merge_datasets.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data for Prediction coffee price task"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import io\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.impute import SimpleImputer\n",
    "import sklearn\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  About the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For price prediction tasks always hard to find an adequate dataset. My task is to collect and preprocess data for such specific tasks.Also create function for filtration data(which will be helpful for people who want to get data and do not have any programming skills)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you have to work with prices you always need to have economic data such a GDP, GDP per capita,  inflation index (CPI) etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>LOCATION</th>\n",
       "      <th>INDICATOR</th>\n",
       "      <th>SUBJECT</th>\n",
       "      <th>MEASURE</th>\n",
       "      <th>FREQUENCY</th>\n",
       "      <th>TIME</th>\n",
       "      <th>Value</th>\n",
       "      <th>Flag Codes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1973-09</td>\n",
       "      <td>6.543496</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1973-10</td>\n",
       "      <td>7.012194</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1973-11</td>\n",
       "      <td>7.966616</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1973-12</td>\n",
       "      <td>7.849054</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1974-01</td>\n",
       "      <td>8.128263</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  LOCATION INDICATOR SUBJECT MEASURE FREQUENCY     TIME     Value Flag Codes\n",
       "0      AUT       CPI     TOT  AGRWTH         M  1973-09  6.543496        NaN\n",
       "1      AUT       CPI     TOT  AGRWTH         M  1973-10  7.012194        NaN\n",
       "2      AUT       CPI     TOT  AGRWTH         M  1973-11  7.966616        NaN\n",
       "3      AUT       CPI     TOT  AGRWTH         M  1973-12  7.849054        NaN\n",
       "4      AUT       CPI     TOT  AGRWTH         M  1974-01  8.128263        NaN"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cpi = pd.read_csv(\"data/data/CPI.csv\")\n",
    "df_cpi.head(5) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Country Name</th>\n",
       "      <th>Country Code</th>\n",
       "      <th>Indicator Name</th>\n",
       "      <th>Indicator Code</th>\n",
       "      <th>1960</th>\n",
       "      <th>1961</th>\n",
       "      <th>1962</th>\n",
       "      <th>1963</th>\n",
       "      <th>1964</th>\n",
       "      <th>1965</th>\n",
       "      <th>...</th>\n",
       "      <th>2008</th>\n",
       "      <th>2009</th>\n",
       "      <th>2010</th>\n",
       "      <th>2011</th>\n",
       "      <th>2012</th>\n",
       "      <th>2013</th>\n",
       "      <th>2014</th>\n",
       "      <th>2015</th>\n",
       "      <th>2016</th>\n",
       "      <th>2017</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Aruba</td>\n",
       "      <td>ABW</td>\n",
       "      <td>GDP (current US$)</td>\n",
       "      <td>NY.GDP.MKTP.CD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>2.791961e+09</td>\n",
       "      <td>2.498933e+09</td>\n",
       "      <td>2.467704e+09</td>\n",
       "      <td>2.584464e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>AFG</td>\n",
       "      <td>GDP (current US$)</td>\n",
       "      <td>NY.GDP.MKTP.CD</td>\n",
       "      <td>5.377778e+08</td>\n",
       "      <td>5.488889e+08</td>\n",
       "      <td>5.466667e+08</td>\n",
       "      <td>7.511112e+08</td>\n",
       "      <td>8.000000e+08</td>\n",
       "      <td>1.006667e+09</td>\n",
       "      <td>...</td>\n",
       "      <td>1.019053e+10</td>\n",
       "      <td>1.248694e+10</td>\n",
       "      <td>1.593680e+10</td>\n",
       "      <td>1.793024e+10</td>\n",
       "      <td>2.053654e+10</td>\n",
       "      <td>2.026425e+10</td>\n",
       "      <td>2.061610e+10</td>\n",
       "      <td>1.921556e+10</td>\n",
       "      <td>1.946902e+10</td>\n",
       "      <td>2.081530e+10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Angola</td>\n",
       "      <td>AGO</td>\n",
       "      <td>GDP (current US$)</td>\n",
       "      <td>NY.GDP.MKTP.CD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>8.417804e+10</td>\n",
       "      <td>7.549239e+10</td>\n",
       "      <td>8.252614e+10</td>\n",
       "      <td>1.041158e+11</td>\n",
       "      <td>1.139232e+11</td>\n",
       "      <td>1.249125e+11</td>\n",
       "      <td>1.267302e+11</td>\n",
       "      <td>1.026212e+11</td>\n",
       "      <td>9.533720e+10</td>\n",
       "      <td>1.242094e+11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Albania</td>\n",
       "      <td>ALB</td>\n",
       "      <td>GDP (current US$)</td>\n",
       "      <td>NY.GDP.MKTP.CD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>1.288135e+10</td>\n",
       "      <td>1.204421e+10</td>\n",
       "      <td>1.192695e+10</td>\n",
       "      <td>1.289087e+10</td>\n",
       "      <td>1.231978e+10</td>\n",
       "      <td>1.277628e+10</td>\n",
       "      <td>1.322824e+10</td>\n",
       "      <td>1.138693e+10</td>\n",
       "      <td>1.188368e+10</td>\n",
       "      <td>1.303935e+10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Andorra</td>\n",
       "      <td>AND</td>\n",
       "      <td>GDP (current US$)</td>\n",
       "      <td>NY.GDP.MKTP.CD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>4.007353e+09</td>\n",
       "      <td>3.660531e+09</td>\n",
       "      <td>3.355695e+09</td>\n",
       "      <td>3.442063e+09</td>\n",
       "      <td>3.164615e+09</td>\n",
       "      <td>3.281585e+09</td>\n",
       "      <td>3.350736e+09</td>\n",
       "      <td>2.811489e+09</td>\n",
       "      <td>2.877312e+09</td>\n",
       "      <td>3.012914e+09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 62 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  Country Name Country Code     Indicator Name  Indicator Code          1960  \\\n",
       "0        Aruba          ABW  GDP (current US$)  NY.GDP.MKTP.CD           NaN   \n",
       "1  Afghanistan          AFG  GDP (current US$)  NY.GDP.MKTP.CD  5.377778e+08   \n",
       "2       Angola          AGO  GDP (current US$)  NY.GDP.MKTP.CD           NaN   \n",
       "3      Albania          ALB  GDP (current US$)  NY.GDP.MKTP.CD           NaN   \n",
       "4      Andorra          AND  GDP (current US$)  NY.GDP.MKTP.CD           NaN   \n",
       "\n",
       "           1961          1962          1963          1964          1965  \\\n",
       "0           NaN           NaN           NaN           NaN           NaN   \n",
       "1  5.488889e+08  5.466667e+08  7.511112e+08  8.000000e+08  1.006667e+09   \n",
       "2           NaN           NaN           NaN           NaN           NaN   \n",
       "3           NaN           NaN           NaN           NaN           NaN   \n",
       "4           NaN           NaN           NaN           NaN           NaN   \n",
       "\n",
       "       ...               2008          2009          2010          2011  \\\n",
       "0      ...       2.791961e+09  2.498933e+09  2.467704e+09  2.584464e+09   \n",
       "1      ...       1.019053e+10  1.248694e+10  1.593680e+10  1.793024e+10   \n",
       "2      ...       8.417804e+10  7.549239e+10  8.252614e+10  1.041158e+11   \n",
       "3      ...       1.288135e+10  1.204421e+10  1.192695e+10  1.289087e+10   \n",
       "4      ...       4.007353e+09  3.660531e+09  3.355695e+09  3.442063e+09   \n",
       "\n",
       "           2012          2013          2014          2015          2016  \\\n",
       "0           NaN           NaN           NaN           NaN           NaN   \n",
       "1  2.053654e+10  2.026425e+10  2.061610e+10  1.921556e+10  1.946902e+10   \n",
       "2  1.139232e+11  1.249125e+11  1.267302e+11  1.026212e+11  9.533720e+10   \n",
       "3  1.231978e+10  1.277628e+10  1.322824e+10  1.138693e+10  1.188368e+10   \n",
       "4  3.164615e+09  3.281585e+09  3.350736e+09  2.811489e+09  2.877312e+09   \n",
       "\n",
       "           2017  \n",
       "0           NaN  \n",
       "1  2.081530e+10  \n",
       "2  1.242094e+11  \n",
       "3  1.303935e+10  \n",
       "4  3.012914e+09  \n",
       "\n",
       "[5 rows x 62 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_gdp = pd.read_csv(\"data/data/GDP.csv\")\n",
    "df_gdp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Country Name</th>\n",
       "      <th>Country Code</th>\n",
       "      <th>Indicator Name</th>\n",
       "      <th>Indicator Code</th>\n",
       "      <th>1960</th>\n",
       "      <th>1961</th>\n",
       "      <th>1962</th>\n",
       "      <th>1963</th>\n",
       "      <th>1964</th>\n",
       "      <th>1965</th>\n",
       "      <th>...</th>\n",
       "      <th>2008</th>\n",
       "      <th>2009</th>\n",
       "      <th>2010</th>\n",
       "      <th>2011</th>\n",
       "      <th>2012</th>\n",
       "      <th>2013</th>\n",
       "      <th>2014</th>\n",
       "      <th>2015</th>\n",
       "      <th>2016</th>\n",
       "      <th>2017</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Aruba</td>\n",
       "      <td>ABW</td>\n",
       "      <td>GDP per capita (current US$)</td>\n",
       "      <td>NY.GDP.PCAP.CD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>27546.899390</td>\n",
       "      <td>24631.434860</td>\n",
       "      <td>24271.940421</td>\n",
       "      <td>25324.720362</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>AFG</td>\n",
       "      <td>GDP per capita (current US$)</td>\n",
       "      <td>NY.GDP.PCAP.CD</td>\n",
       "      <td>59.777327</td>\n",
       "      <td>59.878153</td>\n",
       "      <td>58.492874</td>\n",
       "      <td>78.782758</td>\n",
       "      <td>82.208444</td>\n",
       "      <td>101.290471</td>\n",
       "      <td>...</td>\n",
       "      <td>373.361116</td>\n",
       "      <td>445.893298</td>\n",
       "      <td>553.300289</td>\n",
       "      <td>603.537023</td>\n",
       "      <td>669.009051</td>\n",
       "      <td>638.612543</td>\n",
       "      <td>629.345250</td>\n",
       "      <td>569.577923</td>\n",
       "      <td>561.778746</td>\n",
       "      <td>585.850064</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Angola</td>\n",
       "      <td>AGO</td>\n",
       "      <td>GDP per capita (current US$)</td>\n",
       "      <td>NY.GDP.PCAP.CD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>3868.579014</td>\n",
       "      <td>3347.844900</td>\n",
       "      <td>3531.416878</td>\n",
       "      <td>4299.008136</td>\n",
       "      <td>4539.467689</td>\n",
       "      <td>4804.633826</td>\n",
       "      <td>4707.578098</td>\n",
       "      <td>3683.552607</td>\n",
       "      <td>3308.772828</td>\n",
       "      <td>4170.312280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Albania</td>\n",
       "      <td>ALB</td>\n",
       "      <td>GDP per capita (current US$)</td>\n",
       "      <td>NY.GDP.PCAP.CD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>4370.539647</td>\n",
       "      <td>4114.136545</td>\n",
       "      <td>4094.358832</td>\n",
       "      <td>4437.178068</td>\n",
       "      <td>4247.614308</td>\n",
       "      <td>4413.081697</td>\n",
       "      <td>4578.666728</td>\n",
       "      <td>3952.830781</td>\n",
       "      <td>4131.872341</td>\n",
       "      <td>4537.862492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Andorra</td>\n",
       "      <td>AND</td>\n",
       "      <td>GDP per capita (current US$)</td>\n",
       "      <td>NY.GDP.PCAP.CD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>47785.659086</td>\n",
       "      <td>43339.379875</td>\n",
       "      <td>39736.354063</td>\n",
       "      <td>41098.766942</td>\n",
       "      <td>38391.080867</td>\n",
       "      <td>40619.711298</td>\n",
       "      <td>42294.994727</td>\n",
       "      <td>36038.267604</td>\n",
       "      <td>37231.815671</td>\n",
       "      <td>39146.548836</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 62 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  Country Name Country Code                Indicator Name  Indicator Code  \\\n",
       "0        Aruba          ABW  GDP per capita (current US$)  NY.GDP.PCAP.CD   \n",
       "1  Afghanistan          AFG  GDP per capita (current US$)  NY.GDP.PCAP.CD   \n",
       "2       Angola          AGO  GDP per capita (current US$)  NY.GDP.PCAP.CD   \n",
       "3      Albania          ALB  GDP per capita (current US$)  NY.GDP.PCAP.CD   \n",
       "4      Andorra          AND  GDP per capita (current US$)  NY.GDP.PCAP.CD   \n",
       "\n",
       "        1960       1961       1962       1963       1964        1965  \\\n",
       "0        NaN        NaN        NaN        NaN        NaN         NaN   \n",
       "1  59.777327  59.878153  58.492874  78.782758  82.208444  101.290471   \n",
       "2        NaN        NaN        NaN        NaN        NaN         NaN   \n",
       "3        NaN        NaN        NaN        NaN        NaN         NaN   \n",
       "4        NaN        NaN        NaN        NaN        NaN         NaN   \n",
       "\n",
       "       ...               2008          2009          2010          2011  \\\n",
       "0      ...       27546.899390  24631.434860  24271.940421  25324.720362   \n",
       "1      ...         373.361116    445.893298    553.300289    603.537023   \n",
       "2      ...        3868.579014   3347.844900   3531.416878   4299.008136   \n",
       "3      ...        4370.539647   4114.136545   4094.358832   4437.178068   \n",
       "4      ...       47785.659086  43339.379875  39736.354063  41098.766942   \n",
       "\n",
       "           2012          2013          2014          2015          2016  \\\n",
       "0           NaN           NaN           NaN           NaN           NaN   \n",
       "1    669.009051    638.612543    629.345250    569.577923    561.778746   \n",
       "2   4539.467689   4804.633826   4707.578098   3683.552607   3308.772828   \n",
       "3   4247.614308   4413.081697   4578.666728   3952.830781   4131.872341   \n",
       "4  38391.080867  40619.711298  42294.994727  36038.267604  37231.815671   \n",
       "\n",
       "           2017  \n",
       "0           NaN  \n",
       "1    585.850064  \n",
       "2   4170.312280  \n",
       "3   4537.862492  \n",
       "4  39146.548836  \n",
       "\n",
       "[5 rows x 62 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_gdp_per_c = pd.read_csv(\"data/data/GDP-per-capita.csv\")\n",
    "df_gdp_per_c.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Country Name</th>\n",
       "      <th>Country Code</th>\n",
       "      <th>Indicator Name</th>\n",
       "      <th>Indicator Code</th>\n",
       "      <th>1960</th>\n",
       "      <th>1961</th>\n",
       "      <th>1962</th>\n",
       "      <th>1963</th>\n",
       "      <th>1964</th>\n",
       "      <th>1965</th>\n",
       "      <th>...</th>\n",
       "      <th>2008</th>\n",
       "      <th>2009</th>\n",
       "      <th>2010</th>\n",
       "      <th>2011</th>\n",
       "      <th>2012</th>\n",
       "      <th>2013</th>\n",
       "      <th>2014</th>\n",
       "      <th>2015</th>\n",
       "      <th>2016</th>\n",
       "      <th>2017</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Aruba</td>\n",
       "      <td>ABW</td>\n",
       "      <td>Rural population</td>\n",
       "      <td>SP.RUR.TOTL</td>\n",
       "      <td>26685.0</td>\n",
       "      <td>27297.0</td>\n",
       "      <td>27693.0</td>\n",
       "      <td>27934.0</td>\n",
       "      <td>28108.0</td>\n",
       "      <td>28278.0</td>\n",
       "      <td>...</td>\n",
       "      <td>56978.0</td>\n",
       "      <td>57401.0</td>\n",
       "      <td>57891.0</td>\n",
       "      <td>58231.0</td>\n",
       "      <td>58513.0</td>\n",
       "      <td>58827.0</td>\n",
       "      <td>59121.0</td>\n",
       "      <td>59362.0</td>\n",
       "      <td>59547.0</td>\n",
       "      <td>59692.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>AFG</td>\n",
       "      <td>Rural population</td>\n",
       "      <td>SP.RUR.TOTL</td>\n",
       "      <td>8240568.0</td>\n",
       "      <td>8370722.0</td>\n",
       "      <td>8506983.0</td>\n",
       "      <td>8649584.0</td>\n",
       "      <td>8798513.0</td>\n",
       "      <td>8954113.0</td>\n",
       "      <td>...</td>\n",
       "      <td>20929063.0</td>\n",
       "      <td>21415472.0</td>\n",
       "      <td>21966159.0</td>\n",
       "      <td>22593984.0</td>\n",
       "      <td>23280573.0</td>\n",
       "      <td>23997724.0</td>\n",
       "      <td>24703806.0</td>\n",
       "      <td>25368831.0</td>\n",
       "      <td>25985093.0</td>\n",
       "      <td>26558736.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Angola</td>\n",
       "      <td>AGO</td>\n",
       "      <td>Rural population</td>\n",
       "      <td>SP.RUR.TOTL</td>\n",
       "      <td>5054316.0</td>\n",
       "      <td>5131812.0</td>\n",
       "      <td>5208828.0</td>\n",
       "      <td>5285253.0</td>\n",
       "      <td>5358588.0</td>\n",
       "      <td>5427638.0</td>\n",
       "      <td>...</td>\n",
       "      <td>9077595.0</td>\n",
       "      <td>9237647.0</td>\n",
       "      <td>9398363.0</td>\n",
       "      <td>9559552.0</td>\n",
       "      <td>9720241.0</td>\n",
       "      <td>9878849.0</td>\n",
       "      <td>10032988.0</td>\n",
       "      <td>10183690.0</td>\n",
       "      <td>10329915.0</td>\n",
       "      <td>10472420.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Albania</td>\n",
       "      <td>ALB</td>\n",
       "      <td>Rural population</td>\n",
       "      <td>SP.RUR.TOTL</td>\n",
       "      <td>1114818.0</td>\n",
       "      <td>1146208.0</td>\n",
       "      <td>1180553.0</td>\n",
       "      <td>1214693.0</td>\n",
