{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Sunspots Predicition",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPT5q1TI3CLJpOfrROjR9U0",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/chadgoldsworthy/b359a87389fd107de8af9b7c13dbffcb/sunspots-predicition.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sLANZGDbbuDS"
      },
      "source": [
        "# Imports"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0r9Js8Pnq3O3"
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import csv\n",
        "import os\n",
        "import math\n",
        "\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv1D, Bidirectional, LSTM, Dense, Lambda\n",
        "from tensorflow.keras.callbacks import LearningRateScheduler\n",
        "from tensorflow.keras.optimizers import SGD\n",
        "from tensorflow.keras.losses import Huber"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "79r-sk9-bzLH"
      },
      "source": [
        "# Downloading Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "umls66j2razb"
      },
      "source": [
        "# Download sunspot dataset (kaggle.com/robervalt/sunspots)\n",
        "\n",
        "from google.colab import drive\n",
        "drive.mount('/gdrive')\n",
        "\n",
        "# Upload kaggle.json file into kaggle dir\n",
        "from google.colab import files\n",
        "files.upload()\n",
        "\n",
        "# Set up kaggle dir & copy json file\n",
        "! pip install -q kaggle\n",
        "! mkdir '../gdrive/MyDrive/kaggle' \n",
        "os.environ['KAGGLE_CONFIG_DIR'] = '../gdrive/MyDrive/kaggle'\n",
        "! cp kaggle.json ../gdrive/MyDrive/kaggle\n",
        "# Set permission\n",
        "! chmod 600 ../gdrive/MyDrive/kaggle/kaggle.json\n",
        "\n",
        "# Download & unzip sunpot dataset \n",
        "!kaggle datasets download -d robervalt/sunspots\n",
        "!unzip sunspots.zip  && rm sunspots.zip\n",
        "\n",
        "! mkdir '../gdrive/MyDrive/kaggle/datasets'\n",
        "! cp Sunspots.csv ../gdrive/MyDrive/kaggle/datasets\n",
        "\n",
        "sunspot_path = '/gdrive/MyDrive/kaggle/datasets/Sunspots.csv'\n",
        "\n",
        "print(f'\\n\\nListing {os.getcwd()} dir:\\n')\n",
        "print(os.listdir())"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bVcYEcxf2aD5",
        "outputId": "2d1a0000-d035-4f4f-81d5-be2570556b9e"
      },
      "source": [
        "# # If colab session disconnects\n",
        "\n",
        "# from google.colab import drive\n",
        "# drive.mount('/gdrive')\n",
        "# sunspot_path = '/gdrive/MyDrive/kaggle/datasets/Sunspots.csv'"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /gdrive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CHCILNxMb_H8"
      },
      "source": [
        "# Visualising Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3KmWOkQexICJ"
      },
      "source": [
        "# Function for plotting data:\n",
        "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",
        "    plt.grid(True)\n",
        "    if label: plt.legend()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M6Zm2eKYtp4_",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 367
        },
        "outputId": "02792855-7513-454b-fa00-94b559815ac5"
      },
      "source": [
        "# Build data as numpy arrays\n",
        "time_steps = []\n",
        "sunspots = []\n",
        "\n",
        "with open(sunspot_path) as f:\n",
        "    reader = csv.reader(f, delimiter=',')\n",
        "    next(reader)\n",
        "    for step, _, sunspot in reader: # Columns: #, Date, Monthly Mean \n",
        "        time_steps.append(int(step))\n",
        "        sunspots.append(float(sunspot))\n",
        "\n",
        "time_steps = np.array(time_steps)\n",
        "sunspots = np.array(sunspots)\n",
        "\n",
        "# View data\n",
        "plt.figure(figsize=(20,6))\n",
        "plot_series(time_steps, sunspots)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SA9FJ26pcD4x"
      },
      "source": [
        "# Data Preprocessing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aZTw3rPNusNo"
      },
      "source": [
        "# 90/10 Train/Test split\n",
        "\n",
        "train_split = 3000\n",
        "\n",
        "time_train = time_steps[:train_split]\n",
        "x_train = sunspots[:train_split]\n",
        "\n",
        "time_val = time_steps[train_split:]\n",
        "x_val = sunspots[train_split:]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b_u6V-g90bFH"
      },
      "source": [
        "# Function to build data pipeline\n",
        "def window_data(data, window_size, batch_size, shuffle_buffer):\n",
        "    data = tf.expand_dims(data, -1)\n",
        "    ds = tf.data.Dataset.from_tensor_slices(data)\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",
        "# Function to get forcasted data\n",
        "def data_forecast(model, data, window_size):\n",
        "    data = tf.expand_dims(data, -1)\n",
        "    ds = tf.data.Dataset.from_tensor_slices(data)\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\n",
        "\n",
        "# Function to reset state & reseed\n",
        "def clean_slate(seed=7):\n",
        "    tf.keras.backend.clear_session()\n",
        "    tf.random.set_seed(seed)\n",
        "    np.random.seed(seed)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ssbkRcW4cM6u"
      },
      "source": [
        "# Training with keras' LearningRateScheduler() to find optimal LR"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Af6hFtLjQ3u0",
        "outputId": "d976fd4e-a071-458d-eae1-743408eb082c"
      },
      "source": [
        "# Finding optimal learning rate\n",
        "clean_slate()\n",
        "\n",
        "# Variables\n",
        "window_size = 64\n",
        "batch_size = 256\n",
        "shuffle_buffer = 1000\n",
        "\n",
        "epochs = 100\n",
        "lr_scheduler = LearningRateScheduler(lambda epoch: 1e-8 * 10**(epoch/20))\n",
        "\n",
        "optimizer = SGD(lr=1e-8, momentum=0.9)\n",
        "loss = Huber()\n",
        "\n",
        "# max_range to rescale data in models output Lambda layer\n",
        "max_range = int(math.ceil(x_train.max()/100) * 100)\n",
        "\n",
        "# Training data\n",
        "train_set = window_data(x_train, window_size, batch_size, shuffle_buffer)\n",
        "\n",
        "model = Sequential([\n",
        "                    Conv1D(32, kernel_size=5, strides=1, \n",
        "                           padding='causal', activation='relu',\n",
        "                           input_shape=[None, 1]),\n",
        "                    Bidirectional(LSTM(window_size, return_sequences=True)),\n",
        "                    Bidirectional(LSTM(window_size, return_sequences=True)),\n",
        "                    Dense(32, activation='relu'),\n",
        "                    Dense(16, activation='relu'),\n",
        "                    Dense(1),\n",
        "                    Lambda(lambda x: x * max_range)\n",
        "])\n",
        "\n",
        "model.compile(loss=loss, optimizer=optimizer, metrics=['mae'])\n",
        "\n",
        "lr_scheduler_history = model.fit(train_set, epochs=epochs, callbacks=[lr_scheduler])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "12/12 [==============================] - 13s 541ms/step - loss: 101.6080 - mae: 102.1070\n",
            "Epoch 2/100\n",
            "12/12 [==============================] - 7s 534ms/step - loss: 94.1628 - mae: 94.6616\n",
            "Epoch 3/100\n",
            "12/12 [==============================] - 7s 543ms/step - loss: 83.3264 - mae: 83.8251\n",
            "Epoch 4/100\n",
            "12/12 [==============================] - 7s 540ms/step - loss: 76.9504 - mae: 77.4486\n",
            "Epoch 5/100\n",
            "12/12 [==============================] - 7s 556ms/step - loss: 70.8914 - mae: 71.3895\n",
            "Epoch 6/100\n",
            "12/12 [==============================] - 7s 539ms/step - loss: 65.9163 - mae: 66.4143\n",
            "Epoch 7/100\n",
            "12/12 [==============================] - 7s 539ms/step - loss: 62.9307 - mae: 63.4286\n",
            "Epoch 8/100\n",
            "12/12 [==============================] - 7s 552ms/step - loss: 59.1983 - mae: 59.6962\n",
            "Epoch 9/100\n",
