From e0fa17f755bb71f77de9634f07651fc2a1f24815 Mon Sep 17 00:00:00 2001 From: akhropov Date: Tue, 2 Jun 2020 23:29:45 +0300 Subject: [PATCH] COVID_19 extended tutorial. MLTOOLS-4805. ref:94d41424ae1906c3f7306215042185ca06767be6 --- example_usages/COVID_19.ipynb | 10381 +++++++++++++++++++++++++++++--- 1 file changed, 9705 insertions(+), 676 deletions(-) diff --git a/example_usages/COVID_19.ipynb b/example_usages/COVID_19.ipynb index 335008e..4fb8e3b 100644 --- a/example_usages/COVID_19.ipynb +++ b/example_usages/COVID_19.ipynb @@ -3,7 +3,7 @@ "nbformat_minor": 0, "metadata": { "colab": { - "name": "COVID-19.ipynb", + "name": "COVID_19_2.ipynb", "provenance": [], "collapsed_sections": [] }, @@ -22,16 +22,65 @@ "source": [ "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catboost/tutorials/blob/master/example_usages/COVID_19.ipynb)\n", "\n", - "This is a simple example on how to predict COVID-19 spread using well-known Gradient Boosting Decision Trees libraries (CatBoost, LightGBM and XgBoost). \n", + "This is an example on how to predict COVID-19 spread. The goal is to predict Confirmed cases and Fatalities for future dates based on data for previous dates. \n", "\n", - "The data to train on is taken from https://www.kaggle.com/c/covid19-global-forecasting-week-1/data.\n", - "Training data contains cumulative numbers of Confirmed cases and Fatalities for the different regions for the number of consecutive dates (contained in ``Date`` field) from 2020-01-22 to 2020-03-22.\n", + "## Table of contents\n", + "\n", + "* [General description](#general_description)\n", + "* [Install and import necessary packages ](#install_and_import)\n", + "* [Get main data from Kaggle](#get_main_data)\n", + "* [Feature engineering](#features)\n", + " * [Time delay embedding features](#time_delay_embedding_features)\n", + " * [Day feature](#day_feature)\n", + " * [WeekDay feature](#week_day_feature)\n", + " * [Days since Xth Confirmed case and Xth Fatality features](#days_since_features)\n", + " * ['Distance to Origin' feature](#distance_to_origin_feature)\n", + " * [Functions for merging external data](#functions_for_merging)\n", + " * [Country Area](#country_area_feature)\n", + " * [Country population features](#country_population_features)\n", + " * [Country Population density feature](#country_population_density_feature)\n", + " * [Country Smoking rate feature](#country_smoking_rate_feature)\n", + " * [Country hospital beds per 1000 people](#country_hospital_beds_feature)\n", + " * [Country Health Expenditure](#country_health_expenditure_feature)\n", + "* [Prepare data for training](#prepare_for_training)\n", + "* [Models training](#models_training)\n", + "* [Feature importance in the models](#feature_importance)\n", + "* [Create predictions for eval and test data](#predictions)\n", + "* [Plots with predictions](#plots_with_predictions)\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Kw0fIYKmolzm", + "colab_type": "text" + }, + "source": [ + "## General description \n", + "\n", + "The main data to train on is taken from https://www.kaggle.com/c/covid19-global-forecasting-week-1/data.\n", "Regions are specified using two data fields: ``Province/State`` and ``Country/Region``.\n", - "Data is also enriched with region coordinates in ``Lat`` and ``Long`` fields.\n", + "This data is also enriched with region coordinates in ``Lat`` and ``Long`` fields.\n", + "Train data contains cumulative numbers of Confirmed cases and Fatalities for the different regions for the number of consecutive dates (contained in ``Date`` field) from 2020-01-22 to 2020-03-24. Test data is for dates from 2020-03-12 to 2020-04-23. Test data does not have Confirmed cases and Fatalities.\n", "\n", - "The goal is to predict Confirmed cases and Fatalities for future dates based on data for previous dates. \n", + "We will also add additional data for features from other sources:\n", + "* Country Land area from [World bank data](https://data.worldbank.org/indicator/AG.LND.TOTL.K2)\n", + "* Country Smoking rate from [World bank data](https://data.worldbank.org/indicator/SH.PRV.SMOK)\n", + "* Country Population Age and Sex distribution from [UN World Population Prospects](https://population.un.org/wpp/Download/Standard/CSV/)\n", + "* Country Hospital beds per 1000 people from [World Bank data](https://data.worldbank.org/indicator/sh.med.beds.zs)\n", + "* Country Health expenditure per Capita, PPP (current international $) from [World Bank data](https://data.worldbank.org/indicator/SH.XPD.CHEX.PP.CD) \n", "\n", - "For the sake of simplicity no extensive preprocessing and feature engineering is performed here. If you're interested in more comprehensive analysis of this data look at the Kaggle competition notebooks: https://www.kaggle.com/c/covid19-global-forecasting-week-1/notebooks." + "We will construct models that predict new cases and new fatalities (in the logarithmic scale) for the next day based on time delay embedding features for the last 30 days plus additional features from sources described above. \n", + "\n", + "For forecasting further into the future (more than a single day) we will use an incremental (by days) approach: we will use predicted values of new cases and new fatalities for days with already calculated predictions as input data for time delay embedding features for the next prediction day and then use the same procedure for the new next prediction day and so on." ] }, { @@ -41,7 +90,7 @@ "colab_type": "text" }, "source": [ - "Install and import necessary packages" + "## Install and import necessary packages \n" ] }, { @@ -49,23 +98,20 @@ "metadata": { "id": "IrFU3KqUjYf-", "colab_type": "code", - "outputId": "4a9c57fb-7e43-42d8-f177-44999666e444", + "outputId": "9f7026c0-47f5-4cbe-fbb3-2a38c1d3e8f8", "colab": { "base_uri": "https://localhost:8080/", - "height": 999 + "height": 678 } }, "source": [ - "!pip install kaggle\n", - "\n", - "# install the latest versions of all GBDT packages\n", - "\n", + "!pip install kaggle -U\n", + "!pip install pandas -U\n", "!pip install catboost -U\n", - "!pip install lightgbm -U\n", - "!pip install xgboost -U\n", "\n", "import os\n", "import pathlib\n", + "import re\n", "\n", "import numpy as np\n", "import pandas as pd\n", @@ -73,66 +119,47 @@ "\n", "import geopy.distance\n", "\n", - "import catboost as cb\n", - "import lightgbm as lgb\n", - "import xgboost as xgb" + "import catboost as cb\n" ], - "execution_count": 3, + "execution_count": 1, "outputs": [ { "output_type": "stream", "text": [ - "Requirement already satisfied: kaggle in /usr/local/lib/python3.6/dist-packages (1.5.6)\n", - "Requirement already satisfied: python-slugify in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.0.0)\n", - "Requirement already satisfied: certifi in /usr/local/lib/python3.6/dist-packages (from kaggle) (2019.11.28)\n", - "Requirement already satisfied: tqdm in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.38.0)\n", - "Requirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.21.0)\n", - "Requirement already satisfied: urllib3<1.25,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.24.3)\n", - "Requirement already satisfied: six>=1.10 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.12.0)\n", - "Requirement already satisfied: python-dateutil in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.8.1)\n", - "Requirement already satisfied: text-unidecode>=1.3 in /usr/local/lib/python3.6/dist-packages (from python-slugify->kaggle) (1.3)\n", - "Requirement already satisfied: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (3.0.4)\n", - "Requirement already satisfied: idna<2.9,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (2.8)\n", + "Requirement already up-to-date: kaggle in /usr/local/lib/python3.6/dist-packages (1.5.6)\n", + "Requirement already satisfied, skipping upgrade: certifi in /usr/local/lib/python3.6/dist-packages (from kaggle) (2020.4.5.1)\n", + "Requirement already satisfied, skipping upgrade: six>=1.10 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.12.0)\n", + "Requirement already satisfied, skipping upgrade: urllib3<1.25,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.24.3)\n", + "Requirement already satisfied, skipping upgrade: python-dateutil in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.8.1)\n", + "Requirement already satisfied, skipping upgrade: requests in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.21.0)\n", + "Requirement already satisfied, skipping upgrade: tqdm in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.38.0)\n", + "Requirement already satisfied, skipping upgrade: python-slugify in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.0.0)\n", + "Requirement already satisfied, skipping upgrade: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (3.0.4)\n", + "Requirement already satisfied, skipping upgrade: idna<2.9,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (2.8)\n", + "Requirement already satisfied, skipping upgrade: text-unidecode>=1.3 in /usr/local/lib/python3.6/dist-packages (from python-slugify->kaggle) (1.3)\n", + "Requirement already up-to-date: pandas in /usr/local/lib/python3.6/dist-packages (1.0.3)\n", + "Requirement already satisfied, skipping upgrade: python-dateutil>=2.6.1 in /usr/local/lib/python3.6/dist-packages (from pandas) (2.8.1)\n", + "Requirement already satisfied, skipping upgrade: numpy>=1.13.3 in /usr/local/lib/python3.6/dist-packages (from pandas) (1.18.2)\n", + "Requirement already satisfied, skipping upgrade: pytz>=2017.2 in /usr/local/lib/python3.6/dist-packages (from pandas) (2018.9)\n", + "Requirement already satisfied, skipping upgrade: six>=1.5 in /usr/local/lib/python3.6/dist-packages (from python-dateutil>=2.6.1->pandas) (1.12.0)\n", "Collecting catboost\n", "\u001b[?25l Downloading https://files.pythonhosted.org/packages/94/ec/12b9a42b2ea7dfe5b602f235692ab2b61ee1334ff34334a15902272869e8/catboost-0.22-cp36-none-manylinux1_x86_64.whl (64.4MB)\n", - "\u001b[K |████████████████████████████████| 64.4MB 67kB/s \n", - "\u001b[?25hRequirement already satisfied, skipping upgrade: scipy in /usr/local/lib/python3.6/dist-packages (from catboost) (1.4.1)\n", - "Requirement already satisfied, skipping upgrade: numpy>=1.16.0 in /usr/local/lib/python3.6/dist-packages (from catboost) (1.18.2)\n", - "Requirement already satisfied, skipping upgrade: pandas>=0.24.0 in /usr/local/lib/python3.6/dist-packages (from catboost) (1.0.3)\n", + "\u001b[K |████████████████████████████████| 64.4MB 61kB/s \n", + "\u001b[?25hRequirement already satisfied, skipping upgrade: matplotlib in /usr/local/lib/python3.6/dist-packages (from catboost) (3.2.1)\n", "Requirement already satisfied, skipping upgrade: plotly in /usr/local/lib/python3.6/dist-packages (from catboost) (4.4.1)\n", + "Requirement already satisfied, skipping upgrade: numpy>=1.16.0 in /usr/local/lib/python3.6/dist-packages (from catboost) (1.18.2)\n", "Requirement already satisfied, skipping upgrade: graphviz in /usr/local/lib/python3.6/dist-packages (from catboost) (0.10.1)\n", + "Requirement already satisfied, skipping upgrade: scipy in /usr/local/lib/python3.6/dist-packages (from catboost) (1.4.1)\n", + "Requirement already satisfied, skipping upgrade: pandas>=0.24.0 in /usr/local/lib/python3.6/dist-packages (from catboost) (1.0.3)\n", "Requirement already satisfied, skipping upgrade: six in /usr/local/lib/python3.6/dist-packages (from catboost) (1.12.0)\n", - "Requirement already satisfied, skipping upgrade: matplotlib in /usr/local/lib/python3.6/dist-packages (from catboost) (3.2.1)\n", - "Requirement already satisfied, skipping upgrade: pytz>=2017.2 in /usr/local/lib/python3.6/dist-packages (from pandas>=0.24.0->catboost) (2018.9)\n", - "Requirement already satisfied, skipping upgrade: python-dateutil>=2.6.1 in /usr/local/lib/python3.6/dist-packages (from pandas>=0.24.0->catboost) (2.8.1)\n", - "Requirement already satisfied, skipping upgrade: retrying>=1.3.3 in /usr/local/lib/python3.6/dist-packages (from plotly->catboost) (1.3.3)\n", + "Requirement already satisfied, skipping upgrade: python-dateutil>=2.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->catboost) (2.8.1)\n", "Requirement already satisfied, skipping upgrade: kiwisolver>=1.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->catboost) (1.2.0)\n", - "Requirement already satisfied, skipping upgrade: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->catboost) (2.4.6)\n", + "Requirement already satisfied, skipping upgrade: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->catboost) (2.4.7)\n", "Requirement already satisfied, skipping upgrade: cycler>=0.10 in /usr/local/lib/python3.6/dist-packages (from matplotlib->catboost) (0.10.0)\n", + "Requirement already satisfied, skipping upgrade: retrying>=1.3.3 in /usr/local/lib/python3.6/dist-packages (from plotly->catboost) (1.3.3)\n", + "Requirement already satisfied, skipping upgrade: pytz>=2017.2 in /usr/local/lib/python3.6/dist-packages (from pandas>=0.24.0->catboost) (2018.9)\n", "Installing collected packages: catboost\n", - "Successfully installed catboost-0.22\n", - "Collecting lightgbm\n", - "\u001b[?25l Downloading https://files.pythonhosted.org/packages/0b/9d/ddcb2f43aca194987f1a99e27edf41cf9bc39ea750c3371c2a62698c509a/lightgbm-2.3.1-py2.py3-none-manylinux1_x86_64.whl (1.2MB)\n", - "\u001b[K |████████████████████████████████| 1.2MB 4.5MB/s \n", - "\u001b[?25hRequirement already satisfied, skipping upgrade: numpy in /usr/local/lib/python3.6/dist-packages (from lightgbm) (1.18.2)\n", - "Requirement already satisfied, skipping upgrade: scipy in /usr/local/lib/python3.6/dist-packages (from lightgbm) (1.4.1)\n", - "Requirement already satisfied, skipping upgrade: scikit-learn in /usr/local/lib/python3.6/dist-packages (from lightgbm) (0.22.2.post1)\n", - "Requirement already satisfied, skipping upgrade: joblib>=0.11 in /usr/local/lib/python3.6/dist-packages (from scikit-learn->lightgbm) (0.14.1)\n", - "Installing collected packages: lightgbm\n", - " Found existing installation: lightgbm 2.2.3\n", - " Uninstalling lightgbm-2.2.3:\n", - " Successfully uninstalled lightgbm-2.2.3\n", - "Successfully installed lightgbm-2.3.1\n", - "Collecting xgboost\n", - "\u001b[?25l Downloading https://files.pythonhosted.org/packages/70/91/551d37ba472bcbd70a25e667acc65a18a9d053657b13afcf0f87aa24d7bb/xgboost-1.0.2-py3-none-manylinux1_x86_64.whl (109.7MB)\n", - "\u001b[K |████████████████████████████████| 109.8MB 92kB/s \n", - "\u001b[?25hRequirement already satisfied, skipping upgrade: scipy in /usr/local/lib/python3.6/dist-packages (from xgboost) (1.4.1)\n", - "Requirement already satisfied, skipping upgrade: numpy in /usr/local/lib/python3.6/dist-packages (from xgboost) (1.18.2)\n", - "Installing collected packages: xgboost\n", - " Found existing installation: xgboost 0.90\n", - " Uninstalling xgboost-0.90:\n", - " Successfully uninstalled xgboost-0.90\n", - "Successfully installed xgboost-1.0.2\n" + "Successfully installed catboost-0.22\n" ], "name": "stdout" } @@ -145,6 +172,8 @@ "colab_type": "text" }, "source": [ + "## Get main data from Kaggle \n", + "\n", "As we download from Kaggle we need to provide Kaggle authetification cridentials. We will use the method suggested [here](https://gist.github.com/jayspeidell/d10b84b8d3da52df723beacc5b15cb27#gistcomment-2814834). Replace KAGGLE_USERNAME and KAGGLE_KEY with your own." ] }, @@ -177,7 +206,7 @@ "metadata": { "id": "7X0mwipNgJ7t", "colab_type": "code", - "outputId": "69f3e835-061c-49f8-8251-b1fbfb150cfc", + "outputId": "367366c9-e71c-4721-d656-41bb231813e1", "colab": { "base_uri": "https://localhost:8080/", "height": 206 @@ -190,21 +219,21 @@ "!mkdir -p ${data_folder}\n", "!kaggle competitions download -c {kaggle_data_name} --path {data_folder}" ], - "execution_count": 5, + "execution_count": 3, "outputs": [ { "output_type": "stream", "text": [ "Warning: Looks like you're using an outdated API Version, please consider updating (server 1.5.6 / client 1.5.4)\n", "Downloading submission.csv to /root/datasets/covid19-global-forecasting-week-1\n", - " 0% 0.00/108k [00:00\n", " \n", " \n", + " ForecastId\n", " Province/State\n", " Country/Region\n", " Lat\n", " Long\n", " Date\n", - " ConfirmedCases\n", - " Fatalities\n", - " Day\n", - " Distance_to_Hubei\n", - " WeekDay\n", " \n", " \n", " \n", " \n", " 0\n", - " \n", + " 1\n", + " NaN\n", " Afghanistan\n", " 33.0\n", " 65.0\n", - " 2020-01-22\n", - " 0.0\n", - " 0.0\n", - " 0\n", - " 4435.232618\n", - " 2\n", + " 2020-03-12\n", " \n", " \n", " 1\n", - " \n", + " 2\n", + " NaN\n", " Afghanistan\n", " 33.0\n", " 65.0\n", - " 2020-01-23\n", - " 0.0\n", - " 0.0\n", - " 1\n", - " 4435.232618\n", - " 3\n", + " 2020-03-13\n", " \n", " \n", " 2\n", - " \n", + " 3\n", + " NaN\n", " Afghanistan\n", " 33.0\n", " 65.0\n", - " 2020-01-24\n", - " 0.0\n", - " 0.0\n", - " 2\n", - " 4435.232618\n", - " 4\n", + " 2020-03-14\n", " \n", " \n", " 3\n", - " \n", + " 4\n", + " NaN\n", " Afghanistan\n", " 33.0\n", " 65.0\n", - " 2020-01-25\n", - " 0.0\n", - " 0.0\n", - " 3\n", - " 4435.232618\n", - " 5\n", + " 2020-03-15\n", " \n", " \n", " 4\n", - " \n", + " 5\n", + " NaN\n", " Afghanistan\n", " 33.0\n", " 65.0\n", - " 2020-01-26\n", - " 0.0\n", - " 0.0\n", - " 4\n", - " 4435.232618\n", - " 6\n", + " 2020-03-16\n", " \n", " \n", "\n", "" ], "text/plain": [ - " Province/State Country/Region Lat ... Day Distance_to_Hubei WeekDay\n", - "0 Afghanistan 33.0 ... 0 4435.232618 2\n", - "1 Afghanistan 33.0 ... 1 4435.232618 3\n", - "2 Afghanistan 33.0 ... 2 4435.232618 4\n", - "3 Afghanistan 33.0 ... 3 4435.232618 5\n", - "4 Afghanistan 33.0 ... 4 4435.232618 6\n", - "\n", - "[5 rows x 10 columns]" + " ForecastId Province/State Country/Region Lat Long Date\n", + "0 1 NaN Afghanistan 33.0 65.0 2020-03-12\n", + "1 2 NaN Afghanistan 33.0 65.0 2020-03-13\n", + "2 3 NaN Afghanistan 33.0 65.0 2020-03-14\n", + "3 4 NaN Afghanistan 33.0 65.0 2020-03-15\n", + "4 5 NaN Afghanistan 33.0 65.0 2020-03-16" ] }, "metadata": { "tags": [] }, - "execution_count": 8 + "execution_count": 6 } ] }, { "cell_type": "markdown", "metadata": { - "id": "To5EPZPHn5if", - "colab_type": "text" - }, - "source": [ - "We will split data to train and test by date.\n", - "\n", - "Labels are transformed to the logarithmic scale. Logarithmic scale is natural for this task (model should weight errors between 1000 and 1100 and 100000 and 110000 approximately the same).\n", - "\n", - "We will use RMSE as an objective and evaluation metric for training, that will correspond to RMSLE (Root Mean Square Logarithmic Error) metric.\n", - "\n", - "$$RMSLE(labels, predictions) = \\sqrt{\\frac{1}{n}\\sum_{i=1}^{n}{\\Big(log(labels_i + 1) - log(predictions_i + 1)\\Big)^2}}$$\n", - "\n", - "For prediction values in the original scale model prediction output should be transformed back to the original scale with the reverse transformation (can be preformed using numpy.expm1 function)." - ] - }, - { - "cell_type": "code", - "metadata": { - "id": "CcmpCDfKqPxC", - "colab_type": "code", - "colab": {} - }, - "source": [ - "last_train_date = pd.Timestamp(2020, 3, 11)\n", - "\n", - "# returns train_features_df, train_labels, test_features_df, test_labels\n", - "def get_train_and_test_data(df):\n", - " train_df = df[df['Date'] <= last_train_date].copy()\n", - " test_df = df[df['Date'] > last_train_date].copy()\n", - " \n", - " train_df.sort_values(by=['Date'])\n", - " train_labels = {\n", - " 'ConfirmedCases': np.log1p(train_df['ConfirmedCases']),\n", - " 'Fatalities': np.log1p(train_df['Fatalities'])\n", - " }\n", - " train_df.drop(columns=['ConfirmedCases', 'Fatalities', 'Date'], inplace=True)\n", - " \n", - " test_labels = {\n", - " 'ConfirmedCases': np.log1p(test_df['ConfirmedCases']),\n", - " 'Fatalities': np.log1p(test_df['Fatalities'])\n", - " }\n", - " test_df.drop(columns=['ConfirmedCases', 'Fatalities', 'Date'], inplace=True)\n", - " \n", - " return train_df, train_labels, test_df, test_labels\n", - "\n" - ], - "execution_count": 0, - "outputs": [] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "Dfao1LWwqr9W", - "colab_type": "text" - }, - "source": [ - "Categorical features" - ] - }, - { - "cell_type": "code", - "metadata": { - "id": "cyDExsbCqUVD", - "colab_type": "code", - "colab": {} - }, - "source": [ - "cat_features = ['Province/State', 'Country/Region', 'WeekDay']" - ], - "execution_count": 0, - "outputs": [] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "xlibxWBWrAOV", - "colab_type": "text" - }, - "source": [ - "We'll train models for 1000 iterations.\n" - ] - }, - { - "cell_type": "code", - "metadata": { - "id": "MKfJlg_5qvmY", - "colab_type": "code", - "colab": {} - }, - "source": [ - "iterations = 1000" - ], - "execution_count": 0, - "outputs": [] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "nemw7Ds_rG03", + "id": "mRkKFg29o6y3", "colab_type": "text" }, "source": [ - "Let's train CatBoost first. Categorical features values can be used as is in DataFrame.\n" + "We will concatenate original train and test dataframes for easier time delay embedding features calculations.\n", + "Original train and test data intersect by dates. We will use data from original train in the concatenated dataset for such dates." ] }, { "cell_type": "code", "metadata": { - "id": "BiRDMnL5rDj_", + "id": "4FCBHJhepLw7", "colab_type": "code", - "outputId": "96e59369-8c28-4ff1-90af-e26f799134a4", + "outputId": "bdc9816f-7174-4bcd-caca-2788fdfc097a", "colab": { "base_uri": "https://localhost:8080/", - "height": 696 + "height": 74 } }, "source": [ - "print ('catboost version', cb.__version__)\n", + "last_original_train_date = original_train_df['Date'].max()\n", + "print ('last_original_train_date = ', last_original_train_date)\n", "\n", - "train_df, train_labels, test_df, test_labels = get_train_and_test_data(preprocessed_df)\n", + "print ('original_test_df.shape =', original_test_df.shape)\n", "\n", - "catboost_models = {}\n", + "original_test_wo_train_df = original_test_df.drop(\n", + " index=original_test_df[original_test_df['Date'] <= last_original_train_date].index\n", + ")\n", "\n", - "for prediction_name in ['ConfirmedCases', 'Fatalities']:\n", - " model = cb.CatBoostRegressor(\n", - " has_time=True,\n", - " iterations=iterations\n", - " )\n", + "print ('original_test_wo_train_df.shape =', original_test_wo_train_df.shape)\n", "\n", - " model.fit(\n", - " train_df,\n", - " train_labels[prediction_name],\n", - " eval_set=(test_df, test_labels[prediction_name]),\n", - " cat_features=cat_features,\n", - " verbose=100 # print metrics each 100 iterations\n", - " )\n", - " \n", - " catboost_models[prediction_name] = model\n", - " print ('CatBoost: prediction of %s: RMSLE on test = %s' % (prediction_name, model.evals_result_['validation']['RMSE'][-1]))" + "# recreate index bacause we will need unique index values later \n", + "main_df = pd.concat([original_train_df, original_test_wo_train_df], ignore_index=True)" ], - "execution_count": 12, + "execution_count": 7, "outputs": [ { "output_type": "stream", "text": [ - "catboost version 0.22\n", - "Learning rate set to 0.07608\n", - "0:\tlearn: 1.7234085\ttest: 3.5386691\tbest: 3.5386691 (0)\ttotal: 64.3ms\tremaining: 1m 4s\n", - "100:\tlearn: 0.3685676\ttest: 2.0804581\tbest: 2.0804581 (100)\ttotal: 1.36s\tremaining: 12.1s\n", - "200:\tlearn: 0.2733765\ttest: 1.9753332\tbest: 1.9753332 (200)\ttotal: 2.7s\tremaining: 10.7s\n", - "300:\tlearn: 0.2262019\ttest: 1.9121371\tbest: 1.9120677 (298)\ttotal: 4.04s\tremaining: 9.38s\n", - "400:\tlearn: 0.1976705\ttest: 1.8692683\tbest: 1.8692537 (399)\ttotal: 5.35s\tremaining: 8s\n", - "500:\tlearn: 0.1781107\ttest: 1.8042560\tbest: 1.8042560 (500)\ttotal: 6.66s\tremaining: 6.63s\n", - "600:\tlearn: 0.1632501\ttest: 1.7695398\tbest: 1.7695398 (600)\ttotal: 7.97s\tremaining: 5.29s\n", - "700:\tlearn: 0.1508874\ttest: 1.7541809\tbest: 1.7541809 (700)\ttotal: 9.28s\tremaining: 3.96s\n", - "800:\tlearn: 0.1416807\ttest: 1.7333306\tbest: 1.7333306 (800)\ttotal: 10.6s\tremaining: 2.63s\n", - "900:\tlearn: 0.1347790\ttest: 1.7220938\tbest: 1.7217529 (891)\ttotal: 11.9s\tremaining: 1.31s\n", - "999:\tlearn: 0.1278086\ttest: 1.7130733\tbest: 1.7130733 (999)\ttotal: 13.2s\tremaining: 0us\n", - "\n", - "bestTest = 1.713073301\n", - "bestIteration = 999\n", - "\n", - "CatBoost: prediction of ConfirmedCases: RMSLE on test = 1.7130733010288208\n", - "Learning rate set to 0.07608\n", - "0:\tlearn: 0.5466355\ttest: 1.3958990\tbest: 1.3958990 (0)\ttotal: 13.3ms\tremaining: 13.2s\n", - "100:\tlearn: 0.0996172\ttest: 0.9229031\tbest: 0.9229031 (100)\ttotal: 1.11s\tremaining: 9.88s\n", - "200:\tlearn: 0.0743975\ttest: 0.8680675\tbest: 0.8680650 (199)\ttotal: 2.32s\tremaining: 9.22s\n", - "300:\tlearn: 0.0612372\ttest: 0.8318278\tbest: 0.8313150 (296)\ttotal: 3.54s\tremaining: 8.22s\n", - "400:\tlearn: 0.0540932\ttest: 0.8164244\tbest: 0.8164244 (400)\ttotal: 4.79s\tremaining: 7.16s\n", - "500:\tlearn: 0.0502328\ttest: 0.8050471\tbest: 0.8050471 (500)\ttotal: 6.04s\tremaining: 6.02s\n", - "600:\tlearn: 0.0473407\ttest: 0.8020130\tbest: 0.8020130 (600)\ttotal: 7.36s\tremaining: 4.88s\n", - "700:\tlearn: 0.0446151\ttest: 0.7984937\tbest: 0.7966942 (665)\ttotal: 8.65s\tremaining: 3.69s\n", - "800:\tlearn: 0.0422479\ttest: 0.8002849\tbest: 0.7966942 (665)\ttotal: 9.94s\tremaining: 2.47s\n", - "900:\tlearn: 0.0401878\ttest: 0.7974207\tbest: 0.7966942 (665)\ttotal: 11.2s\tremaining: 1.24s\n", - "999:\tlearn: 0.0379386\ttest: 0.7938151\tbest: 0.7930683 (939)\ttotal: 12.5s\tremaining: 0us\n", - "\n", - "bestTest = 0.7930682959\n", - "bestIteration = 939\n", - "\n", - "Shrink model to first 940 iterations.\n", - "CatBoost: prediction of Fatalities: RMSLE on test = 0.7938150681857505\n" + "last_original_train_date = 2020-03-24 00:00:00\n", + "original_test_df.shape = (12212, 6)\n", + "original_test_wo_train_df.shape = (8520, 6)\n" ], "name": "stdout" } @@ -766,457 +580,9672 @@ { "cell_type": "markdown", "metadata": { - "id": "RyFZzWPf0MB7", + "id": "8rrrBKFJuxpr", "colab_type": "text" }, "source": [ - "Let's look at feature importances.\n", - "For CatBoost use [PredictionValuesChange](https://catboost.ai/docs/concepts/fstr.html#fstr__regular-feature-importance)." + "There are special `Provinces` that are in fact data from Cruise ships with `Country` information. Make this ships ``Countries`` and passengers origin country a ``Province/State`` (for proper hierarchy). " ] }, { "cell_type": "code", "metadata": { - "id": "XDeJNSYuyd--", + "id": "shg5-Qk4sSMS", "colab_type": "code", - "outputId": "1e1e192f-5126-4174-e5bc-30c847e16bc8", + "outputId": "4e99bbdc-bc5f-42c0-8144-b4312f6e011d", "colab": { "base_uri": "https://localhost:8080/", - "height": 394 + "height": 424 } }, "source": [ - "for prediction_name in ['ConfirmedCases', 'Fatalities']:\n", - " print ('\\nCatBoost: prediction of %s. Feature importance. Type=PredictionValuesChange' % prediction_name)\n", - " print (catboost_models[prediction_name].get_feature_importance(type=cb.EFstrType.PredictionValuesChange, prettified=True))" + "from_cruise_ships = main_df['Province/State'].isin(['From Diamond Princess', 'Grand Princess'])\n", + "main_df.loc[from_cruise_ships, ['Province/State','Country/Region']] = main_df.loc[from_cruise_ships, ['Country/Region','Province/State']].values\n", + "main_df[main_df['Country/Region'].isin(['From Diamond Princess', 'Grand Princess'])]" ], - "execution_count": 13, + "execution_count": 8, "outputs": [ { - "output_type": "stream", - "text": [ - "\n", - "CatBoost: prediction of ConfirmedCases. Feature importance. Type=PredictionValuesChange\n", - " Feature Id Importances\n", - "0 Distance_to_Hubei 34.660626\n", - "1 Country/Region 17.132457\n", - "2 Province/State 16.414815\n", - "3 Day 13.621041\n", - "4 Lat 9.746912\n", - "5 Long 6.513599\n", - "6 WeekDay 1.910550\n", - "\n", - "CatBoost: prediction of Fatalities. Feature importance. Type=PredictionValuesChange\n", - " Feature Id Importances\n", - "0 Province/State 49.637990\n", - "1 Distance_to_Hubei 14.761734\n", - "2 Country/Region 12.875072\n", - "3 Long 10.726166\n", - "4 Lat 5.548307\n", - "5 Day 4.907931\n", - "6 WeekDay 1.542799\n" - ], - "name": "stdout" - } - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "vYDU8cAFvfnB", - "colab_type": "text" - }, - "source": [ - "Let's train LightGBM. Categorical features has to be transformed to integer values using enumeration." - ] - }, - { - "cell_type": "code", - "metadata": { - "id": "tTu1KxpUvUN9", - "colab_type": "code", - "outputId": "89b49dae-7901-4c93-f996-662df7c6532a", - "colab": { - "base_uri": "https://localhost:8080/", - "height": 508 - } - }, - "source": [ - "# returns df, dict of cat_feature -> OrdinalEncoder\n", - "def preprocess_for_lightgbm(df, cat_features):\n", - " df = df.copy()\n", - " encoders = {}\n", - " for cat_feature in cat_features:\n", - " encoder = sklearn.preprocessing.OrdinalEncoder()\n", - " df[cat_feature] = encoder.fit_transform([df[cat_feature]])[0]\n", - " encoders[cat_feature] = encoder\n", - " \n", - " return df, encoders\n", - "\n", - "\n", - "print ('lightgbm version', lgb.__version__)\n", - "\n", - "preprocessed_for_lighgbm_df, encoders = preprocess_for_lightgbm(preprocessed_df, cat_features)\n", - "\n", - "train_df, train_labels, test_df, test_labels = get_train_and_test_data(preprocessed_for_lighgbm_df)\n", - "\n", - "lightgbm_models = {}\n", - "\n", - "for prediction_name in ['ConfirmedCases', 'Fatalities']:\n", - " model = lgb.LGBMRegressor(\n", - " objective='rmse',\n", - " n_estimators=iterations\n", - " )\n", - "\n", - " model = model.fit(\n", - " X=train_df,\n", - " y=train_labels[prediction_name],\n", - " categorical_feature=cat_features,\n", - " eval_set=[(test_df, test_labels[prediction_name])],\n", - " verbose=100 # print metrics each 100 iterations\n", - " )\n", - "\n", - " lightgbm_models[prediction_name] = model\n", - " print ('LightGBM: prediction of %s: RMSLE on test = %s' % (prediction_name, model.best_score_['valid_0']['rmse'])) \n", - " \n" - ], - "execution_count": 17, + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastId
567838.0AustraliaFrom Diamond Princess35.4437139.63802020-01-220.00.0NaN
568839.0AustraliaFrom Diamond Princess35.4437139.63802020-01-230.00.0NaN
569840.0AustraliaFrom Diamond Princess35.4437139.63802020-01-240.00.0NaN
570841.0AustraliaFrom Diamond Princess35.4437139.63802020-01-250.00.0NaN
571842.0AustraliaFrom Diamond Princess35.4437139.63802020-01-260.00.0NaN
..............................
24697NaNUSGrand Princess37.6489-122.66552020-04-19NaNNaN9757.0
24698NaNUSGrand Princess37.6489-122.66552020-04-20NaNNaN9758.0
24699NaNUSGrand Princess37.6489-122.66552020-04-21NaNNaN9759.0
24700NaNUSGrand Princess37.6489-122.66552020-04-22NaNNaN9760.0
24701NaNUSGrand Princess37.6489-122.66552020-04-23NaNNaN9761.0
\n", + "