       "      <td>1248887.0</td>\n",
       "      <td>1282417.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1473922.0</td>\n",
       "      <td>1432259.0</td>\n",
       "      <td>1393502.0</td>\n",
       "      <td>1358266.0</td>\n",
       "      <td>1324613.0</td>\n",
       "      <td>1291587.0</td>\n",
       "      <td>1258985.0</td>\n",
       "      <td>1226200.0</td>\n",
       "      <td>1195854.0</td>\n",
       "      <td>1167112.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Andorra</td>\n",
       "      <td>AND</td>\n",
       "      <td>Rural population</td>\n",
       "      <td>SP.RUR.TOTL</td>\n",
       "      <td>5572.0</td>\n",
       "      <td>5609.0</td>\n",
       "      <td>5616.0</td>\n",
       "      <td>5601.0</td>\n",
       "      <td>5554.0</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>...</td>\n",
       "      <td>9221.0</td>\n",
       "      <td>9365.0</td>\n",
       "      <td>9442.0</td>\n",
       "      <td>9442.0</td>\n",
       "      <td>9372.0</td>\n",
       "      <td>9261.0</td>\n",
       "      <td>9157.0</td>\n",
       "      <td>9093.0</td>\n",
       "      <td>9082.0</td>\n",
       "      <td>9120.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 62 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  Country Name Country Code    Indicator Name Indicator Code       1960  \\\n",
       "0        Aruba          ABW  Rural population    SP.RUR.TOTL    26685.0   \n",
       "1  Afghanistan          AFG  Rural population    SP.RUR.TOTL  8240568.0   \n",
       "2       Angola          AGO  Rural population    SP.RUR.TOTL  5054316.0   \n",
       "3      Albania          ALB  Rural population    SP.RUR.TOTL  1114818.0   \n",
       "4      Andorra          AND  Rural population    SP.RUR.TOTL     5572.0   \n",
       "\n",
       "        1961       1962       1963       1964       1965     ...      \\\n",
       "0    27297.0    27693.0    27934.0    28108.0    28278.0     ...       \n",
       "1  8370722.0  8506983.0  8649584.0  8798513.0  8954113.0     ...       \n",
       "2  5131812.0  5208828.0  5285253.0  5358588.0  5427638.0     ...       \n",
       "3  1146208.0  1180553.0  1214693.0  1248887.0  1282417.0     ...       \n",
       "4     5609.0     5616.0     5601.0     5554.0     5482.0     ...       \n",
       "\n",
       "         2008        2009        2010        2011        2012        2013  \\\n",
       "0     56978.0     57401.0     57891.0     58231.0     58513.0     58827.0   \n",
       "1  20929063.0  21415472.0  21966159.0  22593984.0  23280573.0  23997724.0   \n",
       "2   9077595.0   9237647.0   9398363.0   9559552.0   9720241.0   9878849.0   \n",
       "3   1473922.0   1432259.0   1393502.0   1358266.0   1324613.0   1291587.0   \n",
       "4      9221.0      9365.0      9442.0      9442.0      9372.0      9261.0   \n",
       "\n",
       "         2014        2015        2016        2017  \n",
       "0     59121.0     59362.0     59547.0     59692.0  \n",
       "1  24703806.0  25368831.0  25985093.0  26558736.0  \n",
       "2  10032988.0  10183690.0  10329915.0  10472420.0  \n",
       "3   1258985.0   1226200.0   1195854.0   1167112.0  \n",
       "4      9157.0      9093.0      9082.0      9120.0  \n",
       "\n",
       "[5 rows x 62 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_rural = pd.read_csv(\"data/data/rural-population.csv\")\n",
    "df_rural.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Country Name</th>\n",
       "      <th>Country Code</th>\n",
       "      <th>Indicator Name</th>\n",
       "      <th>Indicator Code</th>\n",
       "      <th>1960</th>\n",
       "      <th>1961</th>\n",
       "      <th>1962</th>\n",
       "      <th>1963</th>\n",
       "      <th>1964</th>\n",
       "      <th>1965</th>\n",
       "      <th>...</th>\n",
       "      <th>2008</th>\n",
       "      <th>2009</th>\n",
       "      <th>2010</th>\n",
       "      <th>2011</th>\n",
       "      <th>2012</th>\n",
       "      <th>2013</th>\n",
       "      <th>2014</th>\n",
       "      <th>2015</th>\n",
       "      <th>2016</th>\n",
       "      <th>2017</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Aruba</td>\n",
       "      <td>ABW</td>\n",
       "      <td>Agricultural land (% of land area)</td>\n",
       "      <td>AG.LND.AGRI.ZS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>11.111111</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>AFG</td>\n",
       "      <td>Agricultural land (% of land area)</td>\n",
       "      <td>AG.LND.AGRI.ZS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>57.745918</td>\n",
       "      <td>57.837821</td>\n",
       "      <td>57.914407</td>\n",
       "      <td>58.010906</td>\n",
       "      <td>58.013969</td>\n",
       "      <td>...</td>\n",
       "      <td>58.067580</td>\n",
       "      <td>58.067580</td>\n",
       "      <td>58.069111</td>\n",
       "      <td>58.067580</td>\n",
       "      <td>58.067580</td>\n",
       "      <td>58.067580</td>\n",
       "      <td>58.067580</td>\n",
       "      <td>58.067580</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Angola</td>\n",
       "      <td>AGO</td>\n",
       "      <td>Agricultural land (% of land area)</td>\n",
       "      <td>AG.LND.AGRI.ZS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>45.857063</td>\n",
       "      <td>45.881126</td>\n",
       "      <td>45.897169</td>\n",
       "      <td>45.921232</td>\n",
       "      <td>45.937274</td>\n",
       "      <td>...</td>\n",
       "      <td>46.274164</td>\n",
       "      <td>46.755434</td>\n",
       "      <td>46.835646</td>\n",
       "      <td>47.316917</td>\n",
       "      <td>47.316917</td>\n",
       "      <td>47.477340</td>\n",
       "      <td>47.477340</td>\n",
       "      <td>47.477340</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Albania</td>\n",
       "      <td>ALB</td>\n",
       "      <td>Agricultural land (% of land area)</td>\n",
       "      <td>AG.LND.AGRI.ZS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>44.963504</td>\n",
       "      <td>44.963504</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>44.890511</td>\n",
       "      <td>45.145985</td>\n",
       "      <td>...</td>\n",
       "      <td>43.102190</td>\n",
       "      <td>43.843066</td>\n",
       "      <td>43.843066</td>\n",
       "      <td>43.832117</td>\n",
       "      <td>43.843066</td>\n",
       "      <td>43.332117</td>\n",
       "      <td>42.857299</td>\n",
       "      <td>42.857664</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Andorra</td>\n",
       "      <td>AND</td>\n",
       "      <td>Agricultural land (% of land area)</td>\n",
       "      <td>AG.LND.AGRI.ZS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>55.319149</td>\n",
       "      <td>55.319149</td>\n",
       "      <td>55.319149</td>\n",
       "      <td>55.319149</td>\n",
       "      <td>55.319149</td>\n",
       "      <td>...</td>\n",
       "      <td>44.170213</td>\n",
       "      <td>44.191489</td>\n",
       "      <td>42.063830</td>\n",
       "      <td>42.063830</td>\n",
       "      <td>39.914894</td>\n",
       "      <td>40.021277</td>\n",
       "      <td>40.021277</td>\n",
       "      <td>40.021277</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 62 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  Country Name Country Code                      Indicator Name  \\\n",
       "0        Aruba          ABW  Agricultural land (% of land area)   \n",
       "1  Afghanistan          AFG  Agricultural land (% of land area)   \n",
       "2       Angola          AGO  Agricultural land (% of land area)   \n",
       "3      Albania          ALB  Agricultural land (% of land area)   \n",
       "4      Andorra          AND  Agricultural land (% of land area)   \n",
       "\n",
       "   Indicator Code  1960       1961       1962       1963       1964  \\\n",
       "0  AG.LND.AGRI.ZS   NaN  11.111111  11.111111  11.111111  11.111111   \n",
       "1  AG.LND.AGRI.ZS   NaN  57.745918  57.837821  57.914407  58.010906   \n",
       "2  AG.LND.AGRI.ZS   NaN  45.857063  45.881126  45.897169  45.921232   \n",
       "3  AG.LND.AGRI.ZS   NaN  44.963504  44.963504  45.000000  44.890511   \n",
       "4  AG.LND.AGRI.ZS   NaN  55.319149  55.319149  55.319149  55.319149   \n",
       "\n",
       "        1965  ...        2008       2009       2010       2011       2012  \\\n",
       "0  11.111111  ...   11.111111  11.111111  11.111111  11.111111  11.111111   \n",
       "1  58.013969  ...   58.067580  58.067580  58.069111  58.067580  58.067580   \n",
       "2  45.937274  ...   46.274164  46.755434  46.835646  47.316917  47.316917   \n",
       "3  45.145985  ...   43.102190  43.843066  43.843066  43.832117  43.843066   \n",
       "4  55.319149  ...   44.170213  44.191489  42.063830  42.063830  39.914894   \n",
       "\n",
       "        2013       2014       2015  2016  2017  \n",
       "0  11.111111  11.111111  11.111111   NaN   NaN  \n",
       "1  58.067580  58.067580  58.067580   NaN   NaN  \n",
       "2  47.477340  47.477340  47.477340   NaN   NaN  \n",
       "3  43.332117  42.857299  42.857664   NaN   NaN  \n",
       "4  40.021277  40.021277  40.021277   NaN   NaN  \n",
       "\n",
       "[5 rows x 62 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_agricul = pd.read_csv(\"data/data/agricultural-land.csv\")\n",
    "df_agricul.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1973-09-04</td>\n",
       "      <td>0.6432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1973-09-05</td>\n",
       "      <td>0.6370</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1973-09-06</td>\n",
       "      <td>0.6360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1973-09-07</td>\n",
       "      <td>0.6380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1973-09-10</td>\n",
       "      <td>0.6335</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date   value\n",
       "0  1973-09-04  0.6432\n",
       "1  1973-09-05  0.6370\n",
       "2  1973-09-06  0.6360\n",
       "3  1973-09-07  0.6380\n",
       "4  1973-09-10  0.6335"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_coffee = pd.read_csv(\"data/data/coffee-prices-historical-chart-data.csv\")\n",
    "df_coffee.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Start date</th>\n",
       "      <th>End date</th>\n",
       "      <th>Country</th>\n",
       "      <th>ISO</th>\n",
       "      <th>Location</th>\n",
       "      <th>Latitude</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>Magnitude value</th>\n",
       "      <th>Magnitude scale</th>\n",
       "      <th>Disaster type</th>\n",
       "      <th>Disaster subtype</th>\n",
       "      <th>Disaster name</th>\n",
       "      <th>Disaster No.</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>26/03/1973</td>\n",
       "      <td>26/03/1973</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>East, Central, along Cartinga</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1973-0019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>/03/1974</td>\n",
       "      <td>/03/1974</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Most of the country</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1974-0015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>/05/1974</td>\n",
       "      <td>/05/1974</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>North East (Ceara)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Landslide</td>\n",
       "      <td>Landslide</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1974-0024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>23/10/1974</td>\n",
       "      <td>23/10/1974</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Sao Tome</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Landslide</td>\n",
       "      <td>Landslide</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1974-0053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>01/01/1974</td>\n",
       "      <td>01/01/1974</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Sao Paulo</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Vaccinated</td>\n",
       "      <td>Epidemic</td>\n",
       "      <td>Bacterial disease</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1974-0127</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Start date    End date Country  ISO                       Location  \\\n",
       "0  26/03/1973  26/03/1973  Brazil  BRA  East, Central, along Cartinga   \n",
       "1    /03/1974    /03/1974  Brazil  BRA            Most of the country   \n",
       "2    /05/1974    /05/1974  Brazil  BRA             North East (Ceara)   \n",
       "3  23/10/1974  23/10/1974  Brazil  BRA                       Sao Tome   \n",
       "4  01/01/1974  01/01/1974  Brazil  BRA                      Sao Paulo   \n",
       "\n",
       "  Latitude Longitude  Magnitude value Magnitude scale Disaster type  \\\n",
       "0      NaN       NaN              NaN             Km2         Flood   \n",
       "1      NaN       NaN              NaN             Km2         Flood   \n",
       "2      NaN       NaN              NaN             NaN     Landslide   \n",
       "3      NaN       NaN              NaN             NaN     Landslide   \n",
       "4      NaN       NaN              NaN      Vaccinated      Epidemic   \n",
       "\n",
       "    Disaster subtype Disaster name Disaster No.  \n",
       "0                 --           NaN    1973-0019  \n",
       "1                 --           NaN    1974-0015  \n",
       "2          Landslide           NaN    1974-0024  \n",
       "3          Landslide           NaN    1974-0053  \n",
       "4  Bacterial disease           NaN    1974-0127  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_disaster = pd.read_csv(\"data/data/disaster_in_coffee_region.csv\", encoding=\"ISO-8859-1\")\n",
    "df_disaster.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That is all the data I mined. Perfectly I need daily price data (in my case coffee price), daily weather data, daily consumption and production data\n",
    "As coffee production counties are poor and do not report systematically, I didn't have production data yet."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also, I have trouble with weather data: a website with data is not available. https://governmentshutdown.noaa.gov/\n",
    "\n",
    "Other sites give us At most 3 years of weather data, which does not suffice, because еhree years is inconclusive to describe economic trends; therefore, at least a 25-year period must be analyzed."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data for second Pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_data():\n",
    "    return pd.read_csv(\"data/data/GDP.csv\"), pd.read_csv(\"data/data/GDP-per-capita.csv\"), pd.read_csv(\"data/data/agricultural-land.csv\"), pd.read_csv(\"data/data/rural-population.csv\") \n",
    "\n",
    "def drop_col(df_gdp_per_c, df_gdp, df_agricul, df_rural):\n",
    "    gdp_per_c = df_gdp_per_c.drop(['Country Name','Indicator Name', 'Indicator Code' ], axis=1)\n",
    "    gdp = df_gdp.drop(['Country Name','Indicator Name', 'Indicator Code' ], axis=1)\n",
    "    agricul = df_agricul.drop(['Country Name','Indicator Name', 'Indicator Code' ], axis=1)\n",
    "    rural = df_rural.drop(['Country Name','Indicator Name', 'Indicator Code' ], axis=1)\n",
    "    \n",
    "    return(gdp_per_c, gdp, agricul, rural)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_merged_data():\n",
    "    \"\"\"\n",
    "    () => data frame\n",
    "    \n",
    "    This function merges four datasets for second pipeline. \n",
    "    \"\"\"\n",
    "    gdp_per_c, gdp, agricul, rural = read_data()\n",
    "    gdp_per_c, gdp, agricul, rural = drop_col(gdp_per_c, gdp, agricul, rural)\n",
    "    gdp.index = gdp['Country Code']\n",
    "    gdp_per_c.index = gdp['Country Code']\n",
    "    agricul.index = gdp['Country Code']\n",
    "    rural.index = gdp['Country Code']\n",
    "    gdp_per_c = gdp_per_c.T\n",
    "    agricul = agricul.T\n",
    "    rural = rural.T\n",
    "    gdp = gdp.T\n",
    "    gdp = gdp.drop(['Country Code' ], axis=0)\n",
    "    gdp_per_c = gdp_per_c.drop(['Country Code' ], axis=0)\n",
    "    agricul = agricul.drop(['Country Code' ], axis=0)\n",
    "    rural = rural.drop(['Country Code' ], axis=0)\n",
    "    gdp.rename(columns=lambda x: x+' GDP', inplace=True)\n",
    "    gdp_per_c.rename(columns=lambda x: x+' GDP per capita', inplace=True)\n",
    "    agricul.rename(columns=lambda x: x+' Agricultural land', inplace=True)\n",
    "    rural.rename(columns=lambda x: x+' Rural pop.', inplace=True)\n",
    "    main = gdp.join(gdp_per_c).join(agricul).join(rural)\n",
    "    main.to_csv(\"data/data/ecomonical_index.csv\")\n",
    "    return main\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "main = get_merged_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Country Code\n",
       "ABW GDP           40\n",
       "AFG GDP           19\n",
       "AGO GDP           20\n",
       "ALB GDP           24\n",
       "AND GDP           10\n",
       "ARB GDP            8\n",
       "ARE GDP           15\n",
       "ARG GDP            2\n",
       "ARM GDP           30\n",
       "ASM GDP           43\n",
       "ATG GDP           17\n",
       "AUS GDP            0\n",
       "AUT GDP            0\n",
       "AZE GDP           30\n",
       "BDI GDP            0\n",
       "BEL GDP            0\n",
       "BEN GDP            0\n",
       "BFA GDP            0\n",
       "BGD GDP            0\n",
       "BGR GDP           20\n",
       "BHR GDP           20\n",
       "BHS GDP            0\n",
       "BIH GDP           34\n",
       "BLR GDP           30\n",
       "BLZ GDP            0\n",
       "BMU GDP            4\n",
       "BOL GDP            0\n",
       "BRA GDP            0\n",
       "BRB GDP           14\n",
       "BRN GDP            5\n",
       "                  ..\n",
       "TLA Rural pop.     0\n",
       "TLS Rural pop.     0\n",
       "TMN Rural pop.     0\n",
       "TON Rural pop.     0\n",
       "TSA Rural pop.     0\n",
       "TSS Rural pop.     0\n",