            "12/12 [==============================] - 7s 539ms/step - loss: 57.3339 - mae: 57.8318\n",
            "Epoch 10/100\n",
            "12/12 [==============================] - 7s 557ms/step - loss: 57.4153 - mae: 57.9132\n",
            "Epoch 11/100\n",
            "12/12 [==============================] - 6s 507ms/step - loss: 56.4008 - mae: 56.8987\n",
            "Epoch 12/100\n",
            "12/12 [==============================] - 7s 544ms/step - loss: 56.3723 - mae: 56.8702\n",
            "Epoch 13/100\n",
            "12/12 [==============================] - 7s 539ms/step - loss: 56.4350 - mae: 56.9330\n",
            "Epoch 14/100\n",
            "12/12 [==============================] - 7s 535ms/step - loss: 54.3331 - mae: 54.8309\n",
            "Epoch 15/100\n",
            "12/12 [==============================] - 7s 542ms/step - loss: 53.8237 - mae: 54.3216\n",
            "Epoch 16/100\n",
            "12/12 [==============================] - 7s 532ms/step - loss: 53.2403 - mae: 53.7381\n",
            "Epoch 17/100\n",
            "12/12 [==============================] - 7s 548ms/step - loss: 52.6064 - mae: 53.1043\n",
            "Epoch 18/100\n",
            "12/12 [==============================] - 7s 544ms/step - loss: 52.3515 - mae: 52.8494\n",
            "Epoch 19/100\n",
            "12/12 [==============================] - 7s 534ms/step - loss: 50.8942 - mae: 51.3919\n",
            "Epoch 20/100\n",
            "12/12 [==============================] - 7s 529ms/step - loss: 50.1691 - mae: 50.6667\n",
            "Epoch 21/100\n",
            "12/12 [==============================] - 7s 533ms/step - loss: 49.9756 - mae: 50.4733\n",
            "Epoch 22/100\n",
            "12/12 [==============================] - 6s 511ms/step - loss: 48.7751 - mae: 49.2728\n",
            "Epoch 23/100\n",
            "12/12 [==============================] - 7s 527ms/step - loss: 48.8348 - mae: 49.3325\n",
            "Epoch 24/100\n",
            "12/12 [==============================] - 7s 533ms/step - loss: 47.5432 - mae: 48.0408\n",
            "Epoch 25/100\n",
            "12/12 [==============================] - 7s 541ms/step - loss: 46.8159 - mae: 47.3134\n",
            "Epoch 26/100\n",
            "12/12 [==============================] - 7s 544ms/step - loss: 45.7319 - mae: 46.2294\n",
            "Epoch 27/100\n",
            "12/12 [==============================] - 7s 527ms/step - loss: 44.5210 - mae: 45.0184\n",
            "Epoch 28/100\n",
            "12/12 [==============================] - 7s 545ms/step - loss: 44.1264 - mae: 44.6238\n",
            "Epoch 29/100\n",
            "12/12 [==============================] - 7s 537ms/step - loss: 43.2545 - mae: 43.7519\n",
            "Epoch 30/100\n",
            "12/12 [==============================] - 7s 532ms/step - loss: 42.2074 - mae: 42.7048\n",
            "Epoch 31/100\n",
            "12/12 [==============================] - 7s 546ms/step - loss: 42.5881 - mae: 43.0852\n",
            "Epoch 32/100\n",
            "12/12 [==============================] - 7s 529ms/step - loss: 40.8542 - mae: 41.3515\n",
            "Epoch 33/100\n",
            "12/12 [==============================] - 7s 531ms/step - loss: 39.9963 - mae: 40.4933\n",
            "Epoch 34/100\n",
            "12/12 [==============================] - 7s 540ms/step - loss: 40.3757 - mae: 40.8726\n",
            "Epoch 35/100\n",
            "12/12 [==============================] - 7s 546ms/step - loss: 39.3832 - mae: 39.8800\n",
            "Epoch 36/100\n",
            "12/12 [==============================] - 7s 544ms/step - loss: 39.2144 - mae: 39.7112\n",
            "Epoch 37/100\n",
            "12/12 [==============================] - 7s 546ms/step - loss: 38.1971 - mae: 38.6940\n",
            "Epoch 38/100\n",
            "12/12 [==============================] - 7s 535ms/step - loss: 39.2791 - mae: 39.7760\n",
            "Epoch 39/100\n",
            "12/12 [==============================] - 7s 540ms/step - loss: 36.8235 - mae: 37.3200\n",
            "Epoch 40/100\n",
            "12/12 [==============================] - 7s 540ms/step - loss: 36.9478 - mae: 37.4445\n",
            "Epoch 41/100\n",
            "12/12 [==============================] - 7s 524ms/step - loss: 38.3689 - mae: 38.8655\n",
            "Epoch 42/100\n",
            "12/12 [==============================] - 7s 540ms/step - loss: 37.9598 - mae: 38.4564\n",
            "Epoch 43/100\n",
            "12/12 [==============================] - 7s 530ms/step - loss: 38.2387 - mae: 38.7352\n",
            "Epoch 44/100\n",
            "12/12 [==============================] - 7s 547ms/step - loss: 35.8947 - mae: 36.3910\n",
            "Epoch 45/100\n",
            "12/12 [==============================] - 7s 546ms/step - loss: 34.4049 - mae: 34.9010\n",
            "Epoch 46/100\n",
            "12/12 [==============================] - 7s 539ms/step - loss: 35.0049 - mae: 35.5011\n",
            "Epoch 47/100\n",
            "12/12 [==============================] - 7s 538ms/step - loss: 34.8203 - mae: 35.3164\n",
            "Epoch 48/100\n",
            "12/12 [==============================] - 7s 532ms/step - loss: 34.8484 - mae: 35.3444\n",
            "Epoch 49/100\n",
            "12/12 [==============================] - 7s 526ms/step - loss: 33.6654 - mae: 34.1612\n",
            "Epoch 50/100\n",
            "12/12 [==============================] - 7s 534ms/step - loss: 31.9970 - mae: 32.4924\n",
            "Epoch 51/100\n",
            "12/12 [==============================] - 7s 542ms/step - loss: 30.9553 - mae: 31.4506\n",
            "Epoch 52/100\n",
            "12/12 [==============================] - 7s 545ms/step - loss: 31.1487 - mae: 31.6440\n",
            "Epoch 53/100\n",
            "12/12 [==============================] - 7s 542ms/step - loss: 29.4683 - mae: 29.9634\n",
            "Epoch 54/100\n",
            "12/12 [==============================] - 7s 535ms/step - loss: 29.6891 - mae: 30.1842\n",
            "Epoch 55/100\n",
            "12/12 [==============================] - 7s 530ms/step - loss: 30.5957 - mae: 31.0912\n",
            "Epoch 56/100\n",
            "12/12 [==============================] - 7s 520ms/step - loss: 29.0587 - mae: 29.5541\n",
            "Epoch 57/100\n",
            "12/12 [==============================] - 7s 539ms/step - loss: 29.4311 - mae: 29.9266\n",
            "Epoch 58/100\n",
            "12/12 [==============================] - 7s 531ms/step - loss: 28.2705 - mae: 28.7660\n",
            "Epoch 59/100\n",
            "12/12 [==============================] - 6s 517ms/step - loss: 27.4959 - mae: 27.9911\n",
            "Epoch 60/100\n",
            "12/12 [==============================] - 7s 539ms/step - loss: 28.7602 - mae: 29.2559\n",
            "Epoch 61/100\n",
            "12/12 [==============================] - 7s 531ms/step - loss: 28.6220 - mae: 29.1171\n",
            "Epoch 62/100\n",
            "12/12 [==============================] - 7s 537ms/step - loss: 26.7961 - mae: 27.2908\n",
            "Epoch 63/100\n",
            "12/12 [==============================] - 6s 511ms/step - loss: 27.9910 - mae: 28.4862\n",
            "Epoch 64/100\n",
            "12/12 [==============================] - 7s 545ms/step - loss: 30.0326 - mae: 30.5281\n",
            "Epoch 65/100\n",
            "12/12 [==============================] - 7s 538ms/step - loss: 28.9896 - mae: 29.4850\n",
            "Epoch 66/100\n",
            "12/12 [==============================] - 7s 544ms/step - loss: 30.5024 - mae: 30.9984\n",
            "Epoch 67/100\n",
            "12/12 [==============================] - 7s 544ms/step - loss: 27.4579 - mae: 27.9532\n",
            "Epoch 68/100\n",
            "12/12 [==============================] - 7s 523ms/step - loss: 26.8154 - mae: 27.3103\n",
            "Epoch 69/100\n",
            "12/12 [==============================] - 7s 547ms/step - loss: 30.0546 - mae: 30.5503\n",
            "Epoch 70/100\n",
            "12/12 [==============================] - 7s 531ms/step - loss: 29.2906 - mae: 29.7860\n",
            "Epoch 71/100\n",
            "12/12 [==============================] - 7s 530ms/step - loss: 27.5221 - mae: 28.0174\n",
