notebooks/eda-prelim.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "raw = pd.read_csv('https://raw.githubusercontent.com/learn-co-students/dc-ds-060120/master/mod-4/week-3/Mod_4_Project/time-series/zillow_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "raw = raw.drop('RegionID', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "raw = raw.drop('SizeRank', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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>RegionName</th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>Metro</th>\n",
       "      <th>CountyName</th>\n",
       "      <th>1996-04</th>\n",
       "      <th>1996-05</th>\n",
       "      <th>1996-06</th>\n",
       "      <th>1996-07</th>\n",
       "      <th>1996-08</th>\n",
       "      <th>...</th>\n",
       "      <th>2017-07</th>\n",
       "      <th>2017-08</th>\n",
       "      <th>2017-09</th>\n",
       "      <th>2017-10</th>\n",
       "      <th>2017-11</th>\n",
       "      <th>2017-12</th>\n",
       "      <th>2018-01</th>\n",
       "      <th>2018-02</th>\n",
       "      <th>2018-03</th>\n",
       "      <th>2018-04</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>60657</td>\n",
       "      <td>Chicago</td>\n",
       "      <td>IL</td>\n",
       "      <td>Chicago</td>\n",
       "      <td>Cook</td>\n",
       "      <td>334200.0</td>\n",
       "      <td>335400.0</td>\n",
       "      <td>336500.0</td>\n",
       "      <td>337600.0</td>\n",
       "      <td>338500.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1005500</td>\n",
       "      <td>1007500</td>\n",
       "      <td>1007800</td>\n",
       "      <td>1009600</td>\n",
       "      <td>1013300</td>\n",
       "      <td>1018700</td>\n",
       "      <td>1024400</td>\n",
       "      <td>1030700</td>\n",
       "      <td>1033800</td>\n",
       "      <td>1030600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>75070</td>\n",
       "      <td>McKinney</td>\n",
       "      <td>TX</td>\n",
       "      <td>Dallas-Fort Worth</td>\n",
       "      <td>Collin</td>\n",
       "      <td>235700.0</td>\n",
       "      <td>236900.0</td>\n",
       "      <td>236700.0</td>\n",
       "      <td>235400.0</td>\n",
       "      <td>233300.0</td>\n",
       "      <td>...</td>\n",
       "      <td>308000</td>\n",
       "      <td>310000</td>\n",
       "      <td>312500</td>\n",
       "      <td>314100</td>\n",
       "      <td>315000</td>\n",
       "      <td>316600</td>\n",
       "      <td>318100</td>\n",
       "      <td>319600</td>\n",
       "      <td>321100</td>\n",
       "      <td>321800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>77494</td>\n",
       "      <td>Katy</td>\n",
       "      <td>TX</td>\n",
       "      <td>Houston</td>\n",
       "      <td>Harris</td>\n",
       "      <td>210400.0</td>\n",
       "      <td>212200.0</td>\n",
       "      <td>212200.0</td>\n",
       "      <td>210700.0</td>\n",
       "      <td>208300.0</td>\n",
       "      <td>...</td>\n",
       "      <td>321000</td>\n",
       "      <td>320600</td>\n",
       "      <td>320200</td>\n",
       "      <td>320400</td>\n",
       "      <td>320800</td>\n",
       "      <td>321200</td>\n",
       "      <td>321200</td>\n",
       "      <td>323000</td>\n",
       "      <td>326900</td>\n",
       "      <td>329900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>60614</td>\n",
       "      <td>Chicago</td>\n",
       "      <td>IL</td>\n",
       "      <td>Chicago</td>\n",
       "      <td>Cook</td>\n",
       "      <td>498100.0</td>\n",
       "      <td>500900.0</td>\n",
       "      <td>503100.0</td>\n",
       "      <td>504600.0</td>\n",
       "      <td>505500.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1289800</td>\n",
       "      <td>1287700</td>\n",
       "      <td>1287400</td>\n",
       "      <td>1291500</td>\n",
       "      <td>1296600</td>\n",
       "      <td>1299000</td>\n",
       "      <td>1302700</td>\n",