279 rows × 9 columns

\n", + "
" + ], + "text/plain": [ + " Id Province/State ... Fatalities ForecastId\n", + "567 838.0 Australia ... 0.0 NaN\n", + "568 839.0 Australia ... 0.0 NaN\n", + "569 840.0 Australia ... 0.0 NaN\n", + "570 841.0 Australia ... 0.0 NaN\n", + "571 842.0 Australia ... 0.0 NaN\n", + "... ... ... ... ... ...\n", + "24697 NaN US ... NaN 9757.0\n", + "24698 NaN US ... NaN 9758.0\n", + "24699 NaN US ... NaN 9759.0\n", + "24700 NaN US ... NaN 9760.0\n", + "24701 NaN US ... NaN 9761.0\n", + "\n", + "[279 rows x 9 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 8 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "88Fupigl0Nin", + "colab_type": "text" + }, + "source": [ + "## Feature engineering \n", + "\n", + "### Time delay embedding features \n", + "\n", + "ConfirmedCases and Fatalities are transformed to the logarithmic scale ( log(1+x) to be precise ). \n", + "\n", + "Add log(1+x) - transformed number of new cases and new fatalities. Logarithmic scale is natural for this task as the growth of this values is typically exponential.\n", + "\n", + "Then add their historic values as (LogNewCases_prev_day_X, LogNewFataities_prev_day_X, X = 1 to 30) features.\n", + "\n", + "Some locations contain broken data (Confirmed Cases or Fatalities contain decreasing values), remove them from data." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "Eftr6BCP0M6y", + "colab_type": "code", + "outputId": "e183edfc-289b-4ae0-a009-32a950a05a87", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 679 + } + }, + "source": [ + "main_df.sort_values(by='Date', inplace=True)\n", + "\n", + "location_columns = ['Country/Region','Province/State']\n", + "\n", + "# row selection functions in pandas do not like nans\n", + "for column in location_columns:\n", + " main_df[column].fillna('', inplace=True)\n", + "\n", + "\n", + "days_history_size = 30\n", + "\n", + "\n", + "def is_cumulative(increment_series):\n", + " for v in increment_series:\n", + " if (not np.isnan(v)) and (v < 0):\n", + " return False\n", + " return True\n", + "\n", + "print ('data size before removing bad data = ', len(main_df))\n", + "\n", + "for field in ['LogNewConfirmedCases', 'LogNewFatalities']:\n", + " main_df[field] = np.nan\n", + "\n", + " for prev_day in range(1, days_history_size + 1):\n", + " main_df[field + '_prev_day_%s' % prev_day] = np.nan\n", + "\n", + "\n", + "for location_name, location_df in main_df.groupby(['Country/Region', 'Province/State']):\n", + " for field in ['ConfirmedCases', 'Fatalities']:\n", + " new_values = location_df[field].values\n", + " new_values[1:] -= new_values[:-1]\n", + " if not is_cumulative(new_values):\n", + " print ('%s for %s, %s is not valid cumulative series, drop it' % ((field,) + location_name))\n", + " main_df.drop(index=location_df.index, inplace=True)\n", + " break\n", + " log_new_values = np.log1p(new_values)\n", + " main_df.loc[location_df.index, 'LogNew' + field] = log_new_values\n", + "\n", + " for prev_day in range(1, days_history_size + 1):\n", + " main_df.loc[location_df.index[prev_day:], 'LogNew%s_prev_day_%s' % (field, prev_day)] = (\n", + " log_new_values[:-prev_day]\n", + " )\n", + "\n", + "print ('data size after removing bad data = ', len(main_df))\n", + "main_df.head()" + ], + "execution_count": 9, "outputs": [ { "output_type": "stream", "text": [ - "lightgbm version 2.3.1\n" + "data size before removing bad data = 26412\n", + "ConfirmedCases for Australia, Northern Territory is not valid cumulative series, drop it\n", + "ConfirmedCases for Australia, Queensland is not valid cumulative series, drop it\n", + "ConfirmedCases for Azerbaijan, is not valid cumulative series, drop it\n", + "ConfirmedCases for Bahrain, is not valid cumulative series, drop it\n", + "Fatalities for Canada, Quebec is not valid cumulative series, drop it\n", + "ConfirmedCases for China, Guizhou is not valid cumulative series, drop it\n", + "ConfirmedCases for France, France is not valid cumulative series, drop it\n", + "ConfirmedCases for France, Saint Barthelemy is not valid cumulative series, drop it\n", + "ConfirmedCases for Grand Princess, US is not valid cumulative series, drop it\n", + "ConfirmedCases for Guyana, is not valid cumulative series, drop it\n", + "Fatalities for Iceland, is not valid cumulative series, drop it\n", + "Fatalities for India, is not valid cumulative series, drop it\n", + "ConfirmedCases for Japan, is not valid cumulative series, drop it\n", + "Fatalities for Kazakhstan, is not valid cumulative series, drop it\n", + "ConfirmedCases for Lebanon, is not valid cumulative series, drop it\n", + "ConfirmedCases for Montenegro, is not valid cumulative series, drop it\n", + "Fatalities for Philippines, is not valid cumulative series, drop it\n", + "Fatalities for Russia, is not valid cumulative series, drop it\n", + "Fatalities for Slovakia, is not valid cumulative series, drop it\n", + "ConfirmedCases for US, Nevada is not valid cumulative series, drop it\n", + "ConfirmedCases for US, Utah is not valid cumulative series, drop it\n", + "ConfirmedCases for US, Washington is not valid cumulative series, drop it\n", + "data size after removing bad data = 24366\n" ], "name": "stdout" }, { - "output_type": "stream", - "text": [ - "/usr/local/lib/python3.6/dist-packages/lightgbm/basic.py:1295: UserWarning: categorical_feature in Dataset is overridden.\n", - "New categorical_feature is ['Country/Region', 'Province/State', 'WeekDay']\n", - " 'New categorical_feature is {}'.format(sorted(list(categorical_feature))))\n" - ], - "name": "stderr" - }, - { - "output_type": "stream", - "text": [ - "[100]\tvalid_0's rmse: 1.96629\n", - "[200]\tvalid_0's rmse: 1.87323\n", - "[300]\tvalid_0's rmse: 1.851\n", - "[400]\tvalid_0's rmse: 1.83537\n", - "[500]\tvalid_0's rmse: 1.82451\n", - "[600]\tvalid_0's rmse: 1.81892\n", - "[700]\tvalid_0's rmse: 1.81676\n", - "[800]\tvalid_0's rmse: 1.81196\n", - "[900]\tvalid_0's rmse: 1.81034\n", - "[1000]\tvalid_0's rmse: 1.80805\n", - "LightGBM: prediction of ConfirmedCases: RMSLE on test = 1.8080456616778966\n", - "[100]\tvalid_0's rmse: 0.90422\n", - "[200]\tvalid_0's rmse: 0.861511\n", - "[300]\tvalid_0's rmse: 0.843588\n", - "[400]\tvalid_0's rmse: 0.830418\n", - "[500]\tvalid_0's rmse: 0.824959\n", - "[600]\tvalid_0's rmse: 0.821098\n", - "[700]\tvalid_0's rmse: 0.818259\n", - "[800]\tvalid_0's rmse: 0.815991\n", - "[900]\tvalid_0's rmse: 0.813314\n", - "[1000]\tvalid_0's rmse: 0.811237\n", - "LightGBM: prediction of Fatalities: RMSLE on test = 0.8112365813255957\n" - ], - "name": "stdout" + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastIdLogNewConfirmedCasesLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30LogNewFatalitiesLogNewFatalities_prev_day_1LogNewFatalities_prev_day_2LogNewFatalities_prev_day_3LogNewFatalities_prev_day_4LogNewFatalities_prev_day_5LogNewFatalities_prev_day_6LogNewFatalities_prev_day_7LogNewFatalities_prev_day_8LogNewFatalities_prev_day_9LogNewFatalities_prev_day_10LogNewFatalities_prev_day_11LogNewFatalities_prev_day_12LogNewFatalities_prev_day_13LogNewFatalities_prev_day_14LogNewFatalities_prev_day_15LogNewFatalities_prev_day_16LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30
01.0Afghanistan33.000065.00002020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
55448185.0Cyprus35.126433.42992020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
1430121112.0GuamUS13.4443144.79372020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
56078278.0Czechia49.817515.47302020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
56708371.0DenmarkDenmark56.26399.50182020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
\n", + "
" + ], + "text/plain": [ + " Id ... LogNewFatalities_prev_day_30\n", + "0 1.0 ... NaN\n", + "5544 8185.0 ... NaN\n", + "14301 21112.0 ... NaN\n", + "5607 8278.0 ... NaN\n", + "5670 8371.0 ... NaN\n", + "\n", + "[5 rows x 71 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 9 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "GuxemylRhgSW", + "colab_type": "text" + }, + "source": [ + "### Day feature \n", + "\n", + "Now, add simple linear Day feature" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "qz8fWHG6QZdu", + "colab_type": "code", + "colab": {} + }, + "source": [ + "first_date = min(main_df['Date'])\n", + " \n", + "main_df['Day'] = (main_df['Date'] - first_date).dt.days.astype('int32')" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "yEWL2GAn58II", + "colab_type": "text" + }, + "source": [ + "### WeekDay feature \n", + "\n", + "Let's add WeekDay feature" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "Z7wNg-zD6BPn", + "colab_type": "code", + "colab": {} + }, + "source": [ + "main_df['WeekDay'] = main_df['Date'].transform(lambda d: d.weekday())" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "esXELzAHj77X", + "colab_type": "text" + }, + "source": [ + "### Days since Xth Confirmed case and Xth Fatality features \n", + "\n", + "Add Days since Xth Confirmed case and Xth Fatality (where X = 1, 10, 100) features" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "WUlHXEpRu0L1", + "colab_type": "code", + "colab": {} + }, + "source": [ + "thresholds = [1, 10, 100]\n", + "\n", + "for threshold in thresholds:\n", + " main_df['Days_since_ConfirmedCases=%s' % threshold] = np.nan\n", + " main_df['Days_since_Fatalities=%s' % threshold] = np.nan\n", + "\n", + "for location_name, location_df in main_df.groupby(['Country/Region', 'Province/State']):\n", + " for field in ['ConfirmedCases', 'Fatalities']:\n", + " for threshold in thresholds:\n", + " first_day = location_df['Day'].loc[location_df[field] >= threshold].min()\n", + " if not np.isnan(first_day):\n", + " main_df.loc[location_df.index, 'Days_since_%s=%s' % (field, threshold)] = (\n", + " location_df['Day'].transform(lambda day: -1 if (day < first_day) else (day - first_day))\n", + " )" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "znNwtyobuiYv", + "colab_type": "text" + }, + "source": [ + "### 'Distance to Origin' feature. \n", + "\n", + "Chinese province of Hubei is the original epicentre of the pandemic, so add a feature that represents distance from the region of the sample to the epidemic's origin." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "SpdzCOMzu8GV", + "colab_type": "code", + "colab": {} + }, + "source": [ + "def get_hubei_coords(df):\n", + " for index, row in df.iterrows():\n", + " if row['Province/State'] == 'Hubei':\n", + " return (row['Lat'], row['Long'])\n", + "\n", + " raise Exception('Hubei not found in data')\n", + "\n", + "\n", + "origin_coords = get_hubei_coords(main_df)\n", + "\n", + "main_df['Distance_to_origin'] = main_df.apply(\n", + " lambda row: geopy.distance.distance((row['Lat'], row['Long']), origin_coords).km,\n", + " axis='columns'\n", + ")\n" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "gQD43i1oyMZN", + "colab_type": "text" + }, + "source": [ + "### Functions for merging external data \n", + "\n", + "\n", + "Let's add some data from other sources.\n", + "\n", + "But first, we will need a couple of helper functions with common functionality.\n", + "\n", + "World bank data, UN World Population Project data and the main dataset use different names for some of the countries so we will need to remap country names in external datasets for correct merging." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "ZLcmIiTOydcJ", + "colab_type": "code", + "colab": {} + }, + "source": [ + "def merge_with_column_drop(left_df, right_df, right_df_column='Country'):\n", + " df = pd.merge(\n", + " left=left_df,\n", + " right=right_df,\n", + " how='left',\n", + " left_on='Country/Region',\n", + " right_on=right_df_column\n", + " )\n", + " df.drop(columns=right_df_column, inplace=True)\n", + " \n", + " return df\n", + "\n", + "def remap_country_name_from_world_bank_to_main_df_name(country):\n", + " return {\n", + " 'Bahamas, The': 'The Bahamas',\n", + " 'Brunei Darussalam': 'Brunei',\n", + " 'Congo, Rep.': 'Congo (Brazzaville)',\n", + " 'Congo, Dem. Rep.': 'Congo (Kinshasa)',\n", + " 'Czech Republic': 'Czechia',\n", + " 'Egypt, Arab Rep.': 'Egypt',\n", + " 'Iran, Islamic Rep.': 'Iran',\n", + " 'Korea, Rep.': 'Korea, South',\n", + " 'Kyrgyz Republic': 'Kyrgyzstan',\n", + " 'Russian Federation': 'Russia',\n", + " 'Slovak Republic': 'Slovakia',\n", + " 'St. Lucia': 'Saint Lucia',\n", + " 'St. Vincent and the Grenadines': 'Saint Vincent and the Grenadines',\n", + " 'United States': 'US',\n", + " 'Venezuela, RB': 'Venezuela',\n", + " }.get(country, country)\n", + "\n", + "def remap_country_name_from_un_wpp_to_main_df_name(country):\n", + " return {\n", + " 'Bahamas': 'The Bahamas',\n", + " 'Bolivia (Plurinational State of)': 'Bolivia',\n", + " 'Brunei Darussalam': 'Brunei',\n", + " 'China, Taiwan Province of China': 'Taiwan*',\n", + " 'Congo' : 'Congo (Brazzaville)',\n", + " 'Côte d\\'Ivoire': 'Cote d\\'Ivoire',\n", + " 'Democratic Republic of the Congo': 'Congo (Kinshasa)',\n", + " 'Gambia': 'The Gambia',\n", + " 'Iran (Islamic Republic of)': 'Iran',\n", + " 'Republic of Korea': 'Korea, South',\n", + " 'Republic of Moldova': 'Moldova',\n", + " 'Réunion': 'Reunion',\n", + " 'Russian Federation': 'Russia',\n", + " 'United Republic of Tanzania': 'Tanzania',\n", + " 'United States of America': 'US',\n", + " 'Venezuela (Bolivarian Republic of)': 'Venezuela',\n", + " 'Viet Nam': 'Vietnam'\n", + " }.get(country, country)\n", + "\n", + "# for read_csv\n", + "world_bank_converters={'Country Name': remap_country_name_from_world_bank_to_main_df_name}\n", + "un_wpp_converters={'Location': remap_country_name_from_un_wpp_to_main_df_name}" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "RZrD5vlexX7_", + "colab_type": "text" + }, + "source": [ + "### Country Area \n", + "\n", + "Let's add Country Area feature. We will use data from World Bank for that.\n", + "\n", + "\n" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "Yi6xbUz4qMP2", + "colab_type": "code", + "outputId": "cfff6fd6-95da-4b9e-f47e-ac75eed0a948", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 394 + } + }, + "source": [ + "!mkdir -p area\n", + "os.chdir('area')\n", + "!wget -nc http://api.worldbank.org/v2/en/indicator/AG.LND.TOTL.K2?downloadformat=csv -O area.zip\n", + "!unzip -n area.zip \n", + "!ls -l ." + ], + "execution_count": 15, + "outputs": [ + { + "output_type": "stream", + "text": [ + "--2020-04-19 14:10:51-- http://api.worldbank.org/v2/en/indicator/AG.LND.TOTL.K2?downloadformat=csv\n", + "Resolving api.worldbank.org (api.worldbank.org)... 34.237.118.134\n", + "Connecting to api.worldbank.org (api.worldbank.org)|34.237.118.134|:80... connected.\n", + "HTTP request sent, awaiting response... 200 OK\n", + "Length: 19817 (19K) [application/zip]\n", + "Saving to: ‘area.zip’\n", + "\n", + "\rarea.zip 0%[ ] 0 --.-KB/s \rarea.zip 100%[===================>] 19.35K --.-KB/s in 0.01s \n", + "\n", + "2020-04-19 14:10:51 (1.48 MB/s) - ‘area.zip’ saved [19817/19817]\n", + "\n", + "Archive: area.zip\n", + " inflating: Metadata_Indicator_API_AG.LND.TOTL.K2_DS2_en_csv_v2_937279.csv \n", + " inflating: API_AG.LND.TOTL.K2_DS2_en_csv_v2_937279.csv \n", + " inflating: Metadata_Country_API_AG.LND.TOTL.K2_DS2_en_csv_v2_937279.csv \n", + "total 224\n", + "-rw-r--r-- 1 root root 169544 Apr 9 18:44 API_AG.LND.TOTL.K2_DS2_en_csv_v2_937279.csv\n", + "-rw-r--r-- 1 root root 19817 Apr 19 14:10 area.zip\n", + "-rw-r--r-- 1 root root 30906 Apr 9 18:44 Metadata_Country_API_AG.LND.TOTL.K2_DS2_en_csv_v2_937279.csv\n", + "-rw-r--r-- 1 root root 417 Apr 9 18:44 Metadata_Indicator_API_AG.LND.TOTL.K2_DS2_en_csv_v2_937279.csv\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "sT7s0ACrtkqP", + "colab_type": "code", + "outputId": "786654b1-84d2-44ea-e0b2-594a6870bebe", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 226 + } + }, + "source": [ + "!head API_AG.LND.TOTL.K2_DS2_en_csv_v2_937279.csv\n" + ], + "execution_count": 16, + "outputs": [ + { + "output_type": "stream", + "text": [ + "\"Data Source\",\"World Development Indicators\",\r\n", + "\r\n", + "\"Last Updated Date\",\"2020-04-09\",\r\n", + "\r\n", + "\"Country Name\",\"Country Code\",\"Indicator Name\",\"Indicator Code\",\"1960\",\"1961\",\"1962\",\"1963\",\"1964\",\"1965\",\"1966\",\"1967\",\"1968\",\"1969\",\"1970\",\"1971\",\"1972\",\"1973\",\"1974\",\"1975\",\"1976\",\"1977\",\"1978\",\"1979\",\"1980\",\"1981\",\"1982\",\"1983\",\"1984\",\"1985\",\"1986\",\"1987\",\"1988\",\"1989\",\"1990\",\"1991\",\"1992\",\"1993\",\"1994\",\"1995\",\"1996\",\"1997\",\"1998\",\"1999\",\"2000\",\"2001\",\"2002\",\"2003\",\"2004\",\"2005\",\"2006\",\"2007\",\"2008\",\"2009\",\"2010\",\"2011\",\"2012\",\"2013\",\"2014\",\"2015\",\"2016\",\"2017\",\"2018\",\"2019\",\r\n", + "\"Aruba\",\"ABW\",\"Land area (sq. km)\",\"AG.LND.TOTL.K2\",\"\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"180\",\"\",\r\n", + "\"Afghanistan\",\"AFG\",\"Land area (sq. km)\",\"AG.LND.TOTL.K2\",\"\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"652860\",\"\",\r\n", + "\"Angola\",\"AGO\",\"Land area (sq. km)\",\"AG.LND.TOTL.K2\",\"\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"1246700\",\"\",\r\n", + "\"Albania\",\"ALB\",\"Land area (sq. km)\",\"AG.LND.TOTL.K2\",\"\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"27400\",\"\",\r\n", + "\"Andorra\",\"AND\",\"Land area (sq. km)\",\"AG.LND.TOTL.K2\",\"\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"470\",\"\",\r\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "-bGeu_fzr5Wp", + "colab_type": "code", + "outputId": "9501aff1-64ef-4666-cec2-70949c990311", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 330 + } + }, + "source": [ + "area_df = pd.read_csv('./API_AG.LND.TOTL.K2_DS2_en_csv_v2_937279.csv', skiprows=4, converters=world_bank_converters)\n", + "os.chdir('..')\n", + "area_df.head()" + ], + "execution_count": 17, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Country NameCountry CodeIndicator NameIndicator Code196019611962196319641965196619671968196919701971197219731974197519761977197819791980198119821983198419851986198719881989199019911992199319941995199619971998199920002001200220032004200520062007200820092010201120122013201420152016201720182019Unnamed: 64
0ArubaABWLand area (sq. km)AG.LND.TOTL.K2NaN180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0180.0NaNNaN
1AfghanistanAFGLand area (sq. km)AG.LND.TOTL.K2NaN652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0652860.0NaNNaN
2AngolaAGOLand area (sq. km)AG.LND.TOTL.K2NaN1246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.01246700.0NaNNaN
3AlbaniaALBLand area (sq. km)AG.LND.TOTL.K2NaN27400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.027400.0NaNNaN
4AndorraANDLand area (sq. km)AG.LND.TOTL.K2NaN470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0470.0NaNNaN
\n", + "
" + ], + "text/plain": [ + " Country Name Country Code Indicator Name ... 2018 2019 Unnamed: 64\n", + "0 Aruba ABW Land area (sq. km) ... 180.0 NaN NaN\n", + "1 Afghanistan AFG Land area (sq. km) ... 652860.0 NaN NaN\n", + "2 Angola AGO Land area (sq. km) ... 1246700.0 NaN NaN\n", + "3 Albania ALB Land area (sq. km) ... 27400.0 NaN NaN\n", + "4 Andorra AND Land area (sq. km) ... 470.0 NaN NaN\n", + "\n", + "[5 rows x 65 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 17 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "A-bpLJUxrt7v", + "colab_type": "text" + }, + "source": [ + "We will use the last available data. " + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "7usYQo_RuDCq", + "colab_type": "code", + "colab": {} + }, + "source": [ + "year_columns = [str(year) for year in range(1960, 2020)]\n", + "\n", + "area_df['CountryArea'] = area_df[year_columns].apply(\n", + " lambda row: row[row.last_valid_index()] if row.last_valid_index() else np.nan,\n", + " axis='columns' \n", + ")" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Z9jzp0rDyybx", + "colab_type": "text" + }, + "source": [ + "Drop unneeded columns" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "IUbGQTqJyw1g", + "colab_type": "code", + "outputId": "5bb7eee3-d22e-45a6-e265-5ef2be8231fc", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + } + }, + "source": [ + "area_df = area_df[['Country Name', 'CountryArea']]\n", + "area_df.head()" + ], + "execution_count": 19, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Country NameCountryArea
0Aruba180.0
1Afghanistan652860.0
2Angola1246700.0
3Albania27400.0
4Andorra470.0
\n", + "
" + ], + "text/plain": [ + " Country Name CountryArea\n", + "0 Aruba 180.0\n", + "1 Afghanistan 652860.0\n", + "2 Angola 1246700.0\n", + "3 Albania 27400.0\n", + "4 Andorra 470.0" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 19 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "TEwyah4Nw77K", + "colab_type": "text" + }, + "source": [ + "Now, merge this data with main data by country.\n", + "\n" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "H1XPT-_OyjV4", + "colab_type": "code", + "outputId": "5aec2481-30bd-40e2-fa57-afc32ae3ae42", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 256 + } + }, + "source": [ + "main_df = merge_with_column_drop(\n", + " main_df,\n", + " area_df,\n", + " right_df_column='Country Name'\n", + ")\n", + "main_df.head()" + ], + "execution_count": 20, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastIdLogNewConfirmedCasesLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30...LogNewFatalities_prev_day_1LogNewFatalities_prev_day_2LogNewFatalities_prev_day_3LogNewFatalities_prev_day_4LogNewFatalities_prev_day_5LogNewFatalities_prev_day_6LogNewFatalities_prev_day_7LogNewFatalities_prev_day_8LogNewFatalities_prev_day_9LogNewFatalities_prev_day_10LogNewFatalities_prev_day_11LogNewFatalities_prev_day_12LogNewFatalities_prev_day_13LogNewFatalities_prev_day_14LogNewFatalities_prev_day_15LogNewFatalities_prev_day_16LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30DayWeekDayDays_since_ConfirmedCases=1Days_since_Fatalities=1Days_since_ConfirmedCases=10Days_since_Fatalities=10Days_since_ConfirmedCases=100Days_since_Fatalities=100Distance_to_originCountryArea
01.0Afghanistan33.000065.00002020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN4435.232618652860.0
18185.0Cyprus35.126433.42992020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaN-1.0NaN7177.5291959240.0
221112.0GuamUS13.4443144.79372020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN3851.9426899147420.0
38278.0Czechia49.817515.47302020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02NaNNaNNaNNaNNaNNaN7897.62403877220.0
48371.0DenmarkDenmark56.26399.50182020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0-1.0-1.0NaN7931.06918641990.0
\n", + "

5 rows × 81 columns

\n", + "
" + ], + "text/plain": [ + " Id Province/State ... Distance_to_origin CountryArea\n", + "0 1.0 ... 4435.232618 652860.0\n", + "1 8185.0 ... 7177.529195 9240.0\n", + "2 21112.0 Guam ... 3851.942689 9147420.0\n", + "3 8278.0 ... 7897.624038 77220.0\n", + "4 8371.0 Denmark ... 7931.069186 41990.0\n", + "\n", + "[5 rows x 81 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 20 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "qLqnNUL5zriP", + "colab_type": "text" + }, + "source": [ + "### Country population features \n", + "\n", + "Now, let's add country population data from UN." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "sI11ale0O2EZ", + "colab_type": "code", + "outputId": "0de09604-4414-4309-9357-fb3ecc9b902f", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 245 + } + }, + "source": [ + "!mkdir -p population\n", + "os.chdir('population')\n", + "!wget -nc https://population.un.org/wpp/Download/Files/1_Indicators%20\\(Standard\\)/CSV_FILES/WPP2019_PopulationByAgeSex_Medium.csv\n" + ], + "execution_count": 21, + "outputs": [ + { + "output_type": "stream", + "text": [ + "--2020-04-19 14:10:59-- https://population.un.org/wpp/Download/Files/1_Indicators%20(Standard)/CSV_FILES/WPP2019_PopulationByAgeSex_Medium.csv\n", + "Resolving population.un.org (population.un.org)... 157.150.185.69\n", + "Connecting to population.un.org (population.un.org)|157.150.185.69|:443... connected.\n", + "HTTP request sent, awaiting response... 200 OK\n", + "Length: 118543941 (113M) [application/octet-stream]\n", + "Saving to: ‘WPP2019_PopulationByAgeSex_Medium.csv’\n", + "\n", + "WPP2019_PopulationB 100%[===================>] 113.05M 5.18MB/s in 22s \n", + "\n", + "2020-04-19 14:11:21 (5.25 MB/s) - ‘WPP2019_PopulationByAgeSex_Medium.csv’ saved [118543941/118543941]\n", + "\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "MhD5GFYHYdUA", + "colab_type": "code", + "outputId": "be1bfea0-63a1-43e9-b7ac-cc0dc152dc4e", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + } + }, + "source": [ + "!head ./WPP2019_PopulationByAgeSex_Medium.csv" + ], + "execution_count": 22, + "outputs": [ + { + "output_type": "stream", + "text": [ + "LocID,Location,VarID,Variant,Time,MidPeriod,AgeGrp,AgeGrpStart,AgeGrpSpan,PopMale,PopFemale,PopTotal\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,0-4,0,5,630.044,661.578,1291.622\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,5-9,5,5,516.206,487.335,1003.541\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,10-14,10,5,461.378,423.326,884.704\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,15-19,15,5,414.369,369.363,783.732\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,20-24,20,5,374.109,318.392,692.501\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,25-29,25,5,321.311,272.299,593.61\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,30-34,30,5,276.279,232.168,508.447\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,35-39,35,5,236.792,197.326,434.118\r\n", + "4,Afghanistan,2,Medium,1950,1950.5,40-44,40,5,200.616,167.179,367.795\r\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "BS9ypYhoYndR", + "colab_type": "code", + "outputId": "bcec3c06-0991-4269-936b-92a464fef85e", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + } + }, + "source": [ + "population_df = pd.read_csv(\n", + " 'WPP2019_PopulationByAgeSex_Medium.csv',\n", + " usecols=['Location', 'Time', 'AgeGrp', 'PopMale', 'PopFemale', 'PopTotal'],\n", + " parse_dates=['Time'],\n", + " converters=un_wpp_converters\n", + ")\n", + "os.chdir('..')\n", + "population_df.head()" + ], + "execution_count": 23, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
LocationTimeAgeGrpPopMalePopFemalePopTotal
0Afghanistan1950-01-010-4630.044661.5781291.622
1Afghanistan1950-01-015-9516.206487.3351003.541
2Afghanistan1950-01-0110-14461.378423.326884.704
3Afghanistan1950-01-0115-19414.369369.363783.732
4Afghanistan1950-01-0120-24374.109318.392692.501
\n", + "
" + ], + "text/plain": [ + " Location Time AgeGrp PopMale PopFemale PopTotal\n", + "0 Afghanistan 1950-01-01 0-4 630.044 661.578 1291.622\n", + "1 Afghanistan 1950-01-01 5-9 516.206 487.335 1003.541\n", + "2 Afghanistan 1950-01-01 10-14 461.378 423.326 884.704\n", + "3 Afghanistan 1950-01-01 15-19 414.369 369.363 783.732\n", + "4 Afghanistan 1950-01-01 20-24 374.109 318.392 692.501" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 23 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "m4ONP3hG2UXd", + "colab_type": "text" + }, + "source": [ + "We will use only recent (no earlier than 2014) data\n", + "\n" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "C5Hq9KMHhBQy", + "colab_type": "code", + "outputId": "4cfb40b8-dc76-46a9-ec31-7e0af4d39b45", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + } + }, + "source": [ + "population_df = population_df.loc[\n", + " (population_df['Time'] >= pd.Timestamp(2014,1,1))\n", + " & (population_df['Time'] <= pd.Timestamp(2019,1,1))\n", + "]\n", + "population_df.head()" + ], + "execution_count": 24, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
LocationTimeAgeGrpPopMalePopFemalePopTotal
1344Afghanistan2014-01-010-42803.6772671.5835475.260
1345Afghanistan2014-01-015-92635.8122520.8405156.652
1346Afghanistan2014-01-0110-142356.5242242.1854598.709
1347Afghanistan2014-01-0115-191995.8241894.3803890.204
1348Afghanistan2014-01-0120-241580.8811458.1993039.080
\n", + "
" + ], + "text/plain": [ + " Location Time AgeGrp PopMale PopFemale PopTotal\n", + "1344 Afghanistan 2014-01-01 0-4 2803.677 2671.583 5475.260\n", + "1345 Afghanistan 2014-01-01 5-9 2635.812 2520.840 5156.652\n", + "1346 Afghanistan 2014-01-01 10-14 2356.524 2242.185 4598.709\n", + "1347 Afghanistan 2014-01-01 15-19 1995.824 1894.380 3890.204\n", + "1348 Afghanistan 2014-01-01 20-24 1580.881 1458.199 3039.080" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 24 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "pJEpvz333IKB", + "colab_type": "text" + }, + "source": [ + "Aggregate data by 20-year groups and also compute cumulative population." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "U-AxHnOehRXS", + "colab_type": "code", + "outputId": "b49f8a53-9b6e-4133-eb6f-5af7f90efb1e", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 226 + } + }, + "source": [ + "aggegated_population_df = pd.DataFrame()\n", + "\n", + "for (location, time), group_df in population_df.groupby(['Location', 'Time']):\n", + " # by ['Pop_0-20', 'Pop_20-40', 'Pop_40-60', 'Pop_60-80', 'Pop_80+']\n", + " pop_by_age_groups = [0] * 5 \n", + "\n", + " pop_male = 0\n", + " pop_female = 0\n", + "\n", + " for _, row in group_df.iterrows():\n", + " age_grp_start = int( re.split(r'[\\-\\+]', row['AgeGrp'])[0] )\n", + " pop_by_age_groups[min(age_grp_start // 20, 4)] += row['PopMale'] + row['PopFemale']\n", + " pop_male += row['PopMale']\n", + " pop_female += row['PopFemale']\n", + "\n", + " \"\"\n", + " aggegated_population_df = aggegated_population_df.append(\n", + " {\n", + " 'Location': location,\n", + " 'Time': time,\n", + " 'CountryPop_0-20': pop_by_age_groups[0],\n", + " 'CountryPop_20-40': pop_by_age_groups[1],\n", + " 'CountryPop_40-60': pop_by_age_groups[2],\n", + " 'CountryPop_60-80': pop_by_age_groups[3],\n", + " 'CountryPop_80+': pop_by_age_groups[4],\n", + " 'CountryPopMale': pop_male,\n", + " 'CountryPopFemale': pop_female,\n", + " 'CountryPopTotal': pop_male + pop_female\n", + " },\n", + " ignore_index=True \n", + " )\n", + "\n", + "aggegated_population_df.head()" + ], + "execution_count": 25, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
CountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+LocationTime
016232.00117138.80333370.80419120.8259120.2693811.9381232.31385.459Afghanistan2014-01-01
116727.43717686.16634413.60319506.3679568.8383970.1491282.69785.552Afghanistan2015-01-01
217196.03418186.99435383.02819860.5439972.1884127.6261330.04092.631Afghanistan2016-01-01
317644.12618651.98536296.11120165.54610368.6004286.1911377.37698.398Afghanistan2017-01-01
418078.87919093.04337171.92220432.14910761.9314448.7951426.468102.579Afghanistan2018-01-01
\n", + "
" + ], + "text/plain": [ + " CountryPopFemale CountryPopMale ... Location Time\n", + "0 16232.001 17138.803 ... Afghanistan 2014-01-01\n", + "1 16727.437 17686.166 ... Afghanistan 2015-01-01\n", + "2 17196.034 18186.994 ... Afghanistan 2016-01-01\n", + "3 17644.126 18651.985 ... Afghanistan 2017-01-01\n", + "4 18078.879 19093.043 ... Afghanistan 2018-01-01\n", + "\n", + "[5 rows x 10 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 25 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "3KJP1_TxA2bs", + "colab_type": "text" + }, + "source": [ + "Select the most recent data." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "rNLu6PiBAxDM", + "colab_type": "code", + "outputId": "3acde16d-fcd9-477b-c027-b56b9d897c94", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 417 + } + }, + "source": [ + "aggegated_population_df = aggegated_population_df.sort_values('Time').drop_duplicates(['Location'], keep='last')\n", + "aggegated_population_df.head()" + ], + "execution_count": 26, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
CountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+LocationTime
241734178.72633351.43567530.16115634.44317834.18317763.04412879.9323418.559United Kingdom2019-01-01
2519193337.968196415.223389753.191153429.571127412.45075929.06829424.8423557.260WB region: Middle East and North Africa (exclu...2019-01-01
2375265709.783268213.492533923.275290642.723149895.25569210.04422655.7581519.495UNICEF Regions: West and Central Africa2019-01-01
262714469.63814692.28429161.92214585.7669617.1393621.9701236.095100.952Yemen2019-01-01
26339017.8208843.21417861.0349995.3805186.6382074.569559.20545.242Zambia2019-01-01
\n", + "
" + ], + "text/plain": [ + " CountryPopFemale ... Time\n", + "2417 34178.726 ... 2019-01-01\n", + "2519 193337.968 ... 2019-01-01\n", + "2375 265709.783 ... 2019-01-01\n", + "2627 14469.638 ... 2019-01-01\n", + "2633 9017.820 ... 2019-01-01\n", + "\n", + "[5 rows x 10 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 26 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "YtsznEHyEJP_", + "colab_type": "text" + }, + "source": [ + "Now, drop Time and join with main data" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "5BzM29dEhX0y", + "colab_type": "code", + "outputId": "82de3201-d886-4e1b-e39f-b34cf10e3f04", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 256 + } + }, + "source": [ + "aggegated_population_df.drop(columns='Time', inplace=True)\n", + "\n", + "main_df = merge_with_column_drop(\n", + " main_df,\n", + " aggegated_population_df,\n", + " right_df_column='Location'\n", + ")\n", + "main_df.head()" + ], + "execution_count": 27, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastIdLogNewConfirmedCasesLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30...LogNewFatalities_prev_day_9LogNewFatalities_prev_day_10LogNewFatalities_prev_day_11LogNewFatalities_prev_day_12LogNewFatalities_prev_day_13LogNewFatalities_prev_day_14LogNewFatalities_prev_day_15LogNewFatalities_prev_day_16LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30DayWeekDayDays_since_ConfirmedCases=1Days_since_Fatalities=1Days_since_ConfirmedCases=10Days_since_Fatalities=10Days_since_ConfirmedCases=100Days_since_Fatalities=100Distance_to_originCountryAreaCountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+
01.0Afghanistan33.000065.00002020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN4435.232618652860.018512.03019529.72738041.75720676.36511160.5114620.2681479.526105.087
18185.0Cyprus35.126433.42992020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaN-1.0NaN7177.5291959240.0599.277599.2971198.574275.284382.653308.911193.20238.524
221112.0GuamUS13.4443144.79372020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN3851.9426899147420.0166238.618162826.299329064.91782289.91690159.13482846.12660854.33212915.409
38278.0Czechia49.817515.47302020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02NaNNaNNaNNaNNaNNaN7897.62403877220.05428.5855260.62810689.2132152.9232656.3403107.3602332.215440.375
48371.0DenmarkDenmark56.26399.50182020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0-1.0-1.0NaN7931.06918641990.02902.2052869.6725771.8771287.8641437.5661556.0361224.130266.281
\n", + "

5 rows × 89 columns

\n", + "
" + ], + "text/plain": [ + " Id Province/State ... CountryPop_60-80 CountryPop_80+\n", + "0 1.0 ... 1479.526 105.087\n", + "1 8185.0 ... 193.202 38.524\n", + "2 21112.0 Guam ... 60854.332 12915.409\n", + "3 8278.0 ... 2332.215 440.375\n", + "4 8371.0 Denmark ... 1224.130 266.281\n", + "\n", + "[5 rows x 89 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 27 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Qd3UQl7XHGkf", + "colab_type": "text" + }, + "source": [ + "### Country Population density feature \n", + "\n", + "With both Area and Population data we can now compute Population density feature." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "HlguvbyUHR77", + "colab_type": "code", + "colab": {} + }, + "source": [ + "main_df['CountryPopDensity'] = main_df['CountryPopTotal'] / main_df['CountryArea']" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "17THKbaKIUot", + "colab_type": "text" + }, + "source": [ + "### Country Smoking rate feature \n", + "\n", + "Let's also add Smoking rate by Country data because smoking can influence the severity of respiratory diseases." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "uK8RsydoJA4w", + "colab_type": "code", + "outputId": "1c75995c-9a52-478c-82f2-a611a6fe6e16", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 394 + } + }, + "source": [ + "!mkdir -p smoking\n", + "os.chdir('smoking')\n", + "!wget -nc http://api.worldbank.org/v2/en/indicator/SH.PRV.SMOK?downloadformat=csv -O smoking.zip\n", + "!unzip -n smoking.zip \n", + "!ls -l ." + ], + "execution_count": 29, + "outputs": [ + { + "output_type": "stream", + "text": [ + "--2020-04-19 14:11:51-- http://api.worldbank.org/v2/en/indicator/SH.PRV.SMOK?downloadformat=csv\n", + "Resolving api.worldbank.org (api.worldbank.org)... 107.23.127.3\n", + "Connecting to api.worldbank.org (api.worldbank.org)|107.23.127.3|:80... connected.\n", + "HTTP request sent, awaiting response... 200 OK\n", + "Length: 18137 (18K) [application/zip]\n", + "Saving to: ‘smoking.zip’\n", + "\n", + "\rsmoking.zip 0%[ ] 0 --.-KB/s \rsmoking.zip 100%[===================>] 17.71K --.-KB/s in 0.01s \n", + "\n", + "2020-04-19 14:11:51 (1.41 MB/s) - ‘smoking.zip’ saved [18137/18137]\n", + "\n", + "Archive: smoking.zip\n", + " inflating: Metadata_Indicator_API_SH.PRV.SMOK_DS2_en_csv_v2_937317.csv \n", + " inflating: API_SH.PRV.SMOK_DS2_en_csv_v2_937317.csv \n", + " inflating: Metadata_Country_API_SH.PRV.SMOK_DS2_en_csv_v2_937317.csv \n", + "total 136\n", + "-rw-r--r-- 1 root root 78878 Apr 9 18:50 API_SH.PRV.SMOK_DS2_en_csv_v2_937317.csv\n", + "-rw-r--r-- 1 root root 30906 Apr 9 18:50 Metadata_Country_API_SH.PRV.SMOK_DS2_en_csv_v2_937317.csv\n", + "-rw-r--r-- 1 root root 447 Apr 9 18:50 Metadata_Indicator_API_SH.PRV.SMOK_DS2_en_csv_v2_937317.csv\n", + "-rw-r--r-- 1 root root 18137 Apr 19 14:11 smoking.zip\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "1waTPJazJhR-", + "colab_type": "code", + "outputId": "1c9a78fc-7cb4-46fd-9a62-62ab4846a1d4", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 226 + } + }, + "source": [ + "!head API_SH.PRV.SMOK_DS2_en_csv_v2_937317.csv" + ], + "execution_count": 30, + "outputs": [ + { + "output_type": "stream", + "text": [ + "\"Data Source\",\"World Development Indicators\",\r\n", + "\r\n", + "\"Last Updated Date\",\"2020-04-09\",\r\n", + "\r\n", + "\"Country Name\",\"Country Code\",\"Indicator Name\",\"Indicator Code\",\"1960\",\"1961\",\"1962\",\"1963\",\"1964\",\"1965\",\"1966\",\"1967\",\"1968\",\"1969\",\"1970\",\"1971\",\"1972\",\"1973\",\"1974\",\"1975\",\"1976\",\"1977\",\"1978\",\"1979\",\"1980\",\"1981\",\"1982\",\"1983\",\"1984\",\"1985\",\"1986\",\"1987\",\"1988\",\"1989\",\"1990\",\"1991\",\"1992\",\"1993\",\"1994\",\"1995\",\"1996\",\"1997\",\"1998\",\"1999\",\"2000\",\"2001\",\"2002\",\"2003\",\"2004\",\"2005\",\"2006\",\"2007\",\"2008\",\"2009\",\"2010\",\"2011\",\"2012\",\"2013\",\"2014\",\"2015\",\"2016\",\"2017\",\"2018\",\"2019\",\r\n", + "\"Aruba\",\"ABW\",\"Smoking prevalence, total (ages 15+)\",\"SH.PRV.SMOK\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n", + "\"Afghanistan\",\"AFG\",\"Smoking prevalence, total (ages 15+)\",\"SH.PRV.SMOK\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n", + "\"Angola\",\"AGO\",\"Smoking prevalence, total (ages 15+)\",\"SH.PRV.SMOK\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n", + "\"Albania\",\"ALB\",\"Smoking prevalence, total (ages 15+)\",\"SH.PRV.SMOK\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"34.8\",\"\",\"\",\"\",\"\",\"32.7\",\"\",\"\",\"\",\"\",\"31.2\",\"30.7\",\"30.2\",\"29.8\",\"29.5\",\"29.1\",\"28.7\",\"\",\"\",\"\",\r\n", + "\"Andorra\",\"AND\",\"Smoking prevalence, total (ages 15+)\",\"SH.PRV.SMOK\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"37.4\",\"\",\"\",\"\",\"\",\"36\",\"\",\"\",\"\",\"\",\"34.8\",\"34.5\",\"34.2\",\"34.1\",\"33.9\",\"33.7\",\"33.5\",\"\",\"\",\"\",\r\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "DsUc9YiDJpER", + "colab_type": "code", + "outputId": "0cfcb140-610b-4c8f-e5d3-75b910e11b42", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 504 + } + }, + "source": [ + "smoking_df = pd.read_csv('./API_SH.PRV.SMOK_DS2_en_csv_v2_937317.csv', skiprows=4)\n", + "os.chdir('..')\n", + "smoking_df.head()" + ], + "execution_count": 31, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Country NameCountry CodeIndicator NameIndicator Code196019611962196319641965196619671968196919701971197219731974197519761977197819791980198119821983198419851986198719881989199019911992199319941995199619971998199920002001200220032004200520062007200820092010201120122013201420152016201720182019Unnamed: 64
0ArubaABWSmoking prevalence, total (ages 15+)SH.PRV.SMOKNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
1AfghanistanAFGSmoking prevalence, total (ages 15+)SH.PRV.SMOKNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
2AngolaAGOSmoking prevalence, total (ages 15+)SH.PRV.SMOKNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
3AlbaniaALBSmoking prevalence, total (ages 15+)SH.PRV.SMOKNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN34.8NaNNaNNaNNaN32.7NaNNaNNaNNaN31.230.730.229.829.529.128.7NaNNaNNaNNaN
4AndorraANDSmoking prevalence, total (ages 15+)SH.PRV.SMOKNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN37.4NaNNaNNaNNaN36.0NaNNaNNaNNaN34.834.534.234.133.933.733.5NaNNaNNaNNaN
\n", + "
" + ], + "text/plain": [ + " Country Name Country Code ... 2019 Unnamed: 64\n", + "0 Aruba ABW ... NaN NaN\n", + "1 Afghanistan AFG ... NaN NaN\n", + "2 Angola AGO ... NaN NaN\n", + "3 Albania ALB ... NaN NaN\n", + "4 Andorra AND ... NaN NaN\n", + "\n", + "[5 rows x 65 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 31 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "DimXLlG_J-o7", + "colab_type": "text" + }, + "source": [ + "We will use the last available data from recent years (2010 and later). " + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "wqfNkQuwKGEh", + "colab_type": "code", + "colab": {} + }, + "source": [ + "recent_year_columns = [str(year) for year in range(2010, 2020)]\n", + "\n", + "smoking_df['CountrySmokingRate'] = smoking_df[recent_year_columns].apply(\n", + " lambda row: row[row.last_valid_index()] if row.last_valid_index() else np.nan,\n", + " axis='columns' \n", + ")" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "8f2SYEB-KmWv", + "colab_type": "text" + }, + "source": [ + "Drop unneeded columns" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "ZEMpoYp5Kqlk", + "colab_type": "code", + "outputId": "d2cb66b5-c5d9-44d3-ed98-5230bf3539cc", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + } + }, + "source": [ + "smoking_df = smoking_df[['Country Name', 'CountrySmokingRate']]\n", + "smoking_df.head()" + ], + "execution_count": 33, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Country NameCountrySmokingRate
0ArubaNaN
1AfghanistanNaN
2AngolaNaN
3Albania28.7
4Andorra33.5
\n", + "
" + ], + "text/plain": [ + " Country Name CountrySmokingRate\n", + "0 Aruba NaN\n", + "1 Afghanistan NaN\n", + "2 Angola NaN\n", + "3 Albania 28.7\n", + "4 Andorra 33.5" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 33 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "40PrK13_LGT7", + "colab_type": "text" + }, + "source": [ + "Now, merge this data with main data by country." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "8Mwoe7nBLIww", + "colab_type": "code", + "outputId": "b3c3ca22-c569-4ea6-e2bf-b493d315909b", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 256 + } + }, + "source": [ + "main_df = merge_with_column_drop(\n", + " main_df,\n", + " smoking_df,\n", + " right_df_column='Country Name'\n", + ")\n", + "main_df.head()" + ], + "execution_count": 34, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastIdLogNewConfirmedCasesLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30...LogNewFatalities_prev_day_11LogNewFatalities_prev_day_12LogNewFatalities_prev_day_13LogNewFatalities_prev_day_14LogNewFatalities_prev_day_15LogNewFatalities_prev_day_16LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30DayWeekDayDays_since_ConfirmedCases=1Days_since_Fatalities=1Days_since_ConfirmedCases=10Days_since_Fatalities=10Days_since_ConfirmedCases=100Days_since_Fatalities=100Distance_to_originCountryAreaCountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+CountryPopDensityCountrySmokingRate
01.0Afghanistan33.000065.00002020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN4435.232618652860.018512.03019529.72738041.75720676.36511160.5114620.2681479.526105.0870.058269NaN
18185.0Cyprus35.126433.42992020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaN-1.0NaN7177.5291959240.0599.277599.2971198.574275.284382.653308.911193.20238.5240.12971636.4
221112.0GuamUS13.4443144.79372020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN3851.9426899147420.0166238.618162826.299329064.91782289.91690159.13482846.12660854.33212915.4090.035974NaN
38278.0Czechia49.817515.47302020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02NaNNaNNaNNaNNaNNaN7897.62403877220.05428.5855260.62810689.2132152.9232656.3403107.3602332.215440.3750.138425NaN
48371.0DenmarkDenmark56.26399.50182020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0-1.0-1.0NaN7931.06918641990.02902.2052869.6725771.8771287.8641437.5661556.0361224.130266.2810.13745819.1
\n", + "