       "TTO Rural pop.     0\n",
       "TUN Rural pop.     0\n",
       "TUR Rural pop.     0\n",
       "TUV Rural pop.     0\n",
       "TZA Rural pop.     0\n",
       "UGA Rural pop.     0\n",
       "UKR Rural pop.     0\n",
       "UMC Rural pop.     0\n",
       "URY Rural pop.     0\n",
       "USA Rural pop.     0\n",
       "UZB Rural pop.     0\n",
       "VCT Rural pop.     0\n",
       "VEN Rural pop.     0\n",
       "VGB Rural pop.     0\n",
       "VIR Rural pop.     0\n",
       "VNM Rural pop.     0\n",
       "VUT Rural pop.     0\n",
       "WLD Rural pop.     0\n",
       "WSM Rural pop.     0\n",
       "XKX Rural pop.    58\n",
       "YEM Rural pop.     0\n",
       "ZAF Rural pop.     0\n",
       "ZMB Rural pop.     0\n",
       "ZWE Rural pop.     0\n",
       "Length: 1056, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# amount of missing values\n",
    "main.isnull().sum(axis = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "main = main.drop(main.loc[:, main.isnull().sum() >10], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Country Code\n",
       "AND GDP           10\n",
       "ARB GDP            8\n",
       "ARG GDP            2\n",
       "AUS GDP            0\n",
       "AUT GDP            0\n",
       "BDI GDP            0\n",
       "BEL GDP            0\n",
       "BEN GDP            0\n",
       "BFA GDP            0\n",
       "BGD GDP            0\n",
       "BHS GDP            0\n",
       "BLZ GDP            0\n",
       "BMU GDP            4\n",
       "BOL GDP            0\n",
       "BRA GDP            0\n",
       "BRN GDP            5\n",
       "BWA GDP            0\n",
       "CAF GDP            0\n",
       "CAN GDP            0\n",
       "CHE GDP           10\n",
       "CHL GDP            0\n",
       "CHN GDP            0\n",
       "CIV GDP            0\n",
       "CMR GDP            0\n",
       "COD GDP            0\n",
       "COG GDP            0\n",
       "COL GDP            0\n",
       "CRI GDP            0\n",
       "CSS GDP            0\n",
       "DEU GDP           10\n",
       "                  ..\n",
       "TKM Rural pop.     0\n",
       "TLA Rural pop.     0\n",
       "TLS Rural pop.     0\n",
       "TMN Rural pop.     0\n",
       "TON Rural pop.     0\n",
       "TSA Rural pop.     0\n",
       "TSS Rural pop.     0\n",
       "TTO Rural pop.     0\n",
       "TUN Rural pop.     0\n",
       "TUR Rural pop.     0\n",
       "TUV Rural pop.     0\n",
       "TZA Rural pop.     0\n",
       "UGA Rural pop.     0\n",
       "UKR Rural pop.     0\n",
       "UMC Rural pop.     0\n",
       "URY Rural pop.     0\n",
       "USA Rural pop.     0\n",
       "UZB Rural pop.     0\n",
       "VCT Rural pop.     0\n",
       "VEN Rural pop.     0\n",
       "VGB Rural pop.     0\n",
       "VIR Rural pop.     0\n",
       "VNM Rural pop.     0\n",
       "VUT Rural pop.     0\n",
       "WLD Rural pop.     0\n",
       "WSM Rural pop.     0\n",
       "YEM Rural pop.     0\n",
       "ZAF Rural pop.     0\n",
       "ZMB Rural pop.     0\n",
       "ZWE Rural pop.     0\n",
       "Length: 787, dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "main.isnull().sum(axis = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "main.fillna(main.mean(), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cleaning_data(dataset):\n",
    "    dataset = dataset.drop(dataset.loc[:, dataset.isnull().sum() >10], axis=1)\n",
    "    dataset.fillna(dataset.mean(), inplace=True)\n",
    "    return dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def data_by_county( dataset, countries=None):\n",
    "    \"\"\"\n",
    "    returns ready-to-use dataframe with economic indexes by country\n",
    "    \n",
    "    The data can be used for any economic prediction tasks.   \n",
    "    \n",
    "    \"\"\"\n",
    "\n",
    "    column = []\n",
    "    if countries==None:\n",
    "        return dataset\n",
    "    else:\n",
    "        for country in countries:\n",
    "            for col in dataset.columns:\n",
    "                if country in col:\n",
    "                    column.append(col)\n",
    "    return dataset[column]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# usage of pipelene\n",
    "d = data_by_county(cleaning_data(get_merged_data()), ['BRA', 'ZWE'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Country Code</th>\n",
       "      <th>BRA GDP</th>\n",
       "      <th>BRA GDP per capita</th>\n",
       "      <th>BRA Agricultural land</th>\n",
       "      <th>BRA Rural pop.</th>\n",
       "      <th>ZWE GDP</th>\n",
       "      <th>ZWE GDP per capita</th>\n",
       "      <th>ZWE Agricultural land</th>\n",
       "      <th>ZWE Rural pop.</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1960</th>\n",
       "      <td>210.027470</td>\n",
       "      <td>1.516557e+10</td>\n",
       "      <td>27.959874</td>\n",
       "      <td>38891711.0</td>\n",
       "      <td>280.994586</td>\n",
       "      <td>1.052990e+09</td>\n",
       "      <td>34.917175</td>\n",
       "      <td>3274901.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1961</th>\n",
       "      <td>204.929302</td>\n",
       "      <td>1.523685e+10</td>\n",
       "      <td>18.010108</td>\n",
       "      <td>39315725.0</td>\n",
       "      <td>283.315869</td>\n",
       "      <td>1.096647e+09</td>\n",
       "      <td>28.396019</td>\n",
       "      <td>3374486.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1962</th>\n",
       "      <td>260.225266</td>\n",
       "      <td>1.992629e+10</td>\n",
       "      <td>18.654031</td>\n",
       "      <td>39742281.0</td>\n",
       "      <td>279.440989</td>\n",
       "      <td>1.117602e+09</td>\n",
       "      <td>28.615743</td>\n",
       "      <td>3476215.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1963</th>\n",
       "      <td>291.950589</td>\n",
       "      <td>2.302148e+10</td>\n",
       "      <td>19.323797</td>\n",
       "      <td>40154044.0</td>\n",
       "      <td>280.566213</td>\n",
       "      <td>1.159512e+09</td>\n",
       "      <td>28.835466</td>\n",
       "      <td>3571610.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1964</th>\n",
       "      <td>261.331083</td>\n",
       "      <td>2.121189e+10</td>\n",
       "      <td>19.865784</td>\n",
       "      <td>40536437.0</td>\n",
       "      <td>285.053174</td>\n",
       "      <td>1.217138e+09</td>\n",
       "      <td>29.055189</td>\n",
       "      <td>3668154.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1965</th>\n",
       "      <td>260.964693</td>\n",
       "      <td>2.179004e+10</td>\n",
       "      <td>20.386713</td>\n",
       "      <td>40883136.0</td>\n",
       "      <td>297.363437</td>\n",
       "      <td>1.311436e+09</td>\n",
       "      <td>29.223213</td>\n",
       "      <td>3765439.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1966</th>\n",
       "      <td>315.277382</td>\n",
       "      <td>2.706272e+10</td>\n",
       "      <td>20.948919</td>\n",
       "      <td>41188409.0</td>\n",
       "      <td>281.490800</td>\n",
       "      <td>1.281750e+09</td>\n",
       "      <td>29.494636</td>\n",
       "      <td>3862905.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1967</th>\n",
       "      <td>346.880100</td>\n",
       "      <td>3.059183e+10</td>\n",
       "      <td>21.486718</td>\n",
       "      <td>41456121.0</td>\n",
       "      <td>297.231875</td>\n",
       "      <td>1.397002e+09</td>\n",
       "      <td>29.714360</td>\n",
       "      <td>3960866.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1968</th>\n",
       "      <td>374.083262</td>\n",
       "      <td>3.387588e+10</td>\n",
       "      <td>22.160792</td>\n",
       "      <td>41683417.0</td>\n",
       "      <td>304.982159</td>\n",
       "      <td>1.479600e+09</td>\n",
       "      <td>29.934083</td>\n",
       "      <td>4060357.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1969</th>\n",
       "      <td>403.065252</td>\n",
       "      <td>3.745890e+10</td>\n",
       "      <td>23.184345</td>\n",
       "      <td>41876543.0</td>\n",
       "      <td>348.935805</td>\n",
       "      <td>1.747999e+09</td>\n",
       "      <td>30.153806</td>\n",
       "      <td>4163908.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Country Code     BRA GDP  BRA GDP per capita  BRA Agricultural land  \\\n",
       "1960          210.027470        1.516557e+10              27.959874   \n",
       "1961          204.929302        1.523685e+10              18.010108   \n",
       "1962          260.225266        1.992629e+10              18.654031   \n",
       "1963          291.950589        2.302148e+10              19.323797   \n",
       "1964          261.331083        2.121189e+10              19.865784   \n",
       "1965          260.964693        2.179004e+10              20.386713   \n",
       "1966          315.277382        2.706272e+10              20.948919   \n",
       "1967          346.880100        3.059183e+10              21.486718   \n",
       "1968          374.083262        3.387588e+10              22.160792   \n",
       "1969          403.065252        3.745890e+10              23.184345   \n",
       "\n",
       "Country Code  BRA Rural pop.     ZWE GDP  ZWE GDP per capita  \\\n",
       "1960              38891711.0  280.994586        1.052990e+09   \n",
       "1961              39315725.0  283.315869        1.096647e+09   \n",
       "1962              39742281.0  279.440989        1.117602e+09   \n",
       "1963              40154044.0  280.566213        1.159512e+09   \n",
       "1964              40536437.0  285.053174        1.217138e+09   \n",
       "1965              40883136.0  297.363437        1.311436e+09   \n",
       "1966              41188409.0  281.490800        1.281750e+09   \n",
       "1967              41456121.0  297.231875        1.397002e+09   \n",
       "1968              41683417.0  304.982159        1.479600e+09   \n",
       "1969              41876543.0  348.935805        1.747999e+09   \n",
       "\n",
       "Country Code  ZWE Agricultural land  ZWE Rural pop.  \n",
       "1960                      34.917175       3274901.0  \n",
       "1961                      28.396019       3374486.0  \n",
       "1962                      28.615743       3476215.0  \n",
       "1963                      28.835466       3571610.0  \n",
       "1964                      29.055189       3668154.0  \n",
       "1965                      29.223213       3765439.0  \n",
       "1966                      29.494636       3862905.0  \n",
       "1967                      29.714360       3960866.0  \n",
       "1968                      29.934083       4060357.0  \n",
       "1969                      30.153806       4163908.0  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "clean_d = data_by_county(cleaning_data(get_merged_data()))\n",
    "clean_d.to_csv(\"data/data/clean_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.38156587e-08, 9.97594925e-01, 1.83920739e-09, 2.55830633e-03,\n",
       "        1.84838931e-08, 6.92659679e-02, 2.29686036e-09, 2.15423795e-04],\n",
       "       [1.34148348e-08, 9.97416615e-01, 1.17895593e-09, 2.57363857e-03,\n",
       "        1.85460817e-08, 7.17873570e-02, 1.85882595e-09, 2.20896532e-04],\n",
       "       [1.30388728e-08, 9.98428843e-01, 9.34680700e-10, 1.99133065e-03,\n",
       "        1.40016977e-08, 5.59986559e-02, 1.43382322e-09, 1.74179571e-04],\n",
       "       [1.26655876e-08, 9.98732488e-01, 8.38317347e-10, 1.74198848e-03,\n",
       "        1.21717033e-08, 5.03026800e-02, 1.25095867e-09, 1.54945875e-04],\n",
       "       [1.22997727e-08, 9.98355991e-01, 9.35000224e-10, 1.90788234e-03,\n",
       "        1.34162733e-08, 5.72856490e-02, 1.36750753e-09, 1.72644829e-04],\n",
       "       [1.19546789e-08, 9.98192016e-01, 9.33906433e-10, 1.87283865e-03,\n",
       "        1.36220895e-08, 6.00763027e-02, 1.33870266e-09, 1.72493120e-04],\n",
       "       [1.16368213e-08, 9.98879130e-01, 7.73220168e-10, 1.52025544e-03,\n",
       "        1.03897658e-08, 4.73091023e-02, 1.08864077e-09, 1.42579004e-04],\n",
       "       [1.13271619e-08, 9.98958019e-01, 7.01635919e-10, 1.35372480e-03,\n",
       "        9.70592887e-09, 4.56182636e-02, 9.70304616e-10, 1.29339707e-04],\n",
       "       [1.10322344e-08, 9.99046756e-01, 6.53552495e-10, 1.22930180e-03,\n",
       "        8.99434700e-09, 4.36354539e-02, 8.82797638e-10, 1.19745561e-04],\n",
       "       [1.07484972e-08, 9.98912361e-01, 6.18254399e-10, 1.11671721e-03,\n",
       "        9.30503312e-09, 4.66136937e-02, 8.04108271e-10, 1.11038481e-04],\n",
       "       [1.04798468e-08, 9.99010188e-01, 5.51765475e-10, 9.91998922e-04,\n",
       "        8.59236087e-09, 4.44707776e-02, 7.16871867e-10, 1.00945802e-04],\n",
       "       [1.02223614e-08, 9.99020762e-01, 4.84942882e-10, 8.55324831e-04,\n",
       "        8.26646780e-09, 4.42354812e-02, 6.21148920e-10, 8.92487148e-05],\n",
       "       [9.97522759e-09, 9.98955182e-01, 4.14131416e-10, 7.19817672e-04,\n",
       "        8.25431674e-09, 4.56948574e-02, 5.24934359e-10, 7.71240588e-05],\n",
       "       [9.73959919e-09, 9.99129751e-01, 3.10404385e-10, 5.31938326e-04,\n",
       "        7.28242574e-09, 4.17067727e-02, 3.88815199e-10, 5.85536565e-05],\n",
       "       [9.51070147e-09, 9.99283383e-01, 2.38855780e-10, 4.01197641e-04,\n",
       "        6.39241796e-09, 3.78491424e-02, 2.95324210e-10, 4.53589614e-05],\n",
       "       [9.28683889e-09, 9.99376236e-01, 2.06273641e-10, 3.40875219e-04,\n",
       "        5.77449944e-09, 3.53132007e-02, 2.51739296e-10, 3.95847643e-05],\n",
       "       [9.06970557e-09, 9.99600204e-01, 1.69202977e-10, 2.76039463e-04,\n",
       "        4.48185218e-09, 2.82728681e-02, 2.04613138e-10, 3.28938857e-05],\n",
       "       [8.85719980e-09, 9.99693249e-01, 1.48379126e-10, 2.38967504e-04,\n",
       "        3.80903170e-09, 2.47659165e-02, 1.77710241e-10, 2.92041312e-05],\n",
       "       [8.65015922e-09, 9.99765240e-01, 1.31075770e-10, 2.09319326e-04,\n",
       "        3.23221620e-09, 2.16661334e-02, 1.56567605e-10, 2.62506317e-05],\n",
       "       [8.44780362e-09, 9.99735264e-01, 1.18306100e-10, 1.86426040e-04,\n",
       "        3.32398973e-09, 2.30079588e-02, 1.40317636e-10, 2.40376552e-05],\n",
       "       [8.25023453e-09, 9.99596444e-01, 1.14127004e-10, 1.77947350e-04,\n",
       "        3.96504620e-09, 2.84062729e-02, 1.34955405e-10, 2.36538246e-05],\n",
       "       [8.05878426e-09, 9.99538329e-01, 1.02465724e-10, 1.58188514e-04,\n",
       "        4.08811365e-09, 3.03826153e-02, 1.21071665e-10, 2.17321588e-05],\n",
       "       [7.87366285e-09, 9.99540752e-01, 9.68556861e-11, 1.47474446e-04,\n",
       "        3.92445985e-09, 3.03028580e-02, 1.14062826e-10, 2.09801668e-05],\n",
       "       [7.69366881e-09, 9.99271562e-01, 1.34562095e-10, 2.03378507e-04,\n",
       "        4.75381140e-09, 3.81615300e-02, 1.58946589e-10, 2.99598357e-05],\n",
       "       [7.52660686e-09, 9.99538540e-01, 1.31546179e-10, 1.96800166e-04,\n",
       "        3.64118153e-09, 3.03754464e-02, 1.55318598e-10, 3.00241851e-05],\n",
       "       [7.36812983e-09, 9.99680453e-01, 1.23950186e-10, 1.83352203e-04,\n",
       "        2.91927637e-09, 2.52775966e-02, 1.46511677e-10, 2.89591391e-05],\n",
       "       [7.21830417e-09, 9.99731259e-01, 1.04408160e-10, 1.51285535e-04,\n",
       "        2.58256255e-09, 2.31816108e-02, 1.22546221e-10, 2.47305627e-05],\n",
       "       [7.07660993e-09, 9.99737368e-01, 9.59029454e-11, 1.36734389e-04,\n",
       "        2.46594499e-09, 2.29167247e-02, 1.12393787e-10, 2.31228425e-05],\n",
       "       [6.94242833e-09, 9.99720385e-01, 8.60473972e-11, 1.20511880e-04,\n",
       "        2.46202879e-09, 2.36460680e-02, 1.00743077e-10, 2.10621150e-05],\n",
       "       [6.81571346e-09, 9.99810510e-01, 6.72966454e-11, 9.25655561e-05,\n",
       "        1.96578538e-09, 1.94662684e-02, 7.86711829e-11, 1.66912687e-05],\n",
       "       [6.69437502e-09, 9.99819268e-01, 6.25642906e-11, 8.42967045e-05,\n",
       "        1.86692795e-09, 1.90111382e-02, 7.27879605e-11, 1.56508219e-05],\n",
       "       [6.57928544e-09, 9.99897280e-01, 4.86060985e-11, 6.37980971e-05,\n",
       "        1.37246117e-09, 1.43326710e-02, 5.65082329e-11, 1.21698857e-05],\n",
       "       [6.46888155e-09, 9.99858007e-01, 7.36721846e-11, 9.47316169e-05,\n",
       "        1.57738904e-09, 1.68510389e-02, 8.61324620e-11, 1.85313521e-05],\n",
       "       [6.36333327e-09, 9.99887624e-01, 6.81668689e-11, 8.54804697e-05,\n",
       "        1.37460301e-09, 1.49910850e-02, 7.98198554e-11, 1.72012177e-05],\n",
       "       [6.26106271e-09, 9.99923790e-01, 5.38929887e-11, 6.60875971e-05,\n",
       "        1.11050811e-09, 1.23454609e-02, 6.35414400e-11, 1.36762340e-05],\n",
       "       [6.16129484e-09, 9.99957278e-01, 4.01963422e-11, 4.72330744e-05,\n",
       "        8.16526309e-10, 9.24336034e-03, 4.65696851e-11, 1.00452226e-05],\n",
       "       [6.06348542e-09, 9.99949426e-01, 3.64387329e-11, 4.20742794e-05,\n",
       "        8.73132689e-10, 1.00569711e-02, 4.32669374e-11, 9.19164592e-06],\n",
       "       [5.96826160e-09, 9.99953368e-01, 3.51608232e-11, 3.97446115e-05,\n",
       "        8.24691098e-10, 9.65713028e-03, 4.22000934e-11, 8.94463536e-06],\n",
       "       [5.87629294e-09, 9.99972531e-01, 3.60299887e-11, 3.98343456e-05,\n",
       "        6.23197267e-10, 7.41185460e-03, 4.34547638e-11, 9.23389109e-06],\n",
       "       [5.78802344e-09, 9.99934549e-01, 5.20466493e-11, 5.62118778e-05,\n",
       "        9.48310939e-10, 1.14409324e-02, 6.39962328e-11, 1.34150866e-05],\n",
       "       [5.70461336e-09, 9.99947910e-01, 4.77159055e-11, 5.02980367e-05,\n",
       "        8.35082585e-10, 1.02065890e-02, 5.93935973e-11, 1.23521319e-05],\n",
       "       [5.62544494e-09, 9.99926607e-01, 5.63482555e-11, 5.86143806e-05,\n",
       "        9.79702661e-10, 1.21151648e-02, 7.04221942e-11, 1.45521129e-05],\n",
       "       [5.55046571e-09, 9.99922064e-01, 6.26168098e-11, 6.42583295e-05,\n",
       "        9.98712454e-10, 1.24844300e-02, 7.87194811e-11, 1.60968199e-05],\n",
