q1F2OgzZE3trhvzfgOgct0kW7hho81imLQQddoE9ffE7R3n778I3LZnHFvFR+8upxjgbrwKLxu3dP09TZS4/HyytaCzlXfTmVMnlAVtukE5AtCqob0u5TTcELOGmlc301tYM6RX3krAJPN9QfpSQ3kepWV3jro2sa+tFWEwXJMZC5FL78Niz/POx6QAl2r5eVecqhFnYb/mCaysFs42CrlSyHTRVAGy8BTlGr2Rhak+hg7eeGw5EPCdnEVu3kob9bxg1dT2NsPE5pXyo/ND9K0V9vY1OuF1uUkWf2aLVV/vbv0FGLd+P9vHywlrVFyXxsaS5t0kpVYP2V0dDs5wdl/viv87gMaK0kwRZFvNXMaoMyBflzMoIRULcIlOPY2KPyI2rd1imrtAi6QJ++BNjQfQgh+NENFxBjMfKLbSeCHtbZ4+Hh7WWsKXSSk2TjL5rzzth6hgpvcvAY9MkiwOQCsHFBGi1dbt4rH6RRS+m3oVc0q9KuI6ZKx2tRtB1nWajVoQ5roS5NoJd1mPwx6FjssOmncM1/QtVeKHvDrwWW1YVZoH/4FDSdxNPnZfMvd/LmO+9yrDeZVw6f9aelj5uAOPTQE4u0WuihCCYh1BPUye0sNp3hduMWnuu7mGt7f0D9Jf8GZ3YR89il/GN+HVsP1NBdc0Rp54tvYZ+3gNo2l6rBnxFHm4ijqb529O/0UbUPL4JyU8H4/0+pC1RXrzO7yE2yscpwmJ64XEjIGv4Ye7Jy1PuzRU3EudV1X+FJnH5OUZ0I4LOh9wwUVEl2C1cuSOdvx+rp8Qz9MT7x7imau9x87fIiNi/O5J2yRmobm4npbaDJkuEv8zkl2FNVyV6vGuclhclYzUZeOVwzcL+edvB6wObggzMtLJwVP7ImGKs5vNpr/E2jq8PRLceHv9OMXWnogSy+RTXF3vUrMhKsWEwGSsMp0Hva4c9fgr9+m+f2VfFhZSvzoxvxJubx6QuzuWPdBH0gZpuK4gF/6v+wyV5Sahq6Kp0btI5LMArWqQzhP3wCozWe1jXf454r55G87g740g6wObj9zNe4wv0Grc99XZmBNnyXFw/WEGUysH5uKkIIpNWBu70+9ESl6g+oNs4iOz2tP7Z8rGz4jlIY/vczLInvZKXhCObZl458zKCs6NhoMyla4+wq6dSdojqANUFFitQeGLLp8nkpdPb2sWuQptvR4+E328u5dE4yS7ITuW5xBlLC33Yrx51ImOIimPZU1X1Gi1SwRhm5rDiZVw+fHRjSpiWPuMxxHK5uZXnuKHb+GC1pqb2WZLsFs1FQHc4ORt3NSAy0Y+3X0H2YopTppWwbxoZj5Cfbw2tyaVTt1+Sxl3ni9fdZkmkn2VPD3PmL+cF1C0b/34xGoFNUi1ZyDVdx0dfcQqvlErItOO8SEAbVonDjj/jsFSV8eW2B2uacDV94HZGzip9FPURa/U4aS+7Ca0vmpYM1XFqU7BeA1oRkYmU7750MIdJFSmT1PvZ6cidmVrQmws1/gJ527qv7GrGiG0P+CPZz6L8eNcdobLSJWaIejyWRLqKx605RHQxGKLoSjr2s4pADuKjAidVs5PWPBia4PPGO0s7v3KD6IuYn21mUlcD+A+qmYE+boggXH4OyRQE2zk+jvr2n3yEGfhNHeYcFr2R0oWWKUvG/7TUYDIK0+OgwC/QWXCY7BoOR7KQgDshln1WVH997iNkp9mGTaMZFo0q6EV4Pqzpe558vsiO8bmWbDgdmG3jd0OfG6qu4OJwdPaC5RcdwpXODYXOoBLA5V6uSEoOxJiJufY6yvFvY4V3Iim0FfOLX73K2rYdNC9P9uyU600kS7Ww7EoJjtL0G0XGWDzx5zA+WCT0WUufB5gcwtWv2+9w1I+8/6DqPjTYzSzTQac3QlnUNXQdUUk93E1TsGrA62mxkTaGT14+c9T8ud/Z4eGSH0s59DWsBrl+cgbHtDACpOUVTN3YYki0KsK44hSijwe+sBfyhaQebDBgELMlOYFRi06FdfUZGvDXsGnq7iCUr0Tq0QBiouOOFn4APn2J+gpvK5u7RnYuh0liGRHBU5HNb9HZWxCvzD0kF4fl8X010d/foTS4CG0S7xiDQAW59Dj75pD/UbwhGMwW3/Yrie17nH9fP5WRDJ7EWE+vn9pd1NsY4cRo62Hb07OjtBqtUuYED3vzw+IkW3ACX3QcX3KRs5CMx6DqPizaRKRpoMqv1ulNURzF7vfoBHn1xyKYN81KpaXX528L9z67TNHe5+af1hQP2u2ZRBrmGOnqkmdn5YRIKoeLXXAamRV9c6OSVw7X9P1JNQ99TB3PT40KL241N8wv0TK0JddjobqaxL2bE0DdW3A6ebtZ2vIyUhE9LbyylIzqNx3rXk9VXgTjwtFofNg1dK43scRGtlWUe9mY0qP1cyDZ0UILcMLp4SY61cNflRbx97zrevOfSgTcNm4No2U1dU+vo/9/qfXgxctyQR1HqGDKhR2LtPfDxR0bfz3edd2omF4syudQKtV6PQ9dRRMWoiIGjW/0edB/rilMQQvU07O7t4zc7yllT6ByY8AA47RbWW49zVOSS5ZhAhuF4iBlqcgG4cn4alc3d/T1KNYG+q8Ybuo04QKBnJFipbXOFLWvT291MvSeauSPVvEmdB/mXMvvUHwFJWX14Gl/IxlIO96RQl3WV6iJ14H9Vdqc9SObjeAhocuGLKR820iWU9nNhItpsHOqw1wp0JdDB26Wj2NGrP6DCnEN2StLEwjrHg9WhEpK06zyBVqyil9NeVVZAj0PX6ad4E7ScgbOHBqx22i0szU7k9SNn+