            "Epoch 72/100\n",
            "12/12 [==============================] - 7s 519ms/step - loss: 26.6028 - mae: 27.0979\n",
            "Epoch 73/100\n",
            "12/12 [==============================] - 7s 537ms/step - loss: 26.7584 - mae: 27.2538\n",
            "Epoch 74/100\n",
            "12/12 [==============================] - 7s 543ms/step - loss: 28.5395 - mae: 29.0352\n",
            "Epoch 75/100\n",
            "12/12 [==============================] - 7s 547ms/step - loss: 31.8696 - mae: 32.3656\n",
            "Epoch 76/100\n",
            "12/12 [==============================] - 7s 536ms/step - loss: 35.0251 - mae: 35.5216\n",
            "Epoch 77/100\n",
            "12/12 [==============================] - 7s 536ms/step - loss: 31.9050 - mae: 32.4006\n",
            "Epoch 78/100\n",
            "12/12 [==============================] - 7s 531ms/step - loss: 29.2199 - mae: 29.7151\n",
            "Epoch 79/100\n",
            "12/12 [==============================] - 7s 531ms/step - loss: 26.9459 - mae: 27.4407\n",
            "Epoch 80/100\n",
            "12/12 [==============================] - 7s 541ms/step - loss: 25.3818 - mae: 25.8765\n",
            "Epoch 81/100\n",
            "12/12 [==============================] - 7s 535ms/step - loss: 30.6776 - mae: 31.1732\n",
            "Epoch 82/100\n",
            "12/12 [==============================] - 7s 534ms/step - loss: 34.0797 - mae: 34.5757\n",
            "Epoch 83/100\n",
            "12/12 [==============================] - 7s 544ms/step - loss: 28.4773 - mae: 28.9727\n",
            "Epoch 84/100\n",
            "12/12 [==============================] - 7s 545ms/step - loss: 27.3387 - mae: 27.8342\n",
            "Epoch 85/100\n",
            "12/12 [==============================] - 7s 536ms/step - loss: 27.5773 - mae: 28.0720\n",
            "Epoch 86/100\n",
            "12/12 [==============================] - 7s 550ms/step - loss: 28.2353 - mae: 28.7309\n",
            "Epoch 87/100\n",
            "12/12 [==============================] - 7s 529ms/step - loss: 29.0594 - mae: 29.5553\n",
            "Epoch 88/100\n",
            "12/12 [==============================] - 7s 531ms/step - loss: 34.0806 - mae: 34.5772\n",
            "Epoch 89/100\n",
            "12/12 [==============================] - 7s 532ms/step - loss: 30.7627 - mae: 31.2588\n",
            "Epoch 90/100\n",
            "12/12 [==============================] - 7s 552ms/step - loss: 33.0755 - mae: 33.5722\n",
            "Epoch 91/100\n",
            "12/12 [==============================] - 7s 537ms/step - loss: 34.5036 - mae: 35.0002\n",
            "Epoch 92/100\n",
            "12/12 [==============================] - 7s 541ms/step - loss: 33.5770 - mae: 34.0728\n",
            "Epoch 93/100\n",
            "12/12 [==============================] - 7s 538ms/step - loss: 35.8090 - mae: 36.3058\n",
            "Epoch 94/100\n",
            "12/12 [==============================] - 7s 525ms/step - loss: 36.4856 - mae: 36.9825\n",
            "Epoch 95/100\n",
            "12/12 [==============================] - 7s 533ms/step - loss: 42.7534 - mae: 43.2508\n",
            "Epoch 96/100\n",
            "12/12 [==============================] - 7s 536ms/step - loss: 56.7506 - mae: 57.2486\n",
            "Epoch 97/100\n",
            "12/12 [==============================] - 7s 524ms/step - loss: 53.4703 - mae: 53.9685\n",
            "Epoch 98/100\n",
            "12/12 [==============================] - 7s 521ms/step - loss: 64.4222 - mae: 64.9211\n",
            "Epoch 99/100\n",
            "12/12 [==============================] - 7s 546ms/step - loss: 52.0656 - mae: 52.5635\n",
            "Epoch 100/100\n",
            "12/12 [==============================] - 7s 530ms/step - loss: 68.4292 - mae: 68.9276\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 454
        },
        "id": "ipHGDIiyZtJK",
        "outputId": "299356e2-68d5-44b0-91d3-6e35f595f348"
      },
      "source": [
        "# Visualise change in loss chnage due to learning rate\n",
        "\n",
        "lrs = 1e-8 * 10**(np.arange(100)/20)\n",
        "mae = lr_scheduler_history.history['mae']\n",
        "\n",
        "plt.figure(figsize=(14, 7))\n",
        "plt.semilogx(lrs, mae)\n",
        "plt.axis([1e-8, 1e-3, 20, 70])\n",
        "\n",
        "\n",
        "\n",
        "#_____________________________________________\n",
        "# After ploting...\n",
        "# approximetly 0.4e-5 before getting unstable\n",
        "dot = 0.4e-5\n",
        "plt.scatter(dot, 30.7, c='red', linewidths=5)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.PathCollection at 0x7fc9fb248940>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 104
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1008x504 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zus2bPKJcuep"
      },
      "source": [
        "# Training the final model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gRhXR8sTiI5V",
        "outputId": "04c30450-2908-45ef-fb7d-a26c68ffa9bc"
      },
      "source": [
        "# Retrain model with optimal LR over 400 epochs\n",
        "\n",
        "optimal_lr = 0.4e-5\n",
        "optimizer = SGD(lr=optimal_lr, momentum=0.9)\n",
        "loss = Huber()\n",
        "epochs = 400\n",
        "\n",
        "clean_slate()\n",
        "train_set = window_data(x_train, window_size, batch_size, shuffle_buffer)\n",
        "\n",
        "model = Sequential([\n",
        "                    Conv1D(64, kernel_size=5, strides=1, padding='causal',\n",
        "                           activation='relu', input_shape=[None,1]),\n",
        "                    Bidirectional(LSTM(window_size, return_sequences=True)),\n",
        "                    Bidirectional(LSTM(window_size, return_sequences=True)),\n",
        "                    Dense(32, activation='relu'),\n",
        "                    Dense(16, activation='relu'),\n",
        "                    Dense(1),\n",
        "                    Lambda(lambda x: x*max_range)\n",
        "])\n",
        "\n",
        "model.compile(loss=loss, optimizer=optimizer, metrics=['mae'])\n",
        "\n",
        "optimal_lr_history = model.fit(train_set, epochs=epochs)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/400\n",
            "12/12 [==============================] - 12s 35ms/step - loss: 61.6588 - mae: 62.1568\n",
            "Epoch 2/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 47.5806 - mae: 48.0781\n",
            "Epoch 3/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 41.3920 - mae: 41.8893\n",
            "Epoch 4/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 40.2566 - mae: 40.7537\n",
            "Epoch 5/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 37.7948 - mae: 38.2917\n",
            "Epoch 6/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 36.5132 - mae: 37.0100\n",
            "Epoch 7/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 35.4265 - mae: 35.9228\n",
            "Epoch 8/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 35.3376 - mae: 35.8341\n",
            "Epoch 9/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 33.5201 - mae: 34.0161\n",
            "Epoch 10/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 34.6727 - mae: 35.1688\n",
            "Epoch 11/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 33.6825 - mae: 34.1787\n",
            "Epoch 12/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 34.0072 - mae: 34.5036\n",
            "Epoch 13/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 34.0224 - mae: 34.5187\n",
            "Epoch 14/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 32.6082 - mae: 33.1042\n",
            "Epoch 15/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 32.0112 - mae: 32.5069\n",
            "Epoch 16/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 32.4857 - mae: 32.9816\n",
            "Epoch 17/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 32.8453 - mae: 33.3415\n",
            "Epoch 18/400\n",
            "12/12 [==============================] - 0s 30ms/step - loss: 31.1843 - mae: 31.6798\n",
            "Epoch 19/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 31.0233 - mae: 31.5190\n",