       "      <td>1306400</td>\n",
       "      <td>1308500</td>\n",
       "      <td>1307000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79936</td>\n",
       "      <td>El Paso</td>\n",
       "      <td>TX</td>\n",
       "      <td>El Paso</td>\n",
       "      <td>El Paso</td>\n",
       "      <td>77300.0</td>\n",
       "      <td>77300.0</td>\n",
       "      <td>77300.0</td>\n",
       "      <td>77300.0</td>\n",
       "      <td>77400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>119100</td>\n",
       "      <td>119400</td>\n",
       "      <td>120000</td>\n",
       "      <td>120300</td>\n",
       "      <td>120300</td>\n",
       "      <td>120300</td>\n",
       "      <td>120300</td>\n",
       "      <td>120500</td>\n",
       "      <td>121000</td>\n",
       "      <td>121500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 270 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   RegionName      City State              Metro CountyName   1996-04  \\\n",
       "0       60657   Chicago    IL            Chicago       Cook  334200.0   \n",
       "1       75070  McKinney    TX  Dallas-Fort Worth     Collin  235700.0   \n",
       "2       77494      Katy    TX            Houston     Harris  210400.0   \n",
       "3       60614   Chicago    IL            Chicago       Cook  498100.0   \n",
       "4       79936   El Paso    TX            El Paso    El Paso   77300.0   \n",
       "\n",
       "    1996-05   1996-06   1996-07   1996-08  ...  2017-07  2017-08  2017-09  \\\n",
       "0  335400.0  336500.0  337600.0  338500.0  ...  1005500  1007500  1007800   \n",
       "1  236900.0  236700.0  235400.0  233300.0  ...   308000   310000   312500   \n",
       "2  212200.0  212200.0  210700.0  208300.0  ...   321000   320600   320200   \n",
       "3  500900.0  503100.0  504600.0  505500.0  ...  1289800  1287700  1287400   \n",
       "4   77300.0   77300.0   77300.0   77400.0  ...   119100   119400   120000   \n",
       "\n",
       "   2017-10  2017-11  2017-12  2018-01  2018-02  2018-03  2018-04  \n",
       "0  1009600  1013300  1018700  1024400  1030700  1033800  1030600  \n",
       "1   314100   315000   316600   318100   319600   321100   321800  \n",
       "2   320400   320800   321200   321200   323000   326900   329900  \n",
       "3  1291500  1296600  1299000  1302700  1306400  1308500  1307000  \n",
       "4   120300   120300   120300   120300   120500   121000   121500  \n",
       "\n",
       "[5 rows x 270 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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>RegionName</th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>Metro</th>\n",
       "      <th>CountyName</th>\n",
       "      <th>time</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>60657</td>\n",
       "      <td>Chicago</td>\n",
       "      <td>IL</td>\n",
       "      <td>Chicago</td>\n",
       "      <td>Cook</td>\n",
       "      <td>1996-04-01</td>\n",
       "      <td>334200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>75070</td>\n",
       "      <td>McKinney</td>\n",
       "      <td>TX</td>\n",
       "      <td>Dallas-Fort Worth</td>\n",
       "      <td>Collin</td>\n",
       "      <td>1996-04-01</td>\n",
       "      <td>235700.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>77494</td>\n",
       "      <td>Katy</td>\n",
       "      <td>TX</td>\n",
       "      <td>Houston</td>\n",
       "      <td>Harris</td>\n",
       "      <td>1996-04-01</td>\n",
       "      <td>210400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>60614</td>\n",
       "      <td>Chicago</td>\n",
       "      <td>IL</td>\n",
       "      <td>Chicago</td>\n",
       "      <td>Cook</td>\n",
       "      <td>1996-04-01</td>\n",
       "      <td>498100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79936</td>\n",
       "      <td>El Paso</td>\n",
       "      <td>TX</td>\n",
       "      <td>El Paso</td>\n",
       "      <td>El Paso</td>\n",
       "      <td>1996-04-01</td>\n",