cPuMzR09A7RzgForSSv9wTOko+P3/M/Xix2FU/fMbBU8oZ5qRgNot/sogn0ml6rvyzuqMSmq1CxPg8ZCVb6vHLi/Sw13B1NNEs7xWmjPLoXbsTcWUOSaA9PpIuU9NWf4Jg7hZsumqce+5FKOw9B2w2JgEbRvsYpw2aL+trPaZmiUxmtAaia6MDcBA87S0eocCklsvoD9nnypjbPwoc/K1oJdEevuq5Le9QNSY9y0elnzlUqYiCY2WVuKoeq2vjlGydYlZ8UXLvVjstcOULt8snEkQeHnvUnfYAqA7Ay38GzeyuV/bariV5jDB5MlOSMQUOXXuisJyNB2YXDZkfXiiqNWsQsUUUNlcS3hScWvbMBk7ud02SwpsgJSz+j1jvyRj5uLPgFepe/GmHXsCYXpaH3EIVXTq2mCfgF+uoM2FXeOPwT2JldiO5mdvfmsiBzgg7R8RKT7Fdc7K5qAA51qbHoTlGdfmKckLVCmV0Gcfk8ZdJo7nLzlfXDRLAc3QrOOSpcLBLc8BswRsHjV6uIHY07NxRR2+ZSpVQ762kTcWQ5rMHroAcjIBY9pCbUoeL1Yna302mw++t5DIsWBrokNgQbbyhoES5RKYXERZth1nKYfz3MDWOtO79Ad/k1dNdwGroWttgpVSeqSGnoS5K8tLs8HKoOkh3d54YX78ZlTeP5vtWR0dBhQFZ0dKdK5jvcGY/ZKLBMoQlIF+jnAsWboPagqpjY26Vild0uCpLtzE6xc2Geg1X5SUOP62qCU2/D3Gumfsw+UubCF7ZB8hx46tMqCxIVmnj9kkx+s/0kPXWllPUlszxU7Rz6BXrHWdI1gR4Wx2hPGwa8WGKTMI5motI09DmWZsobOkMqmjYSTRWqBVvunMVqhRBw029h0Scn9LkD8NvQu0O2oVdquW0jFbyaFDSBXhynauW8Hczs8u4DUHeY1/O+TrcYxe8xmdhT/E5RY1sFrTKGdmxT2q0IdIF+bjDnavX6q5Xww3T476XwiyWID5/iqX+4kEduKwl+0Rx/FWSfuiFEkthU+PsXIWU+vPdr/+pvXlWM2Qju+hMcd6dQMpakmVgtXrm9BrvFRLzVHBaTi9Ts+faElNF3tsSC1UG2QWUzjru0sUZ16QF6pInlixdO6HNGJMDk4rehjxLlcrJN3aiGZM1ONlblT7H3tTE3PW6oQG8+BW/9GIqv4YWeJeQlxYwtEiec+MpceL3QcoZaw9RHuABEaPY6YyKpAC79pnqki8tUZpi9T8Bfvowz/SHIX6u09+ZTkLkMrvyxSr45uhViMyB9SaRnoCJ2ci6C/X9QETtCkBIXzT9fkox9ZycnZTqfCtUhCloEjRgQ6RIOgd7cVIcDSHKGINABEnNJ9qgxlNV3jK95s0bP2RNUG9LJT5lELTNAQw81bLG0qQ+r2UhGvHXyxhUMoxks8dDdxOqCJH636zQud5+y/UsJL92jEvCuup9DDx4PLX9hsrCnqoQtVwu0VFBvVDHoU22m0jX0c4VL71WNJi75Oiz7e2XGuOE3qpnxu79SJpmoGNjzKDx9izK3lG5T2nm4IiQmirNQNf5tq/av+lSBKu9bH5U5NM1+JIwmpRW1q7owmQnRYbGhV1apsaWlpY+yp0ZiDjFdKqNwIpEu3b19xHWdxhWXN7mP6L7EIk/oJpdjTR7yk2OmPkoKVOhiVyOrZzvp9XjZe1rLMK75EE68hnftP7PllKqzP+EM0YkQoyUfdZyFljM0m5VJcKodybqGfq5iMKhsxQU3AlJpKgB7fwsv3AkPrVElZiNtbgkkeY56bTiuGlUA5hZV6/v2GzaOXWAMikUfV+eimgPqxxinBHhdnbpBZGVkhHZ8Qg7GI1tJiTFNSKC/c+IsF1NLXcbV4/6MkAgIWzQbDZiNYlSTy5F6N0tzg/hopgJbEnQ1cmGeA5NBsLO0gdWznXjrjmIAPvd2Im81fUBBcgzXLg7xnE0GvmzRuiPg7qQ1Rl1PI1YNnQSmieqmM24Mhn5hDkp7v/FR6KiF6HjIvThiQxuCM0Cg+2gqA2Fk3rwLgh8zEvaBAr3N5fE39AgJT4+qbb7t+/5VzY3KsRXvCNXkkgNeN8uTXMEjXbx9ULl3SGLYYD44eBCL8JCWPz/k4Y+LgMQiGKVrkbsbabRQ2dozctbsZKJp6DEWE0uyE3intIFd5Y388ZW38EpBozmDBz69lNfuWuuPdooIPoFepQrhddvUzUU3uehMnAUfh79/CW56Qtkhpwv2FGUTrT/Wv66xFBJzxzfOQRo64O94FBLlb6mSp/VH/as6WrQkqOgQ7bGJuQAsiWultK5jaM2R0m3wyDrYepdymAVBSklV2UEAzCmTXG/HZAHEgK5FIwn0Ps1EMyZzWDixJflr/VxU4OTDylZufngXzt4qXDHpPP/VdWxamD56RNJk46v3UrkHAJd9FjD1TlFdoM9UsldAwWWRHsVAhIDkooEaemP5+AtPxaarULE+N5lactGY7OgfqcbPsrEUpKTX46Wvswm3waIaQYSCFotebGmizeWhvmNQtqovw3fv4/D8/xtSORPgxYM1xHWpAmq+RsSThhAD2tDZokwjJha5hSbQU6Y4wsVHgEDftDCdbIeNuy8v4vK0TmyphZGx6wcjOkHlW9TsB8AdqwT6VNvQdYGuM7U45/QLdCmVyWW8Qiw2DZDQUefX0KuaQxTofR7ksRfpxYToacPVUkt5QwexshNP1BiiJeKzAEGWQWn2ZXWDaro0HFc3nsvugw//SMMTf0dPT/9TRFevhx++eIRl9kakJa7fuTaZBAj06FFMLi4sGAQT75Q