       "       [5.47969973e-09, 9.99947383e-01, 5.75279010e-11, 5.81649301e-05,\n",
       "        8.11949000e-10, 1.02580800e-02, 7.24545669e-11, 1.48255868e-05],\n",
       "       [5.41285444e-09, 9.99962382e-01, 4.86431850e-11, 4.82420692e-05,\n",
       "        6.78817075e-10, 8.67359265e-03, 6.17915960e-11, 1.25430544e-05],\n",
       "       [5.34984639e-09, 9.99979168e-01, 3.65557441e-11, 3.59852845e-05,\n",
       "        4.98806853e-10, 6.45457665e-03, 4.66755527e-11, 9.56227101e-06],\n",
       "       [5.29059394e-09, 9.99987922e-01, 2.94649195e-11, 2.87651481e-05,\n",
       "        3.74481881e-10, 4.91480019e-03, 3.80400439e-11, 7.82891637e-06],\n",
       "       [5.23483448e-09, 9.99992826e-01, 2.32624333e-11, 2.26317776e-05,\n",
       "        2.84159915e-10, 3.78782580e-03, 2.99741358e-11, 6.32169036e-06],\n",
       "       [5.18189264e-09, 9.99996610e-01, 1.92959113e-11, 1.84930188e-05,\n",
       "        1.92046673e-10, 2.60385886e-03, 2.50749979e-11, 5.31199784e-06],\n",
       "       [5.13087311e-09, 9.99986626e-01, 1.96320043e-11, 1.86543612e-05,\n",
       "        3.74478995e-10, 5.17177924e-03, 2.52754235e-11, 5.51929500e-06],\n",
       "       [5.08134359e-09, 9.99989459e-01, 1.48120058e-11, 1.39563614e-05,\n",
       "        3.25945570e-10, 4.59137262e-03, 1.89582111e-11, 4.26014967e-06],\n",
       "       [5.03299601e-09, 9.99989307e-01, 1.25931932e-11, 1.16718187e-05,\n",
       "        3.21435814e-10, 4.62438424e-03, 1.62041181e-11, 3.68350590e-06],\n",
       "       [4.98593144e-09, 9.99983311e-01, 1.33759055e-11, 1.22660284e-05,\n",
       "        3.92727635e-10, 5.77734790e-03, 1.69869341e-11, 4.00801294e-06],\n",
       "       [4.94040431e-09, 9.99980478e-01, 1.34895375e-11, 1.21067586e-05,\n",
       "        4.15062028e-10, 6.24855379e-03, 1.69345526e-11, 4.09996025e-06],\n",
       "       [4.89674221e-09, 9.99979068e-01, 1.37660482e-11, 1.20629877e-05,\n",
       "        4.19823171e-10, 6.47017826e-03, 1.70504595e-11, 4.23537185e-06],\n",
       "       [4.85506333e-09, 9.99959078e-01, 1.87595056e-11, 1.62617772e-05,\n",
       "        5.73391028e-10, 9.04664884e-03, 2.32352951e-11, 5.91909964e-06],\n",
       "       [4.81552272e-09, 9.99957089e-01, 1.55846403e-11, 1.61555854e-05,\n",
       "        5.73600207e-10, 9.26385099e-03, 1.94625918e-11, 6.09478180e-06],\n",
       "       [4.77791840e-09, 9.99962314e-01, 1.36019194e-11, 1.39394315e-05,\n",
       "        5.25207840e-10, 8.68163517e-03, 1.69865072e-11, 5.44913981e-06]])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import preprocessing\n",
    "normalize = preprocessing.normalize(d)\n",
    "normalize"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data For first pipeline\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1973-09-04</td>\n",
       "      <td>0.6432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1973-09-05</td>\n",
       "      <td>0.6370</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1973-09-06</td>\n",
       "      <td>0.6360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1973-09-07</td>\n",
       "      <td>0.6380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1973-09-10</td>\n",
       "      <td>0.6335</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date   value\n",
       "0  1973-09-04  0.6432\n",
       "1  1973-09-05  0.6370\n",
       "2  1973-09-06  0.6360\n",
       "3  1973-09-07  0.6380\n",
       "4  1973-09-10  0.6335"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_coffee.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_average_monthly_data(df_coffee, month):\n",
    "    date_value = {}\n",
    "    counter = 0\n",
    "    loc = 0\n",
    "    loc_count = 0\n",
    "    for index, row in df_coffee.iterrows():\n",
    "        \n",
    "        if month[counter] in row['date'] :\n",
    "\n",
    "            if month[counter] not in date_value:\n",
    "                date_value[month[counter]] = []\n",
    "            loc_count += 1\n",
    "            loc+= row[' value']\n",
    "        else:\n",
    "\n",
    "            date_value[month[counter]].append(loc/loc_count)\n",
    "            loc = 0\n",
    "            loc_count=0\n",
    "            if '2018-12' in row['date']:\n",
    "                break\n",
    "            loc += row[' value']\n",
    "            loc_count += 1\n",
    "            counter += 1\n",
    "\n",
    "    df = pd.DataFrame.from_dict(date_value, orient='index', columns=[\"coffee price\"])\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = make_average_monthly_data(df_coffee,df_cpi['TIME'] )\n",
    "df = df.sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>LOCATION</th>\n",
       "      <th>INDICATOR</th>\n",
       "      <th>SUBJECT</th>\n",
       "      <th>MEASURE</th>\n",
       "      <th>FREQUENCY</th>\n",
       "      <th>TIME</th>\n",
       "      <th>Value</th>\n",
       "      <th>Flag Codes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1973-09</td>\n",
       "      <td>6.543496</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1973-10</td>\n",
       "      <td>7.012194</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1973-11</td>\n",
       "      <td>7.966616</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1973-12</td>\n",
       "      <td>7.849054</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AUT</td>\n",
       "      <td>CPI</td>\n",
       "      <td>TOT</td>\n",
       "      <td>AGRWTH</td>\n",
       "      <td>M</td>\n",
       "      <td>1974-01</td>\n",
       "      <td>8.128263</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  LOCATION INDICATOR SUBJECT MEASURE FREQUENCY     TIME     Value Flag Codes\n",
       "0      AUT       CPI     TOT  AGRWTH         M  1973-09  6.543496        NaN\n",
       "1      AUT       CPI     TOT  AGRWTH         M  1973-10  7.012194        NaN\n",
       "2      AUT       CPI     TOT  AGRWTH         M  1973-11  7.966616        NaN\n",
       "3      AUT       CPI     TOT  AGRWTH         M  1973-12  7.849054        NaN\n",
       "4      AUT       CPI     TOT  AGRWTH         M  1974-01  8.128263        NaN"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cpi.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "cpi = df_cpi.drop(['SUBJECT', 'INDICATOR', 'MEASURE', 'Flag Codes', 'FREQUENCY' ], axis=1)\n",
    "names = cpi.LOCATION.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "def insert_data(list_of_countries, location, value):\n",
    "    for c in list_of_countries:\n",
    "        loc_df = df_cpi.loc[df_cpi[location] == c]\n",
    "        df.insert(loc=1, column='{} CPI'.format(c), value=[i for i in loc_df[value]]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['AUT' 'BEL' 'CAN' 'CZE' 'DNK' 'FIN' 'FRA' 'DEU' 'GRC' 'HUN' 'ISL' 'IRL'\n",
      " 'ITA' 'JPN' 'KOR' 'LUX' 'MEX' 'NLD' 'NOR' 'POL' 'PRT' 'SVK' 'ESP' 'SWE'\n",
      " 'CHE' 'TUR' 'GBR' 'USA' 'BRA' 'CHL' 'CHN' 'EST' 'IND' 'IDN' 'ISR' 'RUS'\n",
      " 'SVN' 'ZAF' 'OECD' 'OECDE' 'G-7' 'COL' 'LVA' 'SAU' 'EU28' 'EA19' 'ARG'\n",
      " 'LTU' 'CRI' 'G-20']\n"
     ]
    }
   ],
   "source": [
    "print(names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coffee price</th>\n",
       "      <th>COL CPI</th>\n",
       "      <th>IDN CPI</th>\n",
       "      <th>IND CPI</th>\n",
       "      <th>USA CPI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1973-09</th>\n",
       "      <td>0.652700</td>\n",
       "      <td>22.72727</td>\n",
       "      <td>44.00000</td>\n",
       "      <td>19.25078</td>\n",
       "      <td>7.363420</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973-10</th>\n",
       "      <td>0.659552</td>\n",
       "      <td>22.72727</td>\n",
       "      <td>42.30769</td>\n",
       "      <td>21.55440</td>\n",
       "      <td>7.801418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973-11</th>\n",
       "      <td>0.646810</td>\n",
       "      <td>27.27273</td>\n",
       "      <td>30.63584</td>\n",
       "      <td>23.29897</td>\n",
       "      <td>8.254717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973-12</th>\n",
       "      <td>0.670937</td>\n",
       "      <td>27.27273</td>\n",
       "      <td>27.17391</td>\n",
       "      <td>23.81443</td>\n",
       "      <td>8.705882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-01</th>\n",
       "      <td>0.716991</td>\n",
       "      <td>26.08696</td>\n",
       "      <td>42.54144</td>\n",
       "      <td>25.67010</td>\n",
       "      <td>9.389671</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         coffee price   COL CPI   IDN CPI   IND CPI   USA CPI\n",
       "1973-09      0.652700  22.72727  44.00000  19.25078  7.363420\n",
       "1973-10      0.659552  22.72727  42.30769  21.55440  7.801418\n",
       "1973-11      0.646810  27.27273  30.63584  23.29897  8.254717\n",
       "1973-12      0.670937  27.27273  27.17391  23.81443  8.705882\n",
       "1974-01      0.716991  26.08696  42.54144  25.67010  9.389671"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "insert_data(['USA', 'IND', 'IDN', 'COL'], 'LOCATION', 'Value')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Manual technics to rewrite dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Start date</th>\n",
       "      <th>End date</th>\n",
       "      <th>Country</th>\n",
       "      <th>ISO</th>\n",
       "      <th>Location</th>\n",
       "      <th>Latitude</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>Magnitude value</th>\n",
       "      <th>Magnitude scale</th>\n",
       "      <th>Disaster type</th>\n",
       "      <th>Disaster subtype</th>\n",
       "      <th>Disaster name</th>\n",
       "      <th>Disaster No.</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>26/03/1973</td>\n",
       "      <td>26/03/1973</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>East, Central, along Cartinga</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1973-0019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>/03/1974</td>\n",
       "      <td>/03/1974</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Most of the country</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1974-0015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>/05/1974</td>\n",
       "      <td>/05/1974</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>North East (Ceara)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Landslide</td>\n",
       "      <td>Landslide</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1974-0024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>23/10/1974</td>\n",
       "      <td>23/10/1974</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Sao Tome</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Landslide</td>\n",
       "      <td>Landslide</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1974-0053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>01/01/1974</td>\n",
       "      <td>01/01/1974</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Sao Paulo</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Vaccinated</td>\n",
       "      <td>Epidemic</td>\n",
       "      <td>Bacterial disease</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1974-0127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>/07/1975</td>\n",
       "      <td>/07/1975</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Parana</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ï¿½C</td>\n",
       "      <td>Extreme temperature</td>\n",
       "      <td>Cold wave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1975-0039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>18/07/1975</td>\n",
       "      <td>18/07/1975</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Pernambuco, Recife (North East areas)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1975-0042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>/07/1975</td>\n",
       "      <td>/07/1975</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Pernambuco, Recife (North East areas)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Vaccinated</td>\n",
       "      <td>Epidemic</td>\n",
       "      <td>Bacterial disease</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1975-0068</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>/12/1975</td>\n",
       "      <td>/12/1975</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ï¿½C</td>\n",
       "      <td>Extreme temperature</td>\n",
       "      <td>Heat wave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1975-0145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>/05/1977</td>\n",
       "      <td>/05/1977</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Recife</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1977-0061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>17/06/1977</td>\n",
       "      <td>17/06/1977</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Aracaju, Sergipe</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Kph</td>\n",
       "      <td>Storm</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1977-0077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>09/08/1977</td>\n",
       "      <td>09/08/1977</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Rio Grande Do Sul</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1977-0105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>/01/1977</td>\n",
       "      <td>/01/1977</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1977-0232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>//1977</td>\n",
       "      <td>//1977</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>North East</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Drought</td>\n",
       "      <td>Drought</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1977-9006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>/01/1978</td>\n",
       "      <td>/01/1978</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>South Eastern</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1978-0006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>//1978</td>\n",
       "      <td>//1978</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Central and Southern</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Drought</td>\n",
       "      <td>Drought</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1978-9005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>/01/1979</td>\n",
       "      <td>/01/1979</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Minas Gerais, Espirito Santo (East-Central)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1979-0003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>//1979</td>\n",
       "      <td>//1979</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Northeastern</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Drought</td>\n",
       "      <td>Drought</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1979-9012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>/01/1980</td>\n",
       "      <td>/01/1980</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Minas Gerais state</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1980-0008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>/02/1980</td>\n",
       "      <td>/02/1980</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>North, Central</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1980-0031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>25/06/1980</td>\n",
       "      <td>25/06/1980</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Irati town</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Kph</td>\n",
       "      <td>Storm</td>\n",
       "      <td>Convective storm</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1980-0054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>/04/1981</td>\n",
       "      <td>/04/1981</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>NorthEastern</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1981-0041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>02/12/1981</td>\n",
       "      <td>02/12/1981</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Rio de Janeiro state</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1981-0105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>/06/1982</td>\n",
       "      <td>/06/1982</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Parana state</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Kph</td>\n",
       "      <td>Storm</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1982-0071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>/01/1983</td>\n",
       "      <td>/01/1983</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Minas Gerais state</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Km2</td>\n",
       "      <td>Flood</td>\n",
       "      <td>--</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1983-0001</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Start date    End date Country  ISO  \\\n",
       "0   26/03/1973  26/03/1973  Brazil  BRA   \n",
       "1     /03/1974    /03/1974  Brazil  BRA   \n",
       "2     /05/1974    /05/1974  Brazil  BRA   \n",
       "3   23/10/1974  23/10/1974  Brazil  BRA   \n",
       "4   01/01/1974  01/01/1974  Brazil  BRA   \n",
       "5     /07/1975    /07/1975  Brazil  BRA   \n",
       "6   18/07/1975  18/07/1975  Brazil  BRA   \n",
       "7     /07/1975    /07/1975  Brazil  BRA   \n",
       "8     /12/1975    /12/1975  Brazil  BRA   \n",
       "9     /05/1977    /05/1977  Brazil  BRA   \n",
       "10  17/06/1977  17/06/1977  Brazil  BRA   \n",
       "11  09/08/1977  09/08/1977  Brazil  BRA   \n",
       "12    /01/1977    /01/1977  Brazil  BRA   \n",
       "13      //1977      //1977  Brazil  BRA   \n",
       "14    /01/1978    /01/1978  Brazil  BRA   \n",
       "15      //1978      //1978  Brazil  BRA   \n",
       "16    /01/1979    /01/1979  Brazil  BRA   \n",
       "17      //1979      //1979  Brazil  BRA   \n",
       "18    /01/1980    /01/1980  Brazil  BRA   \n",
       "19    /02/1980    /02/1980  Brazil  BRA   \n",
       "20  25/06/1980  25/06/1980  Brazil  BRA   \n",
       "21    /04/1981    /04/1981  Brazil  BRA   \n",
       "22  02/12/1981  02/12/1981  Brazil  BRA   \n",