       "      <td>77300.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   RegionName      City State              Metro CountyName       time  \\\n",
       "0       60657   Chicago    IL            Chicago       Cook 1996-04-01   \n",
       "1       75070  McKinney    TX  Dallas-Fort Worth     Collin 1996-04-01   \n",
       "2       77494      Katy    TX            Houston     Harris 1996-04-01   \n",
       "3       60614   Chicago    IL            Chicago       Cook 1996-04-01   \n",
       "4       79936   El Paso    TX            El Paso    El Paso 1996-04-01   \n",
       "\n",
       "      value  \n",
       "0  334200.0  \n",
       "1  235700.0  \n",
       "2  210400.0  \n",
       "3  498100.0  \n",
       "4   77300.0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "melted = pd.melt(raw, id_vars=['RegionName', 'City', 'State', 'Metro', 'CountyName'], var_name='time')\n",
    "melted['time'] = pd.to_datetime(melted['time'], infer_datetime_format=True)\n",
    "melted = melted.dropna(subset=['value'])\n",
    "melted.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RegionName         0\n",
       "City               0\n",
       "State              0\n",
       "Metro         236023\n",
       "CountyName         0\n",
       "time               0\n",
       "value              0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "melted.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
<<<<<<< HEAD
   "execution_count": 8,
   "metadata": {},
=======

   "execution_count": 12,
   "metadata": {},

>>>>>>> 1ecdb7a68f620e3a15884aebb78f5bb20f09d6e1
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count      14723.00000\n",
       "mean      305870.27100\n",
       "std       387052.96398\n",
       "min        22200.00000\n",
       "25%       138200.00000\n",
       "50%       212900.00000\n",
       "75%       353300.00000\n",
       "max     19314900.00000\n",
       "Name: value, dtype: float64"
      ]
     },

     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "melted['year'] = melted.time.dt.year\n",
    "melted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "melted['year'] = melted.time.dt.year\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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></th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>Metro</th>\n",
       "      <th>CountyName</th>\n",
       "      <th>time</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RegionName</th>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">1001</th>\n",
       "      <th>1996</th>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1996-12-01</td>\n",
       "      <td>113100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1997</th>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1997-12-01</td>\n",
       "      <td>114800.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1998</th>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1998-12-01</td>\n",
       "      <td>117500.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999</th>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1999-12-01</td>\n",
       "      <td>122600.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000</th>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>2000-12-01</td>\n",
       "      <td>129400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">99901</th>\n",
       "      <th>2014</th>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>AK</td>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>Ketchikan Gateway</td>\n",
       "      <td>2014-12-01</td>\n",
       "      <td>267100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>AK</td>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>Ketchikan Gateway</td>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>278600.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>AK</td>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>Ketchikan Gateway</td>\n",
       "      <td>2016-12-01</td>\n",