0XmwOcHeCu5ui1Fi2f+MyvrK+EEPzyfAVLAsHQvQ3urDEYbFPfdo/6AJdZ6pxFqpIgO4WFaHi7hr/D9Mfi15LSmw0RsMYGl2c3onobuZZj+oV+as/vcpH1W0kiA4MtjGET5qiIC6TZHd/6OIAGo6DswjW3sPJpd/EeeYlnnvoO/5uUw+9VUZ1q4u1Sa2IpILhy82GE7NtQBu6bvfw5XM7vWayHDZ/mYApR0su8jVJAVTtn+6m6SXQof9mnJDtd4bqTlGdmY0/0uWEP9V9Yho60F6D0SBIixtDo4sjL9BriOZJrgLg7MlD/OjloySKTsz2MXYFSszB2lmJLco4UKBLCfWaQAf+HH09O7wXsLHpf7jnf3ZQXt/BQ9vL2bw4g/iu05AUpLjaZBDYtchkILXzePD93F2095mYHSn7Ofh9FAMd6SoyatoJdJ9jNCHbbzvXTS46MxtNuNFwzN9ubcICvWOMddG9XuSRreyQi8kqWoQ0RnFNZif17T0km7sx2MYq0HMRLacpSLYPLKPbXqvi7rWb2O5TTfwp8Qs4RAeFpY9z7S/fxmQQfGt1LLRUhK+JxWiYov11Wtb2vsXPmr+i8hgGId0uWtwmCiIV4QKQvki9Vu/vX9d0Ur1O1f8rVOz9GrqvIJfdojtFdWYyCTnKeVR/TGnopmiV/ToeYpK13pW+SJcQk4sq30d01PJ8zzKuWjgL4chndUILNy6bRYqpS/VzHQsJOdBew5wk08Cqiw1aNI+zkF6Pl/0VLThmXwgLbuT2qFew9dTzjdXxpPz5JrDEwfwbxva948Vs82voy1xaN6wzu4bs1uvqoFNGRVZDtyYqLT2gWqdfQ/dp79MFn4Yen0VhSiyOmCgKp/hmqAt0nanFaFIaecMJ9cOcSK3/A4epAAAMgklEQVRvg1H9iHzZoomq0cWoRbKObKFPmNhpWKZaniXNxtBUyk9uWkRMX3voIYs+tCJdi2LbqWrp7m/p1nBCvTrncLi6FZfby/LcRFh3Hyb6eOuC17it9E5oPwu3PqsigKYCn1O0z83czvfVOi1+OhBPTycuLJGLcPGRscQfPQKo6yYusz9JaroQYHLJTrKx79uXT6gUxHjQBbrO1OMs0kwupRO3gw6KRVeNLlxIKensCeLskxJ55AV2sZCSohwVhZA0Wz3G93aq7FrrGJyi4A9dnBOtNbv2mV3qj0FULMSmseeU2laS61D1zUs+h+3E84jmU/DppyHrwnFNf1z4nKIVu7F6O+jAGlSge3u66JZRkYtB95G+WDV50cIX/YrAdEO7DnBOkS8kCLpA15l6nEWqkFjTyYnHXQc2i9ZCF+98aj8rf7SN+d95le+/8NHAZJ+GE4iW07zUu7i/w3zSbFVYqeaAWh6rQNc09ByD6p3qd4w2HFdatxDsPtVEnjOG5Fitzdrab0DhFXDzk5A3Slf5cGOOVgL9xGv0CRNP9m1QY+1uGbCb8HQjzVYSbFFTO77BZGjlhH1ml6by8Db9CBeFl8OXd6qS0RFCF+g6U0/yHNVtyOueuGMrQEMvTovFajZS3drNyvwkPrYog8fePsmPXz7aL9RLXwfgHbGYdcVaer/vplKpmR/GakO3p4HRQpK7BoOg3zGqhSxKKdlzqomSnIAbRYwTbnkGZm8Yz6wnhs/kcuI1quOWsKNvAQBnDr/NZx7bzfbjKqbe1OfCaouwdg79jtGa/eBqU8lk01FDFwLSxlHCIozoxbl0pp7AR9JwaOhdDeDpJT3eyuHv9Rf5klKSaDPz6+3lmIyCNYXJON99HqPMoKBwfn9atm88foE+Rg3dYICEbEytZ8h2bFCOUVebsu07iyir76S5yx16A+zJxmxT/Vjbq6kq/BoHzirhuPWlrWzvuobtx+v54po87qWXGPsobfimAmsiJOYpDd0fsjjNIlymCbqGrjP1JBUCWgLNRH+YPkeU1v4rMBVcCMF3Pzafm5dn8cCbZdz28HZmte7lRNwK7r1qTv9n2JJUITOtDseYBToos4s/dLEjwCFaxPunlO035AbYk43ZCl7lX6hPW0sbMZwig7ne42y5YzW3rszmiR3HMCCJjY1QB6DBZCyG6g+nbwz6NEHX0HWmnigbJGRBV7Mq2DUR4rSSqY2l6jMD6W7BYE3gh9dfwLyMOOZ17iZ6p5srrr0FUgI0TyHUk4LPMTjWKBdQDrHKPRQstLOjtAFvfaXSlpLn8P7BJpz2KPKmOOJhWExadEhCDu6EAuAA+70FbLIdwZwZz8JZCazLMcPzkJvmjOhQ/aQvhsN/7r/pTkcb+jRA19B1IkPGEkhfOPFU9+yVSqPe9eDA9dUfwE8K4Z3/xmAQfGZVLiWe/WC0QM7qoZ8TmKU5Hg09qQBcLSyNrqHX46W98iMwmCAxlz2nminJcUxpb8kR8YX7FW0kOykGi8lA8bJLMbsaoFU161iXr254yYkR7AIUSIbWdevwn5XPImqa3BynGbpA14kM1/5SRXhMFEssXPQVOPFqv4YtJbx8L/T1qp6TmtOU0tdVG7yoIM2OfbZ8YVBJPmNl4c0QHc9FpT8FJL21R8BRwNnOPs40dU0fcwv010QvvIKSXAeHvreR4hKtMqfvf6jVevHvG2l8jtH2at3cMgK6QNeJDNFx49OEg3HhF9VnvXW/Wj70J6jYBRffpZpYbPuBSq1vODZ8VIkv2iY6fnyJTjFJcOk3iaveyXrDPkzNZeAs5PfvngZgVUGEOv4EI+8SWPQp9QqYjQZIXaAyeKu0Zhzv/Jfad7pkY1oTlGMUdIE+AiFduUKIK4UQx4QQpUKIe4Nstwghnta2vyeEyA33QHV0hiVQSz+1E/76HUhbCOu+DStvh/1Pwo6fqn1nrw/+Gb5Il4ncZJZ/AZxFfDfqSeK6zlBryeHBv5Vx47JZke13OZiUYrj+Ia3ZhYbJov5nVXvh3V/Cvt/Bmq9D1vLIjXMwPrOLbj8fllEFuhDCCDwAXAXMAz4lhJg3aLfPA81SytnAz4H7wz1QHZ0R8Wnpf/gktFXCVfer0gCX3KNivvc+rtLFk4uDH+/T+sbjEPVhNMPGH5FFLUb6ePyYmdRYC//6scE/l2lK5jKo2A2vfRvmbYbLvhXpEQ3El2A03YpyTSNC0dAvBEqllOVSyl7gKWDzoH02A09o758F1otp4wHSOS+wxMKqO6C3A+Zfr2zloEw7676t3hesG94JGxWjBP5Yk4oGU7iBo7Hqu99pdfIfNy0iborbkI2bWSUq2StjCVz30Phr7EwWBetVQbaMpZEeybQllLDFTKAiYLkSWDHcPlJKjxCiFUgCGgJ3EkJ8EfgiQHZ29jiHrKMzDCu+DH1uKPnswPVLboWW07DgxpGP3/Bd1SFngnyw8D7++reHKVmxhtWzp0nYXygUbYQVt8PFdwZ3HEeatAVwT2mkRzGtEUOa2g7eQYgbgSullF/Qlv8OWCGlvCNgn0PaPpXacpm2T0OwzwQoKSmRe/bsCcMUdHSmF9Ut3Ty28yRfu6IIW5Se6qETXoQQe6WUJcG2hfJMVQUEZmzM0tYF3UcIYQLigUZ0dM5DMhKs3HfNPF2Y60w5oQj094FCIUSeECIKuBnYMmifLcBt2vsbgTfkaKq/jo6Ojk5YGVWF0GzidwCvAkbgMSnlYSHE94E9UsotwKPA74UQpUATSujr6Ojo6EwhIT0TSilfAl4atO5fA967gJvCOzQdHR0dnbEwzeKSdHR0dHTGiy7QdXR0dGYIukDX0dHRmSHoAl1HR0dnhqALdB0dHZ0ZwqiZopP2xULUA6fHebiTQWUFzhPOx3mfj3OG83Pe5+OcYezzzpFSJgfbEDGBPhGEEHuGS32dyZyP8z4f5wzn57zPxzlDeOetm1x0dHR0Zgi6QNfR0dGZIZyrAv3hSA8gQpyP8z4f5wzn57zPxzlDGOd9TtrQdXR0dHSGcq5q6Do6Ojo6g9AFuo6Ojs4M4ZwT6EKIK4UQx4QQpUKIeyM9nslACJElhHhTCPGREOKwEOKr2nqHEOKvQogT2usEWtRPT4QQRiHEB0KIrdpynhDiPe18P63V5J9RCCEShBDPCiGOCiGOCCFWnSfn+i7t+j4khPijECJ6pp1vIcRjQog6raubb13QcysUv9DmfkAIMebmqeeUQBdCGIEHgKuAecCnhBDnSEv1MeEB7pZSzgNWAv+ozfNeYJuUshDYpi3PNL4KHAlYvh/4uZRyNtAMfD4io5pc/gt4RUpZDCxCzX9Gn2shRCbwT0CJlHIBqtfCzcy88/1b4MpB64Y7t1cBhdrfF4EHx/pl55RABy4ESqWU5VLKXuApYHOExxR2pJQ1Usp92vt21A88EzXXJ7TdngCui8wIJwchxCxgE/CItiyAdcCz2i4zcc7xwCWoJjFIKXullC3M8HOtYQKsWttKG1DDDDvfUsrtqKY/gQx3bjcDv5OKXUCCECJ9LN93rgn0TKAiYLlSWzdjEULkAkuA94BUKWWNtqkWSI3QsCaL/wS+AXi15SSgRUrp0ZZn4vnOA+qBxzVT0yNCiBhm+LmWUlYBPwHOoAR5K7CXmX++YfhzO2H5dq4J9PMKIYQd+BNwp5SyLXCb1rN1xsScCiGuAeqklHsjPZYpxgQsBR6UUi4BOhlkXplp5xpAsxtvRt3QMoAYhpomZjzhPrfnmkCvArIClmdp62YcQggzSpg/KaV8Tlt91vcIpr3WRWp8k8Bq4FohxCmUKW0dyracoD2Sw8w835VApZTyPW35WZSAn8nnGmADcFJKWS+ldAPPoa6BmX6+YfhzO2H5dq4J9PeBQs0THoVyomyJ8JjCjmY7fhQ4IqX8WcCmLcBt2vvbgOenemyThZTym1LKWVLKXNR5fUNKeQvwJnCjttuMmjOAlLIWqBBCzNFWrQc+Ygafa40zwEohhE273n3zntHnW2O4c7sF+IwW7bISaA0wzYSGlPKc+gOuBo4DZcC3Ij2eSZrjxajHsAPAfu3vapRNeRtwAngdcER6rJM0/0uBrdr7fGA3UAo8A1giPb5JmO9iYI92vv8CJJ4P5xr4HnAUOAT8HrDMtPMN/BHlI3CjnsY+P9y5BQQqiq8MOIiKABrT9+mp/zo6OjozhHPN5KKjo6OjMwy6QNfR0dGZIegCXUdHR2eGoAt0HR0dnRmCLtB1dHR0Zgi6QNfR0dGZIegCXUdHR2eG8H8N15AtDsUqJQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# try to concatenate\n"
      ],
      "metadata": {
        "id": "tYL7AGwsFbBp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "milk = pd.read_csv('/content/drive/MyDrive/Network Analytics.csv', index_col=0, parse_dates=True)\n",
        "milk = normalization(milk)\n",
        "print(milk.head())\n",
        "milk = milk.values  # just keep np array here for simplicity."
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GiI4aguNGHaO",
        "outputId": "e1268982-fb91-424c-dd68-0dc26440a8a2"
      },
      "execution_count": 175,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                     OutboundUtilzation (%)\n",
            "Timestamp                                  \n",
            "2017-10-10 07:01:00                0.962485\n",
            "2017-10-10 07:21:00                0.991176\n",
            "2017-10-10 07:26:00                0.972936\n",
            "2017-10-10 07:31:00                0.980330\n",
            "2017-10-10 07:36:00                0.995308\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# https://datamarket.com/data/set/22ox/monthly-milk-production-pounds-per-cow-jan-62-dec-75#!ds=22ox&display=line\n",
        "\n",
        "\n",
        "##\n",
        "# It's a very naive (toy) example to show how to do time series forecasting.\n",
        "# - There are no training-testing sets here. Everything is training set for simplicity.\n",
        "# - There is no input/output normalization.\n",
        "# - The model is simple.\n",
        "##\n",
        "\n",
        "lookback_window = 64  # months.\n",
        "\n",
        "\n",
        "\n",
        "x, y = [], []\n",
        "for i in range(lookback_window, len(milk)):\n",
        "    x.append(milk[i - lookback_window:i])\n",
        "    y.append(milk[i])\n",
        "x = np.array(x)\n",
        "y = np.array(y)\n",
        "\n",
        "print(x.shape)\n",
        "print(y.shape)\n",
        "\n",
        "# noinspection PyArgumentEqualDefault\n",
        "\n",
        "input1_= tf.keras.Input(shape=(lookback_window, 1), name='input1')\n",
        "input2_ = tf.keras.Input((lookback_window, 1), name='input2')\n",
        "\n",
        "\n",
        "input1 = TCN(\n",
        "        kernel_size=4,\n",
        "        use_skip_connections=False,\n",
        "        use_batch_norm=False,\n",
        "        use_weight_norm=False,\n",
        "        use_layer_norm=False\n",
        "        )(input1_)\n",
        "large_size_model = Dense(1, activation='linear')(input1)\n",
        "input2 = TCN(\n",
        "        kernel_size=2,\n",
        "        use_skip_connections=False,\n",
        "        use_batch_norm=False,\n",
        "        use_weight_norm=False,\n",
        "        use_layer_norm=False\n",
        "        )(input2_)\n",
        "small_size_model = Dense(1, activation='linear')(input2)\n",
        "added = tf.keras.layers.concatenate([large_size_model, small_size_model])\n",
        "\n",
        "\n",
        "model_output = Dense(1, activation='linear')(added)\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "model =  tf.keras.models.Model(inputs=[input1_,input2_], outputs=model_output)\n",
        "\n",
        "model.summary()\n",
        "model.compile('adam', 'mae')\n",
        "\n",
        "print('Train...')\n",
        "model.fit([x,x], y, epochs=20, verbose=2, callbacks=[history])\n",
        "\n",
        "p = model.predict([x,x])\n",
        "\n",
        "p = p[:100]\n",
        "y = y[:100]\n",
        "plt.plot(p)\n",
        "plt.plot(y)\n",
        "plt.title('Monthly Milk Production (in pounds)')\n",
        "plt.legend(['predicted', 'actual'])\n",
        "plt.show()\n",
        "\n",
        "\n",
        "plt.plot(history.history['loss'][:10])\n",
        "plt.title('loss')\n",
        "plt.legend(['loss',])\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "lOnvha-BFeV5",
        "outputId": "84931017-93ba-4ae4-be9e-3fa923184afe"
      },
      "execution_count": 176,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(25567, 64, 1)\n",
            "(25567, 1)\n",
            "Model: \"model_2\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                   Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            " input1 (InputLayer)            [(None, 64, 1)]      0           []                               \n",
            "                                                                                                  \n",
            " input2 (InputLayer)            [(None, 64, 1)]      0           []                               \n",
            "                                                                                                  \n",
            " tcn_14 (TCN)                   (None, 64)           181376      ['input1[0][0]']                 \n",
            "                                                                                                  \n",
            " tcn_15 (TCN)                   (None, 64)           91136       ['input2[0][0]']                 \n",
            "                                                                                                  \n",
            " dense_16 (Dense)               (None, 1)            65          ['tcn_14[0][0]']                 \n",
            "                                                                                                  \n",
            " dense_17 (Dense)               (None, 1)            65          ['tcn_15[0][0]']                 \n",
            "                                                                                                  \n",
            " concatenate_2 (Concatenate)    (None, 2)            0           ['dense_16[0][0]',               \n",
            "                                                                  'dense_17[0][0]']               \n",
            "                                                                                                  \n",
            " dense_18 (Dense)               (None, 1)            3           ['concatenate_2[0][0]']          \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 272,645\n",
            "Trainable params: 272,645\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n",
            "Train...\n",
            "Epoch 1/20\n",
            "799/799 - 13s - loss: 0.1813 - 13s/epoch - 17ms/step\n",
            "Epoch 2/20\n",
            "799/799 - 9s - loss: 0.0690 - 9s/epoch - 12ms/step\n",
            "Epoch 3/20\n",
            "799/799 - 9s - loss: 0.0646 - 9s/epoch - 12ms/step\n",
            "Epoch 4/20\n",
            "799/799 - 9s - loss: 0.0613 - 9s/epoch - 12ms/step\n",
            "Epoch 5/20\n",
            "799/799 - 9s - loss: 0.0598 - 9s/epoch - 12ms/step\n",
            "Epoch 6/20\n",
            "799/799 - 9s - loss: 0.0589 - 9s/epoch - 11ms/step\n",
            "Epoch 7/20\n",
            "799/799 - 9s - loss: 0.0580 - 9s/epoch - 12ms/step\n",
            "Epoch 8/20\n",
            "799/799 - 9s - loss: 0.0572 - 9s/epoch - 12ms/step\n",
            "Epoch 9/20\n",
            "799/799 - 10s - loss: 0.0572 - 10s/epoch - 12ms/step\n",
            "Epoch 10/20\n",
            "799/799 - 9s - loss: 0.0569 - 9s/epoch - 11ms/step\n",
            "Epoch 11/20\n",
            "799/799 - 9s - loss: 0.0569 - 9s/epoch - 12ms/step\n",
            "Epoch 12/20\n",
            "799/799 - 9s - loss: 0.0565 - 9s/epoch - 11ms/step\n",
            "Epoch 13/20\n",
            "799/799 - 9s - loss: 0.0561 - 9s/epoch - 12ms/step\n",
            "Epoch 14/20\n",
            "799/799 - 9s - loss: 0.0562 - 9s/epoch - 12ms/step\n",
            "Epoch 15/20\n",
            "799/799 - 9s - loss: 0.0560 - 9s/epoch - 11ms/step\n",
            "Epoch 16/20\n",
            "799/799 - 9s - loss: 0.0556 - 9s/epoch - 11ms/step\n",
            "Epoch 17/20\n",
            "799/799 - 9s - loss: 0.0554 - 9s/epoch - 11ms/step\n",
            "Epoch 18/20\n",
            "799/799 - 9s - loss: 0.0551 - 9s/epoch - 11ms/step\n",
            "Epoch 19/20\n",
            "799/799 - 9s - loss: 0.0549 - 9s/epoch - 12ms/step\n",
            "Epoch 20/20\n",
