SolarModel_ANN_BigQuery_Insert.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/snvssk/data298A/blob/development/SolarModel_ANN_BigQuery_Insert.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fOG5SyLs_u6I"
      },
      "source": [
        "## Libraries"
      ],
      "id": "fOG5SyLs_u6I"
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "QfUqcTPLFUvi"
      },
      "outputs": [],
      "source": [
        "import math\n",
        "import pandas as pd\n",
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt\n",
        "from tensorflow.keras import Model\n",
        "from tensorflow.keras import Sequential\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from tensorflow.keras.layers import Dense, Dropout\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import mean_squared_error\n",
        "from sklearn.metrics import r2_score\n",
        "from math import sqrt\n",
        "from google.cloud import bigquery"
      ],
      "id": "QfUqcTPLFUvi"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7nXx51Gg_yP8"
      },
      "source": [
        "### Mount Drive"
      ],
      "id": "7nXx51Gg_yP8"
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "L0bYBhW2j2Dl",
        "outputId": "1cbe55ce-f0dd-42ce-f46b-a39e33468db9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "id": "L0bYBhW2j2Dl"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zTIM-stc_17I"
      },
      "source": [
        "## Read dataframe and Check Nulls"
      ],
      "id": "zTIM-stc_17I"
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "wRD8-yldc04K",
        "outputId": "3fdf905b-6378-4ae2-91fa-23cb1b33ea17"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        Unnamed: 0    Year  Month   Day  Hour  Minute  Temperature  \\\n",
              "0                0  2018.0    1.0   1.0   0.0     0.0         11.2   \n",
              "1                1  2018.0    1.0   1.0   0.0    30.0         11.2   \n",
              "2                2  2018.0    1.0   1.0   1.0     0.0         11.1   \n",
              "3                3  2018.0    1.0   1.0   1.0    30.0         11.0   \n",
              "4                4  2018.0    1.0   1.0   2.0     0.0         10.9   \n",
              "...            ...     ...    ...   ...   ...     ...          ...   \n",
              "946181      946181  2013.0   12.0  31.0  21.0    30.0          6.0   \n",
              "946182      946182  2013.0   12.0  31.0  22.0     0.0          6.0   \n",
              "946183      946183  2013.0   12.0  31.0  22.0    30.0          5.0   \n",
              "946184      946184  2013.0   12.0  31.0  23.0     0.0          4.0   \n",
              "946185      946185  2013.0   12.0  31.0  23.0    30.0          4.0   \n",
              "\n",
              "        Cloud_Type  Dew_Point  GHI  Relative_Humidity  Solar_Zenith_Angle  \\\n",
              "0              4.0       10.2  0.0              93.63              165.41   \n",
              "1              4.0       10.2  0.0              93.63              165.11   \n",
              "2              4.0       10.1  0.0              93.38              162.29   \n",
              "3              0.0       10.1  0.0              94.00              157.91   \n",
              "4              0.0        9.9  0.0              93.32              152.72   \n",
              "...            ...        ...  ...                ...                 ...   \n",
              "946181         1.0        0.0  0.0              74.05              142.58   \n",
              "946182         1.0        0.0  0.0              74.40              148.35   \n",
              "946183         0.0        0.0  0.0              79.80              153.89   \n",
              "946184         0.0        0.0  0.0              80.32              158.95   \n",
              "946185         0.0        0.0  0.0              86.20              163.08   \n",
              "\n",
              "        Surface_Albedo  Pressure  Precipitable_Water  Wind_Direction  \\\n",
              "0                0.100    1024.0               1.700           294.0   \n",
              "1                0.100    1024.0               1.700           307.0   \n",
              "2                0.100    1024.0               1.700           321.0   \n",
              "3                0.100    1024.0               1.800           332.0   \n",
              "4                0.100    1024.0               1.800           344.0   \n",
              "...                ...       ...                 ...             ...   \n",
              "946181           0.113    1020.0               0.950           174.5   \n",
              "946182           0.113    1020.0               0.977           152.1   \n",
              "946183           0.113    1020.0               0.972           152.1   \n",
              "946184           0.113    1020.0               0.968           166.7   \n",
              "946185           0.113    1020.0               0.980           166.7   \n",
              "\n",
              "        Wind_Speed  Location  \n",
              "0              0.7  117987.0  \n",
              "1              0.7  117987.0  \n",
              "2              0.7  117987.0  \n",
              "3              0.7  117987.0  \n",
              "4              0.7  117987.0  \n",
              "...            ...       ...  \n",
              "946181         0.6  117546.0  \n",
              "946182         0.9  117546.0  \n",
              "946183         1.1  117546.0  \n",
              "946184         1.3  117546.0  \n",
              "946185         1.3  117546.0  \n",
              "\n",
              "[946186 rows x 18 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b3d2132f-a1b5-43c0-8256-fe29d3c1bbad\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>Year</th>\n",
              "      <th>Month</th>\n",
              "      <th>Day</th>\n",
              "      <th>Hour</th>\n",
              "      <th>Minute</th>\n",
              "      <th>Temperature</th>\n",
              "      <th>Cloud_Type</th>\n",
              "      <th>Dew_Point</th>\n",
              "      <th>GHI</th>\n",
              "      <th>Relative_Humidity</th>\n",
              "      <th>Solar_Zenith_Angle</th>\n",
              "      <th>Surface_Albedo</th>\n",
              "      <th>Pressure</th>\n",
              "      <th>Precipitable_Water</th>\n",
              "      <th>Wind_Direction</th>\n",
              "      <th>Wind_Speed</th>\n",
              "      <th>Location</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>11.2</td>\n",
              "      <td>4.0</td>\n",
              "      <td>10.2</td>\n",
              "      <td>0.0</td>\n",
              "      <td>93.63</td>\n",
              "      <td>165.41</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1024.0</td>\n",
              "      <td>1.700</td>\n",
              "      <td>294.0</td>\n",
              "      <td>0.7</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>11.2</td>\n",
              "      <td>4.0</td>\n",
              "      <td>10.2</td>\n",
              "      <td>0.0</td>\n",
              "      <td>93.63</td>\n",
              "      <td>165.11</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1024.0</td>\n",
              "      <td>1.700</td>\n",
              "      <td>307.0</td>\n",
              "      <td>0.7</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>11.1</td>\n",
              "      <td>4.0</td>\n",
              "      <td>10.1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>93.38</td>\n",
              "      <td>162.29</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1024.0</td>\n",
              "      <td>1.700</td>\n",
              "      <td>321.0</td>\n",
              "      <td>0.7</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>10.1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>94.00</td>\n",
              "      <td>157.91</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1024.0</td>\n",
              "      <td>1.800</td>\n",
              "      <td>332.0</td>\n",
              "      <td>0.7</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>10.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>9.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>93.32</td>\n",
              "      <td>152.72</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1024.0</td>\n",
              "      <td>1.800</td>\n",
              "      <td>344.0</td>\n",
              "      <td>0.7</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946181</th>\n",
              "      <td>946181</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>74.05</td>\n",
              "      <td>142.58</td>\n",
              "      <td>0.113</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>0.950</td>\n",
              "      <td>174.5</td>\n",
              "      <td>0.6</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946182</th>\n",
              "      <td>946182</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>74.40</td>\n",
              "      <td>148.35</td>\n",
              "      <td>0.113</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>0.977</td>\n",
              "      <td>152.1</td>\n",
              "      <td>0.9</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946183</th>\n",
              "      <td>946183</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>79.80</td>\n",
              "      <td>153.89</td>\n",
              "      <td>0.113</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>0.972</td>\n",
              "      <td>152.1</td>\n",
              "      <td>1.1</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946184</th>\n",
              "      <td>946184</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>80.32</td>\n",
              "      <td>158.95</td>\n",
              "      <td>0.113</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>0.968</td>\n",
              "      <td>166.7</td>\n",
              "      <td>1.3</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946185</th>\n",
              "      <td>946185</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>86.20</td>\n",
              "      <td>163.08</td>\n",
              "      <td>0.113</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>0.980</td>\n",
              "      <td>166.7</td>\n",
              "      <td>1.3</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>946186 rows × 18 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b3d2132f-a1b5-43c0-8256-fe29d3c1bbad')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-b3d2132f-a1b5-43c0-8256-fe29d3c1bbad button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-b3d2132f-a1b5-43c0-8256-fe29d3c1bbad');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ],
      "source": [
        "df= pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/all_cities_transformed_combined.csv')\n",
        "df"
      ],
      "id": "wRD8-yldc04K"
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "smyUNbQd9VLz",
        "outputId": "e569770f-40d5-44ca-a1eb-0062157ebfa8"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['Unnamed: 0', 'Year', 'Month', 'Day', 'Hour', 'Minute', 'Temperature',\n",
              "       'Cloud_Type', 'Dew_Point', 'GHI', 'Relative_Humidity',\n",
              "       'Solar_Zenith_Angle', 'Surface_Albedo', 'Pressure',\n",
              "       'Precipitable_Water', 'Wind_Direction', 'Wind_Speed', 'Location'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "df.columns"
      ],
      "id": "smyUNbQd9VLz"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IpkF9vC9_5XC"
      },
      "source": [
        "### Check Null and Remove"
      ],
      "id": "IpkF9vC9_5XC"
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DN8llovEd3Qb",
        "outputId": "76d86b55-a13b-415e-86cc-3c5606272750"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Unnamed: 0            0\n",
              "Year                  0\n",
              "Month                 0\n",
              "Day                   0\n",
              "Hour                  0\n",
              "Minute                0\n",
              "Temperature           0\n",
              "Cloud_Type            0\n",
              "Dew_Point             0\n",
              "GHI                   0\n",
              "Relative_Humidity     0\n",
              "Solar_Zenith_Angle    0\n",
              "Surface_Albedo        0\n",
              "Pressure              0\n",
              "Precipitable_Water    0\n",
              "Wind_Direction        0\n",
              "Wind_Speed            0\n",
              "Location              0\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "# nsrdb_na = df.drop(columns='Unnamed: 0')\n",
        "nsrdb_na = df.copy()\n",
        "#null_rows = nsrdb_na[nsrdb_na.isna().any(axis=1)]\n",
        "#null_rows\n",
        "nsrdb_na.isnull().sum()\n",
        "#nsrdb_na.columns\n",
        "#nsrdb_na['GHI'].isnull().values.sum()\n",
        "nsrdb = nsrdb_na.dropna()\n",
        "#print(nsrdb)\n",
        "nsrdb.isnull().sum()"
      ],
      "id": "DN8llovEd3Qb"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6APnUz9Xniz5"
      },
      "source": [
        "## Feature Selection"
      ],
      "id": "6APnUz9Xniz5"
    },
    {
      "cell_type": "code",
      "source": [
        "nsrdb = nsrdb.loc[(nsrdb['Hour'].isin([7, 10, 13, 16, 19])) & (nsrdb['Minute'] == 0)]"
      ],
      "metadata": {
        "id": "_uJe7Jdd5pCz"
      },
      "id": "_uJe7Jdd5pCz",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 500
        },
        "id": "QHrcka2sZRQa",
        "outputId": "da678e01-d8b4-4cd9-dd01-de968125f4f5"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        Unnamed: 0    Year  Month   Day  Hour  Minute  Temperature  \\\n",
              "14              14  2018.0    1.0   1.0   7.0     0.0          9.8   \n",
              "20              20  2018.0    1.0   1.0  10.0     0.0         14.0   \n",
              "26              26  2018.0    1.0   1.0  13.0     0.0         17.6   \n",
              "32              32  2018.0    1.0   1.0  16.0     0.0         15.4   \n",
              "38              38  2018.0    1.0   1.0  19.0     0.0         12.8   \n",
              "...            ...     ...    ...   ...   ...     ...          ...   \n",
              "946152      946152  2013.0   12.0  31.0   7.0     0.0          9.0   \n",
              "946158      946158  2013.0   12.0  31.0  10.0     0.0         13.0   \n",
              "946164      946164  2013.0   12.0  31.0  13.0     0.0         17.0   \n",
              "946170      946170  2013.0   12.0  31.0  16.0     0.0          7.0   \n",
              "946176      946176  2013.0   12.0  31.0  19.0     0.0          7.0   \n",
              "\n",
              "        Cloud_Type  Dew_Point  Relative_Humidity  Solar_Zenith_Angle  \\\n",
              "14             7.0        8.8              93.25               94.85   \n",
              "20             7.0        9.5              74.48               67.81   \n",
              "26             7.0        8.6              55.38               61.32   \n",
              "32             7.0        9.2              66.62               80.55   \n",
              "38             7.0        8.3              74.04              112.83   \n",
              "...            ...        ...                ...                 ...   \n",
              "946152         7.0        8.0             100.00               94.75   \n",
              "946158         3.0        7.0              74.38               67.81   \n",
              "946164         7.0        4.0              49.14               61.44   \n",
              "946170         4.0        1.0              73.84               80.77   \n",
              "946176         4.0        0.0              61.15              112.97   \n",
              "\n",
              "        Surface_Albedo  Pressure  Precipitable_Water  Wind_Direction  \\\n",
              "14               0.100    1025.0               1.800           325.0   \n",
              "20               0.100    1026.0               1.800            24.0   \n",
              "26               0.100    1024.0               1.800            15.0   \n",
              "32               0.100    1023.0               1.800            14.0   \n",
              "38               0.100    1024.0               1.700            72.0   \n",
              "...                ...       ...                 ...             ...   \n",
              "946152           0.111    1020.0               1.233           174.6   \n",
              "946158           0.111    1020.0               1.403            23.4   \n",
              "946164           0.111    1020.0               1.334            33.3   \n",
              "946170           0.113    1020.0               0.999           343.0   \n",
              "946176           0.113    1020.0               0.842           226.1   \n",
              "\n",
              "        Wind_Speed  Location  \n",
              "14             0.7  117987.0  \n",
              "20             1.5  117987.0  \n",
              "26             0.9  117987.0  \n",
              "32             0.8  117987.0  \n",
              "38             1.0  117987.0  \n",
              "...            ...       ...  \n",
              "946152         0.4  117546.0  \n",
              "946158         0.7  117546.0  \n",
              "946164         1.1  117546.0  \n",
              "946170         0.8  117546.0  \n",
              "946176         0.5  117546.0  \n",
              "\n",
              "[98550 rows x 17 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-2a79146e-9941-48a3-b15d-2823c82701e0\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>Year</th>\n",
              "      <th>Month</th>\n",
              "      <th>Day</th>\n",
              "      <th>Hour</th>\n",
              "      <th>Minute</th>\n",
              "      <th>Temperature</th>\n",
              "      <th>Cloud_Type</th>\n",
              "      <th>Dew_Point</th>\n",
              "      <th>Relative_Humidity</th>\n",
              "      <th>Solar_Zenith_Angle</th>\n",
              "      <th>Surface_Albedo</th>\n",
              "      <th>Pressure</th>\n",
              "      <th>Precipitable_Water</th>\n",
              "      <th>Wind_Direction</th>\n",
              "      <th>Wind_Speed</th>\n",
              "      <th>Location</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>14</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>9.8</td>\n",
              "      <td>7.0</td>\n",
              "      <td>8.8</td>\n",
              "      <td>93.25</td>\n",
              "      <td>94.85</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1025.0</td>\n",
              "      <td>1.800</td>\n",
              "      <td>325.0</td>\n",
              "      <td>0.7</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20</th>\n",
              "      <td>20</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>9.5</td>\n",
              "      <td>74.48</td>\n",
              "      <td>67.81</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1026.0</td>\n",
              "      <td>1.800</td>\n",
              "      <td>24.0</td>\n",
              "      <td>1.5</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>26</th>\n",
              "      <td>26</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>17.6</td>\n",
              "      <td>7.0</td>\n",
              "      <td>8.6</td>\n",
              "      <td>55.38</td>\n",
              "      <td>61.32</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1024.0</td>\n",
              "      <td>1.800</td>\n",
              "      <td>15.0</td>\n",
              "      <td>0.9</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32</th>\n",
              "      <td>32</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>15.4</td>\n",
              "      <td>7.0</td>\n",
              "      <td>9.2</td>\n",
              "      <td>66.62</td>\n",
              "      <td>80.55</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1023.0</td>\n",
              "      <td>1.800</td>\n",
              "      <td>14.0</td>\n",
              "      <td>0.8</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>38</th>\n",
              "      <td>38</td>\n",
              "      <td>2018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>12.8</td>\n",
              "      <td>7.0</td>\n",
              "      <td>8.3</td>\n",
              "      <td>74.04</td>\n",
              "      <td>112.83</td>\n",
              "      <td>0.100</td>\n",
              "      <td>1024.0</td>\n",
              "      <td>1.700</td>\n",
              "      <td>72.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>117987.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946152</th>\n",
              "      <td>946152</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>94.75</td>\n",
              "      <td>0.111</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>1.233</td>\n",
              "      <td>174.6</td>\n",
              "      <td>0.4</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946158</th>\n",
              "      <td>946158</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>74.38</td>\n",
              "      <td>67.81</td>\n",
              "      <td>0.111</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>1.403</td>\n",
              "      <td>23.4</td>\n",
              "      <td>0.7</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946164</th>\n",
              "      <td>946164</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>17.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>49.14</td>\n",
              "      <td>61.44</td>\n",
              "      <td>0.111</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>1.334</td>\n",
              "      <td>33.3</td>\n",
              "      <td>1.1</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946170</th>\n",
              "      <td>946170</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>73.84</td>\n",
              "      <td>80.77</td>\n",
              "      <td>0.113</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>0.999</td>\n",
              "      <td>343.0</td>\n",
              "      <td>0.8</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946176</th>\n",
              "      <td>946176</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>61.15</td>\n",
              "      <td>112.97</td>\n",
              "      <td>0.113</td>\n",
              "      <td>1020.0</td>\n",
              "      <td>0.842</td>\n",
              "      <td>226.1</td>\n",
              "      <td>0.5</td>\n",
              "      <td>117546.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>98550 rows × 17 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2a79146e-9941-48a3-b15d-2823c82701e0')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-2a79146e-9941-48a3-b15d-2823c82701e0 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-2a79146e-9941-48a3-b15d-2823c82701e0');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "target = nsrdb['GHI']\n",
        "input= nsrdb.drop(['GHI'], axis=1)\n",
        "input"
      ],
      "id": "QHrcka2sZRQa"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "fiX8XeBjBgeE"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "x_train, x_test, y_train, y_test = train_test_split(input, target, test_size=0.3, random_state=42)"
      ],
      "id": "fiX8XeBjBgeE"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6CYngi19Aayy"
      },
      "outputs": [],
      "source": [
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "tree = DecisionTreeClassifier(criterion='entropy',max_depth=3, random_state=0)\n",
        "forest = RandomForestClassifier(criterion='entropy',max_depth=3, n_estimators=500, random_state=0)"
      ],
      "id": "6CYngi19Aayy"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "r4CDSmcHAa15"
      },
      "outputs": [],
      "source": [
        "forest.fit(x_train, y_train)\n",
        "importances = forest.feature_importances_"
      ],
      "id": "r4CDSmcHAa15"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "UrVEuAPMAa4t"
      },
      "outputs": [],
      "source": [
        "std = np.std([tree.feature_importances_ for tree in forest.estimators_] , axis=0)\n",
        "df = pd.DataFrame({'feature' : x_train.columns, 'importance': importances, 'std': std})\n",
        "df = df.sort_values('importance')"
      ],
      "id": "UrVEuAPMAa4t"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aWW5eMuLAa7i",
        "outputId": "d96af40d-a3c7-4850-9693-59f5eb140056"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "               feature  importance       std\n",
            "4               Minute    0.000000  0.000000\n",
            "15            Location    0.000027  0.000602\n",
            "11            Pressure    0.000062  0.000872\n",
            "0                 Year    0.000455  0.001741\n",
            "2                  Day    0.003226  0.009931\n",
            "10      Surface_Albedo    0.003459  0.015442\n",
            "12  Precipitable_Water    0.003639  0.011958\n",
            "7            Dew_Point    0.007411  0.021134\n",
            "13      Wind_Direction    0.010143  0.027043\n",
