Notebooks/7. Full_CNN_Wavenet.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "Full CNN - Wavenet.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2sN0JP9KEABp",
        "colab_type": "text"
      },
      "source": [
        "# Full CNN predictiong S&P500 Daily Closing Price\n",
        "Full CNNs are becoming more commom to detect trends in NLP and time series modeling. A popular architecture is the WaveNet architecture proposed by Deep Mind in 2016. \n",
        "\n",
        "In WaveNet, every layer has kernel size 2, stride 1, causal padding, and using ReLU activation functions. However, the 2nd layer uses a dilation rate of 2. This means it skips every other input timestep. The next layer uses a dilation rate of 4 which means it skips every 3 input time steps. The fourth layer uses a dilation rate of 8 which means it skips every 7 timesteps out of 8. This patter continues which enables the lower layers to learn shorter term patterns while the deeper layers learn the longer term patterns.\n",
        "\n",
        "The last layer acts as a dense layer to output a single value."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GIenpda_d0c8",
        "colab_type": "text"
      },
      "source": [
        ""
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "vidayERjaO5q"
      },
      "source": [
        "## Setup"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "gqWabzlJ63nL",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "6424168c-1dbe-4ec5-ea35-81383312a4f1"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import tensorflow as tf\n",
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "keras = tf.keras\n",
        "\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "\n",
        "# set style of charts\n",
        "sns.set(style=\"darkgrid\")"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
            "  import pandas.util.testing as tm\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "cg1hfKCPldZG",
        "colab": {}
      },
      "source": [
        "def plot_series(time, series, format=\"-\", start=0, end=None, label=None):\n",
        "    plt.plot(time[start:end], series[start:end], format, label=label)\n",
        "    plt.xlabel(\"Time\")\n",
        "    plt.ylabel(\"Value\")\n",
        "    if label:\n",
        "        plt.legend(fontsize=14)\n",
        "    plt.grid(True)\n",
        "  \n",
        "\n",
        "def seq2seq_window_dataset(series, window_size, batch_size=128,\n",
        "                           shuffle_buffer=1000):\n",
        "    series = tf.expand_dims(series, axis=-1)\n",
        "    ds = tf.data.Dataset.from_tensor_slices(series)\n",
        "    ds = ds.window(window_size + 1, shift=1, drop_remainder=True)\n",
        "    ds = ds.flat_map(lambda w: w.batch(window_size + 1))\n",
        "    ds = ds.shuffle(shuffle_buffer)\n",
        "    ds = ds.map(lambda w: (w[:-1], w[1:]))\n",
        "    return ds.batch(batch_size).prefetch(1)\n",
        "  \n",
        "def sequential_window_dataset(series, window_size):\n",
        "    ds = tf.data.Dataset.from_tensor_slices(series)\n",
        "    ds = ds.window(window_size + 1, shift=window_size, drop_remainder=True)\n",
        "    ds = ds.flat_map(lambda window: window.batch(window_size + 1))\n",
        "    ds = ds.map(lambda window: (window[:-1], window[1:]))\n",
        "    return ds.batch(1).prefetch(1)\n",
        "\n",
        "\n",
        "def model_forecast(model, series, window_size):\n",
        "    ds = tf.data.Dataset.from_tensor_slices(series)\n",
        "    ds = ds.window(window_size, shift=1, drop_remainder=True)\n",
        "    ds = ds.flat_map(lambda w: w.batch(window_size))\n",
        "    ds = ds.batch(32).prefetch(1)\n",
        "    forecast = model.predict(ds)\n",
        "    return forecast"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Zmp1JXKxk9Vb",
        "colab": {}
      },
      "source": [
        "# Read in data\n",
        "spy = pd.read_csv('SPY.csv')\n",
        "\n",
        "# Convert series into datetime type\n",
        "spy['Date'] = pd.to_datetime(spy['Date'])\n",
        "\n",
        "# Save target series\n",
        "series = spy['Close']\n",
        "\n",
        "# Create train data set\n",
        "train_split_date = '2014-12-31'\n",
        "train_split_index = np.where(spy.Date == train_split_date)[0][0]\n",
        "x_train = spy.loc[spy['Date'] <= train_split_date]['Close']\n",
        "\n",
        "# Create test data set\n",