            "Epoch 20/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 31.4390 - mae: 31.9348\n",
            "Epoch 21/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 31.1125 - mae: 31.6081\n",
            "Epoch 22/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 29.9022 - mae: 30.3975\n",
            "Epoch 23/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 30.0626 - mae: 30.5580\n",
            "Epoch 24/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 29.2631 - mae: 29.7584\n",
            "Epoch 25/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 28.9572 - mae: 29.4522\n",
            "Epoch 26/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 29.3743 - mae: 29.8695\n",
            "Epoch 27/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 28.1905 - mae: 28.6854\n",
            "Epoch 28/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 27.6090 - mae: 28.1040\n",
            "Epoch 29/400\n",
            "12/12 [==============================] - 0s 30ms/step - loss: 27.7949 - mae: 28.2898\n",
            "Epoch 30/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 27.6832 - mae: 28.1782\n",
            "Epoch 31/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 28.1744 - mae: 28.6693\n",
            "Epoch 32/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 26.9858 - mae: 27.4799\n",
            "Epoch 33/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 27.4929 - mae: 27.9878\n",
            "Epoch 34/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 26.5901 - mae: 27.0849\n",
            "Epoch 35/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 26.2502 - mae: 26.7449\n",
            "Epoch 36/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 26.3712 - mae: 26.8661\n",
            "Epoch 37/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 27.2398 - mae: 27.7347\n",
            "Epoch 38/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 26.6252 - mae: 27.1202\n",
            "Epoch 39/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 28.6279 - mae: 29.1232\n",
            "Epoch 40/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 28.3043 - mae: 28.7999\n",
            "Epoch 41/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 27.5316 - mae: 28.0266\n",
            "Epoch 42/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 25.9836 - mae: 26.4782\n",
            "Epoch 43/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 25.9095 - mae: 26.4043\n",
            "Epoch 44/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 25.5497 - mae: 26.0438\n",
            "Epoch 45/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 25.2094 - mae: 25.7036\n",
            "Epoch 46/400\n",
            "12/12 [==============================] - 0s 30ms/step - loss: 24.8499 - mae: 25.3442\n",
            "Epoch 47/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 24.7677 - mae: 25.2618\n",
            "Epoch 48/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 24.4533 - mae: 24.9476\n",
            "Epoch 49/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 24.4670 - mae: 24.9613\n",
            "Epoch 50/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 24.0332 - mae: 24.5275\n",
            "Epoch 51/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 23.2473 - mae: 23.7414\n",
            "Epoch 52/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 24.0355 - mae: 24.5297\n",
            "Epoch 53/400\n",
            "12/12 [==============================] - 0s 30ms/step - loss: 23.6013 - mae: 24.0955\n",
            "Epoch 54/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 23.4038 - mae: 23.8978\n",
            "Epoch 55/400\n",
            "12/12 [==============================] - 0s 33ms/step - loss: 23.3017 - mae: 23.7959\n",
            "Epoch 56/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 23.7354 - mae: 24.2297\n",
            "Epoch 57/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 22.9038 - mae: 23.3976\n",
            "Epoch 58/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 22.6016 - mae: 23.0959\n",
            "Epoch 59/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 22.5036 - mae: 22.9978\n",
            "Epoch 60/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 21.8544 - mae: 22.3482\n",
            "Epoch 61/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 21.8788 - mae: 22.3728\n",
            "Epoch 62/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 21.4037 - mae: 21.8972\n",
            "Epoch 63/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 21.7557 - mae: 22.2492\n",
            "Epoch 64/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 20.8764 - mae: 21.3700\n",
            "Epoch 65/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 21.0722 - mae: 21.5656\n",
            "Epoch 66/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 21.1176 - mae: 21.6109\n",
            "Epoch 67/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 20.5804 - mae: 21.0744\n",
            "Epoch 68/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 20.4752 - mae: 20.9684\n",
            "Epoch 69/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 20.5985 - mae: 21.0924\n",
            "Epoch 70/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 21.3691 - mae: 21.8628\n",
            "Epoch 71/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 20.8836 - mae: 21.3766\n",
            "Epoch 72/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 20.6658 - mae: 21.1593\n",
            "Epoch 73/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.9441 - mae: 20.4370\n",
            "Epoch 74/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 20.2714 - mae: 20.7649\n",
            "Epoch 75/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 20.5921 - mae: 21.0853\n",
            "Epoch 76/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 20.2705 - mae: 20.7642\n",
            "Epoch 77/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 20.9819 - mae: 21.4750\n",
            "Epoch 78/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.6246 - mae: 20.1175\n",
            "Epoch 79/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 20.5151 - mae: 21.0088\n",
            "Epoch 80/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 20.5532 - mae: 21.0466\n",
            "Epoch 81/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 25.9132 - mae: 26.4077\n",
            "Epoch 82/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 21.6598 - mae: 22.1530\n",
            "Epoch 83/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 19.9233 - mae: 20.4168\n",
            "Epoch 84/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 19.7839 - mae: 20.2771\n",
            "Epoch 85/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.9927 - mae: 20.4856\n",
            "Epoch 86/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.7967 - mae: 20.2899\n",
            "Epoch 87/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 21.9581 - mae: 22.4513\n",
            "Epoch 88/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 20.7959 - mae: 21.2890\n",
            "Epoch 89/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.4504 - mae: 19.9435\n",
            "Epoch 90/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.0845 - mae: 19.5775\n",
            "Epoch 91/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.9923 - mae: 20.4851\n",
            "Epoch 92/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.5431 - mae: 20.0359\n",
            "Epoch 93/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.3609 - mae: 19.8533\n",
            "Epoch 94/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 20.8451 - mae: 21.3388\n",
            "Epoch 95/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.6780 - mae: 20.1710\n",
            "Epoch 96/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 19.7425 - mae: 20.2352\n",
            "Epoch 97/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 20.3525 - mae: 20.8452\n",
            "Epoch 98/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.9074 - mae: 19.3997\n",
            "Epoch 99/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 20.7885 - mae: 21.2822\n",
            "Epoch 100/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 20.2384 - mae: 20.7314\n",