            "14          Wind_Speed    0.027515  0.050844\n",
            "1                Month    0.043073  0.066045\n",
            "6           Cloud_Type    0.053154  0.073407\n",
            "5          Temperature    0.090706  0.131901\n",
            "8    Relative_Humidity    0.140360  0.163047\n",
            "3                 Hour    0.204011  0.207794\n",
            "9   Solar_Zenith_Angle    0.412759  0.283338\n"
          ]
        }
      ],
      "source": [
        "print(df)"
      ],
      "id": "aWW5eMuLAa7i"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "bH6PQzFNAa-W",
        "outputId": "e142505f-e5ba-4496-e09c-042158d0826f"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-24cd55ed-735d-495f-8fc3-13e0ee5fdabb\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>feature</th>\n",
              "      <th>importance</th>\n",
              "      <th>std</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Solar_Zenith_Angle</td>\n",
              "      <td>0.412759</td>\n",
              "      <td>0.283338</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Hour</td>\n",
              "      <td>0.204011</td>\n",
              "      <td>0.207794</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Relative_Humidity</td>\n",
              "      <td>0.140360</td>\n",
              "      <td>0.163047</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Temperature</td>\n",
              "      <td>0.090706</td>\n",
              "      <td>0.131901</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Cloud_Type</td>\n",
              "      <td>0.053154</td>\n",
              "      <td>0.073407</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Month</td>\n",
              "      <td>0.043073</td>\n",
              "      <td>0.066045</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Wind_Speed</td>\n",
              "      <td>0.027515</td>\n",
              "      <td>0.050844</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Wind_Direction</td>\n",
              "      <td>0.010143</td>\n",
              "      <td>0.027043</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Dew_Point</td>\n",
              "      <td>0.007411</td>\n",
              "      <td>0.021134</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>Precipitable_Water</td>\n",
              "      <td>0.003639</td>\n",
              "      <td>0.011958</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>Surface_Albedo</td>\n",
              "      <td>0.003459</td>\n",
              "      <td>0.015442</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>Day</td>\n",
              "      <td>0.003226</td>\n",
              "      <td>0.009931</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>Year</td>\n",
              "      <td>0.000455</td>\n",
              "      <td>0.001741</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>Pressure</td>\n",
              "      <td>0.000062</td>\n",
              "      <td>0.000872</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>Location</td>\n",
              "      <td>0.000027</td>\n",
              "      <td>0.000602</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>Minute</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-24cd55ed-735d-495f-8fc3-13e0ee5fdabb')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-24cd55ed-735d-495f-8fc3-13e0ee5fdabb button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-24cd55ed-735d-495f-8fc3-13e0ee5fdabb');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "               feature  importance       std\n",
              "0   Solar_Zenith_Angle    0.412759  0.283338\n",
              "1                 Hour    0.204011  0.207794\n",
              "2    Relative_Humidity    0.140360  0.163047\n",
              "3          Temperature    0.090706  0.131901\n",
              "4           Cloud_Type    0.053154  0.073407\n",
              "5                Month    0.043073  0.066045\n",
              "6           Wind_Speed    0.027515  0.050844\n",
              "7       Wind_Direction    0.010143  0.027043\n",
              "8            Dew_Point    0.007411  0.021134\n",
              "9   Precipitable_Water    0.003639  0.011958\n",
              "10      Surface_Albedo    0.003459  0.015442\n",
              "11                 Day    0.003226  0.009931\n",
              "12                Year    0.000455  0.001741\n",
              "13            Pressure    0.000062  0.000872\n",
              "14            Location    0.000027  0.000602\n",
              "15              Minute    0.000000  0.000000"
            ]
          },
          "execution_count": 13,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "ax = df.plot(kind='barh', xerr='std', x='feature', legend=False,figsize=(15, 8))\n",
        "ax.set_ylabel(' ')\n",
        "ax.set_title('Feature Importance')\n",
        "df.sort_values(by=['importance'],ascending = False,ignore_index = True).head(16)"
      ],
      "id": "bH6PQzFNAa-W"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 551
        },
        "id": "X4dukwejAbB3",
        "outputId": "6956e38e-9501-4165-8169-bd26a10e30f1"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-8472ec4b-1076-47ba-9ec3-a91e031d7d18\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>feature</th>\n",
              "      <th>importance</th>\n",
              "      <th>std</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Solar_Zenith_Angle</td>\n",
              "      <td>0.412759</td>\n",
              "      <td>0.283338</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Hour</td>\n",
              "      <td>0.204011</td>\n",
              "      <td>0.207794</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Relative_Humidity</td>\n",
              "      <td>0.140360</td>\n",
              "      <td>0.163047</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Temperature</td>\n",
              "      <td>0.090706</td>\n",
              "      <td>0.131901</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Cloud_Type</td>\n",
              "      <td>0.053154</td>\n",
              "      <td>0.073407</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Month</td>\n",
              "      <td>0.043073</td>\n",
              "      <td>0.066045</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Wind_Speed</td>\n",
              "      <td>0.027515</td>\n",
              "      <td>0.050844</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Wind_Direction</td>\n",
              "      <td>0.010143</td>\n",
              "      <td>0.027043</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Dew_Point</td>\n",
              "      <td>0.007411</td>\n",
              "      <td>0.021134</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>Precipitable_Water</td>\n",
              "      <td>0.003639</td>\n",
              "      <td>0.011958</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>Surface_Albedo</td>\n",
              "      <td>0.003459</td>\n",
              "      <td>0.015442</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>Day</td>\n",
              "      <td>0.003226</td>\n",
              "      <td>0.009931</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>Year</td>\n",
              "      <td>0.000455</td>\n",
              "      <td>0.001741</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>Pressure</td>\n",
              "      <td>0.000062</td>\n",
              "      <td>0.000872</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>Location</td>\n",
              "      <td>0.000027</td>\n",
              "      <td>0.000602</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>Minute</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8472ec4b-1076-47ba-9ec3-a91e031d7d18')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-8472ec4b-1076-47ba-9ec3-a91e031d7d18 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-8472ec4b-1076-47ba-9ec3-a91e031d7d18');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "               feature  importance       std\n",
              "0   Solar_Zenith_Angle    0.412759  0.283338\n",
              "1                 Hour    0.204011  0.207794\n",
              "2    Relative_Humidity    0.140360  0.163047\n",
              "3          Temperature    0.090706  0.131901\n",
              "4           Cloud_Type    0.053154  0.073407\n",
              "5                Month    0.043073  0.066045\n",
              "6           Wind_Speed    0.027515  0.050844\n",
              "7       Wind_Direction    0.010143  0.027043\n",
              "8            Dew_Point    0.007411  0.021134\n",
              "9   Precipitable_Water    0.003639  0.011958\n",
              "10      Surface_Albedo    0.003459  0.015442\n",
              "11                 Day    0.003226  0.009931\n",
              "12                Year    0.000455  0.001741\n",
              "13            Pressure    0.000062  0.000872\n",
              "14            Location    0.000027  0.000602\n",
              "15              Minute    0.000000  0.000000"
            ]
          },
          "execution_count": 14,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df.sort_values(by=['importance'],ascending = False,ignore_index = True).head(16)"
      ],
      "id": "X4dukwejAbB3"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "co7Cbu8ogrRa"
      },
      "outputs": [],
      "source": [
        "\"\"\"from datetime import datetime\n",
        "nsrdb['Date'] = nsrdb.apply(lambda row: datetime.strptime(f\"{int(row.Year)}-{int(row.Month)}-{int(row.Day)}\", '%Y-%m-%d'), axis=1)\n",
        "nsrdb['Date'] = pd.to_datetime(nsrdb['Date'])\n",
        "nsrdb.info()\n",
        "nsrdb = nsrdb.resample('D', on='Date').mean()\n",
        "#nsrdb.drop(['Hour'],\t['Minute'], axis=1)\n",
        "nsrdb\"\"\""
      ],
      "id": "co7Cbu8ogrRa"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E4HR6rJwDgC6"
      },
      "source": [
        "## Train Model with selected features 11"
      ],
      "id": "E4HR6rJwDgC6"
    },
    {
      "cell_type": "code",
      "execution_count": 39,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RiBNXTy9DzIP",
        "outputId": "ef49309c-1036-4958-a2d0-1622b9af6897"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Index(['Unnamed: 0', 'Year', 'Month', 'Day', 'Hour', 'Minute', 'Temperature',\n",
            "       'Cloud_Type', 'Dew_Point', 'GHI', 'Relative_Humidity',\n",
            "       'Solar_Zenith_Angle', 'Surface_Albedo', 'Pressure',\n",
            "       'Precipitable_Water', 'Wind_Direction', 'Wind_Speed', 'Location'],\n",
            "      dtype='object')\n",
            "Index(['Month', 'Day', 'Hour', 'Temperature', 'Cloud_Type', 'Dew_Point',\n",
            "       'Relative_Humidity', 'Solar_Zenith_Angle', 'Wind_Direction',\n",
            "       'Wind_Speed', 'Location'],\n",
            "      dtype='object')\n",
            "x_train shape (48289, 11)\n",
            "x_test shape (29565, 11)\n",
            "y_train shape (48289,)\n",
            "y_text shape (29565,)\n"
          ]
        }
      ],
      "source": [
        "nsrdb= pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/all_cities_transformed_combined.csv')\n",
        "nsrdb = nsrdb.loc[(nsrdb['Hour'].isin([7, 10, 13, 16, 19])) & (nsrdb['Minute'] == 0)]\n",
        "print(nsrdb.columns)\n",
        "X = nsrdb.drop(['Unnamed: 0','GHI','Precipitable_Water', 'Surface_Albedo','Year','Pressure','Minute'], axis =1)\n",
        "print(X.columns)\n",
        "y=nsrdb['GHI']\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_normalized = sc.fit_transform(X)\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "x_train, x_test, y_train, y_test = train_test_split(input_normalized,y, test_size=0.3, random_state=42)\n",
        "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.3, random_state=42)\n",
        "\n",
        "print(\"x_train shape\",x_train.shape)\n",
        "print(\"x_test shape\",x_test.shape)\n",
        "print(\"y_train shape\",y_train.shape)\n",
        "print(\"y_text shape\",y_test.shape)"
      ],
      "id": "RiBNXTy9DzIP"
    },
    {
      "cell_type": "code",
      "execution_count": 40,
      "metadata": {
        "id": "SQVT838gHMON"
      },
      "outputs": [],
      "source": [
        "# https://github.com/tpt5cu/solarIrradiancePredictor/blob/master/Neural_Net.ipynb\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "import pandas as pd\n",
        "import os\n",
        "import numpy as np\n",
        "from glob import glob\n",
        "from matplotlib import pyplot as plt\n",
        "import seaborn as sns\n",
        "from random import randint\n",
        "from glob import glob\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "from tensorflow.keras import regularizers\n",
        "\n",
        "callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=3)\n",
        "model = keras.Sequential()\n",
        "model.add(layers.InputLayer(input_shape=x_train.shape[1:]))\n",
        "model.add(tf.keras.layers.Normalization())\n",
        "model.add(layers.Dense(300, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(200, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(100, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(50, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(25, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(1, activation='relu', kernel_regularizer=regularizers.L1(0.01)))\n",
        "\n",
        "#Compile and train\n",
        "model.compile(optimizer=Adam(), \n",
        "              loss='mse',\n",
        "              metrics=['mae',keras.metrics.RootMeanSquaredError()])"
      ],
      "id": "SQVT838gHMON"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xC4n6_SWHg10"
      },
      "outputs": [],
      "source": [
        "\"\"\"keras.backend.clear_session()\n",
        "tf.keras.backend.clear_session()\n",
        "tf.compat.v1.reset_default_graph\"\"\""
      ],
      "id": "xC4n6_SWHg10"
    },
    {
      "cell_type": "code",
      "execution_count": 41,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Soy8GX1jlJrq",
        "outputId": "9c54ffd5-33e1-41a9-c9a2-ed66af7ee601"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "755/755 [==============================] - 8s 9ms/step - loss: 16698.7656 - mae: 65.1943 - root_mean_squared_error: 129.0118 - val_loss: 3823.3608 - val_mae: 34.3180 - val_root_mean_squared_error: 61.4095\n",
            "Epoch 2/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 3588.4807 - mae: 33.3215 - root_mean_squared_error: 59.4741 - val_loss: 3154.2656 - val_mae: 29.2042 - val_root_mean_squared_error: 55.7073\n",
            "Epoch 3/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 3112.9756 - mae: 29.5416 - root_mean_squared_error: 55.3351 - val_loss: 2890.5842 - val_mae: 28.1139 - val_root_mean_squared_error: 53.2884\n",
            "Epoch 4/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2945.8101 - mae: 28.4727 - root_mean_squared_error: 53.8034 - val_loss: 2761.8696 - val_mae: 26.4743 - val_root_mean_squared_error: 52.0655\n",
            "Epoch 5/100\n",
            "755/755 [==============================] - 10s 13ms/step - loss: 2850.0808 - mae: 27.7449 - root_mean_squared_error: 52.9047 - val_loss: 3234.2554 - val_mae: 34.9648 - val_root_mean_squared_error: 56.4182\n",
            "Epoch 6/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2789.7598 - mae: 27.2347 - root_mean_squared_error: 52.3293 - val_loss: 2793.0852 - val_mae: 26.7419 - val_root_mean_squared_error: 52.3597\n",
            "Epoch 7/100\n",
            "755/755 [==============================] - 8s 11ms/step - loss: 2773.4978 - mae: 27.2871 - root_mean_squared_error: 52.1713 - val_loss: 2688.5105 - val_mae: 26.8470 - val_root_mean_squared_error: 51.3493\n",
            "Epoch 8/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2711.6819 - mae: 26.7766 - root_mean_squared_error: 51.5732 - val_loss: 2629.6226 - val_mae: 26.1750 - val_root_mean_squared_error: 50.7700\n",
            "Epoch 9/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2694.2158 - mae: 26.6667 - root_mean_squared_error: 51.4011 - val_loss: 2926.1055 - val_mae: 31.3762 - val_root_mean_squared_error: 53.6091\n",
            "Epoch 10/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2662.5911 - mae: 26.4145 - root_mean_squared_error: 51.0902 - val_loss: 2606.6672 - val_mae: 26.1553 - val_root_mean_squared_error: 50.5393\n",
            "Epoch 11/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2632.4561 - mae: 26.3857 - root_mean_squared_error: 50.7920 - val_loss: 2632.8132 - val_mae: 26.2769 - val_root_mean_squared_error: 50.7941\n",
            "Epoch 12/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2608.7388 - mae: 25.9892 - root_mean_squared_error: 50.5566 - val_loss: 2695.3967 - val_mae: 26.2957 - val_root_mean_squared_error: 51.4063\n",
            "Epoch 13/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2597.4297 - mae: 26.0022 - root_mean_squared_error: 50.4445 - val_loss: 2519.8132 - val_mae: 25.7180 - val_root_mean_squared_error: 49.6693\n",
            "Epoch 14/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 2577.6924 - mae: 25.8027 - root_mean_squared_error: 50.2480 - val_loss: 2581.9575 - val_mae: 26.8115 - val_root_mean_squared_error: 50.2908\n",
            "Epoch 15/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2547.3311 - mae: 25.6809 - root_mean_squared_error: 49.9449 - val_loss: 2517.7507 - val_mae: 26.2919 - val_root_mean_squared_error: 49.6475\n",
            "Epoch 16/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2541.6775 - mae: 25.7434 - root_mean_squared_error: 49.8876 - val_loss: 2546.2266 - val_mae: 25.5492 - val_root_mean_squared_error: 49.9326\n",
            "Epoch 17/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2528.3135 - mae: 25.5729 - root_mean_squared_error: 49.7532 - val_loss: 2636.5144 - val_mae: 28.2793 - val_root_mean_squared_error: 50.8292\n",
            "Epoch 18/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2515.9829 - mae: 25.5975 - root_mean_squared_error: 49.6283 - val_loss: 2393.7773 - val_mae: 23.9058 - val_root_mean_squared_error: 48.3816\n",
            "Epoch 19/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 2462.2942 - mae: 24.8063 - root_mean_squared_error: 49.0843 - val_loss: 2347.6238 - val_mae: 23.0507 - val_root_mean_squared_error: 47.9017\n",
            "Epoch 20/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2442.5146 - mae: 24.9514 - root_mean_squared_error: 48.8823 - val_loss: 2295.7283 - val_mae: 23.3541 - val_root_mean_squared_error: 47.3570\n",
            "Epoch 21/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2450.4905 - mae: 25.1641 - root_mean_squared_error: 48.9633 - val_loss: 2408.1821 - val_mae: 24.4753 - val_root_mean_squared_error: 48.5292\n",
            "Epoch 22/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2438.0671 - mae: 25.0556 - root_mean_squared_error: 48.8361 - val_loss: 2398.6194 - val_mae: 25.3960 - val_root_mean_squared_error: 48.4305\n",
            "Epoch 23/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2416.9419 - mae: 24.9676 - root_mean_squared_error: 48.6193 - val_loss: 2282.0598 - val_mae: 24.5455 - val_root_mean_squared_error: 47.2120\n",
            "Epoch 24/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2385.3962 - mae: 24.7778 - root_mean_squared_error: 48.2945 - val_loss: 2290.4446 - val_mae: 24.0908 - val_root_mean_squared_error: 47.3010\n",
            "Epoch 25/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2370.7708 - mae: 24.6581 - root_mean_squared_error: 48.1429 - val_loss: 2208.3789 - val_mae: 23.1287 - val_root_mean_squared_error: 46.4263\n",
            "Epoch 26/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2347.9973 - mae: 24.4795 - root_mean_squared_error: 47.9071 - val_loss: 2262.3472 - val_mae: 24.6733 - val_root_mean_squared_error: 47.0051\n",
            "Epoch 27/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2339.8931 - mae: 24.6537 - root_mean_squared_error: 47.8227 - val_loss: 2284.0613 - val_mae: 24.0656 - val_root_mean_squared_error: 47.2362\n",
            "Epoch 28/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2314.6025 - mae: 24.3862 - root_mean_squared_error: 47.5587 - val_loss: 2430.2190 - val_mae: 26.1858 - val_root_mean_squared_error: 48.7591\n",
            "Epoch 29/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2291.3462 - mae: 24.2743 - root_mean_squared_error: 47.3139 - val_loss: 2146.9873 - val_mae: 22.4874 - val_root_mean_squared_error: 45.7644\n",
            "Epoch 30/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2300.0330 - mae: 24.3848 - root_mean_squared_error: 47.4070 - val_loss: 2170.0999 - val_mae: 22.8120 - val_root_mean_squared_error: 46.0166\n",
            "Epoch 31/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2247.9722 - mae: 23.8728 - root_mean_squared_error: 46.8553 - val_loss: 2205.3562 - val_mae: 25.0974 - val_root_mean_squared_error: 46.3992\n",
            "Epoch 32/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2242.7615 - mae: 24.0533 - root_mean_squared_error: 46.8004 - val_loss: 2072.7109 - val_mae: 22.0553 - val_root_mean_squared_error: 44.9483\n",
            "Epoch 33/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2231.9094 - mae: 24.0843 - root_mean_squared_error: 46.6861 - val_loss: 2100.4021 - val_mae: 22.8775 - val_root_mean_squared_error: 45.2562\n",
            "Epoch 34/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2213.3684 - mae: 23.8659 - root_mean_squared_error: 46.4870 - val_loss: 2101.9802 - val_mae: 23.0545 - val_root_mean_squared_error: 45.2731\n",
            "Epoch 35/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2194.3945 - mae: 23.6130 - root_mean_squared_error: 46.2829 - val_loss: 2116.4736 - val_mae: 24.2719 - val_root_mean_squared_error: 45.4345\n",
            "Epoch 36/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2183.2466 - mae: 23.6572 - root_mean_squared_error: 46.1636 - val_loss: 2103.0220 - val_mae: 22.7883 - val_root_mean_squared_error: 45.2861\n",
            "Epoch 37/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2145.9692 - mae: 23.4362 - root_mean_squared_error: 45.7583 - val_loss: 2046.6117 - val_mae: 22.5998 - val_root_mean_squared_error: 44.6596\n",
            "Epoch 38/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2166.2446 - mae: 23.5536 - root_mean_squared_error: 45.9789 - val_loss: 2055.1145 - val_mae: 23.1120 - val_root_mean_squared_error: 44.7543\n",
            "Epoch 39/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2147.5273 - mae: 23.6766 - root_mean_squared_error: 45.7745 - val_loss: 2060.4597 - val_mae: 22.1431 - val_root_mean_squared_error: 44.8131\n",
            "Epoch 40/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2151.0908 - mae: 23.8401 - root_mean_squared_error: 45.8127 - val_loss: 2350.0938 - val_mae: 27.7510 - val_root_mean_squared_error: 47.9360\n"
          ]
        }
      ],
      "source": [
        "history = model.fit(x_train, y_train, epochs=100, batch_size=64, validation_data=(x_train,y_train),callbacks=[callback])"
      ],
      "id": "Soy8GX1jlJrq"
    },
    {
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7vYaIR3aAWAq",
        "outputId": "2432d229-3ea3-4d30-b05c-6604cdd4c18e"
      },
      "id": "7vYaIR3aAWAq",
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " normalization_1 (Normalizat  (None, 11)               23        \n",
            " ion)                                                            \n",
            "                                                                 \n",
            " dense_6 (Dense)             (None, 300)               3600      \n",
            "                                                                 \n",
            " dense_7 (Dense)             (None, 200)               60200     \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 100)               20100     \n",
            "                                                                 \n",
            " dense_9 (Dense)             (None, 50)                5050      \n",
            "                                                                 \n",
            " dense_10 (Dense)            (None, 25)                1275      \n",
            "                                                                 \n",
            " dense_11 (Dense)            (None, 1)                 26        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 90,274\n",
            "Trainable params: 90,251\n",
            "Non-trainable params: 23\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Save model history in Json or CSV"
      ],
      "metadata": {
        "id": "SdwLZuSJAihE"
      },
      "id": "SdwLZuSJAihE"
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "# convert the history.history dict to a pandas DataFrame:     \n",
        "hist_df = pd.DataFrame(history.history) \n",
        "\n",
        "# or save to csv: \n",
        "hist_csv_file = 'history_epochs100_layers7_L2_reg_selectedfeatures_11.csv'\n",
        "with open(hist_csv_file, mode='w') as f:\n",
        "    hist_df.to_csv(f)\n",
        "\n",
        "# save to json:  \n",
        "hist_json_file = 'history_epochs100_layers7_L2_reg_selectedfeatures_11.json' \n",
        "with open(hist_json_file, mode='w') as f:\n",
        "    hist_df.to_json(f)"
      ],
      "metadata": {