       "      <td>284100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017</th>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>AK</td>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>Ketchikan Gateway</td>\n",
       "      <td>2017-12-01</td>\n",
       "      <td>291900.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>AK</td>\n",
       "      <td>Ketchikan</td>\n",
       "      <td>Ketchikan Gateway</td>\n",
       "      <td>2018-04-01</td>\n",
       "      <td>305100.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>325560 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      City State        Metro         CountyName       time  \\\n",
       "RegionName year                                                               \n",
       "1001       1996     Agawam    MA  Springfield            Hampden 1996-12-01   \n",
       "           1997     Agawam    MA  Springfield            Hampden 1997-12-01   \n",
       "           1998     Agawam    MA  Springfield            Hampden 1998-12-01   \n",
       "           1999     Agawam    MA  Springfield            Hampden 1999-12-01   \n",
       "           2000     Agawam    MA  Springfield            Hampden 2000-12-01   \n",
       "...                    ...   ...          ...                ...        ...   \n",
       "99901      2014  Ketchikan    AK    Ketchikan  Ketchikan Gateway 2014-12-01   \n",
       "           2015  Ketchikan    AK    Ketchikan  Ketchikan Gateway 2015-12-01   \n",
       "           2016  Ketchikan    AK    Ketchikan  Ketchikan Gateway 2016-12-01   \n",
       "           2017  Ketchikan    AK    Ketchikan  Ketchikan Gateway 2017-12-01   \n",
       "           2018  Ketchikan    AK    Ketchikan  Ketchikan Gateway 2018-04-01   \n",
       "\n",
       "                    value  \n",
       "RegionName year            \n",
       "1001       1996  113100.0  \n",
       "           1997  114800.0  \n",
       "           1998  117500.0  \n",
       "           1999  122600.0  \n",
       "           2000  129400.0  \n",
       "...                   ...  \n",
       "99901      2014  267100.0  \n",
       "           2015  278600.0  \n",
       "           2016  284100.0  \n",
       "           2017  291900.0  \n",
       "           2018  305100.0  \n",
       "\n",
       "[325560 rows x 6 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zip_by_year = melted.groupby(['RegionName', 'year']).max()\n",
    "zip_by_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "module 'pandas' has no attribute 'describe'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-11-62dfc3dab3d2>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mdir\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdescribe\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/opt/anaconda3/lib/python3.7/site-packages/pandas/__init__.py\u001b[0m in \u001b[0;36m__getattr__\u001b[0;34m(name)\u001b[0m\n\u001b[1;32m    260\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0m_SparseArray\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    261\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 262\u001b[0;31m         \u001b[0;32mraise\u001b[0m \u001b[0mAttributeError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"module 'pandas' has no attribute '{name}'\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    263\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    264\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mAttributeError\u001b[0m: module 'pandas' has no attribute 'describe'"
     ]
    }
   ],
   "source": [
    "dir(pd.describe)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count      14723.00000\n",
       "mean      305870.27100\n",
       "std       387052.96398\n",
       "min        22200.00000\n",
       "25%       138200.00000\n",
       "50%       212900.00000\n",
       "75%       353300.00000\n",
       "max     19314900.00000\n",
       "Name: value, dtype: float64"
      ]
     },

     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('display.float_format', lambda x: '%.5f' % x)\n",