5 rows × 91 columns

\n", + "
" + ], + "text/plain": [ + " Id Province/State ... CountryPopDensity CountrySmokingRate\n", + "0 1.0 ... 0.058269 NaN\n", + "1 8185.0 ... 0.129716 36.4\n", + "2 21112.0 Guam ... 0.035974 NaN\n", + "3 8278.0 ... 0.138425 NaN\n", + "4 8371.0 Denmark ... 0.137458 19.1\n", + "\n", + "[5 rows x 91 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 34 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "xQFM8KLBLr4N", + "colab_type": "text" + }, + "source": [ + "### Country hospital beds per 1000 people \n", + "\n", + "Let's add some information about countries' health system.\n", + "\n", + "We will add the number of hospital beds per 1000 people first." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "6qWbXPD1L-x1", + "colab_type": "code", + "outputId": "3814a0f4-625e-4674-fc63-e58ad9cd2373", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 394 + } + }, + "source": [ + "!mkdir -p hospital_beds\n", + "os.chdir('hospital_beds')\n", + "!wget -nc http://api.worldbank.org/v2/en/indicator/SH.MED.BEDS.ZS?downloadformat=csv -O hospital_beds.zip\n", + "!unzip -n hospital_beds.zip \n", + "!ls -l ." + ], + "execution_count": 35, + "outputs": [ + { + "output_type": "stream", + "text": [ + "--2020-04-19 14:12:00-- http://api.worldbank.org/v2/en/indicator/SH.MED.BEDS.ZS?downloadformat=csv\n", + "Resolving api.worldbank.org (api.worldbank.org)... 107.23.127.3\n", + "Connecting to api.worldbank.org (api.worldbank.org)|107.23.127.3|:80... connected.\n", + "HTTP request sent, awaiting response... 200 OK\n", + "Length: 33566 (33K) [application/zip]\n", + "Saving to: ‘hospital_beds.zip’\n", + "\n", + "\rhospital_beds.zip 0%[ ] 0 --.-KB/s \rhospital_beds.zip 100%[===================>] 32.78K --.-KB/s in 0.01s \n", + "\n", + "2020-04-19 14:12:00 (2.52 MB/s) - ‘hospital_beds.zip’ saved [33566/33566]\n", + "\n", + "Archive: hospital_beds.zip\n", + " inflating: Metadata_Indicator_API_SH.MED.BEDS.ZS_DS2_en_csv_v2_935968.csv \n", + " inflating: API_SH.MED.BEDS.ZS_DS2_en_csv_v2_935968.csv \n", + " inflating: Metadata_Country_API_SH.MED.BEDS.ZS_DS2_en_csv_v2_935968.csv \n", + "total 180\n", + "-rw-r--r-- 1 root root 108825 Apr 9 15:42 API_SH.MED.BEDS.ZS_DS2_en_csv_v2_935968.csv\n", + "-rw-r--r-- 1 root root 33566 Apr 19 14:12 hospital_beds.zip\n", + "-rw-r--r-- 1 root root 30906 Apr 9 15:42 Metadata_Country_API_SH.MED.BEDS.ZS_DS2_en_csv_v2_935968.csv\n", + "-rw-r--r-- 1 root root 403 Apr 9 15:42 Metadata_Indicator_API_SH.MED.BEDS.ZS_DS2_en_csv_v2_935968.csv\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "Kr4el2maMOL9", + "colab_type": "code", + "outputId": "b8265736-c96f-4e66-ae42-65eb56262b91", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 226 + } + }, + "source": [ + "!head API_SH.MED.BEDS.ZS_DS2_en_csv_v2_935968.csv" + ], + "execution_count": 36, + "outputs": [ + { + "output_type": "stream", + "text": [ + "\"Data Source\",\"World Development Indicators\",\r\n", + "\r\n", + "\"Last Updated Date\",\"2020-04-09\",\r\n", + "\r\n", + "\"Country Name\",\"Country Code\",\"Indicator Name\",\"Indicator Code\",\"1960\",\"1961\",\"1962\",\"1963\",\"1964\",\"1965\",\"1966\",\"1967\",\"1968\",\"1969\",\"1970\",\"1971\",\"1972\",\"1973\",\"1974\",\"1975\",\"1976\",\"1977\",\"1978\",\"1979\",\"1980\",\"1981\",\"1982\",\"1983\",\"1984\",\"1985\",\"1986\",\"1987\",\"1988\",\"1989\",\"1990\",\"1991\",\"1992\",\"1993\",\"1994\",\"1995\",\"1996\",\"1997\",\"1998\",\"1999\",\"2000\",\"2001\",\"2002\",\"2003\",\"2004\",\"2005\",\"2006\",\"2007\",\"2008\",\"2009\",\"2010\",\"2011\",\"2012\",\"2013\",\"2014\",\"2015\",\"2016\",\"2017\",\"2018\",\"2019\",\r\n", + "\"Aruba\",\"ABW\",\"Hospital beds (per 1,000 people)\",\"SH.MED.BEDS.ZS\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n", + "\"Afghanistan\",\"AFG\",\"Hospital beds (per 1,000 people)\",\"SH.MED.BEDS.ZS\",\"0.170626997947693\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"0.199000000953674\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"0.275599986314774\",\"\",\"\",\"\",\"\",\"\",\"0.3091000020504\",\"\",\"\",\"0.2497999966\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"0.3\",\"0.4\",\"0.4\",\"0.4\",\"0.4\",\"0.4\",\"0.4\",\"0.4\",\"0.4\",\"0.4\",\"0.4\",\"0.4\",\"0.5\",\"0.5\",\"0.5\",\"0.5\",\"\",\"\",\"\",\"\",\r\n", + "\"Angola\",\"AGO\",\"Hospital beds (per 1,000 people)\",\"SH.MED.BEDS.ZS\",\"2.06146168708801\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"2.72099995613098\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"1.2913000584\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"0.8\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n", + "\"Albania\",\"ALB\",\"Hospital beds (per 1,000 people)\",\"SH.MED.BEDS.ZS\",\"5.1026759147644\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"4.27169990539551\",\"4.18709993362427\",\"4.16069984436035\",\"4.08620023727417\",\"4.13899993896484\",\"4.13880014419556\",\"4.06969976425171\",\"3.97000002861023\",\"3.93549990653992\",\"4.13210010528564\",\"4.0248999596\",\"3.9986999035\",\"4.0134000778\",\"3.8313999176\",\"3.0171000957\",\"3.1900000572\",\"3.1400001049\",\"3.0499999523\",\"3.0499999523\",\"3.0299999714\",\"3.3\",\"3.3\",\"3.1\",\"3.1\",\"3\",\"3.1\",\"3.1\",\"3.1\",\"\",\"2.8\",\"3\",\"2.6\",\"2.9\",\"2.9\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n", + "\"Andorra\",\"AND\",\"Hospital beds (per 1,000 people)\",\"SH.MED.BEDS.ZS\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"3.1800000668\",\"3\",\"3.0899999142\",\"3.2000000477\",\"3.2000000477\",\"2.5899999142\",\"\",\"3.2999999523\",\"\",\"2.7\",\"2.6\",\"2.6\",\"\",\"2.5\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "aAuW8w8uMWtD", + "colab_type": "code", + "outputId": "7330f061-777c-4f41-8995-a8faae3b2332", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 504 + } + }, + "source": [ + "hospital_beds_df = pd.read_csv('./API_SH.MED.BEDS.ZS_DS2_en_csv_v2_935968.csv', skiprows=4)\n", + "os.chdir('..')\n", + "hospital_beds_df.head()" + ], + "execution_count": 37, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Country NameCountry CodeIndicator NameIndicator Code196019611962196319641965196619671968196919701971197219731974197519761977197819791980198119821983198419851986198719881989199019911992199319941995199619971998199920002001200220032004200520062007200820092010201120122013201420152016201720182019Unnamed: 64
0ArubaABWHospital beds (per 1,000 people)SH.MED.BEDS.ZSNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
1AfghanistanAFGHospital beds (per 1,000 people)SH.MED.BEDS.ZS0.170627NaNNaNNaNNaNNaNNaNNaNNaNNaN0.199NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN0.2756NaNNaNNaNNaNNaN0.3091NaNNaN0.2498NaNNaNNaNNaNNaNNaNNaNNaNNaN0.30.400.40.40.40.40.40.40.40.40.40.40.50.50.50.5NaNNaNNaNNaNNaN
2AngolaAGOHospital beds (per 1,000 people)SH.MED.BEDS.ZS2.061462NaNNaNNaNNaNNaNNaNNaNNaNNaN2.721NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN1.2913NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN0.8NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
3AlbaniaALBHospital beds (per 1,000 people)SH.MED.BEDS.ZS5.102676NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN4.27174.18714.16074.08624.1394.13884.06973.97003.93554.13214.02493.99874.01343.83143.01713.193.143.053.053.033.33.303.13.13.03.13.13.1NaN2.83.02.62.92.9NaNNaNNaNNaNNaNNaNNaN
4AndorraANDHospital beds (per 1,000 people)SH.MED.BEDS.ZSNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN3.183.003.093.203.22.59NaN3.3NaN2.72.62.6NaN2.5NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
\n", + "
" + ], + "text/plain": [ + " Country Name Country Code ... 2019 Unnamed: 64\n", + "0 Aruba ABW ... NaN NaN\n", + "1 Afghanistan AFG ... NaN NaN\n", + "2 Angola AGO ... NaN NaN\n", + "3 Albania ALB ... NaN NaN\n", + "4 Andorra AND ... NaN NaN\n", + "\n", + "[5 rows x 65 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 37 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "QhmRv_Z-NjIH", + "colab_type": "text" + }, + "source": [ + "We will use the last available data from recent years (2010 and later). " + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "yZdsXWNyNmxu", + "colab_type": "code", + "colab": {} + }, + "source": [ + "recent_year_columns = [str(year) for year in range(2010, 2020)]\n", + "\n", + "hospital_beds_df['CountryHospitalBedsRate'] = hospital_beds_df[recent_year_columns].apply(\n", + " lambda row: row[row.last_valid_index()] if row.last_valid_index() else np.nan,\n", + " axis='columns' \n", + ")" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "oqZ6aAyIN5AD", + "colab_type": "text" + }, + "source": [ + "Drop unneeded columns" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "UHbZLDwaN95b", + "colab_type": "code", + "outputId": "d91e568a-41b6-4e82-dbca-53d03cc0f101", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + } + }, + "source": [ + "hospital_beds_df = hospital_beds_df[['Country Name', 'CountryHospitalBedsRate']]\n", + "hospital_beds_df.head()" + ], + "execution_count": 39, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Country NameCountryHospitalBedsRate
0ArubaNaN
1Afghanistan0.5
2AngolaNaN
3Albania2.9
4AndorraNaN
\n", + "
" + ], + "text/plain": [ + " Country Name CountryHospitalBedsRate\n", + "0 Aruba NaN\n", + "1 Afghanistan 0.5\n", + "2 Angola NaN\n", + "3 Albania 2.9\n", + "4 Andorra NaN" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 39 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "pUYpnIJEOKx4", + "colab_type": "text" + }, + "source": [ + "Now, merge this data with main data by country." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "l2EcWbVOONxU", + "colab_type": "code", + "outputId": "7ac77b77-8e07-4674-a265-5ab10cde2594", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 256 + } + }, + "source": [ + "main_df = merge_with_column_drop(\n", + " main_df,\n", + " hospital_beds_df,\n", + " right_df_column='Country Name'\n", + ")\n", + "main_df.head()" + ], + "execution_count": 40, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastIdLogNewConfirmedCasesLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30...LogNewFatalities_prev_day_12LogNewFatalities_prev_day_13LogNewFatalities_prev_day_14LogNewFatalities_prev_day_15LogNewFatalities_prev_day_16LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30DayWeekDayDays_since_ConfirmedCases=1Days_since_Fatalities=1Days_since_ConfirmedCases=10Days_since_Fatalities=10Days_since_ConfirmedCases=100Days_since_Fatalities=100Distance_to_originCountryAreaCountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+CountryPopDensityCountrySmokingRateCountryHospitalBedsRate
01.0Afghanistan33.000065.00002020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN4435.232618652860.018512.03019529.72738041.75720676.36511160.5114620.2681479.526105.0870.058269NaN0.5
18185.0Cyprus35.126433.42992020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaN-1.0NaN7177.5291959240.0599.277599.2971198.574275.284382.653308.911193.20238.5240.12971636.43.4
221112.0GuamUS13.4443144.79372020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN3851.9426899147420.0166238.618162826.299329064.91782289.91690159.13482846.12660854.33212915.4090.035974NaNNaN
38278.0Czechia49.817515.47302020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02NaNNaNNaNNaNNaNNaN7897.62403877220.05428.5855260.62810689.2132152.9232656.3403107.3602332.215440.3750.138425NaNNaN
48371.0DenmarkDenmark56.26399.50182020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0-1.0-1.0NaN7931.06918641990.02902.2052869.6725771.8771287.8641437.5661556.0361224.130266.2810.13745819.12.5
\n", + "

5 rows × 92 columns

\n", + "
" + ], + "text/plain": [ + " Id Province/State ... CountrySmokingRate CountryHospitalBedsRate\n", + "0 1.0 ... NaN 0.5\n", + "1 8185.0 ... 36.4 3.4\n", + "2 21112.0 Guam ... NaN NaN\n", + "3 8278.0 ... NaN NaN\n", + "4 8371.0 Denmark ... 19.1 2.5\n", + "\n", + "[5 rows x 92 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 40 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "yxyLUTiLPWGT", + "colab_type": "text" + }, + "source": [ + "### Country Health Expenditure \n", + "\n", + "Now, let's add Health expenditure per capita, PPP (in current international $) feature." + ] + }, + { + "cell_type": "code", + "metadata": { + "colab_type": "code", + "outputId": "d6e3de73-5c6c-4b2f-e018-9c6ea6776fd5", + "id": "zlOWPWCAQIwp", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 394 + } + }, + "source": [ + "!mkdir -p health_expenditure\n", + "os.chdir('health_expenditure')\n", + "!wget -nc http://api.worldbank.org/v2/en/indicator/SH.XPD.CHEX.PP.CD?downloadformat=csv -O health_expenditure.zip\n", + "!unzip -n health_expenditure.zip \n", + "!ls -l ." + ], + "execution_count": 41, + "outputs": [ + { + "output_type": "stream", + "text": [ + "--2020-04-19 14:12:08-- http://api.worldbank.org/v2/en/indicator/SH.XPD.CHEX.PP.CD?downloadformat=csv\n", + "Resolving api.worldbank.org (api.worldbank.org)... 34.237.118.134\n", + "Connecting to api.worldbank.org (api.worldbank.org)|34.237.118.134|:80... connected.\n", + "HTTP request sent, awaiting response... 200 OK\n", + "Length: 45914 (45K) [application/zip]\n", + "Saving to: ‘health_expenditure.zip’\n", + "\n", + "\rhealth_expenditure. 0%[ ] 0 --.-KB/s \rhealth_expenditure. 100%[===================>] 44.84K --.-KB/s in 0.02s \n", + "\n", + "2020-04-19 14:12:08 (1.76 MB/s) - ‘health_expenditure.zip’ saved [45914/45914]\n", + "\n", + "Archive: health_expenditure.zip\n", + " inflating: Metadata_Indicator_API_SH.XPD.CHEX.PP.CD_DS2_en_csv_v2_939222.csv \n", + " inflating: API_SH.XPD.CHEX.PP.CD_DS2_en_csv_v2_939222.csv \n", + " inflating: Metadata_Country_API_SH.XPD.CHEX.PP.CD_DS2_en_csv_v2_939222.csv \n", + "total 224\n", + "-rw-r--r-- 1 root root 143137 Apr 9 22:50 API_SH.XPD.CHEX.PP.CD_DS2_en_csv_v2_939222.csv\n", + "-rw-r--r-- 1 root root 45914 Apr 19 14:12 health_expenditure.zip\n", + "-rw-r--r-- 1 root root 30906 Apr 9 22:50 Metadata_Country_API_SH.XPD.CHEX.PP.CD_DS2_en_csv_v2_939222.csv\n", + "-rw-r--r-- 1 root root 380 Apr 9 22:50 Metadata_Indicator_API_SH.XPD.CHEX.PP.CD_DS2_en_csv_v2_939222.csv\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "-lzQ9G-YQYrf", + "colab_type": "code", + "outputId": "22f1c347-32c1-4b05-d042-6082b34ae6a1", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 226 + } + }, + "source": [ + "!head API_SH.XPD.CHEX.PP.CD_DS2_en_csv_v2_939222.csv" + ], + "execution_count": 42, + "outputs": [ + { + "output_type": "stream", + "text": [ + "\"Data Source\",\"World Development Indicators\",\r\n", + "\r\n", + "\"Last Updated Date\",\"2020-04-09\",\r\n", + "\r\n", + "\"Country Name\",\"Country Code\",\"Indicator Name\",\"Indicator Code\",\"1960\",\"1961\",\"1962\",\"1963\",\"1964\",\"1965\",\"1966\",\"1967\",\"1968\",\"1969\",\"1970\",\"1971\",\"1972\",\"1973\",\"1974\",\"1975\",\"1976\",\"1977\",\"1978\",\"1979\",\"1980\",\"1981\",\"1982\",\"1983\",\"1984\",\"1985\",\"1986\",\"1987\",\"1988\",\"1989\",\"1990\",\"1991\",\"1992\",\"1993\",\"1994\",\"1995\",\"1996\",\"1997\",\"1998\",\"1999\",\"2000\",\"2001\",\"2002\",\"2003\",\"2004\",\"2005\",\"2006\",\"2007\",\"2008\",\"2009\",\"2010\",\"2011\",\"2012\",\"2013\",\"2014\",\"2015\",\"2016\",\"2017\",\"2018\",\"2019\",\r\n", + "\"Aruba\",\"ABW\",\"Current health expenditure per capita, PPP (current international $)\",\"SH.XPD.CHEX.PP.CD\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n", + "\"Afghanistan\",\"AFG\",\"Current health expenditure per capita, PPP (current international $)\",\"SH.XPD.CHEX.PP.CD\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"79.7509994506836\",\"80.9199752807617\",\"87.67626953125\",\"98.6120834350586\",\"111.271469116211\",\"104.657890319824\",\"127.603996276855\",\"143.849624633789\",\"132.271286010742\",\"138.822616577148\",\"143.965316772461\",\"164.160110473633\",\"176.930862426758\",\"183.927856445313\",\"174.73063659668\",\"\",\"\",\"\",\r\n", + "\"Angola\",\"AGO\",\"Current health expenditure per capita, PPP (current international $)\",\"SH.XPD.CHEX.PP.CD\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"71.804817199707\",\"174.036361694336\",\"117.858177185059\",\"127.356391906738\",\"156.64697265625\",\"129.017425537109\",\"134.776672363281\",\"168.597549438477\",\"206.017562866211\",\"233.626068115234\",\"167.820114135742\",\"167.915237426758\",\"162.328979492188\",\"190.890869140625\",\"175.377624511719\",\"185.089279174805\",\"183.510406494141\",\"185.854034423828\",\"\",\"\",\r\n", + "\"Albania\",\"ALB\",\"Current health expenditure per capita, PPP (current international $)\",\"SH.XPD.CHEX.PP.CD\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\r\n", + "\"Andorra\",\"AND\",\"Current health expenditure per capita, PPP (current international $)\",\"SH.XPD.CHEX.PP.CD\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"3049.00610351563\",\"3143.03515625\",\"3284.2294921875\",\"3277.25073242188\",\"3350.19384765625\",\"3741.40380859375\",\"3974.61474609375\",\"4098.1201171875\",\"3983.9091796875\",\"3966.13012695313\",\"3937.48754882813\",\"4010.31762695313\",\"4414.71142578125\",\"4652.24169921875\",\"4963.498046875\",\"5093.17626953125\",\"4990.35546875\",\"5237.24169921875\",\"\",\"\",\r\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "Ib3jSvfLQf7o", + "colab_type": "code", + "outputId": "8db56c99-e5f1-422a-ca6e-eecccb4624eb", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 591 + } + }, + "source": [ + "health_expenditure_df = pd.read_csv('./API_SH.XPD.CHEX.PP.CD_DS2_en_csv_v2_939222.csv', skiprows=4)\n", + "os.chdir('..')\n", + "health_expenditure_df.head()" + ], + "execution_count": 43, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Country NameCountry CodeIndicator NameIndicator Code196019611962196319641965196619671968196919701971197219731974197519761977197819791980198119821983198419851986198719881989199019911992199319941995199619971998199920002001200220032004200520062007200820092010201120122013201420152016201720182019Unnamed: 64
0ArubaABWCurrent health expenditure per capita, PPP (cu...SH.XPD.CHEX.PP.CDNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
1AfghanistanAFGCurrent health expenditure per capita, PPP (cu...SH.XPD.CHEX.PP.CDNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN79.75099980.91997587.67627098.612083111.271469104.657890127.603996143.849625132.271286138.822617143.965317164.160110176.930862183.927856174.730637NaNNaNNaNNaN
2AngolaAGOCurrent health expenditure per capita, PPP (cu...SH.XPD.CHEX.PP.CDNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN71.804817174.036362117.858177127.356392156.646973129.017426134.776672168.597549206.017563233.626068167.820114167.915237162.328979190.890869175.377625185.089279183.510406185.854034NaNNaNNaN
3AlbaniaALBCurrent health expenditure per capita, PPP (cu...SH.XPD.CHEX.PP.CDNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
4AndorraANDCurrent health expenditure per capita, PPP (cu...SH.XPD.CHEX.PP.CDNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN3049.0061043143.0351563284.2294923277.2507323350.1938483741.4038093974.6147464098.1201173983.9091803966.1301273937.4875494010.3176274414.7114264652.2416994963.4980475093.1762704990.3554695237.241699NaNNaNNaN
\n", + "
" + ], + "text/plain": [ + " Country Name Country Code ... 2019 Unnamed: 64\n", + "0 Aruba ABW ... NaN NaN\n", + "1 Afghanistan AFG ... NaN NaN\n", + "2 Angola AGO ... NaN NaN\n", + "3 Albania ALB ... NaN NaN\n", + "4 Andorra AND ... NaN NaN\n", + "\n", + "[5 rows x 65 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 43 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "XwFzpw6yQySh", + "colab_type": "text" + }, + "source": [ + "We will use the last available data from recent years (2010 and later). " + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "YHdi0Py_QzOs", + "colab_type": "code", + "colab": {} + }, + "source": [ + "recent_year_columns = [str(year) for year in range(2010, 2020)]\n", + "\n", + "health_expenditure_df['CountryHealthExpenditurePerCapitaPPP'] = health_expenditure_df[recent_year_columns].apply(\n", + " lambda row: row[row.last_valid_index()] if row.last_valid_index() else np.nan,\n", + " axis='columns' \n", + ")" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "UtuzU6KVRCdf", + "colab_type": "text" + }, + "source": [ + "Drop unneeded columns\n" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "38g2WZe3RDk3", + "colab_type": "code", + "outputId": "80fcef09-cd26-4756-f385-65490ae4c851", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + } + }, + "source": [ + "health_expenditure_df = health_expenditure_df[['Country Name', 'CountryHealthExpenditurePerCapitaPPP']]\n", + "health_expenditure_df.head()" + ], + "execution_count": 45, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Country NameCountryHealthExpenditurePerCapitaPPP
0ArubaNaN
1Afghanistan174.730637
2Angola185.854034
3AlbaniaNaN
4Andorra5237.241699
\n", + "
" + ], + "text/plain": [ + " Country Name CountryHealthExpenditurePerCapitaPPP\n", + "0 Aruba NaN\n", + "1 Afghanistan 174.730637\n", + "2 Angola 185.854034\n", + "3 Albania NaN\n", + "4 Andorra 5237.241699" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 45 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "oh6LBo2ZRWoS", + "colab_type": "text" + }, + "source": [ + "Now, merge this data with main data by country.\n" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "011eb8rxRb08", + "colab_type": "code", + "outputId": "a07af4f9-dc09-4bc7-9df2-63029ac4e1e1", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 256 + } + }, + "source": [ + "main_df = merge_with_column_drop(\n", + " main_df,\n", + " health_expenditure_df,\n", + " right_df_column='Country Name'\n", + ")\n", + "main_df.head()" + ], + "execution_count": 46, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastIdLogNewConfirmedCasesLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30...LogNewFatalities_prev_day_13LogNewFatalities_prev_day_14LogNewFatalities_prev_day_15LogNewFatalities_prev_day_16LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30DayWeekDayDays_since_ConfirmedCases=1Days_since_Fatalities=1Days_since_ConfirmedCases=10Days_since_Fatalities=10Days_since_ConfirmedCases=100Days_since_Fatalities=100Distance_to_originCountryAreaCountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+CountryPopDensityCountrySmokingRateCountryHospitalBedsRateCountryHealthExpenditurePerCapitaPPP
01.0Afghanistan33.000065.00002020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN4435.232618652860.018512.03019529.72738041.75720676.36511160.5114620.2681479.526105.0870.058269NaN0.5174.730637
18185.0Cyprus35.126433.42992020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaN-1.0NaN7177.5291959240.0599.277599.2971198.574275.284382.653308.911193.20238.5240.12971636.43.42430.176025
221112.0GuamUS13.4443144.79372020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN3851.9426899147420.0166238.618162826.299329064.91782289.91690159.13482846.12660854.33212915.4090.035974NaNNaNNaN
38278.0Czechia49.817515.47302020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02NaNNaNNaNNaNNaNNaN7897.62403877220.05428.5855260.62810689.2132152.9232656.3403107.3602332.215440.3750.138425NaNNaNNaN
48371.0DenmarkDenmark56.26399.50182020-01-220.00.0NaN0.0NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0-1.0-1.0NaN7931.06918641990.02902.2052869.6725771.8771287.8641437.5661556.0361224.130266.2810.13745819.12.55509.995605
\n", + "

5 rows × 93 columns

\n", + "
" + ], + "text/plain": [ + " Id ... CountryHealthExpenditurePerCapitaPPP\n", + "0 1.0 ... 174.730637\n", + "1 8185.0 ... 2430.176025\n", + "2 21112.0 ... NaN\n", + "3 8278.0 ... NaN\n", + "4 8371.0 ... 5509.995605\n", + "\n", + "[5 rows x 93 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 46 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "To5EPZPHn5if", + "colab_type": "text" + }, + "source": [ + "## Prepare data for training \n", + "\n", + "Split data to train, validation and test by date.\n", + "\n", + "Construct features (by removing label and non-features data from the original dataframe) and labels for train, validation and test parts." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "fD9v3eMwHJtz", + "colab_type": "code", + "colab": {} + }, + "source": [ + "last_train_date = pd.Timestamp(2020, 3, 11)\n", + "last_eval_date = pd.Timestamp(2020, 3, 24)\n", + "last_test_date = pd.Timestamp(2020, 4, 23)\n", + "\n", + "train_df = main_df[main_df['Date'] <= last_train_date].copy()\n", + "eval_df = main_df[(main_df['Date'] > last_train_date) & (main_df['Date'] <= last_eval_date)].copy()\n", + "test_df = main_df[main_df['Date'] > last_eval_date].copy()\n", + "\n", + "# return features_df, labels\n", + "def preprocess_df(df):\n", + " labels = df[['LogNewConfirmedCases', 'LogNewFatalities']].copy()\n", + " features_df = df.drop(\n", + " columns=['Id', 'ForecastId', 'ConfirmedCases', 'LogNewConfirmedCases', 'Fatalities', 'LogNewFatalities', 'Date']\n", + " ).copy()\n", + " \n", + " return features_df, labels\n", + "\n", + "train_features_df, train_labels = preprocess_df(train_df)\n", + "eval_features_df, eval_labels = preprocess_df(eval_df)\n", + "test_features_df, _ = preprocess_df(test_df)" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "code", + "metadata": { + "id": "CLf7521nJJAM", + "colab_type": "code", + "outputId": "2e597b9e-67a3-4c9b-a471-2948b09096db", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 256 + } + }, + "source": [ + "train_features_df.head()" + ], + "execution_count": 48, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Province/StateCountry/RegionLatLongLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30LogNewFatalities_prev_day_1LogNewFatalities_prev_day_2LogNewFatalities_prev_day_3LogNewFatalities_prev_day_4LogNewFatalities_prev_day_5LogNewFatalities_prev_day_6...LogNewFatalities_prev_day_13LogNewFatalities_prev_day_14LogNewFatalities_prev_day_15LogNewFatalities_prev_day_16LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30DayWeekDayDays_since_ConfirmedCases=1Days_since_Fatalities=1Days_since_ConfirmedCases=10Days_since_Fatalities=10Days_since_ConfirmedCases=100Days_since_Fatalities=100Distance_to_originCountryAreaCountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+CountryPopDensityCountrySmokingRateCountryHospitalBedsRateCountryHealthExpenditurePerCapitaPPP
0Afghanistan33.000065.0000NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN4435.232618652860.018512.03019529.72738041.75720676.36511160.5114620.2681479.526105.0870.058269NaN0.5174.730637
1Cyprus35.126433.4299NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaN-1.0NaN7177.5291959240.0599.277599.2971198.574275.284382.653308.911193.20238.5240.12971636.43.42430.176025
2GuamUS13.4443144.7937NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0NaNNaNNaN3851.9426899147420.0166238.618162826.299329064.91782289.91690159.13482846.12660854.33212915.4090.035974NaNNaNNaN
3Czechia49.817515.4730NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02NaNNaNNaNNaNNaNNaN7897.62403877220.05428.5855260.62810689.2132152.9232656.3403107.3602332.215440.3750.138425NaNNaNNaN
4DenmarkDenmark56.26399.5018NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaNNaN02-1.0-1.0-1.0-1.0-1.0NaN7931.06918641990.02902.2052869.6725771.8771287.8641437.5661556.0361224.130266.2810.13745819.12.55509.995605
\n", + "

5 rows × 86 columns

\n", + "
" + ], + "text/plain": [ + " Province/State ... CountryHealthExpenditurePerCapitaPPP\n", + "0 ... 174.730637\n", + "1 ... 2430.176025\n", + "2 Guam ... NaN\n", + "3 ... NaN\n", + "4 Denmark ... 5509.995605\n", + "\n", + "[5 rows x 86 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 48 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "n_FTuSzhJZEP", + "colab_type": "text" + }, + "source": [ + "Categorical features" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "CKW31V6JJYrs", + "colab_type": "code", + "colab": {} + }, + "source": [ + "cat_features = ['Province/State', 'Country/Region']" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Zttyx5vhaXFI", + "colab_type": "text" + }, + "source": [ + "## Models training \n", + "\n", + "Now, construct models that predict LogNewConfirmedCases and LogNewFatalities for the next day.\n", + "\n", + "We will use RMSE as an objective and evaluation metric for training, that will correspond to RMSLE (Root Mean Square Logarithmic Error) metric for NewConfirmedCases and NewFatalities.\n", + "\n", + "$$RMSLE(labels, predictions) = \\sqrt{\\frac{1}{n}\\sum_{i=1}^{n}{\\Big(log(labels_i + 1) - log(predictions_i + 1)\\Big)^2}}$$\n", + "\n", + "Such metric is natural for this task (model should weight errors between 1000 and 1100 and 100000 and 110000 approximately the same).\n", + "\n", + "For prediction values in the original scale model prediction output should be transformed back to the original scale with the reverse transformation (can be preformed using numpy.expm1 function).\n", + "\n", + "We'll train models for 1000 iterations.\n", + "\n", + "\n" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "dxgp1LOrJ51V", + "colab_type": "code", + "colab": {} + }, + "source": [ + "iterations = 1000" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "G4NpZ6miJpn2", + "colab_type": "text" + }, + "source": [ + "CatBoost allows Categorical features values can be used as is in DataFrame." + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "VrJkAewhJk86", + "colab_type": "code", + "outputId": "c969cdb1-17a5-4936-bd5e-c11ae4a69f4c", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 715 + } + }, + "source": [ + "print ('catboost version', cb.__version__)\n", + "\n", + "catboost_models = {}\n", + "\n", + "for prediction_name in ['LogNewConfirmedCases', 'LogNewFatalities']:\n", + " model = cb.CatBoostRegressor(\n", + " has_time=True,\n", + " iterations=iterations\n", + " )\n", + "\n", + " model.fit(\n", + " train_features_df,\n", + " train_labels[prediction_name],\n", + " eval_set=(eval_features_df, eval_labels[prediction_name]),\n", + " cat_features=cat_features,\n", + " verbose=100 # print metrics each 100 iterations\n", + " )\n", + " \n", + " catboost_models[prediction_name] = model\n", + " print ('CatBoost: prediction of %s: RMSLE on test = %s' % (prediction_name, model.evals_result_['validation']['RMSE'][-1]))" + ], + "execution_count": 51, + "outputs": [ + { + "output_type": "stream", + "text": [ + "catboost version 0.22\n", + "Learning rate set to 0.074929\n", + "0:\tlearn: 0.8444130\ttest: 2.2561020\tbest: 2.2561020 (0)\ttotal: 79.4ms\tremaining: 1m 19s\n", + "100:\tlearn: 0.2849727\ttest: 0.8850558\tbest: 0.8850558 (100)\ttotal: 2.2s\tremaining: 19.6s\n", + "200:\tlearn: 0.2518107\ttest: 0.8737440\tbest: 0.8737323 (198)\ttotal: 4.25s\tremaining: 16.9s\n", + "300:\tlearn: 0.2288005\ttest: 0.8734222\tbest: 0.8722002 (273)\ttotal: 6.27s\tremaining: 14.6s\n", + "400:\tlearn: 0.2078918\ttest: 0.8766954\tbest: 0.8720236 (318)\ttotal: 8.37s\tremaining: 12.5s\n", + "500:\tlearn: 0.1899614\ttest: 0.8815886\tbest: 0.8720236 (318)\ttotal: 10.5s\tremaining: 10.4s\n", + "600:\tlearn: 0.1765715\ttest: 0.8839863\tbest: 0.8720236 (318)\ttotal: 12.6s\tremaining: 8.36s\n", + "700:\tlearn: 0.1650007\ttest: 0.8869404\tbest: 0.8720236 (318)\ttotal: 14.7s\tremaining: 6.26s\n", + "800:\tlearn: 0.1547044\ttest: 0.8882405\tbest: 0.8720236 (318)\ttotal: 16.8s\tremaining: 4.17s\n", + "900:\tlearn: 0.1458812\ttest: 0.8884831\tbest: 0.8720236 (318)\ttotal: 18.9s\tremaining: 2.08s\n", + "999:\tlearn: 0.1379156\ttest: 0.8917756\tbest: 0.8720236 (318)\ttotal: 21s\tremaining: 0us\n", + "\n", + "bestTest = 0.8720236025\n", + "bestIteration = 318\n", + "\n", + "Shrink model to first 319 iterations.\n", + "CatBoost: prediction of LogNewConfirmedCases: RMSLE on test = 0.8917755772136071\n", + "Learning rate set to 0.074929\n", + "0:\tlearn: 0.2846849\ttest: 0.7751307\tbest: 0.7751307 (0)\ttotal: 22.4ms\tremaining: 22.4s\n", + "100:\tlearn: 0.0583554\ttest: 0.3493678\tbest: 0.3493678 (100)\ttotal: 2.08s\tremaining: 18.6s\n", + "200:\tlearn: 0.0461636\ttest: 0.3449327\tbest: 0.3449327 (200)\ttotal: 4.14s\tremaining: 16.4s\n", + "300:\tlearn: 0.0359213\ttest: 0.3416266\tbest: 0.3415318 (297)\ttotal: 6.24s\tremaining: 14.5s\n", + "400:\tlearn: 0.0274293\ttest: 0.3410587\tbest: 0.3410214 (398)\ttotal: 8.34s\tremaining: 12.5s\n", + "500:\tlearn: 0.0217875\ttest: 0.3409957\tbest: 0.3409543 (498)\ttotal: 10.4s\tremaining: 10.4s\n", + "600:\tlearn: 0.0171285\ttest: 0.3406960\tbest: 0.3405870 (550)\ttotal: 12.6s\tremaining: 8.33s\n", + "700:\tlearn: 0.0140178\ttest: 0.3402807\tbest: 0.3402373 (692)\ttotal: 14.7s\tremaining: 6.25s\n", + "800:\tlearn: 0.0115103\ttest: 0.3402232\tbest: 0.3402147 (798)\ttotal: 16.8s\tremaining: 4.17s\n", + "900:\tlearn: 0.0095381\ttest: 0.3404695\tbest: 0.3402147 (798)\ttotal: 18.9s\tremaining: 2.08s\n", + "999:\tlearn: 0.0079940\ttest: 0.3405940\tbest: 0.3402147 (798)\ttotal: 21s\tremaining: 0us\n", + "\n", + "bestTest = 0.3402146803\n", + "bestIteration = 798\n", + "\n", + "Shrink model to first 799 iterations.\n", + "CatBoost: prediction of LogNewFatalities: RMSLE on test = 0.34059397104647504\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "iI8pVMqdJjRb", + "colab_type": "text" + }, + "source": [ + "## Feature importance in the models \n", + "\n", + "Let's look at feature importances.\n", + "For CatBoost use [PredictionValuesChange](https://catboost.ai/docs/concepts/fstr.html#fstr__regular-feature-importance).\n", + "\n", + "We'll print only 25 most important features for clarity.\n" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "3ydN6CMEJMJ5", + "colab_type": "code", + "outputId": "0180e06e-b3c1-42c4-944c-10abc78eb911", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 1000 + } + }, + "source": [ + "for prediction_name in ['LogNewConfirmedCases', 'LogNewFatalities']:\n", + " print ('\\nCatBoost: prediction of %s. Feature importance. Type=PredictionValuesChange' % prediction_name)\n", + " print (\n", + " catboost_models[prediction_name].get_feature_importance(\n", + " type=cb.EFstrType.PredictionValuesChange,\n", + " prettified=True\n", + " ).head(25).to_string()\n", + " )" + ], + "execution_count": 52, + "outputs": [ + { + "output_type": "stream", + "text": [ + "\n", + "CatBoost: prediction of LogNewConfirmedCases. Feature importance. Type=PredictionValuesChange\n", + " Feature Id Importances\n", + "0 LogNewConfirmedCases_prev_day_1 22.109213\n", + "1 LogNewConfirmedCases_prev_day_2 17.254841\n", + "2 Days_since_ConfirmedCases=10 11.929176\n", + "3 Days_since_ConfirmedCases=1 11.340144\n", + "4 LogNewConfirmedCases_prev_day_3 6.377358\n", + "5 Day 2.969185\n", + "6 Days_since_Fatalities=10 2.720522\n", + "7 Days_since_ConfirmedCases=100 1.865486\n", + "8 LogNewConfirmedCases_prev_day_4 1.732866\n", + "9 Distance_to_origin 1.259510\n", + "10 Days_since_Fatalities=100 1.179583\n", + "11 LogNewConfirmedCases_prev_day_10 1.139283\n", + "12 LogNewConfirmedCases_prev_day_24 1.134655\n", + "13 LogNewConfirmedCases_prev_day_5 1.125969\n", + "14 CountryPop_60-80 0.963053\n", + "15 LogNewFatalities_prev_day_3 0.914145\n", + "16 CountryPop_40-60 0.747576\n", + "17 LogNewFatalities_prev_day_2 0.722008\n", + "18 Long 0.608342\n", + "19 Days_since_Fatalities=1 0.587399\n", + "20 LogNewConfirmedCases_prev_day_7 0.583528\n", + "21 Lat 0.536044\n", + "22 LogNewConfirmedCases_prev_day_8 0.527629\n", + "23 Province/State 0.495591\n", + "24 LogNewConfirmedCases_prev_day_27 0.444949\n", + "\n", + "CatBoost: prediction of LogNewFatalities. Feature importance. Type=PredictionValuesChange\n", + " Feature Id Importances\n", + "0 Days_since_Fatalities=10 26.907773\n", + "1 Days_since_Fatalities=1 12.810340\n", + "2 LogNewFatalities_prev_day_2 8.786807\n", + "3 Days_since_Fatalities=100 5.176575\n", + "4 LogNewConfirmedCases_prev_day_9 2.796640\n", + "5 LogNewConfirmedCases_prev_day_2 2.743006\n", + "6 LogNewFatalities_prev_day_3 2.312393\n", + "7 CountryPopDensity 2.179349\n", + "8 LogNewConfirmedCases_prev_day_10 1.811663\n", + "9 CountryPopFemale 1.657966\n", + "10 LogNewFatalities_prev_day_1 1.542263\n", + "11 LogNewConfirmedCases_prev_day_1 1.466262\n", + "12 LogNewConfirmedCases_prev_day_8 1.398892\n", + "13 LogNewConfirmedCases_prev_day_11 1.350219\n", + "14 CountryPop_40-60 1.336757\n", + "15 Distance_to_origin 1.336277\n", + "16 LogNewFatalities_prev_day_6 1.303100\n", + "17 LogNewConfirmedCases_prev_day_3 1.186486\n", + "18 LogNewConfirmedCases_prev_day_7 1.135694\n", + "19 LogNewConfirmedCases_prev_day_6 1.100854\n", + "20 LogNewConfirmedCases_prev_day_14 1.001467\n", + "21 CountryPop_80+ 0.845002\n", + "22 LogNewFatalities_prev_day_9 0.793612\n", + "23 LogNewFatalities_prev_day_24 0.776731\n", + "24 LogNewFatalities_prev_day_26 0.720919\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "jRnM6SwW_PHl", + "colab_type": "text" + }, + "source": [ + "## Create predictions for eval and test data. \n", + "\n", + "For forecasting further into the future (more than a single day) we will use an incremental (by days) approach:\n", + "\n", + "1. Predict Log New Confirmed Cases and Fatalities for the next unknown day.\n", + "2. Confirmed Cases and Fatalities for this day are incremented from the previous day based on predicted New Confirmed Cases and Fatalities.\n", + "3. Use predicted Log New Confirmed Cases and Fatalities for this day to initialize time delay embedding features corresponding to this day for future days.\n", + "\n", + "This procedure is then repeated for the new next prediction day and so on." + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "aTR3pxvsiUUb", + "colab_type": "text" + }, + "source": [ + "Common function for eval and test:" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "ivRAFKbg_Ovv", + "colab_type": "code", + "colab": {} + }, + "source": [ + "def predict_for_dataset(df, features_df, prev_day_df, first_date, last_date, update_features_data):\n", + " df['PredictedLogNewConfirmedCases'] = np.nan\n", + " df['PredictedLogNewFatalities'] = np.nan\n", + " df['PredictedConfirmedCases'] = np.nan\n", + " df['PredictedFatalities'] = np.nan\n", + "\n", + " for day in pd.date_range(first_date, last_date):\n", + " day_df = df[df['Date'] == day]\n", + " day_features_pool = cb.Pool(features_df.loc[day_df.index], cat_features=cat_features)\n", + "\n", + " # predict LogNew* data\n", + " for prediction_type in ['LogNewConfirmedCases', 'LogNewFatalities']:\n", + " # prediction is imprecise and can produce negative values, clip them\n", + " df.loc[day_df.index, 'Predicted' + prediction_type] = np.maximum(\n", + " catboost_models[prediction_type].predict(day_features_pool),\n", + " 0.0\n", + " )\n", + "\n", + " day_predictions_df = df.loc[day_df.index][\n", + " location_columns + ['PredictedLogNewConfirmedCases', 'PredictedLogNewFatalities']\n", + " ]\n", + "\n", + " # update Predicted ConfirmedCases and Fatalities\n", + " for field in ['ConfirmedCases', 'Fatalities']:\n", + " prev_day_field = field if day == first_eval_date else ('Predicted' + field)\n", + " merged_df = day_predictions_df.merge(\n", + " right=prev_day_df[location_columns + [prev_day_field]],\n", + " how='inner',\n", + " on=location_columns\n", + " )\n", + "\n", + " df.loc[day_df.index, 'Predicted' + field] = merged_df.apply(\n", + " lambda row: row[prev_day_field] + np.rint(np.expm1(row['PredictedLogNew' + field])),\n", + " axis='columns'\n", + " ).values\n", + "\n", + " if update_features_data:\n", + " # fill time delay embedding features based on this day for next days\n", + " for next_day in pd.date_range(day + pd.Timedelta(days=1), last_date):\n", + " next_day_features_df = features_df[df['Date'] == next_day]\n", + "\n", + " merged_df = next_day_features_df[location_columns].merge(\n", + " right=day_predictions_df,\n", + " how='inner',\n", + " on=location_columns\n", + " )\n", + "\n", + " prev_day_idx = (next_day - day).days\n", + " for prediction_type in ['LogNewConfirmedCases', 'LogNewFatalities']:\n", + " features_df.loc[next_day_features_df.index, prediction_type + '_prev_day_%s' % prev_day_idx] = (\n", + " merged_df['Predicted' + prediction_type].values\n", + " )\n", + "\n", + " # select by day_df.index again to get Predicted* columns\n", + " prev_day_df = df.loc[day_df.index]\n" + ], + "execution_count": 0, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "l_UO2zbRkPvi", + "colab_type": "text" + }, + "source": [ + "Create predictions for eval data" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "kqBsTO-_kTku", + "colab_type": "code", + "outputId": "6706a429-595a-4601-f40c-35a6b085489e", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 256 + } + }, + "source": [ + "prev_day_df = train_df.loc[train_df['Date'] == last_train_date] \n", + "first_eval_date = last_train_date + pd.Timedelta(days=1)\n", + "\n", + "predict_for_dataset(eval_df, eval_features_df, prev_day_df, first_eval_date, last_eval_date, update_features_data=False)\n", + "\n", + "eval_df.head()" + ], + "execution_count": 54, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastIdLogNewConfirmedCasesLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30...LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30DayWeekDayDays_since_ConfirmedCases=1Days_since_Fatalities=1Days_since_ConfirmedCases=10Days_since_Fatalities=10Days_since_ConfirmedCases=100Days_since_Fatalities=100Distance_to_originCountryAreaCountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+CountryPopDensityCountrySmokingRateCountryHospitalBedsRateCountryHealthExpenditurePerCapitaPPPPredictedLogNewConfirmedCasesPredictedLogNewFatalitiesPredictedConfirmedCasesPredictedFatalities
131006003.0JiangxiChina27.6140115.72212020-03-12935.01.0NaN0.0000000.0000000.00.00.0000000.00.0000000.00.0000000.00.0000000.0000000.0000000.6931470.0000000.00.00.00.00.00.00.00.6931471.3862941.7917592.5649492.6390573.3672963.3672963.7135723.526361...0.00.00.00.00.00.00.00.00.00.00.00.00.00.050350.031.048.0NaN44.0NaN501.2145039388210.0698159.433735624.2591433783.692338099.542416397.382437824.994215315.36226146.4120.15272225.64.2841.1149290.0557870.000000935.01.0
1310122278.0MichiganUS43.3266-84.53612020-03-122.00.0NaN0.0000001.0986120.00.00.0000000.00.0000000.00.0000000.00.0000000.0000000.0000000.0000000.0000000.00.00.00.00.00.00.00.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.000000...0.00.00.00.00.00.00.00.00.00.00.00.00.00.05031.0-1.0-1.0-1.0-1.0NaN11600.9082529147420.0166238.618162826.299329064.91782289.91690159.13482846.12660854.33212915.4090.035974NaNNaNNaN0.8222770.0000003.00.0
1310214652.0Mexico23.6345-102.55282020-03-1212.00.0NaN1.6094380.6931470.00.00.6931470.00.6931470.00.0000000.00.0000000.6931471.3862940.6931470.0000000.00.00.00.00.00.00.00.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.000000...0.00.00.00.00.00.00.00.00.00.00.00.00.00.050313.0-1.00.0NaN-1.0NaN12918.8844421943950.065172.13662403.393127575.52944602.34140260.55328740.66711970.2942001.6740.06562714.01.51035.5850831.5719870.00054912.00.0
131033771.0Nova ScotiaCanada44.6820-63.74432020-03-120.00.0NaN0.0000000.0000000.00.00.0000000.00.0000000.00.0000000.00.0000000.0000000.0000000.0000000.0000000.00.00.00.00.00.00.00.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.000000...0.00.00.00.00.00.00.00.00.00.00.00.00.00.0503NaNNaNNaNNaNNaNNaN11616.3603859093510.018847.50018563.53837411.0387934.16310315.29110057.7607475.0461628.7780.00411414.32.74928.6279300.0552380.0000000.00.0
131046375.0NingxiaChina37.2692106.16552020-03-1275.00.0NaN0.0000000.0000000.00.00.0000000.00.0000000.00.6931470.00.6931470.0000000.6931470.0000000.6931470.00.00.00.00.00.00.00.6931470.0000000.0000000.0000001.3862941.3862941.9459101.7917591.609438...0.00.00.00.00.00.00.00.00.00.00.00.00.00.050350.0NaN44.0NaNNaNNaN896.5088659388210.0698159.433735624.2591433783.692338099.542416397.382437824.994215315.36226146.4120.15272225.64.2841.1149290.0914700.00000075.00.0
\n", + "