            "Epoch 101/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 18.4581 - mae: 18.9508\n",
            "Epoch 102/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.4944 - mae: 19.9876\n",
            "Epoch 103/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.1509 - mae: 19.6436\n",
            "Epoch 104/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.5950 - mae: 20.0876\n",
            "Epoch 105/400\n",
            "12/12 [==============================] - 0s 33ms/step - loss: 19.1908 - mae: 19.6834\n",
            "Epoch 106/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.6020 - mae: 20.0948\n",
            "Epoch 107/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.8276 - mae: 19.3204\n",
            "Epoch 108/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.9875 - mae: 20.4805\n",
            "Epoch 109/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.6702 - mae: 19.1625\n",
            "Epoch 110/400\n",
            "12/12 [==============================] - 0s 30ms/step - loss: 18.6542 - mae: 19.1469\n",
            "Epoch 111/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.3589 - mae: 19.8516\n",
            "Epoch 112/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.3749 - mae: 19.8671\n",
            "Epoch 113/400\n",
            "12/12 [==============================] - 0s 30ms/step - loss: 18.5909 - mae: 19.0828\n",
            "Epoch 114/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.4392 - mae: 19.9320\n",
            "Epoch 115/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 18.9295 - mae: 19.4216\n",
            "Epoch 116/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.2673 - mae: 18.7597\n",
            "Epoch 117/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.1158 - mae: 18.6077\n",
            "Epoch 118/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 19.0543 - mae: 19.5471\n",
            "Epoch 119/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.2895 - mae: 19.7823\n",
            "Epoch 120/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.7021 - mae: 19.1946\n",
            "Epoch 121/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 18.4393 - mae: 18.9322\n",
            "Epoch 122/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.1935 - mae: 19.6857\n",
            "Epoch 123/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.2107 - mae: 19.7028\n",
            "Epoch 124/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 18.3101 - mae: 18.8022\n",
            "Epoch 125/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.1188 - mae: 18.6108\n",
            "Epoch 126/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.0648 - mae: 18.5571\n",
            "Epoch 127/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.6040 - mae: 19.0966\n",
            "Epoch 128/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 18.6571 - mae: 19.1494\n",
            "Epoch 129/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 18.3244 - mae: 18.8166\n",
            "Epoch 130/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.6607 - mae: 19.1537\n",
            "Epoch 131/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 18.8868 - mae: 19.3789\n",
            "Epoch 132/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 18.9996 - mae: 19.4924\n",
            "Epoch 133/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 18.8625 - mae: 19.3549\n",
            "Epoch 134/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.7987 - mae: 19.2912\n",
            "Epoch 135/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.6987 - mae: 19.1912\n",
            "Epoch 136/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.4338 - mae: 18.9266\n",
            "Epoch 137/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 18.4197 - mae: 18.9122\n",
            "Epoch 138/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 18.0831 - mae: 18.5752\n",
            "Epoch 139/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.2794 - mae: 19.7715\n",
            "Epoch 140/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 19.0449 - mae: 19.5380\n",
            "Epoch 141/400\n",
            "12/12 [==============================] - 0s 30ms/step - loss: 19.1072 - mae: 19.5996\n",
            "Epoch 142/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.5318 - mae: 19.0238\n",
            "Epoch 143/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 18.6542 - mae: 19.1466\n",
            "Epoch 144/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.7680 - mae: 19.2610\n",
            "Epoch 145/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 18.6897 - mae: 19.1819\n",
            "Epoch 146/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.6953 - mae: 20.1879\n",
            "Epoch 147/400\n",
            "12/12 [==============================] - 0s 33ms/step - loss: 18.7312 - mae: 19.2233\n",
            "Epoch 148/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.7561 - mae: 18.2480\n",
            "Epoch 149/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.8832 - mae: 18.3748\n",
            "Epoch 150/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 18.6847 - mae: 19.1767\n",
            "Epoch 151/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.8783 - mae: 18.3700\n",
            "Epoch 152/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.5589 - mae: 19.0507\n",
            "Epoch 153/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.5517 - mae: 20.0441\n",
            "Epoch 154/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.7020 - mae: 19.1942\n",
            "Epoch 155/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.6601 - mae: 18.1523\n",
            "Epoch 156/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.9680 - mae: 18.4606\n",
            "Epoch 157/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 18.6624 - mae: 19.1550\n",
            "Epoch 158/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.6403 - mae: 20.1338\n",
            "Epoch 159/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.8522 - mae: 18.3441\n",
            "Epoch 160/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.6235 - mae: 18.1153\n",
            "Epoch 161/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.1517 - mae: 19.6439\n",
            "Epoch 162/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.3161 - mae: 19.8089\n",
            "Epoch 163/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.9767 - mae: 18.4690\n",
            "Epoch 164/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 19.7970 - mae: 20.2892\n",
            "Epoch 165/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.3931 - mae: 19.8859\n",
            "Epoch 166/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.7429 - mae: 20.2364\n",
            "Epoch 167/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 20.4219 - mae: 20.9150\n",
            "Epoch 168/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 19.8078 - mae: 20.3008\n",
            "Epoch 169/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.0750 - mae: 18.5666\n",
            "Epoch 170/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.1216 - mae: 18.6134\n",
            "Epoch 171/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.0611 - mae: 18.5531\n",
            "Epoch 172/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.6826 - mae: 19.1751\n",
            "Epoch 173/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.7890 - mae: 18.2810\n",
            "Epoch 174/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.9851 - mae: 18.4767\n",
            "Epoch 175/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.6234 - mae: 18.1143\n",
            "Epoch 176/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.8918 - mae: 18.3844\n",
            "Epoch 177/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.4707 - mae: 17.9624\n",
            "Epoch 178/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.0990 - mae: 18.5903\n",
            "Epoch 179/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.0209 - mae: 19.5132\n",
            "Epoch 180/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.9085 - mae: 18.4007\n",