        "id": "JgObNEHPAdye"
      },
      "id": "JgObNEHPAdye",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['mae'])\n",
        "plt.plot(history.history['root_mean_squared_error'])\n",
        "plt.title('MAE and RMSE per epoch')\n",
        "plt.ylabel('Error')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['MAE', 'RMSE'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "r7y99rlzDcou",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "76ad567f-533d-4b6a-8488-2c30f7f55139"
      },
      "id": "r7y99rlzDcou",
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred=model.predict(x_test)\n",
        "y_pred"
      ],
      "metadata": {
        "id": "cNoOVxXzDdjE",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "83375e2d-57ab-4ca7-da0f-353b65afb668"
      },
      "id": "cNoOVxXzDdjE",
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "924/924 [==============================] - 4s 4ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[  0.     ],\n",
              "       [744.3603 ],\n",
              "       [  0.     ],\n",
              "       ...,\n",
              "       [423.32278],\n",
              "       [258.19257],\n",
              "       [  0.     ]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "35RiuByMD5If",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "c69fe6c5-b596-4ae9-8812-d1c03498ee0f"
      },
      "id": "35RiuByMD5If",
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import r2_score \n",
        "from sklearn import metrics\n",
        "R_square = r2_score(y_test, y_pred) \n",
        "print('Coefficient of Determination R2', R_square)\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(y_test, y_pred))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(y_test, y_pred))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(y_test, y_pred))) "
      ],
      "metadata": {
        "id": "tR0xHyhkD8Tr",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e7fad537-d6a4-423e-9df7-b8642a7ac288"
      },
      "id": "tR0xHyhkD8Tr",
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Coefficient of Determination R2 0.9717837293999062\n",
            "Model MAE 30.10841577063856\n",
            "Model MSE 2936.9666438303893\n",
            "Model RMSE 54.19378787121629\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Train Model with selected features 12 (Included Surface_Albedo)"
      ],
      "metadata": {
        "id": "-ewkt8vs-tWp"
      },
      "id": "-ewkt8vs-tWp"
    },
    {
      "cell_type": "code",
      "source": [
        "nsrdb= pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/all_cities_transformed_combined.csv')\n",
        "nsrdb = nsrdb.loc[(nsrdb['Hour'].isin([7, 10, 13, 16, 19])) & (nsrdb['Minute'] == 0)]\n",
        "print(nsrdb.columns)\n",
        "X = nsrdb.drop(['Unnamed: 0','GHI','Precipitable_Water','Year','Pressure','Minute'], axis =1)\n",
        "print(X.columns)\n",
        "y=nsrdb['GHI']\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_normalized = sc.fit_transform(X)\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "x_train, x_test, y_train, y_test = train_test_split(input_normalized,y, test_size=0.3, random_state=42)\n",
        "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.3, random_state=42)\n",
        "\n",
        "print(\"x_train shape\",x_train.shape)\n",
        "print(\"x_test shape\",x_test.shape)\n",
        "print(\"y_train shape\",y_train.shape)\n",
        "print(\"y_text shape\",y_test.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w6fkKIUt-t1V",
        "outputId": "bfe75432-c651-4c4d-88ed-76164a0bbbe3"
      },
      "id": "w6fkKIUt-t1V",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Index(['Unnamed: 0', 'Year', 'Month', 'Day', 'Hour', 'Minute', 'Temperature',\n",
            "       'Cloud_Type', 'Dew_Point', 'GHI', 'Relative_Humidity',\n",
            "       'Solar_Zenith_Angle', 'Surface_Albedo', 'Pressure',\n",
            "       'Precipitable_Water', 'Wind_Direction', 'Wind_Speed', 'Location'],\n",
            "      dtype='object')\n",
            "Index(['Month', 'Day', 'Hour', 'Temperature', 'Cloud_Type', 'Dew_Point',\n",
            "       'Relative_Humidity', 'Solar_Zenith_Angle', 'Surface_Albedo',\n",
            "       'Wind_Direction', 'Wind_Speed', 'Location'],\n",
            "      dtype='object')\n",
            "x_train shape (48289, 12)\n",
            "x_test shape (29565, 12)\n",
            "y_train shape (48289,)\n",
            "y_text shape (29565,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=3)\n",
        "model = keras.Sequential()\n",
        "model.add(layers.InputLayer(input_shape=x_train.shape[1:]))\n",
        "model.add(tf.keras.layers.Normalization())\n",
        "model.add(layers.Dense(300, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(200, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(100, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(50, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(25, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(1, activation='relu', kernel_regularizer=regularizers.L1(0.01)))\n",
        "\n",
        "#Compile and train\n",
        "model.compile(optimizer=Adam(), \n",
        "              loss='mse',\n",
        "              metrics=['mae',keras.metrics.RootMeanSquaredError()])"
      ],
      "metadata": {
        "id": "63TJR2o__iNE"
      },
      "id": "63TJR2o__iNE",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history = model.fit(x_train, y_train, epochs=100, batch_size=64, validation_data=(x_train,y_train),callbacks=[callback])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vopZwmI5_iZN",
        "outputId": "758f4077-85dd-43af-fc7c-d3c700ff3244"
      },
      "id": "vopZwmI5_iZN",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "755/755 [==============================] - 8s 9ms/step - loss: 49794.2031 - mae: 114.6843 - root_mean_squared_error: 223.0976 - val_loss: 5468.2808 - val_mae: 47.0545 - val_root_mean_squared_error: 73.7964\n",
            "Epoch 2/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 5134.4282 - mae: 44.8189 - root_mean_squared_error: 71.4968 - val_loss: 4860.6118 - val_mae: 43.2666 - val_root_mean_squared_error: 69.5545\n",
            "Epoch 3/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 4301.7080 - mae: 38.1950 - root_mean_squared_error: 65.4138 - val_loss: 3855.5452 - val_mae: 35.5540 - val_root_mean_squared_error: 61.9108\n",
            "Epoch 4/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 3596.1992 - mae: 32.8260 - root_mean_squared_error: 59.7835 - val_loss: 3446.8262 - val_mae: 31.7981 - val_root_mean_squared_error: 58.5238\n",
            "Epoch 5/100\n",
            "755/755 [==============================] - 9s 12ms/step - loss: 3468.3799 - mae: 31.7684 - root_mean_squared_error: 58.7092 - val_loss: 3365.0989 - val_mae: 31.2805 - val_root_mean_squared_error: 57.8235\n",
            "Epoch 6/100\n",
            "755/755 [==============================] - 8s 10ms/step - loss: 3416.8159 - mae: 31.3724 - root_mean_squared_error: 58.2700 - val_loss: 3433.6128 - val_mae: 30.9758 - val_root_mean_squared_error: 58.4153\n",
            "Epoch 7/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 3379.2263 - mae: 31.0963 - root_mean_squared_error: 57.9481 - val_loss: 3310.7197 - val_mae: 30.9089 - val_root_mean_squared_error: 57.3551\n",
            "Epoch 8/100\n",
            "755/755 [==============================] - 8s 11ms/step - loss: 3347.0295 - mae: 30.9394 - root_mean_squared_error: 57.6707 - val_loss: 3294.7549 - val_mae: 30.1598 - val_root_mean_squared_error: 57.2158\n",
            "Epoch 9/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 3343.5315 - mae: 30.9403 - root_mean_squared_error: 57.6411 - val_loss: 3194.8352 - val_mae: 29.0971 - val_root_mean_squared_error: 56.3367\n",
            "Epoch 10/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 3302.6829 - mae: 30.5209 - root_mean_squared_error: 57.2859 - val_loss: 3326.1782 - val_mae: 31.2552 - val_root_mean_squared_error: 57.4908\n",
            "Epoch 11/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 3293.9885 - mae: 30.4103 - root_mean_squared_error: 57.2111 - val_loss: 3389.6375 - val_mae: 32.6790 - val_root_mean_squared_error: 58.0405\n",
            "Epoch 12/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 3288.5962 - mae: 30.7064 - root_mean_squared_error: 57.1633 - val_loss: 3195.2356 - val_mae: 29.6039 - val_root_mean_squared_error: 56.3410\n",
            "Epoch 13/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 3248.1589 - mae: 30.1529 - root_mean_squared_error: 56.8085 - val_loss: 3233.5642 - val_mae: 30.3194 - val_root_mean_squared_error: 56.6804\n",
            "Epoch 14/100\n",
            "755/755 [==============================] - 8s 11ms/step - loss: 3200.8950 - mae: 30.0778 - root_mean_squared_error: 56.3910 - val_loss: 3025.9824 - val_mae: 28.9186 - val_root_mean_squared_error: 54.8178\n",
            "Epoch 15/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 3054.5242 - mae: 28.8705 - root_mean_squared_error: 55.0768 - val_loss: 2971.3120 - val_mae: 27.9071 - val_root_mean_squared_error: 54.3159\n",
            "Epoch 16/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 2965.8828 - mae: 27.7543 - root_mean_squared_error: 54.2665 - val_loss: 2864.6338 - val_mae: 26.7069 - val_root_mean_squared_error: 53.3267\n",
            "Epoch 17/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2949.7554 - mae: 27.6649 - root_mean_squared_error: 54.1180 - val_loss: 2912.6094 - val_mae: 27.5001 - val_root_mean_squared_error: 53.7739\n",
            "Epoch 18/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 2878.9023 - mae: 26.8758 - root_mean_squared_error: 53.4590 - val_loss: 2838.7817 - val_mae: 26.1930 - val_root_mean_squared_error: 53.0819\n",
            "Epoch 19/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2858.6042 - mae: 26.6419 - root_mean_squared_error: 53.2685 - val_loss: 2776.2527 - val_mae: 25.5177 - val_root_mean_squared_error: 52.4885\n",
            "Epoch 20/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2833.6465 - mae: 26.3635 - root_mean_squared_error: 53.0331 - val_loss: 2750.9839 - val_mae: 25.9619 - val_root_mean_squared_error: 52.2480\n",
            "Epoch 21/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2837.5312 - mae: 26.7594 - root_mean_squared_error: 53.0692 - val_loss: 2724.5613 - val_mae: 25.1861 - val_root_mean_squared_error: 51.9942\n",
            "Epoch 22/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2798.6753 - mae: 26.2267 - root_mean_squared_error: 52.7013 - val_loss: 2775.1582 - val_mae: 25.5739 - val_root_mean_squared_error: 52.4790\n",
            "Epoch 23/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2819.2053 - mae: 26.5715 - root_mean_squared_error: 52.8951 - val_loss: 2788.3508 - val_mae: 26.9902 - val_root_mean_squared_error: 52.6008\n",
            "Epoch 24/100\n",
            "755/755 [==============================] - 8s 10ms/step - loss: 2804.3970 - mae: 26.5260 - root_mean_squared_error: 52.7531 - val_loss: 3234.1021 - val_mae: 34.4655 - val_root_mean_squared_error: 56.6798\n",
            "Epoch 25/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 2770.2998 - mae: 26.0067 - root_mean_squared_error: 52.4281 - val_loss: 2794.6370 - val_mae: 24.8494 - val_root_mean_squared_error: 52.6566\n",
            "Epoch 26/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2761.9709 - mae: 26.2146 - root_mean_squared_error: 52.3464 - val_loss: 2742.6108 - val_mae: 26.1733 - val_root_mean_squared_error: 52.1610\n",
            "Epoch 27/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2764.6179 - mae: 26.2576 - root_mean_squared_error: 52.3704 - val_loss: 2631.6978 - val_mae: 24.3268 - val_root_mean_squared_error: 51.0842\n",
            "Epoch 28/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2737.2661 - mae: 25.9532 - root_mean_squared_error: 52.1067 - val_loss: 2945.3469 - val_mae: 31.3225 - val_root_mean_squared_error: 54.0659\n",
            "Epoch 29/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2734.3030 - mae: 26.2210 - root_mean_squared_error: 52.0768 - val_loss: 2834.0762 - val_mae: 29.3144 - val_root_mean_squared_error: 53.0257\n",
            "Epoch 30/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 2708.5051 - mae: 25.9142 - root_mean_squared_error: 51.8266 - val_loss: 2673.3118 - val_mae: 25.9286 - val_root_mean_squared_error: 51.4855\n",
            "Epoch 31/100\n",
            "755/755 [==============================] - 8s 10ms/step - loss: 2716.8333 - mae: 25.9628 - root_mean_squared_error: 51.9055 - val_loss: 2631.7363 - val_mae: 25.4804 - val_root_mean_squared_error: 51.0789\n",
            "Epoch 32/100\n",
            "755/755 [==============================] - 8s 11ms/step - loss: 2715.3459 - mae: 26.3151 - root_mean_squared_error: 51.8894 - val_loss: 2651.4055 - val_mae: 25.8909 - val_root_mean_squared_error: 51.2674\n",
            "Epoch 33/100\n",
            "755/755 [==============================] - 9s 12ms/step - loss: 2685.6704 - mae: 25.9795 - root_mean_squared_error: 51.6002 - val_loss: 2624.4392 - val_mae: 25.5167 - val_root_mean_squared_error: 51.0028\n",
            "Epoch 34/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2666.3496 - mae: 25.7841 - root_mean_squared_error: 51.4107 - val_loss: 2710.2703 - val_mae: 26.7492 - val_root_mean_squared_error: 51.8357\n",
            "Epoch 35/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2660.2979 - mae: 25.7213 - root_mean_squared_error: 51.3507 - val_loss: 2673.1582 - val_mae: 25.8862 - val_root_mean_squared_error: 51.4742\n",
            "Epoch 36/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2662.3284 - mae: 25.7894 - root_mean_squared_error: 51.3687 - val_loss: 2569.2690 - val_mae: 24.2724 - val_root_mean_squared_error: 50.4532\n",
            "Epoch 37/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2650.4155 - mae: 25.7142 - root_mean_squared_error: 51.2510 - val_loss: 2551.8477 - val_mae: 24.5996 - val_root_mean_squared_error: 50.2798\n",
            "Epoch 38/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 2646.1799 - mae: 25.7778 - root_mean_squared_error: 51.2079 - val_loss: 2581.4092 - val_mae: 25.0680 - val_root_mean_squared_error: 50.5711\n",
            "Epoch 39/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2634.1423 - mae: 25.5441 - root_mean_squared_error: 51.0884 - val_loss: 2649.5552 - val_mae: 26.5694 - val_root_mean_squared_error: 51.2381\n",
            "Epoch 40/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2634.2869 - mae: 25.7731 - root_mean_squared_error: 51.0881 - val_loss: 2591.8577 - val_mae: 25.8862 - val_root_mean_squared_error: 50.6703\n",
            "Epoch 41/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2619.4766 - mae: 25.5140 - root_mean_squared_error: 50.9408 - val_loss: 2557.0315 - val_mae: 23.8860 - val_root_mean_squared_error: 50.3227\n",
            "Epoch 42/100\n",
            "755/755 [==============================] - 8s 10ms/step - loss: 2603.6648 - mae: 25.5286 - root_mean_squared_error: 50.7834 - val_loss: 2556.2898 - val_mae: 24.2649 - val_root_mean_squared_error: 50.3132\n",
            "Epoch 43/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 2600.7310 - mae: 25.5202 - root_mean_squared_error: 50.7526 - val_loss: 2506.0300 - val_mae: 24.7481 - val_root_mean_squared_error: 49.8101\n",
            "Epoch 44/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 2581.6829 - mae: 25.4164 - root_mean_squared_error: 50.5626 - val_loss: 2531.8604 - val_mae: 25.2475 - val_root_mean_squared_error: 50.0673\n",
            "Epoch 45/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2589.9641 - mae: 25.4554 - root_mean_squared_error: 50.6425 - val_loss: 2515.6504 - val_mae: 25.1034 - val_root_mean_squared_error: 49.9017\n",
            "Epoch 46/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2576.9038 - mae: 25.3715 - root_mean_squared_error: 50.5110 - val_loss: 2477.1062 - val_mae: 23.9545 - val_root_mean_squared_error: 49.5127\n",
            "Epoch 47/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2580.0425 - mae: 25.4027 - root_mean_squared_error: 50.5405 - val_loss: 2523.8699 - val_mae: 25.3458 - val_root_mean_squared_error: 49.9807\n",
            "Epoch 48/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 2584.3474 - mae: 25.6242 - root_mean_squared_error: 50.5808 - val_loss: 2582.1777 - val_mae: 25.3556 - val_root_mean_squared_error: 50.5573\n",
            "Epoch 49/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2559.3770 - mae: 25.2078 - root_mean_squared_error: 50.3317 - val_loss: 2513.1646 - val_mae: 25.0732 - val_root_mean_squared_error: 49.8704\n",
            "Epoch 50/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2562.2070 - mae: 25.3712 - root_mean_squared_error: 50.3585 - val_loss: 2586.1758 - val_mae: 27.3984 - val_root_mean_squared_error: 50.5955\n",
            "Epoch 51/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 2569.6191 - mae: 25.6534 - root_mean_squared_error: 50.4302 - val_loss: 2481.8145 - val_mae: 24.1447 - val_root_mean_squared_error: 49.5506\n",
            "Epoch 52/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2555.5256 - mae: 25.3213 - root_mean_squared_error: 50.2878 - val_loss: 2482.1089 - val_mae: 25.1982 - val_root_mean_squared_error: 49.5514\n",
            "Epoch 53/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2556.4241 - mae: 25.3645 - root_mean_squared_error: 50.2945 - val_loss: 2433.5349 - val_mae: 24.2892 - val_root_mean_squared_error: 49.0572\n",
            "Epoch 54/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 2532.0601 - mae: 25.1107 - root_mean_squared_error: 50.0495 - val_loss: 2497.0090 - val_mae: 24.7098 - val_root_mean_squared_error: 49.6976\n",
            "Epoch 55/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2526.6614 - mae: 25.1083 - root_mean_squared_error: 49.9928 - val_loss: 2410.0967 - val_mae: 23.7535 - val_root_mean_squared_error: 48.8123\n",
            "Epoch 56/100\n",
            "755/755 [==============================] - 8s 11ms/step - loss: 2538.2268 - mae: 25.3326 - root_mean_squared_error: 50.1066 - val_loss: 2467.1150 - val_mae: 25.3722 - val_root_mean_squared_error: 49.3905\n",
            "Epoch 57/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 2543.0996 - mae: 25.5712 - root_mean_squared_error: 50.1523 - val_loss: 2492.5168 - val_mae: 25.2852 - val_root_mean_squared_error: 49.6448\n",
            "Epoch 58/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2508.3198 - mae: 25.0801 - root_mean_squared_error: 49.8021 - val_loss: 2486.4890 - val_mae: 25.0970 - val_root_mean_squared_error: 49.5822\n",
            "Epoch 59/100\n",
            "755/755 [==============================] - 8s 11ms/step - loss: 2503.9719 - mae: 25.1294 - root_mean_squared_error: 49.7564 - val_loss: 2388.9827 - val_mae: 23.3695 - val_root_mean_squared_error: 48.5862\n",
            "Epoch 60/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2518.4529 - mae: 25.3365 - root_mean_squared_error: 49.8997 - val_loss: 2461.5466 - val_mae: 25.2611 - val_root_mean_squared_error: 49.3262\n",
            "Epoch 61/100\n",
            "755/755 [==============================] - 9s 12ms/step - loss: 2502.3889 - mae: 25.1199 - root_mean_squared_error: 49.7366 - val_loss: 2419.7444 - val_mae: 24.9099 - val_root_mean_squared_error: 48.8978\n",
            "Epoch 62/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2482.6633 - mae: 25.0032 - root_mean_squared_error: 49.5354 - val_loss: 2449.0317 - val_mae: 24.2901 - val_root_mean_squared_error: 49.1935\n",
            "Epoch 63/100\n",
            "755/755 [==============================] - 8s 10ms/step - loss: 2493.5386 - mae: 25.0943 - root_mean_squared_error: 49.6431 - val_loss: 2720.4399 - val_mae: 30.2003 - val_root_mean_squared_error: 51.8777\n",
            "Epoch 64/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2486.6140 - mae: 25.1558 - root_mean_squared_error: 49.5715 - val_loss: 2422.0767 - val_mae: 24.7080 - val_root_mean_squared_error: 48.9156\n",
            "Epoch 65/100\n",
            "755/755 [==============================] - 8s 10ms/step - loss: 2483.1174 - mae: 25.0253 - root_mean_squared_error: 49.5342 - val_loss: 2427.3850 - val_mae: 24.5116 - val_root_mean_squared_error: 48.9677\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YJl0NTRQlOf4",
        "outputId": "c756e7f6-31df-43c6-f5e0-0291d7c44248"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " normalization_1 (Normalizat  (None, 12)               25        \n",
            " ion)                                                            \n",
            "                                                                 \n",
            " dense_6 (Dense)             (None, 300)               3900      \n",
            "                                                                 \n",
            " dense_7 (Dense)             (None, 200)               60200     \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 100)               20100     \n",
            "                                                                 \n",
            " dense_9 (Dense)             (None, 50)                5050      \n",
            "                                                                 \n",
            " dense_10 (Dense)            (None, 25)                1275      \n",
            "                                                                 \n",
            " dense_11 (Dense)            (None, 1)                 26        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 90,576\n",
            "Trainable params: 90,551\n",
            "Non-trainable params: 25\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "model.summary()"
      ],
      "id": "YJl0NTRQlOf4"
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Save model history into Json"
      ],
      "metadata": {
        "id": "aD4L_SI2DMK9"
      },
      "id": "aD4L_SI2DMK9"
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "# convert the history.history dict to a pandas DataFrame:     \n",
        "hist_df = pd.DataFrame(history.history) \n",
        "\n",
        "# or save to csv: \n",
        "hist_csv_file = 'history_epochs100_layers7_L2_reg_selectedfeatures_12.csv'\n",
        "with open(hist_csv_file, mode='w') as f:\n",
        "    hist_df.to_csv(f)\n",
        "\n",
        "# save to json:  \n",
        "hist_json_file = 'history_epochs100_layers7_L2_reg_selectedfeatures_12.json' \n",
        "with open(hist_json_file, mode='w') as f:\n",
        "    hist_df.to_json(f)"
      ],
      "metadata": {
        "id": "dEanSJlsDMxp"
      },
      "id": "dEanSJlsDMxp",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DxEOZwIUHhAx",
        "outputId": "d0696957-55f3-4cb2-ba17-147ed029241e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "924/924 [==============================] - 2s 2ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[1.7789787e-01],\n",
              "       [6.8605609e+02],\n",
              "       [1.7623480e-01],\n",
              "       ...,\n",
              "       [4.2262793e+02],\n",
              "       [2.3071266e+02],\n",
              "       [1.7800853e-01]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 48
        }
      ],
      "source": [
        "y_pred=model.predict(x_test)\n",
        "y_pred"
      ],
      "id": "DxEOZwIUHhAx"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "yzhePVv3Hg-Z",
        "outputId": "18a4582a-c63d-4117-dd39-c406f343be8e"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "plt.plot(history.history['mae'])\n",
        "plt.plot(history.history['root_mean_squared_error'])\n",
        "plt.title('MAE and RMSE per epoch')\n",
        "plt.ylabel('Error')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['MAE', 'RMSE'], loc='upper right')\n",
        "plt.show()"
      ],
      "id": "yzhePVv3Hg-Z"
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "00yH45-CBUx7",
        "outputId": "e6aea2b4-3754-462c-b622-5ec1c3dd411f"
      },
      "id": "00yH45-CBUx7",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import r2_score \n",
        "from sklearn import metrics\n",
        "R_square = r2_score(y_test, y_pred) \n",
        "print('Coefficient of Determination R2', R_square)\n",
        "print(\"Model MAE Selected Features 7 Layers\", metrics.mean_absolute_error(y_test, y_pred))\n",
        "print(\"Model MSE Selected Features 7 Layers\", metrics.mean_squared_error(y_test, y_pred))\n",
        "print(\"Model RMSE Selected Features 7 Layers\",np.sqrt(metrics.mean_squared_error(y_test, y_pred))) "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8hhXUYKKBIij",
        "outputId": "bee64cf7-cfbd-4e9b-bbc2-f9d7b59be2c0"
      },
      "id": "8hhXUYKKBIij",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Coefficient of Determination R2 0.9736381981154622\n",
            "Model MAE Selected Features 7 Layers 25.81504638758021\n",
            "Model MSE Selected Features 7 Layers 2743.939264811825\n",
            "Model RMSE Selected Features 7 Layers 52.38262369156231\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Train Model with all features and 6 Layers"
      ],
      "metadata": {
        "id": "oPq2HLl4CyZn"
      },
      "id": "oPq2HLl4CyZn"
    },
    {
      "cell_type": "code",
      "source": [
        "nsrdb= pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/all_cities_transformed_combined.csv')\n",