5 rows × 97 columns

\n", + "
" + ], + "text/plain": [ + " Id Province/State ... PredictedConfirmedCases PredictedFatalities\n", + "13100 6003.0 Jiangxi ... 935.0 1.0\n", + "13101 22278.0 Michigan ... 3.0 0.0\n", + "13102 14652.0 ... 12.0 0.0\n", + "13103 3771.0 Nova Scotia ... 0.0 0.0\n", + "13104 6375.0 Ningxia ... 75.0 0.0\n", + "\n", + "[5 rows x 97 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 54 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "OGTU5I5AbuST", + "colab_type": "text" + }, + "source": [ + "Create predictions for test data.\n", + "\n" + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "Mz54gwV5K_qJ", + "colab_type": "code", + "outputId": "f673acbb-46bc-4557-ec43-155fab9ec23d", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 256 + } + }, + "source": [ + "prev_day_df = eval_df.loc[eval_df['Date'] == last_eval_date] \n", + "first_test_date = last_eval_date + pd.Timedelta(days=1)\n", + "\n", + "predict_for_dataset(test_df, test_features_df, prev_day_df, first_test_date, last_test_date, update_features_data=True)\n", + "\n", + "test_df.head()" + ], + "execution_count": 55, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IdProvince/StateCountry/RegionLatLongDateConfirmedCasesFatalitiesForecastIdLogNewConfirmedCasesLogNewConfirmedCases_prev_day_1LogNewConfirmedCases_prev_day_2LogNewConfirmedCases_prev_day_3LogNewConfirmedCases_prev_day_4LogNewConfirmedCases_prev_day_5LogNewConfirmedCases_prev_day_6LogNewConfirmedCases_prev_day_7LogNewConfirmedCases_prev_day_8LogNewConfirmedCases_prev_day_9LogNewConfirmedCases_prev_day_10LogNewConfirmedCases_prev_day_11LogNewConfirmedCases_prev_day_12LogNewConfirmedCases_prev_day_13LogNewConfirmedCases_prev_day_14LogNewConfirmedCases_prev_day_15LogNewConfirmedCases_prev_day_16LogNewConfirmedCases_prev_day_17LogNewConfirmedCases_prev_day_18LogNewConfirmedCases_prev_day_19LogNewConfirmedCases_prev_day_20LogNewConfirmedCases_prev_day_21LogNewConfirmedCases_prev_day_22LogNewConfirmedCases_prev_day_23LogNewConfirmedCases_prev_day_24LogNewConfirmedCases_prev_day_25LogNewConfirmedCases_prev_day_26LogNewConfirmedCases_prev_day_27LogNewConfirmedCases_prev_day_28LogNewConfirmedCases_prev_day_29LogNewConfirmedCases_prev_day_30...LogNewFatalities_prev_day_17LogNewFatalities_prev_day_18LogNewFatalities_prev_day_19LogNewFatalities_prev_day_20LogNewFatalities_prev_day_21LogNewFatalities_prev_day_22LogNewFatalities_prev_day_23LogNewFatalities_prev_day_24LogNewFatalities_prev_day_25LogNewFatalities_prev_day_26LogNewFatalities_prev_day_27LogNewFatalities_prev_day_28LogNewFatalities_prev_day_29LogNewFatalities_prev_day_30DayWeekDayDays_since_ConfirmedCases=1Days_since_Fatalities=1Days_since_ConfirmedCases=10Days_since_Fatalities=10Days_since_ConfirmedCases=100Days_since_Fatalities=100Distance_to_originCountryAreaCountryPopFemaleCountryPopMaleCountryPopTotalCountryPop_0-20CountryPop_20-40CountryPop_40-60CountryPop_60-80CountryPop_80+CountryPopDensityCountrySmokingRateCountryHospitalBedsRateCountryHealthExpenditurePerCapitaPPPPredictedLogNewConfirmedCasesPredictedLogNewFatalitiesPredictedConfirmedCasesPredictedFatalities
16506NaNHeilongjiangChina47.8620127.76152020-03-25NaNNaN2465.0NaN0.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.00.00.0000000.00.00.00.00.00.00.00.00.00.00.0000000.0000000.0000000.00.0...0.00.00.00.0000000.00.00.00.00.00.00.00.00.00.0632NaNNaNNaNNaNNaNNaN2291.4521009388210.0698159.433735624.2591433783.692338099.542416397.382437824.994215315.36226146.4120.15272225.64.2841.1149290.0000000.0016740.00.0
16507NaNKansasUS38.5266-96.72652020-03-25NaNNaN10033.0NaN2.8332133.0445222.0794422.6390572.3978952.8332130.0000002.0794421.3862940.0000001.3862941.6094380.00.00.6931470.00.00.00.00.00.00.00.00.00.00.0000000.0000000.0000000.00.0...0.00.00.00.0000000.00.00.00.00.00.00.00.00.00.063215.012.09.0NaN1.0NaN11746.2648949147420.0166238.618162826.299329064.91782289.91690159.13482846.12660854.33212915.4090.035974NaNNaNNaN2.9452600.04703197.01.0
16508NaNHainanChina19.1959109.74532020-03-25NaNNaN2379.0NaN0.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.00.00.0000000.00.00.00.00.00.00.00.00.00.00.0000000.0000000.0000000.00.0...0.00.00.00.6931470.00.00.00.00.00.00.00.00.00.063263.058.060.0NaN48.0NaN1329.4092369388210.0698159.433735624.2591433783.692338099.542416397.382437824.994215315.36226146.4120.15272225.64.2841.1149290.7031970.019845180.07.0
16509NaNMissouriUS38.4561-92.28842020-03-25NaNNaN10420.0NaN4.2626804.4773373.2958373.0910423.1354942.6390572.0794421.7917590.6931470.6931471.0986120.6931470.00.00.6931470.00.00.00.00.00.00.00.00.00.00.0000000.0000000.0000000.00.0...0.00.00.00.0000000.00.00.00.00.00.00.00.00.00.063215.06.08.0NaN3.0NaN11908.8083779147420.0166238.618162826.299329064.91782289.91690159.13482846.12660854.33212915.4090.035974NaNNaNNaN4.0696440.518520201.02.0
16510NaNHebeiChina39.5490116.13062020-03-25NaNNaN2422.0NaN0.0000000.0000000.6931470.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.00.00.0000000.00.00.00.00.00.00.00.00.00.00.6931471.7917590.6931470.00.0...0.00.00.00.0000000.00.00.00.00.00.00.00.00.00.063263.062.059.0NaN52.0NaN1013.6701829388210.0698159.433735624.2591433783.692338099.542416397.382437824.994215315.36226146.4120.15272225.64.2841.1149290.8015810.012528330.06.0
\n", + "

5 rows × 97 columns

\n", + "
" + ], + "text/plain": [ + " Id Province/State ... PredictedConfirmedCases PredictedFatalities\n", + "16506 NaN Heilongjiang ... 0.0 0.0\n", + "16507 NaN Kansas ... 97.0 1.0\n", + "16508 NaN Hainan ... 180.0 7.0\n", + "16509 NaN Missouri ... 201.0 2.0\n", + "16510 NaN Hebei ... 330.0 6.0\n", + "\n", + "[5 rows x 97 columns]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 55 } ] }, { "cell_type": "markdown", "metadata": { - "id": "2PHeaW7g_AGl", + "id": "kGvuRuKs_Fyz", "colab_type": "text" }, "source": [ - "Let's look at LightGBM models' feature importances. \n", - "1. \"split” means result contains numbers of times the feature is used in a model. \n", - "2. “gain” means result contains total gains of splits which use the feature." + "## Plots with predictions \n", + "\n", + "Let's plot some example graphs with predictions.\n", + "Plots will use data from reconcatenated main_df." ] }, { "cell_type": "code", "metadata": { - "id": "c0B5bbgj_G8m", + "id": "XcXpdsQjTodz", "colab_type": "code", - "outputId": "a31156a0-56aa-4203-d288-dfe0474f8e83", - "colab": { - "base_uri": "https://localhost:8080/", - "height": 1000 - } + "colab": {} }, "source": [ - "for prediction_name in ['ConfirmedCases', 'Fatalities']:\n", - " for importance_type in ['split', 'gain']:\n", - " lgb.plot_importance(lightgbm_models[prediction_name], title='LightGBM: prediction of %s. Feature importance. Type=%s' % (prediction_name, importance_type))" + "main_df = pd.concat([train_df, eval_df, test_df])" ], - "execution_count": 18, - "outputs": [ - { - "output_type": "display_data", - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAEWCAYAAAD7KJTiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "tags": [] - } - }, - { - "output_type": "display_data", - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAEWCAYAAAD7KJTiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDIiIiKSckgEREZGUUzIgIiKScv8PJrMfD1X+aMcAAAAASUVORK5CYII=\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "tags": [] - } - }, - { - "output_type": "display_data", - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAEWCAYAAABL4c8hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "tags": [] - } - }, - { - "output_type": "display_data", - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAEWCAYAAABL4c8hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO3deZgV1ZnH8e8riwooiyDQyiIuKNKI\nimtc2iDuETWMxGUENTGJozETUMmMK0km7RZB1CiRKGqiElExxg2Vxh0VBVxRIygIoqIgIALdvPPH\nOd0Wl15uA7dvV/fv8zz36VpPvaeqbr1Vp6rrmrsjIiIi9d9m+Q5AREREsqOkLSIikhJK2iIiIimh\npC0iIpISStoiIiIpoaQtIiKSEkra9YCZHWxms7OctsjM5uc6prTKXD9m9raZFW1AOVlvk03JzHqa\n2QwzW2Zmv6rr5ccYhprZ84n+5WbWo5rpN2gdNzZm9piZDcl3HJI/ZnaLmV26MWUoadchM5trZodn\nDnf359y95yZaxh1m9vtKhv/EzKaZ2Qoz+zx2n2tmlphvdTxALzOz6WZ2aGL+oWbmZnZ9RrkD4/A7\nNkX8m5q77+7uJTVNF+uwU2K+TbZNaukiYIq7b+XuN2SONLMSM/subqfyzwHVFVjVPpEtd2/l7h9V\nVVa263hDmVn3uH2SdZ65icpsuqnirIm7H+3u4+tqedWJ+9FP8x1HVczsfxLb+jszK0v0v53v+DaU\nu//C3X+3MWUoaTcCZjYMGA1cA3QCOgK/AH4ANE9MerW7twK2Bv4MPGBmTRLj/w2cnHGgGwK8n8PY\nm9Q8VYPSDajpoHReTKTln5fqIrB6oE2iznvkMxALUnf8TEvc7v5/5duacKx6KbHtd893fHnl7vrU\n0QeYCxxeyfAiYH6ify/gDWAZ8A/gPuD3yWmBYcDnwELgzDjuHGANsBpYDvwTaA2sAH5cQ2x3lC8j\n9rcAHCiI/UOB54HHgWPjsHbAZ4STgTuyXAfl8f8P8GVcJ6dlxPFn4NEY9+FAATAR+AKYA/wqMf2W\ncZ6vgXeACzPWZcU6B5rE5f47rtvpQBfg2VjXFXG9Da5km+wGlABLCEn1+IyYbwL+FcudBuxYzTo4\nPpaxJJa5Wxz+DFAGfBfj2KWSeUuAn1Yy/B9xWyyN9dm9qn0iDh+RWA/vACcmyhoKPJ/od2CnaspK\nruPNEmUvBiYA7eK4LYC74/AlwKtAxyz2me4xhqaVjNsVmAx8BcwGTk6MO5bwPfoGmAdckRj3SSxz\nefwcAFwB3F3VcuO6/wPwArAyrpMql1/dtovr+AXg+rguPgIOjMPnEb7bQzL2sVvispYBU4FuifEH\nxvW5NP49MGO5ybj/xrr72Y1xutFx2d8QvhsHJ8q4Im7LO+Py3wb6JcZ3AR4gfEcXl5cZx50FvEv4\njj6RjDvLY8ZQ4v5I+H5PzBh/AzA6Udc/Aq/Eekwi7n9x/P7Ai3GdzwSKahNLLGMHwndsGfAU4buf\n3G8q/S5mHmep5lhe7fJrG7A+G/4hi6RNuPL9GLgAaAacRDhIJjd0KTAyjj8G+BZom7lTxP6j4vTr\nHfAyYkjuTE0IZ7cfAU3isKGEpH0qcF8cdi5wK/B7EkkbmAWcWsVyyuP/E7A5cCghWfZMxLGU0Aqw\nGeHkYTpwWVw3PWJcR8bpi4HnCCcQXYC3qDppXwi8CfQEDNgD2CaOc2CnKrZJM+BDQsJvDvyQ8IVN\nxrwY2BdoSjgo3ltF/XeJ9R0Qy70olt08ji+hkqScmL/S8YQD41ZxnY4CZlS2bRPD/oNwMrQZ4SRl\nBdA5ua0T01asmyrKSq7jC4CXge1jLLcC98RxPyecSLYg7GN7A1tn8b3pTiVJG2hJSDJnxvW+J+FE\nsFdiGxbGOvYBFgEnVFUm2SXtT4Dd4/JaV7f86rZdXMelcd4mhO/QJ4QEsDlwBGEfa5VY78uAQ+L4\n0XyfyNoREuJ/xjhOif3bVBF3MyrZj4DTgW3iNMMIiWeLxLr5jnC8aUJIjC8njhczCScgLQknZwfF\ncQMJ+/dusdxLgBdredwcmqhrZ8K+2ib2NyUkvL0Tdf0U6B1jmVi+TYHtCN/TY+I+MSD2d4jjbyYk\n88o+sxLxvARcSzgWHEQ4OUjuN1l9F6nhWF7l+qjNytNn4z5kl7QPiTudJcY/n7GhV7LuweZzYP/M\nnSL2nw58lrG88jPNlcAhifm+Swz/jnWvgIfGOLYkHPxaEw7OPyAjadewDsp31JaJYROASxNx3JkY\ntx/wSUYZvwVuj90fAUclxp1D1Ul7NjCwiriqS9oHEw5gmyXG30O8cosx35YYdwzwXhXLuRSYkOjf\nLG7vothfQs1J+1u+P5i8Xsk0bWJ9Wle2T1RR7ozydcPGJe13gf6JcZ0JV+dNCQezF4E+tfzedI8x\nJA+iwwknG89lTHsrcHkV5YwCrs8os7ZJe2RifG2XX7Ft4zr+IDGuMC6rY2LYYqBvYr3fmxjXinC1\n3IWQrF/JWNZLwNDK4s5mP4vTfA3skVg3TyXG9QJWxu4DCFfYlbWEPAacnbG/f0strrYr2R8fA34W\nu48D3smoV3FGnKsJJxYXA3dllP0EiRaNLGLpSjh+tUgMuzu532T7XaSGY3lVn3p/b6MRKgA+9bgF\no3kZ0yx299JE/7eEL3FlFgPtk/eh3f1Ad28TxyX3gWvj8BZAP+AaMzs6WZi7ryQ0A19COJN/Ifuq\nVfja3Vck+j8m1Ltcsr7dgAIzW1L+IVzxdozjCzKm/7ia5XYhNNvWVgEwz93XZixnu0T/Z4nu6rZH\nQTLGWOa8jLJq8it3bxM/e5lZEzMrNrN/m9k3hCQK0L6qAszsjPiUevk67V3d9LXQDXgwUe67hOTS\nEbiLcJC818wWmNnVZtasFmW3T9T72ris/TL2jdMIz21gZvuZ2RQz+8LMlhJajza2jpn7ZpXLz8Ki\nRPdKAHfPHJbcjyqW7e7LCU3yBWTsU1Hm/pl5DFmPmQ03s3fNbGmsS2vWXV+Z+/gW8bjSBfg445hU\nrhswOrF+viK0ctVmf880nnAxQvx7V8b4zONBM0I9ugH/kbG9DiKcWGarAPjK3b+tbHkb8F2szbEc\n0INo9dFCYLvyp7qjLrWY3zP6XwJWEZqpsisgeItwD+zYSia5k9B8dnct4kpqa2YtE/1dgQXJEBLd\n84A5iYN1Gw9PVh8Txy9k3fXTtZrlzgN23IB4FwBdMh7g6Uq4Qt6QsrqV98Tt3GUDyyp3KmH7Hk44\n0HYvLz7+XWefMLNuwF+A8wgnXm0ItxWS+1xVMvevTPOAozO21xbu/qm7r3H3K929F+Ee7HHAGVks\ns7plTc1YVit3/2Uc/3fgYaCLu7cm3BOudJ1EKwgnrOUqS76Z+2Z1y9/UKvZzM2tFaBZfQMY+FWXu\nn5n1zdwnDibcqjmZ0DzbhnCbKpt9Yh7QtYon8ecBP89YR1u6+4tZlFuVh4A+ZtabsA/9LWN85vFg\nDeG2xTzClXYylpbuXgwV/461vIpP+cOhC4F2ZtaiiuXV9F3caErada+ZmW2R+GTu6C8RrkzOM7Om\nZjaQcK80W4sI930BcPclwJXAzWY2yMy2MrPNzKwv4Z5PpcxsV8JZaGVPMk8l3A8aU4u4Ml1pZs3j\nweI4wsMblXkFWGZmF5vZlvFMtreZ7RPHTwB+a2ZtzWx74Pxqlnkb8Dsz2zk+RdvHzLaJ49ZZbxmm\nEc6ALzKzZhb+J/lHwL1Z1/Z7E4Bjzax/vMocRjip2piD2FaxjMWEpPN/GeMz69aScND+AsDMziRc\naWejuvUEITH+IZ4YYGYd4j6MmR1mZoXxPwK+IRxM11ZdVI0eAXYxs/+M26WZme1jZrvF8VsRroq+\nM7N9CQfUcl/EZSfrMgM4xMy6mllrwm2YjVn+pnaMmR1kZs2B3xHuKc8jPLS5i5mdGo8ZgwnNwo9U\nU1bmdtyK0Oz7BdDUzC4j/BdJNl4hJLNiM2sZj2s/iONuIXw/dwcws9Zm9h9Zllspd/8OuJ9wUvaK\nu3+SMcnpZtYrJtaRwP3uXka4yPiRmR0ZjyNbWHivw/ax3F/4uv+VkfzsHqf5GHgNuCIevw4gHAvK\n1fRd3GhK2nXvUUKzV/nniuRId19NePjsbMK9u9MJX75VWZY/DugVm38eimVeDfyGcCa9KH5uJdzj\nSSaLi+JZ5QrgSeD2ON064pX40+7+VWUBWHjZxmnVxPgZ4X7ZAsJZ8i/c/b3KJoxftuOAvoQnx78k\nJN/WcZIrCU1gc2LMmU1lSX8iJM0nCUljHOEePYTtMD6ut5MzYlhN+GIeHZd/M3BGVTFXx91nE7bp\nmFjWj4AfxWVsqDsJ6+BTwpPgL2eMX2efcPd3gOsIJ4iLCPdTs73Nsd7+lWE04er2STNbFmPZL47r\nRDjYfkNoNp9K3F7xKueWLGMAwN2XER7Y+glhX/oMuIrwABCEByVHxjguI2z78nm/JT5RHeuyv7tP\nJvynxizCw4/VJb1slr+p/R24nNDEvDexidjdFxO+I8MIyeIi4Dh3/7KaskYDg8zsazO7gXDb4nHC\nv29+THimpcYm9bj8MsJ+vBPhgbf5hPv9uPuDhHVyb2wufovwPQKyOlZUZTxhv63s+34X4d7xZ4SH\n4n4VY5lHuAr+H8LJyTzCw6m1zYOnEe7jLyY8z3Mf3x+fa/oubjRb99ap1EdmNg24xd1vz3csGyte\npd7t7tvnOxaRtLDw8qL57n5JvmOpD8ysK/Ae0Mndv0kMLyEcX26rw1juIzx4enldLE9X2vWQmR1q\nZp1iU9cQwr+rPJ7vuERE8i0+W/IbwtP039Q0fQ6Wv4+Z7RhvMx5FuHqvrNUpJ+rsFX5SKz0JTXkt\nCf/SNMjdF+Y3JBGR/LLwAOsiQhP0UXkKoxPhRTLbEG4F/NLd36irhat5XEREJCXUPC4iIpISah5v\nBNq0aeM77bRTzROm3IoVK2jZssr/YmswGks9ofHUVfWsn6ZPn/6lu3fIdxxJStqNQMeOHXnttdfy\nHUbOlZSUUFRUlO8wcq6x1BMaT11Vz/rJzKp7w2JeqHlcREQkJZS0RUREUkJJW0REJCWUtEVERFJC\nSVtERCQllLRFRERSQklbREQkJZS0RUREUkJJW0REJCWUtEVERFJCSVtERCQllLRFRERSQklbREQk\nJZS0RUREUkJJW0REJCWUtEVERFJCSVtERCQllLRFRERSQklbREQkJZS0RUREUkJJW0REJCWUtEVE\nRFJCSVtERCQllLRFRERSQklbREQkJZS0RUREUkJJW0REJCWUtEVERFJCSVtERCQllLRFRERSQklb\nREQkJZS0RUREUkJJW0REJCWUtEVERFJCSVtERCQllLRFRERSQklbREQkJczd8x2D5FjXHjv5ZieP\nzncYOTessJTr3mya7zByrrHUExpPXVXPqs0tPjZH0dTMzKa7e7+8BVAJXWmLiEi9d9ZZZ7HtttvS\nu3fvimFfffUVAwYMYOedd2bAgAF8/fXX68zz6quv0rRpU+6///6KYZ988glHHHEEu+22G7169WLu\n3LkAnH322eyxxx706dOHQYMGsXz58jqpV20paeeRmZWZ2Qwze9vMZprZMDPTNhERyTB06FAef/zx\ndYYVFxfTv39/PvjgA/r3709xcXHFuLKyMi6++GKOOOKIdeY544wzuPDCC3n33Xd55ZVX2HbbbQG4\n/vrrmTlzJrNmzaJr167ceOONua/UBlCCyK+V7t7X3XcHBgBHA5fnOSYRkXrnkEMOoV27dusMmzRp\nEkOGDAFgyJAhPPTQQxXjxowZw49//OOKpAzwzjvvUFpayoABAwBo1aoVLVq0AGDrrbcGwN1ZuXIl\nZpbT+mwoJe16wt0/B84BzrOgu5k9Z2avx8+BAGZ2p5mdUD6fmf3NzAbmK24RkXxZtGgRnTt3BqBT\np04sWrQIgE8//ZQHH3yQX/7yl+tM//7779OmTRtOOukk9txzTy688ELKysoqxp955pl06tSJ9957\nj/PPP7/uKlILDf/JhxRx94/MrAmwLfA5MMDdvzOznYF7gH7AOOC/gYfMrDVwIDAksywzO4dwEkD7\n9h24rLC0jmqRPx23DA+6NHSNpZ7QeOqqelatpKSkovuzzz5jxYoVFcNKS0vXGV9WVkZJSQlXXHEF\ngwcP5tlnn+Wzzz7j7bffpn379sycOZOSkhLGjh1Lx44dufLKKxkxYgTHHhsedhsyZAinn346N9xw\nA1deeeXGVjcnlLTrr2bAjWbWFygDdgFw96lmdrOZdQB+DEx09/W+Be4+FhgL4elxPZnacDSWekLj\nqavqWbW5pxV93z13Li1btqSoKAzbbrvt6NmzJ507d2bhwoUUFBRQVFTExx9/zNVXXw3Al19+yeuv\nv84ee+zBkUceyTPPPMOpp54KwIIFC3j55ZcryivXrFmzivnrGzWP1yNm1oOQoD8nXE0vAvYgXGE3\nT0x6J3A6cCbw1zoOU0SkXjj++OMZP348AOPHj2fgwHCncM6cOcydO5e5c+cyaNAgbr75Zk444QT2\n2WcflixZwhdffAHAM888Q69evXB3PvzwQyDc03744YfZdddd81OpGihp1xPxyvkW4EYP/zzfGljo\n7muB/wSaJCa/A/g1gLu/U8ehiojUuVNOOYUDDjiA2bNns/322zNu3DhGjBjB5MmT2XnnnXnqqacY\nMWJEtWU0adKEa6+9lv79+1NYWIi787Of/Qx3Z8iQIRQWFlJYWMjChQu57LLL6qhmtaOXq+SRmZUB\nbxKawkuBu4A/ufvaeB97IuDA48B/uXurxLyPAw+5+y01Ladnz54+e/bsXFShXikpKVmvmashaiz1\nhMZTV9WzfqqPL1dp+DdR6jF3b1LNuA+APolBF5d3mFkLoPzhNBERaSTUPJ4yZnY48C4wxt2X5jse\nERGpO7rSThl3fwrolu84RESk7ulKW0REJCWUtEVERFJCSVtERCQllLRFRERSQklbREQkJZS0RURE\nUkJJW0REJCWUtEVERFJCSVtERCQllLRFRERSQklbREQkJZS0RUREUkJJW0REJCWUtEVERFJCSVtE\nRCQllLRFRERSQklbREQkJZS0RUREUkJJW0REJCWUtEVERFJCSVtERCQllLRFRERSQklbREQkJZS0\nRUREUkJJW0REJCWUtEVERFJCSVtERCQllLRFRERSQklbREQkJZS0RUREUkJJW0REJCWa5jsAyb2V\na8roPuJf+Q4j54YVljJU9WxQ8lnXucXHMnv2bAYPHlwx7KOPPmLkyJGcccYZDB48mLlz59K9e3cm\nTJhA27ZtWbp0KaeffjqffPIJpaWlDB8+nDPPPJOPP/6YE088kbVr17JmzRrOP/98fvGLX+SlXpJu\nutKuZ8xseS2mLTKzA3MZj0hj1rNnT2bMmMGMGTOYPn06LVq04MQTT6S4uJj+/fvzwQcf0L9/f4qL\niwG46aab6NWrFzNnzqSkpIRhw4axevVqOnfuzEsvvcSMGTOYNm0axcXFLFiwIM+1kzRS0k63IkBJ\nW6QOPP300+y4445069aNSZMmMWTIEACGDBnCQw89BICZsWzZMtyd5cuX065dO5o2bUrz5s3ZfPPN\nAVi1ahVr167NWz0k3ZS0U8DMfmRm08zsDTN7ysw6mll34BfAf5vZDDM7OL9RijRs9957L6eccgoA\nixYtonPnzgB06tSJRYsWAXDeeefx7rvvUlBQQGFhIaNHj2azzcJhdt68efTp04cuXbpw8cUXU1BQ\nkJ+KSKrpnnY6PA/s7+5uZj8FLnL3YWZ2C7Dc3a/NnMHMzgHOAWjfvgOXFZbWbcR50HHLcA+0oWss\n9YT81rWkpKSie82aNUycOJHjjjuOkpISSktL1xlfVlZGSUkJU6dOpX379vz9739nwYIF/PSnP+W2\n226jZcuWANxwww18+eWXXHrppXTu3Jl27doBsHz58nXKa6gaSz1zSUk7HbYH7jOzzkBzYE5NM7j7\nWGAsQNceO/l1bzb8TT2ssBTVs2HJZ13nnlZU0T1p0iT2228/TjrpJAC22247evbsSefOnVm4cCEF\nBQUUFRVxzTXXMGLECA4+ODR8jRs3jg4dOrDvvvuuU/ajjz7K2rVrKSoKyygpKanobsgaSz1zSc3j\n6TAGuNHdC4GfA1vkOR6RRuWee+6paBoHOP744xk/fjwA48ePZ+DAgQB07dqVp59+GghN6LNnz6ZH\njx7Mnz+flStXAvD111/z/PPP07NnzzquhTQEWZ3CmtmOwHx3X2VmRUAf4E53X5LL4KRCa+DT2D0k\nMXwZsHXdhyPSeKxYsYLJkydz6623VgwbMWIEJ598MuPGjaNbt25MmDABgEsvvZShQ4dSWFiIu3PV\nVVfRvn17Jk+ezLBhwzAz3J3hw4dTWFiYrypJimXb7jQR6GdmOxGaXCcBfweOyVVgjVgLM5uf6P8T\ncAXwDzP7GngG2CGO+ydwv5kNBM539+fqNFKRRqBly5YsXrx4nWHbbLNNxRV1UkFBAU8++eR6wwcM\nGMCsWbNyFqM0Htkm7bXuXmpmJwJj3H2Mmb2Ry8AaK3ev6pbFpEqmfZ/Q6lGtLZs1YXbxsRsbWr1X\nUlKyzn3Ihqqx1BMaV11FspHtPe01ZnYKoWn2kTisWW5CEhERkcpkm7TPBA4A/uDuc8xsB+Cu3IUl\nIiIimbJqHnf3d8zsYqBr7J8DXJXLwERERGRdWV1pm9mPgBnA47G/r5k9nMvAREREZF3ZNo9fAewL\nLAFw9xlAjxzFJCIiIpXI+kE0d1+aMUxvvBcREalD2f7L19tmdirQxMx2Bn4FvJi7sERERCRTtlfa\n5wO7A6sIL1VZCvw6V0GJiIjI+mq80jazJsC/3P0w4H9zH5KIiIhUpsYrbXcvA9aaWes6iEdERESq\nkO097eXAm2Y2GVhRPtDdf5WTqERERGQ92SbtB+JHRERE8iTbN6KNz3UgIiIiUr1sf097DuCZw91d\nL1gRERGpI9k2j/dLdG8B/AfQbtOHIyIiIlXJ6v+03X1x4vOpu48CGv4PNIuIiNQj2TaP75Xo3Yxw\n5Z3tVbqIiIhsAtkm3usS3aXAHODkTR+OiIiIVCXbpH22u3+UHGBmO+QgHhEREalCtu8evz/LYSIi\nIpIj1V5pm9muhB8KaW1mJyVGbU14ilxERETqSE3N4z2B44A2wI8Sw5cBP8tVUCIiIrK+apO2u08C\nJpnZAe7+Uh3FJCIiIpXI9kG0N8zsvwhN5RXN4u5+Vk6iEhERkfVk+yDaXUAn4EhgKrA9oYlcRERE\n6ki2SXsnd78UWBF/PORYYL/chSUiIiKZsk3aa+LfJWbWG2gNbJubkERERKQy2d7THmtmbYFLgYeB\nVsBlOYtKRERE1pPt72nfFjunAvo5ThERkTzIqnnczDqa2Tgzeyz29zKzs3MbmoiIiCRle0/7DuAJ\noCD2vw/8OhcBiYiISOWyTdrt3X0CsBbA3UuBspxFJSIiIuvJNmmvMLNtAAcws/2BpTmLSkRERNaT\n7dPjvyE8Nb6jmb0AdAAG5Swq2aRWrimj+4h/5TuMnBtWWMpQ1bNByXVd5xYfm7OyRXKh2ittM+sK\n4O6vA4cCBwI/B3Z391m5D09EJLeWLFnCoEGD2HXXXdltt9146aWXmDFjBvvvvz99+/alX79+vPLK\nKwD87W9/o0+fPhQWFnLggQcyc+ZMAGbPnk3fvn0rPltvvTWjRo3KZ7WkgarpSvshYK/YfZ+7/zjH\n8YiI1KkLLriAo446ivvvv5/Vq1fz7bffcvLJJ3P55Zdz9NFH8+ijj3LRRRdRUlLCDjvswNSpU2nb\nti2PPfYY55xzDtOmTaNnz57MmDEDgLKyMrbbbjtOPPHEPNdMGqKa7mlbortW/59tZmVmNsPM3jaz\nmWY2zMw2i+P6mdkN1czb3cxOrc3yNrWNicHMiszskYxhd5hZtbcUzOwKMxtey2W9uCExiggsXbqU\nZ599lrPPDv/B2rx5c9q0aYOZ8c0331RMU1AQ/nHmwAMPpG3btgDsv//+zJ8/f70yn376aXbccUe6\ndetWR7WQxqSmK22vojsbK929L4CZbQv8HdgauNzdXwNeq2be7sCpcZ58qQ8x1MjdD8x3DCJpNWfO\nHDp06MCZZ57JzJkz2XvvvRk9ejSjRo3iyCOPZPjw4axdu5YXX1z/3HjcuHEcffTR6w2/9957OeWU\nU+oifGmEzL3qXGxmZcAKwhX3lsC35aMAd/etq5l3ubu3SvT3AF4F2hPujw939+PM7FBgdJzMgUOA\nycBuwBxgPPAg4ZfGWsbpznP3F82sCLgC+BLoDUwHTnd3N7N9YrktgVVA/xh/MVAEbA7c5O63VhH/\nyxkx/Dl++gGlwG/cfUoV8xaV1y8x7A7gEXe/38zmAv3c/Usz6wdc6+5FZnYFsCOwU1xPV7v7X+L8\nFwInx7gfdPfLK1vPieWdA5wD0L59h70vG/WXykJtUDpuCYtW5juK3Gss9YTc17X58s8499xzGTNm\nDL169WLMmDG0bNmS5cuXs8cee3DooYcyZcoUHnnkEa677rqK+d544w1GjRrFDTfcQOvWrSuGr1mz\nhkGDBnH77bfTrl27rONYvnw5rVqt9zVucNJWz8MOO2y6u/fLdxxJ1V5pu3uTTbUgd//IzJqw/g+N\nDAf+y91fMLNWwHfACBJJz8xaAAPc/Tsz2xm4h5A8AfYk/M73AuAF4Adm9gpwHzDY3V81s62BlcDZ\nwFJ338fMNgdeMLMn3X1OJSFnxjAsVMMLzWxX4Ekz28Xdv6uiygeb2YxEf1fgkSqmTeoD7E842XjD\nzP5FOCHZGdiXcML0sJkd4u7PVlWIu48FxgJ07bGTX/dmtv8okF7DCktRPRuWXNf15V8P5I9//CPn\nnnsuAE2aNKG4uJjnn3+eiRMnYmYceuihXH/99RQVFQEwa9YsbrzxRiZPnswuu+yyTnmTJk1iv/32\n46STTqpVHCUlJRXlN2SNpZ65lO3/aefSC8CfzOxXQJv44pZMzYC/mNmbwD+AXolxr7j7fHdfC8wg\nNGv3BBa6+6sA7v5NLPcI4IyYTKcB2xCSYTYOAu6O5b0HfAzsUs30z7l73/IP4V/msjHJ3Ve6+5fA\nFEKiPiJ+3gBeB3atRdwiUvzzbN8AABD0SURBVIVOnTrRpUsXZs+eDYT70b169aKgoICpU6cC8Mwz\nz7DzzuHr9sknn3DSSSdx1113rZewAe655x41jUtO1dnpemweLwM+JzQ7A+DuxfFq8hjCle+Rlcz+\n38AiYA/CiUby6nZVoruM6utkwPnu/sQGVWLTKeX7E6YtMsZl3q9wQtx/rKopX0Q23JgxYzjttNNY\nvXo1PXr04Pbbb2fgwIFccMEFlJaWssUWWzB27FgARo4cyeLFiyuuzJs2bcprr4XHc1asWMHkyZO5\n9VZ9TSV36iRpm1kH4Bbgxni/OTluR3d/E3gz3ofeFZgHbJUoojUw393XmtkQoKZm+9lAZzPbJzaP\nb0VoHn8C+KWZPePua8xsF+BTd19RSRnLMmJ4DjgNeCbO1zUuZ0PMBfYGHgMy/41uoJn9kdA8XkRo\npl8J/M7M/ubuy81sO2CNu3+ezcK2bNaE2Y3gJRIlJSXMPa0o32HkXGOpJ9RNXfv27VuReMsddNBB\nTJ8+fb1pb7vtNm677bb1hgO0bNmSxYsX5yRGkXK5TNpbxmboZoQry7uAP1Uy3a/N7DDCe83fJiSy\ntUCZmc0k/FjJzcBEMzsDeJzwcFyV3H21mQ0GxpjZloSkdzhwG6H5/HULZw5fACdUUcysSmL4c2yi\nLwWGuvuqKuatyZXAODP7HVBSyXKnEB5E+527LwAWmNluwEvxhGc5cDqh1UJERBqJnCXt6h5ic/cS\nYrJy9/OrmOyHGf19Et0XZ5YT+89LdL9KeKAr0//ET7XcfU0lMZxZ03yVxRWHDU10P0cl98Pd/Ypq\nyhzN90/ZJ4en51FMERHZKPXhQTQRERHJQuP4v5FqmFkhoek+aZW775fFvEcCV2UMnuPuen+hiIhs\nco0+aceH4Ppu4LxPEB5uExERyTk1j4uIiKSEkraIiEhKKGmLiIikhJK2iIhISihpi4iIpISStoiI\nSEooaYuIiKSEkraIiEhKKGmLiIikhJK2iIhISihpi4iIpISStoiISEooaYuIiKSEkraIiEhKKGmL\niIikhJK2iIhISihpi4iIpISStoiISEooaYuIiKSEkraIiEhKKGmLiIikhJK2iIhISihpi4iIpISS\ntoiISEooaYuIiKSEkraIiEhKKGmLiIikhJK2iIhISihpi4iIpETTfAcgubdyTRndR/wr32Hk3LDC\nUoaqnvXe3OJjAejevTtbbbUVTZo0oWnTprz22msAjBkzhptuuokmTZpQWFhIUVERkydPZsSIEaxe\nvZrmzZtzzTXX8MMf/pBly5Zx8MEHV5Q9f/58Tj/9dEaNGpWXuonkmpJ2HTKz5e7eKt9xiNQXU6ZM\noX379uv0T5o0iZkzZ7L55pvz4IMPAtC+fXv++c9/UlBQwFtvvcWRRx7Jp59+ylZbbcWMGTMq5t97\n77056aST6rweInVFSVtE6o0///nPjBgxgs033xyAtm3bArDnnntWTLP77ruzcuVKVq1aVTEdwPvv\nv8/nn3++zpW3SEOje9p5ZmZ9zexlM5tlZg+aWds4vMTMrjKzV8zsfTM7OA5vYWYTzOydOP00M+uX\n31qI1J6ZccQRR7D33nszduxYICTe5557jv32249DDz2U9957b735Jk6cyF577bVOwga49957GTx4\nMGZWJ/GL5IOutPPvTuB8d59qZiOBy4Ffx3FN3X1fMzsmDj8cOBf42t17mVlvYEZlhZrZOcA5AO3b\nd+CywtJc1yPvOm4Z7vc2dGmvZ0lJCQBXX301HTp04Ouvv2b48OGsXLmSpUuX8uabb1JcXMx7773H\n5ZdfTs+ePSsS8Zw5c7jkkku4+uqrK8op99e//pXf/va36w1Pg+XLl6cy7tpqLPXMJSXtPDKz1kAb\nd58aB40H/pGY5IH4dzrQPXYfBIwGcPe3zGxWZWW7+1hgLEDXHjv5dW82/E09rLAU1bP+m3ta0XrD\nZs6cyZo1a+jZsyfnn38+hx12GIcddhi///3v6d27Nx06dGD+/Pmcc845TJgwgR/84Afrzd+8eXN+\n/vOf11EtNq2SkhKKioryHUbONZZ65pKax+u3VfFvGTrBkgZkxYoVLFu2rKL7ySefpHfv3pxwwglM\nmTIFCE3la9asoX379ixZsoRjjz2W4uLi9RI2wD333MMpp5xSp3UQyQclgjxy96Vm9rWZHezuzwH/\nCUytYbYXgJOBKWbWCyjMdZwim9qiRYs48cQTASgtLeXUU0/lqKOOYvXq1Zx11ln07t2b5s2bM2LE\nCMyMG2+8kQ8//JCRI0cycuRIAJ588km23XZbACZMmMCjjz6at/qI1BUl7brVwszmJ/r/BAwBbjGz\nFsBHwJk1lHEzMN7M3gHeA94GluYiWJFc6dGjBzNnzlxvePPmzbn77rsr+svvf15yySVccsklVZb3\n0UcfbfIYReojJe065O5V3Y7Yv5JpixLdX/L9Pe3vgNPd/Tsz2xF4Cvi4uuVu2awJs+MLLRqykpKS\nSu+XNjSNpZ4isj4l7fRpQWgabwYYcK67r85zTCIiUgeUtFPG3ZcB+r9sEZFGSE+Pi4iIpISStoiI\nSEooaYuIiKSEkraIiEhKKGmLiIikhJK2iIhISihpi4iIpISStoiISEooaYuIiKSEkraIiEhKKGmL\niIikhJK2iIhISihpi4iIpISStoiISEooaYuIiKSEkraIiEhKKGmLiIikhJK2iIhISihpi4iIpISS\ntoiISEooaYuIiKSEkraIiEhKKGmLiIikhJK2iIhISihpi4iIpISStoiISEooaYuIiKSEkraIiEhK\nKGmLiIikhJK2iIhISihpi4iIpISStoiISEooaYuIiKSEkraIiEhKKGmLiIikhJK2iIhISpi75zsG\nyTEzWwbMznccdaA98GW+g6gDjaWe0HjqqnrWT93cvUO+g0hqmu8ApE7Mdvd++Q4i18zsNdWzYWks\ndVU9JVtqHhcREUkJJW0REZGUUNJuHMbmO4A6ono2PI2lrqqnZEUPoomIiKSErrRFRERSQklbREQk\nJZS0GzgzO8rMZpvZh2Y2It/x1IaZdTGzKWb2jpm9bWYXxOHtzGyymX0Q/7aNw83Mboh1nWVmeyXK\nGhKn/8DMhuSrTtUxsyZm9oaZPRL7dzCzabE+95lZ8zh889j/YRzfPVHGb+Pw2WZ2ZH5qUj0za2Nm\n95vZe2b2rpkd0BC3qZn9d9xv3zKze8xsi4awTc3sr2b2uZm9lRi2ybafme1tZm/GeW4wM6vbGtZz\n7q5PA/0ATYB/Az2A5sBMoFe+46pF/J2BvWL3VsD7QC/gamBEHD4CuCp2HwM8BhiwPzAtDm8HfBT/\nto3dbfNdv0rq+xvg78AjsX8C8JPYfQvwy9h9LnBL7P4JcF/s7hW38ebADnHbN8l3vSqp53jgp7G7\nOdCmoW1TYDtgDrBlYlsObQjbFDgE2At4KzFsk20/4JU4rcV5j8739qxPH11pN2z7Ah+6+0fuvhq4\nFxiY55iy5u4L3f312L0MeJdwMBxIOPAT/54QuwcCd3rwMtDGzDoDRwKT3f0rd/8amAwcVYdVqZGZ\nbQ8cC9wW+w34IXB/nCSznuX1vx/oH6cfCNzr7qvcfQ7wIWEfqDfMrDXhoD8OwN1Xu/sSGuA2Jby8\nakszawq0ABbSALapuz8LfJUxeJNsvzhua3d/2UMGvzNRlqDm8YZuO2Beon9+HJY6sblwT2Aa0NHd\nF8ZRnwEdY3dV9U3DehgFXASsjf3bAEvcvTT2J2OuqE8cvzROn4Z67gB8AdwebwXcZmYtaWDb1N0/\nBa4FPiEk66XAdBrmNoVNt/22i92ZwyVS0pZ6z8xaAROBX7v7N8lx8Ww81f+3aGbHAZ+7+/R8x1IH\nmhKaVv/s7nsCKwjNqRUayDZtS7jK3AEoAFpS/1oCcqIhbL/6TEm7YfsU6JLo3z4OSw0za0ZI2H9z\n9wfi4EWxGY349/M4vKr61vf18APgeDObS7iF8UNgNKEpsfz3AZIxV9Qnjm8NLKb+1xPCldN8d58W\n++8nJPGGtk0PB+a4+xfuvgZ4gLCdG+I2hU23/T6N3ZnDJVLSbtheBXaOT6w2Jzzg8nCeY8pavKc3\nDnjX3f+UGPUwUP606RBgUmL4GfGJ1f2BpbHJ7gngCDNrG6+AjojD6gV3/627b+/u3Qnb6Bl3Pw2Y\nAgyKk2XWs7z+g+L0Hof/JD6JvAOwM+GhnnrD3T8D5plZzzioP/AODWybEprF9zezFnE/Lq9ng9um\n0SbZfnHcN2a2f1xvZyTKEtDT4w39Q3h6833CU6f/m+94ahn7QYRmtlnAjPg5hnCv72ngA+ApoF2c\n3oCbYl3fBPolyjqL8BDPh8CZ+a5bNXUu4vunx3sQDtAfAv8ANo/Dt4j9H8bxPRLz/2+s/2zq6VO3\nQF/gtbhdHyI8PdzgtilwJfAe8BZwF+EJ8NRvU+Aewn36NYSWk7M35fYD+sV19m/gRuKbO/UJH73G\nVEREJCXUPC4iIpISStoiIiIpoaQtIiKSEkraIiIiKaGkLSIikhJK2iIpZ2ZlZjYj8em+AWWcYGa9\nNn10YGYFZnZ/zVNu0mX2NbNj6nKZInWhac2TiEg9t9Ld+25kGScAjxBeAJIVM2vq379Hu0ruvoDv\nXyiSc/GNYn0J/+/7aF0tV6Qu6EpbpAGKv0k81cymm9kTiVdM/szMXjWzmWY2Mb6x60DgeOCaeKW+\no5mVmFm/OE/7+IpVzGyomT1sZs8AT5tZSwu/r/xK/AGQ9X5Fzsy6W/zt5Tj/QxZ+c3mumZ1nZr+J\n875sZu3idCVmNjrG85aZ7RuHt4vzz4rT94nDrzCzu8zsBcKLTEYCg+P8g81sXzN7KS7nxfI3ssV4\nHjCzxy38rvPVibiPMrPX47p6Og6rsb4iuaQrbZH029LMZsTuOcDJwBhgoLt/YWaDgT8Q3kD1gLv/\nBcDMfg+c7e5jzOxhwpvY7o/jqlveXkAfd//KzP6P8MrNs8ysDfCKmT3l7iuqmb834RfbtiC8Deti\nd9/TzK4nvLZyVJyuhbv3NbNDgL/G+a4E3nD3E8zsh4SfbixvZegFHOTuK81sKOHtW+fF+mwNHOzu\npWZ2OPB/wI/jfH1jPKuA2WY2BvgO+AtwiLvPKT+ZILydrLb1FdlklLRF0m+d5nEz601IcJNj8m1C\neO0kQO+YrNsArdiw93VPdvfy31M+gvBjJ8Nj/xZAV8Jvn1dlioffR19mZkuBf8bhbwJ9EtPdA+H3\nm81s65gkDyImW3d/xsy2iQkZ4GF3X1nFMlsD481sZ8KrcZslxj3t7ksBzOwdoBvh1arPevgNazay\nviKbjJK2SMNjwNvufkAl4+4ATnD3mfFqtKiKMkr5/vbZFhnjkleVBvzY3WfXIr5Vie61if61rHtM\nynzHck3vXK7uavd3hJOFE+ODeiVVxFNG9cfFDamvyCaje9oiDc9soIOZHQDh503NbPc4bitgoYWf\nPD0tMc+yOK7cXGDv2F3dQ2RPAOfHX2TCzPbc+PArDI5lHkT4dailwHPEuM2sCPjSM35jPcqsT2u+\n/4nHoVks+2XgEAu/rEWieTyX9RWpkZK2SAPj7qsJifYqM5tJ+HW0A+PoS4FpwAuEX6Aqdy9wYXy4\nakfgWuCXZvYG0L6axf2O0NQ8y8zejv2byndx+bcQfkkK4ApgbzObBRTz/c9BZpoC9Cp/EA24Gvhj\nLK/GFkZ3/wI4B3ggrsP74qhc1lekRvqVLxGpd8ysBBju7q/lOxaR+kRX2iIiIimhK20REZGU0JW2\niIhISihpi4iIpISStoiISEooaYuIiKSEkraIiEhK/D97Q5gKTp393wAAAABJRU5ErkJggg==\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "tags": [] - } - } - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "ovW1PQmNwOSt", - "colab_type": "text" - }, - "source": [ - "And the third GBDT package is XgBoost. XgBoost does not support categorical features out of the box, so we will transform original categorical features to one-hot features." - ] + "execution_count": 0, + "outputs": [] }, { "cell_type": "code", "metadata": { - "id": "54-SINtUwCsH", + "id": "bWFfwZA1_7uI", "colab_type": "code", - "outputId": "83d7e7d5-a758-4fff-9958-b4d668deacff", + "outputId": "09beab14-9b28-43d9-c4ee-e3dae161579f", "colab": { "base_uri": "https://localhost:8080/", - "height": 489 + "height": 1000 } }, "source": [ - "# returns df, dict of cat_feature -> OneHotEncoder\n", - "def preprocess_for_xgboost(df, cat_features):\n", - " df = df.copy()\n", - " encoders = {}\n", - " for cat_feature in cat_features:\n", - " # xgboost doesn't support sparse columns in pandas.DataFrame input\n", - " encoder = sklearn.preprocessing.OneHotEncoder(dtype=np.bool, sparse=False)\n", - "\n", - " ohe_matrix = encoder.fit_transform(df[cat_feature].to_numpy().reshape(-1, 1))\n", - " ohe_df = pd.DataFrame(\n", - " ohe_matrix,\n", - " columns=encoder.get_feature_names([cat_feature])\n", - " )\n", - " df = pd.concat([df, ohe_df], axis=1)\n", - " df.drop(cat_feature, axis=1, inplace=True)\n", + "%matplotlib inline\n", + "import matplotlib.pyplot as plt\n", + "import matplotlib\n", "\n", - " encoders[cat_feature] = encoder\n", - "\n", - " return df, encoders\n", "\n", + "def plot_graph(country_region, province_state, field, log_scale=True):\n", + " location_df = main_df.loc[\n", + " (main_df['Country/Region'] == country_region) & (main_df['Province/State'] == province_state)\n", + " ]\n", "\n", + " if province_state:\n", + " title = '%s for %s, %s' % (field, country_region, province_state)\n", + " else:\n", + " title = '%s for %s' % (field, country_region)\n", + " \n", + " plt.figure(figsize = (16, 10))\n", + " plt.suptitle(title, fontsize=14)\n", + " if log_scale:\n", + " plt.yscale('log')\n", "\n", - "print ('xgboost version', xgb.__version__)\n", + " for sub_field in [field, 'Predicted' + field]:\n", + " plt.plot(location_df['Date'], location_df[sub_field], label=sub_field)\n", "\n", - "preprocessed_for_xgboost_df, encoders = preprocess_for_xgboost(preprocessed_df, cat_features)\n", "\n", - "train_df, train_labels, test_df, test_labels = get_train_and_test_data(preprocessed_for_xgboost_df)\n", + " # add vertical lines splitting train, test and eval parts of the graph\n", + " ax = plt.gca()\n", "\n", - "xgboost_models = {}\n", + " # the x coords of this transformation are data, and the\n", + " # y coord are axes\n", + " transform_for_text = matplotlib.transforms.blended_transform_factory(ax.transData, ax.transAxes)\n", "\n", - "for prediction_name in ['ConfirmedCases', 'Fatalities']:\n", - " model = xgb.XGBRegressor(n_estimators=iterations)\n", + " plt.axvline(x=last_train_date, color='#000000')\n", + " plt.text(first_eval_date, 0.95, 'eval', transform = transform_for_text)\n", + " plt.axvline(x=last_eval_date, color='#000000')\n", + " plt.text(first_test_date, 0.95, 'test', transform = transform_for_text)\n", "\n", - " model = model.fit(\n", - " X=train_df,\n", - " y=train_labels[prediction_name],\n", - " eval_set=[(test_df, test_labels[prediction_name])],\n", - " verbose=False,\n", - " callbacks=[xgb.callback.print_evaluation(period=100)]\n", - " )\n", "\n", - " xgboost_models[prediction_name] = model\n", - " print ('XgBoost: prediction of %s: RMSLE on test = %s' % (prediction_name, model.evals_result()['validation_0']['rmse'][-1])) " - ], - "execution_count": 20, - "outputs": [ - { - "output_type": "stream", - "text": [ - "xgboost version 1.0.2\n", - "[0]\tvalidation_0-rmse:3.40985\n", - "[100]\tvalidation_0-rmse:1.93904\n", - "[200]\tvalidation_0-rmse:1.84852\n", - "[300]\tvalidation_0-rmse:1.80822\n", - "[400]\tvalidation_0-rmse:1.79581\n", - "[500]\tvalidation_0-rmse:1.78997\n", - "[600]\tvalidation_0-rmse:1.78779\n", - "[700]\tvalidation_0-rmse:1.78733\n", - "[800]\tvalidation_0-rmse:1.78607\n", - "[900]\tvalidation_0-rmse:1.78558\n", - "[999]\tvalidation_0-rmse:1.78517\n", - "XgBoost: prediction of ConfirmedCases: RMSLE on test = 1.785172\n", - "[0]\tvalidation_0-rmse:1.20855\n", - "[100]\tvalidation_0-rmse:0.82518\n", - "[200]\tvalidation_0-rmse:0.80351\n", - "[300]\tvalidation_0-rmse:0.80003\n", - "[400]\tvalidation_0-rmse:0.79914\n", - "[500]\tvalidation_0-rmse:0.79899\n", - "[600]\tvalidation_0-rmse:0.79899\n", - "[700]\tvalidation_0-rmse:0.79899\n", - "[800]\tvalidation_0-rmse:0.79899\n", - "[900]\tvalidation_0-rmse:0.79899\n", - "[999]\tvalidation_0-rmse:0.79899\n", - "XgBoost: prediction of Fatalities: RMSLE on test = 0.798991\n" - ], - "name": "stdout" - } - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "WrSAhD8u_qgj", - "colab_type": "text" - }, - "source": [ - "Let's look at XgBoost models' feature importances.\n", + " plt.legend()\n", + " plt.show()\n", "\n", - "1. \"split” means result contains numbers of times the feature is used in a model.\n", - "2. “gain” means result contains total gains of splits which use the feature.\n", - "3. ”cover” means result is the average coverage of splits which use the feature where coverage is defined as the number of samples affected by the split\n", - "\n" - ] - }, - { - "cell_type": "code", - "metadata": { - "id": "zNBo2BNq2Lc2", - "colab_type": "code", - "outputId": "24f2c1e6-9dd4-4580-ad0d-c7c318068db9", - "colab": { - "base_uri": "https://localhost:8080/", - "height": 1000 - } - }, - "source": [ - "for prediction_name in ['ConfirmedCases', 'Fatalities']:\n", - " for importance_type in ['weight', 'gain', 'cover']:\n", - " xgb.plot_importance(\n", - " xgboost_models[prediction_name],\n", - " title='XgBoost: prediction of %s. Feature importance. Type=%s' % (prediction_name, importance_type),\n", - " height=0.9,\n", - " max_num_features=20\n", - " )\n" + "plot_graph('US', 'Kansas', 'ConfirmedCases')\n", + "plot_graph('US', 'Kansas', 'Fatalities')\n", + "plot_graph('China', 'Hubei', 'ConfirmedCases')\n", + "plot_graph('China', 'Hubei', 'Fatalities')\n", + "plot_graph('Germany', '', 'ConfirmedCases', log_scale=False)\n", + "plot_graph('Germany', '', 'Fatalities', log_scale=False)" ], - "execution_count": 21, + "execution_count": 57, "outputs": [ { "output_type": "display_data", "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjsAAAEWCAYAAACADFYuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBEEQBEHQoQlnJwiCIAiCDs3/DyLrTa0vqfL/AAAAAElFTkSuQmCC\n", + "image/png": "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\n", "text/plain": [ - "