    "melted.groupby(['RegionName']).value.max().describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[28658,\n",
       " 94957,\n",
       " 80610,\n",
       " 49885,\n",
       " 60120,\n",
       " 19148,\n",
       " 15042,\n",
       " 97003,\n",
       " 94954,\n",
       " 76513,\n",
       " 74464,\n",
       " 56031,\n",
       " 19149,\n",
       " 97004,\n",
       " 12992,\n",
       " 74471,\n",
       " 72416,\n",
       " 60126,\n",
       " 56024,\n",
       " 10950,\n",
       " 15044,\n",
       " 97005,\n",
       " 27343,\n",
       " 29388,\n",
       " 19147,\n",
       " 8902,\n",
       " 97002,\n",
       " 29384,\n",
       " 97006,\n",
       " 19136,\n",
       " 60119,\n",
       " 56017,\n",
       " 2763,\n",
       " 17090,\n",
       " 23233,\n",
       " 78575,\n",
       " 94951,\n",
       " 10956,\n",
       " 2760,\n",
       " 27332,\n",
       " 23234,\n",
       " 78574,\n",
       " 17102,\n",
       " 33496,\n",
       " 23235,\n",
       " 19137,\n",
       " 70373,\n",
       " 74467,\n",
       " 19150,\n",
       " 29385,\n",
       " 94952,\n",
       " 70372,\n",
       " 60123,\n",
       " 19151,\n",
       " 15045,\n",
       " 60124,\n",
       " 60035,\n",
       " 29365,\n",
       " 27312,\n",
       " 66205,\n",
       " 80538,\n",
       " 15035,\n",
       " 23223,\n",
       " 19125,\n",
       " 31411,\n",
       " 27313,\n",
       " 66204,\n",
       " 2746,\n",
       " 6840,\n",
       " 60070,\n",
       " 19124,\n",
       " 80540,\n",
       " 66203,\n",
       " 15037,\n",
       " 2747,\n",
       " 19123,\n",
       " 17074,\n",
       " 60071,\n",
       " 66202,\n",
       " 92823,\n",
       " 84627,\n",
       " 10940,\n",
       " 31410,\n",
       " 23222,\n",
       " 56026,\n",
       " 55972,\n",
       " 15046,\n",
       " 8901,\n",
       " 97007,\n",
       " 80615,\n",
       " 78566,\n",
       " 19145,\n",
       " 19126,\n",
       " 23220,\n",
       " 31408,\n",
       " 33455,\n",
       " 53925,\n",
       " 66207,\n",
       " 8889,\n",
       " 76442,\n",
       " 80536,\n",
       " 19127,\n",
       " 17078,\n",
       " 23221,\n",
       " 29360,\n",
       " 60067,\n",
       " 66206,\n",
       " 80537,\n",
       " 2748,\n",
       " 15034,\n",
       " 39641,\n",
       " 33498,\n",
       " 78572,\n",
       " 60104,\n",
       " 27355,\n",
       " 25306,\n",
       " 10963,\n",
       " 15057,\n",
       " 60107,\n",
       " 33478,\n",
       " 97018,\n",
       " 76528,\n",
       " 21215,\n",
       " 29403,\n",
       " 10960,\n",
       " 33477,\n",
       " 97017,\n",
       " 97019,\n",
       " 94970,\n",
       " 78578,\n",
       " 21214,\n",
       " 25304,\n",
       " 94973,\n",
       " 76534,\n",
       " 25311,\n",
       " 27358,\n",
       " 21209,\n",
       " 2770,\n",
       " 21212,\n",
       " 60106,\n",
       " 80621,\n",
       " 15071,\n",
       " 97013,\n",
       " 2779,\n",
       " 25302,\n",
       " 60103,\n",
       " 56001,\n",
       " 84720,\n",
       " 19152,\n",
       " 53955,\n",
       " 76541,\n",
       " 97015,\n",
       " 84721,\n",
       " 10973,\n",
       " 10962,\n",
       " 2777,\n",
       " 27349,\n",
       " 19153,\n",
       " 17104,\n",
       " 43725,\n",
       " 33480,\n",
       " 60101,\n",
       " 53954,\n",
       " 97016,\n",
       " 78577,\n",
       " 29401,\n",
       " 60110,\n",
       " 21208,\n",
       " 15061,\n",
       " 10952,\n",
       " 10954,\n",
       " 19142,\n",
       " 23236,\n",
       " 27330,\n",
       " 60115,\n",
       " 23237,\n",
       " 29376,\n",
       " 94947,\n",
       " 60112,\n",
       " 2764,\n",
       " 15050,\n",
       " 23238,\n",
       " 2771,\n",
       " 8904,\n",
       " 19141,\n",
       " 78573,\n",
       " 80620,\n",
       " 94949,\n",
       " 35546,\n",
       " 60118,\n",
       " 10958,\n",
       " 15052,\n",
       " 2762,\n",
       " 19138,\n",
       " 49879,\n",
       " 94945,\n",
       " 76522,\n",
       " 33472,\n",
       " 49866,\n",