            "Epoch 181/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.6682 - mae: 18.1596\n",
            "Epoch 182/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.4984 - mae: 17.9898\n",
            "Epoch 183/400\n",
            "12/12 [==============================] - 1s 38ms/step - loss: 17.7907 - mae: 18.2823\n",
            "Epoch 184/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 18.2369 - mae: 18.7294\n",
            "Epoch 185/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 18.2127 - mae: 18.7052\n",
            "Epoch 186/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.0361 - mae: 18.5278\n",
            "Epoch 187/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3705 - mae: 17.8623\n",
            "Epoch 188/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.9864 - mae: 18.4778\n",
            "Epoch 189/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.6236 - mae: 19.1165\n",
            "Epoch 190/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.1674 - mae: 18.6587\n",
            "Epoch 191/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.8448 - mae: 18.3367\n",
            "Epoch 192/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.8558 - mae: 18.3476\n",
            "Epoch 193/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.6663 - mae: 18.1579\n",
            "Epoch 194/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.7039 - mae: 18.1957\n",
            "Epoch 195/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 18.6033 - mae: 19.0952\n",
            "Epoch 196/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.3753 - mae: 18.8675\n",
            "Epoch 197/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.9811 - mae: 18.4730\n",
            "Epoch 198/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.7592 - mae: 19.2515\n",
            "Epoch 199/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3296 - mae: 17.8207\n",
            "Epoch 200/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.6843 - mae: 18.1758\n",
            "Epoch 201/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.7616 - mae: 18.2542\n",
            "Epoch 202/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.6545 - mae: 18.1458\n",
            "Epoch 203/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.9970 - mae: 18.4892\n",
            "Epoch 204/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.1246 - mae: 18.6163\n",
            "Epoch 205/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.6060 - mae: 18.0980\n",
            "Epoch 206/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.2776 - mae: 18.7695\n",
            "Epoch 207/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.6254 - mae: 18.1168\n",
            "Epoch 208/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.2248 - mae: 17.7160\n",
            "Epoch 209/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.4670 - mae: 17.9591\n",
            "Epoch 210/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 17.2353 - mae: 17.7265\n",
            "Epoch 211/400\n",
            "12/12 [==============================] - 0s 33ms/step - loss: 17.4500 - mae: 17.9414\n",
            "Epoch 212/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.0721 - mae: 18.5641\n",
            "Epoch 213/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 18.4357 - mae: 18.9279\n",
            "Epoch 214/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.8373 - mae: 18.3293\n",
            "Epoch 215/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.5901 - mae: 19.0828\n",
            "Epoch 216/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.7273 - mae: 18.2192\n",
            "Epoch 217/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.8537 - mae: 18.3456\n",
            "Epoch 218/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.6340 - mae: 18.1257\n",
            "Epoch 219/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.9144 - mae: 18.4059\n",
            "Epoch 220/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.4274 - mae: 18.9199\n",
            "Epoch 221/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.1171 - mae: 18.6085\n",
            "Epoch 222/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3801 - mae: 17.8717\n",
            "Epoch 223/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.1517 - mae: 17.6428\n",
            "Epoch 224/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.4509 - mae: 17.9422\n",
            "Epoch 225/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.1719 - mae: 18.6637\n",
            "Epoch 226/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.3536 - mae: 18.8459\n",
            "Epoch 227/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.9932 - mae: 18.4845\n",
            "Epoch 228/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.3238 - mae: 17.8154\n",
            "Epoch 229/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.8748 - mae: 19.3675\n",
            "Epoch 230/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 19.1482 - mae: 19.6402\n",
            "Epoch 231/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.0502 - mae: 18.5422\n",
            "Epoch 232/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 19.0795 - mae: 19.5720\n",
            "Epoch 233/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.8899 - mae: 18.3821\n",
            "Epoch 234/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.1540 - mae: 17.6454\n",
            "Epoch 235/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0876 - mae: 17.5790\n",
            "Epoch 236/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.2668 - mae: 17.7579\n",
            "Epoch 237/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.1163 - mae: 17.6080\n",
            "Epoch 238/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.6902 - mae: 18.1816\n",
            "Epoch 239/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 17.3802 - mae: 17.8722\n",
            "Epoch 240/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.9374 - mae: 17.4289\n",
            "Epoch 241/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.8602 - mae: 18.3517\n",
            "Epoch 242/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 18.2682 - mae: 18.7601\n",
            "Epoch 243/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.7576 - mae: 18.2496\n",
            "Epoch 244/400\n",
            "12/12 [==============================] - 0s 33ms/step - loss: 16.9521 - mae: 17.4433\n",
            "Epoch 245/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 17.7882 - mae: 18.2798\n",
            "Epoch 246/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.1275 - mae: 17.6189\n",
            "Epoch 247/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.6331 - mae: 18.1242\n",
            "Epoch 248/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.4027 - mae: 17.8944\n",
            "Epoch 249/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3506 - mae: 17.8427\n",
            "Epoch 250/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.1548 - mae: 17.6458\n",
            "Epoch 251/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.3682 - mae: 17.8597\n",
            "Epoch 252/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.4332 - mae: 17.9241\n",
            "Epoch 253/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.3625 - mae: 17.8538\n",
            "Epoch 254/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.3121 - mae: 17.8036\n",
            "Epoch 255/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.6063 - mae: 18.0981\n",
            "Epoch 256/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.5781 - mae: 17.0688\n",
            "Epoch 257/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.5599 - mae: 18.0510\n",
            "Epoch 258/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.4957 - mae: 17.9872\n",
            "Epoch 259/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.2687 - mae: 17.7608\n",
            "Epoch 260/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.4837 - mae: 17.9754\n",
            "Epoch 261/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.1006 - mae: 17.5919\n",