       "23    /06/1982    /06/1982  Brazil  BRA   \n",
       "24    /01/1983    /01/1983  Brazil  BRA   \n",
       "\n",
       "                                       Location Latitude Longitude  \\\n",
       "0                 East, Central, along Cartinga      NaN       NaN   \n",
       "1                           Most of the country      NaN       NaN   \n",
       "2                            North East (Ceara)      NaN       NaN   \n",
       "3                                      Sao Tome      NaN       NaN   \n",
       "4                                     Sao Paulo      NaN       NaN   \n",
       "5                                        Parana      NaN       NaN   \n",
       "6         Pernambuco, Recife (North East areas)      NaN       NaN   \n",
       "7         Pernambuco, Recife (North East areas)      NaN       NaN   \n",
       "8                                           NaN      NaN       NaN   \n",
       "9                                        Recife      NaN       NaN   \n",
       "10                             Aracaju, Sergipe      NaN       NaN   \n",
       "11                            Rio Grande Do Sul      NaN       NaN   \n",
       "12                                          NaN      NaN       NaN   \n",
       "13                                   North East      NaN       NaN   \n",
       "14                                South Eastern      NaN       NaN   \n",
       "15                         Central and Southern      NaN       NaN   \n",
       "16  Minas Gerais, Espirito Santo (East-Central)      NaN       NaN   \n",
       "17                                 Northeastern      NaN       NaN   \n",
       "18                           Minas Gerais state      NaN       NaN   \n",
       "19                               North, Central      NaN       NaN   \n",
       "20                                   Irati town      NaN       NaN   \n",
       "21                                 NorthEastern      NaN       NaN   \n",
       "22                         Rio de Janeiro state      NaN       NaN   \n",
       "23                                 Parana state      NaN       NaN   \n",
       "24                           Minas Gerais state      NaN       NaN   \n",
       "\n",
       "    Magnitude value Magnitude scale         Disaster type   Disaster subtype  \\\n",
       "0               NaN             Km2                 Flood                 --   \n",
       "1               NaN             Km2                 Flood                 --   \n",
       "2               NaN             NaN             Landslide          Landslide   \n",
       "3               NaN             NaN             Landslide          Landslide   \n",
       "4               NaN      Vaccinated              Epidemic  Bacterial disease   \n",
       "5               NaN            ï¿½C  Extreme temperature           Cold wave   \n",
       "6               NaN             Km2                 Flood                 --   \n",
       "7               NaN      Vaccinated              Epidemic  Bacterial disease   \n",
       "8               NaN            ï¿½C  Extreme temperature           Heat wave   \n",
       "9               NaN             Km2                 Flood                 --   \n",
       "10              NaN             Kph                 Storm                 --   \n",
       "11              NaN             Km2                 Flood                 --   \n",
       "12              NaN             Km2                 Flood                 --   \n",
       "13              NaN             Km2               Drought            Drought   \n",
       "14              NaN             Km2                 Flood                 --   \n",
       "15              NaN             Km2               Drought            Drought   \n",
       "16              NaN             Km2                 Flood                 --   \n",
       "17              NaN             Km2               Drought            Drought   \n",
       "18              NaN             Km2                 Flood                 --   \n",
       "19              NaN             Km2                 Flood                 --   \n",
       "20              NaN             Kph                 Storm   Convective storm   \n",
       "21              NaN             Km2                 Flood                 --   \n",
       "22              NaN             Km2                 Flood                 --   \n",
       "23              NaN             Kph                 Storm                 --   \n",
       "24              NaN             Km2                 Flood                 --   \n",
       "\n",
       "   Disaster name Disaster No.  \n",
       "0            NaN    1973-0019  \n",
       "1            NaN    1974-0015  \n",
       "2            NaN    1974-0024  \n",
       "3            NaN    1974-0053  \n",
       "4            NaN    1974-0127  \n",
       "5            NaN    1975-0039  \n",
       "6            NaN    1975-0042  \n",
       "7            NaN    1975-0068  \n",
       "8            NaN    1975-0145  \n",
       "9            NaN    1977-0061  \n",
       "10           NaN    1977-0077  \n",
       "11           NaN    1977-0105  \n",
       "12           NaN    1977-0232  \n",
       "13           NaN    1977-9006  \n",
       "14           NaN    1978-0006  \n",
       "15           NaN    1978-9005  \n",
       "16           NaN    1979-0003  \n",
       "17           NaN    1979-9012  \n",
       "18           NaN    1980-0008  \n",
       "19           NaN    1980-0031  \n",
       "20           NaN    1980-0054  \n",
       "21           NaN    1981-0041  \n",
       "22           NaN    1981-0105  \n",
       "23           NaN    1982-0071  \n",
       "24           NaN    1983-0001  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_disaster.head(25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "disaster = df_disaster.drop(['Location', 'Latitude','Longitude','Magnitude value','Magnitude scale', 'Disaster name', 'Disaster subtype', 'Disaster No.', 'Country'],\n",
    "                            axis=1).drop([2392, 2393, 2394, 2395], axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['BRA', 'COL', 'ETH', 'GTM', 'HND', 'IND', 'IDN', 'MEX', 'PER',\n",
       "       'UGA', 'VNM'], dtype=object)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "disaster = disaster.rename(columns={'Disaster type': 'disaster'})\n",
    "disaster.ISO.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Flood', 'Landslide', 'Epidemic', 'Extreme temperature ', 'Storm',\n",
       "       'Drought', 'Earthquake', 'Insect infestation', 'Wildfire',\n",
       "       'Mass movement (dry)', 'Volcanic activity'], dtype=object)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "disaster.disaster.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['1973-09', '1973-10', '1973-11', '1973-12', '1974-01', '1974-02',\n",
       "       '1974-03', '1974-04', '1974-05', '1974-06',\n",
       "       ...\n",
       "       '2018-02', '2018-03', '2018-04', '2018-05', '2018-06', '2018-07',\n",
       "       '2018-08', '2018-09', '2018-10', '2018-11'],\n",
       "      dtype='object', length=543)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "disaster = disaster.drop(0, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Start date</th>\n",
       "      <th>End date</th>\n",
       "      <th>ISO</th>\n",
       "      <th>disaster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>/03/1974</td>\n",
       "      <td>/03/1974</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>/05/1974</td>\n",
       "      <td>/05/1974</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Landslide</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>23/10/1974</td>\n",
       "      <td>23/10/1974</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Landslide</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>01/01/1974</td>\n",
       "      <td>01/01/1974</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Epidemic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>/07/1975</td>\n",
       "      <td>/07/1975</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Extreme temperature</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>18/07/1975</td>\n",
       "      <td>18/07/1975</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>/07/1975</td>\n",
       "      <td>/07/1975</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Epidemic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>/12/1975</td>\n",
       "      <td>/12/1975</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Extreme temperature</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>/05/1977</td>\n",
       "      <td>/05/1977</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>17/06/1977</td>\n",
       "      <td>17/06/1977</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>09/08/1977</td>\n",
       "      <td>09/08/1977</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>/01/1977</td>\n",
       "      <td>/01/1977</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>//1977</td>\n",
       "      <td>//1977</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Drought</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>/01/1978</td>\n",
       "      <td>/01/1978</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>//1978</td>\n",
       "      <td>//1978</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Drought</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>/01/1979</td>\n",
       "      <td>/01/1979</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>//1979</td>\n",
       "      <td>//1979</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Drought</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>/01/1980</td>\n",
       "      <td>/01/1980</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>/02/1980</td>\n",
       "      <td>/02/1980</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>25/06/1980</td>\n",
       "      <td>25/06/1980</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>/04/1981</td>\n",
       "      <td>/04/1981</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>02/12/1981</td>\n",
       "      <td>02/12/1981</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>/06/1982</td>\n",
       "      <td>/06/1982</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>/01/1983</td>\n",
       "      <td>/01/1983</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>/05/1983</td>\n",
       "      <td>/05/1983</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>/12/1983</td>\n",
       "      <td>/12/1983</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Landslide</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>/09/1983</td>\n",
       "      <td>//1983</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Drought</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>/04/1984</td>\n",
       "      <td>/04/1984</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>/05/1984</td>\n",
       "      <td>/05/1984</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Epidemic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>/06/1984</td>\n",
       "      <td>/06/1984</td>\n",
       "      <td>BRA</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2362</th>\n",
       "      <td>16/09/2014</td>\n",
       "      <td>16/09/2014</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2363</th>\n",
       "      <td>30/11/2014</td>\n",
       "      <td>30/11/2014</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2364</th>\n",
       "      <td>24/06/2015</td>\n",
       "      <td>29/06/2015</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2365</th>\n",
       "      <td>25/07/2015</td>\n",
       "      <td>05/08/2015</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2366</th>\n",
       "      <td>07/09/2015</td>\n",
       "      <td>11/09/2015</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2367</th>\n",
       "      <td>14/09/2015</td>\n",
       "      <td>14/09/2015</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2368</th>\n",
       "      <td>/12/2015</td>\n",
       "      <td>/02/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Drought</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2369</th>\n",
       "      <td>28/07/2016</td>\n",
       "      <td>28/07/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2370</th>\n",
       "      <td>13/10/2016</td>\n",
       "      <td>15/10/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2371</th>\n",
       "      <td>02/08/2016</td>\n",
       "      <td>02/08/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2372</th>\n",
       "      <td>18/08/2016</td>\n",
       "      <td>21/08/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2373</th>\n",
       "      <td>09/10/2016</td>\n",
       "      <td>16/10/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2374</th>\n",
       "      <td>12/12/2016</td>\n",
       "      <td>20/12/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2375</th>\n",
       "      <td>01/01/2016</td>\n",
       "      <td>30/09/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Epidemic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2376</th>\n",
       "      <td>30/10/2016</td>\n",
       "      <td>05/11/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2377</th>\n",
       "      <td>12/08/2016</td>\n",
       "      <td>14/08/2016</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2378</th>\n",
       "      <td>27/06/2017</td>\n",
       "      <td>05/07/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2379</th>\n",
       "      <td>07/07/2017</td>\n",
       "      <td>10/07/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2380</th>\n",
       "      <td>17/07/2017</td>\n",
       "      <td>18/07/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2381</th>\n",
       "      <td>03/08/2017</td>\n",
       "      <td>03/08/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2382</th>\n",
       "      <td>24/08/2017</td>\n",
       "      <td>24/08/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2383</th>\n",
       "      <td>15/09/2017</td>\n",
       "      <td>16/09/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2384</th>\n",
       "      <td>04/11/2017</td>\n",
       "      <td>05/11/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2385</th>\n",
       "      <td>08/10/2017</td>\n",
       "      <td>16/10/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2386</th>\n",
       "      <td>19/11/2017</td>\n",
       "      <td>24/11/2017</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2387</th>\n",
       "      <td>06/06/2018</td>\n",
       "      <td>19/06/2018</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2388</th>\n",
       "      <td>23/06/2018</td>\n",
       "      <td>27/06/2018</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2389</th>\n",
       "      <td>18/07/2018</td>\n",
       "      <td>22/07/2018</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2390</th>\n",
       "      <td>17/08/2018</td>\n",
       "      <td>19/08/2018</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Storm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2391</th>\n",
       "      <td>29/08/2018</td>\n",
       "      <td>05/09/2018</td>\n",
       "      <td>VNM</td>\n",
       "      <td>Flood</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2391 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Start date    End date  ISO              disaster\n",
       "1       /03/1974    /03/1974  BRA                 Flood\n",
       "2       /05/1974    /05/1974  BRA             Landslide\n",
       "3     23/10/1974  23/10/1974  BRA             Landslide\n",
       "4     01/01/1974  01/01/1974  BRA              Epidemic\n",
       "5       /07/1975    /07/1975  BRA  Extreme temperature \n",
       "6     18/07/1975  18/07/1975  BRA                 Flood\n",
       "7       /07/1975    /07/1975  BRA              Epidemic\n",
       "8       /12/1975    /12/1975  BRA  Extreme temperature \n",
       "9       /05/1977    /05/1977  BRA                 Flood\n",
       "10    17/06/1977  17/06/1977  BRA                 Storm\n",
       "11    09/08/1977  09/08/1977  BRA                 Flood\n",
       "12      /01/1977    /01/1977  BRA                 Flood\n",
       "13        //1977      //1977  BRA               Drought\n",
       "14      /01/1978    /01/1978  BRA                 Flood\n",
       "15        //1978      //1978  BRA               Drought\n",
       "16      /01/1979    /01/1979  BRA                 Flood\n",
       "17        //1979      //1979  BRA               Drought\n",
       "18      /01/1980    /01/1980  BRA                 Flood\n",
       "19      /02/1980    /02/1980  BRA                 Flood\n",
       "20    25/06/1980  25/06/1980  BRA                 Storm\n",
       "21      /04/1981    /04/1981  BRA                 Flood\n",
       "22    02/12/1981  02/12/1981  BRA                 Flood\n",
       "23      /06/1982    /06/1982  BRA                 Storm\n",
       "24      /01/1983    /01/1983  BRA                 Flood\n",
       "25      /05/1983    /05/1983  BRA                 Flood\n",
       "26      /12/1983    /12/1983  BRA             Landslide\n",
       "27      /09/1983      //1983  BRA               Drought\n",
       "28      /04/1984    /04/1984  BRA                 Flood\n",
       "29      /05/1984    /05/1984  BRA              Epidemic\n",
       "30      /06/1984    /06/1984  BRA                 Flood\n",
       "...          ...         ...  ...                   ...\n",
       "2362  16/09/2014  16/09/2014  VNM                 Storm\n",
       "2363  30/11/2014  30/11/2014  VNM                 Storm\n",
       "2364  24/06/2015  29/06/2015  VNM                 Storm\n",
       "2365  25/07/2015  05/08/2015  VNM                 Flood\n",
       "2366  07/09/2015  11/09/2015  VNM                 Flood\n",
       "2367  14/09/2015  14/09/2015  VNM                 Storm\n",
       "2368    /12/2015    /02/2017  VNM               Drought\n",
       "2369  28/07/2016  28/07/2016  VNM                 Storm\n",
       "2370  13/10/2016  15/10/2016  VNM                 Storm\n",