        "test_split_date = '2019-01-02'\n",
        "test_split_index = np.where(spy.Date == test_split_date)[0][0]\n",
        "x_test = spy.loc[spy['Date'] >= test_split_date]['Close']\n",
        "\n",
        "# Create valid data set\n",
        "valid_split_index = (train_split_index.max(),test_split_index.min())\n",
        "x_valid = spy.loc[(spy['Date'] < test_split_date) & (spy['Date'] > train_split_date)]['Close']"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9Vxt6avR8GEA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 285
        },
        "outputId": "42c9cd33-59e6-4bce-b260-a90e8940b544"
      },
      "source": [
        "# Plot all lines on one chart to see where one segment starts and another ends\n",
        "plt.plot(x_train, label = 'Train')\n",
        "plt.plot(x_valid, label = 'Validate')\n",
        "plt.plot(x_test, label = 'Test')\n",
        "plt.legend()\n",
        "print(x_train.index.max(),x_valid.index.min(),x_valid.index.max(),x_test.index.min(),x_test.index.max())"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "5521 5522 6527 6528 6949\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iWPZW0cX8Kgl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Reshape values\n",
        "x_train_values = x_train.values.reshape(-1, 1)\n",
        "x_valid_values = x_valid.values.reshape(-1, 1)\n",
        "x_test_values = x_test.values.reshape(-1, 1)\n",
        "\n",
        "#  Create Scaler Object\n",
        "x_train_scaler = MinMaxScaler(feature_range=(0, 1))\n",
        "\n",
        "# Fit x_train values\n",
        "normalized_x_train = x_train_scaler.fit_transform(x_train_values)\n",
        "\n",
        "# Fit x_valid values\n",
        "normalized_x_valid = x_train_scaler.transform(x_valid_values)\n",
        "\n",
        "# Fit x_test values\n",
        "normalized_x_test = x_train_scaler.transform(x_test_values)\n",
        "\n",
        "# All values normalized to training data\n",
        "spy_normalized_to_traindata = x_train_scaler.transform(series.values.reshape(-1, 1))\n",
        "\n",
        "# Example of how to iverse\n",
        "# inversed = scaler.inverse_transform(normalized_x_train).flatten()"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "kfPTmghd8OJe"
      },
      "source": [
        "## Fully Convolutional Forecasting - Wavenet"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "4-cPF5CX8OJf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "813b8274-4472-4988-a81c-39d146db60c5"
      },
      "source": [
        "keras.backend.clear_session()\n",
        "tf.random.set_seed(42)\n",
        "np.random.seed(42)\n",
        "\n",
        "window_size = 20\n",
        "train_set = seq2seq_window_dataset(normalized_x_train, window_size,\n",
        "                                   batch_size=128)\n",
        "\n",
        "model = keras.models.Sequential()\n",
        "model.add(keras.layers.InputLayer(input_shape=[None, 1]))\n",
        "for dilation_rate in (1, 2, 4, 8, 16, 32):\n",
        "    model.add(\n",
        "      keras.layers.Conv1D(filters=32,\n",
        "                          kernel_size=2,\n",
        "                          strides=1,\n",
        "                          dilation_rate=dilation_rate,\n",
        "                          padding=\"causal\",\n",
        "                          activation=\"relu\")\n",
        "    )\n",
        "\n",
        "model.add(keras.layers.Conv1D(filters=1, kernel_size=1))\n",
        "lr_schedule = keras.callbacks.LearningRateScheduler(\n",
        "    lambda epoch: 1e-4 * 10**(epoch / 30))\n",
        "\n",
        "optimizer = keras.optimizers.Nadam(lr=1e-4)\n",
        "\n",
        "model.compile(loss=keras.losses.Huber(),\n",
        "              optimizer=optimizer,\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "history = model.fit(train_set, epochs=100, callbacks=[lr_schedule])"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0801 - mae: 0.3453\n",
            "Epoch 2/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0578 - mae: 0.2904\n",
            "Epoch 3/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0255 - mae: 0.1873\n",
            "Epoch 4/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0107 - mae: 0.1092\n",
            "Epoch 5/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0086 - mae: 0.0990\n",
            "Epoch 6/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0063 - mae: 0.0821\n",
            "Epoch 7/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0043 - mae: 0.0657\n",