        "nsrdb = nsrdb.loc[(nsrdb['Hour'].isin([7, 10, 13, 16, 19])) & (nsrdb['Minute'] == 0)]\n",
        "print(nsrdb.columns)\n",
        "X = nsrdb.drop(['Unnamed: 0','GHI'], axis =1)\n",
        "print(X.columns)\n",
        "y=nsrdb['GHI']\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_normalized = sc.fit_transform(X)\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "x_train, x_test, y_train, y_test = train_test_split(input_normalized,y, test_size=0.3, random_state=42)\n",
        "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.3, random_state=42)\n",
        "\n",
        "print(\"x_train shape\",x_train.shape)\n",
        "print(\"x_test shape\",x_test.shape)\n",
        "print(\"y_train shape\",y_train.shape)\n",
        "print(\"y_text shape\",y_test.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "emU_aetnC74L",
        "outputId": "58902b27-b7d0-4a52-efaf-5b72fb2cb020"
      },
      "id": "emU_aetnC74L",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Index(['Unnamed: 0', 'Year', 'Month', 'Day', 'Hour', 'Minute', 'Temperature',\n",
            "       'Cloud_Type', 'Dew_Point', 'GHI', 'Relative_Humidity',\n",
            "       'Solar_Zenith_Angle', 'Surface_Albedo', 'Pressure',\n",
            "       'Precipitable_Water', 'Wind_Direction', 'Wind_Speed', 'Location'],\n",
            "      dtype='object')\n",
            "Index(['Year', 'Month', 'Day', 'Hour', 'Minute', 'Temperature', 'Cloud_Type',\n",
            "       'Dew_Point', 'Relative_Humidity', 'Solar_Zenith_Angle',\n",
            "       'Surface_Albedo', 'Pressure', 'Precipitable_Water', 'Wind_Direction',\n",
            "       'Wind_Speed', 'Location'],\n",
            "      dtype='object')\n",
            "x_train shape (48289, 16)\n",
            "x_test shape (29565, 16)\n",
            "y_train shape (48289,)\n",
            "y_text shape (29565,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=3)\n",
        "model = keras.Sequential()\n",
        "model.add(layers.InputLayer(input_shape=x_train.shape[1:]))\n",
        "model.add(tf.keras.layers.Normalization())\n",
        "#model.add(layers.Dense(300, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(200, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(100, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(50, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(25, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(1, activation='relu', kernel_regularizer=regularizers.L1(0.01)))\n",
        "\n",
        "#Compile and train\n",
        "model.compile(optimizer=Adam(), \n",
        "              loss='mse',\n",
        "              metrics=['mae',keras.metrics.RootMeanSquaredError()])"
      ],
      "metadata": {
        "id": "4SS3TyU8C77q"
      },
      "id": "4SS3TyU8C77q",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history = model.fit(x_train, y_train, epochs=100, batch_size=64, validation_data=(x_train,y_train),callbacks=[callback])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iyiWwzeCC7-s",
        "outputId": "58305080-2805-4ce2-ff20-388fc6db738b"
      },
      "id": "iyiWwzeCC7-s",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "755/755 [==============================] - 5s 6ms/step - loss: 22022.1328 - mae: 77.8051 - root_mean_squared_error: 148.3035 - val_loss: 3923.0000 - val_mae: 35.0414 - val_root_mean_squared_error: 62.4001\n",
            "Epoch 2/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 3616.2117 - mae: 33.2420 - root_mean_squared_error: 59.8910 - val_loss: 3221.1489 - val_mae: 30.3645 - val_root_mean_squared_error: 56.4962\n",
            "Epoch 3/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 3097.3333 - mae: 29.9501 - root_mean_squared_error: 55.3873 - val_loss: 2968.5850 - val_mae: 28.8828 - val_root_mean_squared_error: 54.2108\n",
            "Epoch 4/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2759.8052 - mae: 27.1137 - root_mean_squared_error: 52.2488 - val_loss: 2663.7498 - val_mae: 27.1233 - val_root_mean_squared_error: 51.3208\n",
            "Epoch 5/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2643.1396 - mae: 26.1985 - root_mean_squared_error: 51.1195 - val_loss: 2486.3716 - val_mae: 24.0124 - val_root_mean_squared_error: 49.5620\n",
            "Epoch 6/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2556.8997 - mae: 25.5567 - root_mean_squared_error: 50.2686 - val_loss: 2504.3657 - val_mae: 25.3464 - val_root_mean_squared_error: 49.7431\n",
            "Epoch 7/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2508.1294 - mae: 25.0338 - root_mean_squared_error: 49.7808 - val_loss: 2552.9124 - val_mae: 27.0140 - val_root_mean_squared_error: 50.2285\n",
            "Epoch 8/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2479.6562 - mae: 24.8654 - root_mean_squared_error: 49.4932 - val_loss: 2385.6570 - val_mae: 24.3191 - val_root_mean_squared_error: 48.5338\n",
            "Epoch 9/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2442.2783 - mae: 24.7250 - root_mean_squared_error: 49.1131 - val_loss: 2335.9055 - val_mae: 23.1101 - val_root_mean_squared_error: 48.0171\n",
            "Epoch 10/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2399.8049 - mae: 24.2725 - root_mean_squared_error: 48.6775 - val_loss: 2416.8506 - val_mae: 25.9603 - val_root_mean_squared_error: 48.8516\n",
            "Epoch 11/100\n",
            "755/755 [==============================] - 5s 6ms/step - loss: 2371.1143 - mae: 24.0536 - root_mean_squared_error: 48.3803 - val_loss: 2349.8796 - val_mae: 23.5950 - val_root_mean_squared_error: 48.1595\n",
            "Epoch 12/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 2350.8076 - mae: 23.8380 - root_mean_squared_error: 48.1683 - val_loss: 2236.9253 - val_mae: 21.7587 - val_root_mean_squared_error: 46.9697\n",
            "Epoch 13/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2334.5818 - mae: 23.7051 - root_mean_squared_error: 47.9977 - val_loss: 2205.5837 - val_mae: 22.3793 - val_root_mean_squared_error: 46.6336\n",
            "Epoch 14/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2296.1206 - mae: 23.7682 - root_mean_squared_error: 47.5937 - val_loss: 2205.1628 - val_mae: 22.2830 - val_root_mean_squared_error: 46.6276\n",
            "Epoch 15/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2282.2754 - mae: 23.6240 - root_mean_squared_error: 47.4462 - val_loss: 2172.3010 - val_mae: 22.1874 - val_root_mean_squared_error: 46.2722\n",
            "Epoch 16/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 2250.8599 - mae: 23.3446 - root_mean_squared_error: 47.1124 - val_loss: 2317.1323 - val_mae: 23.2019 - val_root_mean_squared_error: 47.8092\n",
            "Epoch 17/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2247.7517 - mae: 23.2461 - root_mean_squared_error: 47.0776 - val_loss: 2167.4900 - val_mae: 22.7907 - val_root_mean_squared_error: 46.2162\n",
            "Epoch 18/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2210.1326 - mae: 23.1583 - root_mean_squared_error: 46.6743 - val_loss: 2215.7019 - val_mae: 23.8564 - val_root_mean_squared_error: 46.7335\n",
            "Epoch 19/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2182.1541 - mae: 22.9574 - root_mean_squared_error: 46.3717 - val_loss: 2245.6382 - val_mae: 23.2628 - val_root_mean_squared_error: 47.0503\n",
            "Epoch 20/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2181.9448 - mae: 23.1409 - root_mean_squared_error: 46.3678 - val_loss: 2101.0410 - val_mae: 23.0728 - val_root_mean_squared_error: 45.4862\n",
            "Epoch 21/100\n",
            "755/755 [==============================] - 3s 5ms/step - loss: 2169.5083 - mae: 23.1284 - root_mean_squared_error: 46.2318 - val_loss: 2036.4354 - val_mae: 21.8925 - val_root_mean_squared_error: 44.7689\n",
            "Epoch 22/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2140.5710 - mae: 22.8268 - root_mean_squared_error: 45.9162 - val_loss: 2073.9146 - val_mae: 22.2372 - val_root_mean_squared_error: 45.1833\n",
            "Epoch 23/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2119.3376 - mae: 22.6517 - root_mean_squared_error: 45.6824 - val_loss: 1993.3501 - val_mae: 21.6614 - val_root_mean_squared_error: 44.2810\n",
            "Epoch 24/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2082.9963 - mae: 22.5395 - root_mean_squared_error: 45.2810 - val_loss: 2004.1885 - val_mae: 20.9789 - val_root_mean_squared_error: 44.4007\n",
            "Epoch 25/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2072.7441 - mae: 22.4888 - root_mean_squared_error: 45.1657 - val_loss: 2060.4985 - val_mae: 22.0591 - val_root_mean_squared_error: 45.0291\n",
            "Epoch 26/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2068.7012 - mae: 22.4081 - root_mean_squared_error: 45.1196 - val_loss: 1994.9742 - val_mae: 22.5967 - val_root_mean_squared_error: 44.2940\n",
            "Epoch 27/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 2034.6232 - mae: 22.4144 - root_mean_squared_error: 44.7384 - val_loss: 1993.7776 - val_mae: 22.1528 - val_root_mean_squared_error: 44.2784\n",
            "Epoch 28/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 2034.0938 - mae: 22.6607 - root_mean_squared_error: 44.7304 - val_loss: 1952.1099 - val_mae: 21.4058 - val_root_mean_squared_error: 43.8040\n",
            "Epoch 29/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 2009.0760 - mae: 22.3063 - root_mean_squared_error: 44.4487 - val_loss: 1920.7168 - val_mae: 21.2980 - val_root_mean_squared_error: 43.4425\n",
            "Epoch 30/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1992.9797 - mae: 22.1314 - root_mean_squared_error: 44.2653 - val_loss: 2013.6687 - val_mae: 22.3614 - val_root_mean_squared_error: 44.4981\n",
            "Epoch 31/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1980.8657 - mae: 22.1579 - root_mean_squared_error: 44.1268 - val_loss: 1818.6935 - val_mae: 20.4289 - val_root_mean_squared_error: 42.2484\n",
            "Epoch 32/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1956.1792 - mae: 22.0838 - root_mean_squared_error: 43.8445 - val_loss: 1832.4248 - val_mae: 21.4152 - val_root_mean_squared_error: 42.4090\n",
            "Epoch 33/100\n",
            "755/755 [==============================] - 5s 6ms/step - loss: 1944.2189 - mae: 21.9359 - root_mean_squared_error: 43.7058 - val_loss: 1846.5886 - val_mae: 22.6225 - val_root_mean_squared_error: 42.5736\n",
            "Epoch 34/100\n",
            "755/755 [==============================] - 6s 7ms/step - loss: 1918.0680 - mae: 21.6602 - root_mean_squared_error: 43.4040 - val_loss: 1799.2422 - val_mae: 20.3663 - val_root_mean_squared_error: 42.0120\n",
            "Epoch 35/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1924.7471 - mae: 21.9938 - root_mean_squared_error: 43.4792 - val_loss: 1782.0040 - val_mae: 20.4385 - val_root_mean_squared_error: 41.8047\n",
            "Epoch 36/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1880.0403 - mae: 21.3230 - root_mean_squared_error: 42.9600 - val_loss: 1787.0647 - val_mae: 21.5303 - val_root_mean_squared_error: 41.8633\n",
            "Epoch 37/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1871.7527 - mae: 21.6953 - root_mean_squared_error: 42.8614 - val_loss: 1793.8800 - val_mae: 22.1900 - val_root_mean_squared_error: 41.9424\n",
            "Epoch 38/100\n",
            "755/755 [==============================] - 4s 6ms/step - loss: 1859.1211 - mae: 21.5561 - root_mean_squared_error: 42.7116 - val_loss: 1765.5936 - val_mae: 21.0275 - val_root_mean_squared_error: 41.6013\n",
            "Epoch 39/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1841.8878 - mae: 21.3677 - root_mean_squared_error: 42.5073 - val_loss: 1820.4678 - val_mae: 22.4490 - val_root_mean_squared_error: 42.2532\n",
            "Epoch 40/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1836.2056 - mae: 21.4603 - root_mean_squared_error: 42.4378 - val_loss: 1685.7152 - val_mae: 20.0507 - val_root_mean_squared_error: 40.6250\n",
            "Epoch 41/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1814.2008 - mae: 21.2610 - root_mean_squared_error: 42.1754 - val_loss: 1736.5015 - val_mae: 21.1885 - val_root_mean_squared_error: 41.2431\n",
            "Epoch 42/100\n",
            "755/755 [==============================] - 4s 6ms/step - loss: 1818.6143 - mae: 21.4356 - root_mean_squared_error: 42.2252 - val_loss: 1685.8004 - val_mae: 20.1948 - val_root_mean_squared_error: 40.6208\n",
            "Epoch 43/100\n",
            "755/755 [==============================] - 7s 10ms/step - loss: 1785.4551 - mae: 21.1361 - root_mean_squared_error: 41.8285 - val_loss: 1787.3666 - val_mae: 22.8611 - val_root_mean_squared_error: 41.8505\n",
            "Epoch 44/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 1770.2261 - mae: 21.1034 - root_mean_squared_error: 41.6441 - val_loss: 1722.0471 - val_mae: 20.9947 - val_root_mean_squared_error: 41.0605\n",
            "Epoch 45/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1781.8342 - mae: 21.2790 - root_mean_squared_error: 41.7806 - val_loss: 1632.3418 - val_mae: 19.4429 - val_root_mean_squared_error: 39.9501\n",
            "Epoch 46/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1744.7012 - mae: 20.8449 - root_mean_squared_error: 41.3309 - val_loss: 1667.0994 - val_mae: 21.4053 - val_root_mean_squared_error: 40.3807\n",
            "Epoch 47/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1761.0376 - mae: 21.1454 - root_mean_squared_error: 41.5259 - val_loss: 1694.3119 - val_mae: 21.8306 - val_root_mean_squared_error: 40.7131\n",
            "Epoch 48/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1737.4634 - mae: 20.9147 - root_mean_squared_error: 41.2384 - val_loss: 1679.5815 - val_mae: 21.6411 - val_root_mean_squared_error: 40.5295\n",
            "Epoch 49/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1698.8954 - mae: 20.5572 - root_mean_squared_error: 40.7656 - val_loss: 1647.6608 - val_mae: 21.2968 - val_root_mean_squared_error: 40.1307\n",
            "Epoch 50/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1691.8120 - mae: 20.4292 - root_mean_squared_error: 40.6756 - val_loss: 1557.7708 - val_mae: 19.7229 - val_root_mean_squared_error: 38.9916\n",
            "Epoch 51/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1691.6313 - mae: 20.5945 - root_mean_squared_error: 40.6704 - val_loss: 1621.9952 - val_mae: 21.0670 - val_root_mean_squared_error: 39.8038\n",
            "Epoch 52/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 1670.5389 - mae: 20.6820 - root_mean_squared_error: 40.4074 - val_loss: 1594.2236 - val_mae: 20.1895 - val_root_mean_squared_error: 39.4502\n",
            "Epoch 53/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1648.5410 - mae: 20.3301 - root_mean_squared_error: 40.1309 - val_loss: 1514.2196 - val_mae: 19.8092 - val_root_mean_squared_error: 38.4194\n",
            "Epoch 54/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 1631.2505 - mae: 20.2428 - root_mean_squared_error: 39.9115 - val_loss: 1557.4224 - val_mae: 19.5084 - val_root_mean_squared_error: 38.9741\n",
            "Epoch 55/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 1621.6727 - mae: 20.1346 - root_mean_squared_error: 39.7880 - val_loss: 1588.5564 - val_mae: 21.1761 - val_root_mean_squared_error: 39.3687\n",
            "Epoch 56/100\n",
            "755/755 [==============================] - 4s 6ms/step - loss: 1632.0087 - mae: 20.5569 - root_mean_squared_error: 39.9146 - val_loss: 1499.2739 - val_mae: 19.8163 - val_root_mean_squared_error: 38.2144\n",
            "Epoch 57/100\n",
            "755/755 [==============================] - 10s 13ms/step - loss: 1585.7614 - mae: 20.0575 - root_mean_squared_error: 39.3279 - val_loss: 1472.4528 - val_mae: 18.7924 - val_root_mean_squared_error: 37.8583\n",
            "Epoch 58/100\n",
            "755/755 [==============================] - 8s 10ms/step - loss: 1583.0383 - mae: 20.0063 - root_mean_squared_error: 39.2902 - val_loss: 1868.3651 - val_mae: 25.5850 - val_root_mean_squared_error: 42.7658\n",
            "Epoch 59/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1585.1509 - mae: 20.0475 - root_mean_squared_error: 39.3137 - val_loss: 1401.0895 - val_mae: 18.0442 - val_root_mean_squared_error: 36.8966\n",
            "Epoch 60/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1554.4760 - mae: 19.8143 - root_mean_squared_error: 38.9182 - val_loss: 1588.7480 - val_mae: 19.7107 - val_root_mean_squared_error: 39.3549\n",
            "Epoch 61/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1553.1257 - mae: 19.9784 - root_mean_squared_error: 38.8977 - val_loss: 1554.6853 - val_mae: 19.3351 - val_root_mean_squared_error: 38.9166\n",
            "Epoch 62/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1544.5426 - mae: 19.7986 - root_mean_squared_error: 38.7840 - val_loss: 1414.7739 - val_mae: 18.0882 - val_root_mean_squared_error: 37.0720\n",
            "Epoch 63/100\n",
            "755/755 [==============================] - 4s 6ms/step - loss: 1512.2853 - mae: 19.6115 - root_mean_squared_error: 38.3626 - val_loss: 1438.6637 - val_mae: 19.5700 - val_root_mean_squared_error: 37.3893\n",
            "Epoch 64/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1528.1143 - mae: 19.7173 - root_mean_squared_error: 38.5651 - val_loss: 1471.3551 - val_mae: 19.0788 - val_root_mean_squared_error: 37.8204\n",
            "Epoch 65/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1470.3115 - mae: 19.3558 - root_mean_squared_error: 37.8047 - val_loss: 1344.9631 - val_mae: 17.6411 - val_root_mean_squared_error: 36.1069\n",
            "Epoch 66/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1476.1532 - mae: 19.2784 - root_mean_squared_error: 37.8785 - val_loss: 1621.0560 - val_mae: 20.7931 - val_root_mean_squared_error: 39.7438\n",
            "Epoch 67/100\n",
            "755/755 [==============================] - 5s 6ms/step - loss: 1485.7402 - mae: 19.4513 - root_mean_squared_error: 38.0015 - val_loss: 1450.6129 - val_mae: 18.9385 - val_root_mean_squared_error: 37.5347\n",
            "Epoch 68/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 1465.1187 - mae: 19.4907 - root_mean_squared_error: 37.7263 - val_loss: 1335.7678 - val_mae: 18.8334 - val_root_mean_squared_error: 35.9698\n",
            "Epoch 69/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1466.3741 - mae: 19.4065 - root_mean_squared_error: 37.7400 - val_loss: 1346.5162 - val_mae: 18.0554 - val_root_mean_squared_error: 36.1149\n",
            "Epoch 70/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 1431.1169 - mae: 19.1567 - root_mean_squared_error: 37.2663 - val_loss: 1353.8536 - val_mae: 18.4550 - val_root_mean_squared_error: 36.2131\n",
            "Epoch 71/100\n",
            "755/755 [==============================] - 9s 11ms/step - loss: 1417.1953 - mae: 19.0639 - root_mean_squared_error: 37.0760 - val_loss: 1298.2959 - val_mae: 17.9607 - val_root_mean_squared_error: 35.4344\n",
            "Epoch 72/100\n",
            "755/755 [==============================] - 8s 11ms/step - loss: 1402.3929 - mae: 18.8604 - root_mean_squared_error: 36.8724 - val_loss: 1281.0704 - val_mae: 17.9690 - val_root_mean_squared_error: 35.1872\n",
            "Epoch 73/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1401.1786 - mae: 19.0160 - root_mean_squared_error: 36.8523 - val_loss: 1261.2054 - val_mae: 17.5201 - val_root_mean_squared_error: 34.8999\n",
            "Epoch 74/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1401.6550 - mae: 18.9736 - root_mean_squared_error: 36.8551 - val_loss: 1330.5261 - val_mae: 19.5437 - val_root_mean_squared_error: 35.8753\n",
            "Epoch 75/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1387.5037 - mae: 18.9682 - root_mean_squared_error: 36.6594 - val_loss: 1351.1763 - val_mae: 20.1699 - val_root_mean_squared_error: 36.1590\n",
            "Epoch 76/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1358.8076 - mae: 18.6572 - root_mean_squared_error: 36.2624 - val_loss: 1299.3223 - val_mae: 17.9986 - val_root_mean_squared_error: 35.4310\n",
            "Epoch 77/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 1367.9312 - mae: 18.9093 - root_mean_squared_error: 36.3844 - val_loss: 1278.9852 - val_mae: 19.8505 - val_root_mean_squared_error: 35.1387\n",
            "Epoch 78/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 1346.0464 - mae: 18.6945 - root_mean_squared_error: 36.0785 - val_loss: 1247.3252 - val_mae: 18.3783 - val_root_mean_squared_error: 34.6814\n",
            "Epoch 79/100\n",
            "755/755 [==============================] - 4s 6ms/step - loss: 1333.0955 - mae: 18.5507 - root_mean_squared_error: 35.8946 - val_loss: 1328.8715 - val_mae: 18.0836 - val_root_mean_squared_error: 35.8339\n",
            "Epoch 80/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1323.8341 - mae: 18.4261 - root_mean_squared_error: 35.7622 - val_loss: 1188.7377 - val_mae: 17.6684 - val_root_mean_squared_error: 33.8190\n",
            "Epoch 81/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1320.4196 - mae: 18.5110 - root_mean_squared_error: 35.7109 - val_loss: 1161.3857 - val_mae: 17.1988 - val_root_mean_squared_error: 33.4083\n",
            "Epoch 82/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1299.7426 - mae: 18.3335 - root_mean_squared_error: 35.4168 - val_loss: 1173.5315 - val_mae: 17.5204 - val_root_mean_squared_error: 33.5863\n",
            "Epoch 83/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 1283.5729 - mae: 18.3180 - root_mean_squared_error: 35.1841 - val_loss: 1207.7239 - val_mae: 17.4625 - val_root_mean_squared_error: 34.0880\n",
            "Epoch 84/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1269.2583 - mae: 17.9898 - root_mean_squared_error: 34.9768 - val_loss: 1179.2983 - val_mae: 17.9212 - val_root_mean_squared_error: 33.6637\n",
            "Epoch 85/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1262.0720 - mae: 18.1226 - root_mean_squared_error: 34.8698 - val_loss: 1211.3918 - val_mae: 17.2130 - val_root_mean_squared_error: 34.1339\n",
            "Epoch 86/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 1254.4855 - mae: 17.9954 - root_mean_squared_error: 34.7579 - val_loss: 1138.5795 - val_mae: 16.7987 - val_root_mean_squared_error: 33.0467\n",
            "Epoch 87/100\n",
            "755/755 [==============================] - 6s 8ms/step - loss: 1219.6223 - mae: 17.8092 - root_mean_squared_error: 34.2486 - val_loss: 1244.0229 - val_mae: 17.7519 - val_root_mean_squared_error: 34.6004\n",
            "Epoch 88/100\n",
            "755/755 [==============================] - 5s 7ms/step - loss: 1230.0397 - mae: 17.9016 - root_mean_squared_error: 34.3960 - val_loss: 1076.2628 - val_mae: 16.2572 - val_root_mean_squared_error: 32.0811\n",
            "Epoch 89/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 1207.3104 - mae: 17.7110 - root_mean_squared_error: 34.0604 - val_loss: 1168.5532 - val_mae: 18.2186 - val_root_mean_squared_error: 33.4850\n",
            "Epoch 90/100\n",
            "755/755 [==============================] - 4s 6ms/step - loss: 1211.2941 - mae: 17.8221 - root_mean_squared_error: 34.1149 - val_loss: 1261.2357 - val_mae: 17.9557 - val_root_mean_squared_error: 34.8372\n",
            "Epoch 91/100\n",
            "755/755 [==============================] - 5s 6ms/step - loss: 1211.8822 - mae: 17.8920 - root_mean_squared_error: 34.1197 - val_loss: 1188.3527 - val_mae: 19.9708 - val_root_mean_squared_error: 33.7711\n",
            "Epoch 92/100\n",
            "755/755 [==============================] - 4s 6ms/step - loss: 1188.9752 - mae: 17.6227 - root_mean_squared_error: 33.7785 - val_loss: 1091.7360 - val_mae: 17.7599 - val_root_mean_squared_error: 32.3048\n",
            "Epoch 93/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1157.5823 - mae: 17.3926 - root_mean_squared_error: 33.3062 - val_loss: 1095.0836 - val_mae: 16.6338 - val_root_mean_squared_error: 32.3522\n",
            "Epoch 94/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1159.4113 - mae: 17.3736 - root_mean_squared_error: 33.3295 - val_loss: 1040.9679 - val_mae: 16.4850 - val_root_mean_squared_error: 31.5008\n",
            "Epoch 95/100\n",
            "755/755 [==============================] - 4s 6ms/step - loss: 1166.3977 - mae: 17.6717 - root_mean_squared_error: 33.4302 - val_loss: 1019.6111 - val_mae: 15.9022 - val_root_mean_squared_error: 31.1559\n",
            "Epoch 96/100\n",
            "755/755 [==============================] - 5s 6ms/step - loss: 1141.4087 - mae: 17.2929 - root_mean_squared_error: 33.0503 - val_loss: 1034.2616 - val_mae: 16.1353 - val_root_mean_squared_error: 31.3853\n",
            "Epoch 97/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 1121.4508 - mae: 17.0471 - root_mean_squared_error: 32.7434 - val_loss: 1096.5732 - val_mae: 16.7035 - val_root_mean_squared_error: 32.3589\n",
            "Epoch 98/100\n",
            "755/755 [==============================] - 7s 9ms/step - loss: 1110.4376 - mae: 16.9855 - root_mean_squared_error: 32.5703 - val_loss: 1041.8781 - val_mae: 16.0016 - val_root_mean_squared_error: 31.4985\n",
            "Epoch 99/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1108.8771 - mae: 17.0409 - root_mean_squared_error: 32.5421 - val_loss: 993.3206 - val_mae: 15.7773 - val_root_mean_squared_error: 30.7131\n",
            "Epoch 100/100\n",