" + "
" ] }, "metadata": { - "tags": [] + "tags": [], + "needs_background": "light" } }, { "output_type": "display_data", "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjQAAAEWCAYAAABxBw2jAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCNo94dAEQRAEQdDu+f8B+15NYcbHVoMAAAAASUVORK5CYII=\n", + "image/png": "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\n", "text/plain": [ - "
" + "
" ] }, "metadata": { - "tags": [] + "tags": [], + "needs_background": "light" } }, { "output_type": "display_data", "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAEWCAYAAACaMLagAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6UAAAKGCAYAAACybj1qAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdeXhU5d3/8c89k0lCFhJCWATEsMq+gyCLiEUQoUrrxmLdKYKt6KN1b/X56WOpFhdQKVWLVlCsaxVpFQVREJQ1soOAksgaIGTf5v79cSZhCCEESHKyvF/Xda4zZ53vnDnAfLjPuY+x1goAAAAAADd43C4AAAAAAFB7EUoBAAAAAK4hlAIAAAAAXEMoBQAAAAC4hlAKAAAAAHANoRQAAAAA4BpCKQDUcMaYR40x+4wx1hhzY2B6vdt1BTPGrDfGPOp2HWfLGOMxxvzNGJMSON6DXahhcOC9489mneogcD6nl8N+FhtjZpRHTQCA00coBQAXGGMaGWOeM8b8YIzJMcYkG2MWGGNGlPP7dJL0J0kTJZ0jaZ6kpyVdVJ7vUxGMMd2MMfOMMXsDx2i7MWa2Maaz27WVYoSkmySNknO8l5X3G5TTcVkWqC+lvOs7lZMFQGPMVcYYtx6e/itJD7j03gBQ6xFKAaCSGWMSJK2WNEzOD+Eukn4hab6kmeX8dq0D4w+stXuttVnW2nRr7UnDiDEmtJxrOG3GmJGSVkiKknS9pHaSrpO0R9KfXSztVFpL2mOtXRY43rlnshNjjO8k88vluFhrcwP1uRUCqxRr7SFrbZrbdQBAbUUoBYDK92Jg3Mta+7a1dou1dpO1doacgCpJMsY0N8a8b4xJCwzvGWOaBS1/NHDZ63WBFtc0Y8wHhZdkBi6HfT+wur+wFar45buBVraPjTH3GWOSJCUZYxICl3deZ4z50hiTZYxZY4zpYozpZIxZZozJMMZ8bYxpEfzhjDGjjDGrjDHZxpidxpgngoOuMaahMebDwD5/NMbcXGz7CEn/kPRfa+3l1trPrLU7rbUrrbUPSBoXWM9rjHkl8B5Zxphtxpg/GGM8QfvqbIz53Bhz1BiTboxZZ4y5OGh5B2PM/MCx22+MedMY07is2xere7akZyQ1Dxy7XYH5YcaYZ41zCXW2MWa5MWZA0HaFl9KOMMZ8a4zJlfMfFsX3X6bjEqSrMWaFMSbTGLPSGNOjhPcsPFduDHy+SwLnVIYxZlHwd2uMaRX43vYGlq8OhOQKYUq4NLd43UHzRxljtgaO7yJjTMsSlpd2TnL5LgC4iFAKAJXIGBMnabikF6y1J9wLZ609EljPI+lDSY0kXRwYmkj6wBhjgjZJkHStpNGSLpXUXdITgWVPS7ot8PqcwHAyF8kJxMMlXRI0/zFJUwP7PSLpTUnTJT0kqY+kcEnPB32+YZLmSJohqaOkmyVdJen/gvY5W06L4i8kXSnpN4HPUWiYpHidpOWv8BjJ+TcsWdI1ktoHanpQzuWzhebKaUXsI6mbpEclZQdqPUfSEknrA8t/IacF8sOgYHvS7Utwp6T/lZQk51j3Dsz/i5zv6GY5x/F7Sf8JvH+wqZIeltP6uaKE/Zf1uBR6UtL9knrIuUx3TrFzp7gwOS33N0vqJylWx7fcR0laIGmopK6S3pX0njGmXSn7rAxhci5Rv0lO3V45dRmpzOckAMBN1loGBgYGhkoa5IQbK2n0KdYbKqlAUkLQvJaS/JJ+EZh+VE5Aigla5yFJ24Omr3L+qj9u349KWh80PVvSAUlhQfMSAnX+NmjeyMC8XwXNu1FSetD0EkmPFHu/KyWlSzKS2gb20T9o+XmBz/poYPoPgXXqncHx/bOkhUHTRyXdcJJ1/1fS58Xm1Qu8d59TbX+Sfd4jaVfQdKSkXEm/CZrnlfSDpMcD04MD7/nrU+y7TMclaH/Dgub1D8xrVmyd+KDv0Uo6P2ibcZJyJJlS3mu5pIdP8ztaHDgm6cWGrOBztfi5dYq6SzqfCv+clHpOBtU043TPNwYGBgaG8hloKQWAylVaS1Ww9pJ+ttbuKpxhrd0h6WdJHYLW+9Famxo0/bOkhmdQ13prbU4J8xODXu8LjL8vNi8ycGmpJPWU9FDgUtD0wOWXc+WEs8ZyPpdf0reFO7DW/hiou1BZj5GMMRMDl6YeCLzXXZKaB60yTdLLxpgvjDEPFWvV6ylpULFadweWtSrD9mXRSpJP0tLCGdbaAknf6PjvUZJWnmJfZT4uAcHfXeHxLe3cyLHWbim2TaicoC5jTKQx5i/GmI3GmMOB49VLxx/vsponp+U5eLj3DPYjnfx8Kjy+pzonAQAuI5QCQOXaJqdlp/1Z7CO4c5q8Epadyd/tGSeZH7x/W8o8T9D4MR0fNrpIaiOnNbb4diXZGhiXeoyMMddKelZOS++wwHu9KCdIOW9i7aNywskHki6UlBh0D6tHTudSxcNRG0kfl2H7s1X8GJzsOyhUpuMSpLTvqST5xaaLb/O0pKslPSLncu9ucsLgmXSMlWqt3R48SNpbbB2/TgziJXYApdLPp7KekwAAlxBKAaASWWsPSfqvpDuMMVHFlxtjYgMvN0lqYpyeeguXtZRzX+nGiq/0jK2W1K544AgM+ZI2y/m3p0/hBsaY5nI+V6FPJR2Ucz/kCYKO0QBJK6y1M6y1qwPBplXx9a2126y1z1trL5f0iqRbg2rtKKe1uXitaWXYvix+kHOpav+g+r1y7n083e+xrMelogyQ9Lq19l1rbaKce2dPON7l6ICkCGNM3aB53UpY72Tn06bArFOdkwAAlxFKAaDyTZbTArTSGHO1MeZ8Y0w7Y8ztOnbJ5cLA6znGmF7GmF5yOmtZLekLV6oum/+VNNYY87/G6aW3nXGeP/kXSQpcHvofSX8zxvQzxnST09KZVbgDa22GnOA33Dg94w41Tm/APYwx/0/OcZCclsMexpjLjDFtjDGFLXiSJGNMHWPMC4EeWxOMMRfICVaFYfAFSTGS5hljLjDGtDTG/MIYM8sYE12G7U8p8FlekjTVOL3rtg9MN9KxXphPZ19lOS4VZauk0YH36yzpDTkdXVWUFXJaj580xrQ2xvxa0qQS1suX9GzQ+fSapA1y/gxJpzgnAQDuI5QCQCUL3BvaQ9JncnpcTZQTNH8paUJgHSvpCjmtRYsCw15JVwaWVUnW2v9KulxOb8HfBob7Jf0UtNqNknbK+cwfybm/b1ex/XwopzUxU0742SLpX5LOldPhjyT9TdLbge2/k9M501+DdlMg537I2YHt35dzL+fdgff4WU4Lpl9OUN4gJ6jmBIZStz8N98m5h/IfktYq0MuxtXbPae6nrMelotwtab+kr+T0wrs88LqIcR43VC7nZ+CqgnFyOv36Xs6fjUdKWDVHTo/Tr8sJsh45nXHZwH7Kck4CAFxkqvBvGwAAUI0YY16T1Nhae8JzVgEAOJkQtwsAAADVX+C5oEN0/HNuAQA4JVpKAQAAAACu4Z5SAAAAAIBrCKUAAAAAANcQSgEAAAAAriGUAgAAAABcQygFAAAAALiGUAoAAAAAcA2hFAAAAADgGkIpAAAAAMA1hFIAAAAAgGsIpQAAAAAA1xBKAQAAAACuIZQCAAAAAFxDKAUAAAAAuIZQCgAAAABwDaEUAAAAAOAaQikAAAAAwDWEUgAAAACAawilAAAAAADXEEoBAAAAAK4hlAIAAAAAXEMoBQAAAAC4hlAKAAAAAHANoRQAAAAA4BpCKQAAAADANYRSAAAAAIBrCKUAAAAAANcQSgEAAAAAriGUAgAAAABcQygFAAAAALiGUAoAAAAAcA2hFAAAAADgGkIpAAAAAMA1hFIAAAAAgGsIpQAAAAAA1xBKAQAAAACuIZQCAAAAAFxDKAUAAAAAuIZQCgAAAABwDaEUAAAAAOAaQikAAAAAwDWEUgAAAACAawilAAAAAADXEEoBAAAAAK4hlAIAAAAAXEMoBQAAAAC4hlAKAAAAAHANoRQAAAAA4BpCKQAAAADANYRSAAAAAIBrCKUAAAAAANcQSgEAAAAAriGUAgAAAABcQygFAAAAALiGUAoAAAAAcA2hFAAAAADgGkIpAAAAAMA1hFIAAAAAgGtC3C5AkuLj421CQoLbZQAAUKNs2bJFknT++ee7XAnOFt8lgOpu1apVB621DUpaViVCaUJCglauXOl2GQAA1CiDBw+WJC1evNjVOnD2+C4BVHfGmB9PtozLdwEAAAAAriGUAgAAAABcQygFAAAAALimStxTWpK8vDwlJSUpOzvb7VIAhYeHq1mzZvL5fG6XAgAAANQoroZSY8woSaNat259wrKkpCRFR0crISFBxpjKLw4IsNYqJSVFSUlJatGihdvlAAAAADWKq5fvWms/stZOiImJOWFZdna26tevTyCF64wxql+/Pq32AAAAQAWo0veUEkhRVXAuAgAAABWjSodSAAAAAEDNRig9hb179+q6665Tq1at1LNnT40YMUJbt2497f189dVX6tixo7p166bk5GRdddVVFVDtiaKioopeb926VSNGjFCbNm3Uo0cPXXPNNdq3b1+l1AEAAAAAJamyve9WBdZajR49WjfccIPeeustSdK6deu0b98+tW3b9rT2NWfOHD3wwAMaP368JOmdd945YZ38/HyFhFTMV5Kdna3LL79c06ZN06hRoyRJixcv1oEDB9SoUaMKeU8AAAAAOBVaSkuxaNEi+Xw+TZw4sWhe165dNWDAAN17773q1KmTOnfurHnz5klyQt7gwYN11VVXqV27dho3bpystXr55Zf19ttv65FHHtG4ceO0a9cuderUSZI0e/Zs/fKXv9SQIUN0ySWXaPbs2bryyis1dOhQJSQkaMaMGZo2bZq6d++uvn376tChQ5KkH374QcOHD1fPnj01cOBAbd68WZK0c+dO9evXT507d9bDDz9cVPfcuXPVr1+/okAqSYMHD1anTp20a9cuDRw4UD169FCPHj20bNkySdKePXs0aNAgdevWTZ06ddJXX30lSfr000/Vr18/9ejRQ1dffbXS09MlSffff786dOigLl266J577qmorwUAAABADVItWkof+2iDNv58tFz32aFJXf1pVMdS11m/fr169ux5wvz33ntPa9eu1bp163Tw4EH17t1bgwYNkiStWbNGGzZsUJMmTdS/f38tXbpUt956q77++muNHDlSV111lXbt2nXc/lavXq3ExETFxcVp9uzZWr9+vdasWaPs7Gy1bt1aU6dO1Zo1a3TXXXfp9ddf15QpUzRhwgTNnDlTbdq00YoVKzRp0iR98cUXuvPOO3X77bfrN7/5jV544YVTfhZJatiwoT777DOFh4dr27ZtGjNmjFauXKm5c+dq2LBheuihh1RQUKDMzEwdPHhQjz/+uBYuXKjIyEhNnTpV06ZN0+TJk/X+++9r8+bNMsboyJEjp/mNAAAAAKiNqkUorWq+/vprjRkzRl6vV40aNdJFF12k7777TnXr1lWfPn3UrFkzSVK3bt20a9cuDRgwoNT9DR06VHFxcUXTF198saKjoxUdHa2YmJii1s3OnTsrMTFR6enpWrZsma6++uqibXJyciRJS5cu1bvvvitJuv7663Xfffed8vPk5eXpjjvu0Nq1a+X1eovume3du7duvvlm5eXl6corr1S3bt305ZdfauPGjerfv78kKTc3V/369VNMTIzCw8N1yy23aOTIkRo5cmRZDycAAACAWqxahNJTtWhWlI4dO5Z472dpwsLCil57vV7l5+efcpvIyMiT7sPj8RRNezwe5efny+/3KzY2VmvXri1xfyU9vqRjx4768ssvS1z/mWeeUaNGjbRu3Tr5/X6Fh4dLkgYNGqQlS5Zo/vz5uvHGG3X33XerXr16Gjp0qN58880T9vPtt9/q888/1zvvvKMZM2boiy++OOVnBwAAAFC7cU9pKYYMGaKcnBzNmjWraF5iYqJiY2M1b948FRQU6MCBA1qyZIn69OlTaXXVrVtXLVq00L/+9S9JTodM69atkyT179+/qFOmOXPmFG0zduxYLVu2TPPnzy+at2TJEq1fv16pqak655xz5PF49M9//lMFBQWSpB9//FGNGjXSbbfdpltvvVWrV69W3759tXTpUm3fvl2SlJGRoa1btyo9PV2pqakaMWKEnnnmmaJ6AAAAAKA0roZSY8woY8ys1NRUN8s4KWOM3n//fS1cuFCtWrVSx44d9cADD2js2LHq0qWLunbtqiFDhugvf/mLGjduXKm1zZkzR6+88oq6du2qjh076sMPP5QkPffcc3rhhRfUuXNnJScnF61fp04dffzxx5o+fbratGmjDh066MUXX1SDBg00adIkvfbaa+ratas2b95c1HK7ePFide3aVd27d9e8efN05513qkGDBpo9e7bGjBmjLl26qF+/ftq8ebPS0tI0cuRIdenSRQMGDNC0adMq9XgAAAAAqJ6MtdbtGtSrVy+7cuXK4+Zt2rRJ7du3d6ki4ESckwCqm8GDB0ty/pMR1RvfJYDqzhizylrbq6RlXL4LAAAAAHANoRQAAJSLhIQEHTx40O0yarUjR47oxRdfPKNtn332WWVmZpZzRQBwaoRSAACAGoJQCqA6qhaPhAEAABXrjTfe0PPPP6/c3FxdcMEF6tKli3bt2qWnnnpKkjR79mytXLlSM2bM0JVXXqndu3crOztbd955pyZMmOBy9Sh0//3364cfflC3bt00dOhQNWzYUG+//bZycnI0evRoPfbYY8rIyNA111yjpKQkFRQU6JFHHtG+ffv0888/6+KLL1Z8fLwWLVrk9kcBUIvQUgoAQC23adMmzZs3T0uXLtXatWvl9XoVFRWl999/v2idefPm6brrrpMkvfrqq1q1apVWrlyp559/XikpKW6VjmL+/Oc/q1WrVlq7dq2GDh2qbdu26dtvv9XatWu1atUqLVmyRP/5z3/UpEkTrVu3TuvXr9fw4cP1+9//Xk2aNNGiRYsIpAAqHS2lAADUcp9//rlWrVql3r17S5KysrLUsGFDtWzZUsuXL1ebNm20efNm9e/fX5L0/PPPFwXW3bt3a9u2bapfv75r9aNkn376qT799FN1795dkpSenq5t27Zp4MCB+p//+R/dd999GjlypAYOHOhypQBUkCflpBUbjgaN06Xc9MA47cTp3EzpjpWSp3q2ORJKS+H1etW5c2fl5+erffv2eu211xQREXFG+7rxxhs1cuRIXXXVVbr11lt19913q0OHDiWuu3jxYoWGhurCCy88rfdISEjQypUrFR8fr71792rKlCn67rvvFBsbq0aNGunZZ59V27ZtT7v2r776ShMnTpTP59P8+fN155136p133jnt/ZyuqKgopaenS5K2bt2qKVOmaNu2bYqOjlbr1q01ffp0NWrUqMLrAICazlqrG264QU8++eRx81999VW9/fbbateunUaPHi1jjBYvXqyFCxfqm2++UUREhAYPHqzs7GyXKkdprLV64IEH9Nvf/vaEZatXr9Ynn3yihx9+WJdccon++Mc/ulAhUENYK+VmSNlHpKzDxw/ZR4+Fy+DXRePA/Pyssr1XaJQzhEVJoZFSaLRUt6kzz58veUIr9rNWEEJpKerUqaO1a9dKksaNG6eZM2fq7rvvLlqen5+vkJDTP4Qvv/xyqcsXL16sqKio0w6lhay1Gj16tG644Qa99dZbkqR169Zp3759ZxRK58yZowceeEDjx4+XpBID6Zkei7LIzs7W5ZdfrmnTpmnUqFGSnGN04MABQikAlINLLrlEV1xxhe666y41bNhQhw4dUlpamkaPHq0nnnhCa9as0dSpUyVJqampqlevniIiIrR582YtX77c5eoRLDo6WmlpaZKkYcOG6ZFHHtG4ceMUFRWl5ORk+Xw+5efnKy4uTuPHj1dsbGzR75LCbePj4938CIB7CsNl1qFjoTLz0IlBs6R5/ryT79d4pLBoKazusXFEvBTX0gmT4XUDywqXRx9bL7zusRDqi6y2LaGnQigto4EDByoxMVGLFy/WI488onr16mnz5s3atGmT7r//fi1evFg5OTmaPHmyfvvb38paq9/97nf67LPPdO655yo09Nj/WgwePFhPP/20evXqpf/85z968MEHVVBQoPj4eL3yyiuaOXOmvF6v3njjDU2fPl3t2rXTxIkT9dNPP0lyesfr37+/UlJSNGbMGCUnJ6tfv36y1kqSFi1aJJ/Pp4kTJxa9Z9euXSU5gfUPf/iDFixYIGOMHn74YV177bVavHixHn30UcXHx2v9+vXq2bOn3njjDb3yyit6++239d///lcLFizQE088oZEjR2r9+vWaPXu23nvvPaWnp6ugoEA33XSTPvjgA2VkZGjbtm265557lJubq3/+858KCwvTJ598ori4OP3www+aPHmyDhw4oIiICP39739Xu3bttHPnTo0dO1bp6em64oorimqfO3eu+vXrVxRIC4+hJO3atUvXX3+9MjIyJEkzZszQhRdeqD179ujaa6/V0aNHlZ+fr5deekkDBw7Up59+qj/96U/KyclRq1at9I9//ENRUVG6//779e9//1shISG69NJL9fTTT1fMiQQAVVCHDh30+OOP69JLL5Xf75fP59MLL7yg8847T+3bt9fGjRvVp08fSdLw4cM1c+ZMtW/fXueff7769u3rcvUIVr9+ffXv31+dOnXSZZddprFjx6pfv36SnCuQ3njjDW3fvl333nuvPB6PfD6fXnrpJUnShAkTNHz48KJ7S4Fqze+XclKljBQpM0XKPOiMMwLjrCMlh8/SwqUvUqpTzxki6kkNzpci4o7NKxzCYwOvY53XoZGSMZX32auh6hFKF9wv7f2+fPfZuLN02Z/LtGp+fr4WLFig4cOHS3IueVm/fr1atGihWbNmKSYmRt99951ycnLUv39/XXrppVqzZo22bNmijRs3at++ferQoYNuvvnm4/Z74MAB3XbbbVqyZIlatGihQ4cOKS4uThMnTlRUVJTuueceSdLYsWN11113acCAAfrpp580bNgwbdq0SY899pgGDBigP/7xj5o/f75eeeUVSSoKlSV57733tHbtWq1bt04HDx5U7969NWjQIEnSmjVrtGHDBjVp0kT9+/fX0qVLdeutt+rrr78uuvR4165dx+1v9erVSkxMVFxcnGbPnq3169drzZo1ys7OVuvWrTV16lStWbNGd911l15//XVNmTJFEyZM0MyZM9WmTRutWLFCkyZN0hdffKE777xTt99+u37zm9/ohRdeKHqP0j5Pw4YN9dlnnyk8PFzbtm3TmDFjtHLlSs2dO1fDhg3TQw89pIKCAmVmZurgwYN6/PHHtXDhQkVGRmrq1KmaNm2aJk+erPfff1+bN2+WMUZHjhwp03kBADXJtddeq2uvvfaE+R9//PFx02FhYVqwYEGJ+yj+bwTcMXfu3OOm77zzzuOmW7VqpWHDhp2w3e9+9zv97ne/q9DagNOWl+UEyOwjx8bZqSfOCx5nBoKoLSh5nyF1jg+T8W0DQbN4wIw7Nj88VvKFV+5nr0WqRyh1SVZWlrp16ybJaSm95ZZbtGzZMvXp00ctWrSQ5HQikJiYWHRJa2pqqrZt26YlS5ZozJgx8nq9atKkiYYMGXLC/pcvX65BgwYV7SsuLq7EOhYuXKiNGzcWTR89elTp6elasmSJ3nvvPUnS5Zdfrnr16p3yM3399ddFdTVq1EgXXXSRvvvuO9WtW1d9+vRRs2bNJEndunXTrl27NGDAgFL3N3To0OPqvvjiixUdHa3o6GjFxMQUtW527txZiYmJSk9P17Jly3T11VcXbZOTkyNJWrp0qd59911J0vXXX6/77rvvlJ8nLy9Pd9xxR1FvkVu3bpUk9e7dWzfffLPy8vJ05ZVXqlu3bvryyy+1cePGoo46cnNz1a9fP8XExCg8PFy33HKLRo4cqZEjR57yfQEAAFAG+TnH7qHMTj35cLLAWZBT+v5Do4+1SNaJleq3ks7tI0XGO5fIRtSXIus748Lp0DPrIwYVx9VQaowZJWlU69atS1+xjC2a5S34ntJgkZGRRa+ttZo+ffoJ/+P4ySeflFsdfr9fy5cvV3h42f53pmPHjmfUEVFYWFjRa6/Xq/z8/FNuE3wsiu/D4/EUTXs8HuXn58vv9ys2NrbE4ypJpoRLGzp27Kgvv/yyxPWfeeYZNWrUSOvWrZPf7y86RoMGDdKSJUs0f/583Xjjjbr77rtVr149DR06VG+++eYJ+/n222/1+eef65133tGMGTP0xRdfnPKzAwAA1Ej+gqCeXQPj4I55Cofc4r3Fpp24XkFu6e9lPFJ4jHP/ZGG4bNjuWMgMjwl6HTyu52zjpY2tJnD1W7TWfiTpo169et3mZh1nY9iwYXrppZc0ZMgQ+Xw+bd26VU2bNtWgQYP0t7/9TTfccIP279+vRYsWaezYscdt27dvX02aNEk7d+487vLd6OhoHT16tGi9Sy+9VNOnT9e9994rSVq7dq26deumQYMGae7cuXr44Ye1YMECHT58WJI0ZMgQPfjgg5o1a1bRA80TExOVmpqqgQMHFtV16NAhLVmyRE899ZQ2b95cKcerbt26atGihf71r3/p6quvlrVWiYmJ6tq1q/r376+33npL48eP15w5c4q2GTt2rJ588knNnz9fl19+uSRpyZIliouLU2pqqpo1ayaPx6PXXntNBQXOZRo//vijmjVrpttuu005OTlavXq1HnroIU2ePFnbt29X69atlZGRoeTkZDVp0kSZmZkaMWKE+vfvr5YtW1bKsQAAACgzv9+5HNWf74TGglxnyM9xhoIcKT83MM5xluVmSHmZzjg33XlsSNHrwLKigBkUQvMyy1ZTSLjTIU9hZz2h0VLdZsU664k61oFPeGHIDAx1Yp1tud+y1uO/Fs7Srbfeql27dqlHjx6y1qpBgwb64IMPNHr0aH3xxRfq0KGDmjdvXtTJQLAGDRpo1qxZ+tWvfiW/3190f+SoUaN01VVX6cMPP9T06dP1/PPPa/LkyerSpYvy8/M1aNAgzZw5U3/60580ZswYdezYURdeeKGaN28uyWltfP/99zVlyhRNnTpV4RhufAEAACAASURBVOHhSkhI0LPPPqsBAwbom2++UdeuXWWM0V/+8hc1bty40kKp5PTme/vtt+vxxx9XXl6errvuOnXt2lXPPfecxo4dq6lTpx7X0VGdOnX08ccfa8qUKZoyZYp8Pp+6dOmi5557TpMmTdKvf/1rvf766xo+fHhRy+3ixYv11FNPyefzKSoqSq+//roaNGig2bNna8yYMUWXDD/++OOKjo7WFVdcoezsbFlrNW3atEo7FgCqNmut8gqs8v1+Z1zgV4HfqsBa5RdY+a1Vgb9wrKLXziD5rZW1zn78xcZWx5b7A9O2aPrYOk4dkgJT1uq4+TZo/rG1nO0Ppjt/1324NrloubNO4cp+nbv7Ix23Ax1bsfBnog3euNg6x2b5T7nK8TOPfRBTfH5pghYf267w89gy78+U5TMVm3fiNiff/+nxn3KN9P27FB7ikb4K/jeq2HuX6Xsq9qKE7/2EeaVNn9brkrYtPj6dZTrF+mUd6/hp6w+cz4WvbdB00PIS1y2cHzxd0jr+ktcpHPx+J3wGh9DyONeMJ/BIkUjJF+GMw6Kl6HMCjxiJOhYyg6dPGAI9woZUz8ePoOoxJ/5DU/l69eplV65cedy8TZs2qX379i5VBJyIcxIoXwV+q8OZucrMKVBOfoFy8v3KyS9Qdp4zzsnzF83LLbAqKPAr32+LwmG+3wmL+QXOvAK/X3n+wLICqzx/YF7Q+nkFx6ZzA+PCeXmB1862zjjf7/6/kWdj79z7JUmNx5Z8G4xXBfoh/PrKLAlnaPBsp5f5xTdGnmLNihL4L4qiFq3g6bN8bQL7K3GZOfWy0vZ1JmPjOTZW0LSM5PEWW89z/GudZH7xdTzeEtYpNnhCAut5j70umvZK3lBnCAmTvGHOOCTs+HmhEYHwGQiiIWG0SsI1xphV1tpeJS2jpRQAUC6stcrMLVBqVp6OZObpSFauUtJzdSAtRwfSc5xx4ZCeo5T0HJ1t5vN6jEI8Rj6vRyFe53WIx3nt83pOWO7zehTu88gXHqIQj0ehIc76Pq9HPu+x9UKL9udRaIjH2W9gHa/HyGuMPIGx1+O8DvEYeYyRxzh1eYzzA7lwnlFgbIxMYL4zlhS0rHDdwt+NReNi80zgx7gp/B2u4N+azotxX8XKSJr7PxcFlh/7MWqcL01JR785YTt73I9Wc8K2JU0Hrxu0UgmrmONeH9vPie95UqaEiWL7KekznPD6hPpOUe9J1zlZoWVe4ZRBIfezXyrEa6SHPjvFdmX4DkoNmGWrBwDKE6EUACBJyi/w62h2vo5k5upodr7Ss/OVnuMMGYFxWnbw68Lw6YxTs3KVV1ByyvR5jRpEhalBdJiaxIar67kxahAVpvjoMEWGhijM51FYiFdhIR6FhXgU7vMWzQsNccJgUdj0HAugJQcjFKrj80qSWjaIOvlKDTpUUjU4G6G+wE82HkkBoAaq0qHUWssPDlQJVeEyd6AsrLVKy8lXamaeUrOOH44cNy/3WItmZp6OZuUpLefUPW77vEZRYSGKDAtRdLhP9SJ8atsoSjF1QhUb4VNsHZ9iI3yKqROqmDo+xUeFqkF0mGLq+Pj7HAAAlKjKhtLw8HClpKSofv36/JCBq6y1SklJKfMjeYCzYa1VTr5f2XnH7q3MziucLtDhzDwdyshVSnqOUjJynSE9R4cycnUwPVeHM3NVUMo1saFej+oWBUefGtcN1/mNohUT4VNsnVDF1AlRbESo6tYJUVSYT5FhXkUHxlHhIQoL8Vbi0QAAALVBlQ2lzZo1U1JSkg4cOOB2KYDCw8PVrFkzt8tANZKSnqNvdx7Sip2HtHxHipIPZ5W6vpWUV+B07FNW0eEhio8KU1xkqJrHRah783qqH+m0WNat44TO2Do+xQQCaEwdn+r4vPxHHwAAqFKqbCj1+Xxq0aKF22UAQJnsP5qtFTsPacXOFK3YcUjb9qdLcu7p63lePfVtWd/p+KYUPq9RmM+r8MC9lOE+j8JDvAoPmhcb4VP9qFDFRYbSagkAAGqEKhtKAaAyFfht4B7LXB0OdNpzOMO5/zIrcOlsVm6BsvIKTpjem5qtXSnOg8YjQ73qlRCn0T2a6oIW9dW5aYxCQzwufzoAAICqi1AKoMax1mrLvjQlH87S0ew8Hc1yeoo9mp2vo1l5Opqdp7TA6yNZeTqc4fQ2WxqPcVo964Q6LZd1fMfGbRtFa+wFzXVBi/rq2KSuQryEUAAAgLIilAKoEXLyC7RixyEt3LRPCzfu08+p2SesE+7zqG64c79l3XCnQ5+E+EjVi3B6iq0X4VNsRKAX2YhQ1YvwqW64TxFhXoV6PdyLCQAAUAEIpQCqrSOZuVq0Zb8WbtyvL7ceUHpOvur4vBrYJl5TftFW5zeOLgqg0eE+LqMFAACoggilAKqVPalZ+uT7vfp0w16t/PGwCvxWDaPDNKprEw3t0FAXtopXuI8OgAAAAKoLV0OpMWaUpFGtW7d2swwAVdze1Gx98v0ezf9+j1b9eFiS1K5xtCYNbqVftG+kzk1j5PFwaS0AAEB15GootdZ+JOmjXr163eZmHQCqnr2p2Vqwfo/mJ+7RyqAges+lbTWi8zlq2SDK5QoBAABQHrh8F0CVcSQzVx8l7tG/1yZr5Y+HZa0TRP9naFuN6HKOWhFEAQAAahxCKQBX5RX49eWWA3p3dZI+37RfuQV+tWkYpSmXtNXlXRqrdcNot0sEAABABSKUAnDFhp9T9e6qZP17XbIOpueqfmSoxvVtrl/3aKaOTery+BUAAIBaglAKoNLsT8vWh2t+1rurk7R5b5pCvR5d0r6hftWjmQaf30A+L49sAQAAqG0IpQAqVFZugT7duFfvrk7W19sOyG+lbufG6v9d2Umjupyj2IhQt0sEAACAiwilAMqd32+1fEeK3luTrAXf71FGboGaxtbRpMGtNbpHUzosAgAAQBFCKYBys31/mt5dnawP1iRrT2q2osNCNLJLE43u0VR9EuJ4ligAAABOQCgFUC7+u2GvJr6xSh5jNKhNvB4c0V5DOzRSuM/rdmkAAACowgilAM7aTymZuudf69SlaYxevqG3GkSHuV0SAAAAqgm6ugRwVnLyCzR57moZSTPG9iCQAgAA4LTQUgrgrDwxf5O+T07VrOt76ty4CLfLAQAAQDVDSymAM/Zx4s96/ZsfdeuAFrq0Y2O3ywEAAEA1RCgFcEZ2HszQ/e9+r+7NY3XfZe3cLgcAAADVFKEUwGnLzivQ5DmrFeI1mjG2h3xe/ioBAADAmeGeUgCn7bGPNmrjnqP6x4291TS2jtvlAAAAoBqjeQPAaflwbbLe/PYnTbyolS5u19DtcgAAAFDNuRpKjTGjjDGzUlNT3SwDQBlt35+uB977Xr0T6umeS9u6XQ4AAABqAFdDqbX2I2vthJiYGDfLAFAGWbnOfaThPq+mj+mhEO4jBQAAQDngnlIAZfLovzdo6/40vXZTHzWOCXe7HAAAANQQNHUAOKU9qVmat3K3bh3QQoPaNnC7HAAAANQghFIAp7R8R4ok6YpuTV2uBAAAADUNoRTAKa3YcUjR4SFqf05dt0sBAABADUMoBXBKy3ek6IIWcfJ6jNulAAAAoIYhlAIo1d7UbO1KydQFLeq7XQoAAABqIEIpgFKt2OncT9q3JaEUAAAA5Y9QCqBUy3ccUnRYiDo04X5SAAAAlD9CKYBSrdiRot7cTwoAAIAKQigFcFL7j2Zrx8EM9W0Z53YpAAAAqKEIpQBOavnOQ5JEJ0cAAACoMIRSACe1YkeKosJC1JH7SQEAAFBBCKUATmr5jhT1SqinEC9/VQAAAKBi8EsTQIn2p2XrhwMZPAoGAAAAFYpQCqBE3xbdT0onRwAAAKg4hFIAJVq+I0WRoV51ahrjdikAAACowQilAEq0Ysch9UyIk4/7SQEAAFCB+LUJ4AQH03O0bX86zycFAABAhSOUAjhB4f2kdHIEAACAiuZqKDXGjDLGzEpNTXWzDADFLN+RoohQrzpzPykAAAAqmKuh1Fr7kbV2QkwMP3yBqmTFjkPqeV497icFAABAheMXJ4DjHMrI1ZZ9aVy6CwAAgEpBKAVwnG93pkgSnRwBAACgUhBKARxn+Y5DCvd51LlprNulAAAAoBYglAI4zvIdKep1XpxCQ/jrAQAAABWPX50AihzOyNXmvWm6oAWX7gIAAKByEEoBFFlR+HzSVnRyBAAAgMpBKAVQZMXOFIWFeNSlGY9pAgAAQOUglAIosjzwfNKwEK/bpQAAAKCWIJQCkCQdyczV5r1HdUELLt0FAABA5SGUApAkfbvzkKzl+aQAAACoXIRSAJKcTo7CQjzqei7PJwUAAEDlIZQCkOQ8n7R781iF+7ifFAAAAJWHUApAqVl52rjnqPq25H5SAAAAVC5CKQB9F7iflE6OAAAAUNkIpQC0fEeKQkM86t6c+0kBAABQuULcLgBAxdm+P01P/XeL1vx0pNT1jmTlqdu53E8KAACAykcoBWqgA2k5enbhVr313W5F+Lwa1qmxfF5T6jZXdGtaSdUBAAAAxxBKgRokMzdfL3+1U3/78gfl5Pt1fd/z9LshrVU/Kszt0gAAAIASEUqBGqDAb/XOqt3666dbtT8tR8M7NtZ9l7VTi/hIt0sDAAAASkUoBaoxa60Wbz2gP3+yWVv2pal781i9OK6HeiXEuV0aAAAAUCaEUqCaOpqdpwfe+17zE/fovPoRenFcD13WqbGMKf3eUQAAAKAqIZQC1dD65FRNnrtaSYezdO+w83XbwJYKDeEJTwAAAKh+CKVANWKt1evf/Kgn5m9S/ahQzZvQl0t1AQAAUK0RSoFq4mh2nu57J1EL1u/VkHYN9deru6peZKjbZQEAAABnhVAKVAOJSUd0x9w1Sj6SpQdHtNOtA1rK4+HeUQAAAFR/hFKgCrPW6rVlu/TEJ5vUICpMb/+2r3qex+W6AAAAqDlcDaXGmFGSRrVu3drNMoAqKT0nX/e8vU7/2bBXl7RrqKe5XBcAAAA1kKvddVprP7LWToiJiXGzDKDKOZqdp+tfWaHPNu3TQyPa6+UbehFIAQAAUCNx+S5QxaRm5un6V1do056jemFsDw3v1NjtkgAAAIAKQygFqpBDGbka//IKbd+frpnje+qS9o3cLgkAAACoUIRSoIo4mJ6j8S+v0M6DGfr7Db10UdsGbpcEAAAAVDhCKVAF7E/L1ri/r9Duw5l69cbe6t863u2SAAAAgEpBKAVctjc1W2P/vlx7j2Zr9k191LdlfbdLAgAAACoNoRRwUfKRLI39+3KlpOfq9Zv7qFcCzyAFAABA7UIoBVyy+1Cmxvx9uVKz8vT6LX3Uo3k9t0sCAAAAKh2hFHDBDwfSdf3LK5SRW6A5t16gLs1i3S4JAAAAcAWhFKhka346rJtnfyePMZp72wXq2CTG7ZIAAAAA1xBKgUq0aMt+TXpjtRpEh+n1m/soIT7S7ZIAAAAAVxFKgUry7qok3fduos5vHK3ZN/VRg+gwt0sCAAAAXEcoBSqYtVazluzQkws2q3/r+po5vqeiw31ulwUAAABUCYRSoAL5/VZPfLJJr3y9UyO7nKO/XtNVYSFet8sCAAAAqgxCKVBBcvP9uveddfpw7c+68cIE/XFkB3k8xu2yAAAAgCqFUApUgPScfN3+xip9te2g/jD8fN1+USsZQyAFAAAAiiOUAuUsO69A415eofXJqXrqqi66ute5bpcEAAAAVFmEUqCcPfbRBq3bfUQzx/fU8E6N3S4HAAAAqNI8bhcA1CTvr0nSm9/u1qTBrQikAAAAQBkQSoFysn1/mh58b736tIjT3UPbul0OAAAAUC0QSoFykJmbr0lzVisi1KvpY7orxMsfLQAAAKAsuKcUKAePfLBB2/an6583X6BGdcPdLgcAAACoNmjOAc7S2yt3693VSfrdkDYa0Cbe7XIAAACAaoVQCpyFzXuP6pEP1uvCVvV15yVt3C4HAAAAqHYIpcAZSs9x7iOtW8en567rLq/HuF0SAAAAUO0QSoEzYK3Vg+99r10HM/T8dd3VIDrM7ZIAAACAaolQCpyBud/+pH+v+1l3D22rfq3qu10OAAAAUG0RSoHTtD45VY99tFGD2jbQpMGt3S4HAAAAqNYIpcBpSM3M06Q5qxUXEapnrukqD/eRAgAAAGeF55QCZeT3W9399lr9fCRL837bV/WjuI8UAAAAOFu0lAJl9OLi7fp88349fHl79Twvzu1yAAAAgBqBUAqUwVfbDuivn23VFd2a6IYLE9wuBwAAAKgxCKXAKSQfydLv31yjtg2j9eSvOssY7iMFAAAAyguhFChFTn6BJr2xSvkFVi+N76GIUG7DBgAAAMoTv7CBUjz20UatS0rVzPE91bJBlNvlAAAAADUOLaXASfxr5W7NXfGTJl7USsM7NXa7HAAAAKBGIpQCJdjwc6oe/mC9+rWsr3subet2OQAAAECNRSgFiknNzNPEN1apXkSopo/trhAvf0wAAACAiuLqr21jzChjzKzU1FQ3ywCK+P1Wd729VntTs/Xi+B6KjwpzuyQAAACgRnM1lFprP7LWToiJiXGzDKDIrK926IvN+/XHkR3Uo3k9t8sBAAAAajyuSwQC9qdl6/nPt2loh0Ya3/c8t8sBAAAAagVCKRDwzGfblJvv10Mj2ssY43Y5AAAAQK1AKAUkbd2Xpnnf/aTr+52nhPhIt8sBAAAAag1CKSDpyU82KTIsRL8f0sbtUgAAAIBahVCKWu/rbQe1aMsB3XFxa9WLDHW7HAAAAKBWIZSiVvP7rf7vk01qGltHN1yY4HY5AAAAQK1DKEWt9v6aZG3cc1R/GH6+wn1et8sBAAAAah1CKWqtrNwCPf3pFnVtFqNRXZq4XQ4AAABQKxFKUWu9unSn9qRm68ER7eXx8AgYAAAAwA2EUtRKB9Jy9OKi7bq0QyNd0LK+2+UAAAAAtRahFLXSc59vVU6+X/df1s7tUgAAAIBajVCKWmf7/jS9+e1ujbuguVo2iHK7HAAAAKBWI5Si1vnzgs2K8Hn1+0vauF0KAAAAUOsRSlGrLPvhoBZu2q9JF7dW/agwt8sBAAAAaj1CKWoNv9/q/z7ZpKaxdXRT/wS3ywEAAAAgQilqkY+/36P1yUd177DzFe7zul0OAAAAABFKUUtYa/XyVzvUqkGkftm1idvlAAAAAAgglKJWWP3TYSUmpeqm/i3k8Ri3ywEAAAAQQChFrfDq17tUNzxEv+rR1O1SAAAAAAQhlKLGSz6Spf9s2KsxFzRXRGiI2+UAAAAACEIoRY33+je7JEm/6ZfgZhkAAAAASkAoRY2WmZuvt77drWEdG6lpbB23ywEAAABQDKEUNdp7q5OVmpWnm/u3cLsUAAAAACUglKLG8vut/rF0pzo3jVHP8+q5XQ4AAACAEhBKUWN9tf2gfjiQoZsHJMgYHgMDAAAAVEWEUtRY/1i6Uw2iw3R55yZulwIAAADgJAilqJG270/X4i0HdH3f8xQawmkOAAAAVFX8WkeN9NqyXQr1ejT2guZulwIAAACgFIRS1DipmXl6Z1WSrujWRPFRYW6XAwAAAKAUhFLUOPNW/qSsvALdxGNgAAAAgCqPUIoaJb/Ar9eW/ai+LePUoUldt8sBAAAAcAqEUtQon23cp+QjWbSSAgAAANUEoRQ1yqtLd+rcuDr6RftGbpcCAAAAoAwIpagxvk9K1Xe7DuuGfgnyeozb5QAAAAAoA0Ipaox/LN2pyFCvrul9rtulAAAAACgjQilqhP1p2foo8Wdd3etc1Q33uV0OAAAAgDIilKJGeOXrnSrwW914YYLbpQAAAAA4DYRSVHtHMnP1xjc/6vIuTZQQH+l2OQAAAABOA6EU1d7sZbuUkVugSYNbuV0KAAAAgNNEKEW1lp6Tr38s3aVftG+o9ufUdbscAAAAAKeJUIpqbe6KH5WalafJF7d2uxQAAAAAZ4BQimorO69Af/9qp/q3rq/uzeu5XQ4AAACAM0AoRbX1r1VJOpCWQyspAAAAUI0RSlEt5RX4NXPxD+rePFb9WtZ3uxwAAAAAZ4hQimrpw7U/K/lIlu64uLWMMW6XAwAAAOAMEUpR7RT4rV5cvF3tz6mrIe0aul0OAAAAgLNAKEW1898Ne7XjQIYmX9yKVlIAAACgmiOUolqx1uqFRdvVMj5Sl3U6x+1yAAAAAJwlQimqlcVbDmjDz0c1cXAreT20kgIAAADVHaEU1Ya1VjMWbVfT2Doa3b2p2+UAAAAAKAeEUlQbK3Ye0qofD+u3F7WUz8upCwAAANQE/LJHtfHCou2KjwrTNb3OdbsUAAAAAOWEUIpqYd3uI/pq20HdOrCFwn1et8sBAAAAUE4IpagWXli0XTF1fBrf9zy3SwEAAABQjgilqPL2p2Xr04379Jt+5ykqLMTtcgAAAACUI0IpqrzE3amSpIvaNnC5EgAAAADljVCKKi8xOVUeI3VsEuN2KQAAAADKGaEUVV5i0hG1bRStOqF0cAQAAADUNIRSVGnWWn2flKrOTWklBQAAAGoiQimqtOQjWUrJyFWXc2PdLgUAAABABSCUokr7Psnp5KgLLaUAAABAjUQoRZWWmJwqn9eo3TnRbpcCAAAAoAJUSCg1xkQaY1YaY0ZWxP5ReyQmHVG7xnUVFkInRwAAAEBNVKZQaox51Riz3xizvtj84caYLcaY7caY+4MW3Sfp7fIsFLWPtVaJSanq3IxLdwEAAICaqqwtpbMlDQ+eYYzxSnpB0mWSOkgaY4zpYIwZKmmjpP3lWCdqoV0pmUrLzldXQikAAABQY4WUZSVr7RJjTEKx2X0kbbfW7pAkY8xbkq6QFCUpUk5QzTLGfGKt9RffpzFmgqQJktS8efMzrR81WGLSEUlS56b0vAsAAADUVGUKpSfRVNLuoOkkSRdYa++QJGPMjZIOlhRIJclaO0vSLEnq1auXPYs6UEN9n5SqsBCP2jaKcrsUAAAAABXkbEJpqay1sytq36gdEpNS1bFJXYV46SQaAAAAqKnO5td+sqRzg6abBeYBZ63Ab7X+51R1acaluwAAAEBNdjah9DtJbYwxLYwxoZKuk/Tv8ikLtd0PB9KVmVugLnRyBAAAANRoZX0kzJuSvpF0vjEmyRhzi7U2X9Idkv4raZOkt621GyquVNQmiUmpkkQoBQAAAGq4sva+O+Yk8z+R9Em5VgRI+j7piCJDvWoRTydHAAAAQE1GDzKoktYlpapT0xh5PcbtUgAAAABUIEIpqpy8Ar827jnKpbsAAABALUAoRZWzZW+acvP99LwLAAAA1AKuhlJjzChjzKzU1FQ3y0AV830ynRwBAAAAtYWrodRa+5G1dkJMDOEDxyQmpSqmjk/N4yLcLgUAAABABePyXVQ5iUlH1KVZjIyhkyMAAACgpiOUokrJzivQlr1p6tyU1nMAAACgNiCUokrZtOeo8v2WTo4AAACAWoJQiiqFTo4AAACA2oVQiiolMSlV8VGhOicm3O1SAAAAAFQCQimqFKeTo1g6OQIAAABqCUIpqoyMnHxt359OJ0cAAABALUIoRZWx4eej8lup67mEUgAAAKC2IJSiyvj/7d17kJ3lYR/+77O6ottK6GYMxlwFJqljY+IEg8F24/oyQ1wn007SS+okHSbtL7+26TWdznQybX/T/txpm7bptKOmdi7TJu24N9M0dXBrJNuY2I5jYxBokbkYBOyuENpdIXYlrZ7+sQdHkdEFtLvPe875fGbOaPWe25fzvjroq+d53+fBZ44kSb7XSCkAAAwNpZTO+ObBqVw2ujY7NrrIEQAADIumpbSUclcpZffU1FTLGHTEg89MWQoGAACGTNNSWmu9p9Z69+ioIjLspl4+kScOvZS3XrG5dRQAAGAZmb5LJzx0cGG03JV3AQBguCildMKDzyyUUtN3AQBguCildMI3Dx7JlZeuy+Z1q1tHAQAAlpFSSid842kXOQIAgGGklNLcC0fncvDIy0opAAAMIaWU5h78zkWOXHkXAACGjVJKc998ZiqlJN97+abWUQAAgGWmlNLcg89M5Zpt67Nx7arWUQAAgGWmlNLU3Mn5fO3bL+b7rjB1FwAAhpFSSlO/9eBzOfzS8fzxt1/eOgoAANCAUkoztdb8uy88ket2bMi7r9/WOg4AANBA01JaSrmrlLJ7amqqZQwa+cqTL+bhZ6fzU7ddnVJK6zgAAEADTUtprfWeWuvdo6PWpxxGn/jCE9m8blU+auouAAAMLdN3aeLpw8fyO/uez59655W5ZPWK1nEAAIBGlFKa+NX7n8xIKfmzt765dRQAAKAhpZRld3TuZP7jV57Oh//IZbls9JLWcQAAgIaUUpbdp776dGbmTuanbr+6dRQAAKAxpZRldepUzSfvfzI3X7k5b3vT5tZxAACAxpRSltX/eXQiT71wzCgpAACQRCllmX3ii0/kjaNr88HveUPrKAAAQAcopSybR56bzv3feiE/8a6rsnKFQw8AAFBKWUaf+MITuWTVivzY97+pdRQAAKAjlFKWxaGjc/nvX382P/qOy7N53erWcQAAgI5QSlkW//6Bb+f4/Kn85G0ucAQAAPwBpZQlN3dyPr/+wFN57w3bc+32Da3jAAAAHaKUsuT+xzeey6Gjc5aBAQAAvkvTUlpKuauUsntqaqplDJZQrTWf+OITuX7Hhtx+3bbWcQAAgI5pWkprrffUWu8eHR1tGYMl9OUnDufhZ6fzU7dfnVJK6zgAAEDHmL7Lkvq3n388W9atykfffnnrKAAAQAcppSyZrz55OJ99ZCI/ffvVWbtqRes4AABABymlLIlaa/7hbz+aHRvXuMARAABwVkopS+LefeP5vadezF/5oV1Zt3pl6zgAAEBHKaUsupPzp/Lxz+zPNdvX50/eckXrOAAAQIcppSy6T/3eMzkwcTR/8wM3ZuUKhxgAAHB2GgOL6uXj8/ln2vHWVgAAG1xJREFUnx3L26/cnA98z87WcQAAgI5TSllUn7z/iYxPz+Vvf+gt1iUFAADOSyll0bz40vH86/u+lR96y4688+pLW8cBAAD6gFLKovlXnzuQl+ZO5m984MbWUQAAgD6hlLIonnnxWH7tS0/lR2++Ije8YWPrOAAAQJ9QSlkU//R3xlJK8nPv39U6CgAA0EeUUi7avmen81+/fjAfu+2qvHHzJa3jAAAAfUQp5aJ9/DOPZtPaVfmLd17XOgoAANBnlFIuyv3fOpT79k/m/3nvtRldt6p1HAAAoM80LaWllLtKKbunpqZaxuB1qrXmH/32o3nj6Nr8xK1XtY4DAAD0oaaltNZ6T6317tHR0ZYxeJ1++6Hn8+AzU/m59+/K2lUrWscBAAD6kOm7vG7/5WsHc/nmS/IjN1/ROgoAANCnlFJel7mT87n/W4fyvht3ZMVIaR0HAADoU0opr8tXn3wxx47P5z03bG8dBQAA6GNKKa/LnrHJrF4xkluv3do6CgAA0MeUUl6X+/ZP5J1XX5p1q1e2jgIAAPQxpZTX7NkjL2ds/KipuwAAwEVTSnnN9oxNJknu3KWUAgAAF0cp5TW7b/9ELt98Sa7bsaF1FAAAoM8ppbwmx0+eyhcPvJA7b9ieUiwFAwAAXByllNfka99+MUfnTpq6CwAALAqllNfkvv2TWbWi5LbrtrWOAgAADACllNfkvv0TueXNl2bDGkvBAAAAF08p5YKNT8/m0edncqelYAAAgEWilHLB9uxfWArG+qQAAMBiUUq5YPeNTeQNm9bmhp0bW0cBAAAGhFLKBTk5fyqff+xQ7txlKRgAAGDxKKVckN9/+khmZk+augsAACwqpZQLct/+iawYKbntekvBAAAAi0cp5YLsGZvMO67ckk1rV7WOAgAADBCllPOamJnNQwenLQUDAAAsuqaltJRyVyll99TUVMsYnMfesUNJLAUDAAAsvqaltNZ6T6317tHR0ZYxOI89Y5PZvnFNbrpsU+soAADAgDF9l3OaP1Xz+ccmLQUDAAAsCaWUc/r600dy5NgJU3cBAIAloZRyTnvGJjNSktuvsxQMAACw+JRSzmnP/om8/cot2bxudesoAADAAFJKOasXjs7lwYNTec8uU3cBAICloZRyVnsfm0ytsT4pAACwZJRSzmrP/sls27A63/tGS/YAAABLQynlVZ06VbP3sUO54/rtGRmxFAwAALA0VrYOwPI7MX8q//yzj2V69sRZH3N09mQOv3Tc1F0AAGBJKaVD6BtPH8kvfe5ANq5ZmZUrzj4Kes329XnPrh3LmAwAABg2SukQGp+eS5L8p5+5NW+5bFPjNAAAwDBzTukQmpiZTZLs3LS2cRIAAGDYKaVDaHx6LqtWlGxZt6p1FAAAYMgppUNoYmY2OzauTSmuqgsAALSllA6hiem5bN+4pnUMAAAApXQYTczMZucmpRQAAGhPKR1C49Nz2bHRRY4AAID2lNIhM3tiPlMvnzBSCgAAdIJSOmQmZxbWKDVSCgAAdIFSOmReWaN0h5FSAACgA5TSITM+baQUAADoDqV0yExML4yUOqcUAADoAqV0yIzPzGXVipIt61a3jgIAAKCUDpvx6dls37AmIyOldRQAAACldNhMzsxlxybnkwIAAN2glA6Z8enZ7NjofFIAAKAbmpbSUspdpZTdU1NTLWMMlYmZuew0UgoAAHRE01Jaa72n1nr36OhoyxhDY/bEfI4cO2GkFAAA6AzTd4fI5MzCGqVGSgEAgK5QSofIxMzCGqXbrVEKAAB0hFI6RCameyOlG42UAgAA3aCUDpHx6YWR0h1GSgEAgI5QSofIxMxcVo6UXLpudesoAAAASZTSoTI+PZcdG9dkZKS0jgIAAJBEKR0qEzOz2e7KuwAAQIcopUNkYnouO61RCgAAdIhSOkTGZ2Zd5AgAAOgUpXRIzJ2cz5FjJywHAwAAdIpSOiReWaPUSCkAANAlSumQmJh5pZQaKQUAALpDKR0SE9OzSZIdLnQEAAB0iFI6JF4ZKd1ppBQAAOgQpXRIjE/PZuVIyaXrVreOAgAA8B1K6ZCYmJnL9o1rMjJSWkcBAAD4DqV0SIxPzzqfFAAA6ByldEhMzsy58i4AANA5SumQGJ+ezU5rlAIAAB2jlA6BuZPzefHYiezYaKQUAADoFqV0CEx+ZzkYI6UAAEC3KKVDYHx6oZQaKQUAALpGKR0CkzOzSZIdRkoBAICOUUqHgJFSAACgq5TSITAxM5sVIyVb169uHQUAAOAPUUqHwPj0XLZvWJORkdI6CgAAwB+ilA6BiZk5V94FAAA6SSkdAhPTs9nufFIAAKCDlNIhYKQUAADoKqV0wB0/eSqHXzqenZuMlAIAAN2jlA64yaOvLAdjpBQAAOgepXTAjU/PJomRUgAAoJOU0gE3Mb0wUrrdSCkAANBBSumAm5gxUgoAAHSXUjrgxqdns2KkZOv61a2jAAAAfBeldMBNTM9l+4Y1GRkpraMAAAB8F6V0wI3PzGWHNUoBAICOalpKSyl3lVJ2T01NtYwx0CamZ7Njo/NJAQCAbmpaSmut99Ra7x4dHW0ZY6BNGCkFAAA6zPTdAXb85Kkcful4dhopBQAAOkopHWCTRxfWKDVSCgAAdJVSOsAmpl9Zo1QpBQAAukkpHWDj072RUtN3AQCAjlJKB9jkzMJIqem7AABAVymlA2x8ei4rRkq2rldKAQCAblJKB9jEzGy2bVidFSOldRQAAIBXpZQOsPHpuezc5HxSAACgu5TSATYxM5cdG03dBQAAukspHWAT07PZYaQUAADoMKV0QB0/eSovvHTcSCkAANBpSumAOnR0YY1S55QCAABdppQOqPHp3hqlRkoBAIAOU0oH1MSMkVIAAKD7lNIBNfHKSOkmI6UAAEB3KaUDamJmLiMl2bpeKQUAALpLKR1Q49Oz2b5xTVaMlNZRAAAAzkopHVATM3PZsdH5pAAAQLcppQNqfHouO51PCgAAdJxSOqAmZ2az3UgpAADQcUrpADoxfyqHjh43UgoAAHSeUjqADh1dWKPUOaUAAEDXKaUDaHx6oZQaKQUAALpOKR1A49OzSYyUAgAA3aeUDqCJGSOlAABAf1BKB9DE9GxGSrJ1g1IKAAB0m1I6gCam57Jtw5qsGCmtowAAAJyTUjqAxmdms3OT80kBAIDuU0oH0MT0XHZsNHUXAADoPqV0AE3MzGaHkVIAAKAPKKUD5ujcybzw0vG8QSkFAAD6gFI6YPbsn0ytya3Xbm0dBQAA4LyU0gFz777ns2Xdqtx85ebWUQAAAM5LKR0gJ+ZP5f88OpH33bgzK1fYtQAAQPdpLgPkK08ezvTsybz/pp2towAAAFwQpXSA3LtvPGtWjuSOXdtaRwEAALggSumAqLXm3n3juf26bVm3emXrOAAAABdEKR0Qjz4/k2defNnUXQAAoK8opQPi3n3jKSX5o29RSgEAgP6hlA6Ie/eN5+1v2pztG9e0jgIAAHDBlNIB8NzUy/nmwan8kKm7AABAn1FKB8Bn940nSf6YUgoAAPQZpXQA/M6+8Vy9bX2u3b6hdRQAAIDXRCntczOzJ/LA4y/k/TftTCmldRwAAIDXRCntc3vGJnNivloKBgAA6EtKaZ+7d994tq5fnZuv3NI6CgAAwGumlPaxE/On8rlHJ/K+G3dkxYipuwAAQP9RSvvYl584nOnZk5aCAQAA+pZS2sfu3TeeNStH8u7rt7WOAgAA8LoopX2q1pp7943n3ddvy7rVK1vHAQAAeF2U0j6177npHDzysqvuAgAAfU0p7VP37htPKcn7blRKAQCA/qWU9ql7943n5iu3ZPvGNa2jAAAAvG5KaR969sjLefjZaVN3AQCAvqeU9qHPPjKeJEopAADQ95TSPnTvvvFcs219rt2+oXUUAACAi6KU9pnp2RN54PEXjJICAAADQSntM/ftn8yJ+aqUAgAAA0Ep7TP37hvP1vWr8/Yrt7SOAgAAcNGU0j5ycv5U9uyfyHtv3JEVI6V1HAAAgIumlPaRbzwzlenZk7lz1/bWUQAAABaFUtpH9o5NppTk9uu2tY4CAACwKBa9lJZS3lJK+TellE+VUv7CYr/+MNv72GTeesXmbFm/unUUAACARXFBpbSU8olSykQp5aEztn+wlLK/lHKglPLzSVJrfaTW+jNJ/mSS2xY/8nCaOnYi33j6SO683igpAAAwOC50pPRXknzw9A2llBVJ/lWSDyW5KcmPl1Ju6t33w0l+K8n/XLSkQ+4LBw7lVE3ucD4pAAAwQC6olNZa9yY5fMbmdyY5UGt9vNZ6PMlvJvlI7/GfrrV+KMmfPttrllLuLqV8tZTy1cnJydeXfojsHZvMxrUr87Y3bW4dBQAAYNGsvIjnXp7k6dN+/0ySHyilvCfJjyRZk3OMlNZadyfZnSS33HJLvYgcA6/Wmr2PTea2a7dl5QrXpgIAAAbHxZTSV1VrvS/JfYv9usPswMTRPDc1m//3fabuAgAAg+Viht0OJnnTab+/oreNRbZnbGF68x27XOQIAAAYLBdTSr+S5PpSytWllNVJfizJpxcnFqf7/GOHcs329bliy7rWUQAAABbVhS4J8xtJvpTkhlLKM6WUn661nkzys0k+k+SRJP+p1vrw0kUdTrMn5vO7T7yQO643dRcAABg8F3ROaa31x8+y/X/Gsi9L6itPHs7siVO501IwAADAAHIp147bOzaZ1StG8gPXXNo6CgAAwKJTSjtu79ihfP/VW7Ju9aJfKBkAAKA5pbTDnp+azf7xGeeTAgAAA6tpKS2l3FVK2T01NdUyRmftfeyVpWCUUgAAYDA1LaW11ntqrXePjo62jNFZe8cms2Pjmtz4ho2towAAACwJ03c7av5UzRcOHMq7r9+eUkrrOAAAAEtCKe2obx6cypFjJ3LHrm2towAAACwZpbSj9o5NppTk3S5yBAAADDCltKP2jk3me984mkvXr24dBQAAYMkopR00PXsiv//0EVN3AQCAgaeUdtD9Bw5l/lS1PikAADDwlNIO2jN2KBvWrMzNb97SOgoAAMCSUko7ptaavWOTufXarVm1wu4BAAAGm9bTMY8feikHj7ycO3aZugsAAAy+pqW0lHJXKWX31NRUyxidsndsMklyp/NJAQCAIdC0lNZa76m13j06OtoyRqfsHZvMVVvX5cqt61pHAQAAWHKm73bI3Mn5PPD4YVN3AQCAoaGUdshXn3wxL5+YtxQMAAAwNJTSDtk7NplVK0puvXZr6ygAAADLQintkD1jk3nHm7dk/ZqVraMAAAAsC6W0I8anZ/Po8zO5c9eO1lEAAACWjVLaEd9ZCsZFjgAAgCGilHbEnrHJbN+4Jm+5bGPrKAAAAMtGKe2A+VM1XzhwKHdcvz2llNZxAAAAlo1S2gEPPnMkR46dyJ03mLoLAAAMF6W0A/aOHUopybuv29Y6CgAAwLJSSjtgz9hE3nrF5mxZv7p1FAAAgGXVtJSWUu4qpeyemppqGaOpqWMn8vWnj+TO642SAgAAw6dpKa213lNrvXt0dLRljKa+cOBQTtU4nxQAABhKpu82tmdsIpvWrsz3XbG5dRQAAIBlp5Q2VGvN3rFDuf36bVm5wq4AAACGjybU0Nj40Tw/PZs7d5m6CwAADCeltKE9YxNJkjuUUgAAYEgppQ3tHTuUXTs35LLRS1pHAQAAaEIpbeTY8ZP58hOHTd0FAACGmlLayO8+fjjH50+ZugsAAAw1pbSRPWOTWbtqJN9/1aWtowAAADSjlDayZ2wyt16zNWtXrWgdBQAAoBmltIFvv3AsTxx6ydRdAABg6CmlDex5bDJJXOQIAAAYekppA3vHJnPFlkty9bb1raMAAAA0pZQus+MnT+X+A4dy567tKaW0jgMAANBU01JaSrmrlLJ7amqqZYxl9XtPvZiXjs+bugsAAJDGpbTWek+t9e7R0dGWMZbV3scms3Kk5NZrt7aOAgAA0Jzpu8tsz/7JvOPNW7Jx7arWUQAAAJpTSpfRxMxs9j03nTtvMHUXAAAgUUqX1efHDiVJ7rheKQUAAEiU0mW1Z2wy2zasyU2XbWodBQAAoBOU0mUyf6rm849N5o7rt2VkxFIwAAAAiVK6bB46OJUXj51wPikAAMBplNJl8jv7nk8pye3XbWsdBQAAoDOU0mUwe2I+v/nlp/NHb9yRrRvWtI4DAADQGUrpMvgfDz6XF146np+87erWUQAAADpFKV1itdZ88otP5PodG/Kua7e2jgMAANApSukS+72nXszDz07nY7ddlVJcdRcAAOB0SukS++T9T2bT2pX56Nsvbx0FAACgc5TSJfTc1Mv5Xw89nx9755VZt3pl6zgAAACdo5QuoV//0lOptebP/uCbW0cBAADoJKV0icyemM9vfPnbef9NO/OmS9e1jgMAANBJSukS+fTXn82Lx07kY++yDAwAAMDZNC2lpZS7Sim7p6amWsZYdLXWfPL+J3PjGzbmB6+5tHUcAACAzmpaSmut99Ra7x4dHW0ZY9F9+YnDeeS56XzsXZaBAQAAOBfTd5fAr9z/ZDavW5WPvM0yMAAAAOeilC6yg0dezmcefj4//s4rc8nqFa3jAAAAdJpSush+/UtPpZSSP2MZGAAAgPNSShfRy8cXloH5wPfszOWbL2kdBwAAoPOU0kX0375+MFMvWwYGAADgQimli6TWml/54pO56bJN+f6rtrSOAwAA0BeU0kXypcdfyP7xmXzsNsvAAAAAXCildJH8yhefzKXrV+eHv++NraMAAAD0jZWtA3TdC0fn8nc//fC5H1STzz4ynr/4nuuydpVlYAAAAC6UUnoeJ+ZrHn1u+ryP+yOXj+YnbrUMDAAAwGuhlJ7HG0bX5n//tfe0jgEAADCQnFMKAABAM0opAAAAzSilAAAANKOUAgAA0IxSCgAAQDNKKQAAAM0opQAAADSjlAIAANBM01JaSrmrlLJ7amqqZQwAAAAaaVpKa6331FrvHh0dbRkDAACARkzfBQAAoBmlFAAAgGaUUgAAAJpRSgEAAGhGKQUAAKAZpRQAAIBmlFIAAACaUUoBAABoRikFAACgGaUUAACAZpRSAAAAmlFKAQAAaEYpBQAAoJlSa22dIaWUySRPtc5BU9uSHGodgmVhX3Mmx8TwsK9ZCo6r4WJ/968311q3v9odnSilUEr5aq31ltY5WHr2NWdyTAwP+5ql4LgaLvb3YDJ9FwAAgGaUUgAAAJpRSumK3a0DsGzsa87kmBge9jVLwXE1XOzvAeScUgAAAJoxUgoAAEAzSikAAADNKKW8qlLKm0opnyul7CulPFxK+cu97ZeWUu4tpTzW+3VLb/ufLqU8WEr5Zinl/lLK9532Wh8spewvpRwopfz8Od7zz/Ve97FSyp/rbVtXSvmtUsqjvRz/6BzPf0fv/Q+UUv5FKaX0tv+J3nNPlVJcQvwMfbqv/79SytOllKNnbP9YKWWylPL13u3PX+znM4y6ckz0tv+vUso3ejn+TSllxVme/6rvU0r52d62WkrZthifzyDp0339iVLKRCnloTO2/0Ip5eBpf/4/fLGfD69Pl46r0+7/9JnHzBn3+w55nfp0f/se6Zpaq5vbd92SXJbk5t7PG5OMJbkpyceT/Hxv+88n+f97P78ryZbezx9K8ru9n1ck+VaSa5KsTvKNJDe9yvtdmuTx3q9bej9vSbIuyXt7j1md5PNJPnSWzF9O8oNJSpLffuVxSd6S5IYk9yW5pfVn27Vbn+7rH+zlPnrG9o8l+aXWn2m/37pyTPTu29T7tST5z0l+7FWef9b3SfL2JFcleTLJttafbddu/bave/ffkeTmJA+dsf0Xkvz11p+pW7eOq979P5LkP5x5zJx2v++QIdrfvcf4HunYzUgpr6rW+lyt9Wu9n2eSPJLk8iQfSfKrvYf9apI/3nvM/bXWF3vbH0hyRe/ndyY5UGt9vNZ6PMlv9l7jTB9Icm+t9XDvde5N8sFa67Fa6+d673E8yddOe+3vKKVcloW/0DxQF75Vfu20bI/UWvdfxMcx0PptX/fuf6DW+tzF/Hdzdl05JnqvPd17zMos/CXl1a7Od9b3qbX+fq31ydf+KQyHPtzXqbXuTXL49f0Xsxy6dFyVUjYk+atJ/sE5IvsOuQh9uL99j3SQUsp5lVKuysK/FP5ukp2nlYHnk+x8laf8dBZGKpOFL6WnT7vvmd62M533caWUzUnuSvK/z/L8Zy7gfTiHPtnX5/OjvWlBnyqlvOl1PJ/TdOGYKKV8JslEkpkkn3qtz+fC9Mm+Pp+f7f35/8QrUwVpqwPH1d9P8k+SHDtHTN8hi6RP9vf5+B5pQCnlnHr/4vSfk/yV0/4VO0nSG5GsZzz+vVn4gvlbi5xjZZLfSPIvaq2PL+Zrs2BA9vU9Sa6qtb41C/9y+qvneTzn0JVjotb6gSxMD1uT5H2L+dosGJB9/a+TXJvkbUmey8JfTGmo9XFVSnlbkmtrrf91MV6PcxuQ/e17pBGllLMqpazKwpfLv6+1/pfe5vHeVNlXpsxOnPb4tyb55SQfqbW+0Nt8MMnpo1VXJDlYSvmB004i/+GzPe603+9O8lit9Rd777XitOf/vd5jrzjH8zmHPtvXZ1VrfaHWOtf77S8neceFfgb8YR07JlJrnU3y35N8pHdRjVee/zMX8nzOrs/29VnVWsdrrfO11lNJ/m0WpgLSSEeOq1uT3FJKeTLJF5LsKqXc5ztk8fXZ/j4r3yMN1Q6c2OrWvVsWLjTxa0l+8Yzt/zh/+KT1j/d+vjLJgSTvOuPxK7NwAvrV+YOT1r/nVd7v0iRPZOGE9S29ny/t3fcPsvBFN3KezGde6OjDZ9x/X1zoaCD29WmvdeaFji477eePJnmg9efbj7euHBNJNryyT3uv9R+T/OyrPP+87xMXKRmIfX3a61yV775Ayel//n8uyW+2/nyH9daV4+p8x8xreR/fIYOzv8/1GN8jDY+j1gHcunlLcnsWplk8mOTrvduHk2zNwnl+jyX5bP6gTPxykhdPe+xXT3utD2fhSmzfSvJ3zvGeP9X7kjqQ5Cd7267o5XjktNf+82d5/i1JHuq9zy8lKb3tH83C+QZzScaTfKb159ulW5/u64/39ump3q+/0Nv+D5M83Psf2eeS3Nj68+3HW4eOiZ1JvtLL8VCSf5lk5Vme/6rvk+Qv9Y6Rk0meTfLLrT/fLt36dF//Rham1Z3o7duf7m3/9STf7L3Gp3PaXy7dhvO4OuP+q3Luq7H6Dhmu/e17pGO3V/7SDgAAAMvOOaUAAAA0o5QCAADQjFIKAABAM0opAAAAzSilAAAANKOUAgAA0IxSCgAAQDP/F+Ge5gQ1DgoAAAAAAElFTkSuQmCC\n", "text/plain": [ - "
" + "
" ] }, "metadata": { - "tags": [] + "tags": [], + "needs_background": "light" } }, { "output_type": "display_data", "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAisAAAEWCAYAAACnotfGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n", + "image/png": "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\n", "text/plain": [ - "
" + "
" ] }, "metadata": { - "tags": [] + "tags": [], + "needs_background": "light" } }, { "output_type": "display_data", "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiQAAAEWCAYAAABWqYxLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOydd5hURfa/30MUGB1ARMmISA6ziHEV\nQcSALIZVkFUBMeEuKwYMa0RcFWVZEcUICpgQZQW/riIqjD8UUEEGMYC4MoqIgARhSDJwfn9UdXOn\n6e7paXpmmuG8z3OfvreqbtWp27fvPV3hU6KqGIZhGIZhlCblStsAwzAMwzAMc0gMwzAMwyh1zCEx\nDMMwDKPUMYfEMAzDMIxSxxwSwzAMwzBKHXNIDMMwDMModcwhMYz9BBHpLCI/BY6/EpHOSeRziogs\nTalxiZXbXERyRGSziFxX0uV7G/qLyEeB4zwRaRInfVLX+EBDRN4RkX6lbYdReojIUyJy177kYQ6J\nUeyISIaI5IrIJYGwg0XkRxG5MME8VES2+BfIryLyiohULz6rwdt8enGWsS+oamtVzS4snb92TQPn\nzVbV5sVqXHRuAWap6sGqOjoyUkSyRWS7/45D24nxMhSR8SLyz2QNUtUMVf0+Vl6JXuNkEZHG/vsJ\n1nlRivKskCo7C0NVz1bVCSVVXjz8fXRladsRCxG5PfBdbxeRXYHjr0rbvmRR1YGqet++5GEOiVHs\nqGoecA0wSkQO88EPA/NV9fUiZNVeVTOAJkANYGhKDS1hRKR8adtQwjQCCnvgDvJOQmibWxKGpQHV\nA3VuX5qGiGO/ezfsL3ar6gOh7xoYCMwNfPetS9u+UkVVbbOtRDZgPPAK0BlYBxwRiDsU+D9gE/AZ\n8E/go0C8Ak0Dx38FZgSO6wJvAuuB74CrAnGVgVHAz34bBVT2cbWAt4CN/tzZOEf9BWA3sA3IA25J\noH6dgZ+A24FfgVzgkoj6Pwm8DWwBTvd2TwHWAsuB6wLpq/hzNgBfAzcDPwXic4HT/X55X+7/gM3A\nAqAB8P/8tdvi69E7ZGcgn5ZAtr8GXwE9I2weA/zX5/sJcFSca9DT57HR59nSh88EdgHbvR3Nopyb\nDVwZJfw14BfgN1+f1j78amAn8LvP8/98+G2B6/A1cH4gr/5Eua/i5BW8xuUCea8DJgM1fdxBwIs+\nfCPuHj48gXumsbehQpS4FsB7uPtyKdArEHcOsBD3e1kBDA3E/ejzzPPbiTjn/cVY5fprfz/wMe6e\nbxqv/Hjfnb/GHwOP+GvxPXCSD18BrAH6RdxjT/myNgMfAo0C8Sf56/mb/zwpotyg3S9R8D573Kd7\n1Je9CffbOCWQx1D/XU705X8FdAzENwD+g/uNrgvl6eMGAN/gfqPvBu1O8JnYH38/4n7fUyLiRwOP\nBur6IPCpr8c0/P3n408A5vhrvgjonMQz+kjcb2wz8D7utx+8b6L+FgPf4z8jnoU3+e97FXB5oeUX\n1WDbbEt2w7VqrMK9rC+PiJvkt6pAK//wiOqQ+HxmAMMC8f8PeAL3YsjyD4/TfNwwYB5QGzjM/2jv\n83EP4h6GFf12CiA+Lhf/MgqU8wXwlxj16wzkA//GOUGn4hyB5j5+vP8h/xH3cquKezjeDVTCtfx8\nD5zp0w/HOUg1cQ/FL4ntkNwMLAaaAwK0Bw6NvHYBO3/y+xVxDtzt3obT/MMoaPM64DigAu6BPylG\n/Zv5+nbz+d7i867k47OJ4nAEzo8aj3voH8wexzInEDce/xAMhF2Ec/TK4RywLUAdH9ef2PdVtLyC\n13gw7j6q7215GnjFx12Dc6ir4pzDY4BDEvhNNCaKQwJUw/0GLvfX/Q+4302rwHfY1texHbAaOC9W\nniTmkPwItPblZcYrP953569xvj+3PO7PxY+4l1tl4AzcPZYRuO6bgU4+/lH2vKRr4l72l3k7+vjj\nQ2PYXZEo9xFwKe5PTwXcS/IX4KDAtdkOdPf2PgjM83HlcS/3R/x3chBwso87F3d/t/T53gnMKeIz\nsX+grnVw92p1f1wB9zI/JlDXlUAbb8uU0HcK1MP9Trv7e6KbPz7Mxz+Bc1SibV8E7JkL/Av3LDgZ\n5/gE75uEfovseRYO899Jd2ArUCPu9SjKxbPNtn3dcF73ViAzEFYe9++0eSAsWgvJJv8D2gUsAer5\nuAY+7OBA+geB8X7/f0D3QNyZQK7fH4b7p9E0iq25RDgkhdQt9COsFgibDNzl98cDEwNxxwM/RuTx\nD+B5v/89cFYg7mpiOyRLgXNj2BXPITkF93AuF4h/Bf+P29s8NhDXHVgSo5y7gMmB43K4B2hnf5xN\n4Q7JVvY8KD+Pkqa6r09mwL5/xsrTp8kJXRv2zSH5BugaiKuDu28r4B7Uc4B2Rfw9NPY2BF8QQ3CO\n1OyItE8D98TIZxTwSESeRXVIgg5+UcsPf7f+Gi8LxLX1ZR0eCFsHZAWu+6RAXAbu99wA54h8GlHW\nXKB/NLsTuc98mg24LuDQtXk/ENcK2Ob3T8T9uYnWgvUOcEXE/b6VIrSSRLkf38G37gI9gK8j6jU8\nws7fcc/PW4EXIvJ+l0BLVAK2NMQ9v6oGwl4M3jeJ/hZxz5htEffgGuCEeDakfX+bUXYQkUtxD8L3\ngYcCUYfhHuorAmHB/RAdVLU67l/Kk8BsETkI9294vapuDqT9AfevAR//Q0RcXb8/AvcvZ4aIfC8i\ntyVRtSAbVHVLjLKgYL0aAXVFZGNow7VUHB6wO5g+WIdIGuAcr6JSF1ihqrsjyqkXOP4lsL8V98KI\nlVfYRp/nioi8CuM6Va3utw4iUl5EhovI/0RkE85BANfVFhUR6etn84SuaZt46YtAI+CNQL7f4F6c\nh+O6+N4FJonIzyLysIhULELetQL1/pcv6/iIe+MS4Ahfx+NFZJaIrBWR33BjEfa1jpH3ZszyE2B1\nYH8bgKpGhgXvo3DZ6sacrcfdT5G/Xdj7/oz2rCiAiAwRkW9E5Ddfl0wKXq/Ie/wgPyi4AfCDquZH\nybYR8Gjg+qzHtU4W5X6PZAKuNQf/+UJEfOTzoCKuHo2AiyK+r5NxTnOihJ6jW6OVl8RvcV3EdYv3\n7ABsUKtRQohIbVyz51W45u1eInKKj16L88zrB05pECsvVd0JjMX1d7bBjQupKSIHB5I1xP07x8c3\nioj72ee1WVVvUtUmuPEPN4pI11BRRa0nUENEqkUrK0qeK4DlgRdRdXUzULr7+FUUvA4N45S7Ajgq\nCXt/BhpEDAYMXrui5hW+ziIiOPuTySvEX3BN46fjXiKNQ9n7zwLfkYg0Ap4FBuGa9avjurqEwins\n+14BnB3xfR2kqitVdaeq3quqrXBjHnoAfRMoM15ZH0aUlaGq1/r4l3Fjphqoaiau2zHqNfFswXUn\nhYjmWETem/HKTzXh+1xEMnBdNaExX40i0kben5H1jbwnTsF1H/bCdRlUx3WdJnJPrAAaxpixtAK4\nJuIaVVHVOQnkG4upQDsRaYO7h16KiI98HuzEdaWtwLWQBG2ppqrDITwlNy/GFhpovgr3HK0ao7zC\nfov7jDkkRknxODBVVWep6ircA+JZEamsqrtwg8aGikhVEWlBnIe5n51yOe5f1vequgLXXP6giBwk\nIu2AK3DNjeC6IO4UkcNEpBZuzMaLPq8eItLUvzx/w/3jDbUWrMaN6ygq94pIJf8g7IEbCBaNT4HN\nInKriFTx/0DaiMixPn4y8A8RqSEi9YG/xylzLHCfiBztZxu0E5FDE6jHJ7h/LreISEVxmht/wo3n\nKSqTgXNEpKtvHbgJ2IH7bpLlYJ/HOtwL9YGI+Mi6VcO9kNYCiMjlOKc1EQr7vp8C7vdOD/5+Otfv\ndxGRtv7e3IR7UeyOnVWhvAU0E5HL/PdSUUSOFZGWPv5g3L/Z7SJyHO5lEWKtLztYlxygk4g0FJFM\nXNfgvpSfarqLyMkiUgm4DzeGYwVuAHgzEfmLiFQQkd64roq34uQV+T0ejPvDsxaoICJ3A4ckaNen\nuBf1cBGp5p8vf/RxT+F+n60BRCRTRC5KMN+oqOp24HWcw/mpqv4YkeRSEWnlnYZhwOv++fki8CcR\nOdM/Rw4Sp1tU3+c7UAvOXgturX2aH4D5uOdwJXFT7v8UKLuw3+I+Yw6JUeyIyHm45sObQ2GqOhb3\n7+duHzQI53X/gmumfAV38wdZJCJ5uP7ffrjZE+t9XB+cx/4z8Aaur/t9H/dP3A/tC9zAz899GMDR\nuC6kPFzf9BOqOsvHPYhzZDaKyBBfl68koKcShV+8fT/j/t0MVNUl0RL6B0kP3CDc5bh/OmP9dQC4\nF9csuxw3iDey+TbIv3EOwQzcC3EcbpYOuD7yCb4evSJs+B330Dnbl/8E0DeWzfFQ1aW4ZubHfF5/\nAv7ky0iWibhrsBI3Y2ZeRPw4oJWv21RV/RoYifsuV+PGL3ycYFkF8ooS/yiuVWKGiGz2thzv447A\nvUg24bpyPsR/X/7f6VMJ2gC4ljvc4M+LcffSL7huzso+yV+BYd6Ou3HffejcrfiZJ74uJ6jqe8Cr\nuN/AAuK/0BMpP9W8DNyD6/Y4Bt9toarrcL+Rm3AvwluAHqr6a5y8HgUuFJENIjIa15U2HfgWdy9t\nJ4FuHl/+Ltx93BQ3ePYn3PgaVPUN3DWZ5LswvsT9joCEnhWxmIC7b6P93l/AjdX4Bdd1fZ23ZQWu\n9eJ2nOO1Ave8Leo7/hLcuJl1uGfkq+x5Dhf2W9xnQrMJDCOtEJGHcNOC+5W2LYniWxdeVNX6haU1\nDMMhIuNxg6zvLG1b0gERaYgbtH+Eqm4KhGfjni9jS9CWV3GD2O8pifKshcRIC0Skhe9mEN8EfQWu\npcMwDOOAQNxYrhtxs442FZa+GMo/VkSOEpFyInIWrtUlWmthsVBi0sKGUQgH47pp6uKa2kfipuMa\nhmGUecQNhl+N6xY5q5TMOAI3nu9QXPfUtaq6sKQKty4bwzAMwzBKHeuyMQzDMAyj1LEuG6NMUb16\ndW3atGnhCUuBLVu2UK1atcITlgLpbBukt31mW3Kks22Q3val2rYFCxb8qqqHFZ6yeDGHxChTHH74\n4cyfP7+0zYhKdnY2nTt3Lm0zopLOtkF622e2JUc62wbpbV+qbROReCrQJYZ12RiGYRiGUeqYQ2IY\nhmEYRqljDolhGIZhGKWOOSSGYRiGYZQ6pkNilCkaNmmq5Xo9WtpmROWmtvmMXJye48jT2TZIb/vM\ntuRIZ9ugdOzLHX5OQumKYVDrAlXtmLIMk8RaSAzDMAwjTVi6dClZWVnh7ZBDDmHUqFH07t07HHbx\nxReTlZVV4Lwff/yRjIwM/vWvf4XDBgwYQO3atWnTJtEFr0uX9HVPjTKFiOzCrbRbEbcU+ETgEVXd\nlyXaDcMwyhTNmzcnJycHgF27dlGvXj3OP/98rr/++nCaXr167eVk3HjjjZx99tkFwvr378+gQYPo\n27dv8RueAswhMUqKbaqaBSAitXHLjR+CW3LcMAzDiOCDDz7gqKOOolGjRuEwVSU7O5v7778/HDZ1\n6lSOPPLIvcTSOnXqRG5ubkmZu89Yl41R4qjqGuBqYJBf3bexiMwWkc/9dhKAiEwUkfNC54nISyJy\nbmnZbRiGUZJMmjSJPn36FAibPXs2NWrU4OijjwYgLy+Phx56iHvu2f//29mgVqNEEJE8Vc2ICNsI\nNAc2A7tVdbuIHA28oqodReRU4AZVPU9EMoEc4GhVzY/I52qcg0OtWocdc/eoZ0uiSkXm8Cqweltp\nWxGddLYN0ts+sy050tk2KB372tbLDO/v3LmTCy+8kOeff56aNWuGwx955BFq1arFZZddBsCTTz5J\nixYt6NKlC+PHj6dKlSr07t07nP6XX37hH//4B88//3zMcrt06ZIWg1rNITFKhEIcku3A40AWsAto\npqpVfZqvgM7An4GmqjokXjk2yyY50tk2SG/7zLbkSGfboPRn2UybNo0xY8YwY8aMcFh+fj716tXj\n8ccf56KLLgLglFNOYcWKFQBs3LiRcuXKMWzYMAYNGuTyzM2lR48efPnllzHLTZdZNul7NxhlGhFp\ngnM+1uDGkawG2uO6EbcHkk4ELgUuBi4vYTMNwzBKhVdeeWWv7pr333+fFi1acNhhe9bBmz17dnh/\n6NChZGRkhJ2R/Q0bQ2KUOCJyGPAU8Li6JrpMYJWfcXMZUD6QfDxwPYCqfl3CphqGYZQ4W7Zs4b33\n3uOCCy4oEB5tTEk8+vTpw4knnsjSpUupX78+48aNS7WpKcW6bIwSIcq03xeAf6vqbj9uZAqgwHTg\nb8HuHRGZDkxV1acKK6d58+a6dOnS4qjCPnMgrR6aatLZPrMtOdLZNkhv+8qqMJp12RglgqqWjxO3\nDGgXCLo1tCMiVYGjgVcSKWfbzl00vu2/SdmYqEqiYRiGkXqsy8ZIW0TkdOAb4DFV/a0kyty+fTvH\nHXcc7du3p3Xr1uGpdDNnzqRDhw60adOGfv36kZ/vJvqMGDEirJ7Ypk0bypcvz/r160vCVMMwjDKF\nOSRG2qKq76tqI1UdVVJlVq5cmZkzZ7Jo0SJycnKYPn06c+bMoV+/fkyaNIkvv/ySRo0aMWHCBABu\nvvlmcnJyyMnJ4cEHH+TUU08tMEXPMAzDSAxzSFKAiOwSkRwR+UpEFonITSJSzsd1FJHRcc5tLCJ/\nKTlrU2uDiHQWkbciwsaLyIWFnDdUROJO4Y1yzpxkbCxiGWRkuOErO3fuZOfOnZQvX55KlSrRrFkz\nALp168aUKVP2OjfaqHjDMAwjMcwhSQ3bVDVLVVsD3YCz8ZLoqjpfVa+Lc25joFQdkjSxoVBU9aSS\nKGfXrl1kZWVRu3ZtunXrxnHHHUd+fj7z588H4PXXXw/P+w+xdetWpk+fzp///OeSMNEwDKPMYbNs\nUkCk6JfX2PgMqAWcCgxR1R5eeTSk2qVAJ+A9oCWwHJgAvIGbgRJalGCQqs4Rkc7AUOBXoA2wALhU\nVVVEjvX5VgN2AF2BrcBwnKhYZWCMqj4dw/55ETY86beOuBkxN6rqrBjndg7VLxA2HnhLVV8XkVyg\no6r+KiIdgX+pamcRGQocBTT11+lhVX3Wn38z0Mvb/Yaq3hPtOgfKS4lSa1AlEZwk81133cV1113H\n1q1befrpp9m5cycdO3Zk7ty5jB07Npx25syZvP/++zzwwAMx88/Lywu3vqQb6WwbpLd9ZltypLNt\nkN72pdq2dFFqtVk2xYCqfi8i5YHaEVFDcFNaPxaRDJwA2G0EXuh+Vkm3oIw6zjEA+APQGvgZ+Bj4\no4h8CrwK9FbVz0TkEGAbcAXwm6oeKyKVgY9FZIaqLo9icqQNN7lqaFsRaQHMEJFmqro9yrkAp4hI\nTuC4IfBWjLRB2gEn4ByphSLyX5yzdTRwHCDAmyLSSVX/X6xMVPUZ4BlwSq3JqivmXtJ5r7DPP/+c\ndevWMWTIEP72t78BMGPGDHbs2FFg2t2jjz7KoEGD4k7FO5CmEaaadLbPbEuOdLYN0tu+dLZtX7Au\nm5LlY+DfInIdUD1yTRZPReBZEVkMvAa0CsR9qqo/eQGxHFxXS3OcqNhnAKq6yed7BtDXOwqfAIfi\nXvSJcDLwos9vCfAD0CxO+tm+yyrLr+j7ZoLlTFPVbar6KzAL54Sc4beFwOdAiyLYvc+sXbuWjRs3\nArBt2zbee+89WrRowZo1awDYsWMHDz30EAMHDgyf89tvv/Hhhx9y7rm27p9hGEayWAtJMRAhi94y\nFK6qw30rQHdci8WZUU6/gdgy6jsC+7uI//0J8HdVfTepSqSOfPY4vgdFxEX2FyrO7gdjdS8VN6tW\nraJfv37s2rWL3bt306tXL3r06MHNN9/MW2+9xe7du7n22ms57bTTwue88cYbnHHGGXst/W0YhmEk\njjkkKSZSFl1EgnFHqepiYLEf99ECWAEcHMgiE/jJK5j2o6CMejSWAnVE5FjfZXMwrsvmXeBaEZmp\nqjtFpBmwUlW3RMljc4QNs4FLgJn+vIa+nGTIBY4B3sEtkBfkXBF5ENdl0xnXdbQNuE9EXlLVPBGp\nB+xU1TVJll8k2rVrx8KFC/cKHzFiBCNGjIh6Tv/+/enfv38xW2YYhlG2MYckNVTxXSMFZNGjpLte\nRLoAu4GvcC/p3cAuEVmEW7flCWCKiPTFyahHcyDCqOrvItIbeExEquBe6KcDY3FdOp+L84rWAufF\nyOaLKDY86buN8oH+qrojxrmFcS8wTkTuA7KjlDsLN6j1PlX9GfhZRFoCc70zl4dbXC8hh6RKxfIs\nNcVVwzCM/Q5zSFJAIbLo2fgXsar+PUay0yKO95JRD+bjjwcF9j/DDQ6N5Ha/xUVVd0axIaGVdSPt\n8mH9A/uziTL+RFWHxsnzUfbMRgqGFzqsvCjS8bP/1o6+ffuyevVqRISrr76awYMHk5OTw8CBA9m+\nfTsVKlTgiSee4LjjjkNVGTx4MG+//TZVq1Zl/PjxdOjQIaGyDMMwjPiYQ2IcsFSoUIGRI0fSoUMH\nNm/ezDHHHEO3bt245ZZbuOeeezj77LN5++23ueWWW8jOzuadd95h2bJlLFu2jE8++YRrr72WTz75\npLSrYRiGUSYwh6QMISKPAD+EpNZF5F1ghape6Y8n4Lpz1gZO26Gqx8fJczxuCu9m4KFAVFPcINTl\nQCXgfeBOVd2Ywvo8B/QA1qhqm1TlG6JOnTrUqVMHgIMPPpiWLVuycuVKRIRNmzYBbgZN3bp1AZg2\nbRp9+/ZFRDjhhBPYuHEjq1atCudhGIZhJI85JGWLj3GCYqO8dH0t4JBAfDPgz6o6r6gZ+9k64Rk7\nIpKN0y6ZLyKVgAeBaTghuFQxHngcmJjCPKOSm5vLwoULOf744xk1ahRnnnkmQ4YMYffu3cyZ4xTr\nV65cSYMGDcLn1K9fn5UrV5pDYhiGkQLMISlbzAEe8futgS9xM3Bq4JRbWwIqIh8CGTjV1/6qukpE\njgLGAIf5tFd5DZIwfmBqA5zoWhg/sPYW4DsRaa+qi0Rkqk97EPCoqj4jIgOAdqp6vc/vKqCVqt4Q\nrTKq+v9EpHFhlY5QauXuttHkXfYmOzsbcHojgwcP5sorr+Tzzz9n9OjRXHHFFZx66qnMmjWLCy64\ngJEjR7Ju3ToWLlwYXul3w4YNLFiwgLy8vITKy8vLC5eZbqSzbZDe9pltyZHOtkF625fOtu0LJh1f\nxhCR5bhWirNxmh71gLnAb8C/cN0s56rqWj8750xVHSAiHwADVXWZiByP0wI5LdBlczxuavC1fjpz\nNr6FJFD2VOAVVX1VRGqq6no/8+czb9MOYBHQwk9FngNc46dCx6pPY5wMfUJdNg2bNNVyvfYaDxuV\n3OHnsHPnTnr06MGZZ57JjTfeCEBmZiYbN25ERFBVMjMz2bRpE9dccw2dO3cOL6DXvHlzsrOzE24h\nSWd1xXS2DdLbPrMtOdLZNkhv+1Jtm4ikhXS8KbWWPeYAJ/ltrt9Cxytx0uzv+WnKdwL1vYz9ScBr\nPvxpIPiWvQvIVNWBGt+DlcD+dX4a8TxcS8nRqpoHzAR6eEn6ivGckeJGVbniiito2bJl2BkBqFu3\nLh9++CHg1qg5+mgnFNuzZ08mTpyIqjJv3jwyMzOtu8YwDCNFWJdN2eNjnHPRFtdlswK4CdiEm55b\nT1VPDJ7g17/Z6GXfo/EZcEyo1SNaAr92T1vgG7/g3unAiaq61bemhFRax+KmIi8Bnk+yjinh448/\n5oUXXqBt27ZkZbmqP/DAAzz77LMMHjyY/Px8DjroIJ555hkAunfvzttvv03Tpk2pWrUqzz9fquYb\nhmGUKcwhKXvMwS3i972q7gLWi0h13JiSa4C/i8iJqjpXRCoCzVT1KxFZLiIXqeprXkitnaou8nlO\nxw1o/a+InKGqm4MF+nzux83o+UJEzgU2eGekBQGNFFX9REQaAB0oqLdS4px88snEavBZsGDBXmEi\nwpgxY4rbLMMwjAMSc0jKHotxs2tejgjLUNU1InIhMFpEMnHf/yicauwlOHXWO3GKs5Nw4z0A8I7K\nwbjVd7v74JdEZAdQGTftN7S63HRgoIh8g5Ocj5zVMxnIUtUN8SoiIq/gJOVrichPwD2qOi7eOabU\nahiGsX9iDkkZw7eKHBIR1j+wnwN0inLecuCsKOHBc58DnvOHnePYsAM3qDYWJ7NnNlBMVLVPYWki\nSUSpNdccFsMwjLTDBrUaJYaIVBeRb4FtqvpBadqyYsUKunTpQqtWrWjdujWPPlpwZs7IkSMREX79\n9VfAiaK1a9eOrKwsOnbsyEcffVQaZhuGYZRZrIXEKDG8imszABHJU9UMETkUiOacdFXVdT5tZ+B3\nVZ2TKltiyca3atWKFStWMGPGDBo2bLjHmK5d6dmzJyLCF198Qa9evViyZEmcEgzDMIyiYC0kRqmi\nqutUNSvKti6QrDNu5lDKqFOnTnhhvKBsPMANN9zAww8/jF9tGICMjIzw8ZYtWwrEGYZhGPuOOSRG\n2iAifxKRT0RkoYi8LyKHe2G0gcANIpIjIqekutygbPy0adOoV68e7du33yvdG2+8QYsWLTjnnHN4\n7rnnouRkGIZhJIsptRqlQqjLJiKsBk4PRUXkSqClqt4kIkOBPFX9V4y8gtLxx9w96tm4Zbetlxne\nD8nGX3rppRx33HHccMMNjBgxgoyMDC6++GKefvppMjMzC5y/aNEiJk6cyMiRI4tU57y8PDIyMgpP\nWAqks22Q3vaZbcmRzrZBetuXatu6dOmSFkqt5pAYpUIMh6QtMBKnElsJWK6qZxXmkARJRDo+NMsm\nUjZ+8eLFdO3alapVqwLw008/UbduXT799FOOOOKIAnk0adKETz/9lFq1aiVa5QNKijrVpLN9Zlty\npLNtkN72mXS8YRQ/jwGPq2pbnIjbQYWkT5posvFt27ZlzZo15ObmkpubS/369fn888854ogj+O67\n78Iiap9//jk7duzg0EMPLS7zDMMwDjhslo2RTmTi1tsB6BcI30yEtsq+Eks2vnv37lHTT5kyhYkT\nJ1KxYkWqVKnCq6++agNbDcMwUog5JEZpUdWrr4b4NzAUt8DfBtwifEf6uP8DXveS9H9X1dmxMk1U\nqTWebHyI3Nzc8P6tt97KrVQOQ20AACAASURBVLfeWmi+hmEYRnKYQ2KUCqoaq7twWpS035Lgujem\n1GoYhrF/YmNIjAOOWCqtd911V1iN9YwzzuDnn38GYMmSJZx44olUrlyZf/2r0HG1hmEYRhKYQ2Kk\nHBHJK20b4hFSaf3666+ZN28eY8aM4euvv+bmm2/miy++ICcnhx49ejBs2DAAatasyejRoxkyZEgp\nW24YhlF2MYfEOOCIpdJ6yCF7xs0G1Vhr167NscceS8WKFUvFXsMwjAMBG0NilAgikgU8BVQF/gcM\nUNUNIpINfAJ0AaoDV6jqbBGpCowH2gBLgbrA31R1firtCqq0Atxxxx1MnDiRzMxMZs2alcqiDMMw\njDiYMJqRcmKInn2BmyHzoYgMAw5R1eu9Q7LAK7J2B25U1dNFZAhwtKpeIyJtgBzghGgOSbJKrUGV\n1k6dOhVI89JLL/H7779z+eWXh8PGjx9PlSpV6N27dxGviONAUn5MNelsn9mWHOlsG6S3fWVVqdVa\nSIxiR0Qygeqq+qEPmgC8FkjyH/+5AGjs908GHgVQ1S+9QxMVVX0GeAacUuvIxfFv69xLOodVWgcO\nHBgWRgvSpEkTunfvzoQJE8Jh2dnZZGRkJK2QeCApP6aadLbPbEuOdLYN0tu+dLZtX7AxJEY6sMN/\n7qIEnORoKq0Ay5YtC+9PmzaNFi1aFLcphmEYhsdaSIxiR1V/E5ENInKKFzW7DPiwkNM+BnoBs0Sk\nFdA2VfbEUmkdN24cS5cupVy5cjRq1IinnnoKgF9++YWOHTuyadMmypUrx6hRo/j6668LDII1DMMw\n9g1zSIziIJoKaz/gKT9Y9Xvg8qhn7uEJYIKIfA0sAb4CfkuFcbFUWmPJxh9xxBH89NNPUeMMwzCM\n1GAOiZFy4qiwnhAlbefA/q/sGUOyHbhUVbeLyFHA+8APhZWdqHS8YRiGkV6YQ2KkK1Vx3TUVAQH+\nqqq/F3ZSPOn43OHnsGLFCvr27cvq1asREa6++moGDx7M+vXr6d27N7m5uTRu3JjJkydTo0YNVJXB\ngwfz9ttvU7VqVcaPHx/WMDEMwzBShw1qNdISVd2sqh1Vtb2qtlPVd1KRbyyV1uHDh9O1a1eWLVtG\n165dGT58OADvvPMOy5YtY9myZTzzzDNce+21qTDDMAzDiMAckjKEiDwiItcHjt8VkbGB45Eisvcc\n1/h5jheRC6OEZ4vIUhH5QkSWiMjjIlJ932pQIP8GIjJLRL4Wka9EZHAq8o2l0jpt2jT69esHQL9+\n/Zg6dSrgZtv07dsXEeGEE05g48aNrFq1KhWmGIZhGAHMISlbfAycBCAi5YBaQOtA/EnAnBSWd4mq\ntsOtxLuDKCv17gP5wE2q2go39uRvfrZNygiqtK5evZo6deoAbhDr6tWrAVi5ciUNGjQIn1O/fn1W\nrlyZSjMMwzAMbAxJWWMO8Ijfbw18CdQRkRrAVqAloCLyIZAB/Ar0V9VVfuDoGOAwn/YqVV0SzFxE\n7gMaAFcEw1X1dxG5BfhORNqr6iIRmerTHgQ8qqrPiMgAoJ2qXu/zuwpopao3RFZEVVcBq/z+ZhH5\nBqgHfB2ZNkKplbvb5ke9ONnZ2eH9kErrlVdeyeeff05+fn6B+F27dpGdnc26detYuHAh+fkuzw0b\nNrBgwQLy8oq+fmBeXl6BMtKJdLYN0ts+sy050tk2SG/70tm2fcEckjKEqv4sIvki0hDXGjIX9xI/\nETdl9hucw3Kuqq4Vkd7A/cAAnNLpQFVdJiLH46bdnhbKW0RGAAcDl6uqhhaeC5S9S0QWAS2ARbi1\nataLSBXgMxGZAkwG7hCRm1V1J27q7zWF1UtEGgN/wK15E63eCSm15l7SGSCqSmu9evVo3rw5derU\nYdWqVdStW5fOnTvTrl07atWqFVZF3LJlCz179gy3phSFdFZXTGfbIL3tM9uSI51tg/S2L51t2xes\ny6bsMQfnjIQckrmB45W4xereE5Ec4E6gvohk+PjXfPjTQPCNexeQqaoDNf7iR0Ev5TrvoMzDtZQc\nrap5wEygh4i0ACqq6uJ4lfG2TQGuV9VNCV2BOMRSae3Zs2dYJn7ChAmce+654fCJEyeiqsybN4/M\nzMyknBHDMAwjPtZCUvYIjSNpi+uyWQHcBGwCsoF6qnpi8AQROQTYqKpZMfL8DDhGRGqq6vpoCUSk\nvC/zGxHpDJwOnKiqW/0Cegf5pGOB23FiZ8/Hq4if8jsFeElV/xMvbaLEUmm97bbb6NWrF+PGjaNR\no0ZMnjwZcGJpb7/9Nk2bNqVq1ao8/3xckw3DMIwkMYek7DEHGAJ8r6q7gPV+9ktrXPfI30XkRFWd\n61/4zVT1KxFZLiIXqepr4vpj2qnqIp/ndOBd4L8icoaqbg4W6PO5H1ihql+IyLnABu+MtCAgiKaq\nn4hIA6ADbjBsVLwN44BvVPXfKbkyxFZpBfjggw+i2cGYMWNSVbxhGIYRA3NIyh6LcbNrXo4Iy1DV\nNX4K72i/Am8FYBROlv0S4EkRuROoCEzCjQUBwDsqBwNvikhIY/0lEdkBVMYpqZ7rw6cDA/1A1KW4\nbpsgk4EsVd0Qpx5/xK15s9h3IwHcrqpvx6u8KbUahmHsn5hDUsbwrSKHRIT1D+znAJ2inLccOCtK\nePDc54Dn/GHnODbsAM6OY+bJ7JkNFCuPjyg4JiUhYim15pqTYhiGkdbYoFajxBCR6iLyLbBNVffu\nHylmBgwYQO3atWnTpk2B8Mcee4wWLVrQunVrbrnlFgA+/fRTsrKyyMrKon379rzxxhslba5hGMYB\nhbWQGCWGqm4EmgXDRORQIJpz0lVV16Wy/P79+zNo0CD69u0bDps1axbTpk1j0aJFVK5cmTVr1gDQ\npk0b5s+fT4UKFVi1ahXt27fnT3/6ExUq2E/GMAyjOLAWkjLE/igdr6rrVDUryrZORM7yZXwnIrcV\nNe9IOnXqRM2aNQuEPfnkk9x2221UrlwZgNq1awNQtWrVsPOxfft2InVXDMMwjNRiDknZosxIx/tp\nxGNwY1FaAX1SLR0P8O233zJ79myOP/54Tj31VD777LNw3CeffELr1q1p27YtTz31lLWOGIZhFCP2\nhC1blBnpeOA44DtV/d6nnYSbxZOUdHxIZvmXX35hy5Yt4ePffvuNxYsXM3z4cJYsWULPnj15+eWX\nwy0iY8aM4YcffuD222+nWrVqVKpUKYqpiZHOcs/pbBukt31mW3Kks22Q3vals237hKraVoY2YDnQ\nEKc5MhC4D+iOm0Y7F+e0HObT9gae8/sf4NRUAY4HZvr98cCFwAjgKUB8eDbQMaLsqUBvv1/Tf1bB\nOUaH4pyg/+EUWvG2tI1RjwuBsYHjy4DHC6t/gyOP0ka3vrXXFmL58uXaunXr8PGZZ56pM2fODB83\nadJE16xZo5F06dJFP/vss73Ci8KsWbP26fziJJ1tU01v+8y25Ehn21TT275U2wbM1zR4f1mXTdmj\nTEnHFzfnnXces2bNAlz3ze+//06tWrVYvnx5eEG9H374gSVLltC4ceNStNQwDKNsY102ZY+yIh2/\nEufIhKjvw5KmT58+ZGdn8+uvv1K/fn3uvfdeBgwYwIABA2jTpg2VKlViwoQJiAgfffQRw4cPp2LF\nipQrV44nnniCWrVq7UvxhmEYRhzMISl7lAnpeJwTdLSIHIlzRC4G/lJY5eMptb7yyitRw1988cW9\nwi677DIuu+yywoozDMMwUoR12ZQ9QtLx8yLCflPVNbixGQ/57pQc/KwcnHT8FT78K/bIwANOOh54\nFicdX8UHvyQiX+BaYqpRUDq+gpeOH0506fiPNY50vKrmA4NwjtA3wGRV/SqxS2AYhmHsb1gLSRlD\ny4h0vM/nbSDu2jWRxJOOHzBgAG+99Ra1a9fmyy+/BGDo0KE8++yzHHbYYYBb+bd79+7h83788Uda\ntWrF0KFDGTJkSFFMMQzDMIqAtZAYJUZpS8f379+f6dOn7xV+ww03kJOTQ05OTgFnBODGG2/k7LPj\n+VaGYRhGKrAWkjKEiDwC/KCqo/zxu7hxHVf645HASlX9dxHyHA+8paqvR4Rn42bi7AAq4Vb7vVOd\nPHxUtIjS8cAu3CDYNoACA1R1bqK2R9KpUydyc3MTTj916lSOPPJIqlWrlmyRhmEYRoJYC0nZYr9T\natU40vHAo8B0VW0BtMeNJUk5jz/+OO3atWPAgAFs2OCGteTl5fHQQw9xzz33FEeRhmEYRgQhkSuj\nDCAidYFPVLWBiLTFzbapgxNA2wqsBs4EHiZBpdZgC0mEUusHwBBVne/LLg98B5ynKVBqFZFM3KDb\nJoVon0QqtR5z96hn90rTtl4m4JRa//GPf/D8827G8fr168nMzEREeO6551i3bh233norTz75JC1a\ntKBLly6MHz+eKlWq0Lt378K+grjk5eWRkZGxT3kUF+lsG6S3fWZbcqSzbZDe9qXati5duixQ1Y4p\nyzBZSluZzbbUbpQdpdYs4FNf/kJc1021wupfVKXWIMG4k08+WRs1aqSNGjXSzMxMrVGjhj722GNR\nz0uUA0n5MdWks31mW3Kks22q6W1fWVVqtTEkZY+gUuu/gXp+/zecnscZOKVWgPLAqgil1lA+lQN5\n3oVrebm6kLIjlVrP9/shpdZ5IhJSav2G+EqtFXBaJX9Xp13yKHCbtyVlrFq1ijp1nCjtG2+8QZs2\nbQCYPXt2OM3QoUPJyMhg0KBBqSzaMAzDCGAOSdmjrCi1/gT8pKqf+OPXcQ5J0kRTas3OziYnJwcR\noXHjxjz99NP7UoRhGIaRJOaQlD3KhFKrqv4iIitEpLmqLsXNutlrpd+iEE2p9YorroiSsiBDhw7d\nl2INwzCMBDCHpOwRUmp9OSIsQ1XXiMiFwGg/aLQCMAqnzHoJ8KSI3AlUBCYBIYcE76gcjFNqDYl1\nvCQiO3DdO+9TUKl1oO+WWUp0pdYsjaPU6vm7L6MS8D1weWGVjycdbxiGYaQv5pCUMbRsKbXmAEUa\n+R1NqTWWSmuIkSNHMmTIENauXUutWrVQVQYPHszbb79N1apVGT9+PB06dCiKGYZhGEYRSUiHRESO\nEpHKfr+ziFznuwEMI2FKU6k1lkrrihUrmDFjBg0bNgyHvfPOOyxbtoxly5bxzDPPcO2115akqYZh\nGAckiQqjTQF2iUhT4BncrImX459ilDQi8oiIXB84fldExgaOR4rIjUXMc7zv5okMzxaRpSLyhYgs\nEZHHC3NSVXWjqjZT1YsC+RwqIjlRtkNFJFdEFvvj+UWxO5JOnTpRs2bNvcJvuOEGHn74YQKzi5g2\nbRp9+/ZFRDjhhBPYuHEjq1at2pfiDcMwjEJI1CHZrW711fOBx1T1ZpzglpFelDWlVoAu/jjloj3T\npk2jXr16tG/fvkD4ypUradCgQfi4fv36rFy5MtXFG4ZhGAESdUh2ikgfoB/wlg+rWDwmGfvAHCA0\npbc1btrvZhGp4bvcWgIqIh+KyALfglIHwt1y0334bD87pgAicp9vMSkfDFfV34FbgIYi0t6nnerz\n+sorqSIiA0RkVCC/q/z6OyXO1q1beeCBBxg2bFhpFG8YhmFEkOig1stxqp/3q+pyETkSeKH4zDKS\nQVV/FpF8EWmIaw2ZixNGOxEnjPYNbjDpuaq6VkR646brDsB1xQ1U1WUicjzwBHBaKG8RGQEcDFyu\nqhrs4vBl7xKRRUAL3OycAaq6XkSqAJ+JyBTc7Jo7RORmVd2Ju6+uiVclYIaIKPC0qj4TLVGEdDx3\nt80vEJ+dnQ042fgtW7aQnZ3N999/z7fffkvz5s0BWLt2La1bt+bJJ59ERHj33XfJz3f5LFu2jB9+\n+IG8vLw4phZOXl5e2JZ0I51tg/S2z2xLjnS2DdLbvnS2bZ9IVNIVJwHevLSlZW0r9Ht6CbgYmIBb\nkK478E/gZpy42CbcGjE5uOnAM3CS7tsC4TnANz6/8TgH45mIcrLZWzp+Gnuk44f68xbhnKETfPiz\nuK6/FsBnhdSlnv+s7fPpVFj9o0nHh4gnG9+oUSNdu3atqqq+9dZbetZZZ+nu3bt17ty5euyxx0Y9\np6gcSFLUqSad7TPbkiOdbVNNb/vKqnR8orNs/uRfUtP9cZaIvJmYy2OUMJFKrfNwLSQnAbOBr3TP\nOI22qnoGrutuoxYcw9EykGdYqTVWoXGUWtvj1qIJKrX2x7WOxFNqRVVX+s81wBvAcYlfhoL06dOH\nE088kaVLl1K/fn3GjRsXM2337t1p0qQJTZs25aqrruKJJ55ItljDMAwjQRLtshmKexlkg9OHEJEm\nxWSTsW+UCaVWEakGlFPVzX7/DCDpAR/RVFqD5ObmBstmzJgxyRZlGIZhJEGiDslOVf0tYtzA7mKw\nx9h3yopS6+HAG/6eqwC8rKp7C4lEYEqthmEY+yeJOiRfichfgPIicjRwHamdPmqkCC0jSq2q+j1u\nDEyRiFRqzTXnxDAMY78g0Wm/f8c1+e/A/fP+Dbg+7hmGEUFpKbUOGDCA2rVr06ZNm3DY+vXr6dat\nG0cffTTdunVjwwbXWPPSSy/Rrl072rZty0knncSiRYtiZWsYhmGkkEIdEj9Y8b+qeoeqHuu3O1V1\newnYZ5QhtIhKrakqN5ps/PDhw+natSvLli2ja9euDB8+HIAjjzySDz/8kMWLF3PXXXdx9dVXp8oM\nwzAMIw6FOiS+C2C3H3NgpDHpLh0fDS1EqVVEyovIQhF5q7C8YhFNNn7atGn069cPgH79+jF16lQA\nTjrpJGrUqAHACSecwE8//ZRssYZhGEYRSLTLJg9YLCLjRGR0aCtOw4yk2O+k4xNgME7QLaWsXr2a\nOnXc6gdHHHEEq1ev3ivNuHHjOPvseENhDMMwjFSR6KDW//jNSG/msGewaEg6vo6I1AC2EpCOx4mh\n/Qr0V9VVInIUMAY4zKe9SlWXBDMXkftwCyteEQxX1d9F5BbgOxFpr6qLRGSqT3sQ8KiqPiMiA3DT\nia/3+V0FtFLVG6JVRkTqA+fgphTHbNmJp9QaTaUVID8/v4DS4a5duwocL1y4kMcee4zRo0enTBEx\nndUV09k2SG/7zLbkSGfbIL3tS2fb9onSVmazLbUbsBxoiNMcGQjch1Nr/SNOSn4OcJhP2xt4zu9/\nABzt948HZvr98cCFwAjgKUB8eDZ7K7VOZY9Sa03/WQXnGB2Kc4L+B1T0cXOAtnHq8jpwDG5Gz1uJ\n1D9SqTVEpEprs2bN9Oeff1ZV1Z9//lmbNWsWjlu0aJE2adJEly5dqqnkQFJ+TDXpbJ/ZlhzpbJtq\nett3oCu1LheR7yO3Ivg9RskxB9c1E1rLZm7geCXQBnhPRHKAO4H6IpLh41/z4U9TcDXnu4BMVR3o\nb95YBIVqrvNr28zDtZQcrap5wEyghxdMq6iqi6NmJNIDWKOqC4pW/cTo2bMnEyZMAGDChAmce66T\nUPnxxx+54IILeOGFF2jWrFlxFG0YhmFEIdEum+DS7wcBFwExZcSNUiVSOn4FcBNuDZts3PowJwZP\nEJFD8NLxMfIMS8er6vpoCeJIx28VkWwKSsffDiwhvnT8H4GeXoTtIOAQEXlRVS+Nc05U+vTpQ3Z2\nNr/++iv169fn3nvv5bbbbqNXr16MGzeORo0aMXnyZACGDRvGunXr+Otf/wpAhQoVmD9/flGLNAzD\nMIpIQg6J+hkPAUaJyALg7tSbZOwjZUI6XlX/AfzD598ZGJKMMwKxZeM/+GBvKZSxY8cyduzYKKkN\nwzCM4iQhh0REOgQOy+FaTBJtXTFKlrIiHZ8UJh1vGIaxf5KoUzEysJ+PGzjZK/XmGPuKlhHp+Ij8\nsvELOxZGUDo+JBv/yCOPMHbsWESEtm3b8vzzzzNw4EA+/PBDMjOdvM748ePJyorVY2UYhmEUN4k6\nJFeoW1skjIgcWQz2GGUY33X0KbBIS0g6fuXKlYwePZqvv/6aKlWq0KtXLyZNmgTAiBEjuPDCvTTf\nDMMwjFIgUWG01xMMM0qRdFdq1aJJx9cVkU9FZJGIfCUi9xbF7iD5+fls27aN/Px8tm7dSt26dZPN\nyjAMwygm4jokItJCRP4MZIrIBYGtP3tmTRjpw36n1KoxpOOBVcBpqtoeyALOEpET4ue2N/Xq1WPI\nkCE0bNiQOnXqkJmZyRlnnAHAHXfcQbt27bjhhhvYsWNHUbM2DMMwUkhI5Cp6pJstcR7QE3gzELUZ\nmKSqqXy5GfuIiNQFPlHVBiLSFjfbpg5OAG0rsBo4E3iYBJVaRWQ8TpTs9Qil1g9wM1/m+7LLA98B\n52mKlFoD9aoKfARcq6qfRIkPKrUec/eoZwFoWy+TzZs3c88993D33XeTkZHB0KFDOfXUU+nQoQM1\na9Zk586djBw5krp164bXtiku8vLyyMjIKNYykiWdbYP0ts9sS450tg3S275U29alS5cFqtqx8JTF\nTCLqaTg9iVJXcbMtoe+qLCm1lgdycGspPZRI/YNKraqqkydP1gEDBmiICRMm6LXXXqtBZs2apeec\nc44WNweS8mOqSWf7zLbkSGfbVNPbvrKq1JrooNaFIvI3XPN/uKtGVQckeL5RcgSVWv8N1PP7v+GU\nWs/AKbWCe+GvilBqDeVTOZDnXbiWl6sLKTtSqfV8vx9Sap0nIiGl1m+Io9QK4RlDWX5syhsi0kZV\nvyzEhgI0bNiQefPmsXXrVqpUqcIHH3xAx44dWbVqFXXq1EFVmTp1Km3atClKtoZhGEaKSdQheQGn\nrHkmMAynWZHyFViNlFBWlFrDqOpGEZmFm5ZcJIfk+OOP58ILL6RDhw5UqFCBP/zhD1x99dWcffbZ\nrF27FlUlKyuLp556qijZGoZhGCkmUYekqapeJCLnquoEEXkZmF2chhlJUyaUWkXkMGCnd0aqAN2A\nh5K5IPfeey/33ltwks7MmTOTycowDMMoJhJ1SHb6z40i0gb4BahdPCYZ+0hZUWqtA0zwLS/lgMmq\n+lZhlTelVsMwjP2TRB2SZ0SkBm4swZu4wYm2jk0aomVEqVVVvwD+EC9NNCKVWpcuXUrv3r3D8d9/\n/z3Dhg2jS5cuDBw4kLy8PBo3bsxLL73EIYccEitbwzAMo5hJSBhNVceq6gZV/VBVm6hqbVW1Tnej\nSIhIdRH5FtimJaTU2rx5c3JycsjJyWHBggVUrVqV888/nyuvvJLhw4ezePFizj//fEaMGFES5hiG\nYRgxSMghEZHDRWSciLzjj1uJyBXFa1rxISK7vBrolyLymte52Nc8e4rIbamwLyLfOiIyQ0TKicho\nb/NiEfksJN8vIrcnmFdC6WKcmysitQLHnUUkbheKiHQUkdGhYy2aUuuhydoaiw8++ICjjjqKRo0a\n8e2339Kpk2so6tatG1OmTEl1cYZhGEYRSFQ6fjxuUGNIc/tb4PqYqdOfbeoUQdsAv+P0OsKISJFX\nMlbVN1V1eKoMDHAW7tr3xl3/dqraFjgf2OjTJOpoJO2QJIOqzlfV6wpJE1WpVVXXpdqeSZMm0adP\nHwBat27NtGlOWPa1115jxYoVqS7OMAzDKAKJOiS1VHUysBtAVfOBXcVmVckyG2jq//HPFpE3ga9F\n5CARed63RiwUkS4AIjJPRMJy7H5Nl44i0l9EHvdh431rxhwR+V4Ca8GIyK0+z0UiMtyHHSUi00Vk\ngbehRcC+s4B3cIM8V6lq6Dv4SVU3+Dyq+FaFl3x+U31eX3kVU2Kku1TcejE5IvK0H0BaZETkOBGZ\n66/THBFp7sPDrSgiMlREnvPX63sRuS5w/l72+vA8EbnfX6t5InJ4MvYB/P7777z55ptcdJFrnHnu\nued44oknOOaYY9i8eTOVKlVKNmvDMAwjBcSVjg8ncjoSfwbeU9UO4tYUeUhVTy1m+4oFEclT1Qzf\nEjIFNyvkG+C/QBtVXS4iNwGtVXWAdxBmAM2Aa4HqqnqPiNQBslW1ubj1fTqq6iBxcuvVcK0aLYA3\nVbWpiJyNGxh8up8SW1NV14vIB8BAVV0mIscDD6rqad5BWKCqWSJSHyefvhGnqvqiqi4M1idQv1C+\nVXAaIqeq6rpgOhFpiZOQv0BVd4rIE8A8VZ0Y45rl4pYMCDmiGcASVe0hTsdkq6rmi8jpOIn3P4vT\nIxni0wzFibJ1AQ7Gzb45wpcdy14Feqrq/4nIw8AmVf1nFNtiSseH+Oijj5g2bVrUsSIrVqzggQce\n4Mknn4xW9ZRxIElRp5p0ts9sS450tg3S274DXTq+A05w6zf/+S2u66DUpWaT2XAv1Ry/PQZUws0a\nmRVI8wZucbfQ8WycbkY94CsfNhi43+/3Bx7XPXLrlwTO3ew/R+LWiAnakgFsC9iTA3zj404Cng6k\nrYybvTICWA909eF5EXkOxU3ZXeS/sxMi0wGDgJ8DZS4Fhsa5Zrm4lrLQcWfcGjfglFjfwImWLcY5\nKpFphgJ3BM7/BqhfiL072OM09wbGFvbdRkrHh+jdu7c+99xz4ePVq1erququXbv0sssu03Hjxmlx\ncyBJUaeadLbPbEuOdLZNNb3tOyCl40Wkoar+qKqfi8ipQHOcPPhSVd0Z79w0Z5tGqJKKk0zfUtiJ\nqrpSRNaJSDvcS3JgjKTB5WMlRhpw3WaxVFLPxrXehMregeu+eUdEVuMWPiwwW0Xiq6QWSApMUNV/\nxLEtUe7DOXPni0hjnCJsNILXZBdQoRB7d/ofSzh9MsZt2bKF9957j6effjoc9sorrzBmzBgALrjg\nAi6//PJksjYMwzBSRGFjSKYG9l9V1a9U9cv93BlJlNk4sTBEpBluwbqlPu5V4BYgU51eRqK8B1wu\nflaP76rYBCwXkYt8mIhIe5++K05wDBHpIG41X0SkHK615gefbqc4tVSATGKopEak+wC4UERqh2wR\nkUZFqEuQTNw6OeBaiop6bix7U0K1atVYt24dmZl7unAGDx7Mt99+y7fffsvw4cNDDqlhGIZRShTm\nkASf0k2K05A05AmgcXfAIAAAIABJREFUnIgsxjkg/X0LBcDrwMU4xdGEUdXpOGG5+SKSg5N4B+f4\nXCEii3CqqeeKk07frntk2msD/yciXwJfAPnA4z7uGeALP1h1Oq7l4RtgOAVVUsPpVPVr4E5ghoh8\ngXOW6hSlPgEeBh4UkYUUvRUjnr1FpkrF8uQOP4dcU2s1DMPYryjs5aEx9vdrNDAANBCWTaCrQVW3\nA1Hb8VV1NRHXTlXH48aOoAF108jy1E0NHh4Rv5dKqohcihtIG0oznUD3TcT5twK3BoKiqqRGplPV\nV3HOVqGoauOI42z89VLVubgBvyHujJJmaMT5weV1Y9kbvG6v4xzBuISUWs0hMQzD2L8orIWkvYhs\nEpHNQDu/v0lENovIppIw8EBFVV/U4tE1OWBYunQpWVlZ4e2QQw5h1KhRvPbaa7Ru3Zpy5coxf/78\n0jbTMAzDoJAWElVNSpfC2H8RkU9ws3mCXKaqi0vDnn0hJBsPsGvXLurVq8f555/P1q1b+c9//sM1\n11xTyhYahmEYIRIVRttvEZOJLxKqejwwGgitslsBP37Ii5qVylx1ERkrIq2SPT8oG9+yZUuaN2+e\nSvMMwzCMfaTMOySYTHyR8AJsdwAnq2o73KyXoswkKhZU9Uo/EDcpgrLxhmEYRvqRlK7Dfsxs3FiY\nzjjtjA1AC68p8iTQETd75UZVnSUi84ArVPUrCCvWDgHaUFCVdZM/9wjgFj8AExG5FbgUJ7n/jqre\nJiJHAWOAw4CtOKG0Jd6+s4B7/WcBmXifX1j+HSfOdomITMUJkx0EPKqqz8RIdylwHU4E7hPgr6oa\nTf6/Nk6RNc+XnRfa91zkVV2r+2sz22uPvIBTpwUYpKpzRCQDmAbUACoCd6rqNJ/+HZzy7Em4KcPn\nAjuBucDNqpotIg8Cu1X1jtC1V9W9Bn1EKLVyd9t8srOzw/E7d+5kypQp9OjRo0D4xo0bWbBgAXl5\neZQEeXl5BcpPJ9LZNkhv+8y25Ehn2yC97Utn2/aJ0lZmK+4Nr06Kc76m4aTfO+NE0I70cTcBz/n9\nFsCPuBf8Df+/vXOPt6qq9vj3x0tRFETxTSJdjfCFYppdM9AyS3t4M7EsxfCWeq+W5TPL0G73es2u\nJma+UtJLKWqCejMfPNI0H6AHFJO0ALUU0dTER6CO+8cY+5x1Nvu8D+esA+P7+awPc80195xjrbV1\njzPnGL8JnBX1W+CCcLCqKuv1+GzTSOCpqP8EcB+wXpwPjn9nANtFeU9gZpR7A3VR3hpXRq3D1V13\nrb6fwnml3/64UurG1e2A9wO3AH3j/GLgiCaeV298luZp4CrgU4Vrs4EfRfmTwF1RXg9YN8rbEap/\n8cw3jPImwFN4Kvkw3PEbFdemAl+K8g64iutHgUeAfoWxd2/pfVeUWotMmzbNPvaxj1k1H/nIR+yh\nhx5apX51sTYpP3Y2ZbYvbWsfZbbNrNz2rZVKrWsIlZkC8BmSn+F/lT9onm4LsDcuIY+ZPSFpCZ7G\nOhVPvf0ecChNp51OM5/NeLywAdxHgavM7I3o928xY/Ah4PqCEFclgHRPfOYCM3tWvkHdvnHMkPR5\nM2ukyhqcIOngKA/FHYLqnXL3A0YDD8W4/YEXat2Imb0j6QDgA/G58yWNtoa03V/Fv3NxxwJ89uMi\nSaNwRdVKCrCA/5S0Dz5LtBVQeT6LzKyuui8zWyDpGuBWXL11RS0728Ivf/nLXK5JkiQpOWuDQ5Iy\n8W2UiQ+P+UHgQUl34jMlE+Ny5V6LUu4nAkuBXeIe34r6w/GlqdHmm+gtLthXLSPfv3C+Ex4zs2lr\n7G2OWrLxN910E8cffzzLli3jwAMPZNSoUdx+++0dHSpJkiTpAGtDUGtrSJn4QNKWknYrVI0qjN0U\nA2mIefkyvuxTqX8hnJGxQIvS9JL+BRgM7ANMkjSopc80Ry3Z+IMPPphnn32Wf/zjHyxdujSdkSRJ\nkhKQDomTMvEN9AXOk/RE2D0O39W4OS4Gjox7GkHD7NMUYPd4rkcATzTxeQAkbRL3cbSZ/THu+cct\njN2IinR8kiRJ0rNY45dsLGXiK+etkok3syV43Eqta2MK5RdpiPt4Ep/FqXBqoc1eTQxVLx1vZucV\n6rcv1F9Ya+zmqJaOf+WVVzj66KN57LHHkMSVV17JBRdcwMKFC+uvDxo0qF5ALUmSJOke1niHpCdg\nZv/b3TasqXz961/ngAMO4IYbbmDFihW88cYbXHddg1/2rW99q9FyTpIkSdI9rLYlm1RIbZcdi2PZ\nokuQ9ICkFZKWx7uqk7RT5b11oR0TJZ0U5bMlfbQz+n311Ve5++67mTBhAgD9+vVj0KCGkBQzY+rU\nqZmBkyRJUgJWZwxJKqSWjOpnbi4T/1dcH+RTkf3zdnfYVrDpTDO7qzP6WrRoEUOGDOGoo45i1113\n5eijj+b11xuSq+655x4222wztttuu84YLkmSJOkA8gzP1dCxtLwSTyHpGDzGYCoFhdSo63aFVEnX\n0aCQuq2ZHV91L+cAJwOP0rJCanW71iqkEmmxuwMDqKFkamZvxrN4ABhLY7XU3ni8yhhc2+QnZnap\nqlRpzWz7GmNeBqwws/MknY0HpX7ZzHaUtG4T72g88GlcFO29wE1mdkr0OQGPI3kFmAf8I97ZMOBK\nXCRtGXCUmT0taSIu5HZevNdbzewGSWcCn8JTgu8DvmY1vrBVSq2jz7zgcnbaaiALFy7kuOOOY9Kk\nSYwcOZJJkyax/vrr85WvfAWA888/n6222opDDz201uvodJYvX86AAauENJWCMtsG5bYvbWsfZbYN\nym1fZ9s2duzYuWbWLfuUNWJ1Ka6RCqnQBoXUuL4Y/7EeRtNKprOprZb6VVyaHdwhmQNsW/3Mmxjz\nfcB9cf5IPM/HWnhH44E/46m96+KpwUPxGabFeOpuXzyluvLObgGOjPJXcEE5cI2Tkwrv9ZDiM47y\nNRRUY5s6ikqtzz33nG2zzTZW4e6777ZPfvKTZma2cuVK23TTTe2ZZ56xrmJtUn7sbMpsX9rWPsps\nm1m57VtTlVpX55JNRSF1TvyI/SzqqxVS/xdcIRX/UasopB4SbVpUSDVPa22tQmodcCkNaa+NFFLx\nH+fT8dmVGZL2a2LsEyLN9X4aFFKrKSqk1sX58Cb6q2aR1VAyDWqppe4PHBHjPABsXLCp+Mxr8RLw\nsqTDcNn2NwrXmnpHADPM7FXzLKXHcZ2RPYDfmtnfzGwl7jRW2Av4RZSvib6bY2zEuTyKZ/7s0EL7\nRmy++eYMHTq0PqNmxowZjBzpGwbfddddjBgxgq233rotXSZJkiSridWZZZMKqW1USK2iOSXTWmqp\nAo43s0YqX2Fri88cTwn+CT7z0V4bO+37FEtFF+NLdM/Esk6tZ9wskyZN4vDDD2fFihUMHz6cq666\nCsjdf5MkScpGdwujpUJq53E7cGxlbEnbS1q/hc8UuQk4N/op0tw7qsVDwEckbRRBtJ8rXLsPF5kj\n+rynmX4qzseLMcN1SDNtm2TUqFHMmTOH+fPnM23aNDbaaCMAJk+ezDHHNOXnJkmSJF1NdzskqZDa\neVyBL5s8HPZfShtmLMzsNTP7b1t1M7vm3lGtfv4C/Ce+F869eDzJq3H5eNxZnI9LzDepAGtmrwCX\n4/E5t+OOToukUmuSJEkPpbuDWLrzwDNxTutuO9a0AxgQ//bBA1kP7qqxi0GtZWNtCpLrbMpsX9rW\nPspsm1m57VtTg1rXaqVWS4XU1cXEEDdbF5fEn9ZdhqR0fJIkSc9grXZIWoukzYELgA/g2hpLgW+Y\nbwDXnv4ewFNzKwwAJrbHQZK0BfBzPO33D3h8Rz88u2mCeaZLW/vcHU9PPqGtnwUws5Oq64q6NF1J\nSscnSZL0DNIhaQF5atBNeLbMYVG3C55m3C6HxFwhtTjGRFzcrdb4fcysOfXUisoswJ/MbFSIpN2J\np0xPaYd9c3CHZrXSinvrEBXp+MmTJwMuHd+vX7/662YuHT9z5szVZUKSJEnSSro7qLUnMBZYaWaX\nVCrMbB7wO0k/LOx7Mw48zVbSrZW2ki4KVdPKXjVnSXo4PjMi1EuPAU6MPWQ+LGmypEtiJuVcSU9G\nAC7y/XaeqpzjDsltRYPNlWAfBLaKz4yW9FtJcyXdHrMqSPqApPkx7g8jGLbRPURm0LRod3+kYlf2\nn7lS0mxJf5bUqtmU6PseSTfjQbhE/3MlLQjV1UrboyT9UdKDki6XdFGTHdcgpeOTJEl6DjlD0jI7\n4gJk1fwLMArYBVdXfUjS3a3o70Uz203Scbg66dGSLiGk06Feen1r4ENm9o6kV/EsoQtw/ZN5ZrYs\nZkLeZ2aPh2NDfH5dXPDt65EGPAmXnl8WjtMPcKXUq3AJ/d/LZe9rcRbwiJl9VtK+wNVx3+DKrWOB\nDYCFkn7ayiWi3YAdrUGs7SvmAnb98ed4I77sdBYuLPcqMAtXkV0FNZaO58yd3mb27NksXLiQuXPn\nMn78eMaPH8+kSZM49thjG0nH77HHHsyePbsVJnec5cuXd9lYbaXMtkG57Uvb2keZbYNy21dm2zpE\nd0fVlv3A96E5v0b9+fgPaeX8GnxvlzH4XiyV+ovwVFnwFNitorwnDbLvEwnp9DifTEisx/lQ4OEo\nXwscFOUPAZdGeRjwJi57/yrwi6jfEd/zpy6OR/FA00HAksIYO9MgF19/D7gTMLzQ7hlgw7D5jEL9\nH4Ctm3mOywt9z6q6NhHf82Ze2P5BXJDu6qr3cFFL7yul49tHmW0zK7d9aVv7KLNtZuW2b03Nsskl\nm5ZZgP+V3lrepvFSWLW6aC2V1VrUry2Y2TPA0pih2IOGJZpGKrNEDAm+2d1oSZ/GFVwXmO+8PMrM\ndjKz/dtwP83RXqXW+ntTY9XbXXAHqM2KrLVI6fgkSZKeQzokLTMTWKcqtmFnPNtmnKTeEc+xDx63\nsQQYKWkdSYNwJdiWeA1f9miOK/A9Za63ht2C61Vmi5jZi8Bp+J48C4EhkvYK2/tK2sFceOw1SZUA\n28Oq+wmKSq1j8CWnv7finlpLU6q3D+CKrxvHstPn29N5RTp+5513pq6ujm9/+9tASscnSZKUjYwh\naQEzM0kHAxdIOhV4C196+QaerjsPMOAUM3seQNJUXGF0EU3EPVRxC3CDpM/gaqa1uBmP+bgqxqhW\nma1mGr4Usicuu36hpIH4O78An/mZAFwu6V3gtzQoqhaZCFwZ6qpvAEe24n7awm+AY0L1diGhemtm\nz0X20e9x569VQiH9+/ZmYUGptSIdX00l8yZJkiQpB+mQtAIz+yueQlvNyXFUtz8F34enun5YoTwH\nj6fAXM9k50LTWnu87IIHsz4R5x/HY0Eq/S3G40Uq5xafqbBPjT4XmFkla+Y0ItXXzGYDs6P8Nzye\no/peJlad71jdpur6gOq+4/wf+NJTrc8UHbDxwO7NjQHw5sp3GHba/6V8fJIkSQ8jHZIeQDgLxxJL\nJ9BpKrMHSjod/x4soW07/fYIUqk1SZKkZ5AOSQ/AzM7BN/Hr7H6vwzfM6xQkbYzvcFzNfmb2Ukf6\nNrPJePZRm0il1iRJkp5BaYNaJb0Tgl2PSbpe0nqd0OenY7ahU5G0haQ7QrTswoJY2kOSto02325l\nX61q18RnB0j6qaQ/hfjaXEn/2t7+2oqZvVTI5hmFx6r8rqPOSHupKLVOmDABcKXWQYMGFe1l6tSp\nGdyaJElSAkrrkABvxg/bjsAKXM20Hkltnt0xs5tjtqGzqci3jwO2BHY2s52Ag/GATIDWOhrtdkjw\nTJyXge3MbLewa3AH+uvRpFJrkiRJz0Ee+1g+VNiMTdIxeNDnVOD7+I/uiKj7KR7s+DbwTTObJel+\nfGO5BfH52cBJeNDn7mb275Im44Jhu+P7yJxiZjdE+1OBLwHvAreZ2WmS3gv8BBiCZ5v8ayXAVNJ1\nuKroAcC2ZtYoUyZUUE/GRckWmNnhkqbhgmfrAj82s8uaaPclXBSsH54Ke1wh7bc4xnvx/Wv+ycze\nrXF9DC6+dlCcX4SL4UyWdCbwKaA/cB/wtcgumh1jjsWF1CaY2T0xWzU5nudC3An7NzObI+koPN34\nFTwD6R/xvIcBV+KqtsuAo8zs6abeg6ReuKjcvrgY20rgyso7qrq3olLr6DMvuJydthrIwoULOe64\n45g0aRIjR45k0qRJrL/++o2UWrfaaisOPbRWvHLns3z5cgYM6PL9BVtFmW2DctuXtrWPMtsG5bav\ns20bO3bsXDNrMWlgtdPdymxNHTQoe/YBpuNBnWNwUa1t49q38B8pcAflafwH/kTgrKjfAlgY5fGE\n2if+g3o9Pks0Engq6j+B/yivF+eD498Z+MwDeCrtzCj3BuqivDWeElwH/AjYtfp+CueVfvvjKcIb\nV7cD3o+nBPeN84vxXXhrPa9PAzc18zzH0LSC7OBC/TXAp6I8G/hRlD9Jg7LsSTQoxO6IO4O7x7N+\nGnfa+gH3Fp73LYT6LC5bP62F93AI8Ouo3xx3Qg9p6XuTSq3to8y2mZXbvrStfZTZNrNy25dKrV1P\nf0l1eCrq08DPov5Ba9gDZW9cLAzz2YolwPb4TMoh0eZQYJW/qoNpZvaumT2O794Lrhp6lZm9Ef3+\nTdIAXKb9+rDpUvzHF9w5eSDaPgu8D58heBeYIakpYbQTJM3DdTeGArXWDfbDVWIfinH3A4Y30V8j\nJJ0RMTh/bUXzsZIekPQoPiOxQ+Har+Lfubg8PfhzvxbAzB4D5kf9nsBsM1tmZitoHDC7F/CLKF8T\nfVSo9R72xkXg3jXXd5nVivtoRCq1JkmS9BzKnGXzpnlgZD2SoCA73hRm9hdJL4Wi6jiq4k8KFKXP\n1UyXvYBXqu0JGsm3m+tq3AbcJmkpruHRKPOkSi79jVgaqSWXLuDnZnZ6M7ZVeBzYRVKv+BH/AfAD\nScvjek1J+9iI72J8KeuZECMr2tJaqfuO0Nr30GYqSq0rVqxg+PDhXHXVVUAqtSZJkpSNMs+QtIai\nrPn2wHvwmAbwv85PAQaa2fzaH6/JncBRlaweSYPNpdIXSfp81ElSRXSsXr5d0m6StoxyLzzGZUm0\nWxkS6NC0XHp1uxnAIZI2rdgiaZtaRpvZU/hs0n/IdwGuOBuVH/imJO0rzseLMRN0CC1zLyEUJ2kk\nsFPUNyf3fh8N8vSHU1v8rXqMz0Xm0maEiFxbqSi1zp8/n2nTprHRRhsBrtR6zDFN+alJkiRJV1Pm\nGZLWcDHw01hqeBuPiaj8tX0D8GM8CLbVmNlvJI0C5khagccxfBv/Ef2ppO8AfYFrYzmkKN++KS7F\nvk6cP4jHagBcBsyX9DAeQ7GKXHp1O/Og1u8Ad4SDsxL4NxqcnGqOBn4IPCXpJXz331Pivp6pJWlv\nZq9IujzqnwceasVjuhj4uaTHgSdwGfpXrXm59+OBqySdTAS1tjDGjbjT9Dge1PowtaXtG1EtHZ8k\nSZL0DErrkFhk2FTVzaax7PhbNPHDZmZLqbo/K4hrmdn4psazGkJkEbdyQLEuMmCK8u2/ofHuu8XP\nnwqcWqhqSi69UTtrg3hZzOR8rZnrTUnafwf4To36MYXyizTEkLwFfMnM3orsnrsIJ8kKcu9VfS3B\n41Oq68dXnVck5t+VdJKZLQ/BtQfx7KNmqZaOr6XU+utf/5rp06fTq1cvNt10UyZPnsyWW27ZUtdJ\nkiTJaqS0DklPwDpHvr0nsh4wK5ZlhKcir1gN49way0v9gO9HcGubqKXUusMOO/D97/vE2YUXXsjZ\nZ5/NJZdc0smmJ0mSJG2hp8eQdIgerAb7vKQ343hD0uOSduoiNdjFeDxIH3yX42PM7Lb29tccZjbG\nXBxvZMxutYmmlFo33HDD+javv/56JVg6SZIk6UbWaoeEnqsG+ztgfTPrj6c5/7OZPUrXqMECjLUG\nifj7ihcqAbVloDml1jPOOIOhQ4cyZcoUzj777G62NEmSJCmtUmtXkGqwbVODjXEWx/29WHyOuDbL\nR/Gg231pm/Jrb+C/497eBS43s0mSRgP/AwwAXsSDlp+rYVO7lFoBpkyZwooVKzjqqJZibDvO2qT8\n2NmU2b60rX2U2TYot32p1LoGHqQaLLRBDTauL8admTrggagz4NDqcaPcGuXXY/GsqD6Vz+OZTPcB\nQ6JuXOU9NHe0Vqm1wpIlS2yHHXawrmBtUn7sbMpsX9rWPspsm1m57Uul1jWTVINtnxpsZclmzzh/\nB0/Trb/eRuXXj+JS9G/HPf4t7nFH4M6w6zu4M9ZqmlJqffLJJ+vbTJ8+nREjRrSl2yRJkmQ1sLZn\n2aQabNvUYJviLYslnk5UfhW+pLRXB+yqqdR69NFHs3DhQnr16sU222yTGTZJkiQlYG13SFpDRQ12\nZierwZ4paUo4C4NjlmSRpM+b2fVyz2hnM5uHz1qcC64GCzxvZn8tqMFWxl4pqa+ZraQVarDRbgYw\nXdL5ZvaCpMHABua6Ie2hlvJrU7NHxefxNUmzzOztsGEhMETSXmb2+0gx3t4iZqe1VJRai9x4441N\ntE6SJEm6i7V9yaY1XAz0iuWH61hVDfYwfPmm1ZgLqN2Mq8HW4cGw4I7PhFhmWQB8RtIQVlWDvUVS\nZVO7t1lVDXYKPqPSJ9Rgz6G2GuyUWEqqqMHOx52DLWgnZvYKUFF+vZ3WKb9egS+ZzY97/6K5rskh\nwH9HXR2+pNUs/fv2rhdFS5IkSXoOa/UMiaUabOW8LWqww2rUDag6b5Pya8SOfDOOYvs6YJ/W2FWh\nWqk1SZIk6Rms1Q5JT8DWXjXYTmHYsGFssMEG9O7dmz59+jBnzhzGjRtXH+j6yiuvMGjQIOrq6lro\nKUmSJFmdpEPSBJI2By4APoBvFLcU+IaZ/bGT+h8DrLAqYbFWfnYL4Oe49kZlg75+eLbQhIgNaWuf\nu+PpvifE+QPAOlXNvmwuwNbaPpeb2QBJw4APmdkvWmg/DLjVXKiu05g1axabbLJJ/fl11zVMBn3r\nW99i4MCBnTlckiRJ0g7SIalBBJTehGefHBZ1u+Bpu53ikOB6J8txrY3q8ftUUmCboKLaCvAnMxsV\n4mJ34inIU9pqjJnNwR2ayvmezTRvK8OALwLNOiRdjZkxdepUZs6c2d2mJEmSrPVkUGttxgIrzaw+\nHzSyXX4n6Yex982jksaBz3ZIurXSVtJFksZHebGksyQ9HJ8ZETMBxwAnxl46H5Y0WdIlMTNxrqQn\nI6CV2L/mqco57pA02j8m0m4fBLaKz4yW9FtJcyXdHrMqSPqApPkx7g8jOLbRPUgaLGlatLs/UpuR\nNFHSlZJmS/qzpBNa+TzPAT4cY54oaZike+KZPCxplWBVSXdLGlU4/104hW1CEvvvvz+jR4/msssu\na3TtnnvuYbPNNmO77WrJsyRJkiRdSc6Q1GZHXLirmn8BRgG7AJvgYmJ3t6K/F81sN0nHASeZ2dGS\nLsEVU88DkDQBF/76kJm9I+lVPOvmAlxPZJ6ZLYuZkPeZ2ePh2BCfXxcXUPt6pMhOAj4TnxkH/AD4\nCnAVLkn/e7mMfC3OAh4xs89K2he4Ou4bXK12LLABsFDST1uxRHRa3PdBYet6wMfM7C1J2wG/xOX1\ni/wMV739RqRbrxtO4SqosXQ8Z+70NrNnzwbg3HPPZciQIbz88sucdNJJvPnmm+yyi/s1559/Pnvs\nsUd929XN8uXLu2ystlJm26Dc9qVt7aPMtkG57SuzbR0hHZK2sTfwy5iNWCrpt3iMyd9b+FxRnfRf\nmml3vTXsIXMlLmd/AQ2OBBRUW4P3RurwtsD/mdl8STvSoHIKLj3/nKRBuMbI7+OzvwAOauI+Pwdg\nZjMlbSypskXu/0Xa8z8kvYAvYz3bwv1X0xe4KGZA3sGVb6u5HviupJPj/ic31ZmZXYanMvOe4f9k\nP3q0D4sPH7NKu3nz5rFy5UrGjBnD22+/zbhx45g7dy5bb90mAdh2M3v2bMaMWdWuMlBm26Dc9qVt\n7aPMtkG57SuzbR0hl2xqswCXU28tb9P4WVYrorZWnbReIdbMnsGdnn2BPWhYommk2krEkADvBUZL\n+jQNKqeVHXl3MrP923A/zVFUnm3pfpriRDxIeBd8ZqRfdYOQ1b8T+AztjIt5/fXXee211+rLd9xx\nBzvu6PGyd911FyNGjOgyZyRJkiRpnnRIajMTWCeWAgCIOIpXgHGSekc8xz543MYSYKSkdWIWoqm9\nZYq8hi97NMcV+D46xZmT/YC7qhuGrsdp+B439SqnYXtfSTuEaNlrkioBq4c1MW5FnbaSDfSimbU0\nC9Qc1fc6EHjOzN4FvozP4NTiCuBC4CEze7mtgy5dupS9996bXXbZhT322IMDDzyQAw5wmZdrr72W\nL3zhC23tMkmSJFlN5JJNDczMJB0MXCDpVOAtfJfbbwADgHn4DrenmNnzAJKm4uqki4BHWjHMLcAN\nkj4DHN9Em5vxpZqrYoxq1dZqpgET8WWdQ4ALJQ3E3/MF+MzPBOBySe8CvwVerdHPRODKUG59Aziy\nFffTHPOBd0JxdTKufnujpCPw2Z6aeweZ2VxJf6dhuapF+vftzcIQRRs+fDjz5tUMO2Hy5MltMD9J\nkiRZ3aRD0gRm9ld8qaCak+Oobn8Kvq9Ndf2wQnkOnu5L6JnsXGh6T42xdsGDWZ+I84/TWLV1MR4r\nUjm3+EyFWiqnC8yskjVzGpHqW1Sojd12P1vjXiZWnTerF1JRcI2g132rLhfv/dRa9yNpS3wW7w5a\nSSq1JkmS9EzSISkp4SwcSyydQKepth4o6XT83S/BM1lKR8ye/AD4ZizttItaSq3f/e53mT59Or16\n9WLTTTdl8uTJbLnllp1nfJIkSdJmMoakpJjZOWa2jZn9rpP7vS4CXXc0swPNbFlH+osMnLoax8Yd\ntPNqMxtqZtd3pB9wpda6urr6XX9PPvlk5s+fT11dHQcddBBnn312R4dIkiRJOsga5ZBI2lzStZL+\nFIJgvw4Ni86MCHeUAAASY0lEQVTqf0wtEa9WfnYLSXeEKNib8aP9uKSrQzekPX3uLunC9ny2mT7f\nCdvmNSVaVsTMXgL+qZDRUzlequr3CkkjO9PW9rLhhhvWl19//XUiNTpJkiTpRtaYJRsp5d47iTcj\njRhJHwf+C/hIRzs1s6M72kd7qCi1SuJrX/saX/2qJ06dccYZXH311QwcOJBZs2Z1h2lJkiRJAXkc\nZM8n9Dommtk+VfUCzsX1Owz4DzO7LtJZi+qhFwFzzGyypMX45nWfwkW8Po9n2tyPa28swzNjJkT9\nrsC90f5DoY7aC3eE9orz63AF1DcobCAXaql/M7NzJY0G/gfP5HkRGG9mz0n6AK5c+i7uwHzCzHYs\n3oOkwbiY2vAY46shkjYReE/Uvwe4wMyanFVRbIgX5c8Dh5vZZ+P8ZNx5Wge4ycy+V/xM3PNFeADr\nM8BK4Eozu0HS7LB1TtUYhwAHmdl4SZOBN+N5booLoh0B7AU8YGbjm7C5qNQ6+swLLmenrXzDvGXL\nljVSaj3hhBPqlVoBpkyZwooVKzjqqKOaeiSdxvLlyxkwYMBqH6c9lNk2KLd9aVv7KLNtUG77Otu2\nsWPHzjWzarXsrsfM1ogDOAE4v0b95/Af8d74bMnTwBb4bMethXYX4Q4AeIrv8VE+DrgiyhPxH9XK\nZyYDtwK94/x7+I7AAPsDN0a5N1AX5WHAY1FeF5iFZ5z0xWdehsS1cfiPOXg68V5RPqfw+fp7wKXi\nvxflfQvjTYx+18Hl7l8C+jbzHN8B6oAn8JTg0YX7uQwXXesV971PXFse/x4C/Dqubw68DBwS12YD\nuxfbFz4zufA8r40xPoMr4O4U/c0FRrX0PRi67Xttm1NvtVp873vfsx/+8IeN6pYsWWI77LBDzfad\nzaxZs7pknPZQZtvMym1f2tY+ymybWbnt62zb8D/Gu/13fI2KIWmCerl3M1uKa298oBWfK8q9D2um\nXbXc+xFRbo3c+1JcIGw+8D4a5N7rgO8AWzch916LvYFrwOXegVXk3s3F0ypy703xpnkMyAh8menq\nmGXaP45HgIfxPW2qd6XbO57Hu+b6LO1ZC7kl/gN5FFhqZo+aZ9ksoPn3sApNKbU++eST9W2mT5/O\niBEj2mFmkiRJ0pmsMTEk+A/WIW1ov1rk3iUV5d4rKbs15d4lbQLcG3Lvi3CNkL2KnYdD0lHaJfdu\nvgHfJsAQfNbiv8zs0k6wp7hO2NRzf5fGdr9LG7+vS5cu5eCDDwbg7bff5otf/CIHHHAAn/vc51i4\ncCG9evVim2224ZJLLmmhpyRJkmR1syY5JDOB/5T0VfPN1qrl3n8ODMbFwk7Gl0hGSloH6I9LsreU\nYvsasGELbSpy79dYY7n3c6sbmtmLoTdyOh44OkTSXuEI9AW2N7MFkl6TtKeZPUDLcu/fL8q9dySD\nRNIIfLnpJTwg9/uSppjZcklbASvN7IXCR+4FjoxnPQRfUqo1o7NU0vtxifuD8efa6TSl1HrjjTeu\njuGSJEmSDrDGOCRmKfdO58i9948lI/BZkSPDsbojnIjfh5OzHPgSvgRU4Ubc+XocD2p9uAlbT8Nj\nUJbhWUKdFp1VlI5PkiRJeg5rjEMCKfdO58i9N7XRHWb2Y+DHNeorEvHvSjopZlA2xjcefDSujSm0\nvwG4oUY/4wvlxTR+TuOr29eiWjo+lVqTJEl6BmtDUGuXEc7CjfgSDOBy72Z2Tge7PjDEyh4DPgz8\nRwf7W53cGjMs9wDfr8xGdSep1JokSVJ+eqxDUlAUfUzS9ZLW64Q+Px1ORbuwJuTeCyqtvSRdGDY/\nKukhSdtGm28302+93Dtwr7VT7l3SV2LcBaEW++d4hsslLVQH5d7D1jFh60hgsaRbO9pnZ5NKrUmS\nJOWjxzokNKSn7gisAI4pXpTU5uUoM7u5E2YzalFRaR0HbAnsbGY74QGdr0SbJh2SKlrbrhGStgbO\nAPY2sx3woNP9zFVZ5+ACaC8110dPpKLUOnr0aC677LL6+jPOOIOhQ4cyZcqUnCFJkiQpAT1WqbVK\n7fMYPLZjKvB9XJBrRNT9FNgdT/P9ppnNknQ/MMHMFsTnZwMn4TELu5vZv4dq6N/js5vjwbA3RPtT\n8YDOd4HbzOw0Se8FfoL/0L8B/GsljqSg0noAsK2ZNQqIDbXWk/F4iwVmdrikacBQPC32x2Z2WRPt\nvoSLwvXDtU6OK2T3FMfYDddJGV19Pe7/AWAsMCiezT2ShuHaJutH0383s/sii2ciria7I67V8qUI\nLD4AD8Z9A89aGm7tVJKV9N14zsvwINm5ZnZejXtLpdYOUmbboNz2pW3to8y2QbntS6XWkh00qIP2\nAaYDx+LBp6/jP/oA36JB7XQErtK6LnAicFbUbwEsjPJ44CJrUA29Hp9FGgk8FfWfwJVP14vzwfHv\nDGC7KO8JzLRVVVq3xjN/6oAfAbtW30/hvNJvfzwTaOPqdsD78cyfvnF+MXBEE8+rNz5L8zSeAfSp\nwrXZwI+i/EngriivB6wb5e0INb94zq/G/fQCfo+Loq2LOw7b4Rk6U2mnkiwuXlcXfW4APElBJbep\nI5Va20eZbTMrt31pW/sos21m5bYvlVrLRyU9dQ7+I/uzqH/QzBZFeW9cEwTz2YolwPb4D2VFRO1Q\namR8BNPMVUcfp0Hd9KPAVWb2RvT7N0kDgA8B14dNl+KODhRUWs3sWVyR9XR8dmWGpP2aGPsESfPw\n/XOGsqoqKniK7WjgoRh3P3ymYRXMZ0UOiPv+I3B+zE5UqKVM2xdPN34Ud86Ku/U+aGbPmquo1sVn\nRgCLzOzJ+JL/b6F9W5Vk/xmYbmaVlOlbaj6lZkil1iRJkp5DT077rd+VtkIEJ75eu3kDZvYXSS+F\ncNo4quJPChSVQpuLfOwFvFJtT9BIpdXM/gHcBtwmaSmeqjuj6j7G4I7PXmb2RiypVCuaVmz6uZmd\nXuPaKoST8CDwoKQ78ZmSiXG5ljLtibi8/S5xj28VumuX+msTdGZf9aRSa5IkSc+hJzskraGiXjpT\n0vZ4jMLCuHYdrkEy0HwvmdZyJ3BmKJa+IWlwzJIskvR5M7s+9n7Z2czmUVBpjTiO583sr7Ez7s5A\nZeyVkvqa2UpgIPBy9D8C+GBh/GK7GcB0Seeb2QsRp7GBmS2pNlrSlsDmZvZwVI3CZ4yaYyDwrLm+\nyJH4sk9zPAEMk/ReM/sT8IXCtbYqyd4LXCrpv/Dv6UH45n6tJpVakyRJeg49ecmmNVwM9Iolh+vw\n3Xwrf43fgMuwT21Lh2b2G1yNdU4sk5wUlw4HJsQyywLgMzVUWjcFbgk9kfl4oO1Fce0yYL6kKfiM\nSh9Jf8B3972/YEJ9u1hK+g6uojofd5a2oDZ9gfMkPRF2jwO+3sLtXoxLwc/Dl2OanX0ys7fw4NL/\nk/QwjVVcJwKjw85zaEFJ1swewp/zfHxG6VFqq742on/f3vWiaEmSJEnPocfOkFhk2FTVzSaUS+P8\nLaBm+oT5zr99quom48GsqyiDFsczTw0+p+r6IjxGo57IgCmqtP6GxpvsFT9/KnBqoeoTrWlnZtfh\nzlazxKzJvk1cG1Mov0jEkJjZkzRWpj016mfT+Dn/e6H8G9x5qR6jPUqy55nZxNCYuRuPb0mSJEnW\nQHqsQ9ITMLP/bblV0gyXSRqJx8/8vLDclCRJkqxhpEOyBiLpATyNtsiXzezR7rCnvZjZF7vbhiRJ\nkqRrSIdkDcTM9uxuG5IkSZKkLazpQa1JkiRJkvQAeqx0fJLUQtJrNKR2l41NcLn9MlJm26Dc9qVt\n7aPMtkG57ets27YxsyGd2F+7yCWbZE1joZVhT4YaSJqTtrWPMtuXtrWPMtsG5bavzLZ1hFyySZIk\nSZKk20mHJEmSJEmSbicdkmRNo03y8l1M2tZ+ymxf2tY+ymwblNu+MtvWbjKoNUmSJEmSbidnSJIk\nSZIk6XbSIUmSJEmSpNtJhyRZI5B0gKSFkp6SdFoXjXmlpBdi9+ZK3WBJd0p6Mv7dKOol6cKwb76k\n3QqfOTLaPymp2V2Q22DbUEmzJD0uaYGkr5fMvnUlPShpXth3VtRvK+mBsOM6Sf2ifp04fyquDyv0\ndXrUL5T08c6wL/rtLekRSbeWyTZJiyU9KqlO0pyoK8V7jX4HSbohdhb/g6S9ymCfpPfFM6scf5f0\njTLYFn2eGP8tPCbpl/HfSCm+c12GmeWRR48+gN7An4DhQD9gHjCyC8bdB9gNeKxQdy5wWpRPA/47\nyp8EbgMEfBB4IOoHA3+OfzeK8kadYNsWwG5R3gD4IzCyRPYJGBDlvsADMe5U4LCovwQ4NsrHAZdE\n+TDguiiPjPe9DrBtfA96d9L7/SbwC+DWOC+FbcBiYJOqulK81+j758DRUe4HDCqTfdF/b+B5YJsy\n2AZsBSwC+he+a+PL8p3rqqPbDcgjj44ewF7A7YXz04HTu2jsYTR2SBYCW0R5C1yoDeBS4AvV7YAv\nAJcW6hu160Q7pwMfK6N9wHrAw8CeuPpkn+r3CtwO7BXlPtFO1e+62K6DNm0NzAD2BW6Nscpi22JW\ndUhK8V6BgfgPq8poX6G//YF7y2Ib7pA8gzs5feI79/GyfOe66sglm2RNoPIfc4Vno6472MzMnovy\n88BmUW7KxtVue0zn7orPQpTGvlgSqQNeAO7E/5p7xczerjFWvR1x/VVg49Vo3wXAKcC7cb5xiWwz\n4A5JcyV9NerK8l63BZYBV8Vy1xWS1i+RfRUOA34Z5W63zcz+ApwHPA08h3+H5lKe71yXkA5Jkqwm\nzP9E6da8ekkDgBuBb5jZ34vXuts+M3vHzEbhsxF7ACO6y5Yikg4CXjCzud1tSxPsbWa7AZ8A/k3S\nPsWL3fxe++DLmD81s12B1/FlkHq6+3sXcRifBq6vvtZdtkXcymdwh25LYH3ggK62o7tJhyRZE/gL\nMLRwvnXUdQdLJW0BEP++EPVN2bjabJfUF3dGppjZr8pmXwUzewWYhU9JD5JU2WOrOFa9HXF9IPDS\narLvn4FPS1oMXIsv2/y4JLZV/prGzF4AbsKdubK812eBZ83sgTi/AXdQymIfuCP3sJktjfMy2PZR\nYJGZLTOzlcCv8O9hKb5zXUU6JMmawEPAdhGR3g+fjr25m2y5GahE3R+Jx25U6o+IyP0PAq/GNPHt\nwP6SNoq/kvaPug4hScDPgD+Y2f+U0L4hkgZFuT8e3/IH3DE5pAn7KnYfAsyMv2ZvBg6LrINtge2A\nBztim5mdbmZbm9kw/Ls008wOL4NtktaXtEGljL+PxyjJezWz54FnJL0vqvYDHi+LfcEXaFiuqdjQ\n3bY9DXxQ0nrx327luXX7d65L6e4gljzy6IwDj4j/Ix6HcEYXjflLfL13Jf6X4QR8HXcG8CRwFzA4\n2gr4Sdj3KLB7oZ+vAE/FcVQn2bY3PvU8H6iL45Mlsm9n4JGw7zHgzKgfjv8P9Cl8Sn2dqF83zp+K\n68MLfZ0Rdi8EPtHJ73gMDVk23W5b2DAvjgWV73pZ3mv0OwqYE+92Gp6JUgr78KWQl4CBhbqy2HYW\n8ET893ANninT7d+5rjxSOj5JkiRJkm4nl2ySJEmSJOl20iFJkiRJkqTbSYckSZIkSZJuJx2SJEmS\nJEm6nXRIkiRJkiTpdvq03CRJkmTtRtI7eOpnhc+a2eJuMidJ1kgy7TdJkqQFJC03swFdOF4fa9jD\nJEnWCnLJJkmSpINI2kLS3ZLqJD0m6cNRf4CkhyXNkzQj6gZLmiZpvqT7Je0c9RMlXSPpXuCaULO9\nUdJDcfxzN95ikqx2cskmSZKkZfrHzsTge44cXHX9i/jW8D+Q1BtYT9IQ4HJgHzNbJGlwtD0LeMTM\nPitpX+BqXN0UYCS+ed6bkn4BnG9mv5P0Hlye/P2r8R6TpFtJhyRJkqRl3jTfmbgpHgKujA0Np5lZ\nnaQxwN1mtgjAzP4WbfcGPhd1MyVtLGnDuHazmb0Z5Y8CI31rEwA2lDTAzJZ33m0lSXlIhyRJkqSD\nmNndkvYBDgQmS/of4OV2dPV6odwL+KCZvdUZNiZJ2ckYkiRJkg4iaRtgqZldDlwB7AbcD+wTu65S\nWLK5Bzg86sYAL5rZ32t0ewdwfGGM5mZokqTHkzMkSZIkHWcMcLKklcBy4AgzWybpq8CvJPUCXgA+\nBkzEl3fmA2/QsI18NScAP4l2fYC7gWNW610kSTeSab9JkiRJknQ7uWSTJEmSJEm3kw5JkiRJkiTd\nTjokSZIkSZJ0O+mQJEmSJEnS7aRDkiRJkiRJt5MOSZIkSZIk3U46JEmSJEmSdDv/D3utW+4XYIau\nAAAAAElFTkSuQmCC\n", + "image/png": "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\n", "text/plain": [ - "
" + "
" ] }, "metadata": { - "tags": [] + "tags": [], + "needs_background": "light" } }, { "output_type": "display_data", "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAicAAAEWCAYAAAC9njdIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", + "image/png": "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\n", "text/plain": [ - "
" + "
" ] }, "metadata": { - "tags": [] + "tags": [], + "needs_background": "light" } } ]