            "Epoch 262/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.3013 - mae: 17.7927\n",
            "Epoch 263/400\n",
            "12/12 [==============================] - 0s 30ms/step - loss: 17.3468 - mae: 17.8386\n",
            "Epoch 264/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.8352 - mae: 18.3266\n",
            "Epoch 265/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.0350 - mae: 19.5270\n",
            "Epoch 266/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.2116 - mae: 18.7048\n",
            "Epoch 267/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.5407 - mae: 18.0315\n",
            "Epoch 268/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0222 - mae: 17.5133\n",
            "Epoch 269/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.7853 - mae: 18.2769\n",
            "Epoch 270/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.2589 - mae: 17.7504\n",
            "Epoch 271/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.5565 - mae: 18.0479\n",
            "Epoch 272/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.1991 - mae: 17.6911\n",
            "Epoch 273/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.0503 - mae: 19.5427\n",
            "Epoch 274/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.1601 - mae: 18.6527\n",
            "Epoch 275/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.1648 - mae: 18.6569\n",
            "Epoch 276/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.1485 - mae: 18.6401\n",
            "Epoch 277/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 19.0619 - mae: 19.5548\n",
            "Epoch 278/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.6993 - mae: 18.1914\n",
            "Epoch 279/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.6103 - mae: 18.1018\n",
            "Epoch 280/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.4088 - mae: 17.9005\n",
            "Epoch 281/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.5145 - mae: 18.0060\n",
            "Epoch 282/400\n",
            "12/12 [==============================] - 1s 36ms/step - loss: 17.4501 - mae: 17.9422\n",
            "Epoch 283/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.8778 - mae: 17.3693\n",
            "Epoch 284/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.6969 - mae: 19.1885\n",
            "Epoch 285/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.4925 - mae: 18.9848\n",
            "Epoch 286/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 18.2135 - mae: 18.7051\n",
            "Epoch 287/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.2787 - mae: 17.7698\n",
            "Epoch 288/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.1277 - mae: 17.6192\n",
            "Epoch 289/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.2500 - mae: 17.7424\n",
            "Epoch 290/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.2018 - mae: 17.6936\n",
            "Epoch 291/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.1728 - mae: 17.6640\n",
            "Epoch 292/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3442 - mae: 17.8355\n",
            "Epoch 293/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.1209 - mae: 17.6122\n",
            "Epoch 294/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3371 - mae: 17.8285\n",
            "Epoch 295/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 18.0822 - mae: 18.5741\n",
            "Epoch 296/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.9335 - mae: 17.4251\n",
            "Epoch 297/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.8607 - mae: 17.3513\n",
            "Epoch 298/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.1008 - mae: 17.5924\n",
            "Epoch 299/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.9922 - mae: 17.4834\n",
            "Epoch 300/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.8650 - mae: 17.3560\n",
            "Epoch 301/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.4217 - mae: 17.9133\n",
            "Epoch 302/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3454 - mae: 17.8367\n",
            "Epoch 303/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.3453 - mae: 17.8362\n",
            "Epoch 304/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0672 - mae: 17.5583\n",
            "Epoch 305/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.8688 - mae: 17.3602\n",
            "Epoch 306/400\n",
            "12/12 [==============================] - 1s 36ms/step - loss: 16.9847 - mae: 17.4758\n",
            "Epoch 307/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.7968 - mae: 18.2887\n",
            "Epoch 308/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.9532 - mae: 17.4444\n",
            "Epoch 309/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.2153 - mae: 17.7067\n",
            "Epoch 310/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0384 - mae: 17.5294\n",
            "Epoch 311/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.5559 - mae: 18.0470\n",
            "Epoch 312/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 18.2080 - mae: 18.6997\n",
            "Epoch 313/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.2996 - mae: 17.7909\n",
            "Epoch 314/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.4528 - mae: 17.9446\n",
            "Epoch 315/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 19.7545 - mae: 20.2471\n",
            "Epoch 316/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.5381 - mae: 18.0300\n",
            "Epoch 317/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.4084 - mae: 18.9010\n",
            "Epoch 318/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.1213 - mae: 17.6128\n",
            "Epoch 319/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.8513 - mae: 17.3427\n",
            "Epoch 320/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 16.9497 - mae: 17.4402\n",
            "Epoch 321/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.3138 - mae: 17.8056\n",
            "Epoch 322/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.2443 - mae: 17.7354\n",
            "Epoch 323/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.9062 - mae: 17.3970\n",
            "Epoch 324/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 17.3314 - mae: 17.8226\n",
            "Epoch 325/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.1982 - mae: 17.6893\n",
            "Epoch 326/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 17.6690 - mae: 18.1601\n",
            "Epoch 327/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.2606 - mae: 17.7522\n",
            "Epoch 328/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.1276 - mae: 17.6185\n",
            "Epoch 329/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.1854 - mae: 17.6774\n",
            "Epoch 330/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.9642 - mae: 17.4556\n",
            "Epoch 331/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.8865 - mae: 17.3777\n",
            "Epoch 332/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.0440 - mae: 17.5353\n",
            "Epoch 333/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 17.0519 - mae: 17.5425\n",
            "Epoch 334/400\n",
            "12/12 [==============================] - 1s 36ms/step - loss: 16.7324 - mae: 17.2239\n",
            "Epoch 335/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.5188 - mae: 18.0102\n",
            "Epoch 336/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.3516 - mae: 18.8440\n",
            "Epoch 337/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 18.0025 - mae: 18.4941\n",
            "Epoch 338/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.3690 - mae: 17.8602\n",
            "Epoch 339/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.0300 - mae: 17.5210\n",
            "Epoch 340/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.1321 - mae: 17.6233\n",
            "Epoch 341/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0201 - mae: 17.5106\n",
            "Epoch 342/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.4350 - mae: 17.9270\n",
            "Epoch 343/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.3045 - mae: 17.7955\n",