            "755/755 [==============================] - 4s 5ms/step - loss: 1099.5945 - mae: 16.9984 - root_mean_squared_error: 32.3948 - val_loss: 1090.8765 - val_mae: 16.8508 - val_root_mean_squared_error: 32.2579\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ADH9N-WtC8Bg",
        "outputId": "94bf83c5-4a59-47de-da79-6b9440d24063"
      },
      "id": "ADH9N-WtC8Bg",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " normalization_2 (Normalizat  (None, 16)               33        \n",
            " ion)                                                            \n",
            "                                                                 \n",
            " dense_12 (Dense)            (None, 200)               3400      \n",
            "                                                                 \n",
            " dense_13 (Dense)            (None, 100)               20100     \n",
            "                                                                 \n",
            " dense_14 (Dense)            (None, 50)                5050      \n",
            "                                                                 \n",
            " dense_15 (Dense)            (None, 25)                1275      \n",
            "                                                                 \n",
            " dense_16 (Dense)            (None, 1)                 26        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 29,884\n",
            "Trainable params: 29,851\n",
            "Non-trainable params: 33\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Save model history into Json"
      ],
      "metadata": {
        "id": "vqdJBW4xEp8W"
      },
      "id": "vqdJBW4xEp8W"
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "# convert the history.history dict to a pandas DataFrame:     \n",
        "hist_df = pd.DataFrame(history.history) \n",
        "\n",
        "# or save to csv: \n",
        "hist_csv_file = 'history_epochs100_layers6_L2_reg_AllFeatures.csv'\n",
        "with open(hist_csv_file, mode='w') as f:\n",
        "    hist_df.to_csv(f)\n",
        "\n",
        "# save to json:  \n",
        "hist_json_file = 'history_epochs100_layers6_L2_reg_AllFeatures_12.json' \n",
        "with open(hist_json_file, mode='w') as f:\n",
        "    hist_df.to_json(f)"
      ],
      "metadata": {
        "id": "NLdItJMXC8EG"
      },
      "id": "NLdItJMXC8EG",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred=model.predict(x_test)\n",
        "y_pred"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gG8CSSkuC8HZ",
        "outputId": "acbc47e8-d865-4ca7-d581-9852fd2fb748"
      },
      "id": "gG8CSSkuC8HZ",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "924/924 [==============================] - 2s 2ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[  0.     ],\n",
              "       [711.9729 ],\n",
              "       [  0.     ],\n",
              "       ...,\n",
              "       [422.84656],\n",
              "       [241.76868],\n",
              "       [  0.     ]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['mae'])\n",
        "plt.plot(history.history['root_mean_squared_error'])\n",
        "plt.title('MAE and RMSE per epoch')\n",
        "plt.ylabel('Error')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['MAE', 'RMSE'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "Ns9Me_KVC8Kb",
        "outputId": "199a2a3b-c693-47d4-cf1d-e78dd48f9493"
      },
      "id": "Ns9Me_KVC8Kb",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "lhV9CeL6EmF_",
        "outputId": "39e843bc-6809-4aff-c497-ccdc995b6500"
      },
      "id": "lhV9CeL6EmF_",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import r2_score \n",
        "from sklearn import metrics\n",
        "R_square = r2_score(y_test, y_pred) \n",
        "print('Coefficient of Determination R2', R_square)\n",
        "print(\"Model MAE All Features 6 Layers\", metrics.mean_absolute_error(y_test, y_pred))\n",
        "print(\"Model MSE All Features 6 Layers\", metrics.mean_squared_error(y_test, y_pred))\n",
        "print(\"Model RMSE All Features 6 Layers\",np.sqrt(metrics.mean_squared_error(y_test, y_pred))) "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jIM_Vr3WEmJP",
        "outputId": "9c2ba396-8f1e-42b7-c87f-01c13309f2fd"
      },
      "id": "jIM_Vr3WEmJP",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Coefficient of Determination R2 0.9757875426018217\n",
            "Model MAE All Features 6 Layers 22.652943466150397\n",
            "Model MSE All Features 6 Layers 2520.218945709209\n",
            "Model RMSE All Features 6 Layers 50.20178229614173\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Train Model with selected features and 8 Layers less neurons in hidden layers, batch size = 32, epochs=200"
      ],
      "metadata": {
        "id": "qfOc00mRFaZ6"
      },
      "id": "qfOc00mRFaZ6"
    },
    {
      "cell_type": "code",
      "source": [
        "nsrdb= pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/all_cities_transformed_combined.csv')\n",
        "nsrdb = nsrdb.loc[(nsrdb['Hour'].isin([7, 10, 13, 16, 19])) & (nsrdb['Minute'] == 0)]\n",
        "print(nsrdb.columns)\n",
        "X = nsrdb.drop(['Unnamed: 0','GHI','Surface_Albedo','Precipitable_Water','Year','Pressure','Minute'], axis =1)\n",
        "print(X.columns)\n",
        "y=nsrdb['GHI']\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_normalized = sc.fit_transform(X)\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "x_train, x_test, y_train, y_test = train_test_split(input_normalized,y, test_size=0.3, random_state=42)\n",
        "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.3, random_state=42)\n",
        "\n",
        "print(\"x_train shape\",x_train.shape)\n",
        "print(\"x_test shape\",x_test.shape)\n",
        "print(\"y_train shape\",y_train.shape)\n",
        "print(\"y_text shape\",y_test.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WWjLIfNTFbUD",
        "outputId": "50ea52c5-02dd-4331-f2aa-573be8a6b159"
      },
      "id": "WWjLIfNTFbUD",
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Index(['Unnamed: 0', 'Year', 'Month', 'Day', 'Hour', 'Minute', 'Temperature',\n",
            "       'Cloud_Type', 'Dew_Point', 'GHI', 'Relative_Humidity',\n",
            "       'Solar_Zenith_Angle', 'Surface_Albedo', 'Pressure',\n",
            "       'Precipitable_Water', 'Wind_Direction', 'Wind_Speed', 'Location'],\n",
            "      dtype='object')\n",
            "Index(['Month', 'Day', 'Hour', 'Temperature', 'Cloud_Type', 'Dew_Point',\n",
            "       'Relative_Humidity', 'Solar_Zenith_Angle', 'Wind_Direction',\n",
            "       'Wind_Speed', 'Location'],\n",
            "      dtype='object')\n",
            "x_train shape (48289, 11)\n",
            "x_test shape (29565, 11)\n",
            "y_train shape (48289,)\n",
            "y_text shape (29565,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=3)\n",
        "model = keras.Sequential()\n",
        "model.add(layers.InputLayer(input_shape=x_train.shape[1:]))\n",
        "model.add(tf.keras.layers.Normalization())\n",
        "model.add(layers.Dense(100, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(80, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(70, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(60, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(50, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(25, activation=\"relu\", kernel_regularizer=regularizers.L1(0.01)))\n",
        "model.add(layers.Dense(1, activation='relu', kernel_regularizer=regularizers.L1(0.01)))\n",
        "\n",
        "#Compile and train\n",
        "model.compile(optimizer=Adam(), \n",
        "              loss='mse',\n",
        "              metrics=['mae',keras.metrics.RootMeanSquaredError()])"
      ],
      "metadata": {
        "id": "-QrVxfloFbXf"
      },
      "id": "-QrVxfloFbXf",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history = model.fit(x_train, y_train, epochs=200, batch_size=32, validation_data=(x_train,y_train),callbacks=[callback])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EdLjwlIUFbay",
        "outputId": "8e028424-a381-47eb-ad12-e52d8d6a46ce"
      },
      "id": "EdLjwlIUFbay",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/200\n",
            "1510/1510 [==============================] - 9s 5ms/step - loss: 39969.1602 - mae: 99.5369 - root_mean_squared_error: 199.9098 - val_loss: 5111.2207 - val_mae: 44.1509 - val_root_mean_squared_error: 71.4592\n",
            "Epoch 2/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 5061.8423 - mae: 44.5831 - root_mean_squared_error: 71.1133 - val_loss: 4761.9502 - val_mae: 42.8933 - val_root_mean_squared_error: 68.9728\n",
            "Epoch 3/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 4859.3701 - mae: 43.2842 - root_mean_squared_error: 69.6749 - val_loss: 5374.8149 - val_mae: 45.7918 - val_root_mean_squared_error: 73.2796\n",
            "Epoch 4/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 4709.1836 - mae: 42.2956 - root_mean_squared_error: 68.5871 - val_loss: 4560.8628 - val_mae: 40.6242 - val_root_mean_squared_error: 67.4964\n",
            "Epoch 5/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 4604.0449 - mae: 41.2846 - root_mean_squared_error: 67.8153 - val_loss: 4679.5845 - val_mae: 42.7308 - val_root_mean_squared_error: 68.3700\n",
            "Epoch 6/200\n",
            "1510/1510 [==============================] - 10s 6ms/step - loss: 4451.5557 - mae: 40.0573 - root_mean_squared_error: 66.6810 - val_loss: 4281.5596 - val_mae: 38.9309 - val_root_mean_squared_error: 65.3934\n",
            "Epoch 7/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 4384.0337 - mae: 39.3657 - root_mean_squared_error: 66.1720 - val_loss: 4774.2886 - val_mae: 42.1178 - val_root_mean_squared_error: 69.0575\n",
            "Epoch 8/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 4323.7593 - mae: 38.7908 - root_mean_squared_error: 65.7146 - val_loss: 4176.2983 - val_mae: 37.5413 - val_root_mean_squared_error: 64.5829\n",
            "Epoch 9/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 4295.6021 - mae: 38.5042 - root_mean_squared_error: 65.4999 - val_loss: 4955.2075 - val_mae: 43.9521 - val_root_mean_squared_error: 70.3551\n",
            "Epoch 10/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 4065.6902 - mae: 36.9290 - root_mean_squared_error: 63.7202 - val_loss: 3848.5955 - val_mae: 36.4075 - val_root_mean_squared_error: 61.9934\n",
            "Epoch 11/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3895.8760 - mae: 35.9618 - root_mean_squared_error: 62.3734 - val_loss: 3724.2263 - val_mae: 34.1103 - val_root_mean_squared_error: 60.9817\n",
            "Epoch 12/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3730.7942 - mae: 34.7505 - root_mean_squared_error: 61.0352 - val_loss: 5425.2856 - val_mae: 49.4563 - val_root_mean_squared_error: 73.6193\n",
            "Epoch 13/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3720.3875 - mae: 35.0050 - root_mean_squared_error: 60.9495 - val_loss: 5085.8159 - val_mae: 49.0719 - val_root_mean_squared_error: 71.2755\n",
            "Epoch 14/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 3619.5217 - mae: 34.2981 - root_mean_squared_error: 60.1160 - val_loss: 3619.8833 - val_mae: 34.9486 - val_root_mean_squared_error: 60.1187\n",
            "Epoch 15/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3585.9065 - mae: 33.9910 - root_mean_squared_error: 59.8355 - val_loss: 3369.8726 - val_mae: 33.1730 - val_root_mean_squared_error: 58.0019\n",
            "Epoch 16/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3564.8792 - mae: 34.1067 - root_mean_squared_error: 59.6591 - val_loss: 4820.1313 - val_mae: 44.5545 - val_root_mean_squared_error: 69.3857\n",
            "Epoch 17/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3560.0754 - mae: 33.7448 - root_mean_squared_error: 59.6185 - val_loss: 3653.6689 - val_mae: 34.4428 - val_root_mean_squared_error: 60.3983\n",
            "Epoch 18/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 3463.9636 - mae: 33.1958 - root_mean_squared_error: 58.8065 - val_loss: 3283.0583 - val_mae: 31.4639 - val_root_mean_squared_error: 57.2476\n",
            "Epoch 19/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3490.4961 - mae: 33.3547 - root_mean_squared_error: 59.0311 - val_loss: 3601.8547 - val_mae: 37.0981 - val_root_mean_squared_error: 59.9664\n",
            "Epoch 20/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3338.1580 - mae: 32.3711 - root_mean_squared_error: 57.7255 - val_loss: 3117.1113 - val_mae: 29.3887 - val_root_mean_squared_error: 55.7774\n",
            "Epoch 21/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3146.9866 - mae: 30.7857 - root_mean_squared_error: 56.0442 - val_loss: 2910.0103 - val_mae: 27.9847 - val_root_mean_squared_error: 53.8883\n",
            "Epoch 22/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 3045.7778 - mae: 29.7553 - root_mean_squared_error: 55.1334 - val_loss: 2957.1128 - val_mae: 29.2578 - val_root_mean_squared_error: 54.3234\n",
            "Epoch 23/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 3009.0542 - mae: 29.4643 - root_mean_squared_error: 54.7990 - val_loss: 3144.4812 - val_mae: 31.7808 - val_root_mean_squared_error: 56.0206\n",
            "Epoch 24/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2951.1443 - mae: 28.9531 - root_mean_squared_error: 54.2677 - val_loss: 3155.4661 - val_mae: 33.9539 - val_root_mean_squared_error: 56.1187\n",
            "Epoch 25/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2921.2808 - mae: 28.7004 - root_mean_squared_error: 53.9915 - val_loss: 3039.1360 - val_mae: 29.8513 - val_root_mean_squared_error: 55.0718\n",
            "Epoch 26/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2943.2668 - mae: 28.9150 - root_mean_squared_error: 54.1944 - val_loss: 2824.7966 - val_mae: 27.7680 - val_root_mean_squared_error: 53.0899\n",
            "Epoch 27/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2903.6147 - mae: 28.4574 - root_mean_squared_error: 53.8269 - val_loss: 3013.6589 - val_mae: 31.4885 - val_root_mean_squared_error: 54.8395\n",
            "Epoch 28/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2900.2275 - mae: 28.6580 - root_mean_squared_error: 53.7950 - val_loss: 2791.0320 - val_mae: 27.1801 - val_root_mean_squared_error: 52.7703\n",
            "Epoch 29/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2906.8706 - mae: 28.6302 - root_mean_squared_error: 53.8564 - val_loss: 2765.5510 - val_mae: 27.6935 - val_root_mean_squared_error: 52.5277\n",
            "Epoch 30/200\n",
            "1510/1510 [==============================] - 9s 6ms/step - loss: 2864.6660 - mae: 28.1412 - root_mean_squared_error: 53.4627 - val_loss: 2780.4058 - val_mae: 26.6970 - val_root_mean_squared_error: 52.6685\n",
            "Epoch 31/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2850.6367 - mae: 28.1992 - root_mean_squared_error: 53.3309 - val_loss: 2734.2998 - val_mae: 26.6028 - val_root_mean_squared_error: 52.2287\n",
            "Epoch 32/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2837.2190 - mae: 28.1899 - root_mean_squared_error: 53.2046 - val_loss: 2686.8052 - val_mae: 26.6066 - val_root_mean_squared_error: 51.7717\n",
            "Epoch 33/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 2852.1865 - mae: 28.3795 - root_mean_squared_error: 53.3448 - val_loss: 3161.0317 - val_mae: 32.3594 - val_root_mean_squared_error: 56.1647\n",
            "Epoch 34/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 2830.2031 - mae: 28.0203 - root_mean_squared_error: 53.1379 - val_loss: 2712.4202 - val_mae: 27.3050 - val_root_mean_squared_error: 52.0177\n",
            "Epoch 35/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 2810.5806 - mae: 28.1801 - root_mean_squared_error: 52.9526 - val_loss: 2723.6885 - val_mae: 27.2172 - val_root_mean_squared_error: 52.1257\n",
            "Epoch 36/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 2796.9414 - mae: 27.7513 - root_mean_squared_error: 52.8234 - val_loss: 2677.6211 - val_mae: 26.9760 - val_root_mean_squared_error: 51.6816\n",
            "Epoch 37/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2809.1453 - mae: 27.9848 - root_mean_squared_error: 52.9386 - val_loss: 2862.1995 - val_mae: 27.0022 - val_root_mean_squared_error: 53.4371\n",
            "Epoch 38/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2795.8005 - mae: 27.8542 - root_mean_squared_error: 52.8121 - val_loss: 2774.8718 - val_mae: 27.9236 - val_root_mean_squared_error: 52.6136\n",
            "Epoch 39/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 2788.7175 - mae: 27.7271 - root_mean_squared_error: 52.7448 - val_loss: 2923.7891 - val_mae: 30.3781 - val_root_mean_squared_error: 54.0097\n",
            "Epoch 40/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2808.1316 - mae: 27.8593 - root_mean_squared_error: 52.9282 - val_loss: 2683.3865 - val_mae: 26.9621 - val_root_mean_squared_error: 51.7362\n",
            "Epoch 41/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2784.5911 - mae: 27.8679 - root_mean_squared_error: 52.7051 - val_loss: 3015.7993 - val_mae: 32.3854 - val_root_mean_squared_error: 54.8544\n",
            "Epoch 42/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2772.4551 - mae: 27.7528 - root_mean_squared_error: 52.5896 - val_loss: 2925.5459 - val_mae: 26.9984 - val_root_mean_squared_error: 54.0253\n",
            "Epoch 43/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2756.8157 - mae: 27.5506 - root_mean_squared_error: 52.4405 - val_loss: 2706.9421 - val_mae: 27.0259 - val_root_mean_squared_error: 51.9625\n",
            "Epoch 44/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2751.8062 - mae: 27.5762 - root_mean_squared_error: 52.3927 - val_loss: 2831.2002 - val_mae: 29.7246 - val_root_mean_squared_error: 53.1447\n",
            "Epoch 45/200\n",
            "1510/1510 [==============================] - 7s 5ms/step - loss: 2725.3403 - mae: 27.1595 - root_mean_squared_error: 52.1392 - val_loss: 2721.0059 - val_mae: 26.6098 - val_root_mean_squared_error: 52.0974\n",
            "Epoch 46/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2734.9033 - mae: 27.3493 - root_mean_squared_error: 52.2306 - val_loss: 2686.7241 - val_mae: 27.7365 - val_root_mean_squared_error: 51.7671\n",
            "Epoch 47/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2731.8118 - mae: 27.4610 - root_mean_squared_error: 52.2006 - val_loss: 2639.7102 - val_mae: 25.3592 - val_root_mean_squared_error: 51.3106\n",
            "Epoch 48/200\n",
            "1510/1510 [==============================] - 8s 5ms/step - loss: 2728.3931 - mae: 27.3938 - root_mean_squared_error: 52.1675 - val_loss: 2629.9375 - val_mae: 26.8726 - val_root_mean_squared_error: 51.2150\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oqqZkXiqFbeD",
        "outputId": "d0929381-fe1b-4b21-fc5c-5bd7f744d53d"
      },
      "id": "oqqZkXiqFbeD",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " normalization_3 (Normalizat  (None, 11)               23        \n",
            " ion)                                                            \n",
            "                                                                 \n",
            " dense_17 (Dense)            (None, 100)               1200      \n",
            "                                                                 \n",
            " dense_18 (Dense)            (None, 80)                8080      \n",
            "                                                                 \n",
            " dense_19 (Dense)            (None, 70)                5670      \n",
            "                                                                 \n",
            " dense_20 (Dense)            (None, 60)                4260      \n",
            "                                                                 \n",
            " dense_21 (Dense)            (None, 50)                3050      \n",
            "                                                                 \n",
            " dense_22 (Dense)            (None, 25)                1275      \n",
            "                                                                 \n",
            " dense_23 (Dense)            (None, 1)                 26        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 23,584\n",
            "Trainable params: 23,561\n",
            "Non-trainable params: 23\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Save model history into Json"
      ],
      "metadata": {
        "id": "5ux-djp_H1MN"
      },
      "id": "5ux-djp_H1MN"
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "# convert the history.history dict to a pandas DataFrame:     \n",
        "hist_df = pd.DataFrame(history.history) \n",
        "\n",
        "# or save to csv: \n",
        "hist_csv_file = 'history_epochs200_layers8_L2_reg_SelectedFeatures_batch32_DifferentLayerStructure.csv'\n",
        "with open(hist_csv_file, mode='w') as f:\n",
        "    hist_df.to_csv(f)\n",
        "\n",
        "# save to json:  \n",
        "hist_json_file = 'history_epochs200_layers8_L2_reg_SelectedFeatures_batch32_DifferentLayerStructure.json' \n",
        "with open(hist_json_file, mode='w') as f:\n",
        "    hist_df.to_json(f)"
      ],
      "metadata": {
        "id": "iAcb9946HI6g"
      },
      "id": "iAcb9946HI6g",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred=model.predict(x_test)\n",
        "y_pred"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IFSPndSJHJCa",
        "outputId": "138d14ee-4840-4f77-f288-893ea613e280"
      },
      "id": "IFSPndSJHJCa",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "924/924 [==============================] - 2s 2ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[  1.0206823],\n",
              "       [712.1481   ],\n",
              "       [  1.0206823],\n",
              "       ...,\n",
              "       [432.1627   ],\n",
              "       [228.7662   ],\n",
              "       [  1.0206823]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['mae'])\n",
        "plt.plot(history.history['root_mean_squared_error'])\n",
        "plt.title('MAE and RMSE per epoch')\n",
        "plt.ylabel('Error')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['MAE', 'RMSE'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "9Bbr5htDHJKa",
        "outputId": "903a9bb3-65d7-43e6-c836-f214e0993d74"
      },
      "id": "9Bbr5htDHJKa",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "jIffg1poHJUB",
        "outputId": "110db3a1-d775-4418-bc95-fa2e03de5714"
      },
      "id": "jIffg1poHJUB",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import r2_score \n",
        "from sklearn import metrics\n",
        "R_square = r2_score(y_test, y_pred) \n",
        "print('Coefficient of Determination R2', R_square)\n",
        "print(\"Model MAE Selected Features 8 Layers\", metrics.mean_absolute_error(y_test, y_pred))\n",
        "print(\"Model MSE Selected  Features 8 Layers\", metrics.mean_squared_error(y_test, y_pred))\n",
        "print(\"Model RMSE Selected  Features 8 Layers\",np.sqrt(metrics.mean_squared_error(y_test, y_pred))) "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LjP3VmvfINsT",
        "outputId": "61a3aaf1-8f61-4200-f263-0c3ebe54a6f5"
      },
      "id": "LjP3VmvfINsT",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Coefficient of Determination R2 0.9727355964196602\n",
            "Model MAE Selected Features 8 Layers 27.58347840250274\n",
            "Model MSE Selected  Features 8 Layers 2837.8889972483316\n",
            "Model RMSE Selected  Features 8 Layers 53.27184056561526\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Upload model history into Bigquery"
      ],
      "metadata": {
        "id": "jMaPh3ru_9Xg"
      },
      "id": "jMaPh3ru_9Xg"
    },
    {
      "cell_type": "code",
      "source": [
        "from google.cloud import bigquery\n",
        "from google.oauth2 import service_account\n",
        "\n",
        "credentials = service_account.Credentials.from_service_account_file('/content/drive/Shareddrives/DATA298/BigQuery/data298-347103-eaa0e3e59658.json')\n",
        "\n",
        "project_id = 'data298-347103'\n",
        "# Construct a BigQuery client object.\n",
        "#OAuth2ClientId = AdwordsModel.OAuth2ClientId\n",
        "client = bigquery.Client(credentials=credentials,project=project_id)\n",
        "\n",
        "# Set table_id to the ID of the table model to fetch.\n",
        "table_id = 'data298-347103.projectReports.ModelPerformanceResults'\n",
        "\n",
        "# Make an API request.\n",
        "table = client.get_table(table_id)  \n",
        "\n",
        "# View table properties\n",
        "print(f\"Got table {table_id}.\")\n",
        "#print(f\"Table schema: {table.schema}\")\n",
        "#print(f\"Table description: {table.description}\")\n",
        "#print(f\"Table has {table.num_rows} rows\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NwYsO86Bzx3q",
        "outputId": "e64eeb4e-adde-41eb-9348-06e533e70b0b"
      },
      "id": "NwYsO86Bzx3q",
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Got table data298-347103.projectReports.ModelPerformanceResults.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import auth\n",