            "Epoch 8/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0028 - mae: 0.0499\n",
            "Epoch 9/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0016 - mae: 0.0344\n",
            "Epoch 10/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 8.7841e-04 - mae: 0.0231\n",
            "Epoch 11/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 5.4238e-04 - mae: 0.0181\n",
            "Epoch 12/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 3.4260e-04 - mae: 0.0148\n",
            "Epoch 13/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 2.1213e-04 - mae: 0.0120\n",
            "Epoch 14/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 1.5977e-04 - mae: 0.0112\n",
            "Epoch 15/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.0624e-04 - mae: 0.0093\n",
            "Epoch 16/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 1.9150e-04 - mae: 0.0117\n",
            "Epoch 17/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 8.7944e-05 - mae: 0.0093\n",
            "Epoch 18/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.6471e-04 - mae: 0.0110\n",
            "Epoch 19/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 7.3722e-05 - mae: 0.0088\n",
            "Epoch 20/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.7585e-04 - mae: 0.0107\n",
            "Epoch 21/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 9.7792e-05 - mae: 0.0100\n",
            "Epoch 22/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 2.8390e-04 - mae: 0.0118\n",
            "Epoch 23/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.2592e-04 - mae: 0.0112\n",
            "Epoch 24/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 5.4576e-04 - mae: 0.0160\n",
            "Epoch 25/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 1.3403e-04 - mae: 0.0117\n",
            "Epoch 26/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 5.8778e-05 - mae: 0.0079\n",
            "Epoch 27/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 3.9416e-04 - mae: 0.0147\n",
            "Epoch 28/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.5803e-04 - mae: 0.0124\n",
            "Epoch 29/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0011 - mae: 0.0263\n",
            "Epoch 30/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.3417e-04 - mae: 0.0116\n",
            "Epoch 31/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 5.3456e-05 - mae: 0.0077\n",
            "Epoch 32/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.2482e-04 - mae: 0.0095\n",
            "Epoch 33/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 4.9868e-04 - mae: 0.0168\n",
            "Epoch 34/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0018 - mae: 0.0344\n",
            "Epoch 35/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.2301e-04 - mae: 0.0118\n",
            "Epoch 36/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 1.1546e-04 - mae: 0.0105\n",
            "Epoch 37/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 6.8428e-04 - mae: 0.0172\n",
            "Epoch 38/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 2.2019e-04 - mae: 0.0146\n",
            "Epoch 39/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0030 - mae: 0.0316\n",
            "Epoch 40/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 5.2569e-04 - mae: 0.0214\n",
            "Epoch 41/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 9.9280e-04 - mae: 0.0251\n",
            "Epoch 42/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.3694e-04 - mae: 0.0125\n",
            "Epoch 43/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 2.5758e-04 - mae: 0.0126\n",
            "Epoch 44/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0019 - mae: 0.0289\n",
            "Epoch 45/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 8.5546e-04 - mae: 0.0238\n",
            "Epoch 46/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0030 - mae: 0.0507\n",
            "Epoch 47/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0016 - mae: 0.0340\n",
            "Epoch 48/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 7.3252e-04 - mae: 0.0249\n",
            "Epoch 49/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0023 - mae: 0.0422\n",
            "Epoch 50/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 1.7402e-04 - mae: 0.0137\n",
            "Epoch 51/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0022 - mae: 0.0353\n",
