LTE_Cell_Traffic_Grow_and_Congestion_Forecasting.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "LTE Cell Traffic Grow and Congestion Forecasting.ipynb",
      "provenance": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyMRb0g+04oKunLbsi1mlKII"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# LTE Cell Traffic Grow and Congestion Forecasting"
      ],
      "metadata": {
        "id": "4iBCys5kiCn7"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## I. Introduction"
      ],
      "metadata": {
        "id": "nWBgg9WHoDuJ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "This code was writen based on paper:\n",
        "* D. Chmieliauskas and D. Guršnys, \"LTE Cell Traffic Grow and Congestion Forecasting,\" 2019 Open Conference of Electrical, Electronic and Information Sciences (eStream), 2019, pp. 1-5, doi: 10.1109/eStream.2019.8732145. PDF found on: https://www.researchgate.net/publication/345446849_LTE_Cell_Traffic_Grow_and_Congestion_Forecasting\n",
        "\n",
        "The purpose of this work was to provide a replicable code for the [paperswithcode.com](https://paperswithcode.com/). For more details, go to: https://paperswithcode.com/paper/lte-cell-traffic-grow-and-congestion.\n",
        "\n",
        "**Notes:**\n",
        "1. The study and data cleaning steps are beyond the scope of this work, being done beforehand and omitted from the code.\n",
        "2. The results shown in this work were obtained using real data from a cellular network, but due to compliance reasons, these data can't be shared.\n",
        "3. There are some differences between the method proposed by the authors and this work. This may have occurred due to lack of information in the paper or our judgment that the changes could obtain better accuracy. In any case, these differences were mentioned throughout the code."
      ],
      "metadata": {
        "id": "xNKvU8yDk78S"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "from scipy import stats\n",
        "from sklearn import metrics\n",
        "from datetime import datetime, timedelta\n",
        "from itertools import cycle\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "plt.rcParams[\"figure.figsize\"] = (8,5)\n",
        "\n",
        "from fbprophet import Prophet"
      ],
      "metadata": {
        "id": "YWT73XAEiZnH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Data Transform"
      ],
      "metadata": {
        "id": "R_c-jLp_S6rX"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The paper does not mention any form of data normalization, but in our studies we've seen that the Prophet achieve better accuracy when we apply the Box-Cox method for this solution."
      ],
      "metadata": {
        "id": "J1CWPS_1CX_m"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Creating inverse Box-Cox\n",
        "def inverse_boxcox(y, lambda_):\n",
        "    return np.exp(y) if lambda_ == 0 else np.exp(np.log(lambda_ * y + 1) / lambda_)"
      ],
      "metadata": {
        "id": "t8UJgaURLTr0"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Evaluation Metrics"
      ],
      "metadata": {
        "id": "k1XHAgx9Pkv-"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "As evaluation metrics, the authors worked with:\n",
        "* Explained Variance Score\n",
        "* R² Score\n",
        "* Mean Average Percentage Error (MAPE)\n",
        "\n",
        "Despite being widely used, the MAPE metric has several flaws, among them the high sensitivity to outliers, which are frequently found in the data of our study. For this reason, we will use SMAPE instead, which is a slightly more robust metric. For more details, see the work of Chris Tofallis in the study entitled \"*A better measure of relative prediction accuracy for model selection and model estimation*\"."
      ],
      "metadata": {
        "id": "fJtwZJU3QENv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def smape_metric(y_true, y_pred):\n",
        "  smape = 1/len(y_true) * np.sum(2 * np.abs(y_true-y_pred) / (np.abs(y_true) + np.abs(y_pred)))\n",
        "  return smape\n",
        "\n",
        "metric_cols = ['sMAPE', 'EVS', 'R2']\n",
        "\n",
        "def timeseries_evaluation_metrics_func(y_true, y_pred):\n",
        "  sMAPE = smape_metric(y_true, y_pred)\n",
        "  EVS = metrics.explained_variance_score(y_true, y_pred)\n",
        "  R2 = metrics.r2_score(y_true, y_pred)\n",
        "\n",
        "  return [sMAPE, EVS, R2]"
      ],
      "metadata": {
        "id": "iuae3_2BPgN-"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## II. Total Daily Traffic Forecasting"
      ],
      "metadata": {
        "id": "g-ZaLdLri1bR"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Getting Data"
      ],
      "metadata": {
        "id": "mbgo2aHtVAej"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Note:** In the original paper authors used 180 days, where 5 months was used for training and one month for forecast testing. In this work we used 1.5 years for training, mantaining 28 days for forecast testing. By doing this, we can take advantage of the annual seasonality, not originally used by the authors."
      ],
      "metadata": {
        "id": "ljGVPlVUv_qG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Opening daily cell data\n",
        "daily_traffic = pd.read_csv('df_daily.csv', sep=';')\n",
        "daily_traffic.head(10)"
      ],
      "metadata": {
        "id": "ej3W5M31jAI3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        },
        "outputId": "dff5128f-5911-4d49-c5dc-f91fc95000cf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Cell  Band(MHz)        Date   Data_Volume\n",
              "0  cell0001       1800  2019-01-01  9.406620e+06\n",
              "1  cell0001       1800  2019-01-02  3.229836e+06\n",
              "2  cell0001       1800  2019-01-03  5.869033e+06\n",
              "3  cell0001       1800  2019-01-04  6.525841e+06\n",
              "4  cell0001       1800  2019-01-05  7.415637e+06\n",
              "5  cell0001       1800  2019-01-06  1.244459e+07\n",
              "6  cell0001       1800  2019-01-07  1.201881e+07\n",
              "7  cell0001       1800  2019-01-08  1.411982e+07\n",
              "8  cell0001       1800  2019-01-09  1.780909e+07\n",
              "9  cell0001       1800  2019-01-10  1.164014e+07"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-734ff40b-017f-422e-9d4a-bc11a366a3ae\">\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>Cell</th>\n",
              "      <th>Band(MHz)</th>\n",
              "      <th>Date</th>\n",
              "      <th>Data_Volume</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-01</td>\n",
              "      <td>9.406620e+06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-02</td>\n",
              "      <td>3.229836e+06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-03</td>\n",
              "      <td>5.869033e+06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-04</td>\n",
              "      <td>6.525841e+06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-05</td>\n",
              "      <td>7.415637e+06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-06</td>\n",
              "      <td>1.244459e+07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-07</td>\n",
              "      <td>1.201881e+07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-08</td>\n",
              "      <td>1.411982e+07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-09</td>\n",
              "      <td>1.780909e+07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>1800</td>\n",
              "      <td>2019-01-10</td>\n",
              "      <td>1.164014e+07</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-734ff40b-017f-422e-9d4a-bc11a366a3ae')\"\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-734ff40b-017f-422e-9d4a-bc11a366a3ae 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-734ff40b-017f-422e-9d4a-bc11a366a3ae');\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": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Renaming columns for Prophet standard, where ds its the date\n",
        "# and y it's a target variable Data_Volume\n",
        "daily_traffic.columns = ['Cell', 'Band(MHz)', 'ds', 'y']\n",
        "daily_traffic['ds'] = pd.to_datetime(daily_traffic['ds'])"
      ],
      "metadata": {
        "id": "pDq85Q8jkKQi"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Checking cell quantity and data range\n",
        "list_cells = daily_traffic.Cell.unique()\n",
        "print(f'Dataset have {len(list_cells)} unique cells',\n",
        "      f'with date between {min(daily_traffic.ds)} and {max(daily_traffic.ds)}.')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "B8v0FTx4jZBA",
        "outputId": "ccea31b1-4324-4cf3-fd06-d02d3c93ef9e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Dataset have 159 unique cells with date between 2019-01-01 00:00:00 and 2021-01-28 00:00:00.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Splitting dataset\n",
        "n_days_forecast = 28\n",
        "train_ini, train_end = '2019-01-01', '2020-07-01'\n",
        "val_ini, val_end = train_end, datetime.strptime(train_end, \"%Y-%m-%d\") + timedelta(days=n_days_forecast)\n",
        "\n",
        "df_daily_train = daily_traffic[(daily_traffic['ds'] >= train_ini) & (daily_traffic['ds'] < train_end)].copy()\n",
        "df_daily_val = daily_traffic[(daily_traffic['ds'] >= val_ini) & (daily_traffic['ds'] < val_end)].copy()"
      ],
      "metadata": {
        "id": "VdA0TMXu-S2m"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Forecasting"
      ],
      "metadata": {
        "id": "9Ou00AYgVEYE"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Note:** Authors do not make it clear whether they used the standard/automatic hyperparameters or whether they did tuning. In this work, we customized all the seasonality in order to get better performance. In this way it's possible to test several configurations and see which one performs best."
      ],
      "metadata": {
        "id": "csz38sgMwVIL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Parameters\n",
        "daily_params = [# Config01\n",
        "                [{'changepoint_prior_scale': 0.05},                                     # Trend sensibility\n",
        "                  {'fourier_order': 3, 'mode': 'multiplicative', 'prior_scale': 0.1},   # Weekly Hyperparameters\n",
        "                  {'fourier_order': 4, 'mode': 'additive', 'prior_scale': 1}],          # Yearly Hyperparameters\n",
        "\n",
        "                # Config02\n",
        "                [{'changepoint_prior_scale': 0.05},\n",
        "                  {'fourier_order': 3, 'mode': 'multiplicative', 'prior_scale': 0.05},\n",
        "                  {'fourier_order': 4, 'mode': 'additive', 'prior_scale': 1}],\n",
        "\n",
        "                # Config03\n",
        "                [{'changepoint_prior_scale': 0.05},\n",
        "                  {'fourier_order': 3, 'mode': 'multiplicative', 'prior_scale': 0.1},\n",
        "                  {'fourier_order': 4, 'mode': 'additive', 'prior_scale': 0.1}]\n",
        "]"
      ],
      "metadata": {
        "id": "aVK3uULk1yHt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "final_result_daily = pd.DataFrame()\n",
        "\n",
        "for cell in list_cells[:]:\n",
        "  #print(cell)\n",
        "  n_config = 1\n",
        "  # Filtering cell\n",
        "  df_train_temp = df_daily_train[df_daily_train['Cell']==cell].copy()\n",
        "  df_val_temp = df_daily_val[df_daily_val['Cell']==cell].copy()\n",
        "\n",
        "  # Applying Box-Cox\n",
        "  df_train_temp['y'], lambda_prophet = stats.boxcox(df_train_temp['y'])\n",
        "\n",
        "  # Executing all configurations\n",
        "  for params in daily_params:\n",
        "    config = 'conf_' + str(n_config).zfill(len(str(len(daily_params))))\n",
        "\n",
        "    # Instantiating the model\n",
        "    m = Prophet(yearly_seasonality=False,\n",
        "                weekly_seasonality=False,\n",
        "                daily_seasonality=False,\n",
        "                uncertainty_samples=0,      # Zeroed to speed up forecast\n",
        "                **params[0])\n",
        "\n",
        "    # Weekly seasonality\n",
        "    m.add_seasonality(name='weekly', period=7, **params[1])\n",
        "\n",
        "    # Yearly seasonality\n",
        "    m.add_seasonality(name='yearly', period=365.25, **params[2])\n",
        "\n",
        "    # Fit\n",
        "    m.fit(df_train_temp)\n",
        "\n",
        "    # Create future dataset\n",
        "    future = m.make_future_dataframe(n_days_forecast, freq='d')\n",
        "\n",
        "    # Forecast\n",
        "    forecast = m.predict(future)\n",
        "\n",
        "    # Applying inverse Box-Cox in forecast data\n",
        "    forecast['yhat'] = inverse_boxcox(forecast['yhat'], lambda_prophet)\n",
        "\n",
        "    # Insert config number\n",
        "    forecast['configuration'] = config\n",
        "\n",
        "    # Mergin forecasted and actual data\n",
        "    temp_result_daily = pd.merge(df_val_temp,\n",
        "                                 forecast[['ds', 'yhat', 'configuration']],\n",
        "                                 left_on='ds',\n",
        "                                 right_on='ds',\n",
        "                                 how='inner')\n",
        "\n",
        "    final_result_daily = pd.concat([final_result_daily, temp_result_daily])\n",
        "\n",
        "    n_config += 1"
      ],
      "metadata": {
        "id": "U1UBgMpWVY7V"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Evaluating"
      ],
      "metadata": {
        "id": "hebcqhW8VTuF"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### LTE Traffic Forecasting Accuracy Overall Result"
      ],
      "metadata": {
        "id": "Gk-wpajGOwr0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Eveluating results for each configuration\n",
        "evaluating_cols = metric_cols + ['Configuration']\n",
        "overall_result_daily = pd.DataFrame(columns= evaluating_cols)\n",
        "full_eval = []\n",
        "daily_configurations = final_result_daily.configuration.unique()\n",
        "\n",
        "for configuration in daily_configurations:\n",
        "  temp_df = final_result_daily[final_result_daily['configuration']==configuration]\n",
        "  temp_eval = timeseries_evaluation_metrics_func(temp_df.y, temp_df.yhat)\n",
        "  temp_eval.append(configuration)\n",
        "  full_eval.append(temp_eval)\n",
        "\n",
        "overall_result_daily = pd.concat([overall_result_daily,\n",
        "                                  pd.DataFrame(full_eval, columns= evaluating_cols)])\n",
        "overall_result_daily.round(5)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "Ty0tunW0KMHN",
        "outputId": "56a9e2c0-50d8-4d00-c7ef-8ff86af24234"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     sMAPE      EVS       R2 Configuration\n",
              "0  0.29390  0.85330  0.85234        conf_1\n",
              "1  0.29032  0.87967  0.87854        conf_2\n",
              "2  0.29395  0.88292  0.88147        conf_3"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ac056ee4-866f-4b68-946b-cfc70298e201\">\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>sMAPE</th>\n",
              "      <th>EVS</th>\n",
              "      <th>R2</th>\n",
              "      <th>Configuration</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.29390</td>\n",
              "      <td>0.85330</td>\n",
              "      <td>0.85234</td>\n",
              "      <td>conf_1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.29032</td>\n",
              "      <td>0.87967</td>\n",
              "      <td>0.87854</td>\n",
              "      <td>conf_2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.29395</td>\n",
              "      <td>0.88292</td>\n",
              "      <td>0.88147</td>\n",
              "      <td>conf_3</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ac056ee4-866f-4b68-946b-cfc70298e201')\"\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-ac056ee4-866f-4b68-946b-cfc70298e201 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-ac056ee4-866f-4b68-946b-cfc70298e201');\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": 72
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Eveluating results for each configuration\n",
        "evaluating_cols_1 = metric_cols + ['Configuration', 'cell']\n",
        "individual_result_daily = pd.DataFrame(columns=evaluating_cols_1)\n",
        "ind_eval = []\n",
        "daily_configurations = final_result_daily.configuration.unique()\n",
        "cells_eval = final_result_daily['Cell'].unique()\n",
        "\n",
        "for cell in cells_eval:\n",
        "  temp_cell_df = final_result_daily[final_result_daily['Cell']==cell]\n",
        "\n",
        "  for configuration in daily_configurations:\n",
        "    temp_df = temp_cell_df[temp_cell_df['configuration']==configuration]\n",
        "    temp_eval = timeseries_evaluation_metrics_func(temp_df.y, temp_df.yhat)\n",
        "    temp_eval.append(configuration)\n",
        "    temp_eval.append(cell)\n",
        "    ind_eval.append(temp_eval)\n",
        "\n",
        "individual_result_daily = pd.concat([individual_result_daily,\n",
        "                                  pd.DataFrame(ind_eval, columns= evaluating_cols_1)])"
      ],
      "metadata": {
        "id": "sE2psBdmv0ay"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Note:** After run all desired configurations, pick one that best performed. This one will be used further, in the final solution."