        "auth.authenticate_user()\n",
        "print('Authenticated')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "axXKraVxHjk-",
        "outputId": "5fba1cc9-5321-487f-b050-fac20869faf6"
      },
      "id": "axXKraVxHjk-",
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Authenticated\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#BigQuery Client\n",
        "import json\n",
        "from pprint import pprint\n",
        "project_id = 'data298-347103'\n",
        "client = bigquery.Client(project=project_id)\n",
        "#BigQuery Lookup for Address\n",
        "insert_sql = \"\"\"INSERT INTO `data298-347103.projectReports.ModelPerformanceResults` \n",
        "                           VALUES(@datevalue, @model_config,@model_name,@target_objective,@iteration,@epoch,@training_metrics,@validation_metrics,@execution_time,@github_link)\"\"\"\n",
        "\n",
        "hist_df = pd.DataFrame(history.history) \n",
        "metrics_results = json.loads(hist_df.to_json())\n",
        "print(type(metrics_results))\n",
        "\n",
        "for i in range(0, 1):\n",
        "  epoch_number = i+1\n",
        "  idx = str(i)\n",
        "  #print(metrics_results[0])\n",
        "  training_metrics = {'loss':metrics_results['loss'][idx],'mae':metrics_results['mae'][idx],'root_mean_squared_error':metrics_results['root_mean_squared_error'][idx]}\n",
        "  validation_metrics = {'val_loss':metrics_results['val_loss'][idx],'val_mae':metrics_results['val_mae'][idx],'val_root_mean_squared_error':metrics_results['val_root_mean_squared_error'][idx] }\n",
        "  print(type(json.dumps(training_metrics)))\n",
        "  print(type(json.dumps(validation_metrics)))\n",
        "  job_config = bigquery.QueryJobConfig(\n",
        "  query_parameters=[\n",
        "      bigquery.ScalarQueryParameter(\"datevalue\",\"TIMESTAMP\",'2022-12-06 02:08:28.801628 UTC'),\n",
        "      bigquery.ScalarQueryParameter(\"model_config\", \"JSON\", model.to_json()),\n",
        "      bigquery.ScalarQueryParameter(\"model_name\", \"STRING\", 'ANN'),\n",
        "      bigquery.ScalarQueryParameter(\"target_objective\", \"STRING\", 'SolarEnergyEstimation'),\n",
        "      bigquery.ScalarQueryParameter(\"iteration\", \"INTEGER\", 0),\n",
        "      bigquery.ScalarQueryParameter(\"epoch\", \"INTEGER\", epoch_number),\n",
        "      bigquery.ScalarQueryParameter(\"training_metrics\", \"JSON\", json.dumps(training_metrics)),\n",
        "      bigquery.ScalarQueryParameter(\"validation_metrics\", \"JSON\", json.dumps(validation_metrics)),\n",
        "      bigquery.ScalarQueryParameter(\"execution_time\", \"INTEGER\", 0),\n",
        "      bigquery.ScalarQueryParameter('github_link','STRING','https://github.com/snvssk/data298A/blob/main/SolarEnergy/MLModels/ANN.ipynb')\n",
        "  ]\n",
        "  )\n",
        "  #print(insert_sql)       \n",
        "  insert_result = client.query(insert_sql,job_config=job_config) \n",
        "  print(insert_result)\n",
        "  \n",
        "  pprint(vars(insert_result))\n",
        "  #print(insert_result.query['query'])\n",
        "  print(insert_result.errors)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Pk_p_IB-qpN6",
        "outputId": "3feaf34f-ab07-489c-f5d4-46355dd0e119"
      },
      "id": "Pk_p_IB-qpN6",
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'dict'>\n",
            "<class 'str'>\n",
            "<class 'str'>\n",
            "QueryJob<project=data298-347103, location=us-east1, id=e80b80d7-fa0b-49f8-bc59-55478c29def4>\n",
            "{'_client': <google.cloud.bigquery.client.Client object at 0x7f8749287700>,\n",
            " '_completion_lock': <unlocked _thread.lock object at 0x7f87492ea960>,\n",
            " '_configuration': <google.cloud.bigquery.job.query.QueryJobConfig object at 0x7f87492eaee0>,\n",
            " '_done_callbacks': [],\n",
            " '_done_timeout': None,\n",
            " '_exception': None,\n",
            " '_job_retry': <google.api_core.retry.Retry object at 0x7f87b2b25e50>,\n",
            " '_polling_thread': None,\n",
            " '_properties': {'configuration': {'jobType': 'QUERY',\n",
            "                                   'query': {'destinationTable': {'datasetId': 'projectReports',\n",
            "                                                                  'projectId': 'data298-347103',\n",
            "                                                                  'tableId': 'ModelPerformanceResults'},\n",
            "                                             'parameterMode': 'NAMED',\n",
            "                                             'priority': 'INTERACTIVE',\n",
            "                                             'query': 'INSERT INTO '\n",
            "                                                      '`data298-347103.projectReports.ModelPerformanceResults` \\n'\n",
            "                                                      '                           '\n",
            "                                                      'VALUES(@datevalue, '\n",
            "                                                      '@model_config,@model_name,@target_objective,@iteration,@epoch,@training_metrics,@validation_metrics,@execution_time,@github_link)',\n",
            "                                             'queryParameters': [{'name': 'datevalue',\n",
            "                                                                  'parameterType': {'type': 'TIMESTAMP'},\n",
            "                                                                  'parameterValue': {'value': '2022-12-06 '\n",
            "                                                                                              '02:08:28.801628 '\n",
            "                                                                                              'UTC'}},\n",
            "                                                                 {'name': 'model_config',\n",
            "                                                                  'parameterType': {'type': 'JSON'},\n",
            "                                                                  'parameterValue': {'value': '{\"class_name\": '\n",
            "                                                                                              '\"Sequential\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"name\": '\n",
            "                                                                                              '\"sequential_1\", '\n",
            "                                                                                              '\"layers\": '\n",
            "                                                                                              '[{\"class_name\": '\n",
            "                                                                                              '\"InputLayer\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"batch_input_shape\": '\n",
            "                                                                                              '[null, '\n",
            "                                                                                              '11], '\n",
            "                                                                                              '\"dtype\": '\n",
            "                                                                                              '\"float32\", '\n",
            "                                                                                              '\"sparse\": '\n",
            "                                                                                              'false, '\n",
            "                                                                                              '\"ragged\": '\n",
            "                                                                                              'false, '\n",
            "                                                                                              '\"name\": '\n",
            "                                                                                              '\"input_2\"}}, '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Normalization\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"name\": '\n",
            "                                                                                              '\"normalization_1\", '\n",
            "                                                                                              '\"trainable\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"dtype\": '\n",
            "                                                                                              '\"float32\", '\n",
            "                                                                                              '\"axis\": '\n",
            "                                                                                              '[-1], '\n",
            "                                                                                              '\"mean\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"variance\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Dense\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"name\": '\n",
            "                                                                                              '\"dense_6\", '\n",
            "                                                                                              '\"trainable\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"dtype\": '\n",
            "                                                                                              '\"float32\", '\n",
            "                                                                                              '\"units\": '\n",
            "                                                                                              '300, '\n",
            "                                                                                              '\"activation\": '\n",
            "                                                                                              '\"relu\", '\n",
            "                                                                                              '\"use_bias\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"kernel_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"GlorotUniform\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"seed\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '\"bias_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Zeros\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{}}, '\n",
            "                                                                                              '\"kernel_regularizer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"L1\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"l1\": '\n",
            "                                                                                              '0.009999999776482582}}, '\n",
            "                                                                                              '\"bias_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"activity_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"kernel_constraint\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"bias_constraint\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Dense\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"name\": '\n",
            "                                                                                              '\"dense_7\", '\n",
            "                                                                                              '\"trainable\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"dtype\": '\n",
            "                                                                                              '\"float32\", '\n",
            "                                                                                              '\"units\": '\n",
            "                                                                                              '200, '\n",
            "                                                                                              '\"activation\": '\n",
            "                                                                                              '\"relu\", '\n",
            "                                                                                              '\"use_bias\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"kernel_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"GlorotUniform\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"seed\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '\"bias_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Zeros\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{}}, '\n",
            "                                                                                              '\"kernel_regularizer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"L1\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"l1\": '\n",
            "                                                                                              '0.009999999776482582}}, '\n",
            "                                                                                              '\"bias_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"activity_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"kernel_constraint\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"bias_constraint\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Dense\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"name\": '\n",
            "                                                                                              '\"dense_8\", '\n",
            "                                                                                              '\"trainable\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"dtype\": '\n",
            "                                                                                              '\"float32\", '\n",
            "                                                                                              '\"units\": '\n",
            "                                                                                              '100, '\n",
            "                                                                                              '\"activation\": '\n",
            "                                                                                              '\"relu\", '\n",
            "                                                                                              '\"use_bias\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"kernel_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"GlorotUniform\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"seed\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '\"bias_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Zeros\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{}}, '\n",
            "                                                                                              '\"kernel_regularizer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"L1\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"l1\": '\n",
            "                                                                                              '0.009999999776482582}}, '\n",
            "                                                                                              '\"bias_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"activity_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"kernel_constraint\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"bias_constraint\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Dense\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"name\": '\n",
            "                                                                                              '\"dense_9\", '\n",
            "                                                                                              '\"trainable\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"dtype\": '\n",
            "                                                                                              '\"float32\", '\n",
            "                                                                                              '\"units\": '\n",
            "                                                                                              '50, '\n",
            "                                                                                              '\"activation\": '\n",
            "                                                                                              '\"relu\", '\n",
            "                                                                                              '\"use_bias\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"kernel_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"GlorotUniform\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"seed\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '\"bias_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Zeros\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{}}, '\n",
            "                                                                                              '\"kernel_regularizer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"L1\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"l1\": '\n",
            "                                                                                              '0.009999999776482582}}, '\n",
            "                                                                                              '\"bias_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"activity_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"kernel_constraint\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"bias_constraint\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Dense\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"name\": '\n",
            "                                                                                              '\"dense_10\", '\n",
            "                                                                                              '\"trainable\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"dtype\": '\n",
            "                                                                                              '\"float32\", '\n",
            "                                                                                              '\"units\": '\n",
            "                                                                                              '25, '\n",
            "                                                                                              '\"activation\": '\n",
            "                                                                                              '\"relu\", '\n",
            "                                                                                              '\"use_bias\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"kernel_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"GlorotUniform\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"seed\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '\"bias_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Zeros\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{}}, '\n",
            "                                                                                              '\"kernel_regularizer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"L1\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"l1\": '\n",
            "                                                                                              '0.009999999776482582}}, '\n",
            "                                                                                              '\"bias_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"activity_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"kernel_constraint\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"bias_constraint\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Dense\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"name\": '\n",
            "                                                                                              '\"dense_11\", '\n",
            "                                                                                              '\"trainable\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"dtype\": '\n",
            "                                                                                              '\"float32\", '\n",
            "                                                                                              '\"units\": '\n",
            "                                                                                              '1, '\n",
            "                                                                                              '\"activation\": '\n",
            "                                                                                              '\"relu\", '\n",
            "                                                                                              '\"use_bias\": '\n",
            "                                                                                              'true, '\n",
            "                                                                                              '\"kernel_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"GlorotUniform\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"seed\": '\n",
            "                                                                                              'null}}, '\n",
            "                                                                                              '\"bias_initializer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"Zeros\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{}}, '\n",
            "                                                                                              '\"kernel_regularizer\": '\n",
            "                                                                                              '{\"class_name\": '\n",
            "                                                                                              '\"L1\", '\n",
            "                                                                                              '\"config\": '\n",
            "                                                                                              '{\"l1\": '\n",
            "                                                                                              '0.009999999776482582}}, '\n",
            "                                                                                              '\"bias_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"activity_regularizer\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"kernel_constraint\": '\n",
            "                                                                                              'null, '\n",
            "                                                                                              '\"bias_constraint\": '\n",
            "                                                                                              'null}}]}, '\n",
            "                                                                                              '\"keras_version\": '\n",
            "                                                                                              '\"2.9.0\", '\n",
            "                                                                                              '\"backend\": '\n",
            "                                                                                              '\"tensorflow\"}'}},\n",
            "                                                                 {'name': 'model_name',\n",
            "                                                                  'parameterType': {'type': 'STRING'},\n",
            "                                                                  'parameterValue': {'value': 'ANN'}},\n",
            "                                                                 {'name': 'target_objective',\n",
            "                                                                  'parameterType': {'type': 'STRING'},\n",
            "                                                                  'parameterValue': {'value': 'SolarEnergyEstimation'}},\n",
            "                                                                 {'name': 'iteration',\n",
            "                                                                  'parameterType': {'type': 'INT64'},\n",
            "                                                                  'parameterValue': {'value': '0'}},\n",
            "                                                                 {'name': 'epoch',\n",