       " 53960,\n",
       " 97022,\n",
       " 78583,\n",
       " 29407,\n",
       " 25309,\n",
       " 21211,\n",
       " 17113,\n",
       " 10964,\n",
       " 60108,\n",
       " 53963,\n",
       " 33473,\n",
       " 97023,\n",
       " 80631,\n",
       " 78582,\n",
       " 29406,\n",
       " 27357,\n",
       " 21210,\n",
       " 17112,\n",
       " 15063,\n",
       " 10965,\n",
       " 2769,\n",
       " 56011,\n",
       " 2744,\n",
       " 27316,\n",
       " 17073,\n",
       " 33461,\n",
       " 8867,\n",
       " 80512,\n",
       " 60091,\n",
       " 33462,\n",
       " 19119,\n",
       " 17070,\n",
       " 31401,\n",
       " 15009,\n",
       " 74370,\n",
       " 80513,\n",
       " 53951,\n",
       " 19116,\n",
       " 2726,\n",
       " 2724,\n",
       " 15010,\n",
       " 8865,\n",
       " 92808,\n",
       " 80514,\n",
       " 60089,\n",
       " 33460,\n",
       " 17068,\n",
       " 2725,\n",
       " 12962,\n",
       " 10913,\n",
       " 6820,\n",
       " 19118,\n",
       " 55985,\n",
       " 17057,\n",
       " 15021,\n",
       " 12972,\n",
       " 19107,\n",
       " 21152,\n",
       " 80525,\n",
       " 92807,\n",
       " 60084,\n",
       " 49841,\n",
       " 15022,\n",
       " 10924,\n",
       " 21155,\n",
       " 80526,\n",
       " 33463,\n",
       " 92804,\n",
       " 60085,\n",
       " 55987,\n",
       " 10925,\n",
       " 8876,\n",
       " 27301,\n",
       " 21154,\n",
       " 72331,\n",
       " 92805,\n",
       " 49855,\n",
       " 60090,\n",
       " 80515,\n",
       " 55992,\n",
       " 27310,\n",
       " 27282,\n",
       " 15015,\n",
       " 10917,\n",
       " 2721,\n",
       " 74427,\n",
       " 78521,\n",
       " 90803,\n",
       " 2718,\n",
       " 8859,\n",
       " 17047,\n",
       " 19094,\n",
       " 23188,\n",
       " 39564,\n",
       " 31404,\n",
       " 55940,\n",
       " 60034,\n",
       " 78520,\n",
       " 84663,\n",
       " 90802,\n",
       " 2719,\n",
       " 15001,\n",
       " 19095,\n",
       " 17046,\n",
       " 21140,\n",
       " 27283,\n",
       " 21162,\n",
       " 43701,\n",
       " 45750,\n",
       " 60093,\n",
       " 17067,\n",
       " 19114,\n",
       " 21161,\n",
       " 10918,\n",
       " 15012,\n",
       " 80516,\n",
       " 33458,\n",
       " 39601,\n",
       " 31405,\n",
       " 29356,\n",
       " 19115,\n",
       " 17066,\n",
       " 21160,\n",
       " 2723,\n",
       " 80517,\n",
       " 49849,\n",
       " 31406,\n",
       " 21163,\n",
       " 17065,\n",
       " 15014,\n",
       " 8869,\n",
       " 10916,\n",
       " 2720,\n",
       " 10927,\n",
       " 47805,\n",
       " 19120,\n",
       " 15027,\n",
       " 33446,\n",
       " 23230,\n",
       " 27320,\n",
       " 2740,\n",
       " 12979,\n",
       " 15026,\n",
       " 10928,\n",
       " 33445,\n",
       " 80530,\n",
       " 23231,\n",
       " 31419,\n",
       " 8880,\n",
       " 55981,\n",
       " 55983,\n",
       " 60073,\n",
       " 33444,\n",
       " 98948,\n",
       " 27326,\n",
       " 19130,\n",
       " 8887,\n",
       " 12981,\n",
       " 2738,\n",
       " 55976,\n",
       " 35490,\n",
       " 53932,\n",
       " 15025,\n",
       " 92806,\n",
       " 92821,\n",
       " 33449,\n",
       " 60068,\n",
       " 53923,\n",
       " 80542,\n",
       " 76444,\n",
       " 98953,\n",
       " 10941,\n",
       " 2745,\n",
       " 27317,\n",
       " 60069,\n",
       " 80543,\n",
       " 17087,\n",
       " 8882,\n",
       " 19134,\n",
       " 23228,\n",
       " 29369,\n",
       " 10930,\n",
       " 15024,\n",
       " 60074,\n",
       " 80528,\n",
       " 98951,\n",
       " 19135,\n",
       " 23229,\n",
       " 2743,\n",
       " 98947,\n",
       " 29372,\n",
       " 19131,\n",
       " 60081,\n",
       " 19111,\n",
       " 17062,\n",
       " 80521,\n",
       " 33469,\n",
       " 47802,\n",
       " 8873,\n",
       " 10920,\n",
       " 17061,\n",
       " 76424,\n",
       " 47803,\n",