            "Epoch 52/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0019 - mae: 0.0364\n",
            "Epoch 53/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0033 - mae: 0.0503\n",
            "Epoch 54/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0039 - mae: 0.0484\n",
            "Epoch 55/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0034 - mae: 0.0585\n",
            "Epoch 56/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0013 - mae: 0.0313\n",
            "Epoch 57/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0017 - mae: 0.0353\n",
            "Epoch 58/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 4.1435e-04 - mae: 0.0197\n",
            "Epoch 59/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0023 - mae: 0.0401\n",
            "Epoch 60/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0037 - mae: 0.0528\n",
            "Epoch 61/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0041 - mae: 0.0608\n",
            "Epoch 62/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0028 - mae: 0.0441\n",
            "Epoch 63/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0029 - mae: 0.0448\n",
            "Epoch 64/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0026 - mae: 0.0483\n",
            "Epoch 65/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0024 - mae: 0.0441\n",
            "Epoch 66/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0016 - mae: 0.0306\n",
            "Epoch 67/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0067 - mae: 0.0746\n",
            "Epoch 68/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0050 - mae: 0.0673\n",
            "Epoch 69/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0099 - mae: 0.0778\n",
            "Epoch 70/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0139 - mae: 0.1331\n",
            "Epoch 71/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0220 - mae: 0.1656\n",
            "Epoch 72/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0221 - mae: 0.1660\n",
            "Epoch 73/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0218 - mae: 0.1640\n",
            "Epoch 74/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0219 - mae: 0.1659\n",
            "Epoch 75/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0214 - mae: 0.1634\n",
            "Epoch 76/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0216 - mae: 0.1628\n",
            "Epoch 77/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0216 - mae: 0.1628\n",
            "Epoch 78/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0212 - mae: 0.1616\n",
            "Epoch 79/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0210 - mae: 0.1608\n",
            "Epoch 80/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0209 - mae: 0.1597\n",
            "Epoch 81/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0206 - mae: 0.1588\n",
            "Epoch 82/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0202 - mae: 0.1571\n",
            "Epoch 83/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0200 - mae: 0.1554\n",
            "Epoch 84/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0197 - mae: 0.1538\n",
            "Epoch 85/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0198 - mae: 0.1535\n",
            "Epoch 86/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0192 - mae: 0.1520\n",
            "Epoch 87/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0187 - mae: 0.1491\n",
            "Epoch 88/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0187 - mae: 0.1480\n",
            "Epoch 89/100\n",
            "43/43 [==============================] - 1s 16ms/step - loss: 0.0182 - mae: 0.1466\n",
            "Epoch 90/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0179 - mae: 0.1453\n",
            "Epoch 91/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0176 - mae: 0.1424\n",
            "Epoch 92/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0170 - mae: 0.1417\n",
            "Epoch 93/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0173 - mae: 0.1420\n",
            "Epoch 94/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0167 - mae: 0.1398\n",
            "Epoch 95/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0164 - mae: 0.1381\n",
            "Epoch 96/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0161 - mae: 0.1367\n",