      ],
      "metadata": {
        "id": "7XrOu3cZlRVG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "daily_prediction = final_result_daily[final_result_daily['configuration']=='conf_3'].copy().reset_index(drop=True)\n",
        "daily_prediction.drop(['configuration'], axis=1, inplace=True)"
      ],
      "metadata": {
        "id": "-bazAA39k_KO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Mean absolute percentage error vs forecast day\n",
        "The chart shows the calculated forecast results for all LTE cells value for every forecast day from 1 to 28"
      ],
      "metadata": {
        "id": "855TpbxmnLD-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def daily_metrics(dataset, metric):\n",
        "  daily_metric = []\n",
        "\n",
        "  for day in dataset['ds'].unique():\n",
        "    temp_d_forecast = dataset[dataset['ds']==day]\n",
        "    performance = metric(temp_d_forecast.y, temp_d_forecast.yhat)\n",
        "    daily_metric.append(performance)\n",
        "\n",
        "  return daily_metric"
      ],
      "metadata": {
        "id": "YanndU_KQNhO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "daily_smape = daily_metrics(daily_prediction, smape_metric)\n",
        "\n",
        "# Plotting\n",
        "x = range(1, len(daily_smape)+1)\n",
        "y = daily_smape\n",
        "plt.fill_between( x, y)\n",
        "plt.title('Symmetric mean absolute percentage error vs forecast day', fontsize=14)\n",
        "plt.xlabel('Forecast Day', fontsize=12)\n",
        "plt.ylabel('Symmetric mean  absolute precenteage error', fontsize=12);"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 354
        },
        "id": "0rmKwsJVnKH7",
        "outputId": "b8163b0d-02ea-46d0-acb6-25080f3493cb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAFRCAYAAABzDARaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdd5hdZbn38e+dSIJUEZCSEAhFioAgMeA5ghwUCCJgAaUpIBLxiPhSDqIUIRIVRFAhkIQWeugQIIJASCC99zozmSSTXieTNvV+/3jWJis7U9bM7DUtv8917Wv2fla7Z7d7r2c9xdwdERERabvaNXcAIiIiki4lexERkTZOyV5ERKSNU7IXERFp45TsRURE2jglexERkTZOyV62YWZXmtmG5o6jJTGzu8xsesrHGGBm76R5DJE0mNkFZjbPzCrMbEBzx9MczGwfM3MzO725Y6mJkn0WM9vXzB4xs0IzKzWz5Wb2kZmd2dyxNZSZDTWzhxOu/hJwaJrxSG7U83Vt06Iv2gubO44d1BPAa8DBwG+bOZYGM7NDovdRt+aOJQ2fa+4AWqDXgF2Aq4E84EvAt4C9mzOopmBmO7n7ZmBzc8cirZ+ZdXD3suaOo7WIPn/lWWUNeg6b6rk3sy8Qvhvfd/fFjdiP3itpc3fdohvwBcCB79Syzp3A9GrKRwD/iu4PAN4BfgcsA4qBvxJqUu4CVkTlv8vahwO/At4CNgFzgf8BOgPvAxuBycDXsrb7L2BYtM1i4FFgj1gsnnU7BDg9uv9dYCxQBnwPuBLYkLX/7wJjCD8CVgNvAzvX8PxcCWwAzgFmRzENAvYELgTmRc/Hs8DnY9sZcAuQHx1nGnB51r7/CsyJlhcC98XjiJ7b6cDF0X5KgDeBfep43ZPu9xfAwmi9bfYLHAd8BKyP/v8pwP/Elp8WPYdbgOXAg0CH2PIBwDuxx0OBh7Pi/Gydml7XaNkxwLvR/78CeBHYv5b//5Bo+0uB4VGMs4Gzstardb9s+74vAlZE5QcCz0fvnU2E93D8uTkPmBAddz7QO+u5KQRuB/pFz28R8H9Zy+PPQ2FUfhjhs7SM8NmZCHwv63/aj/D+3AwsAK6KXuu7YuvsCfSP/ucSwmetWx3vqQ7AvVGsm4BxwNmx5adT/edvKOHzez+wEhiX8P1T7XZZMX05OuZxWeU9gVXATtHtX8ASoBRYBPy1hv8x8z/Eb6dHy35I+Axn9nEbYFmv2V3Ak8A64JW6vsti3xM3Eb5HSqPn9y/1+CwfFL0n1kTHmA1cHPv+jd+G1vL6fp2t79lJwLlZ/397Qo3H/CiWeYTvt3ax17OcrM8l4b0/tbb3VkNvTZJEW8uNUNNREr3Za0pmnYEKoHus7Mjohf5q9HgA4UupL3AUcAlQBbwH/CX60F0bbXNSbD8evcEvAY4AXiB8sN8HLoi2Gxx/MxCSzIboA3AEcDIwCng1Wr4nMDL6UO0f3dqz9YM6DTiLUHW/L1nJHugR/b/3EL7sjwduBnap4fm5MnoTfwicBHyD8MXxIeFHwvGEHzBrgZuy3uRzouN1JSSejcC5sXXuAP6bkJy+S0i8f4otvyt6Lt6IjvMNwhd4vzpe96T7HQqcGK07AxgUW2ca8Fz0eh8O/AD4RrSsU/S/9AWOJnypLwP+Htt+APVL9jW9rgcQvrjvjY51fPS8jyH6oqnm/z8kei8UAT+O/oeHCF9SnaJ16txvFF8JIbEfS3hv7kr4ohsBnEpIwD8kSvbA2YTPylXRsv+J3gf3x+IrJPxQuC56bn8TxZt5fveNHv8ieh72jcq/SvicHRdtdxshqR4V2/d7hB9m3wBOIPxgKyFK9oTkMpzwI6d7tJ8/RTEfUMt76nlgNOFL/dAo9jK2fkecTvWfv6HR8f8evQ5Hk+z9s912NcQ1lqzkTUiuj0T3byIk59OALoTke1UN++pA+E7w6DXdPyo7CagE7iZ8Z11G+Pz8Jus1XU9IgIcTvrtq/S6LtvsL4cfBz6PtvgH8bz0+y28DH0Tvja6E75se0bKvR//L2dH/8sUa/u/dCD/8XiG8z88GZrFtst8J6BXt8xDC52odcHVsP7OBW2KP20XP/W9TyW9p7LQ134AfEX71bYneaPcDJ2et8w7QN/b4XmB87PGA6EVrHysbD0zJ2k8hcHPssbPtr9Rjo7IbY2WnR2X7RI+fAZ7I2u8J0Tpfih4PZfvEkdnPj7LKr2TbZD8CGFiP5+/KaL9HxsruJ3z442fCA9iauHYlJJZTs/b1D2BwLce6FsiLPb4ret32jJXdFl8n4f9Q3X4rgS6xsm9G/+cR0eP1wBU17K83IeG1i5VdSTgz2SX7+ajlNUuyTi/go6yyvaJYu9cQ3yHR8ttiZe0INUv3JN1vFN9KoGNsnWsISaja2hXgE+COrLLvE770LfY5eTFrnXnA7VmfnQsTvLajM9ux9Uf6KbHlB0Wv9V3R4zOiWD6ftZ/JxL6os5YdRvhx3yWr/E22JtXTqf7zN5SsM7uE75/ttqshtusJP4Azz22XKNb/ih7/i/CDx+raV7T+PsSSXFT2PDAka727gKLY40Lg7ax1av0uIyTZLcC1SWKLts/+LE8F/ljH56CuWpuehMS9W6zs8uznoZrt/gp8GHt8MzAr9vic6DXdO+n/V5+bGuhlcffXCNWO5wH/JvyyHW1mf4it9hhwsZl93szaAz8lVNnEzXT3ytjj5YTqQbLKvpRVNjVrOYRf/9llme1OAi43sw2ZGyFBQ/jSqcv4OpafSPjw10epu8+JPV4OLHP3VVllmf/hGGBn4L2s/+NXxP4HM7vQzIab2bJo+YOEL6u4Be5eHHu8hO2f420k3O9id18YezyG8CV5dPT4AeBxMxtiZreZ2VGxdY8GRrt7VaxsOOEs6PDaYmuAk4DTsp7HRdGyut4PozJ3oljHEF6b+ux3uruXxh6fSEhC8dc+O97bsvb7AuEH4P6x9aZmbZfkdd3VzO4zs5lmtjbadze2vrZHEV7Dzz4D7r4o2nc8vl2AlVkxHkvNz+fXCDUCM7O2Obeabar7/E3Iepz0/ZO9XXUGEr7fTo0eXwLMd/eR0eMBhAQ718z6mNm5ZlbfPHE0W7+D4vF2MrM9YmXZ/3td32XHAB2p5fsowWf5n8DtZjbKzO4xs5Pq+b9B+P+munu819Ko7JXM7FozG29mK6NYbsiK5WngUDP7r+jxz4E33X11A2KqkxroVcPdtxCqej4AepnZ48BdZna/h0Yk7xKu9/yIcP35C4QvqLjyrMdeQ1n2B6k8a3lNZe1ifx8nvKmzJWkwszHBOvVVkfW4rv898/c8QrVbXDmAmZ1C+KK6m/ChWQecT6g12G79Go6znXrst1bufpeZPU/4dX428Eczu9bdn6xr0xrKqwgJI26nBKG0I7w/b65m2fJqypJKut/6vp/aEZ77V6pZtjJ2v16va+R+QjXtzYQz402Es8cO9YxvOVuTY9z6WrZxQhVudtzZjV+re77q8xzG3z91bufuK8zsA0LV+ifR3+djyyea2SGE9/C3CQlpipmdmfVjo6Fqi7eu77Ljattxks+yuz9hZu8Tqvi/A4w0s7+4+131+zdqZ2Y/IdRM3ky43LYe+DXh8l4mlpVmNgj4uZnNiWI9L5dxxCnZJzOT8FztDJS5e6Y/6c8Jyf71rLPJpjQR+Iq759WyThnhem5DTCJ86B9r4PZJzCRUXx3s7kNqWOe/CWfXf8oUmNnBOTh20v12MrODojM/CNdv2xGu1QHg7vMISeVfZvYo4Rryk9E6PzazdrEvzG8SXpf8GuJaSbhOHvdVQvVnRnWv60TC9cEFntWyO4FTgCEAZmaE//HVRu53EvBTM9unhrP7iYRr6LW9f5MoZ/vn4pvAM1FtHWa2M+EMcW60fDbhNTyJUIuBmXUmnPnG49sPqHL3goSxTCL8UNvf3T+u/7+ynYa8f2rzHPCwmfUnJNBtuiy6ewnhdX81+p4bTahBmEsyswifq7hvEqrxS2rZrtbvMjObRfie+Dbhc5Yt0WfZ3YsIDS77m9nvCN0F7yI8n1D3d+Us4Eoz29XdMz9YTsla55vAGHf/rGusmVVXE/QY4bkuILTD+LCOYzeYqvFjzGzvqBr2cjM73sy6mtlFhEYkH7l7/Jf844Qued9j+yr8pnQv0N3M+prZiWZ2uJl9z8z6xdYpjNY5xMLgD/V53XsDF0VVXseY2VfM7AYz2yVX/0D0BXA/cL+Z/Tz6H06IqsF6RqvNJSTcy8zsUDP7FaEKsrGS7ncz8HQU1zcIjaXedfd50eWcPmZ2evQcn0z4sM+Mtn2EkEAeMbOjzexcwvW7h919Uw1xDQHOMbPzzexIM3uAcD05rpDtX9c+hMZ7L5nZydH/9B0z629mu9fxXPwqqgY9knBWcjChNTSN2O8LhMZMb5nZqdF255vZ/0TLewGXmlkvMzvWzI6KYrivjlizFQLfNrP9zWyvqGwu8AMz+5qZHUdIcjtnNoguNb0P9DWzU8zsBOApQg1A5gz0Q0JV8ltmdk70nfANM7vbzKo728fd5xLOlgdE/8uhZtbNzG42sx/W8/+Chr1/avMmoZboCUKr/c+SuJndaGaXRMc5nNBQNtMDIqm/A9+yMBjVl83sMkKju7pe01q/y6LviX8CfzGzq8zsMDPrHn1mIcFn2cz+aWY9ouUnEGp+Mp/TFYTP+dlmtp+Z7VlDnC8Qai+fjL4PzyS0DYqbC3wtes8cYWZ3EPJFtg8IjU//CAzIUe1J9dJoCNBab4TrQX8mdJNZS/jQzyNcj92uZSbhCzmfrMYsZDWkisreiV7MeNlotm11vE0jI6pv/HJUVHZsrKwboVXxekLV2DSgV2z5lwnXlDJfYoeQ1dAvtu6VbN/17nzC9cBSQovsQdTR9S6r7Gai7lCxsr+ybaNGI7SyzpzlryR8EM6MrfOXqHwD8Drhmr7Hlt9FVrfI6uKpJuZE+yU0zFlE+EJ4i62tvjsQvgAKo9iXEM4c4l2GMl2nStnadSrekG2b9wzhy7hP9HyvIlRNZq+z3esalR9BOFtYG8U6h9C6vkMN//8h0faXEaoct0TbnJO1Xq37zY4vtl1nwmBN66JYJ7Hte/os4NNo2XrCtdzrYssLiTVkjcqGEmucSKj+nEc4wy+Myg4mJOuNhGR1M1mfQ0K7gLej/3lh9H7JJ9YtFtidkGSKCGd/iwjVxYfV8p7aKXrfFETbLCN8bk6Klp9O9Z+/bf6verx/qt2ulvieiY5/fVb5NYQz7JLotRhG1Hivhv1s9x0VlWe63mWer+q63t1czf7q+i5rB9wae14XAb3r8Vl+KHqfbInWG0jU4yRanuleW0ntXe9Ojp6nUkJvjvPizwPhO+EJwmdlXXT/TrK+B6N17yRctjsk6evXkFumRaY0gJnNBJ53997NHYtIQ1m4Rjsf+Lq719Vgs00zs30IP9Yu8aj6XyRNFi75He7uqY7Sqmv2DWBm+xKucx1CGOhDRFohMzuDcOY+jdC6vzehJuW95oxL2r7oMsExwM8I7WFS1WTX7KPrJHPMLM/Mbq1lvR9Z1vjEZvb7aLs5ZnZ200RcqxWEa42/9Jq7FIlIy7cTYcCoaYTq/E3Aab614ZVIWt4idCN80t3fTftgTVKNb6Ev+lzgTMJ1r3GEarKZWevtTuje04FwzW68mR1DGJazO6GRyofAl33bPuwiIiJSg6Y6s+9OGMWowEM/9YGE4V+z/YnQInNLrOwCwghupe4+nzA5Tfe0AxYREWkr6kz2ZtbezJ42s46NOE4nto62BeHsvlPWcb4GHFRNdUad24qIiEjN6myg5+6VZnYWoWtAKqL+wQ8Qur00dB89CV2j2HXXXU866qij6thCRESk7ZgwYcIqd9+3umVJW+M/CNxtZn/0+o/KBWGow/iAIJ3ZdijX3QljTQ81Mwh9XweZ2fkJtgXA3fsT+jbTrVs3Hz9+h+5BJCIiOxgzW1DTsqTJ/jeEBHyjma0kNr6xu2dPGFKdccARZtaVkKgvJozMlNlHMWFwhkzAQwkDLow3s83AC9EIYgcSBvYYmzBuERGRHV7SZH95Yw7iYSz56whDU7YndDWYYWa9CKOoDapl2xlm9jJhZLUK4NdqiS8iIpJcmxxBT9X4IiKyozGzCe7erbplibremdlO0cQPBWa2Jfp7t5nVZ6pIERERaQZJq/HvI/RtvxZYQJhg4g5gD8K8wSIiItJCJU32FwFfdffV0eM5ZjaRMNuPkr2IiEgLlnQEPatnuYiIiLQQSZP9K8DbZna2mR1tZj2AN4GX0wtNREREciFpNf4twO1AH0Jf98WE8e3vSSkuERERyZE6k300Y91jQE93vzP9kERERCSX6qzGjwawSXVsfBERkaZWVlHFf2Ysa+4wmkTSa/aZsfHVr15ERFq9DaUVXP30OK4fOIk1G8uaO5zUJU32vwH+D1hvZovMbGHmlmJsIiIiObdqQymX9B/Np/NWsaW8igEjC5s7pNQ1ydj4IiIiLcHC1Zv42ZNjKFy96bOyp0cW0vO0Q9mtY9KU2PokbaD3JHCMu5emH5KIiEjuTV9czJVPjWPVhm1TWfHmcgaOXcgvTj20mSJLX9IGepXAzumHIyIiknsj81Zxcf/R2yX6jMc+LaC0ou1OqJr0mv0/gJfN7FtmdpiZHZq5pRmciIhIY70zdQlXPjWODaUVNa6zfH0pb01a0oRRNa2kFygejv6emVXuhPnpRUREWpwBI+Zz9zszSTKbe99h+fzopM60b9f2RoJPdGbv7u1quCnRi4hIi+Pu/O392dz1drJED1CwamOb7XeftBofADM7yMxOSSsYERGRxqqorOJ3r02lz8f59d720WH5eNJfB61IomRvZl3MbAQwG/gwKrvQzB5PMzgREZH62FxWyS+fncDL44satP3UomJG5q+ue8VWJumZfT/gXWB3oDwq+4Dtr+GLiIg0i7Uby7js8dF8NHtFo/bz6ND61wi0dEmTfXfgr+5eRWiUh7sXA3umFZiIiEhSS9Zt5qJ+o5i4cF2j9zU8bxVTixq/n5YkabJfDhweLzCzYwANlysiIs1q7vISfvjISPJWbMjZPtva2X3SZH8/8I6ZXQV8zswuAV4C7k0tMmk2bbFxioi0TTOWFPPjfqNYtn5LTvf73oxl5K/M3Y+H5pa0692ThIlwLgIWAT8D7nD351OMTZrJ+AVrKa/UjMYi0rJNX1zMpY+NYd2m8rpXrid36D+sIOf7bS6Ju965+1vu/l13/4q7n+Pub6YZmDSf4fNWMWdZSXOHISJSo6lF67j0sdEUb859os94fVIRy4pzW2PQXOrVz74xzKyHmc0xszwzu7Wa5dea2TQzm2xmw6M2AZjZIWa2OSqfbGZ9myrmHdWEBWuZtKhtNU4RkbZj8qJ1XPb4GNZvqXn421wor3Qe/zT3Z/dVVc7kJv6ObZJkH82c1wc4BzgGuCSTzGNecPfj3P0E4D7ggdiyfHc/Ibpd2xQx76gqKquYtHAtk3PQolVEJNcmLlzLTx8fQ0nKiT7jhbELWbepLGf7q6pybn9rOoMmN+04/E11Zt8dyHP3AncvAwYCF8RXcPf1sYe7EnXxk6Y1e1kJG8sqmbxobXOHIiKyjQkL1vCzJ8ZSUsuENrm2qaySZ0YtyMm+qqqc296czgtjmr4jW1Ml+06Ehn0ZRVHZNszs12aWTzizvz62qKuZTTKzYWZ2arqh7tgmLAhJPn/lRtZvSe9amIhIfYwrDIm+tpnr0vLUiPlsKmvccTOJ/sWxzdNjPelwuR3NrLeZFZhZcVR2lpldl8tg3L2Pux8G/A64PSpeCnRx9xOBG4EXzGyPamLsaWbjzWz8ypUrcxnWDmX8gq1n9FMXFTdjJCIiwZiC1Vzx5Fg2ljXPfPNrN5Xz0rhFda9Yg+ZO9JD8zP5B4FjgMrZWr88AfpVw+8XAQbHHnaOymgwEvg/g7qXuvjq6PwHIB76cvYG793f3bu7ebd99900YlmSbULjms/uqyheR5jYqfzVXPjWOTc2U6DMe+6SgQV2SQ6Kf1qyJHpIn+x8Al7r7KKAKwN0XU01VfA3GAUeYWVcz6wBcDAyKr2BmR8QengvMi8r3jRr4YWaHAkcAbafzYwuyZN1mlsS6mTR1a1ERkbgReau4asBYNpc3b6IHWFK8pd6N6rYm+obXCuRK0mRfBnwuXmBm+wKJpgZy9wrgOuB9YBbwsrvPMLNeZnZ+tNp1ZjbDzCYTquuviMpPA6ZG5a8C17r7GiTn4lX4AJMXFWs0PRFpFp/MXcnPB4xjS3nLGeCr77B8qqqSfSdWVTl/eKNlJHrISuC1eAV42sxuADCzA4B/EKrbE3H3wcDgrLI7Y/d/W8N2rwGvJT2ONFy8Ch9g1YZSFq/bTOe9dmmmiERkRzR0zgp6PjuBsoqWk+gB5q3YwEezV3DmMfvVul4m0Q9sxHX+XEt6Zv8HYD4wDfgCoYp9CXB3SnFJM8g+swdV5Ys0Rpqju7VVH89eQc9nWl6iz3hkaF6tNZ4tMdFDwjP7qG/8DcANUfX9Klf9bpuyobSCWUvXb1c+eeE6vnf8gc0QkUjr9t70ZTwxvICXf/kNzKy5w0ld8eZyHvxgLp9rZ+y7e0e+tEdH9t1t5+hvR76wy051Pg8fzVrOr56bSFkLnptj0sJ1jJm/hlMO3Xu7ZS010UPCZB81jIvbPXrRSoGl0Tz30opNXriO6i5FTWljczqLNIWitZu45dUprN9SwesTF/Ojkzo32bGLN5dz3QsT6XXBsXTdZ9cmOWbByg384pnxFKzcWOM6O7U39t2tI/vu3pF9d985/CDYPfO4I+s2lXH7m9Mpr2z555GPDs3fLtlXVTm/f30aL41veYkekl+zzyN0uYv/LMu8IlVmNgj4X3dfnsvgpOmMX1B9m8dpi4spr6xip/ZNNo2CSKtWXlnF9S9O+mzc9r/8ezZnfmU/9th5pyY5/r3vzebTeas4/6Hh/P3HX+Wsr+yf6vGGz1vF/z4/oc5x6ssrnSXFW6IeP617DI9hc1cyY0kxXzlwT6DlJ3pIfs3+GuAFQre3nQn93J8F/hc4jvCjoU8aAUrTmFDN9XqALeVVmgFPpB4e+GAuE2NzS6zaUMqDH8xtkmOPK1zz2VCsJaUV9Hx2An97fzaVCVuQ14e788yoQq54amzqE9K0RI8OzQdCor/19aktOtFD8mR/N3CNu+e7e5m75xES/R3uPhu4Ejg9nRAlbZVVzqRaJr5RIz2RZD6Zu/KzJBD39MjCatvE5FJpRSW/f33aduV9Ps7nyqfGsmZj7iZzKa+s4rY3p3PnWzNS+SHRGgyetpSClRu49fWpvDy+qLnDqVPSZN8OOCSrrAvQPrq/keSXBKSFmb1sfa3jTU9Rshep04qSLdz48uRql1U53PnW9FTHreg7tIC8FRuqXfbpvFWc99BwphU1vvp87cYyfvrEmGaZzKUlqXL4cb9RrSLRQ/Jk/w9gSDQ+/rVmdg/wUVQO8F1gVBoBSvpqqsLP0Jm9SO0qq5wbXprMqg01nz2PK1zLm5NrGyW84fJWbKDPx3m1rrN43WZ+1HckLzeipfjc5SVc0GcEows0rhlQ6+vd0iRK9u5+H/BzYH/C1LQHAle7+73R8jfd/ZzUopRUjS+sPdnnrdxAiWbAE6nRo0PzGJFX94Civd+dnfPZJKuqnD+8Pi1Rd7WyiipueW0qv399GqUV9RuCdsjs5fzwkZEsXLOpoaFKM0rcxNrd33P3q939HHf/ubu/l2Zg0nTqOrN3JyfVfyJt0bjCNTyQsAHeqg2l/OODeTk9/svjFzG2sH5n2i+OXciP+45i8brNda7r7vT/JJ+rnx7fLNPLSm4kTvZmdoKZ/cbM7o7GtO9lZr3SDE7St7R4c6IP/CRV5YtsZ+3GMq5/cVK1Y1TU5OlRhcxelpvGeitLSvnz4FkN2nZKUTHnPTScEXmralyntKKSm1+Zyp8Hz0bDqLVuSeez7wmMAM4gzDV/HHATcHh6oUlTqKsKP0PX7UW25e7836tTWRqbKTKJyirnzjdn5KSxXq93Zjaq29uaqLHdo0Pzt4tnZUkplz42htcmto4GaFK7pGf2twA93P0HwObo74WALuS2cnVV4WdMXrROM+CJxDw1opAPZzVsHLGxhWt4q57TpWb7ePYK3p7SuH1AaFV+73uzufa5CZ+1zZm5ZD3f7zMi8feDtHxJk/2X3P3T6H6VmbVz938D56UUlzSRmkbOy7aypLTeZzCSHnenvAWPH97WTSsq5i//blj1eUbvwbMa3PB1Y2kFt785vVHHz/b+jOVc8PAIBoyYz48eHZno8p60HkmTfZGZHRLdnwtcYGanEua5l1ZqY2kFs5YmHx1PVfktx/C8VTt8P+fmUrKlnOtenNjoMdxXlpTyjw8b1ljvwQ/mppKMC1Zt5K63Z7K5vH4t9aXlS5rs7wOOju73Ap4DhqApblu1yYvW1Wv0KyX7lsHd+eeH83hkaB5b9KXcpNydP7wxnQWrc9P9bMDIwnoPRz2tqJgnR8zPyfFlx5G0n/2AqNqe6O9ewF7u/miawUm6kjbOy5hcy5C60nRGFaxm/IK1LF9fykstcCrNtuylcYtycp08o7LKuaMeI+tVVFZx6+tT69X6XwTq1/VubzP7qZndEs1vv4eZNd28jZJzSa/XZ0xbXEyFrhM3u3/Gqn51dt905i4v4a63Z+R8v2Pnr2FQwh8QT40oZMaSdMfYl7Ypade7bwFzgMuAO6LiIwCd2bdSlVVe7zP1zeWVzF1e/djb0jRGF6xmzPytP9KWry9l4Fhdu0/b5rJKfv38RLaUp/Njt/e7dTfWW7RmU+LBe0Sy1Wds/J+4ew8g06lzDNA9lagkdXOXl1DSgNGwdN2+eT00ZPsGXY8MzdfZfcrufnsG82qYZCYXVpSUblNjk83duf3N6Wo4Jw2WNNkf4u4fRfczV4vK0Ex3rdb4BvafnbxI/W6by/jCNdWOv76ipJQXdXafmkFTljCwCdpGPFVLY723py5l2NyVqccgbVfSZD/TzM7OKvsOsP3kydIqTKjnWNoZUxZpjPzm8q8hNc9qprP73FqweiNPDp/P5Y+P4caXqjAAzh0AACAASURBVJ+2Ntcqq7zaaXDXbSqjVwptBWTHkvTM/CbgHTN7F/i8mfUjDKhzQWqRSaoaemY/d0UJG0or2K2jKnWa0qSFa/mkljO7lSWlvDBmIT//ZtcmjKrtqKisYsKCtQyZvYKPZq+ocV74tI2JGutdcEKnz8r+Mnh2q5pKVVqmpF3vRgPHAzOAJ4H5QHd3H5f0QGbWw8zmmFmemd1azfJrzWyamU02s+Fmdkxs2e+j7eZUU8Mg9bR8/RaK1jZsQA53mFqk6/ZN7V8f1T34yqPDdHZfH8Wbynlr8mJ+O3ASJ93zIT/pP5p+nxQ0W6LP6P3urM9mlxuVv5qXxqt7pTReotMzM7vZ3e8nDK4TL7/R3R9IsH17oA9wJlAEjDOzQe4+M7baC+7eN1r/fOABoEeU9C8GvgIcCHxoZl92d32rNVB9+9dnm7xoHf912D45ikbqMrVoHR/Pqft67cqSUp4fs5CrdXZfLXcnf+VGhsxezoezVjBhwdp6DSrVVEJjvbncdNaR3PaGrpRKbiSti70TuL+a8tsJSbku3YE8dy8AMLOBhEsAnyV7d493Ht2VrQ0BLwAGunspMN/M8qL9jUoYu2Spb//6bFPUIr9J/eujmq/VZ+s7LJ/LTu7Czju1TzGi1qWyyrnv/dm8N31Zzka+S9tTIwpZWVJKwaqNzR2KtBG1JnszOyO6297M/gew2OJDgaTjPHYC4nVRRcDJ1Rzv18CNQAfCdLqZbUdnbdsJabDGzmSl7ndNZ/ri4nrNrLaypJTnRi/gF6cemmJUrcujQ/PoN6ygucOol4oq581GzoonElfXmf0T0d+dCdfqMxxYBvwml8G4ex+gj5ldSqg1uCLptmbWE+gJ0KVLl1yG1aZsKqto9Ahcy9eXsrR4Mwfs+fkcRSU1qa5ffV36DivgspMP5vMddHY/YcFaHmzgZDMibUmtDfTcvau7dwWez9yPboe6+3+5+6CEx1kMHBR73Dkqq8lA4Pv12dbd+7t7N3fvtu+++yYMa8dT38lvatyPxslP3ayl63l/Rv3nS1+1oZTnxyxIIaLWZf2Wcn47cFKLvC4v0tSStsb/Wea+mbWL3xIeZxxwhJl1NbMOhAZ32/xQMLMjYg/PBTI/xwcBF5tZRzPrShimd2zC40qWCY1snJcxWS3yU/dwLf3q69J3WD6by3bcNqzuzu1vTG9wrxORtibp2PhfM7NRZrYRKI9uFdHfOrl7BXAd8D4wC3jZ3WeYWa+o5T3AdWY2w8wmE67bXxFtOwN4mdCY7z3g12qJ33AN7V+fTWf26Zq7vITB05c2ePtVG8p26LP71yYuTjy5jMiOIGlr/KeBt4GfAw1qzurug4HBWWV3xu7/tpZtewO9G3Jc2aqqypm4MDfJftriYiqrnPbtrO6Vpd4eGpJHwllPa9R3WD6XntyFXTq0jAGQnh29gJlLiun9/eNol+L7pnDVRu58a3pq+xdpjZJ+CxwM3OZJJ12WFmnuihJKttR/8pvqbCqrZN6KEo7af4+c7E+2yluxgXemNv6sdNWGMp4fvZBrTmvelvkVlVXc8+4sBowsjEqM3t8/NpWEX1ZRxfUDJ7FpB76EIVKdpNfc3wDOSjMQSV9jB9PJpqr8dPT5uPFn9Rl9h+WzqSw3P/AaomRLOb94Znws0cOLYxfyx0EzthsDPhf+/sEcphZp/gaRbEmT/c7AG2b2HzN7Jn5LMzjJrcb2r8+m/va5N3/VRt6aXFtHlfpZvbGM50Y3z7X7RWs28aNHRzK0mtH/nh29gLvfnpnThD983qpW159epKkkrcafSWy0O2