       "2371  02/08/2016  02/08/2016  VNM                 Storm\n",
       "2372  18/08/2016  21/08/2016  VNM                 Storm\n",
       "2373  09/10/2016  16/10/2016  VNM                 Flood\n",
       "2374  12/12/2016  20/12/2016  VNM                 Flood\n",
       "2375  01/01/2016  30/09/2016  VNM              Epidemic\n",
       "2376  30/10/2016  05/11/2016  VNM                 Flood\n",
       "2377  12/08/2016  14/08/2016  VNM                 Flood\n",
       "2378  27/06/2017  05/07/2017  VNM                 Flood\n",
       "2379  07/07/2017  10/07/2017  VNM                 Flood\n",
       "2380  17/07/2017  18/07/2017  VNM                 Storm\n",
       "2381  03/08/2017  03/08/2017  VNM                 Flood\n",
       "2382  24/08/2017  24/08/2017  VNM                 Storm\n",
       "2383  15/09/2017  16/09/2017  VNM                 Storm\n",
       "2384  04/11/2017  05/11/2017  VNM                 Storm\n",
       "2385  08/10/2017  16/10/2017  VNM                 Flood\n",
       "2386  19/11/2017  24/11/2017  VNM                 Flood\n",
       "2387  06/06/2018  19/06/2018  VNM                 Storm\n",
       "2388  23/06/2018  27/06/2018  VNM                 Flood\n",
       "2389  18/07/2018  22/07/2018  VNM                 Storm\n",
       "2390  17/08/2018  19/08/2018  VNM                 Storm\n",
       "2391  29/08/2018  05/09/2018  VNM                 Flood\n",
       "\n",
       "[2391 rows x 4 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "disaster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "def disaster_to_timeline(disaster_df, month, countries, disasters):\n",
    "    # BRA 0-10\n",
    "    # COL 11-21\n",
    "    # ETH 22-32\n",
    "    # GTM 33-43\n",
    "    # HND 44-54\n",
    "    # IND 55-65\n",
    "    # IDN 66-76\n",
    "    # MEX 77-87\n",
    "    # PER 88-98\n",
    "    # UGA 99-109\n",
    "    # VNM 110-120\n",
    "    disaster= {'Flood':0, 'Landslide':1, 'Epidemic':2, 'Extreme temperature ': 3, 'Storm': 4,\n",
    "       'Drought': 5, 'Earthquake': 6, 'Insect infestation': 7, 'Wildfire': 8,\n",
    "       'Mass movement (dry)': 9, 'Volcanic activity': 10}\n",
    "    contr = {'BRA':0, 'COL': 11, 'ETH':22, 'GTM':33, 'HND':44, 'IND':55, 'IDN':66, 'MEX':77, 'PER':88,\n",
    "       'UGA':99, 'VNM':110}\n",
    "\n",
    "    date_value = {}\n",
    "    for i in month:\n",
    "        date_value[i] = np.zeros(121)\n",
    "    for index, row in disaster_df.iterrows():\n",
    "        \n",
    "        loc_row = row['Start date'].split('/')\n",
    "        loc_end = row['End date'].split('/')\n",
    "        if loc_row[1]=='' or loc_end[1]=='':\n",
    "            for y in [int(loc_row[2]), int(loc_end[2])]:\n",
    "                for i in range(1, 13):\n",
    "                    if '{0}-{1}'.format(y, str(i).zfill(2)) in date_value:\n",
    "                        date_value['{0}-{1}'.format(y, str(i).zfill(2))][disaster[row['disaster']]+contr[row['ISO']]] = 1\n",
    "        elif loc_row[2] == loc_end[2]:\n",
    "            for i in range(int(loc_row[1]), int(loc_end[1])+1):\n",
    "                 if '{0}-{1}'.format(loc_end[2], str(i).zfill(2)) in date_value:\n",
    "                        date_value['{0}-{1}'.format(loc_end[2], str(i).zfill(2))][disaster[row['disaster']]+contr[row['ISO']]] = 1\n",
    "        elif loc_row[2] != loc_end[2]:\n",
    "            \n",
    "            for i in range(int(loc_row[1]), 13):\n",
    "                 if '{0}-{1}'.format(loc_row[2], str(i).zfill(2)) in date_value:\n",
    "                        date_value['{0}-{1}'.format(loc_row[2], str(i).zfill(2))][disaster[row['disaster']]+contr[row['ISO']]] = 1\n",
    "            for i in range(1,  int(loc_end[1])+1):\n",
    "                 if '{0}-{1}'.format(loc_end[2], str(i).zfill(2)) in date_value:\n",
    "                        date_value['{0}-{1}'.format(loc_end[2], str(i).zfill(2))][disaster[row['disaster']]+contr[row['ISO']]] = 1\n",
    "        \n",
    "\n",
    "    col = ['{0} {1}'.format(c, d) for c in countries for d in disasters]\n",
    "    df = pd.DataFrame.from_dict(date_value, orient='index', columns=col)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "data = disaster_to_timeline(disaster, df.index, ['BRA', 'COL', 'ETH', 'GTM', 'HND', 'IND', 'IDN', 'MEX', 'PER',\n",
    "       'UGA', 'VNM'], ['Flood', 'Landslide', 'Epidemic', 'Extreme temperature ', 'Storm',\n",
    "       'Drought', 'Earthquake', 'Insect infestation', 'Wildfire',\n",
    "       'Mass movement (dry)', 'Volcanic activity'])\n",
    "data = data.sort_index()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coffee price</th>\n",
       "      <th>COL CPI</th>\n",
       "      <th>IDN CPI</th>\n",
       "      <th>IND CPI</th>\n",
       "      <th>USA CPI</th>\n",
       "      <th>BRA Flood</th>\n",
       "      <th>BRA Landslide</th>\n",
       "      <th>BRA Epidemic</th>\n",
       "      <th>BRA Extreme temperature</th>\n",
       "      <th>BRA Storm</th>\n",
       "      <th>...</th>\n",
       "      <th>VNM Landslide</th>\n",
       "      <th>VNM Epidemic</th>\n",
       "      <th>VNM Extreme temperature</th>\n",
       "      <th>VNM Storm</th>\n",
       "      <th>VNM Drought</th>\n",
       "      <th>VNM Earthquake</th>\n",
       "      <th>VNM Insect infestation</th>\n",
       "      <th>VNM Wildfire</th>\n",
       "      <th>VNM Mass movement (dry)</th>\n",
       "      <th>VNM Volcanic activity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1973-09</th>\n",
       "      <td>0.652700</td>\n",
       "      <td>22.727270</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>19.250780</td>\n",
       "      <td>7.363420</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973-10</th>\n",
       "      <td>0.659552</td>\n",
       "      <td>22.727270</td>\n",
       "      <td>42.307690</td>\n",
       "      <td>21.554400</td>\n",
       "      <td>7.801418</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973-11</th>\n",
       "      <td>0.646810</td>\n",
       "      <td>27.272730</td>\n",
       "      <td>30.635840</td>\n",
       "      <td>23.298970</td>\n",
       "      <td>8.254717</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973-12</th>\n",
       "      <td>0.670937</td>\n",
       "      <td>27.272730</td>\n",
       "      <td>27.173910</td>\n",
       "      <td>23.814430</td>\n",
       "      <td>8.705882</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-01</th>\n",
       "      <td>0.716991</td>\n",
       "      <td>26.086960</td>\n",
       "      <td>42.541440</td>\n",
       "      <td>25.670100</td>\n",
       "      <td>9.389671</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-02</th>\n",
       "      <td>0.756550</td>\n",
       "      <td>26.086960</td>\n",
       "      <td>46.703300</td>\n",
       "      <td>25.304880</td>\n",
       "      <td>10.023310</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-03</th>\n",
       "      <td>0.732176</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>47.282610</td>\n",
       "      <td>27.254510</td>\n",
       "      <td>10.392610</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-04</th>\n",
       "      <td>0.714795</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>49.206350</td>\n",
       "      <td>28.011750</td>\n",
       "      <td>10.091740</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-05</th>\n",
       "      <td>0.730982</td>\n",
       "      <td>23.076920</td>\n",
       "      <td>49.206350</td>\n",
       "      <td>28.964860</td>\n",
       "      <td>10.706150</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-06</th>\n",
       "      <td>0.725355</td>\n",
       "      <td>23.076920</td>\n",
       "      <td>43.216080</td>\n",
       "      <td>29.182160</td>\n",
       "      <td>10.859730</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-07</th>\n",
       "      <td>0.705859</td>\n",
       "      <td>18.518520</td>\n",
       "      <td>36.842110</td>\n",
       "      <td>28.074870</td>\n",
       "      <td>11.512410</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-08</th>\n",
       "      <td>0.633591</td>\n",
       "      <td>23.076920</td>\n",
       "      <td>36.842110</td>\n",
       "      <td>29.973710</td>\n",
       "      <td>10.864750</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-09</th>\n",
       "      <td>0.544295</td>\n",
       "      <td>22.222220</td>\n",
       "      <td>35.185180</td>\n",
       "      <td>34.642230</td>\n",
       "      <td>11.946900</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-10</th>\n",
       "      <td>0.575364</td>\n",
       "      <td>25.925930</td>\n",
       "      <td>35.135140</td>\n",
       "      <td>31.884060</td>\n",
       "      <td>12.061400</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-11</th>\n",
       "      <td>0.602961</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>35.398230</td>\n",
       "      <td>27.842810</td>\n",
       "      <td>12.200440</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1974-12</th>\n",
       "      <td>0.594642</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>33.760680</td>\n",
       "      <td>25.395500</td>\n",
       "      <td>12.337660</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-01</th>\n",
       "      <td>0.560886</td>\n",
       "      <td>24.137930</td>\n",
       "      <td>22.480620</td>\n",
       "      <td>23.543890</td>\n",
       "      <td>11.802580</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-02</th>\n",
       "      <td>0.534572</td>\n",
       "      <td>27.586210</td>\n",
       "      <td>19.475650</td>\n",
       "      <td>21.735600</td>\n",
       "      <td>11.228810</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-03</th>\n",
       "      <td>0.513710</td>\n",
       "      <td>26.666670</td>\n",
       "      <td>19.926200</td>\n",
       "      <td>16.771650</td>\n",
       "      <td>10.251050</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-04</th>\n",
       "      <td>0.480641</td>\n",
       "      <td>25.806450</td>\n",
       "      <td>17.375890</td>\n",
       "      <td>14.154550</td>\n",
       "      <td>10.208330</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-05</th>\n",
       "      <td>0.520343</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>18.439720</td>\n",
       "      <td>11.192930</td>\n",
       "      <td>9.465020</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-06</th>\n",
       "      <td>0.561429</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>17.192980</td>\n",
       "      <td>8.992805</td>\n",
       "      <td>9.387755</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-07</th>\n",
       "      <td>0.631441</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>17.832170</td>\n",
       "      <td>4.175365</td>\n",
       "      <td>9.716599</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-08</th>\n",
       "      <td>0.846624</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>18.881120</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8.600000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-09</th>\n",
       "      <td>0.814557</td>\n",
       "      <td>24.242420</td>\n",
       "      <td>20.890410</td>\n",
       "      <td>-4.471808</td>\n",
       "      <td>7.905138</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-10</th>\n",
       "      <td>0.809605</td>\n",
       "      <td>20.588240</td>\n",
       "      <td>18.666670</td>\n",
       "      <td>-5.623788</td>\n",
       "      <td>7.436399</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-11</th>\n",
       "      <td>0.774787</td>\n",
       "      <td>17.142860</td>\n",
       "      <td>19.607840</td>\n",
       "      <td>-4.839765</td>\n",
       "      <td>7.378641</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975-12</th>\n",
       "      <td>0.839729</td>\n",
       "      <td>17.142860</td>\n",
       "      <td>19.488820</td>\n",
       "      <td>-6.175299</td>\n",
       "      <td>6.936416</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1976-01</th>\n",
       "      <td>0.931620</td>\n",
       "      <td>16.666670</td>\n",
       "      <td>20.886080</td>\n",
       "      <td>-8.565737</td>\n",
       "      <td>6.717850</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1976-02</th>\n",
       "      <td>0.993883</td>\n",
       "      <td>16.216220</td>\n",
       "      <td>22.257050</td>\n",
       "      <td>-10.726180</td>\n",
       "      <td>6.285714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-06</th>\n",
       "      <td>1.358814</td>\n",
       "      <td>8.600918</td>\n",
       "      <td>3.454303</td>\n",
       "      <td>6.130268</td>\n",
       "      <td>0.997326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-07</th>\n",
       "      <td>1.454600</td>\n",
       "      <td>8.960837</td>\n",
       "      <td>3.207983</td>\n",
       "      <td>6.463878</td>\n",
       "      <td>0.827139</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-08</th>\n",
       "      <td>1.411904</td>\n",
       "      <td>8.096013</td>\n",
       "      <td>2.793067</td>\n",
       "      <td>5.303030</td>\n",
       "      <td>1.062875</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-09</th>\n",
       "      <td>1.522881</td>\n",
       "      <td>7.270965</td>\n",
       "      <td>3.073888</td>\n",
       "      <td>4.135338</td>\n",
       "      <td>1.463784</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-10</th>\n",
       "      <td>1.556048</td>\n",
       "      <td>6.483710</td>\n",
       "      <td>3.306737</td>\n",
       "      <td>3.345725</td>\n",
       "      <td>1.635988</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-11</th>\n",
       "      <td>1.607762</td>\n",
       "      <td>5.966340</td>\n",
       "      <td>3.579051</td>\n",
       "      <td>2.592592</td>\n",
       "      <td>1.692537</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12</th>\n",
       "      <td>1.410190</td>\n",
       "      <td>5.747127</td>\n",
       "      <td>3.024636</td>\n",
       "      <td>2.230483</td>\n",
       "      <td>2.074622</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01</th>\n",
       "      <td>1.486925</td>\n",
       "      <td>5.470340</td>\n",
       "      <td>3.494580</td>\n",
       "      <td>1.858736</td>\n",
       "      <td>2.500042</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02</th>\n",
       "      <td>1.455253</td>\n",
       "      <td>5.185071</td>\n",
       "      <td>3.829649</td>\n",
       "      <td>2.621723</td>\n",
       "      <td>2.737958</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-03</th>\n",
       "      <td>1.413870</td>\n",
       "      <td>4.692643</td>\n",
       "      <td>3.612121</td>\n",
       "      <td>2.611940</td>\n",
       "      <td>2.380612</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-04</th>\n",
       "      <td>1.358942</td>\n",
       "      <td>4.661792</td>\n",
       "      <td>4.172417</td>\n",
       "      <td>2.214022</td>\n",
       "      <td>2.199690</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-05</th>\n",
       "      <td>1.332045</td>\n",
       "      <td>4.365290</td>\n",
       "      <td>4.332685</td>\n",
       "      <td>1.090909</td>\n",
       "      <td>1.874878</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-06</th>\n",
       "      <td>1.242455</td>\n",
       "      <td>3.990044</td>\n",
       "      <td>4.368815</td>\n",
       "      <td>1.083032</td>\n",
       "      <td>1.633488</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-07</th>\n",
       "      <td>1.325250</td>\n",
       "      <td>3.399115</td>\n",
       "      <td>3.875350</td>\n",
       "      <td>1.785714</td>\n",
       "      <td>1.727978</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-08</th>\n",
       "      <td>1.335409</td>\n",
       "      <td>3.869025</td>\n",
       "      <td>3.820027</td>\n",
       "      <td>2.517986</td>\n",
       "      <td>1.938974</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-09</th>\n",
       "      <td>1.328024</td>\n",
       "      <td>3.968971</td>\n",
       "      <td>3.723786</td>\n",
       "      <td>2.888087</td>\n",
       "      <td>2.232964</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10</th>\n",
       "      <td>1.261205</td>\n",
       "      <td>4.046722</td>\n",
       "      <td>3.583088</td>\n",
       "      <td>3.237410</td>\n",
       "      <td>2.041129</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11</th>\n",
       "      <td>1.259505</td>\n",
       "      <td>4.117425</td>\n",
       "      <td>3.304803</td>\n",
       "      <td>3.971119</td>\n",
       "      <td>2.202583</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-12</th>\n",
       "      <td>1.231900</td>\n",
       "      <td>4.085457</td>\n",
       "      <td>3.606661</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.109082</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01</th>\n",
       "      <td>1.240905</td>\n",
       "      <td>3.672924</td>\n",
       "      <td>3.251524</td>\n",
       "      <td>5.109489</td>\n",
       "      <td>2.070508</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02</th>\n",
       "      <td>1.209011</td>\n",
       "      <td>3.372025</td>\n",
       "      <td>3.181535</td>\n",
       "      <td>4.744525</td>\n",
       "      <td>2.211795</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03</th>\n",
       "      <td>1.197571</td>\n",
       "      <td>3.136882</td>\n",
       "      <td>3.400406</td>\n",
       "      <td>4.363636</td>\n",
       "      <td>2.359711</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-04</th>\n",
       "      <td>1.172529</td>\n",
       "      <td>3.129549</td>\n",
       "      <td>3.413076</td>\n",
       "      <td>3.971119</td>\n",
       "      <td>2.462744</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05</th>\n",
       "      <td>1.203136</td>\n",
       "      <td>3.158812</td>\n",
       "      <td>3.229062</td>\n",
       "      <td>3.956835</td>\n",