            "Epoch 97/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0159 - mae: 0.1357\n",
            "Epoch 98/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0157 - mae: 0.1348\n",
            "Epoch 99/100\n",
            "43/43 [==============================] - 1s 15ms/step - loss: 0.0155 - mae: 0.1329\n",
            "Epoch 100/100\n",
            "43/43 [==============================] - 1s 14ms/step - loss: 0.0150 - mae: 0.1321\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "GfWVZ8k-8OJf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "7bf89dbf-8a13-42af-aa7f-88fcbacc7fea"
      },
      "source": [
        "# Plot learning rate\n",
        "plt.semilogx(history.history[\"lr\"], history.history[\"loss\"])\n",
        "plt.axis([1e-4, 1e-1, 0, .1])"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.0001, 0.1, 0.0, 0.1)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "WVrxlzbk8OJg",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "47e4eb05-cb9f-4b3a-d5c3-596732f8d8e6"
      },
      "source": [
        "# Clear any background data that may have stored\n",
        "keras.backend.clear_session()\n",
        "tf.random.set_seed(42)\n",
        "np.random.seed(42)\n",
        "\n",
        "# Create input sequence batches\n",
        "window_size = 30\n",
        "train_set = seq2seq_window_dataset(normalized_x_train, window_size,\n",
        "                                   batch_size=128)\n",
        "valid_set = seq2seq_window_dataset(normalized_x_valid, window_size,\n",
        "                                   batch_size=128)\n",
        "\n",
        "# Create model object and set up wavenet architecture\n",
        "model = keras.models.Sequential()\n",
        "model.add(keras.layers.InputLayer(input_shape=[None, 1]))\n",
        "for dilation_rate in (1, 2, 4, 8, 16, 32):\n",
        "    model.add(\n",
        "      keras.layers.Conv1D(filters=128,\n",
        "                          kernel_size=2,\n",
        "                          strides=1,\n",
        "                          dilation_rate=dilation_rate,\n",
        "                          padding=\"causal\",\n",
        "                          activation=\"relu\")\n",
        "    )\n",
        "\n",
        "# Add final dense layer\n",
        "model.add(keras.layers.Conv1D(filters=1, kernel_size=1))\n",
        "\n",
        "# Set Nadam as optimizer\n",
        "optimizer = keras.optimizers.Nadam(lr=5e-4)\n",
        "\n",
        "# Compile model\n",
        "model.compile(loss=keras.losses.Huber(),\n",
        "              optimizer=optimizer,\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# save best model \n",
        "model_checkpoint = keras.callbacks.ModelCheckpoint(\n",
        "    \"my_checkpoint.h5\", save_best_only=True)\n",
        "\n",
        "# set early stopping\n",
        "early_stopping = keras.callbacks.EarlyStopping(patience=50)\n",
        "\n",
        "# sav history\n",
        "history = model.fit(train_set, epochs=500,\n",
        "                    validation_data=valid_set,\n",
        "                    callbacks=[early_stopping, model_checkpoint])"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/500\n",
            "43/43 [==============================] - 1s 31ms/step - loss: 0.0051 - mae: 0.0703 - val_loss: 0.0063 - val_mae: 0.1019\n",
            "Epoch 2/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 0.0013 - mae: 0.0258 - val_loss: 0.0023 - val_mae: 0.0593\n",
            "Epoch 3/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 1.4693e-04 - mae: 0.0122 - val_loss: 6.7912e-04 - val_mae: 0.0303\n",
            "Epoch 4/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 1.6410e-04 - mae: 0.0117 - val_loss: 0.0271 - val_mae: 0.2276\n",
            "Epoch 5/500\n",
            "43/43 [==============================] - 1s 21ms/step - loss: 0.0017 - mae: 0.0279 - val_loss: 0.0015 - val_mae: 0.0482\n",
            "Epoch 6/500\n",
            "43/43 [==============================] - 1s 21ms/step - loss: 7.2446e-05 - mae: 0.0090 - val_loss: 2.9009e-04 - val_mae: 0.0183\n",
            "Epoch 7/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 4.9042e-04 - mae: 0.0164 - val_loss: 0.0068 - val_mae: 0.1125\n",
            "Epoch 8/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 1.2063e-04 - mae: 0.0115 - val_loss: 7.7890e-04 - val_mae: 0.0335\n",
            "Epoch 9/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 4.1206e-04 - mae: 0.0156 - val_loss: 0.0383 - val_mae: 0.2709\n",