mdGjtyXjYl+9x7eEgeue4p1m9YAZefcnCTXrufsGANPZ+ZwOqNNU/gMmBkITu1N/7w3aMxa1yV/uoNpdzwctPMTifSGiX69Lv73WkHIulasX4Li9bkthvS3OUlbCytYFfNgJcTC1Zv5M0cntVnrN5YxrOjFvDLbx2W831X581Ji7nl1amUVVbVue5jn86nfbt2/K7HkQ1O+O7O/706lZUlpQ3aXmRHkLQaHzM708yeMLO3o8fdYsPpSguXqy53cVUeWuVLbjzycX5qA8D0+6Qg9Wv3VVXO3/8zh//30uREiT6j77B8HvhgboOP+8yoBQyZvaLB24vsCJL2s/8N8ChhoJvTouLNwD0pxSU5luvGeRmqys+NRWs28drEotT2v2ZjGc+MSu/a/eaySq57cSIPNXAgoIeG5PHPBgxrO2vpenoPntWgY4rsSJKe2f8/4Dvu/lcg85N9NnBkKlFJzk3I8fX6DLXIz41HhuZTkfKwrv0/KWBjae7P7les38JP+o9i8LRljdrPgx/Opc/HyX8sbC6r5PoXJ1FWkbwWQWRHlTTZ787WWesy30g7ATW3vpEWY3NZZaMnv6nJFA2b22iL123m1QmL6l6xkdI4u5+xpJgL+ozIWXe3v70/h37D8hOte8+7M5m3YkNOjivS1iVN9p8At2aVXQ98nNtwJA2TF61L7axxafEWlq/fksq+dxR9h+ZTXtk041X1/yQ/Z2f3/5mxjIv6jmJpcW5f/7/8ezaPf1p7F7r3pi/j+TELc3pckbYsabL/DfADMysEdjezOcCPCWPYSwuXVhV+xiRV5TfYsuItvDQu/bP6jLWbynl6VGGj9uHu9BuWzy+fm5DaSHX3vDuLASPmV7tsafFmbn19airHFWmrkna9W2pmXwe+Thg6dxEw1t11sawVSKMlftzkRevocez+qR6jreo7LL9eLddz4bFPCvhal70wwjU5d3A8/I3dr3IP1+wy9z2s//6MZbw6Ib3GhBl3vT2Tz7Vvx+WnHPxZWWWVc8NLk1m3KdEcXCISSZTszewEYLW7jyWaXtbMDjKzL7r7lDQDlMapqnImppzsp6hFfoO8PWUJL4xt+qrotZvKubj/6CY/bkPc/uZ0PtfOuLh7FyD8OBpdkG5NlUhblHQ0lOeA87PKOgDPAsfnNCLJqXkrNrA+R5Pf1GRq0TrNgFcPazaWccdb03l3asOnsN2R/P6NabRvZxz2pd0a1R9fZEeWNNl3cfdtWsy4e76ZHZLziCSncj1EbnU2llWSt2IDR+6/e+rHau0+mLmc378+jVUbNNpbUu5wy2tT2XvXDqkNOiTS1iVtoFdkZl+LF0SPGz8Pp6RqQkqD6WSbvKjxx/lo1nIKVrbNrlTrt5Rz8ytTuOaZ8Ur0DeAepuwVkYZJemb/IPCWmd0H5AOHATcDvdMKTHIj7cZ5GZMXFfOTrzds28oq5x8fzuWhIXnst0dHXur5DQ7ZZ9fcBtiMhs9bxS2vTmFJjruoiYgklbQ1/mNmtg64GjiI0Br/Jnd/Nc3gpHFWlGxh4ZpNTXKshg6bu25TGb8dOJlhc8M0qMvXl3LJY6MZ2PMUDt67dSf8jaUV/PXfs3m2maaYFRHJSDxdmbu/ArySYiySY01VhQ8wZ9l6NpVV1Gsa1RlLirn2uQnbzca3tHgLl/QfzUu//AYHfXGXXIfaJMYVruHmV6awYHXT/NgSEalN0olwzMyuMbOPzGxqVHaamf043fCkMZqqCh+iGfDqMWTqG5OK+OEjI2ucdndJ8RYu7j+aRU1UM5ErW8or+fPgWfy43yglehFpMZI20OtFqMJ/DOgSlRUBv0sjKMmNpkz2kGyc/LKKKu4aNIMbXppCaR0TmCxet5lLHhvN4nXV/yBoaaYWreN7Dw2n/ycFuBqNi0gLkjTZXwl8z90HsnUinPnAoWkEJY23uaySGU0813xd1+1XrN/CpY+NZsDIwsT7LFq7mUv6j2ZJC074ZRVVPPDBXH7wyEjyNDGLiLRASZN9eyDzLZZJ9rvFyqSF+XjOitSnTM1W23S34wvXcO5DwxtU27BwzSYufWw0y1pga/ZFazbxg0dG8K+P5qkPuIi0WEmT/b+BB8ysI4Rr+MCfgLfTCkwazt3rNS94riwp3sKKrBnw3J2nRxZycf/RrCxpeP/ywtUh4WfvvznNXraeHz06MrXpg0VEciVpsr8B2B8oBvYknNEfTD2u2ZtZDzObY2Z5ZpY9XS5mdqOZzTSzqVFDwINjyyrNbHJ0G5T0mDuqoXNWNlsCilflby6r5KaXp/DHQTNyUstQsGojFz82mhUlzZ/wxxeu4cd9R7GiET9gRESaSp3J3szaAxcClxIa550CHObuP3D3kiQHifbRBzgHOAa4xMyOyVptEtDN3Y8HXgXuiy3b7O4nRLfsMfolxt3515B5zXb8TLJfuHoTP3x0JK9PWpzT/Res3Milj41pVC1BYw2ZvZzLHh+T+pwDIiK5Umeyd/dK4AF33+LuK9x9nLsvq+dxugN57l7g7mXAQOCCrON87O6Zvkqjgc71PIYAo/JXN+v88pMXrWPonBWc9/BwZi1Np3Yhb8UGLnt8dLMMO/vahCKueWZCnT0JRERakqTV+G+b2XmNOE4nwqh7GUVRWU2uJrQTyNjZzMab2Wgz+351G5hZz2id8StXrmxEqK3bQ0Oa/lp93LjCNVw1YBzFm9Odb3zu8g1c/vgY1mxsuvHSH/+0gJtemaKGeCLS6iQd7mxn4FUzG0VI2p9927n7z3IZkJldDnQDvhUrPtjdF5vZocAQM5vm7vnx7dy9P9AfoFu3bjvkt/H4wjWMKljdrDGUVzbdUz97WQmXPjaaF685hb127ZDacdyd+96fw6ND8+teWUSkBUqa7KdHt4ZaTBhTP6NzVLYNM/sOcBvwLXf/rI7W3RdHfwvMbChwImFCHol5uBla4De32ctKuOzxMbxwzcl8YZfcJ/yKyipue2M6L41fVPfKIiItVNKJcO5u5HHGAUeYWVdCkr+Y0ODvM2Z2ItAP6OHuK2LlewGb3L3UzPYB/pttG+8JYajaoXN2zMsXM5eu56dPjKX3D47luE57EnqGNt6W8kquf3ES/5m5PCf7ExFpLolnLTGzM4BLgAMJ89gPdPePkmzr7hVmdh3wPmGAnifdfYaZ9QLGu/sg4G+EgXpeib6sF0Yt748G+plZFaGNwV/dfWbi/3AH8fDHzdcCvyWYtriY8x8ewZH77c5F3TpzwQmd2Hf3jg3e3/ot5Vzz9HjGzF+TwyhFRJqHeYJBvM3sJkKf+qeABYQueFcB97n731ONsAG6devm48ePb+4wmszc5SWc9eAnzR1Gi/K5dsbpR36Ji7p15oyjvsRO7ZO2RYWVJaVc8eRYZqbUm0BE5Of/3ZU7z8vugd44ZjbB3btVtyzpmf2NwBnu/tl1ezN7FvgAaHHJfkfTHKPltXQVVc6Hs5bz4azl7L1rB75/YicuPKkzRx+wR63bLVy9iZ8+OUYz1olIm5J88nHIzigFxFrlS/OYv2ojb09Z0txhtGirN5bxxPD5PDF8Psd22oOLTjqIC044cLsGfbOWrudnT45t1gF7RETSkDTZ3wU8YWZ3EfrIHwTcAfzRzD6rH3V3jTTSxB4dmoe6fSc3ffF6pi+eQe93Z3HmMftx4UmdOfWIfZi4cB1XPz2OEo2KJyJtUNJk3y/6ewnhbD7T3PmyaJlF5e1zGp3UqmjtJl6fmNvhaHcUZZVVvDttKe9OW8p+e3Rk3aZyjYonIm1W0mTfNdUopEH6DSto8mls26Ll61VtLyJtW9J+9gvSDkTqZ8X6LRroRUREEkneH0lalMc+LaBM1c4iIpKAkn0rtGZjGc+NXtjcYYiISCuhZN8KPTl8PpvLK5s7DBERaSXqlezN7CAzOyWtYKRuxZvLeXpkYXOHISIirUiiZG9mXcxsBDAb+DAqu9DMHk8zONneMyMLKSlVX3AREUku6Zl9P+BdYHegPCr7ADgzjaCkehtLK3hixPzmDkNERFqZpP3suwPnunuVmTmAuxeb2Z7phSbZnh+zgHWbyuteUUREJCbpmf1y4PB4gZkdA6hJeBPZUl7JY5/qrF5EROovabK/H3jHzK4CPmdmlwAvAfemFpls4+XxizRBi4iINEjSEfSeNLPVwC+BRcDPgDvc/c00g2vtNpVVcOdbMzDge189kP86bO96zaueUVZRRd+h+bkPUEREdgiJkr2ZnezubwFvZZV3d/exqUTWyi1cvYmez45n9rISAF6ZUMReu+xEj2MP4LzjD+DkQ/emfTurYy/BG5OKWFK8Jc1wRUSkDUvaQO8DYI9qyt8Dvpi7cNqGT+au5DcvTqJ487aN6dZuKufFsQt5cexC9tmtI989bn/O++qBnNRlL9rVkPgrKqt4RGf1IiLSCLUm+2iuegt3zdg6tS3AYYA6fMe4O32HFfC392fXOcf8qg2lPDNqAc+MWsD+e+zMuccfwPeOP4ATDvoC4akO3pm6lAWrN6UcuYiItGV1ndlXEOapz9yPqwJ65zyiVmpjaQW3vDqVd6ctrfe2y9Zv4Ynh83li+Hw67/V5zj3+AM47/kCOOWAPHv44L4VoRURkR1JXsu9KOJsfBpwWK3dgpbtvTiuw1qRw1UZ++ewE5iwvafS+itZupt+wAvoNK2C/PTpqrnUREWm0WpN9bB77g5sgllbp4zkr+O2Lk1i/JfdXNJToRUQkF5K2xn+mpmXu/rPchdN6uDuPDM3n/v/Mweu4Pi8iItKcknb6zs+6bQTOAdYkPZCZ9TCzOWaWZ2a3VrP8RjObaWZTzewjMzs4tuwKM5sX3a5Iesy0bCit4FfPTeRv7yvRi4hIy5d0UJ27s8vM7Angj0m2N7P2QB/CxDlFwDgzG+TuM2OrTQK6ufsmM/sVcB/wEzP7YnScboS2AhOibdcmOXauzV+1kZ7PjGfeig3NcXgREZF6q/9wbltNBr6VcN3uQJ67F7h7GTAQuCC+grt/7O6ZPmajgc7R/bOBD9x9TZTgPwB6NCLuBhsyeznnPzxciV5ERFqVpNfsz8gq2gW4GJhZzerV6UQYZjejCDi5lvWvBv5dy7adEh43J6qqnIc/zuPBD+eq2l5ERFqdpCPoPZH1eCPhzP6S3IYDZnY5oco+aa1BZrueQE+ALl265DSmMfPX8MAHc3O6TxERkaaS9Jp910YeZzFwUOxx56hsG2b2HeA24FvuXhrb9vSsbYdWE2N/oD9At27dcnr+7TqdFxGRVqzGa/Zm1i7JLeFxxgFHmFlXM+tAuAQwKOt4JwL9gPPdfUVs0fvAWWa2l5ntBZwVlYmIiEgCtZ3Zx4fKrY5Fy9vXdRB3rzCz6whJuj3wpLvPMLNewHh3HwT8DdgNeCUaG36hu5/v7mvM7E+EHwwAvdw9cZc/ERGRHV1tyb6xVffbcPfBwOCssjtj979Ty7ZPAk/mMh4REZEdRY3JPjZU7meiavv9gOXuXpVmYCIiIpIbia65m9ke0ZC5WwgN5jab2dNmtmeq0YmIiEijJW1g9y9gV+BY4PPAcYS+9v9KKS4RERHJkaT97HsAh8ZGuJtrZlcRxskXERGRFizpmf0WYN+ssn0AzcEqIiLSwiU9s38c+MDMHgAWEOa3v4FoEBsRERFpuZIm+97AEuBS4MDo/n2oO5yIiEiLl3S4XCckdiV3ERGRViZp17tLzOzo6P6XzWyYmX1sZkelG56IiIg0VtIGevcAmSFq/04YunYY8EgaQYmIiEjuJL1mv6+7LzeznYFvAhcC5cCq1CITERGRnEia7Fea2eGEwXTGuXupme1CmAxHREREWrCkyf5PwASgEvhJVPYdYEoaQYmIiEjuJG2NP8DMXo7uZ0bRG02Yl15ERERasKRn9gAdgHPNLNPPfrC7r00nLBEREcmVpF3vzgAKgeuBrwO/Aeab2bfTC01ERERyIemZ/cNAT3d/OVNgZhcBfQD1tRcREWnBkvazPxB4LavsDWD/3IYjIiIiuZY02T8L/Dqr7FfAM7kNR0RERHKtxmp8M/sU8OhhO+BaM7sFWAx0AvYjtMgXERGRFqy2a/aPZz1+LM1AREREJB01Jnt3f7opAxEREZF0JO5nb2b7Ad2BfYgNk+vumvZWRESkBUvaz/77QD7QC+hH6GffD/hp0gOZWQ8zm2NmeWZ2azXLTzOziWZWYWYXZi2rNLPJ0W1Q0mOKiIhI8jP7e4Cr3P0VM1vr7iea2VXAV5JsbGbtCX3yzwSKgHFmNsjdZ8ZWWwhcCdxczS42u/sJCWMVERGRmKRd77q4+ytZZU8DP0u4fXcgz90L3L0MGAhcEF/B3QvdfSpQlXCfIiIikkDSZL8iumYPUGhm3wAOA9on3L4TsCj2uCgqS2pnMxtvZqOjSwoiIiKSUNJq/MeAbxJG0XsQ+JhwBv73lOLKdrC7LzazQ4EhZjbN3fPjK5hZT6AnQJcuXZooLBERkZYv6RS398buP2NmQ4Fd3X1WwuMsBg6KPe4clSXi7oujvwXRsU8kNBiMr9Mf6A/QrVs3z96HiIjIjippNf423H1hPRI9wDjgCDPramYdgIuBRK3qzWwvM+sY3d8H+G9gZu1biYiISEaDkn19uXsFcB3wPjALeNndZ5hZLzM7H8DMvm5mRcBFQD8zmxFtfjQw3symEC4f/DWrFb+IiIjUIvGgOo3l7oOBwVlld8bujyNU72dvNxI4LvUARURE2qgmObMXERGR5qNkLyIi0sYp2YuIiLRxSvYiIiJtnJK9iIhIG1efKW73BI4EdouXu/uQXAclIiIiuZMo2ZvZlYRZ6zYAm2KLHDg092GJiIhIriQ9s+8NXOju/04zGBEREcm9pNfsPwf8J81AREREJB1Jk/29wO1mpgZ9IiIirUzSavwbgP2BW8xsdXyBu2s+WRERkRYsabK/PNUoREREJDVJ57MflnYgIiIiko769LM/ATgV2AewTHl85joRERFpeRI1uDOznsAI4Azgd4QpZ28CDk8vNBEREcmFpK3rbwF6uPsPgM3R3wuB8tQiExERkZxImuy/5O6fRverzKxdNMDOeSnFJSIiIjmS9Jp9kZkd4u6FwFzgAjNbBZSlFpmIiIjkRNJkfx9wNFAI9AJeBToA16cTloiIiORK0q53A2L3/21mewEd3H1DWoGJiIhIbiQe/tbM9jazn5rZLe5eBuxhZp1TjE1ERERyIGnXu28Bc4DLgDui4iOAR1OKS0RERHIk6Zn9P4CfuHsPoCIqGwN0TyUqERERyZmkyf4Qd/8ouu/R3zLqMQKfiIiINI+kyX6mmZ2dVfYdYFrSA5lZDzObY2Z5ZnZrNctPM7OJZlZhZhdmLbvCzOZFtyuSHlNERESSn5nfBLxjZu8CnzezfoQBdS5IsrGZtQf6AGcCRcA4Mxvk7jNjqy0ErgRuztr2i8AfgW6EWoUJ0bZrE8YuIiKyQ0t0Zu/uo4GvAjOAJ4H5QHd3H5fwON2BPHcviFryDyTrh4K7F7r7VKAqa9uzgQ/cfU2U4D8AeiQ8roiIyA4v8TV3d19MGFynIToBi2KPi4CTG7FtpwbGISIissNJlOzNbE/CaHknArvFl7n7WSnEVW/RzHw9Abp06dLM0YiIiLQcSc/sXwHaA28AmxtwnMXAQbHHnaOypNuenrXt0OyV3L0/0B+gW7dunr1cRERkR5U02Z8C7BNdb2+IccARZtaVkLwvBi5NuO37wJ+jIXoBzgJ+38A4REREdjhJu94NB45q6EHcvQK4jpC4ZwEvu/sMM+tlZucDmNnXzawIuAjoZ2Yzom3XAH8i/GAYB/SKykRERCSBpGf2VwKDzWwMsDy+wN17JdmBuw8GBmeV3Rm7P45QRV/dtk8SegGIiIhIPSVN9r0J19wLgT1i5bo2LiIi0sIlTfYXA19296VpBiMiIiK5l/SafQFQnmYgIiIiko6kZ/bPAoPM7CG2v2Y/JOdRiYiISM4kTfa/jv7+OavcgUNzF46IiIjkWqJk7+5d0w5ERERE0pH0mr2IiIi0Ukr2IiIibZySvYiISBunZC8iItLGJUr2ZnaWmX05q+xIMzsznbBEREQkV5Ke2fcBSrLKSqJyERERacGSJvsvVTNU7lJg/xzHIyIiIjmWeLhcMzsjq+x0YH5uwxEREZFcSzqC3l3A62b2BJAPHAZcFd1ERESkBUt0Zu/ubwFnAbsC50Z/z47KRUREpAVLemaPu48FxqYYi4iIiKSgxmRvZre5e+/ofq+a1nP3O9MITERERHKjtjP7zrH7B9WwjucwFhEREUlBjcne3X8FYGbtCPPZj3D30qYKTERERHKjzgZ67l4FvKVELyIi0jol7Wf/iZmdkmokIiIikoqkrfEXAP82s7eARcSu1auBnoiISMuW9Mz+88CbhCTfmdBgL3NLxMx6mNkcM8szs1urWd7RzF6Klo8xs0Oi8kPMbLOZTY5ufZMeU0RERBKe2bt7o0bKM7P2hElzzgSKgHFmNsjdZ8ZWuxpY6+6Hm9nFwL3AT6Jl+e5+QmNiEBER2VElneJ2TQ3lKxIepzuQ5+4F7l4GDAQuyFrnAuDp6P6rwLfNzBLuX0RERGqQtBp/p+wCM9sJaJ9w+06Ea/0ZRVFZteu4ewVQDOwdLetqZpPMbJiZnZrwmCIiIkId1fhm9inhOv3OZvZJ1uLOwMi0AotZCnRx99VmdhLwppl9xd3XZ8XaE+gJ0KVLlyYIS0REpHWo65r944ABXweeiJU7sBwYkvA4i9m2MV/nqKy6dYrM7HPAnsBqd3egFMDdJ5hZPvBlYHx8Y3fvD/QH6Natm0b2ExERidSa7N39aQAzG+3usxtxnHHAEWbWlZDULwYuzVpnEHAFMAq4EBji7m5m+wJr3L3SzA4FjgAKGhGLiIjIDiVpP/s5ZnYNcAmwj7sfb2anAfu7+8t1bezuFWZ2HfA+4Tr/k+4+I5pgZ7y7DyLUHDxrZnnAGsIPAoDTgF5mVg5UAde6e7UNBkVERGR7SZN9L0K3uX8AmX7uRcCDQJ3JHsDdBwODs8rujN3fAlxUzXavAa8ljFNERESyJG2NfyXwPXcfyNbR8+YDh6YRlIiIiORO0mTfHtgQ3c8k+91iZSIiItJCJU32g4EHzKwjQDTYuAJO+AAADAVJREFUzZ+At9MKTERERHIjabK/ETiAMNDNnoQz+oOB36UUl4iIiORI0rHx1wM/MLP9gC7AIndflmpkIiIikhNJz+wzNhP6ybczswPN7MAUYhIREZEcSnRmb2bfIYxOdzBhRL0MJ/n4+CIiItIMkp7ZPwH8mXC9fqfYrUNKcYmIiEiOJB1UZ2fgKXevTDMYERERyb2kZ/YPArdofnkREZHWJ+mZ/WuEce1/b2ar4gvcXaPoiYiItGBJk/2rwKfAK4QW+SIiItJKJE32XYET3b0qzWBEREQk95Jes38LOCPNQERERCQdSc/sOwKDzOxTYHl8gbv/LOdRiYiISM4kTfYzopuIiIi0MknHxr877UBEREQkHUnP7DGzg4GvEuax/4y7v5DroERERCR3ko6N/3vgDmAm23a9c0DJXkREpAVLemZ/E9DN3WemGYyIiIjkXtKud6uBwhTjEBERkZQkPbP/f0B/M/sHsCK+wN0X5jwqERERyZmkyb4DcBZwaVa55rMXERFp4ZJW4z8C/AHYA81nLyIi0qokTfafI8xnv8HdK+O3pAcysx5mNsfM8szs1mqWdzSzl6LlY8zskNiy30flc8zs7KTHFBERkeTJ/n7g1obOZ29m7YE+wDnAMcAlZnZM1mpXA2vd/XDgQeDeaNtjgIuBrwA9gEei/YmIiEgCSZP99cBdwAYzWxi/Jdy+O5Dn7gXuXgYMBC7IWucC4Ono/qvAt6MfFxcAA9291N3nA3nR/kRERCSBpA30Lm/kcToBi2KPi4CTa1rH3SvMrBjYOyofnbVtp+wDmFlPoCdAly5dGhnutg7eZ1duP/fonO5TRER2XF85cM8mPV7SZD/d3VenGkkjuXt/oD9At27dPJf77vSFz/OLUw/N5S5FRESaTNJq/IVm9tb/b+/OY+Ws6jCOfx9AIKHYQlnKXjaNWgIoApqqmKhRUAEDSJHNxKBCjQb/AIWw2gRRDCBGAlIDFIqAgFVRWRRBUYRCZU3ZBLvQspTKDpU+/nHOleFyZ+50Y+bOPJ9kcmfOO++8vzk5ub/3nPfMeyTtJ2l5ZuDPA7ZoeL15LRvyPZLWAEZTbubTzr4RERHRRLvJfjxwI3AMsEDSeZImLsNxbge2l7R1PVk4EJgx6D0zgMPq8/2AP9p2LT+wztbfGtge+McyHDsiIqKvtZXsbT9l+2zbHwQ+RLmL3sWSHpV0Sl0Rr9X+/wUmA38AHgAut31f3ffz9W0XAGMlPQwcDRxb970PuJyyCM/vgaOW5Sd/ERER/a7tJW4bjKuPdwJ3UibL3SXpdNunNdvJ9rXAtYPKTmh4/gqwf5N9pwBTliPWiIiIvtfuErfvo8zIPwh4kfITuR1tz63bTwXuBpom+4iIiOiMdnv2NwPTgf1tv+V6ue3H6iI5ERER0WXaTfbjbC9p9YbGIfmIiIjoHi0n6En6gKQJA4le0oaSLpH0T0nnShr19oQZERERy2u42fhnUibjDfgZ8C7KzWsmAKevorgiIiJiJRluGP89wC0AksZQFrKZYPtBSTOAW4EjV22IERERsSKG69mvAbxWn+8OLLD9IIDtOcCYVRhbRERErATD9ezvo/z2/XLKXe9uGNggaTPgP6sutOU3c+bMpyU9PsSmDYCn3+54RqjUVftSV+1LXbUvddW+1FXR9AZ3KnekbbKx3BL314CB14GJtmfXbUcDu9n+4sqNddWRdIftXTodx0iQumpf6qp9qav2pa7al7oaXsueve2/SNqSMinvQdvPN2z+LWVd+oiIiOhiw/7Ovib4mUOUz14lEUVERMRK1e6qd73ivE4HMIKkrtqXumpf6qp9qav2pa6G0fKafURERIx8/dazj4iI6Dt9kewlfVrSbEkPSzq20/F0M0mPSbpH0ixJd3Q6nm4jaaqkJyXd21C2vqTrJT1U/67XyRi7RZO6OknSvNq+Zknas5MxdgtJW0j6k6T7Jd0n6Zu1PG1rkBZ1lbbVQs8P40taHXgQ+CQwF7gdmGT7/o4G1qUkPQbsYju/WR2CpI8CLwAX2Z5Qy04HFtk+rZ5Mrmf7mE7G2Q2a1NVJwAu2f9jJ2LqNpE2ATWzfKWldyqTofYDDSdt6kxZ1dQBpW031Q89+V+Bh24/afo3yc8G9OxxTjFC2bwYWDSreG7iwPr+Q8o+n7zWpqxiC7Sds31mfPw88AGxG2tZbtKiraKEfkv1mwJyG13NJw2jFwHWSZko6otPBjBAb236iPl8AbNzJYEaAyZLursP8fT8sPZik8cDOwG2kbbU0qK4gbaupfkj2sWwm2n4/ZdGjo+pQbLTJ5bpYb18bWzE/BbYFdgKeAM7obDjdpS4b/kvgW7afa9yWtvVmQ9RV2lYL/ZDs5wFbNLzevJbFEGzPq3+fBK6mXAaJ1hbW64gD1xOf7HA8Xcv2Qtuv214KnE/a1/9JegcleV1i+6panLY1hKHqKm2rtX5I9rcD20vaWtKalAV9ZnQ4pq4kaZ064QVJ6wCfAu5tvVdQ2tNh9flhwK86GEtXG0hc1b6kfQEgScAFwAO2f9SwKW1rkGZ1lbbVWs/PxgeoP8E4E1gdmGp7SodD6kqStqH05qHcSvnS1NWbSZoO7EFZZWshcCJwDWVlyC2Bx4EDbPf9xLQmdbUHZZjVwGPAVxuuSfetuujYLcA9wNJa/F3Ktei0rQYt6moSaVtN9UWyj4iI6Gf9MIwfERHR15LsIyIielySfURERI9Lso+IiOhxSfYRERE9Lsk+IiKixyXZR4xAdSnilyW90PDYtNNxDUfSTZK+0mL7eElu+E4LJf1G0iffzjgjek2SfcTI9Tnboxoe85dlZ0lrrKrAVoIxtkcBOwLXA1dLOryzIUWMXEn2ET1E0lqSzpQ0vz7OlLRW3baHpLmSjpG0APi5pNUkHSvpEUnPSLpc0voNnzdR0q2SFkuaM5BwJe0l6S5Jz9Xykxr2WVvStPp5iyXdLmljSVOAjwDn1F77OcN9H9sLbJ8FnAR8X9Jq9RgDMT8v6X5J+9byNSUtkrRDQzwbSXpJ0oYrXsMRI1OSfURvOQ7YnXLb0B0pi4Ec37B9HLA+sBVwBPANyhrpHwM2BZ4FfgIgaSvgd8CPgQ3rZ86qn/MicCgwBtgL+LqkgbXWDwNGUxagGgt8DXjZ9nGU25xOriMRk5fhe10FbAS8u75+hHLiMBo4GZgmaRPbrwGXAQc37DsJuNH2U8twvIiekmQfMXJdU3vOiyVdU8u+BJxi+8ma3E4GDmnYZylwou1Xbb9MScTH2Z5r+1VKD3q/OsR/EHCD7em2l9h+xvYsANs32b7H9lLbdwPTKScMAEsoSX67ugrZzMHLtS6HgUsU69fjX2F7fj3+L4CHeGOVswuBSXXBFOr3v3gFjx8xonXzNbuIaG0f2zcMKtuUsmDKgMdr2YCnbL/S8HoryvXwpQ1lrwMbU3rmjwx1YEm7AacBE4A1gbWAK+rmi+u+l0kaA0yjnFAsWYbvNthm9e+ievxDgaOB8bV8FGXBHWzfJuklYA9JTwDbkZUuo8+lZx/RW+ZTEviALXmjVwxlRbBGc4DP2B7T8Fjb9ry6bdsmx7mUkkC3sD0aOBcQQB0FONn2e4EPA5+lDPkPdfx27UtZy312vbxwPjAZGGt7DGU5UzW8/0LKUP4hwJWDTnAi+k6SfURvmQ4cL2lDSRsAJ1B61s2cC0ypCZS639512yXAJyQdIGkNSWMl7VS3rQsssv2KpF0pQ/7Uz/i4pB0krQ48RxnWHxg5WAhs0+6XqRP7JlOWx/2O7aXAOpSThqfqe75MGWFoNI1ygnAwcFG7x4voVUn2Eb3le8AdwN2U9b7vrGXNnEXpoV8n6Xng78BuALb/DewJfJsyfD6LMukP4EjglLrPCZQ11weMA66kJPoHgD/zxjXzsyhzAp6VdHaLuBZLerF+hz2B/W1PrXHdD5wB/I1y8rAD8NfGnW3Pqd/dlEmBEX0t69lHRE+SNBWYb/v4Yd8c0eMyQS8ieo6k8cAXgJ07G0lEd8gwfkT0FEmnUibs/cD2vzodT0Q3yDB+REREj0vPPiIioscl2UdERPS4JPuIiIgel2QfERHR45LsIyIielySfURERI/7H5Uk5ii2wTYnAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### R² score and explained variance score vs forecast day"
      ],
      "metadata": {
        "id": "vaBKFjHknP0V"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "daily_evs = daily_metrics(daily_prediction, metrics.explained_variance_score)\n",
        "daily_r2 = daily_metrics(daily_prediction, metrics.r2_score)\n",
        "\n",
        "# Plotting\n",
        "x = range(1, len(daily_r2)+1)\n",
        "plt.plot(x, daily_evs, '--')\n",
        "plt.plot(x, daily_r2, '--o')\n",
        "plt.title('R² score and explained variance score vs forecast day', fontsize=14)\n",
        "plt.xlabel(\"Forecast Day\", fontsize=12)\n",
        "plt.legend(['explained_variance_score', 'r2_score'])\n",
        "plt.grid();"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 354
        },
        "id": "jDPOrDwenAME",
        "outputId": "0c50a450-1968-4a8c-aeef-039feaa2880e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Scatter of forecasting accuracy (sMAPE) vs traffic volume"
      ],
      "metadata": {
        "id": "49NSxHVcX32G"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "cell_traffic = []\n",
        "cell_smape = []\n",
        "\n",
        "for cell in daily_prediction.Cell.unique():\n",
        "  temp = daily_prediction[daily_prediction['Cell']==cell].copy()\n",