            "799/799 - 9s - loss: 0.0542 - 9s/epoch - 12ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uLG0O8SJvarg"
      },
      "source": [
        "# 3 muti step pred test"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 177,
      "metadata": {
        "id": "GV7mQW4Svarn",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e3909972-4656-479b-d09f-158bcf591607"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                     OutboundUtilzation (%)\n",
            "Timestamp                                  \n",
            "2017-10-10 07:01:00                 96.2442\n",
            "2017-10-10 07:21:00                 99.1131\n",
            "2017-10-10 07:26:00                 97.2892\n",
            "2017-10-10 07:31:00                 98.0286\n",
            "2017-10-10 07:36:00                 99.5263\n",
            "(8539, 12, 1)\n",
            "(8539, 3)\n"
          ]
        }
      ],
      "source": [
        "\n",
        "milk = pd.read_csv('/content/drive/MyDrive/Network Analytics.csv', index_col=0, parse_dates=True)\n",
        "\n",
        "print(milk.head())\n",
        "\n",
        "lookback_window = 12  # months.\n",
        "pred_window = 3  # months.\n",
        "\n",
        "# milk = milk.values # just keep np array here for simplicity.\n",
        "\n",
        "x, y = [], []\n",
        "for i in range(lookback_window, len(milk)-pred_window,pred_window):\n",
        "    x.append(milk[i - lookback_window:i])\n",
        "    y.append(milk[i:i+pred_window])\n",
        "\n",
        "x,y\n",
        "x = np.array(x,dtype=object).astype(np.float32)\n",
        "y = np.array(y,dtype=object).astype(np.float32)\n",
        "y = y.squeeze()\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "np.set_printoptions(threshold=5)\n",
        "print(x.shape)\n",
        "print(y.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 178,
      "metadata": {
        "id": "Xsu5-Ryivuh0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "7a80105c-3900-4c08-8f73-6e5243d67c1d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_37\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " tcn_16 (TCN)                (None, 64)                91136     \n",
            "                                                                 \n",
            " dense_19 (Dense)            (None, 3)                 195       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 91,331\n",
            "Trainable params: 91,331\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Train...\n",
            "Epoch 1/100\n",
            "267/267 - 3s - loss: 27.3210 - 3s/epoch - 12ms/step\n",
            "Epoch 2/100\n",
            "267/267 - 2s - loss: 10.6117 - 2s/epoch - 6ms/step\n",
            "Epoch 3/100\n",
            "267/267 - 2s - loss: 10.2302 - 2s/epoch - 6ms/step\n",
            "Epoch 4/100\n",
            "267/267 - 2s - loss: 9.5603 - 2s/epoch - 6ms/step\n",
            "Epoch 5/100\n",
            "267/267 - 2s - loss: 9.2849 - 2s/epoch - 6ms/step\n",
            "Epoch 6/100\n",
            "267/267 - 2s - loss: 9.0120 - 2s/epoch - 6ms/step\n",
            "Epoch 7/100\n",
            "267/267 - 2s - loss: 8.8014 - 2s/epoch - 6ms/step\n",
            "Epoch 8/100\n",
            "267/267 - 2s - loss: 8.7372 - 2s/epoch - 6ms/step\n",
            "Epoch 9/100\n",
            "267/267 - 2s - loss: 8.6924 - 2s/epoch - 6ms/step\n",
            "Epoch 10/100\n",
            "267/267 - 2s - loss: 8.8315 - 2s/epoch - 6ms/step\n",
            "Epoch 11/100\n",
            "267/267 - 2s - loss: 8.7208 - 2s/epoch - 6ms/step\n",
            "Epoch 12/100\n",
            "267/267 - 2s - loss: 8.8769 - 2s/epoch - 6ms/step\n",
            "Epoch 13/100\n",
            "267/267 - 2s - loss: 8.5349 - 2s/epoch - 6ms/step\n",
            "Epoch 14/100\n",
            "267/267 - 2s - loss: 8.6645 - 2s/epoch - 6ms/step\n",
            "Epoch 15/100\n",
            "267/267 - 2s - loss: 8.3761 - 2s/epoch - 6ms/step\n",
            "Epoch 16/100\n",
            "267/267 - 2s - loss: 8.7095 - 2s/epoch - 6ms/step\n",
            "Epoch 17/100\n",
            "267/267 - 2s - loss: 8.4804 - 2s/epoch - 6ms/step\n",
            "Epoch 18/100\n",
            "267/267 - 2s - loss: 8.3403 - 2s/epoch - 6ms/step\n",
            "Epoch 19/100\n",
            "267/267 - 2s - loss: 8.4196 - 2s/epoch - 6ms/step\n",
            "Epoch 20/100\n",
            "267/267 - 2s - loss: 8.3883 - 2s/epoch - 6ms/step\n",
            "Epoch 21/100\n",
            "267/267 - 2s - loss: 8.3911 - 2s/epoch - 6ms/step\n",
            "Epoch 22/100\n",
            "267/267 - 2s - loss: 8.4466 - 2s/epoch - 6ms/step\n",
            "Epoch 23/100\n",
            "267/267 - 2s - loss: 8.3580 - 2s/epoch - 6ms/step\n",
            "Epoch 24/100\n",
            "267/267 - 2s - loss: 8.2890 - 2s/epoch - 6ms/step\n",
            "Epoch 25/100\n",
            "267/267 - 2s - loss: 8.2548 - 2s/epoch - 6ms/step\n",
            "Epoch 26/100\n",
            "267/267 - 2s - loss: 8.2946 - 2s/epoch - 6ms/step\n",
            "Epoch 27/100\n",
            "267/267 - 2s - loss: 8.3043 - 2s/epoch - 6ms/step\n",
            "Epoch 28/100\n",
            "267/267 - 2s - loss: 8.2807 - 2s/epoch - 6ms/step\n",
            "Epoch 29/100\n",
            "267/267 - 2s - loss: 8.3066 - 2s/epoch - 6ms/step\n",
            "Epoch 30/100\n"
          ]
        },
        {
          "output_type": "error",
          "ename": "KeyboardInterrupt",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-178-fb4cd9b6585d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     16\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     17\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'Train...'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 18\u001b[0;31m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepochs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m100\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/keras/utils/traceback_utils.py\u001b[0m in \u001b[0;36merror_handler\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     62\u001b[0m     \u001b[0mfiltered_tb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     63\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 64\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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     65\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;31m# pylint: disable=broad-except\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     66\u001b[0m       \u001b[0mfiltered_tb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_process_traceback_frames\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__traceback__\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[1;32m   1382\u001b[0m                 _r=1):\n\u001b[1;32m   1383\u001b[0m               \u001b[0mcallbacks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_train_batch_begin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1384\u001b[0;31m               \u001b[0mtmp_logs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0miterator\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   1385\u001b[0m               \u001b[0;32mif\u001b[0m \u001b[0mdata_handler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshould_sync\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1386\u001b[0m                 \u001b[0mcontext\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0masync_wait\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;32m/usr/local/lib/python3.7/dist-packages/tensorflow/python/util/traceback_utils.py\u001b[0m in \u001b[0;36merror_handler\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    148\u001b[0m     \u001b[0mfiltered_tb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    149\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 