            "Epoch 344/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.7466 - mae: 17.2376\n",
            "Epoch 345/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.0443 - mae: 17.5354\n",
            "Epoch 346/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.6572 - mae: 18.1483\n",
            "Epoch 347/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.5076 - mae: 17.9990\n",
            "Epoch 348/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.7625 - mae: 17.2537\n",
            "Epoch 349/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.6073 - mae: 17.0982\n",
            "Epoch 350/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.4887 - mae: 16.9798\n",
            "Epoch 351/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0204 - mae: 17.5118\n",
            "Epoch 352/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.4854 - mae: 17.9764\n",
            "Epoch 353/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.4413 - mae: 17.9324\n",
            "Epoch 354/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 16.9259 - mae: 17.4164\n",
            "Epoch 355/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.9806 - mae: 17.4718\n",
            "Epoch 356/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.7902 - mae: 17.2812\n",
            "Epoch 357/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0878 - mae: 17.5786\n",
            "Epoch 358/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 16.8135 - mae: 17.3047\n",
            "Epoch 359/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 16.7737 - mae: 17.2648\n",
            "Epoch 360/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.7547 - mae: 17.2456\n",
            "Epoch 361/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.1696 - mae: 17.6603\n",
            "Epoch 362/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.9719 - mae: 17.4630\n",
            "Epoch 363/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.7661 - mae: 18.2579\n",
            "Epoch 364/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 17.2186 - mae: 17.7097\n",
            "Epoch 365/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.9351 - mae: 17.4268\n",
            "Epoch 366/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0255 - mae: 17.5158\n",
            "Epoch 367/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 18.0210 - mae: 18.5131\n",
            "Epoch 368/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3042 - mae: 17.7955\n",
            "Epoch 369/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.2166 - mae: 17.7079\n",
            "Epoch 370/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.7068 - mae: 17.1980\n",
            "Epoch 371/400\n",
            "12/12 [==============================] - 1s 36ms/step - loss: 17.3552 - mae: 17.8468\n",
            "Epoch 372/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.6206 - mae: 18.1123\n",
            "Epoch 373/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.9062 - mae: 17.3978\n",
            "Epoch 374/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.9994 - mae: 17.4906\n",
            "Epoch 375/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.5663 - mae: 18.0583\n",
            "Epoch 376/400\n",
            "12/12 [==============================] - 0s 31ms/step - loss: 18.1252 - mae: 18.6176\n",
            "Epoch 377/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.2748 - mae: 17.7657\n",
            "Epoch 378/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.6244 - mae: 17.1157\n",
            "Epoch 379/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 17.3168 - mae: 17.8085\n",
            "Epoch 380/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.7497 - mae: 18.2416\n",
            "Epoch 381/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 17.7601 - mae: 18.2511\n",
            "Epoch 382/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 17.8250 - mae: 18.3168\n",
            "Epoch 383/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.7628 - mae: 17.2540\n",
            "Epoch 384/400\n",
            "12/12 [==============================] - 1s 36ms/step - loss: 16.8415 - mae: 17.3331\n",
            "Epoch 385/400\n",
            "12/12 [==============================] - 0s 32ms/step - loss: 17.0955 - mae: 17.5859\n",
            "Epoch 386/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 17.0871 - mae: 17.5780\n",
            "Epoch 387/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.2350 - mae: 17.7267\n",
            "Epoch 388/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.7974 - mae: 17.2877\n",
            "Epoch 389/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.6248 - mae: 17.1159\n",
            "Epoch 390/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.5438 - mae: 17.0346\n",
            "Epoch 391/400\n",
            "12/12 [==============================] - 1s 31ms/step - loss: 16.7868 - mae: 17.2775\n",
            "Epoch 392/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.4653 - mae: 16.9562\n",
            "Epoch 393/400\n",
            "12/12 [==============================] - 1s 35ms/step - loss: 16.8771 - mae: 17.3677\n",
            "Epoch 394/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.8039 - mae: 17.2948\n",
            "Epoch 395/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 16.5141 - mae: 17.0043\n",
            "Epoch 396/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.3689 - mae: 17.8597\n",
            "Epoch 397/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 17.0572 - mae: 17.5485\n",
            "Epoch 398/400\n",
            "12/12 [==============================] - 1s 33ms/step - loss: 16.3011 - mae: 16.7917\n",
            "Epoch 399/400\n",
            "12/12 [==============================] - 1s 32ms/step - loss: 17.2427 - mae: 17.7342\n",
            "Epoch 400/400\n",
            "12/12 [==============================] - 1s 34ms/step - loss: 16.6234 - mae: 17.1147\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OXaYufrhaOtE",
        "outputId": "bf9a5acf-e5a9-46f4-9a08-a24a1bf6ddeb"
      },
      "source": [
        "# Predict on entire dataset\n",
        "model_forecast = data_forecast(model, sunspots, window_size)\n",
        "\n",
        "# Slice validation\n",
        "forecast_val = model_forecast[train_split-window_size+1:,-1,0]\n",
        "\n",
        "print(forecast_val.shape == x_val.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "True\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gM-8DzWocyy7"
      },
      "source": [
        "# Visualising the results"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 497
        },
        "id": "v4btT5wGaZoe",
        "outputId": "d2c6e29c-ea66-4de3-b712-e5d5aecdebba"
      },
      "source": [
        "# Plot val and forecasted data\n",
        "plt.figure(figsize=(18,8))\n",
        "plot_series(time_val, x_val, label='x_val')\n",
        "plot_series(time_val, forecast_val, label='forecast_val')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1296x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 515
        },
        "id": "6eprZhZLbtQG",
        "outputId": "30d2e021-4ae3-4412-bb9d-46520e10404d"
      },
      "source": [
        "# Visualising model loss\n",
        "\n",
        "fig, (loss, loss_zoom) = plt.subplots(2,1, figsize=(18,8))\n",
        "\n",
        "loss_list = optimal_lr_history.history['loss']\n",
        "epoch_list = range(len(loss_list))\n",
        "\n",
        "loss.set_xlabel('Epochs')\n",
        "loss.set_ylabel('Loss')\n",
        "loss.plot(epoch_list, loss_list)\n",
        "\n",
        "# Zooming in\n",
        "loss_list_zoom = loss_list[199:]\n",
        "epoch_list_zoom = epoch_list[199:]\n",
        "\n",
        "loss_zoom.set_xlabel('Epochs')\n",
        "loss_zoom.set_ylabel('Loss')\n",
        "loss_zoom.plot(epoch_list_zoom, loss_list_zoom)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7f2e06c3b978>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 71
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1296x576 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sA3Sze8vdBNN"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}