            "Epoch 10/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 0.0010 - mae: 0.0266 - val_loss: 0.0212 - val_mae: 0.1999\n",
            "Epoch 11/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 4.5591e-04 - mae: 0.0189 - val_loss: 0.0104 - val_mae: 0.1384\n",
            "Epoch 12/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 9.7139e-05 - mae: 0.0102 - val_loss: 6.6265e-04 - val_mae: 0.0305\n",
            "Epoch 13/500\n",
            "43/43 [==============================] - 1s 22ms/step - loss: 4.2417e-04 - mae: 0.0168 - val_loss: 0.0159 - val_mae: 0.1725\n",
            "Epoch 14/500\n",
            "43/43 [==============================] - 1s 22ms/step - loss: 1.3172e-04 - mae: 0.0117 - val_loss: 0.0057 - val_mae: 0.1013\n",
            "Epoch 15/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 1.8136e-04 - mae: 0.0127 - val_loss: 0.0142 - val_mae: 0.1628\n",
            "Epoch 16/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 8.5269e-05 - mae: 0.0097 - val_loss: 6.1685e-04 - val_mae: 0.0288\n",
            "Epoch 17/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 5.0830e-04 - mae: 0.0167 - val_loss: 0.0193 - val_mae: 0.1905\n",
            "Epoch 18/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 2.3995e-04 - mae: 0.0146 - val_loss: 0.0032 - val_mae: 0.0766\n",
            "Epoch 19/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 7.2669e-04 - mae: 0.0213 - val_loss: 0.0121 - val_mae: 0.1499\n",
            "Epoch 20/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 8.5001e-05 - mae: 0.0097 - val_loss: 0.0051 - val_mae: 0.0956\n",
            "Epoch 21/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 8.9331e-05 - mae: 0.0098 - val_loss: 8.9359e-04 - val_mae: 0.0371\n",
            "Epoch 22/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 6.0122e-04 - mae: 0.0177 - val_loss: 0.0058 - val_mae: 0.1052\n",
            "Epoch 23/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 3.4116e-04 - mae: 0.0168 - val_loss: 0.0039 - val_mae: 0.0852\n",
            "Epoch 24/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 7.2091e-05 - mae: 0.0090 - val_loss: 9.6271e-04 - val_mae: 0.0381\n",
            "Epoch 25/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 5.6756e-05 - mae: 0.0078 - val_loss: 6.4974e-04 - val_mae: 0.0304\n",
            "Epoch 26/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 7.7530e-05 - mae: 0.0086 - val_loss: 0.0021 - val_mae: 0.0616\n",
            "Epoch 27/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 8.2067e-04 - mae: 0.0212 - val_loss: 0.0191 - val_mae: 0.1894\n",
            "Epoch 28/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 2.6345e-04 - mae: 0.0149 - val_loss: 0.0142 - val_mae: 0.1629\n",
            "Epoch 29/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 9.8311e-05 - mae: 0.0104 - val_loss: 0.0042 - val_mae: 0.0856\n",
            "Epoch 30/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 3.1633e-04 - mae: 0.0143 - val_loss: 0.0181 - val_mae: 0.1840\n",
            "Epoch 31/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 1.0492e-04 - mae: 0.0107 - val_loss: 8.0268e-04 - val_mae: 0.0349\n",
            "Epoch 32/500\n",
            "43/43 [==============================] - 1s 21ms/step - loss: 1.6934e-04 - mae: 0.0114 - val_loss: 0.0038 - val_mae: 0.0848\n",
            "Epoch 33/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 4.4508e-04 - mae: 0.0180 - val_loss: 0.0157 - val_mae: 0.1712\n",
            "Epoch 34/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 1.0991e-04 - mae: 0.0108 - val_loss: 0.0010 - val_mae: 0.0404\n",
            "Epoch 35/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 8.5702e-05 - mae: 0.0090 - val_loss: 0.0070 - val_mae: 0.1123\n",
            "Epoch 36/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 5.2285e-05 - mae: 0.0076 - val_loss: 7.9207e-04 - val_mae: 0.0333\n",
            "Epoch 37/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 1.0056e-04 - mae: 0.0088 - val_loss: 0.0136 - val_mae: 0.1590\n",
            "Epoch 38/500\n",
            "43/43 [==============================] - 1s 21ms/step - loss: 1.8189e-04 - mae: 0.0125 - val_loss: 0.0033 - val_mae: 0.0784\n",
            "Epoch 39/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 6.6754e-05 - mae: 0.0086 - val_loss: 0.0012 - val_mae: 0.0421\n",
            "Epoch 40/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 6.3343e-05 - mae: 0.0076 - val_loss: 0.0078 - val_mae: 0.1191\n",