150\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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    151\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    152\u001b[0m       \u001b[0mfiltered_tb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_process_traceback_frames\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__traceback__\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwds)\u001b[0m\n\u001b[1;32m    913\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    914\u001b[0m       \u001b[0;32mwith\u001b[0m \u001b[0mOptionalXlaContext\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_jit_compile\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;32m--> 915\u001b[0;31m         \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\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    916\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    917\u001b[0m       \u001b[0mnew_tracing_count\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexperimental_get_tracing_count\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;32m/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m_call\u001b[0;34m(self, *args, **kwds)\u001b[0m\n\u001b[1;32m    945\u001b[0m       \u001b[0;31m# In this case we have created variables on the first call, so we run the\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    946\u001b[0m       \u001b[0;31m# defunned version which is guaranteed to never create variables.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 947\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_stateless_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# pylint: disable=not-callable\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    948\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_stateful_fn\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    949\u001b[0m       \u001b[0;31m# Release the lock early so that multiple threads can perform the call\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   2955\u001b[0m        filtered_flat_args) = self._maybe_define_function(args, kwargs)\n\u001b[1;32m   2956\u001b[0m     return graph_function._call_flat(\n\u001b[0;32m-> 2957\u001b[0;31m         filtered_flat_args, captured_inputs=graph_function.captured_inputs)  # pylint: disable=protected-access\n\u001b[0m\u001b[1;32m   2958\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2959\u001b[0m   \u001b[0;34m@\u001b[0m\u001b[0mproperty\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m_call_flat\u001b[0;34m(self, args, captured_inputs, cancellation_manager)\u001b[0m\n\u001b[1;32m   1852\u001b[0m       \u001b[0;31m# No tape is watching; skip to running the function.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1853\u001b[0m       return self._build_call_outputs(self._inference_function.call(\n\u001b[0;32m-> 1854\u001b[0;31m           ctx, args, cancellation_manager=cancellation_manager))\n\u001b[0m\u001b[1;32m   1855\u001b[0m     forward_backward = self._select_forward_and_backward_functions(\n\u001b[1;32m   1856\u001b[0m         \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36mcall\u001b[0;34m(self, ctx, args, cancellation_manager)\u001b[0m\n\u001b[1;32m    502\u001b[0m               \u001b[0minputs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    503\u001b[0m               \u001b[0mattrs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mattrs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 504\u001b[0;31m               ctx=ctx)\n\u001b[0m\u001b[1;32m    505\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    506\u001b[0m           outputs = execute.execute_with_cancellation(\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/execute.py\u001b[0m in \u001b[0;36mquick_execute\u001b[0;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[1;32m     53\u001b[0m     \u001b[0mctx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mensure_initialized\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[1;32m     54\u001b[0m     tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,\n\u001b[0;32m---> 55\u001b[0;31m                                         inputs, attrs, num_outputs)\n\u001b[0m\u001b[1;32m     56\u001b[0m   \u001b[0;32mexcept\u001b[0m \u001b[0mcore\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_NotOkStatusException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     57\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mname\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ],
      "source": [
        "\n",
        "# noinspection PyArgumentEqualDefault\n",
        "model = Sequential([\n",
        "    TCN(input_shape=(lookback_window, 1),\n",
        "        kernel_size=2,\n",
        "        use_skip_connections=False,\n",
        "        use_batch_norm=False,\n",
        "        use_weight_norm=False,\n",
        "        use_layer_norm=False\n",
        "        ),\n",
        "    Dense(pred_window, activation='linear')\n",
        "])\n",
        "\n",
        "model.summary()\n",
        "model.compile('adam', 'mae')\n",
        "\n",
        "print('Train...')\n",
        "model.fit(x, y, epochs=100, verbose=2)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yKZRoqq0vqEf"
      },
      "outputs": [],
      "source": [
        "\n",
        "p = model.predict(x)\n",
        "p = p.flatten()\n",
        "y = y.flatten()\n",
        "p = p[:100]\n",
        "y = y[:100]\n",
        "plt.plot(p)\n",
        "plt.plot(y)\n",
        "plt.title('Monthly Milk Production (in pounds)')\n",
        "plt.legend(['predicted', 'actual'])\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1pmCgKyXvarn"
      },
      "outputs": [],
      "source": [
        "x.shape,y.shape"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# TCN summary\n",
        "## 1.  训练集即测试集（仅仅是过拟合）\n",
        "## 2.  3-step-tcn 是将3步变成三维进行预测，从原理上有点小问题，即通过1-12预测13-15；但最好还是return_sequence = true,然后迭代预测，此时也是通过1-12预测13-15，没有新的点加入（即非自回归预测模型 ACF PACF）。"
      ],
      "metadata": {
        "id": "_0K-YV2fBcKj"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# lstm 1 step"
      ],
      "metadata": {
        "id": "ewXTKFw9_4Up"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![Imgur](https://i.imgur.com/FAGMvlc.png)\n",
        "https://en.wikipedia.org/wiki/Long_short-term_memory\n",
        "\n",
        "hidden state(also known as output of lstm units)restricted from minus one to one!!"
      ],
      "metadata": {
        "id": "l6g--2OLNDfb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "milk = pd.read_csv('/content/drive/MyDrive/Network Analytics.csv', index_col=0, parse_dates=True)\n",
        "milk = normalization(milk)\n",
        "print(milk.head())\n",
        "milk = milk.values  # just keep np array here for simplicity."
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AV3RauNGCbJT",
        "outputId": "0a581c59-2ef7-4ca3-f2d8-d3053444222e"
      },
      "execution_count": 179,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                     OutboundUtilzation (%)\n",
            "Timestamp                                  \n",
            "2017-10-10 07:01:00                0.962485\n",
            "2017-10-10 07:21:00                0.991176\n",
            "2017-10-10 07:26:00                0.972936\n",
            "2017-10-10 07:31:00                0.980330\n",
            "2017-10-10 07:36:00                0.995308\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "# milk = pd.read_csv('/content/drive/MyDrive/Network Analytics.csv', index_col=0, parse_dates=True)\n",
        "# milk = normalization(milk)  #数据标准归一化\n",
        "# print(milk.head())\n",
        "\n",
        "lookback_window = 64  # months.\n",
        "\n",
        "# milk = milk.values  # just keep np array here for simplicity.\n",