       "      <td>2.801012</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-06</th>\n",
       "      <td>1.161443</td>\n",
       "      <td>3.198665</td>\n",
       "      <td>3.122109</td>\n",
       "      <td>3.928571</td>\n",
       "      <td>2.871548</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-07</th>\n",
       "      <td>1.110571</td>\n",
       "      <td>3.120464</td>\n",
       "      <td>3.184615</td>\n",
       "      <td>5.614035</td>\n",
       "      <td>2.949515</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08</th>\n",
       "      <td>1.031822</td>\n",
       "      <td>3.101674</td>\n",
       "      <td>3.202217</td>\n",
       "      <td>5.614035</td>\n",
       "      <td>2.699180</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-09</th>\n",
       "      <td>0.999289</td>\n",
       "      <td>3.223470</td>\n",
       "      <td>2.882841</td>\n",
       "      <td>5.614035</td>\n",
       "      <td>2.276972</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10</th>\n",
       "      <td>1.148500</td>\n",
       "      <td>3.331643</td>\n",
       "      <td>3.159351</td>\n",
       "      <td>5.226481</td>\n",
       "      <td>2.522470</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-11</th>\n",
       "      <td>1.121786</td>\n",
       "      <td>3.267785</td>\n",
       "      <td>3.229766</td>\n",
       "      <td>4.861111</td>\n",
       "      <td>2.176601</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>543 rows × 126 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         coffee price    COL CPI    IDN CPI    IND CPI    USA CPI  BRA Flood  \\\n",
       "1973-09      0.652700  22.727270  44.000000  19.250780   7.363420        0.0   \n",
       "1973-10      0.659552  22.727270  42.307690  21.554400   7.801418        0.0   \n",
       "1973-11      0.646810  27.272730  30.635840  23.298970   8.254717        0.0   \n",
       "1973-12      0.670937  27.272730  27.173910  23.814430   8.705882        0.0   \n",
       "1974-01      0.716991  26.086960  42.541440  25.670100   9.389671        0.0   \n",
       "1974-02      0.756550  26.086960  46.703300  25.304880  10.023310        0.0   \n",
       "1974-03      0.732176  25.000000  47.282610  27.254510  10.392610        1.0   \n",
       "1974-04      0.714795  24.000000  49.206350  28.011750  10.091740        0.0   \n",
       "1974-05      0.730982  23.076920  49.206350  28.964860  10.706150        0.0   \n",
       "1974-06      0.725355  23.076920  43.216080  29.182160  10.859730        0.0   \n",
       "1974-07      0.705859  18.518520  36.842110  28.074870  11.512410        0.0   \n",
       "1974-08      0.633591  23.076920  36.842110  29.973710  10.864750        0.0   \n",
       "1974-09      0.544295  22.222220  35.185180  34.642230  11.946900        0.0   \n",
       "1974-10      0.575364  25.925930  35.135140  31.884060  12.061400        0.0   \n",
       "1974-11      0.602961  25.000000  35.398230  27.842810  12.200440        0.0   \n",
       "1974-12      0.594642  25.000000  33.760680  25.395500  12.337660        0.0   \n",
       "1975-01      0.560886  24.137930  22.480620  23.543890  11.802580        0.0   \n",
       "1975-02      0.534572  27.586210  19.475650  21.735600  11.228810        0.0   \n",
       "1975-03      0.513710  26.666670  19.926200  16.771650  10.251050        0.0   \n",
       "1975-04      0.480641  25.806450  17.375890  14.154550  10.208330        0.0   \n",
       "1975-05      0.520343  25.000000  18.439720  11.192930   9.465020        0.0   \n",
       "1975-06      0.561429  25.000000  17.192980   8.992805   9.387755        0.0   \n",
       "1975-07      0.631441  25.000000  17.832170   4.175365   9.716599        1.0   \n",
       "1975-08      0.846624  25.000000  18.881120   0.000000   8.600000        0.0   \n",
       "1975-09      0.814557  24.242420  20.890410  -4.471808   7.905138        0.0   \n",
       "1975-10      0.809605  20.588240  18.666670  -5.623788   7.436399        0.0   \n",
       "1975-11      0.774787  17.142860  19.607840  -4.839765   7.378641        0.0   \n",
       "1975-12      0.839729  17.142860  19.488820  -6.175299   6.936416        0.0   \n",
       "1976-01      0.931620  16.666670  20.886080  -8.565737   6.717850        0.0   \n",
       "1976-02      0.993883  16.216220  22.257050 -10.726180   6.285714        0.0   \n",
       "...               ...        ...        ...        ...        ...        ...   \n",
       "2016-06      1.358814   8.600918   3.454303   6.130268   0.997326        0.0   \n",
       "2016-07      1.454600   8.960837   3.207983   6.463878   0.827139        0.0   \n",
       "2016-08      1.411904   8.096013   2.793067   5.303030   1.062875        0.0   \n",
       "2016-09      1.522881   7.270965   3.073888   4.135338   1.463784        0.0   \n",
       "2016-10      1.556048   6.483710   3.306737   3.345725   1.635988        0.0   \n",
       "2016-11      1.607762   5.966340   3.579051   2.592592   1.692537        0.0   \n",
       "2016-12      1.410190   5.747127   3.024636   2.230483   2.074622        0.0   \n",
       "2017-01      1.486925   5.470340   3.494580   1.858736   2.500042        0.0   \n",
       "2017-02      1.455253   5.185071   3.829649   2.621723   2.737958        0.0   \n",
       "2017-03      1.413870   4.692643   3.612121   2.611940   2.380612        0.0   \n",
       "2017-04      1.358942   4.661792   4.172417   2.214022   2.199690        0.0   \n",
       "2017-05      1.332045   4.365290   4.332685   1.090909   1.874878        1.0   \n",
       "2017-06      1.242455   3.990044   4.368815   1.083032   1.633488        1.0   \n",
       "2017-07      1.325250   3.399115   3.875350   1.785714   1.727978        0.0   \n",
       "2017-08      1.335409   3.869025   3.820027   2.517986   1.938974        0.0   \n",
       "2017-09      1.328024   3.968971   3.723786   2.888087   2.232964        0.0   \n",
       "2017-10      1.261205   4.046722   3.583088   3.237410   2.041129        0.0   \n",
       "2017-11      1.259505   4.117425   3.304803   3.971119   2.202583        0.0   \n",
       "2017-12      1.231900   4.085457   3.606661   4.000000   2.109082        0.0   \n",
       "2018-01      1.240905   3.672924   3.251524   5.109489   2.070508        0.0   \n",
       "2018-02      1.209011   3.372025   3.181535   4.744525   2.211795        1.0   \n",
       "2018-03      1.197571   3.136882   3.400406   4.363636   2.359711        1.0   \n",
       "2018-04      1.172529   3.129549   3.413076   3.971119   2.462744        0.0   \n",
       "2018-05      1.203136   3.158812   3.229062   3.956835   2.801012        0.0   \n",
       "2018-06      1.161443   3.198665   3.122109   3.928571   2.871548        0.0   \n",
       "2018-07      1.110571   3.120464   3.184615   5.614035   2.949515        0.0   \n",
       "2018-08      1.031822   3.101674   3.202217   5.614035   2.699180        0.0   \n",
       "2018-09      0.999289   3.223470   2.882841   5.614035   2.276972        0.0   \n",
       "2018-10      1.148500   3.331643   3.159351   5.226481   2.522470        0.0   \n",
       "2018-11      1.121786   3.267785   3.229766   4.861111   2.176601        0.0   \n",
       "\n",
       "         BRA Landslide  BRA Epidemic  BRA Extreme temperature   BRA Storm  \\\n",
       "1973-09            0.0           0.0                       0.0        0.0   \n",
       "1973-10            0.0           0.0                       0.0        0.0   \n",
       "1973-11            0.0           0.0                       0.0        0.0   \n",
       "1973-12            0.0           0.0                       0.0        0.0   \n",
       "1974-01            0.0           1.0                       0.0        0.0   \n",
       "1974-02            0.0           0.0                       0.0        0.0   \n",
       "1974-03            0.0           0.0                       0.0        0.0   \n",
       "1974-04            0.0           0.0                       0.0        0.0   \n",
       "1974-05            1.0           0.0                       0.0        0.0   \n",
       "1974-06            0.0           0.0                       0.0        0.0   \n",
       "1974-07            0.0           0.0                       0.0        0.0   \n",
       "1974-08            0.0           0.0                       0.0        0.0   \n",
       "1974-09            0.0           0.0                       0.0        0.0   \n",
       "1974-10            1.0           0.0                       0.0        0.0   \n",
       "1974-11            0.0           0.0                       0.0        0.0   \n",
       "1974-12            0.0           0.0                       0.0        0.0   \n",
       "1975-01            0.0           0.0                       0.0        0.0   \n",
       "1975-02            0.0           0.0                       0.0        0.0   \n",
       "1975-03            0.0           0.0                       0.0        0.0   \n",
       "1975-04            0.0           0.0                       0.0        0.0   \n",
       "1975-05            0.0           0.0                       0.0        0.0   \n",
       "1975-06            0.0           0.0                       0.0        0.0   \n",
       "1975-07            0.0           1.0                       1.0        0.0   \n",
       "1975-08            0.0           0.0                       0.0        0.0   \n",
       "1975-09            0.0           0.0                       0.0        0.0   \n",
       "1975-10            0.0           0.0                       0.0        0.0   \n",
       "1975-11            0.0           0.0                       0.0        0.0   \n",
       "1975-12            0.0           0.0                       1.0        0.0   \n",
       "1976-01            0.0           0.0                       0.0        0.0   \n",
       "1976-02            0.0           0.0                       0.0        0.0   \n",
       "...                ...           ...                       ...        ...   \n",
       "2016-06            0.0           0.0                       0.0        0.0   \n",
       "2016-07            0.0           0.0                       0.0        0.0   \n",
       "2016-08            0.0           0.0                       0.0        0.0   \n",
       "2016-09            0.0           0.0                       0.0        0.0   \n",
       "2016-10            0.0           0.0                       0.0        0.0   \n",
       "2016-11            0.0           0.0                       0.0        0.0   \n",
       "2016-12            0.0           1.0                       0.0        0.0   \n",
       "2017-01            0.0           1.0                       0.0        0.0   \n",
       "2017-02            0.0           1.0                       0.0        0.0   \n",
       "2017-03            0.0           1.0                       0.0        0.0   \n",
       "2017-04            0.0           1.0                       0.0        0.0   \n",
       "2017-05            0.0           0.0                       0.0        0.0   \n",
       "2017-06            0.0           0.0                       0.0        0.0   \n",
       "2017-07            0.0           1.0                       0.0        0.0   \n",
       "2017-08            0.0           1.0                       0.0        0.0   \n",
       "2017-09            0.0           1.0                       0.0        0.0   \n",
       "2017-10            0.0           1.0                       0.0        0.0   \n",
       "2017-11            0.0           1.0                       0.0        0.0   \n",
       "2017-12            0.0           1.0                       0.0        0.0   \n",
       "2018-01            0.0           1.0                       0.0        0.0   \n",
       "2018-02            0.0           1.0                       0.0        0.0   \n",
       "2018-03            0.0           0.0                       0.0        0.0   \n",
       "2018-04            0.0           0.0                       0.0        0.0   \n",
       "2018-05            0.0           0.0                       0.0        0.0   \n",
       "2018-06            0.0           0.0                       0.0        0.0   \n",
       "2018-07            0.0           0.0                       0.0        0.0   \n",
       "2018-08            0.0           0.0                       0.0        0.0   \n",
       "2018-09            0.0           0.0                       0.0        0.0   \n",
       "2018-10            0.0           0.0                       0.0        0.0   \n",
       "2018-11            1.0           0.0                       0.0        0.0   \n",
       "\n",
       "                 ...            VNM Landslide  VNM Epidemic  \\\n",
       "1973-09          ...                      0.0           0.0   \n",
       "1973-10          ...                      0.0           0.0   \n",
       "1973-11          ...                      0.0           0.0   \n",
       "1973-12          ...                      0.0           0.0   \n",
       "1974-01          ...                      0.0           0.0   \n",
       "1974-02          ...                      0.0           0.0   \n",
       "1974-03          ...                      0.0           0.0   \n",
       "1974-04          ...                      0.0           0.0   \n",
       "1974-05          ...                      0.0           0.0   \n",
       "1974-06          ...                      0.0           0.0   \n",
       "1974-07          ...                      0.0           0.0   \n",
       "1974-08          ...                      0.0           0.0   \n",
       "1974-09          ...                      0.0           0.0   \n",
       "1974-10          ...                      0.0           0.0   \n",
       "1974-11          ...                      0.0           0.0   \n",
       "1974-12          ...                      0.0           0.0   \n",
       "1975-01          ...                      0.0           0.0   \n",
       "1975-02          ...                      0.0           0.0   \n",
       "1975-03          ...                      0.0           0.0   \n",
       "1975-04          ...                      0.0           0.0   \n",
       "1975-05          ...                      0.0           0.0   \n",
       "1975-06          ...                      0.0           0.0   \n",
       "1975-07          ...                      0.0           0.0   \n",
       "1975-08          ...                      0.0           0.0   \n",
       "1975-09          ...                      0.0           0.0   \n",
       "1975-10          ...                      0.0           0.0   \n",
       "1975-11          ...                      0.0           0.0   \n",
       "1975-12          ...                      0.0           0.0   \n",
       "1976-01          ...                      0.0           0.0   \n",
       "1976-02          ...                      0.0           0.0   \n",
       "...              ...                      ...           ...   \n",
       "2016-06          ...                      0.0           1.0   \n",
       "2016-07          ...                      0.0           1.0   \n",
       "2016-08          ...                      0.0           1.0   \n",
       "2016-09          ...                      0.0           1.0   \n",
       "2016-10          ...                      0.0           0.0   \n",
       "2016-11          ...                      0.0           0.0   \n",
       "2016-12          ...                      0.0           0.0   \n",
       "2017-01          ...                      0.0           0.0   \n",
       "2017-02          ...                      0.0           0.0   \n",
       "2017-03          ...                      0.0           0.0   \n",
       "2017-04          ...                      0.0           0.0   \n",
       "2017-05          ...                      0.0           0.0   \n",
       "2017-06          ...                      0.0           0.0   \n",
       "2017-07          ...                      0.0           0.0   \n",
       "2017-08          ...                      0.0           0.0   \n",
       "2017-09          ...                      0.0           0.0   \n",
       "2017-10          ...                      0.0           0.0   \n",
       "2017-11          ...                      0.0           0.0   \n",
       "2017-12          ...                      0.0           0.0   \n",
       "2018-01          ...                      0.0           0.0   \n",
       "2018-02          ...                      0.0           0.0   \n",
       "2018-03          ...                      0.0           0.0   \n",
       "2018-04          ...                      0.0           0.0   \n",
       "2018-05          ...                      0.0           0.0   \n",
       "2018-06          ...                      0.0           0.0   \n",
       "2018-07          ...                      0.0           0.0   \n",
       "2018-08          ...                      0.0           0.0   \n",
       "2018-09          ...                      0.0           0.0   \n",
       "2018-10          ...                      0.0           0.0   \n",
       "2018-11          ...                      0.0           0.0   \n",
       "\n",
       "         VNM Extreme temperature   VNM Storm  VNM Drought  VNM Earthquake  \\\n",
       "1973-09                       0.0        0.0          0.0             0.0   \n",
       "1973-10                       0.0        0.0          0.0             0.0   \n",
       "1973-11                       0.0        1.0          0.0             0.0   \n",
       "1973-12                       0.0        0.0          0.0             0.0   \n",
       "1974-01                       0.0        0.0          0.0             0.0   \n",
       "1974-02                       0.0        0.0          0.0             0.0   \n",
       "1974-03                       0.0        0.0          0.0             0.0   \n",
       "1974-04                       0.0        0.0          0.0             0.0   \n",
       "1974-05                       0.0        0.0          0.0             0.0   \n",