        "  cell_traffic.append(temp.y.sum())\n",
        "  cell_smape.append(smape_metric(temp.y, temp.yhat))\n",
        "\n",
        "# Plotting\n",
        "plt.scatter(np.array(cell_traffic)/1024/1024, cell_smape, marker='+')\n",
        "plt.title('Accuracy (sMAPE) vs traffic volume per cell', fontsize=14)\n",
        "plt.xlabel('Traffic per cell in 28 days, GB', fontsize=12)\n",
        "plt.ylabel('Symmetric mean absolute precenteage error', fontsize=12);"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 354
        },
        "id": "KexE3tkIYaK7",
        "outputId": "8c7edd36-9e63-4c78-bab8-8c3162c92a0d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## III. Hourly Traffic Trend"
      ],
      "metadata": {
        "id": "Hqnb7WtWlH_a"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Getting Data"
      ],
      "metadata": {
        "id": "5NqRYnIhgnu4"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "In the second step, using the available 28 days of hourly data"
      ],
      "metadata": {
        "id": "MPZgxRX9Q-B6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Opening daily cell data\n",
        "hourly_traffic = pd.read_csv('df_hourly.csv', sep=';')"
      ],
      "metadata": {
        "id": "3jlItFoognvE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculating PRB usage\n",
        "hourly_traffic['PRB_Util_Qty_True'] = hourly_traffic['PRB_Utilization_Percent']*hourly_traffic['PRB_Available_DL']\n",
        "hourly_traffic['Datetime'] = pd.to_datetime(hourly_traffic['Datetime'])"
      ],
      "metadata": {
        "id": "4pcPk0M8P_ay"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Renaming columns for Prophet standard, where 'ds' its the date\n",
        "# and 'y' it's a target variable Data_Volume(GB)\n",
        "hourly_traffic.columns = ['Cell', 'ds', 'y', 'PRB_Available_DL', 'PRB_Utilization_Percent', 'PRB_Util_Qty_True']"
      ],
      "metadata": {
        "id": "gNgnEY-WgnvE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Checking cell quantity and data range\n",
        "list_cells = hourly_traffic.Cell.unique()\n",
        "print(f'Dataset have {len(list_cells)} unique cells',\n",
        "      f'with date between {min(hourly_traffic.ds)} and {max(hourly_traffic.ds)}.')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "24a434e3-d49f-4e5d-ac0d-bc087d7d9486",
        "id": "8BvAeAZkgnvF"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Dataset have 163 unique cells with date between 2019-01-01 00:00:00 and 2021-01-28 23:00:00.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Splitting dataset\n",
        "# We will use the same period from daily traffic, except train_ini witch we will use last 30 days\n",
        "train_ini = datetime.strptime(train_end, \"%Y-%m-%d\") - timedelta(days=30)\n",
        "n_hour_forecast = (val_end - datetime.strptime(val_ini, \"%Y-%m-%d\")).days * 7\n",
        "\n",
        "df_hourly_train = hourly_traffic[(hourly_traffic['ds'] >= train_ini) & (hourly_traffic['ds'] < train_end)].copy()\n",
        "df_hourly_val = hourly_traffic[(hourly_traffic['ds'] >= val_ini) & (hourly_traffic['ds'] < val_end)].copy()\n",
        "\n",
        "df_hourly_train.head()"
      ],
      "metadata": {
        "id": "ijUqjvzkgnvF",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "756cb3de-9640-4ee2-fc02-826644c67483"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "             Cell                  ds             y  PRB_Available_DL  \\\n",
              "1760928  cell0001 2020-06-01 00:00:00  1.158804e+06             100.0   \n",
              "1760929  cell0001 2020-06-01 01:00:00  4.515076e+05             100.0   \n",
              "1760930  cell0001 2020-06-01 02:00:00  6.314205e+05             100.0   \n",
              "1760931  cell0001 2020-06-01 03:00:00  1.551132e+05             100.0   \n",
              "1760932  cell0001 2020-06-01 04:00:00  1.360913e+04             100.0   \n",
              "\n",
              "         PRB_Utilization_Percent  PRB_Util_Qty_True  \n",
              "1760928                 0.096667           9.666667  \n",
              "1760929                 0.051667           5.166667  \n",
              "1760930                 0.053333           5.333333  \n",
              "1760931                 0.050000           5.000000  \n",
              "1760932                 0.050000           5.000000  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-dbaf99f6-bf40-4017-9eef-0d6e97161d7d\">\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>Cell</th>\n",
              "      <th>ds</th>\n",
              "      <th>y</th>\n",
              "      <th>PRB_Available_DL</th>\n",
              "      <th>PRB_Utilization_Percent</th>\n",
              "      <th>PRB_Util_Qty_True</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1760928</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>2020-06-01 00:00:00</td>\n",
              "      <td>1.158804e+06</td>\n",
              "      <td>100.0</td>\n",
              "      <td>0.096667</td>\n",
              "      <td>9.666667</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1760929</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>2020-06-01 01:00:00</td>\n",
              "      <td>4.515076e+05</td>\n",
              "      <td>100.0</td>\n",
              "      <td>0.051667</td>\n",
              "      <td>5.166667</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1760930</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>2020-06-01 02:00:00</td>\n",
              "      <td>6.314205e+05</td>\n",
              "      <td>100.0</td>\n",
              "      <td>0.053333</td>\n",
              "      <td>5.333333</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1760931</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>2020-06-01 03:00:00</td>\n",
              "      <td>1.551132e+05</td>\n",
              "      <td>100.0</td>\n",
              "      <td>0.050000</td>\n",
              "      <td>5.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1760932</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>2020-06-01 04:00:00</td>\n",
              "      <td>1.360913e+04</td>\n",
              "      <td>100.0</td>\n",
              "      <td>0.050000</td>\n",
              "      <td>5.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-dbaf99f6-bf40-4017-9eef-0d6e97161d7d')\"\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-dbaf99f6-bf40-4017-9eef-0d6e97161d7d 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-dbaf99f6-bf40-4017-9eef-0d6e97161d7d');\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": 82
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Note:** As we can se on topic *B. LTE network level throughput to traffic relation*, one of the factors related to the occupation of cell resources is the volume of traffic generated by users. In our studies, we verified that the behavior of this volume throughout the day can be differentiated between weekdays and weekends.\n",
        "\n",
        "In order to better capture this difference, two conditional columns were included in the dataset, one indicating weekdays and the other indicating weekends. Two custom conditional seasonalities were also added to the algorithm to reflect this difference."