        "\n",
        "x, y = [], []\n",
        "for i in range(lookback_window, len(milk)):\n",
        "    x.append(milk[i - lookback_window:i])\n",
        "    y.append(milk[i])\n",
        "x = np.array(x)\n",
        "y = np.array(y)\n",
        "\n",
        "print(x.shape)\n",
        "print(y.shape)\n"
      ],
      "metadata": {
        "id": "3-1jZY5m_9pr",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7660d7c4-27fc-4c72-ff36-04d2400b87a9"
      },
      "execution_count": 180,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(25567, 64, 1)\n",
            "(25567, 1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y\n"
      ],
      "metadata": {
        "id": "ZXy_21_0Kf-m"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "# noinspection PyArgumentEqualDefault\n",
        "model = Sequential([tf.keras.layers.LSTM(1,input_shape=(lookback_window, 1)),\n",
        "      ])\n",
        "\n",
        "model.summary()\n",
        "model.compile('adam', 'mae')\n",
        "\n",
        "print('Train...')\n",
        "model.fit(x, y, epochs=20, verbose=2, callbacks=[history])\n"
      ],
      "metadata": {
        "id": "0DM4ZXetBsKB",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e58afc56-620c-48d0-eda1-4e998aa05322"
      },
      "execution_count": 181,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_38\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " lstm_38 (LSTM)              (None, 1)                 12        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 12\n",
            "Trainable params: 12\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Train...\n",
            "Epoch 1/20\n",
            "799/799 - 5s - loss: 0.4999 - 5s/epoch - 7ms/step\n",
            "Epoch 2/20\n",
            "799/799 - 3s - loss: 0.1915 - 3s/epoch - 4ms/step\n",
            "Epoch 3/20\n",
            "799/799 - 3s - loss: 0.1309 - 3s/epoch - 4ms/step\n",
            "Epoch 4/20\n",
            "799/799 - 3s - loss: 0.0943 - 3s/epoch - 4ms/step\n",
            "Epoch 5/20\n",
            "799/799 - 3s - loss: 0.0870 - 3s/epoch - 4ms/step\n",
            "Epoch 6/20\n",
            "799/799 - 3s - loss: 0.0830 - 3s/epoch - 4ms/step\n",
            "Epoch 7/20\n",
            "799/799 - 3s - loss: 0.0795 - 3s/epoch - 4ms/step\n",
            "Epoch 8/20\n",
            "799/799 - 3s - loss: 0.0765 - 3s/epoch - 4ms/step\n",
            "Epoch 9/20\n",
            "799/799 - 3s - loss: 0.0738 - 3s/epoch - 4ms/step\n",
            "Epoch 10/20\n",
            "799/799 - 3s - loss: 0.0715 - 3s/epoch - 4ms/step\n",
            "Epoch 11/20\n",
            "799/799 - 3s - loss: 0.0694 - 3s/epoch - 4ms/step\n",
            "Epoch 12/20\n",
            "799/799 - 3s - loss: 0.0677 - 3s/epoch - 4ms/step\n",
            "Epoch 13/20\n",
            "799/799 - 3s - loss: 0.0663 - 3s/epoch - 4ms/step\n",
            "Epoch 14/20\n",
            "799/799 - 3s - loss: 0.0651 - 3s/epoch - 4ms/step\n",
            "Epoch 15/20\n",
            "799/799 - 3s - loss: 0.0642 - 3s/epoch - 4ms/step\n",
            "Epoch 16/20\n",
            "799/799 - 3s - loss: 0.0635 - 3s/epoch - 4ms/step\n",
            "Epoch 17/20\n",
            "799/799 - 3s - loss: 0.0629 - 3s/epoch - 4ms/step\n",
            "Epoch 18/20\n",
            "799/799 - 3s - loss: 0.0625 - 3s/epoch - 4ms/step\n",
            "Epoch 19/20\n",
            "799/799 - 3s - loss: 0.0622 - 3s/epoch - 4ms/step\n",
            "Epoch 20/20\n",
            "799/799 - 3s - loss: 0.0620 - 3s/epoch - 4ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fdc5f34e650>"
            ]
          },
          "metadata": {},
          "execution_count": 181
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y[:100]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nJ7zo2n3LErX",
        "outputId": "2222aafd-a0af-4b4f-bacc-d7a5c221de97"
      },
      "execution_count": 182,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.51176079],\n",
              "       [0.48188243],\n",
              "       [0.87708541],\n",
              "       ...,\n",
              "       [0.16329294],\n",
              "       [0.36659919],\n",
              "       [0.43514831]])"
            ]
          },
          "metadata": {},
          "execution_count": 182
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "p = model.predict(x)\n",
        "p = p[:100,-1]\n",
        "y = y[:100]\n",
        "plt.plot(p)\n",
        "plt.plot(y)\n",
        "plt.title('Monthly Milk Production (in pounds)')\n",
        "plt.legend(['predicted', 'actual'])\n",
        "plt.show()\n",
        "\n",
        "\n",
        "plt.plot(history.history['loss'][:10])\n",
        "plt.title('loss')\n",
        "plt.legend(['loss',])\n",
        "plt.show()\n"
      ],
      "metadata": {
        "id": "6W87oa5CBx40",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 545
        },
        "outputId": "64fc51f0-e9c3-44b4-b298-5d01c09b41d9"
      },
      "execution_count": 183,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "因为只有12个可调参数，根本学习不到特征\n"
      ],
      "metadata": {
        "id": "4_SFlCBcDj9A"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# lstm 3 step"
      ],
      "metadata": {
        "id": "ArSGGCxyEklT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def normalization(data):\n",
        "    _range = np.max(data) - np.min(data)\n",
        "    return ((data - np.min(data)) / _range)"
      ],
      "metadata": {
        "id": "lixdbF27Lz0q"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JdycwSY-J_PS"
      },
      "outputs": [],
      "source": [
        "\n",
        "milk = pd.read_csv('/content/drive/MyDrive/Network Analytics.csv', index_col=0, parse_dates=True)\n",
        "milk = normalization(milk)  #数据标准归一化\n",
        "print(milk.head())\n",
        "\n",
        "lookback_window = 64  # months.\n",
        "pred_window = 3  # months.\n",
        "\n",
        "milk = milk.values # just keep np array here for simplicity.\n",
        "\n",
        "x, y = [], []\n",
        "for i in range(lookback_window, len(milk)-pred_window,pred_window):\n",
        "    x.append(milk[i - lookback_window:i])\n",
        "    y.append(milk[i:i+pred_window])\n",
        "\n",
        "x,y\n",
        "x = np.array(x,dtype=object).astype(np.float32)\n",
        "y = np.array(y,dtype=object).astype(np.float32)\n",
        "y = y.squeeze()\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "np.set_printoptions(threshold=5)\n",
        "print(x.shape)\n",
        "print(y.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "aWMoCLpoKDXx"
      },
      "outputs": [],
      "source": [
        "\n",
        "# noinspection PyArgumentEqualDefault\n",
        "model = Sequential([tf.keras.layers.LSTM(3,input_shape=(lookback_window, 1))\n",
        "\n",
        "])\n",
        "\n",
        "model.summary()\n",
        "model.compile('adam', 'mae')\n",
        "\n",
        "print('Train...')\n",
        "model.fit(x, y, epochs=100, verbose=2)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "cnCcmTMsKNDT"
      },
      "outputs": [],
      "source": [
        "\n",
        "p = model.predict(x)\n",
        "p = p.flatten()\n",
        "y = y.flatten()\n",
        "plt.plot(p)\n",
        "plt.plot(y)\n",
        "plt.title('Monthly Milk Production (in pounds)')\n",
        "plt.legend(['predicted', 'actual'])\n",
        "plt.show()"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "machine_shape": "hm",
      "name": "TCN_report.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "uLG0O8SJvarg"
      ],
      "mount_file_id": "18SGD7Wr29fEgcn8tVcF_W3SIITmFOMN4",
      "authorship_tag": "ABX9TyNsITzmTN6jaMPi6U6L9zy/",
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}