            "                                                                  'parameterType': {'type': 'INT64'},\n",
            "                                                                  'parameterValue': {'value': '1'}},\n",
            "                                                                 {'name': 'training_metrics',\n",
            "                                                                  'parameterType': {'type': 'JSON'},\n",
            "                                                                  'parameterValue': {'value': '{\"loss\": '\n",
            "                                                                                              '16698.765625, '\n",
            "                                                                                              '\"mae\": '\n",
            "                                                                                              '65.1942749023, '\n",
            "                                                                                              '\"root_mean_squared_error\": '\n",
            "                                                                                              '129.0117950439}'}},\n",
            "                                                                 {'name': 'validation_metrics',\n",
            "                                                                  'parameterType': {'type': 'JSON'},\n",
            "                                                                  'parameterValue': {'value': '{\"val_loss\": '\n",
            "                                                                                              '3823.3608398438, '\n",
            "                                                                                              '\"val_mae\": '\n",
            "                                                                                              '34.3180427551, '\n",
            "                                                                                              '\"val_root_mean_squared_error\": '\n",
            "                                                                                              '61.4094581604}'}},\n",
            "                                                                 {'name': 'execution_time',\n",
            "                                                                  'parameterType': {'type': 'INT64'},\n",
            "                                                                  'parameterValue': {'value': '0'}},\n",
            "                                                                 {'name': 'github_link',\n",
            "                                                                  'parameterType': {'type': 'STRING'},\n",
            "                                                                  'parameterValue': {'value': 'https://github.com/snvssk/data298A/blob/main/SolarEnergy/MLModels/ANN.ipynb'}}],\n",
            "                                             'useLegacySql': False}},\n",
            "                 'etag': '5+gvDhoPl2G/4hLXma8W2A==',\n",
            "                 'id': 'data298-347103:us-east1.e80b80d7-fa0b-49f8-bc59-55478c29def4',\n",
            "                 'jobReference': {'jobId': 'e80b80d7-fa0b-49f8-bc59-55478c29def4',\n",
            "                                  'location': 'us-east1',\n",
            "                                  'projectId': 'data298-347103'},\n",
            "                 'kind': 'bigquery#job',\n",
            "                 'principal_subject': 'user:ritanjali.jena@sjsu.edu',\n",
            "                 'selfLink': 'https://bigquery.googleapis.com/bigquery/v2/projects/data298-347103/jobs/e80b80d7-fa0b-49f8-bc59-55478c29def4?location=us-east1',\n",
            "                 'statistics': {'creationTime': 1670292826447.0,\n",
            "                                'query': {'statementType': 'INSERT'},\n",
            "                                'startTime': 1670292826571.0},\n",
            "                 'status': {'state': 'RUNNING'},\n",
            "                 'user_email': 'ritanjali.jena@sjsu.edu'},\n",
            " '_query_results': None,\n",
            " '_result': None,\n",
            " '_result_set': False,\n",
            " '_retry': <google.api_core.retry.Retry object at 0x7f87b2b0ad00>,\n",
            " '_retry_do_query': <function query_jobs_insert.<locals>.do_query at 0x7f87491301f0>,\n",
            " '_transport_timeout': None}\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# #BigQuery Client\n",
        "# import json\n",
        "# project_id = 'data298-347103'\n",
        "# client = bigquery.Client(project=project_id)\n",
        "# #BigQuery Lookup for Address\n",
        "# insert_sql = \"\"\"INSERT INTO `data298-347103.projectReports.ModelPerformanceResults` \n",
        "#                            VALUES(current_timestamp(), @model_config,@model_name,@target_objective,@iteration,@epoch,@training_metrics,@validation_metrics,@execution_time,@github_link,)\"\"\"\n",
        "\n",
        "# metrics_results = json.loads(hist_df.to_json())\n",
        "# print(type(metrics_results))\n",
        "\n",
        "# for i in range(0, 10):\n",
        "#   epoch_number = i+1\n",
        "#   idx = str(i)\n",
        "#   #print(metrics_results[0])\n",
        "#   training_metrics = {'loss':metrics_results['loss'][idx],'mae':metrics_results['mae'][idx],'root_mean_squared_error':metrics_results['root_mean_squared_error'][idx]}\n",
        "#   validation_metrics = {'val_loss':metrics_results['val_loss'][idx],'val_mae':metrics_results['val_mae'][idx],'val_root_mean_squared_error':metrics_results['val_root_mean_squared_error'][idx] }\n",
        "#   job_config = bigquery.QueryJobConfig(\n",
        "#   query_parameters=[\n",
        "#       bigquery.ScalarQueryParameter(\"model_config\", \"JSON\", model.to_json()),\n",
        "#       bigquery.ScalarQueryParameter(\"model_name\", \"STRING\", 'ANN'),\n",
        "#       bigquery.ScalarQueryParameter(\"target_objective\", \"STRING\", 'SolarEnergyEstimation'),\n",
        "#       bigquery.ScalarQueryParameter(\"iteration\", \"INTEGER\", 0),\n",
        "#       bigquery.ScalarQueryParameter(\"epoch\", \"INTEGER\", epoch_number),\n",
        "#       bigquery.ScalarQueryParameter(\"training_metrics\", \"JSON\", training_metrics),\n",
        "#       bigquery.ScalarQueryParameter(\"validation_metrics\", \"JSON\", validation_metrics),\n",
        "#       bigquery.ScalarQueryParameter(\"execution_time\", \"INTEGER\", 0),\n",
        "#       bigquery.ScalarQueryParameter('github_link','STRING','https://github.com/snvssk/data298A/blob/main/SolarEnergy/MLModels/ANN.ipynb')\n",
        "#   ]\n",
        "#   )\n",
        "          \n",
        "#   insert_result = client.query(insert_sql,job_config=job_config) \n",
        "#   print(insert_result)\"\"\""
      ],
      "metadata": {
        "id": "MrmsLad7zFEQ"
      },
      "id": "MrmsLad7zFEQ",
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(insert_result.errors)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CEpUvF-Qx6DM",
        "outputId": "67046d07-d222-4e6e-e013-b8486a197b52"
      },
      "id": "CEpUvF-Qx6DM",
      "execution_count": 51,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[{'reason': 'invalidQuery', 'location': 'query', 'message': 'Syntax error: Unexpected \")\" at [2:185]'}]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dywLD7SPvL71"
      },
      "source": [
        "## Model Artifacts save into folder for deployement\n",
        "### Delete saved model before saving another time"
      ],
      "id": "dywLD7SPvL71"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yb96cdnzgHHe"
      },
      "outputs": [],
      "source": [
        "rm -rf '/content/drive/MyDrive/my_model'"
      ],
      "id": "yb96cdnzgHHe"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9SohDE8Qft1F"
      },
      "outputs": [],
      "source": [
        "model.save('/content/drive/MyDrive/ANN_5Points_FeatureSelected_1000epochs')"
      ],
      "id": "9SohDE8Qft1F"
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['mae'])\n",
        "plt.plot(history.history['root_mean_squared_error'])\n",
        "plt.title('mae and rmse per epoch')\n",
        "plt.ylabel('Error')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['MAE', 'RMSE'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "qmMxwpqGA0pb"
      },
      "id": "qmMxwpqGA0pb",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred=model.predict(x_test)\n",
        "y_pred"
      ],
      "metadata": {
        "id": "EOKaF1QGBAqZ"
      },
      "id": "EOKaF1QGBAqZ",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7559ee4b",
        "outputId": "ebe85bbb-1fef-44cd-8467-105baff5d75d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Coefficient of Determination R2 0.977652089983716\n",
            "Model MAE 21.81954326999383\n",
            "Model MSE 2326.142501515815\n",
            "Model RMSE 48.230099538730116\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import r2_score \n",
        "from sklearn import metrics\n",
        "R_square = r2_score(y_test, y_pred) \n",
        "print('Coefficient of Determination R2', R_square)\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(y_test, y_pred))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(y_test, y_pred))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(y_test, y_pred))) "
      ],
      "id": "7559ee4b"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vySLwRgqntdZ"
      },
      "source": [
        "### The model validated with 30% data and the validation MAE we got 11.68 w/m2, and RMSE 33.17 w/m2. Might be the validation dataset location is different than the training dataset. So, we got bit high MAE and RMSE. We will validate the model with a new location and observe the errors. However, it still performed surprisingly well. "
      ],
      "id": "vySLwRgqntdZ"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vLiP5DIGdRnH"
      },
      "source": [
        "# Test with new location"
      ],
      "id": "vLiP5DIGdRnH"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "akhlVwgbdqUz"
      },
      "source": [
        "## Cupertino"
      ],
      "id": "akhlVwgbdqUz"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "sywhERmAeFjI"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "df =pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/cupertino_test.csv')\n",
        "#df"
      ],
      "id": "sywhERmAeFjI"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_nkQD5aNWpmX",
        "outputId": "2e857fa5-8b1a-4ab9-e65a-ee1923b6a205"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "115/115 [==============================] - 0s 2ms/step\n",
            "Model MAE 28.456212397666828\n",
            "Model MSE 3054.3011576390622\n",
            "Model RMSE 55.265732218428646\n"
          ]
        }
      ],
      "source": [
        "null = df[df.isna().any(axis=1)]\n",
        "df = df.dropna()\n",
        "df.isnull().sum()\n",
        "\n",
        "df = df.loc[(df['Hour'].isin([7, 10, 13, 16, 19])) & (df['Minute'] == 0)]\n",
        "\n",
        "input_test = df.drop(['Unnamed: 0','GHI','Precipitable_Water', 'Surface_Albedo','Year','Pressure','Minute'], axis =1)\n",
        "input_test\n",
        "\n",
        "target_test=df['GHI']\n",
        "target_test\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_nor_test = sc.fit_transform(input_test)\n",
        "y_pred_test=model.predict(input_nor_test)\n",
        "y_pred_test\n",
        "\n",
        "from sklearn import metrics\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(target_test, y_pred_test))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(target_test, y_pred_test))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(target_test, y_pred_test)))"
      ],
      "id": "_nkQD5aNWpmX"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4MIPKvSmfOsM"
      },
      "source": [
        "## FosterCity"
      ],
      "id": "4MIPKvSmfOsM"
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "kQcSEWasnY7k"
      },
      "id": "kQcSEWasnY7k",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "xKQl1q9FfcOz",
        "outputId": "55cbd103-e740-4601-b22e-b77327eb5245"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-b4c79dc8-3bca-424f-b460-8ad6aabe7221\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>Year</th>\n",
              "      <th>Month</th>\n",
              "      <th>Day</th>\n",
              "      <th>Hour</th>\n",
              "      <th>Minute</th>\n",
              "      <th>Temperature</th>\n",
              "      <th>Cloud_Type</th>\n",
              "      <th>Dew_Point</th>\n",
              "      <th>GHI</th>\n",
              "      <th>Relative_Humidity</th>\n",
              "      <th>Solar_Zenith_Angle</th>\n",
              "      <th>Surface_Albedo</th>\n",
              "      <th>Pressure</th>\n",
              "      <th>Precipitable_Water</th>\n",
              "      <th>Wind_Direction</th>\n",
              "      <th>Wind_Speed</th>\n",
              "      <th>Location</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>5.3</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>74.37</td>\n",
              "      <td>165.21</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1028.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>29.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>5.1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>75.41</td>\n",
              "      <td>164.97</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1028.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>30.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.0</td>\n",
              "      <td>74.07</td>\n",
              "      <td>162.23</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1029.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>30.0</td>\n",
              "      <td>4.9</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.0</td>\n",
              "      <td>74.07</td>\n",
              "      <td>157.91</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1029.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>31.0</td>\n",
              "      <td>4.9</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.2</td>\n",
              "      <td>0.0</td>\n",
              "      <td>71.60</td>\n",
              "      <td>152.76</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1029.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>31.0</td>\n",
              "      <td>4.9</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35037</th>\n",
              "      <td>35037</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>11.1</td>\n",
              "      <td>7.0</td>\n",
              "      <td>9.7</td>\n",
              "      <td>0.0</td>\n",
              "      <td>91.09</td>\n",
              "      <td>142.44</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1030.0</td>\n",
              "      <td>1.1</td>\n",
              "      <td>297.0</td>\n",
              "      <td>2.5</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35038</th>\n",
              "      <td>35038</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>10.6</td>\n",
              "      <td>7.0</td>\n",
              "      <td>10.1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>97.02</td>\n",
              "      <td>148.20</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1030.0</td>\n",
              "      <td>1.2</td>\n",
              "      <td>300.0</td>\n",
              "      <td>2.9</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35039</th>\n",
              "      <td>35039</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>10.4</td>\n",
              "      <td>7.0</td>\n",
              "      <td>10.1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>98.32</td>\n",
              "      <td>153.71</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1030.0</td>\n",
              "      <td>1.2</td>\n",
              "      <td>303.0</td>\n",
              "      <td>3.1</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35040</th>\n",
              "      <td>35040</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>10.2</td>\n",
              "      <td>7.0</td>\n",
              "      <td>10.2</td>\n",
              "      <td>0.0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>158.76</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1030.0</td>\n",
              "      <td>1.2</td>\n",
              "      <td>307.0</td>\n",
              "      <td>3.3</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35041</th>\n",
              "      <td>35041</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>10.3</td>\n",
              "      <td>7.0</td>\n",
              "      <td>10.3</td>\n",
              "      <td>0.0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>162.87</td>\n",
              "      <td>0.09</td>\n",
              "      <td>1030.0</td>\n",
              "      <td>1.2</td>\n",
              "      <td>311.0</td>\n",
              "      <td>3.3</td>\n",
              "      <td>120198.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>35042 rows × 18 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b4c79dc8-3bca-424f-b460-8ad6aabe7221')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-b4c79dc8-3bca-424f-b460-8ad6aabe7221 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-b4c79dc8-3bca-424f-b460-8ad6aabe7221');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "       Unnamed: 0    Year  Month   Day  Hour  Minute  Temperature  Cloud_Type  \\\n",
              "0               0  2019.0    1.0   1.0   0.0     0.0          5.3         0.0   \n",
              "1               1  2019.0    1.0   1.0   0.0    30.0          5.1         0.0   \n",
              "2               2  2019.0    1.0   1.0   1.0     0.0          5.0         0.0   \n",
              "3               3  2019.0    1.0   1.0   1.0    30.0          5.0         0.0   \n",
              "4               4  2019.0    1.0   1.0   2.0     0.0          4.9         0.0   \n",
              "...           ...     ...    ...   ...   ...     ...          ...         ...   \n",
              "35037       35037  2020.0   12.0  31.0  21.0    30.0         11.1         7.0   \n",
              "35038       35038  2020.0   12.0  31.0  22.0     0.0         10.6         7.0   \n",
              "35039       35039  2020.0   12.0  31.0  22.0    30.0         10.4         7.0   \n",
              "35040       35040  2020.0   12.0  31.0  23.0     0.0         10.2         7.0   \n",
              "35041       35041  2020.0   12.0  31.0  23.0    30.0         10.3         7.0   \n",
              "\n",
              "       Dew_Point  GHI  Relative_Humidity  Solar_Zenith_Angle  Surface_Albedo  \\\n",
              "0            1.1  0.0              74.37              165.21            0.09   \n",
              "1            1.1  0.0              75.41              164.97            0.09   \n",
              "2            0.8  0.0              74.07              162.23            0.09   \n",
              "3            0.8  0.0              74.07              157.91            0.09   \n",
              "4            0.2  0.0              71.60              152.76            0.09   \n",
              "...          ...  ...                ...                 ...             ...   \n",
              "35037        9.7  0.0              91.09              142.44            0.09   \n",
              "35038       10.1  0.0              97.02              148.20            0.09   \n",
              "35039       10.1  0.0              98.32              153.71            0.09   \n",
              "35040       10.2  0.0             100.00              158.76            0.09   \n",
              "35041       10.3  0.0             100.00              162.87            0.09   \n",
              "\n",
              "       Pressure  Precipitable_Water  Wind_Direction  Wind_Speed  Location  \n",
              "0        1028.0                 0.3            29.0         5.0  120198.0  \n",
              "1        1028.0                 0.3            30.0         5.0  120198.0  \n",
              "2        1029.0                 0.3            30.0         4.9  120198.0  \n",
              "3        1029.0                 0.3            31.0         4.9  120198.0  \n",
              "4        1029.0                 0.3            31.0         4.9  120198.0  \n",
              "...         ...                 ...             ...         ...       ...  \n",
              "35037    1030.0                 1.1           297.0         2.5  120198.0  \n",
              "35038    1030.0                 1.2           300.0         2.9  120198.0  \n",
              "35039    1030.0                 1.2           303.0         3.1  120198.0  \n",
              "35040    1030.0                 1.2           307.0         3.3  120198.0  \n",
              "35041    1030.0                 1.2           311.0         3.3  120198.0  \n",
              "\n",
              "[35042 rows x 18 columns]"
            ]
          },
          "execution_count": 107,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df =pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/fostercity_test.csv')\n",
        "df"
      ],
      "id": "xKQl1q9FfcOz"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "acrDE_aoV9Nv",
        "outputId": "887d77bd-7473-4aea-e679-74c235c3d797"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "115/115 [==============================] - 0s 2ms/step\n",
            "Model MAE 27.817592678298688\n",
            "Model MSE 3151.8031092700794\n",
            "Model RMSE 56.14092187762933\n"
          ]
        }
      ],
      "source": [
        "null = df[df.isna().any(axis=1)]\n",
        "df = df.dropna()\n",
        "df.isnull().sum()\n",
        "\n",
        "df = df.loc[(df['Hour'].isin([7, 10, 13, 16, 19])) & (df['Minute'] == 0)]\n",
        "\n",
        "input_test = df.drop(['Unnamed: 0','GHI','Precipitable_Water', 'Surface_Albedo','Year','Pressure','Minute'], axis =1)\n",
        "input_test\n",
        "\n",
        "target_test=df['GHI']\n",
        "target_test\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_nor_test = sc.fit_transform(input_test)\n",
        "y_pred_test=model.predict(input_nor_test)\n",
        "y_pred_test\n",
        "\n",
        "from sklearn import metrics\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(target_test, y_pred_test))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(target_test, y_pred_test))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(target_test, y_pred_test)))"
      ],
      "id": "acrDE_aoV9Nv"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LAljUqz5fR22",
        "outputId": "f075e7a5-63c3-4763-eb0c-9f11409675d7"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "115/115 [==============================] - 0s 2ms/step\n",
            "Model MAE 29.19396081330025\n",
            "Model MSE 3869.01764842444\n",
            "Model RMSE 62.201428025604365\n"
          ]
        }
      ],
      "source": [
        "\"\"\"df = df.loc[(df['Hour'].isin([7, 10, 13, 16, 19])) & (df['Minute'] == 0)]\n",
        "df = df.drop(['Unnamed: 0'],axis =1)\n",
        "null = df[df.isna().any(axis=1)]\n",
        "df = df.dropna()\n",
        "df.isnull().sum()\n",
        "\n",
        "target_test = df['GHI']#keeping only target(ghi) data here\n",
        "input_test= df.drop(['GHI'], axis=1)\n",
        "input_test.head()\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_nor_test = sc.fit_transform(input_test)\n",
        "y_pred_test=model.predict(input_nor_test)\n",
        "y_pred_test\n",
        "\n",
        "from sklearn import metrics\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(target_test, y_pred_test))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(target_test, y_pred_test))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(target_test, y_pred_test)))\"\"\""
      ],
      "id": "LAljUqz5fR22"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CtKColO3fhYv"
      },
      "source": [
        "### LosGatos"
      ],
      "id": "CtKColO3fhYv"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "qo96d6gofR6w",
        "outputId": "c64a2c5d-9dcb-40a1-e41c-ad816bcf0f2b"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-304da446-8491-4a01-a584-d082bedc4b72\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>Year</th>\n",
              "      <th>Month</th>\n",
              "      <th>Day</th>\n",
              "      <th>Hour</th>\n",
              "      <th>Minute</th>\n",
              "      <th>Temperature</th>\n",
              "      <th>Cloud_Type</th>\n",
              "      <th>Dew_Point</th>\n",
              "      <th>GHI</th>\n",
              "      <th>Relative_Humidity</th>\n",
              "      <th>Solar_Zenith_Angle</th>\n",
              "      <th>Surface_Albedo</th>\n",
              "      <th>Pressure</th>\n",
              "      <th>Precipitable_Water</th>\n",
              "      <th>Wind_Direction</th>\n",
              "      <th>Wind_Speed</th>\n",