      ],
      "metadata": {
        "id": "P_M5OGb_uQbj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Weekend function\n",
        "def is_weekend(ds):\n",
        "    date = pd.to_datetime(ds)\n",
        "    return (date.dayofweek == 5 or date.dayofweek == 6)"
      ],
      "metadata": {
        "id": "82IrFn0bv10N"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Creating weekend/weekday columns\n",
        "df_hourly_train['weekend'] = df_hourly_train['ds'].apply(is_weekend)\n",
        "df_hourly_train['weekday'] = ~df_hourly_train['ds'].apply(is_weekend)\n",
        "df_hourly_val['weekend'] = df_hourly_val['ds'].apply(is_weekend)\n",
        "df_hourly_val['weekday'] = ~df_hourly_val['ds'].apply(is_weekend)"
      ],
      "metadata": {
        "id": "sIM5NPtywqTm"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Forecasting"
      ],
      "metadata": {
        "id": "A8LyOtgnsQT0"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Notes:**\n",
        "1. As we're only interested in hourly traffic distribution between weekdays and weekends, we will use **growth as flat** and choose two days forecast, been one for weekday and one for weekend.\n",
        "2. In this step we also add more than one configuration to check which one fits best in hourly distribution."
      ],
      "metadata": {
        "id": "sfm1z2ctzS18"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Parameters\n",
        "hourly_params = [\n",
        "                 # Config01\n",
        "                 [{'growth': 'flat'},                                                     # Trend growth\n",
        "                  {'fourier_order': 8, 'mode': 'multiplicative', 'prior_scale': 0.1}],    # Daily Hyperparameters\n",
        "\n",
        "                 # Config02\n",
        "                 [{'growth': 'flat'},\n",
        "                  {'fourier_order': 8, 'mode': 'multiplicative', 'prior_scale': 0.05}],\n",
        "\n",
        "                 # Config03\n",
        "                 [{'growth': 'flat'},\n",
        "                  {'fourier_order': 8, 'mode': 'multiplicative', 'prior_scale': 0.01}]\n",
        "]"
      ],
      "metadata": {
        "id": "YU9NujzFsQUO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "models_hourly = pd.DataFrame()\n",
        "\n",
        "for cell in list_cells[:]:\n",
        "\n",
        "  n_config = 1\n",
        "  # Filtering cell\n",
        "  df_train_temp = df_hourly_train[df_hourly_train['Cell']==cell].copy()\n",
        "  df_val_temp = df_hourly_val[df_hourly_val['Cell']==cell].copy()\n",
        "\n",
        "  # Applying Box-Cox\n",
        "  df_train_temp['y'], lambda_prophet = stats.boxcox(df_train_temp['y'])\n",
        "\n",
        "  # Executing all configurations\n",
        "  for params in hourly_params:\n",
        "    config = 'conf_' + str(n_config).zfill(len(str(len(hourly_params))))\n",
        "\n",
        "    # Instantiating the model\n",
        "    m = Prophet(yearly_seasonality=False,\n",
        "                weekly_seasonality=False,\n",
        "                daily_seasonality=False,\n",
        "                uncertainty_samples=0,      # Zeroed to speed up forecast\n",
        "                **params[0])\n",
        "\n",
        "    # Daily seasonality\n",
        "    m.add_seasonality(name='daily_weekend', period=1, condition_name='weekend', **params[1])\n",
        "    m.add_seasonality(name='daily_weekday', period=1, condition_name='weekday', **params[1])\n",
        "\n",
        "    # Fit\n",
        "    m.fit(df_train_temp)\n",
        "\n",
        "    # Create future dataset\n",
        "    date_range = pd.date_range(start='2017-01-6', periods=48, freq='H')\n",
        "    future = pd.DataFrame(date_range, columns=['ds'])\n",
        "\n",
        "    # Creating conditional columns for future dataset\n",
        "    future['weekend'] = future['ds'].apply(is_weekend)\n",
        "    future['weekday'] = ~future['ds'].apply(is_weekend)\n",
        "\n",
        "    # Forecast\n",
        "    forecast = m.predict(future)\n",
        "\n",
        "    # Applying inverse Box-Cox in forecast data\n",
        "    forecast['yhat'] = inverse_boxcox(forecast['yhat'], lambda_prophet)\n",
        "\n",
        "    # Adding information columns\n",
        "    forecast['Cell'] = cell\n",
        "    forecast['weekday'] = ~forecast['ds'].apply(is_weekend)\n",
        "    forecast['hour'] = forecast['ds'].dt.hour\n",
        "    forecast['configuration'] = config\n",
        "\n",
        "    models_hourly = pd.concat([models_hourly,\n",
        "                               forecast[['Cell', 'weekday', 'hour', 'yhat', 'configuration']]])\n",
        "\n",
        "    n_config += 1"
      ],
      "metadata": {
        "id": "jFqUef1osQUO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "As we can see below, there are differences in behavior between weekdays and weekend. We will use this in the traffic distribution."
      ],
      "metadata": {
        "id": "U-mQl8oVHeBt"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "m.plot_components(forecast);"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 657
        },
        "id": "TrX_nQinYOpN",
        "outputId": "b3270d5f-905c-43f9-904f-58d32a5921ca"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 648x648 with 3 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculating cell proportional hourly traffic per day and configuration\n",
        "models_hourly['yhat%'] = models_hourly['yhat'] / models_hourly.groupby(['Cell', 'weekday', 'configuration'])['yhat'].transform('sum')\n",
        "models_hourly.drop(columns='yhat', inplace=True)"
      ],
      "metadata": {
        "id": "6sy8SzoMVUNH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Peak daily traffic"
      ],
      "metadata": {
        "id": "pBrbl71jSh9b"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "To obtain peak daily traffic, authors suggest proportionally distributing the hourly prediction values within the predicted 30 days daily traffic. According to them, this method accurately follows real traffic pattern but removes random peaks or decreases and allows to forecast traffic volume for each hour."
      ],
      "metadata": {
        "id": "utyyFsOxTf5U"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Getting the quantity and configurations name\n",
        "n_conf_hourly = len(models_hourly.configuration.unique())\n",
        "config_name = models_hourly.configuration.unique()\n",
        "\n",
        "# Creating hourly dataset\n",
        "final_result_hourly = daily_prediction.drop('y', axis=1)\n",
        "\n",
        "# Creating hourly column for each cell and configuration\n",
        "hours_24 = cycle(list(range(0, 24)))\n",
        "configs = cycle(np.repeat(config_name, 24))\n",
        "\n",
        "final_result_hourly = pd.DataFrame(np.repeat(final_result_hourly.values, 24 * n_conf_hourly, axis=0),\n",
        "                                   columns=final_result_hourly.columns)\n",
        "final_result_hourly['hour'] = [next(hours_24) for hour in range(len(final_result_hourly))]\n",
        "final_result_hourly['configuration'] = [next(configs) for config in range(len(final_result_hourly))]\n",
        "\n",
        "# Adding weekday status\n",
        "final_result_hourly['weekday'] = ~final_result_hourly['ds'].apply(is_weekend)\n",
        "final_result_hourly.rename(columns={'yhat':'daily_data_volume'}, inplace=True)"
      ],
      "metadata": {
        "id": "aXibxWtZFcbT"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "We joined daily and hour dataframe considering cell, hour, configuration and weekend/weekday status."
      ],
      "metadata": {
        "id": "p-ebJS-20Y5I"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Merging daily prediction with hourly prediction\n",
        "final_result_hourly = final_result_hourly.merge(models_hourly,\n",
        "                                                on=['Cell', 'weekday', 'hour', 'configuration'],\n",
        "                                                how='left')\n",
        "\n",
        "# Distributing the hourly prediction values within the predicted daily traffic\n",
        "final_result_hourly['yhat'] = final_result_hourly['daily_data_volume'] * final_result_hourly['yhat%']\n",
        "\n",
        "final_result_hourly.rename(columns={'ds':'day'}, inplace=True)\n",
        "\n",
        "# Creating timestamp\n",
        "final_result_hourly['ds'] = final_result_hourly['day'] + final_result_hourly.hour.astype('timedelta64[h]')"
      ],
      "metadata": {
        "id": "QadpqdvD1qYY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "final_result_hourly = pd.merge(final_result_hourly[['Cell', 'day', 'ds', 'yhat',\n",
        "                                                    'configuration', 'daily_data_volume', 'yhat%']],\n",
        "                               df_hourly_val.drop(['weekend', 'weekday'], axis=1),\n",
        "                               left_on=['Cell', 'ds'],\n",
        "                               right_on=['Cell', 'ds'],\n",
        "                               how='inner')\n",
        "final_result_hourly = final_result_hourly.sort_values(by = ['Cell', 'configuration', 'ds']).copy()"
      ],
      "metadata": {
        "id": "2-fId3AEIAZs"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Evaluating"
      ],
      "metadata": {
        "id": "patg6T-g0BsT"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Note:** Authors didn't perform precision measurements in this phase of study, nevertheless, I've inserted the measurements presented in the previous topic, with this, it's possible to choose the model configuration that best fit the validation data."
      ],
      "metadata": {
        "id": "i6SM2p7KcJU9"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### LTE Traffic Forecasting Accuracy Overall Result"
      ],
      "metadata": {
        "id": "A4xRRpnq0E_g"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Eveluating results for each configuration\n",
        "evaluating_cols = metric_cols + ['Configuration']\n",
        "overall_result_hourly = pd.DataFrame(columns= evaluating_cols)\n",
        "full_eval = []\n",
        "hourly_configurations = final_result_hourly.configuration.unique()\n",
        "\n",
        "for configuration in hourly_configurations:\n",
        "  temp_df = final_result_hourly[final_result_hourly['configuration']==configuration].copy()\n",
        "\n",
        "  temp_df.dropna(inplace=True)   # Delete this line\n",
        "\n",
        "  temp_eval = timeseries_evaluation_metrics_func(temp_df.y, temp_df.yhat)\n",
        "  temp_eval.append(configuration)\n",
        "  full_eval.append(temp_eval)\n",
        "\n",
        "overall_result_hourly = pd.concat([overall_result_hourly,\n",
        "                                  pd.DataFrame(full_eval, columns= evaluating_cols)])\n",
        "overall_result_hourly"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "outputId": "29ca8e9f-2654-43d7-c0e7-1bb61733c056",
        "id": "PattQlU90E_h"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      sMAPE       EVS        R2 Configuration\n",
              "0  0.478260  0.820387  0.819381        conf_1\n",
              "1  0.482540  0.820672  0.819665        conf_2\n",
              "2  0.631291  0.642894  0.641886        conf_3"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-d9dcbb54-edc9-4477-9dc7-93867e11a9f9\">\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>sMAPE</th>\n",
              "      <th>EVS</th>\n",
              "      <th>R2</th>\n",
              "      <th>Configuration</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.478260</td>\n",
              "      <td>0.820387</td>\n",
              "      <td>0.819381</td>\n",
              "      <td>conf_1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.482540</td>\n",
              "      <td>0.820672</td>\n",
              "      <td>0.819665</td>\n",
              "      <td>conf_2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.631291</td>\n",
              "      <td>0.642894</td>\n",
              "      <td>0.641886</td>\n",
              "      <td>conf_3</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d9dcbb54-edc9-4477-9dc7-93867e11a9f9')\"\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-d9dcbb54-edc9-4477-9dc7-93867e11a9f9 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-d9dcbb54-edc9-4477-9dc7-93867e11a9f9');\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": 93
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Eveluating results for each configuration\n",
        "evaluating_cols_1 = metric_cols + ['Configuration', 'cell']\n",
        "individual_result_hourly = pd.DataFrame(columns=evaluating_cols_1)\n",
        "ind_eval = []\n",
        "hourly_configurations = final_result_hourly.configuration.unique()\n",
        "cells_eval = final_result_hourly['Cell'].unique()\n",
        "\n",
        "for cell in cells_eval:\n",
        "  temp_cell_df = final_result_hourly[final_result_hourly['Cell']==cell]\n",
        "\n",
        "  for configuration in hourly_configurations:\n",
        "    temp_df = temp_cell_df[temp_cell_df['configuration']==configuration]\n",
        "    temp_eval = timeseries_evaluation_metrics_func(temp_df.y, temp_df.yhat)\n",
        "    temp_eval.append(configuration)\n",
        "    temp_eval.append(cell)\n",
        "    ind_eval.append(temp_eval)\n",
        "\n",
        "individual_result_hourly = pd.concat([individual_result_hourly,\n",
        "                                  pd.DataFrame(ind_eval, columns= evaluating_cols_1)])"
      ],
      "metadata": {
        "id": "X8JO6cbiORKC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Note:** After run all desired configurations, pick one that best performed. This one will be used further, in the final solution."