       "1974-06                       0.0        0.0          0.0             0.0   \n",
       "1974-07                       0.0        0.0          0.0             0.0   \n",
       "1974-08                       0.0        0.0          0.0             0.0   \n",
       "1974-09                       0.0        0.0          0.0             0.0   \n",
       "1974-10                       0.0        0.0          0.0             0.0   \n",
       "1974-11                       0.0        0.0          0.0             0.0   \n",
       "1974-12                       0.0        0.0          0.0             0.0   \n",
       "1975-01                       0.0        0.0          0.0             0.0   \n",
       "1975-02                       0.0        0.0          0.0             0.0   \n",
       "1975-03                       0.0        0.0          0.0             0.0   \n",
       "1975-04                       0.0        0.0          0.0             0.0   \n",
       "1975-05                       0.0        0.0          0.0             0.0   \n",
       "1975-06                       0.0        0.0          0.0             0.0   \n",
       "1975-07                       0.0        0.0          0.0             0.0   \n",
       "1975-08                       0.0        0.0          0.0             0.0   \n",
       "1975-09                       0.0        0.0          0.0             0.0   \n",
       "1975-10                       0.0        0.0          0.0             0.0   \n",
       "1975-11                       0.0        0.0          0.0             0.0   \n",
       "1975-12                       0.0        0.0          0.0             0.0   \n",
       "1976-01                       0.0        0.0          0.0             0.0   \n",
       "1976-02                       0.0        0.0          0.0             0.0   \n",
       "...                           ...        ...          ...             ...   \n",
       "2016-06                       0.0        0.0          0.0             0.0   \n",
       "2016-07                       0.0        1.0          0.0             0.0   \n",
       "2016-08                       0.0        1.0          0.0             0.0   \n",
       "2016-09                       0.0        0.0          0.0             0.0   \n",
       "2016-10                       0.0        1.0          0.0             0.0   \n",
       "2016-11                       0.0        0.0          0.0             0.0   \n",
       "2016-12                       0.0        0.0          0.0             0.0   \n",
       "2017-01                       0.0        0.0          1.0             0.0   \n",
       "2017-02                       0.0        0.0          1.0             0.0   \n",
       "2017-03                       0.0        0.0          0.0             0.0   \n",
       "2017-04                       0.0        0.0          0.0             0.0   \n",
       "2017-05                       0.0        0.0          0.0             0.0   \n",
       "2017-06                       0.0        0.0          0.0             0.0   \n",
       "2017-07                       0.0        1.0          0.0             0.0   \n",
       "2017-08                       0.0        1.0          0.0             0.0   \n",
       "2017-09                       0.0        1.0          0.0             0.0   \n",
       "2017-10                       0.0        0.0          0.0             0.0   \n",
       "2017-11                       0.0        1.0          0.0             0.0   \n",
       "2017-12                       0.0        0.0          0.0             0.0   \n",
       "2018-01                       0.0        0.0          0.0             0.0   \n",
       "2018-02                       0.0        0.0          0.0             0.0   \n",
       "2018-03                       0.0        0.0          0.0             0.0   \n",
       "2018-04                       0.0        0.0          0.0             0.0   \n",
       "2018-05                       0.0        0.0          0.0             0.0   \n",
       "2018-06                       0.0        1.0          0.0             0.0   \n",
       "2018-07                       0.0        1.0          0.0             0.0   \n",
       "2018-08                       0.0        1.0          0.0             0.0   \n",
       "2018-09                       0.0        0.0          0.0             0.0   \n",
       "2018-10                       0.0        0.0          0.0             0.0   \n",
       "2018-11                       0.0        0.0          0.0             0.0   \n",
       "\n",
       "         VNM Insect infestation  VNM Wildfire  VNM Mass movement (dry)  \\\n",
       "1973-09                     0.0           0.0                      0.0   \n",
       "1973-10                     0.0           0.0                      0.0   \n",
       "1973-11                     0.0           0.0                      0.0   \n",
       "1973-12                     0.0           0.0                      0.0   \n",
       "1974-01                     0.0           0.0                      0.0   \n",
       "1974-02                     0.0           0.0                      0.0   \n",
       "1974-03                     0.0           0.0                      0.0   \n",
       "1974-04                     0.0           0.0                      0.0   \n",
       "1974-05                     0.0           0.0                      0.0   \n",
       "1974-06                     0.0           0.0                      0.0   \n",
       "1974-07                     0.0           0.0                      0.0   \n",
       "1974-08                     0.0           0.0                      0.0   \n",
       "1974-09                     0.0           0.0                      0.0   \n",
       "1974-10                     0.0           0.0                      0.0   \n",
       "1974-11                     0.0           0.0                      0.0   \n",
       "1974-12                     0.0           0.0                      0.0   \n",
       "1975-01                     0.0           0.0                      0.0   \n",
       "1975-02                     0.0           0.0                      0.0   \n",
       "1975-03                     0.0           0.0                      0.0   \n",
       "1975-04                     0.0           0.0                      0.0   \n",
       "1975-05                     0.0           0.0                      0.0   \n",
       "1975-06                     0.0           0.0                      0.0   \n",
       "1975-07                     0.0           0.0                      0.0   \n",
       "1975-08                     0.0           0.0                      0.0   \n",
       "1975-09                     0.0           0.0                      0.0   \n",
       "1975-10                     0.0           0.0                      0.0   \n",
       "1975-11                     0.0           0.0                      0.0   \n",
       "1975-12                     0.0           0.0                      0.0   \n",
       "1976-01                     0.0           0.0                      0.0   \n",
       "1976-02                     0.0           0.0                      0.0   \n",
       "...                         ...           ...                      ...   \n",
       "2016-06                     0.0           0.0                      0.0   \n",
       "2016-07                     0.0           0.0                      0.0   \n",
       "2016-08                     0.0           0.0                      0.0   \n",
       "2016-09                     0.0           0.0                      0.0   \n",
       "2016-10                     0.0           0.0                      0.0   \n",
       "2016-11                     0.0           0.0                      0.0   \n",
       "2016-12                     0.0           0.0                      0.0   \n",
       "2017-01                     0.0           0.0                      0.0   \n",
       "2017-02                     0.0           0.0                      0.0   \n",
       "2017-03                     0.0           0.0                      0.0   \n",
       "2017-04                     0.0           0.0                      0.0   \n",
       "2017-05                     0.0           0.0                      0.0   \n",
       "2017-06                     0.0           0.0                      0.0   \n",
       "2017-07                     0.0           0.0                      0.0   \n",
       "2017-08                     0.0           0.0                      0.0   \n",
       "2017-09                     0.0           0.0                      0.0   \n",
       "2017-10                     0.0           0.0                      0.0   \n",
       "2017-11                     0.0           0.0                      0.0   \n",
       "2017-12                     0.0           0.0                      0.0   \n",
       "2018-01                     0.0           0.0                      0.0   \n",
       "2018-02                     0.0           0.0                      0.0   \n",
       "2018-03                     0.0           0.0                      0.0   \n",
       "2018-04                     0.0           0.0                      0.0   \n",
       "2018-05                     0.0           0.0                      0.0   \n",
       "2018-06                     0.0           0.0                      0.0   \n",
       "2018-07                     0.0           0.0                      0.0   \n",
       "2018-08                     0.0           0.0                      0.0   \n",
       "2018-09                     0.0           0.0                      0.0   \n",
       "2018-10                     0.0           0.0                      0.0   \n",
       "2018-11                     0.0           0.0                      0.0   \n",
       "\n",
       "         VNM Volcanic activity  \n",
       "1973-09                    0.0  \n",
       "1973-10                    0.0  \n",
       "1973-11                    0.0  \n",
       "1973-12                    0.0  \n",
       "1974-01                    0.0  \n",
       "1974-02                    0.0  \n",
       "1974-03                    0.0  \n",
       "1974-04                    0.0  \n",
       "1974-05                    0.0  \n",
       "1974-06                    0.0  \n",
       "1974-07                    0.0  \n",
       "1974-08                    0.0  \n",
       "1974-09                    0.0  \n",
       "1974-10                    0.0  \n",
       "1974-11                    0.0  \n",
       "1974-12                    0.0  \n",
       "1975-01                    0.0  \n",
       "1975-02                    0.0  \n",
       "1975-03                    0.0  \n",
       "1975-04                    0.0  \n",
       "1975-05                    0.0  \n",
       "1975-06                    0.0  \n",
       "1975-07                    0.0  \n",
       "1975-08                    0.0  \n",
       "1975-09                    0.0  \n",
       "1975-10                    0.0  \n",
       "1975-11                    0.0  \n",
       "1975-12                    0.0  \n",
       "1976-01                    0.0  \n",
       "1976-02                    0.0  \n",
       "...                        ...  \n",
       "2016-06                    0.0  \n",
       "2016-07                    0.0  \n",
       "2016-08                    0.0  \n",
       "2016-09                    0.0  \n",
       "2016-10                    0.0  \n",
       "2016-11                    0.0  \n",
       "2016-12                    0.0  \n",
       "2017-01                    0.0  \n",
       "2017-02                    0.0  \n",
       "2017-03                    0.0  \n",
       "2017-04                    0.0  \n",
       "2017-05                    0.0  \n",
       "2017-06                    0.0  \n",
       "2017-07                    0.0  \n",
       "2017-08                    0.0  \n",
       "2017-09                    0.0  \n",
       "2017-10                    0.0  \n",
       "2017-11                    0.0  \n",
       "2017-12                    0.0  \n",
       "2018-01                    0.0  \n",
       "2018-02                    0.0  \n",
       "2018-03                    0.0  \n",
       "2018-04                    0.0  \n",
       "2018-05                    0.0  \n",
       "2018-06                    0.0  \n",
       "2018-07                    0.0  \n",
       "2018-08                    0.0  \n",
       "2018-09                    0.0  \n",
       "2018-10                    0.0  \n",
       "2018-11                    0.0  \n",
       "\n",
       "[543 rows x 126 columns]"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df= df.join(data, how='left', lsuffix='_left', rsuffix='_right')\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv(\"data/data/data_for_renges.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Trying simple prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see Linear Regression works bad on this data.\n",
    "But PCA seems helpful here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = df.iloc[:, 1::]\n",
    "y = df.iloc[:, 0]\n",
    "train_data, test_data, train_y, test_y = train_test_split(x, y, test_size = 0.2, random_state = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "def lin_reg(train_data, test_data, train_y, test_y):\n",
    "    start_time = time.time()\n",
    "    log_reg = LinearRegression().fit(train_data, train_y)\n",
    "    lr_train_predicted = log_reg.predict(train_data)\n",
    "    lr_test_predicted = log_reg.predict(test_data)\n",
    "\n",
    "    perf = time.time() - start_time\n",
    "    return lr_train_predicted, lr_test_predicted, perf\n",
    "pred_train, pred_test, performance = lin_reg(train_data, test_data, train_y, test_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f6812144128>]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(pred_train)\n",
    "plt.plot(train_y)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "scaler.fit(train_data)\n",
    "train = scaler.transform(train_data)\n",
    "test = scaler.transform(test_data)\n",
    "def pca_(n):\n",
    "    pca = PCA(n_components=n)\n",
    "    pca.fit(train)\n",
    "    return pca"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "pca50 = pca_(75)\n",
    "train_ = pca50.transform(train)\n",
    "test_ = pca50.transform(test)\n",
    "_train, _test, performance = lin_reg(train_, test_, train_y, test_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f6812370e10>]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(train_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f6811b7d7f0>]"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f6811b44630>]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(pred_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f6811aef898>]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(_train)\n",
    "plt.plot(train_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.29939936574500603"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r2_score(train_y, _train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3894856157727974"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r2_score(test_y, pred_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4833867277874738"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r2_score(train_y, pred_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "PCA here in not usable : it is bad to do dimentionality reduction with 126 dimentions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "ada = AdaBoostRegressor().fit(train_data, train_y)\n",
    "train_predicted = ada.predict(train_data)\n",
    "test_predicted = ada.predict(test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5549474673544836"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ada.score(train_data, train_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.44346267772460063"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ada.score(test_data, test_y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Without weather it is hard to make good prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Coffee price - mentions in media dependency**\n",
    "\n",
    "I have done an experiment which aims to discover dependency between mentions a word “coffee” and similar words in media.\n",
    "The experiment didn’t show any results because of a leak of data.\n",
    "Where to find the data:\n",
    "\n",
    "    -scrap the reuters.com archive(https://www.reuters.com/resources/archive/) (scraped less than 30 days before the site was down)\n",
    "    \n",
    "    -use a framework to access and analyze Global Database of Events, Language, and Tone (GDELT) 1.0 or 2.0  (github.com/linwoodc3/gdeltPyR)\n",
    "    \n",
    "Why experiment was not a successful ?\n",
    "For this task not enough scraped data from reuters.com,\n",
    "search for keywords (‘coffee’ and similar) in Global Database of Events didn’t show any results, even on International Coffee Day. Data collected there has some specification on events, not on the media.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data description\n",
    "\n",
    "\n",
    "- **GDP** - GDP (current US)\n",
    "- **GDP per capita** - (current US)\n",
    "- **Agricultural land** - (% of land area)\n",
    "- **Rural population**\n",
    "- **ICO** - mounthly\n",
    "- **Disaster** - Geophysical(Volcanic activity, Mass movement(dry), Landslide, Earthquake); Climatological (Wildfire, Drought); Biological (Epidemia, Insect infestation); Meteorogical (Extreme temperature, Storm, Fog)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data sources :\n",
    "\n",
    "- www.ico.org - International Coffe Organization\n",
    "- www.emdat.be - Université Catholique de Louvain's disaster database\n",
    "- https://data.oecd.org/\n",
    "- https://data.worldbank.org/\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Results:\n",
    "\n",
    "- clean data in csv format (\"data/data/clean_data.csv\")\n",
    "- monthly data in csv format (\"data/data/data_for_renges.csv\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}