              "      <th>Location</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.3</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-0.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>73.88</td>\n",
              "      <td>165.55</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1016.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>29.0</td>\n",
              "      <td>1.9</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>3.1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-0.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>74.93</td>\n",
              "      <td>165.20</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1016.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>26.0</td>\n",
              "      <td>1.9</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-1.3</td>\n",
              "      <td>0.0</td>\n",
              "      <td>73.65</td>\n",
              "      <td>162.32</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1016.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>24.0</td>\n",
              "      <td>1.9</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>2.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-1.3</td>\n",
              "      <td>0.0</td>\n",
              "      <td>73.72</td>\n",
              "      <td>157.90</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1017.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>19.0</td>\n",
              "      <td>1.9</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.8</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-1.8</td>\n",
              "      <td>0.0</td>\n",
              "      <td>71.84</td>\n",
              "      <td>152.68</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1017.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>15.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35037</th>\n",
              "      <td>35037</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>10.7</td>\n",
              "      <td>7.0</td>\n",
              "      <td>7.5</td>\n",
              "      <td>0.0</td>\n",
              "      <td>80.58</td>\n",
              "      <td>142.72</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>235.0</td>\n",
              "      <td>1.1</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35038</th>\n",
              "      <td>35038</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>10.2</td>\n",
              "      <td>7.0</td>\n",
              "      <td>7.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>85.47</td>\n",
              "      <td>148.50</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>251.0</td>\n",
              "      <td>1.4</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35039</th>\n",
              "      <td>35039</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>9.6</td>\n",
              "      <td>7.0</td>\n",
              "      <td>7.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>88.96</td>\n",
              "      <td>154.05</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1018.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>260.0</td>\n",
              "      <td>1.5</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35040</th>\n",
              "      <td>35040</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>9.1</td>\n",
              "      <td>7.0</td>\n",
              "      <td>8.6</td>\n",
              "      <td>0.0</td>\n",
              "      <td>96.76</td>\n",
              "      <td>159.12</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1018.0</td>\n",
              "      <td>1.1</td>\n",
              "      <td>269.0</td>\n",
              "      <td>1.6</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35041</th>\n",
              "      <td>35041</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>8.6</td>\n",
              "      <td>0.0</td>\n",
              "      <td>97.41</td>\n",
              "      <td>163.24</td>\n",
              "      <td>0.12</td>\n",
              "      <td>1018.0</td>\n",
              "      <td>1.1</td>\n",
              "      <td>275.0</td>\n",
              "      <td>1.6</td>\n",
              "      <td>116662.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>35042 rows × 18 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-304da446-8491-4a01-a584-d082bedc4b72')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-304da446-8491-4a01-a584-d082bedc4b72 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-304da446-8491-4a01-a584-d082bedc4b72');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "       Unnamed: 0    Year  Month   Day  Hour  Minute  Temperature  Cloud_Type  \\\n",
              "0               0  2019.0    1.0   1.0   0.0     0.0          3.3         0.0   \n",
              "1               1  2019.0    1.0   1.0   0.0    30.0          3.1         0.0   \n",
              "2               2  2019.0    1.0   1.0   1.0     0.0          2.9         0.0   \n",
              "3               3  2019.0    1.0   1.0   1.0    30.0          2.9         0.0   \n",
              "4               4  2019.0    1.0   1.0   2.0     0.0          2.8         0.0   \n",
              "...           ...     ...    ...   ...   ...     ...          ...         ...   \n",
              "35037       35037  2020.0   12.0  31.0  21.0    30.0         10.7         7.0   \n",
              "35038       35038  2020.0   12.0  31.0  22.0     0.0         10.2         7.0   \n",
              "35039       35039  2020.0   12.0  31.0  22.0    30.0          9.6         7.0   \n",
              "35040       35040  2020.0   12.0  31.0  23.0     0.0          9.1         7.0   \n",
              "35041       35041  2020.0   12.0  31.0  23.0    30.0          9.0         7.0   \n",
              "\n",
              "       Dew_Point  GHI  Relative_Humidity  Solar_Zenith_Angle  Surface_Albedo  \\\n",
              "0           -0.9  0.0              73.88              165.55            0.12   \n",
              "1           -0.9  0.0              74.93              165.20            0.12   \n",
              "2           -1.3  0.0              73.65              162.32            0.12   \n",
              "3           -1.3  0.0              73.72              157.90            0.12   \n",
              "4           -1.8  0.0              71.84              152.68            0.12   \n",
              "...          ...  ...                ...                 ...             ...   \n",
              "35037        7.5  0.0              80.58              142.72            0.12   \n",
              "35038        7.9  0.0              85.47              148.50            0.12   \n",
              "35039        7.9  0.0              88.96              154.05            0.12   \n",
              "35040        8.6  0.0              96.76              159.12            0.12   \n",
              "35041        8.6  0.0              97.41              163.24            0.12   \n",
              "\n",
              "       Pressure  Precipitable_Water  Wind_Direction  Wind_Speed  Location  \n",
              "0        1016.0                 0.3            29.0         1.9  116662.0  \n",
              "1        1016.0                 0.3            26.0         1.9  116662.0  \n",
              "2        1016.0                 0.3            24.0         1.9  116662.0  \n",
              "3        1017.0                 0.3            19.0         1.9  116662.0  \n",
              "4        1017.0                 0.3            15.0         2.0  116662.0  \n",
              "...         ...                 ...             ...         ...       ...  \n",
              "35037    1018.0                 1.0           235.0         1.1  116662.0  \n",
              "35038    1018.0                 1.0           251.0         1.4  116662.0  \n",
              "35039    1018.0                 1.0           260.0         1.5  116662.0  \n",
              "35040    1018.0                 1.1           269.0         1.6  116662.0  \n",
              "35041    1018.0                 1.1           275.0         1.6  116662.0  \n",
              "\n",
              "[35042 rows x 18 columns]"
            ]
          },
          "execution_count": 109,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df =pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/losgatos_test.csv')\n",
        "df"
      ],
      "id": "qo96d6gofR6w"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vAHlW9h1WEee",
        "outputId": "70bdb928-cd97-456c-d8a3-d482858b4bff"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "115/115 [==============================] - 0s 2ms/step\n",
            "Model MAE 28.612866300687397\n",
            "Model MSE 3027.5059819944336\n",
            "Model RMSE 55.022776938231985\n"
          ]
        }
      ],
      "source": [
        "null = df[df.isna().any(axis=1)]\n",
        "df = df.dropna()\n",
        "df.isnull().sum()\n",
        "\n",
        "df = df.loc[(df['Hour'].isin([7, 10, 13, 16, 19])) & (df['Minute'] == 0)]\n",
        "\n",
        "input_test = df.drop(['Unnamed: 0','GHI','Precipitable_Water', 'Surface_Albedo','Year','Pressure','Minute'], axis =1)\n",
        "input_test\n",
        "\n",
        "target_test=df['GHI']\n",
        "target_test\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_nor_test = sc.fit_transform(input_test)\n",
        "y_pred_test=model.predict(input_nor_test)\n",
        "y_pred_test\n",
        "\n",
        "from sklearn import metrics\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(target_test, y_pred_test))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(target_test, y_pred_test))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(target_test, y_pred_test)))"
      ],
      "id": "vAHlW9h1WEee"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 70
        },
        "id": "GoaJP4zpfR9f",
        "outputId": "832fe28a-e90b-42d8-9e90-53218727a164"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'df = df.loc[(df[\\'Hour\\'].isin([7, 10, 13, 16, 19])) & (df[\\'Minute\\'] == 0)]\\ndf = df.drop([\\'Unnamed: 0\\'],axis =1)\\nnull = df[df.isna().any(axis=1)]\\ndf = df.dropna()\\ndf.isnull().sum()\\n\\ntarget_test = df[\\'GHI\\']#keeping only target(ghi) data here\\ninput_test= df.drop([\\'GHI\\'], axis=1)\\ninput_test.head()\\n\\nfrom sklearn.preprocessing import StandardScaler\\nsc = StandardScaler()\\ninput_nor_test = sc.fit_transform(input_test)\\ny_pred_test=model.predict(input_nor_test)\\ny_pred_test\\n\\nfrom sklearn import metrics\\nprint(\"Model MAE\", metrics.mean_absolute_error(target_test, y_pred_test))\\nprint(\"Model MSE\", metrics.mean_squared_error(target_test, y_pred_test))\\nprint(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(target_test, y_pred_test)))'"
            ]
          },
          "execution_count": 111,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "\"\"\"df = df.loc[(df['Hour'].isin([7, 10, 13, 16, 19])) & (df['Minute'] == 0)]\n",
        "df = df.drop(['Unnamed: 0'],axis =1)\n",
        "null = df[df.isna().any(axis=1)]\n",
        "df = df.dropna()\n",
        "df.isnull().sum()\n",
        "\n",
        "target_test = df['GHI']#keeping only target(ghi) data here\n",
        "input_test= df.drop(['GHI'], axis=1)\n",
        "input_test.head()\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_nor_test = sc.fit_transform(input_test)\n",
        "y_pred_test=model.predict(input_nor_test)\n",
        "y_pred_test\n",
        "\n",
        "from sklearn import metrics\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(target_test, y_pred_test))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(target_test, y_pred_test))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(target_test, y_pred_test)))\"\"\""
      ],
      "id": "GoaJP4zpfR9f"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6L8noB8_f3EA"
      },
      "source": [
        "### Saratoga"
      ],
      "id": "6L8noB8_f3EA"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "FaMTW1t7f5Sq",
        "outputId": "0bd1848a-0917-4882-a60f-1a122f2fd140"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-98cb70a1-0c40-4ab0-a3a8-7d1622894bbd\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Year</th>\n",
              "      <th>Month</th>\n",
              "      <th>Day</th>\n",
              "      <th>Hour</th>\n",
              "      <th>Minute</th>\n",
              "      <th>Temperature</th>\n",
              "      <th>Cloud Type</th>\n",
              "      <th>Dew Point</th>\n",
              "      <th>GHI</th>\n",
              "      <th>Relative Humidity</th>\n",
              "      <th>Solar Zenith Angle</th>\n",
              "      <th>Surface Albedo</th>\n",
              "      <th>Pressure</th>\n",
              "      <th>Precipitable Water</th>\n",
              "      <th>Wind Direction</th>\n",
              "      <th>Wind Speed</th>\n",
              "      <th>Location</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2020</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>8.3</td>\n",
              "      <td>7</td>\n",
              "      <td>8.3</td>\n",
              "      <td>0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>165.59</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.0</td>\n",
              "      <td>319</td>\n",
              "      <td>1.7</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2020</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>30</td>\n",
              "      <td>8.5</td>\n",
              "      <td>7</td>\n",
              "      <td>8.5</td>\n",
              "      <td>0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>165.23</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.1</td>\n",
              "      <td>327</td>\n",
              "      <td>1.6</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2020</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>8.7</td>\n",
              "      <td>7</td>\n",
              "      <td>8.7</td>\n",
              "      <td>0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>162.35</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.1</td>\n",
              "      <td>336</td>\n",
              "      <td>1.6</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2020</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>30</td>\n",
              "      <td>8.8</td>\n",
              "      <td>7</td>\n",
              "      <td>8.8</td>\n",
              "      <td>0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>157.92</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.1</td>\n",
              "      <td>336</td>\n",
              "      <td>1.5</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2020</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>8.9</td>\n",
              "      <td>7</td>\n",
              "      <td>8.9</td>\n",
              "      <td>0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>152.69</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.1</td>\n",
              "      <td>337</td>\n",
              "      <td>1.4</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17515</th>\n",
              "      <td>2020</td>\n",
              "      <td>12</td>\n",
              "      <td>31</td>\n",
              "      <td>21</td>\n",
              "      <td>30</td>\n",
              "      <td>10.2</td>\n",
              "      <td>7</td>\n",
              "      <td>7.5</td>\n",
              "      <td>0</td>\n",
              "      <td>83.23</td>\n",
              "      <td>142.70</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>0.9</td>\n",
              "      <td>323</td>\n",
              "      <td>1.1</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17516</th>\n",
              "      <td>2020</td>\n",
              "      <td>12</td>\n",
              "      <td>31</td>\n",
              "      <td>22</td>\n",
              "      <td>0</td>\n",
              "      <td>9.7</td>\n",
              "      <td>7</td>\n",
              "      <td>7.8</td>\n",
              "      <td>0</td>\n",
              "      <td>88.11</td>\n",
              "      <td>148.49</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.0</td>\n",
              "      <td>294</td>\n",
              "      <td>1.4</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17517</th>\n",
              "      <td>2020</td>\n",
              "      <td>12</td>\n",
              "      <td>31</td>\n",
              "      <td>22</td>\n",
              "      <td>30</td>\n",
              "      <td>9.1</td>\n",
              "      <td>7</td>\n",
              "      <td>7.8</td>\n",
              "      <td>0</td>\n",
              "      <td>91.73</td>\n",
              "      <td>154.04</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.0</td>\n",
              "      <td>298</td>\n",
              "      <td>1.6</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17518</th>\n",
              "      <td>2020</td>\n",
              "      <td>12</td>\n",
              "      <td>31</td>\n",
              "      <td>23</td>\n",
              "      <td>0</td>\n",
              "      <td>8.6</td>\n",
              "      <td>7</td>\n",
              "      <td>8.5</td>\n",
              "      <td>0</td>\n",
              "      <td>99.42</td>\n",
              "      <td>159.12</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.0</td>\n",
              "      <td>302</td>\n",
              "      <td>1.7</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17519</th>\n",
              "      <td>2020</td>\n",
              "      <td>12</td>\n",
              "      <td>31</td>\n",
              "      <td>23</td>\n",
              "      <td>30</td>\n",
              "      <td>8.5</td>\n",
              "      <td>7</td>\n",
              "      <td>8.5</td>\n",
              "      <td>0</td>\n",
              "      <td>100.00</td>\n",
              "      <td>163.25</td>\n",
              "      <td>0.11</td>\n",
              "      <td>1006</td>\n",
              "      <td>1.0</td>\n",
              "      <td>311</td>\n",
              "      <td>1.7</td>\n",
              "      <td>116661</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>17520 rows × 17 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-98cb70a1-0c40-4ab0-a3a8-7d1622894bbd')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-98cb70a1-0c40-4ab0-a3a8-7d1622894bbd button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-98cb70a1-0c40-4ab0-a3a8-7d1622894bbd');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "       Year  Month  Day  Hour  Minute  Temperature  Cloud Type  Dew Point  \\\n",
              "0      2020      1    1     0       0          8.3           7        8.3   \n",
              "1      2020      1    1     0      30          8.5           7        8.5   \n",
              "2      2020      1    1     1       0          8.7           7        8.7   \n",
              "3      2020      1    1     1      30          8.8           7        8.8   \n",
              "4      2020      1    1     2       0          8.9           7        8.9   \n",
              "...     ...    ...  ...   ...     ...          ...         ...        ...   \n",
              "17515  2020     12   31    21      30         10.2           7        7.5   \n",
              "17516  2020     12   31    22       0          9.7           7        7.8   \n",
              "17517  2020     12   31    22      30          9.1           7        7.8   \n",
              "17518  2020     12   31    23       0          8.6           7        8.5   \n",
              "17519  2020     12   31    23      30          8.5           7        8.5   \n",
              "\n",
              "       GHI  Relative Humidity  Solar Zenith Angle  Surface Albedo  Pressure  \\\n",
              "0        0             100.00              165.59            0.11      1006   \n",
              "1        0             100.00              165.23            0.11      1006   \n",
              "2        0             100.00              162.35            0.11      1006   \n",
              "3        0             100.00              157.92            0.11      1006   \n",
              "4        0             100.00              152.69            0.11      1006   \n",
              "...    ...                ...                 ...             ...       ...   \n",
              "17515    0              83.23              142.70            0.11      1006   \n",
              "17516    0              88.11              148.49            0.11      1006   \n",
              "17517    0              91.73              154.04            0.11      1006   \n",
              "17518    0              99.42              159.12            0.11      1006   \n",
              "17519    0             100.00              163.25            0.11      1006   \n",
              "\n",
              "       Precipitable Water  Wind Direction  Wind Speed  Location  \n",
              "0                     1.0             319         1.7    116661  \n",
              "1                     1.1             327         1.6    116661  \n",
              "2                     1.1             336         1.6    116661  \n",
              "3                     1.1             336         1.5    116661  \n",
              "4                     1.1             337         1.4    116661  \n",
              "...                   ...             ...         ...       ...  \n",
              "17515                 0.9             323         1.1    116661  \n",
              "17516                 1.0             294         1.4    116661  \n",
              "17517                 1.0             298         1.6    116661  \n",
              "17518                 1.0             302         1.7    116661  \n",
              "17519                 1.0             311         1.7    116661  \n",
              "\n",
              "[17520 rows x 17 columns]"
            ]
          },
          "execution_count": 136,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df =pd.read_csv('/content/drive/Shareddrives/DATA298/WeatherData/FinalWeatherData/saratoga_2020.csv')\n",
        "df"
      ],
      "id": "FaMTW1t7f5Sq"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DM4ge1vKYf3T",
        "outputId": "f709b167-8a3a-446d-9fd5-ac15ec30d8c6"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "58/58 [==============================] - 0s 2ms/step\n",
            "Model MAE 27.304809762353766\n",
            "Model MSE 2987.8395232790194\n",
            "Model RMSE 54.66113357111266\n"
          ]
        }
      ],
      "source": [
        "null = df[df.isna().any(axis=1)]\n",
        "df = df.dropna()\n",
        "df.isnull().sum()\n",
        "\n",
        "df = df.loc[(df['Hour'].isin([7, 10, 13, 16, 19])) & (df['Minute'] == 0)]\n",
        "\n",
        "input_test =df.drop(['Surface Albedo','Precipitable Water','GHI','Year','Pressure','Minute'], axis=1)\n",
        "input_test\n",
        "\n",
        "target_test=df['GHI']\n",
        "target_test\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_nor_test = sc.fit_transform(input_test)\n",
        "y_pred_test=model.predict(input_nor_test)\n",
        "y_pred_test\n",
        "\n",
        "from sklearn import metrics\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(target_test, y_pred_test))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(target_test, y_pred_test))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(target_test, y_pred_test)))"
      ],
      "id": "DM4ge1vKYf3T"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7jLKzLOkf6Zx",
        "outputId": "141e1458-b4d7-4f1d-888f-734c8938b20a"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "58/58 [==============================] - 0s 2ms/step\n",
            "Model MAE 28.052558402361935\n",
            "Model MSE 3582.6169744593362\n",
            "Model RMSE 59.854966163713904\n"
          ]
        }
      ],
      "source": [
        "\"\"\"df = df.loc[(df['Hour'].isin([7, 10, 13, 16, 19])) & (df['Minute'] == 0)]\n",
        "#df = df.drop(['Unnamed: 0'],axis =1)\n",
        "null = df[df.isna().any(axis=1)]\n",
        "df = df.dropna()\n",
        "df.isnull().sum()\n",
        "\n",
        "target_test = df['GHI']#keeping only target(ghi) data here\n",
        "input_test= df.drop(['GHI'], axis=1)\n",
        "input_test.head()\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "sc = StandardScaler()\n",
        "input_nor_test = sc.fit_transform(input_test)\n",
        "y_pred_test=model.predict(input_nor_test)\n",
        "y_pred_test\n",
        "\n",
        "from sklearn import metrics\n",
        "print(\"Model MAE\", metrics.mean_absolute_error(target_test, y_pred_test))\n",
        "print(\"Model MSE\", metrics.mean_squared_error(target_test, y_pred_test))\n",
        "print(\"Model RMSE\",np.sqrt(metrics.mean_squared_error(target_test, y_pred_test)))\"\"\""
      ],
      "id": "7jLKzLOkf6Zx"
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [
        "fOG5SyLs_u6I",
        "zTIM-stc_17I",
        "6APnUz9Xniz5",
        "-ewkt8vs-tWp",
        "oPq2HLl4CyZn",
        "qfOc00mRFaZ6",
        "dywLD7SPvL71",
        "vySLwRgqntdZ",
        "vLiP5DIGdRnH",
        "CtKColO3fhYv",
        "6L8noB8_f3EA"
      ],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.7"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}