      ],
      "metadata": {
        "id": "5zYj-opd0E_i"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "hourly_prediction = final_result_hourly[final_result_hourly['configuration']=='conf_2'].copy().reset_index(drop=True)\n",
        "hourly_prediction.drop(['configuration'], axis=1, inplace=True)\n",
        "\n",
        "# Calculating daily traffic %\n",
        "hourly_prediction['y%'] = hourly_prediction['y'] / hourly_prediction.groupby(['Cell', 'day'])['y'].transform('sum')"
      ],
      "metadata": {
        "id": "8BQatL9L0E_j"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### LTE cell traffic  compared to traffic trend modeled"
      ],
      "metadata": {
        "id": "shLVRwk32AQQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "example_cell = hourly_prediction[hourly_prediction['Cell']=='cell0012']\n",
        "example_cell = example_cell.head(24*7)\n",
        "\n",
        "plt.plot(example_cell.ds, example_cell['yhat%'], '--', label='Trend')\n",
        "plt.plot(example_cell.ds, example_cell['y%'], label='Real Traffic Proportion')\n",
        "\n",
        "plt.title('Proportional traffic comparison', fontsize=14)\n",
        "plt.xlabel(\"Time\", fontsize=10)\n",
        "plt.ylabel(\"Proportion of daily traffic per hour\", fontsize=10)\n",
        "plt.xticks(rotation=90)\n",
        "plt.legend()\n",
        "plt.grid();"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 398
        },
        "id": "ubR-jtOaLZGD",
        "outputId": "9b56a936-8f98-4bb9-eab5-19ba3692059d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## IV. LTE Cell Capacity Evaluation"
      ],
      "metadata": {
        "id": "jmU-L6m2l7Z-"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Espectral efficiency"
      ],
      "metadata": {
        "id": "U5ci8-Qa3kV4"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Each cell has a unique behavior regard to reaching the threshold. This is defined by the spectral efficiency. The authors suggest to evaluate the cells individually and, through linear regression, verify the traffic demand that each cell supports before reaching the PRB utilization threshold."
      ],
      "metadata": {
        "id": "PyKmiDntI6h8"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Example of LTE cells with different spectral efficiency"
      ],
      "metadata": {
        "id": "yGjE5Rzk3PvH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "cell_a, cell_b = 'cell0015', 'cell0020'\n",
        "\n",
        "cell_a_x = df_hourly_train[df_hourly_train['Cell']==cell_a].PRB_Utilization_Percent*100\n",
        "cell_a_y = (df_hourly_train[df_hourly_train['Cell']==cell_a].y)/1024/1024\n",
        "cell_b_x = df_hourly_train[df_hourly_train['Cell']==cell_b].PRB_Utilization_Percent*100\n",
        "cell_b_y = (df_hourly_train[df_hourly_train['Cell']==cell_b].y)/1024/1024\n",
        "\n",
        "plt.plot(cell_a_x, cell_a_y, 'o', mfc='none', label='Cell1', c='k')\n",
        "m, b = np.polyfit(cell_a_x, cell_a_y, 1)\n",
        "plt.plot(cell_a_x, m*cell_a_x+b, c='k')\n",
        "\n",
        "plt.plot(cell_b_x, cell_b_y, '^', mfc='none', label='Cell2', c='g')\n",
        "m, b = np.polyfit(cell_b_x, cell_b_y, 1)\n",
        "plt.plot(cell_b_x, m*cell_b_x+b, c='g')\n",
        "\n",
        "plt.title('Cells with different spectral efficiency', fontsize=14)\n",
        "plt.xlabel(\"Average DL PRB Usage per Cell, %\", fontsize=10)\n",
        "plt.ylabel(\"Downlink traffic volume, GB\", fontsize=10)\n",
        "plt.xticks(rotation=90)\n",
        "plt.legend(title=\"Cell Name\", bbox_to_anchor=(1, 0.5))\n",
        "plt.grid();"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 353
        },
        "id": "ZI9KdME-SOpy",
        "outputId": "c31676ca-a3c2-4422-b1d7-8b7cd9a6b60c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Calculation of spectral efficiency for each cell using regression"
      ],
      "metadata": {
        "id": "vtlhj2pYjk8J"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "prb_util_prediction = pd.DataFrame()\n",
        "cell_coefficients = []\n",
        "\n",
        "for cell in list_cells[:]:\n",
        "  df_train_temp = df_hourly_train[df_hourly_train['Cell']==cell].copy()\n",
        "\n",
        "  # Regression between Data Volume and PRB_Utilization_Qty on train dataset\n",
        "  x = df_train_temp['y']\n",
        "  y = df_train_temp['PRB_Util_Qty_True']\n",
        "  var_lin = np.polyfit(x, y, 1) # Getting the coefficients\n",
        "  cell_coefficients.append([cell, var_lin[0], var_lin[1]])\n",
        "\n",
        "  # Calculating PRB usage\n",
        "  df_test_temp = hourly_prediction[hourly_prediction['Cell']==cell].copy()\n",
        "  df_test_temp['PRB_Util_Qty_Pred'] = np.array([var_lin[0]*val+var_lin[1] for val in df_test_temp.yhat])\n",
        "\n",
        "  prb_util_prediction = pd.concat([prb_util_prediction, df_test_temp])\n",
        "\n",
        "prb_util_prediction['PRB_Util_Perc_Pred'] = prb_util_prediction['PRB_Util_Qty_Pred']/prb_util_prediction['PRB_Available_DL']"
      ],
      "metadata": {
        "id": "rctHdsj8gDvW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Congestion Threshold"
      ],
      "metadata": {
        "id": "I5l2JIScSadb"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Authors found a strong correlation between the utilization of PRBs (physical resource block) and average user throughput. They suggest that hourly samples that had 20 or fewer unused PRBs were considered congested:\n",
        "\n",
        "> *LTE 800MHz cell with 10MHz bandwidth can be considered as loaded if utilization is 60% or 20 PRB unused, and LTE 1800MHz with 20MHz bandwidth PRB utilization is above 80% (20 PRB left)* - Chmieliauskas & Guršnys, 2020\n",
        "\n",
        "To simplify our calculations, we adopt as congested hourly samples over 70% PRB utilization, as shown below.\n",
        "\n",
        "|Bandwidth (MHz)|PRB Available|Threshold (70% usage)|PRB unused|\n",
        "|:---: |:---: |:---: |:---: |\n",
        "|1,4|\t6\t|4\t|2  |\n",
        "|3\t|15\t|10\t|5  |\n",
        "|5\t|25\t|17\t|8  |\n",
        "|10\t|50\t|35\t|15 |\n",
        "|15\t|75\t|52\t|23 |\n",
        "|20\t|100|70\t|30 |"
      ],
      "metadata": {
        "id": "mobFlCFz1s-M"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculating congestion samples\n",
        "cong_threshold = 0.6\n",
        "\n",
        "prb_util_prediction['Cong_True'] = prb_util_prediction.PRB_Utilization_Percent >= cong_threshold\n",
        "prb_util_prediction['Cong_Pred'] = prb_util_prediction.PRB_Util_Qty_Pred / prb_util_prediction.PRB_Available_DL >= cong_threshold"
      ],
      "metadata": {
        "id": "ocwIdU2mXG-A"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Renaming and organizing columns\n",
        "prb_util_prediction.rename(columns={'yhat':'daily_data_vol_pred',\n",
        "                                    'yhat%': 'daily_data_vol_pred%',\n",
        "                                    'y': 'daily_data_vol_true',\n",
        "                                    'y%': 'daily_data_vol_true%',\n",
        "                                    'PRB_Utilization_Percent': 'PRB_Util_%_True'}, inplace=True)\n",
        "\n",
        "prb_util_prediction['weekend'] = prb_util_prediction['ds'].apply(is_weekend)\n",
        "\n",
        "prb_util_prediction = prb_util_prediction[['Cell', 'day', 'ds', 'weekend', 'daily_data_vol_true',\n",
        "                                           'daily_data_vol_true%', 'daily_data_vol_pred',\n",
        "                                           'daily_data_vol_pred%', 'PRB_Available_DL',\n",
        "                                           'PRB_Util_Qty_True', 'PRB_Util_%_True',\n",
        "                                           'PRB_Util_Qty_Pred', 'PRB_Util_Perc_Pred',\n",
        "                                           'Cong_True', 'Cong_Pred']]"
      ],
      "metadata": {
        "id": "HIi_vm0ztf_P"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Evaluating"
      ],
      "metadata": {
        "id": "0obrIn853JSe"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The ultimate purpose of this study is to identify the cells that will show the highest congestion sampling. As mentioned before, there are cells that have greater use characteristics on weekdays, such as cells in the city center and others with greater use on weekends, such as cells in residential neighborhoods. As a filter, we will consider as congested the cells that present an average of 1 or more daily samples of congestion, as follows:\n",
        "\n",
        "$\n",
        "Congestioned \\ cell =   \n",
        "  \\begin{cases}\n",
        "    1  & \\quad \\text{for weekdays } n \\text{ cong. samples} \\geq n \\text{ weekdays or}    \\\\\n",
        "    & \\quad \\text{weekends } n \\text{ cong. samples} \\geq n \\text{ weekends}    \\\\\n",
        "    0  & \\quad \\text{otherwise}\n",
        "  \\end{cases}\n",
        "$"
      ],
      "metadata": {
        "id": "GdZx7O6HxwIG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pred_congestion_eval = pd.pivot_table(prb_util_prediction, values=['Cong_Pred', 'Cong_True'],\n",
        "                                      index='Cell', columns='weekend', aggfunc='sum')\n",
        "pred_congestion_eval = pred_congestion_eval.droplevel(0, axis=1).reset_index()\n",
        "pred_congestion_eval.columns = ['Cell', 'Cong_pred_weekday','Cong_pred_weekend','Cong_true_weekday','Cong_true_weekend']\n",
        "\n",
        "pred_congestion_eval.head(15)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 520
        },
        "id": "H69L3wWgBc8j",
        "outputId": "85eee5f5-1442-40a2-db5f-8f7934a0bb5f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        Cell  Cong_pred_weekday  Cong_pred_weekend  Cong_true_weekday  \\\n",
              "0   cell0001                  0                  0                  0   \n",
              "1   cell0002                  0                  0                  0   \n",
              "2   cell0003                  0                  0                  0   \n",
              "3   cell0004                  0                  0                  0   \n",
              "4   cell0006                  0                  0                  1   \n",
              "5   cell0007                  0                  0                 61   \n",
              "6   cell0008                  0                  0                  6   \n",
              "7   cell0009                  0                  0                  0   \n",
              "8   cell0010                  0                  0                  0   \n",
              "9   cell0011                  0                  0                  1   \n",
              "10  cell0012                 14                  0                 37   \n",
              "11  cell0013                  0                  0                  0   \n",
              "12  cell0014                206                 89                 98   \n",
              "13  cell0015                  0                  1                  0   \n",
              "14  cell0016                  0                  1                  2   \n",
              "\n",
              "    Cong_true_weekend  \n",
              "0                   0  \n",
              "1                   1  \n",
              "2                   0  \n",
              "3                   0  \n",
              "4                   0  \n",
              "5                   0  \n",
              "6                   0  \n",
              "7                   0  \n",
              "8                   0  \n",
              "9                   0  \n",
              "10                 29  \n",
              "11                  0  \n",
              "12                 58  \n",
              "13                  0  \n",
              "14                  0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-bf77dc64-2753-4174-b46b-29a7a931442b\">\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>Cell</th>\n",
              "      <th>Cong_pred_weekday</th>\n",
              "      <th>Cong_pred_weekend</th>\n",
              "      <th>Cong_true_weekday</th>\n",
              "      <th>Cong_true_weekend</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>cell0001</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>cell0002</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>cell0003</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>cell0004</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>cell0006</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>cell0007</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>61</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>cell0008</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>6</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>cell0009</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>cell0010</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>cell0011</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>cell0012</td>\n",
              "      <td>14</td>\n",
              "      <td>0</td>\n",
              "      <td>37</td>\n",
              "      <td>29</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>cell0013</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>cell0014</td>\n",
              "      <td>206</td>\n",
              "      <td>89</td>\n",
              "      <td>98</td>\n",
              "      <td>58</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>cell0015</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>cell0016</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-bf77dc64-2753-4174-b46b-29a7a931442b')\"\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-bf77dc64-2753-4174-b46b-29a7a931442b 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-bf77dc64-2753-4174-b46b-29a7a931442b');\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": 113
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## V. Conclusions"
      ],
      "metadata": {
        "id": "kqWMj4wvPnjE"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "As stated at the beginning of this study, the main objective of this work is to replicate the work done by D. Chmieliauskas and D. Guršnys in order to make the code available on paperswithcode.com. Some improvements were suggested throughout this work, such as data transformation through box-cox, the utilization of a 1.5-year data history to capture annual seasonality, the differentiation of data between weekdays and weekends to better capture individual cell characteristics, and finally, the use of the sMAPE indicator due to its robustness to outliers. The results obtained in this work were not satisfactory; however, since there was no tuning performed, it is possible to improve them."
      ],
      "metadata": {
        "id": "W0R4TOpTokMh"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "---"
      ],
      "metadata": {
        "id": "aLVsj0DCooFd"
      }
    }
  ]
}