            "Epoch 41/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 5.4097e-05 - mae: 0.0078 - val_loss: 0.0012 - val_mae: 0.0417\n",
            "Epoch 42/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 2.0506e-04 - mae: 0.0106 - val_loss: 0.0374 - val_mae: 0.2665\n",
            "Epoch 43/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 8.3310e-04 - mae: 0.0233 - val_loss: 0.0015 - val_mae: 0.0473\n",
            "Epoch 44/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 5.4527e-05 - mae: 0.0078 - val_loss: 9.1504e-04 - val_mae: 0.0362\n",
            "Epoch 45/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 3.6202e-05 - mae: 0.0063 - val_loss: 0.0010 - val_mae: 0.0387\n",
            "Epoch 46/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 3.7259e-05 - mae: 0.0064 - val_loss: 0.0019 - val_mae: 0.0555\n",
            "Epoch 47/500\n",
            "43/43 [==============================] - 1s 21ms/step - loss: 7.8858e-05 - mae: 0.0084 - val_loss: 0.0087 - val_mae: 0.1260\n",
            "Epoch 48/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 7.4282e-04 - mae: 0.0204 - val_loss: 0.0019 - val_mae: 0.0583\n",
            "Epoch 49/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 9.9730e-05 - mae: 0.0103 - val_loss: 0.0028 - val_mae: 0.0681\n",
            "Epoch 50/500\n",
            "43/43 [==============================] - 1s 19ms/step - loss: 3.8612e-05 - mae: 0.0065 - val_loss: 5.3254e-04 - val_mae: 0.0262\n",
            "Epoch 51/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 3.3897e-05 - mae: 0.0060 - val_loss: 2.9028e-04 - val_mae: 0.0183\n",
            "Epoch 52/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 5.0814e-05 - mae: 0.0069 - val_loss: 0.0035 - val_mae: 0.0772\n",
            "Epoch 53/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 3.7157e-05 - mae: 0.0064 - val_loss: 7.7238e-04 - val_mae: 0.0325\n",
            "Epoch 54/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 4.8618e-05 - mae: 0.0066 - val_loss: 0.0074 - val_mae: 0.1156\n",
            "Epoch 55/500\n",
            "43/43 [==============================] - 1s 20ms/step - loss: 1.2372e-04 - mae: 0.0101 - val_loss: 0.0228 - val_mae: 0.2064\n",
            "Epoch 56/500\n",
            "43/43 [==============================] - 1s 21ms/step - loss: 6.7090e-04 - mae: 0.0195 - val_loss: 0.0124 - val_mae: 0.1500\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "eNwWZB0d8OJh",
        "colab": {}
      },
      "source": [
        "# reload best model checkpoint\n",
        "model = keras.models.load_model(\"my_checkpoint.h5\")"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "PgYwn9VM8OJi",
        "colab": {}
      },
      "source": [
        "# Use best model to get forecasts\n",
        "cnn_forecast = model_forecast(model, spy_normalized_to_traindata[:,  np.newaxis], window_size)\n",
        "cnn_forecast = cnn_forecast[x_test.index.min() - window_size:-1,-1,0]"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q7z-qvEJQEiu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 33
        },
        "outputId": "c849b08c-c1fc-4a72-e1cc-14ca11340c63"
      },
      "source": [
        "# Scale back to normal\n",
        "cnn_unscaled_forecast = x_train_scaler.inverse_transform(cnn_forecast.reshape(-1,1)).flatten()\n",
        "cnn_unscaled_forecast.shape"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(422,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "MCgshNPx8OJi",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 410
        },
        "outputId": "f6b54ea8-940f-41d7-a01b-2d204923fdbc"
      },
      "source": [
        "# Plot results\n",
        "plt.figure(figsize=(10, 6))\n",
        "\n",
        "plt.ylabel('Dollars $')\n",
        "plt.xlabel('Timestep in Days')\n",
        "plt.title(f'Full CNN Model {window_size}')\n",
        "plot_series(x_test.index, x_test)\n",
        "plot_series(x_test.index, cnn_unscaled_forecast)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "epK1gFEN8OJj",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 33
        },
        "outputId": "c700b5bc-55d7-49ba-de1e-2132195c9b6d"
      },
      "source": [
        "# Calculate MAE\n",
        "keras.metrics.mean_absolute_error(x_test, cnn_unscaled_forecast).numpy()"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "7.9851394"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RAoWAxgGQM_M",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 13,
      "outputs": []
    }
  ]
}