       " 55991,\n",
       " 15019,\n",
       " 23225,\n",
       " 10921,\n",
       " 8872,\n",
       " 21158,\n",
       " 92801,\n",
       " 33467,\n",
       " 8879,\n",
       " 10926,\n",
       " 27302,\n",
       " 29349,\n",
       " 19106,\n",
       " 80524,\n",
       " 15017,\n",
       " 10923,\n",
       " 60083,\n",
       " 33470,\n",
       " 8886,\n",
       " 2739,\n",
       " 33442,\n",
       " 92831,\n",
       " 23226,\n",
       " 19128,\n",
       " 60076,\n",
       " 33441,\n",
       " 80534,\n",
       " 29374,\n",
       " 23227,\n",
       " 19129,\n",
       " 8884,\n",
       " 60077,\n",
       " 33440,\n",
       " 80535,\n",
       " 76437,\n",
       " 98944,\n",
       " 33471,\n",
       " 21157,\n",
       " 27298,\n",
       " 90755,\n",
       " 92802,\n",
       " 76543,\n",
       " 60102,\n",
       " 33483,\n",
       " 95018,\n",
       " 39701,\n",
       " 45840,\n",
       " 27403,\n",
       " 84780,\n",
       " 95019,\n",
       " 60184,\n",
       " 39702,\n",
       " 29451,\n",
       " 15106,\n",
       " 84781,\n",
       " 97067,\n",
       " 70438,\n",
       " 62232,\n",
       " 78626,\n",
       " 76577,\n",
       " 62234,\n",
       " 60185,\n",
       " 33556,\n",
       " 29450,\n",
       " 27401,\n",
       " 99123,\n",
       " 97068,\n",
       " 78629,\n",
       " 70433,\n",
       " 33558,\n",
       " 60187,\n",
       " 47901,\n",
       " 37658,\n",
       " 39705,\n",
       " 11023,\n",
       " 13068,\n",
       " 2827,\n",
       " 78639,\n",
       " 97062,\n",
       " 37659,\n",
       " 62231,\n",
       " 60180,\n",
       " 11020,\n",
       " 95014,\n",
       " 62230,\n",
       " 78624,\n",
       " 60181,\n",
       " 11021,\n",
       " 2825,\n",
       " 97064,\n",
       " 70437,\n",
       " 33559,\n",
       " 29449,\n",
       " 2822,\n",
       " 11010,\n",
       " 13057,\n",
       " 76579,\n",
       " 60190,\n",
       " 56088,\n",
       " 45845,\n",
       " 37615,\n",
       " 21236,\n",
       " 56037,\n",
       " 60131,\n",
       " 70364,\n",
       " 80601,\n",
       " 78552,\n",
       " 92883,\n",
       " 94928,\n",
       " 99022,\n",
       " 2812,\n",
       " 21239,\n",
       " 47850,\n",
       " 27406,\n",
       " 80602,\n",
       " 76504,\n",
       " 94931,\n",
       " 92880,\n",
       " 99021,\n",
       " 2813,\n",
       " 11001,\n",
       " 27377,\n",
       " 39663,\n",
       " 68317,\n",
       " 80603,\n",
       " 11003,\n",
       " 92882,\n",
       " 60130,\n",
       " 45801,\n",
       " 13061,\n",
       " 15108,\n",
       " 2818,\n",
       " 95020,\n",
       " 78628,\n",
       " 60191,\n",
       " 45844,\n",
       " 27407,\n",
       " 31501,\n",
       " 95023,\n",
       " 97070,\n",
       " 70435,\n",
       " 60188,\n",
       " 29455,\n",
       " 13063,\n",
       " 2816,\n",
       " 97071,\n",
       " 60189,\n",
       " 31503,\n",
       " 27405,\n",
       " 2817,\n",
       " 2814,\n",
       " 21237,\n",
       " 35611,\n",
       " 78636,\n",
       " 92881,\n",
       " 78642,\n",
       " 54028,\n",
       " 60171,\n",
       " 33542,\n",
       " 95035,\n",
       " 78643,\n",
       " 99109,\n",
       " 15122,\n",
       " 11024,\n",
       " 33541,\n",
       " 45827,\n",
       " 70454,\n",
       " 29466,\n",
       " 23325,\n",
       " 13074,\n",
       " 62218,\n",
       " 60169,\n",
       " 33540,\n",
       " 43777,\n",
       " 95037,\n",
       " 78645,\n",
       " 23320,\n",
       " 11030,\n",
       " 13077,\n",
       " 60174,\n",
       " 29464,\n",
       " 78640,\n",
       " 29445,\n",
       " 37642,\n",
       " 78655,\n",
       " 2840,\n",
       " 23314,\n",
       " 37643,\n",
       " 33545,\n",
       " 60164,\n",
       " 78654,\n",
       " 70458,\n",
       " 95030,\n",
       " 15135,\n",
       " 23315,\n",
       " 33544,\n",
       " 70452,\n",
       " 60165,\n",
       " 95033,\n",
       " 97080,\n",
       " 78641,\n",
       " 23324,\n",
       " 2838,\n",
       " 13073,\n",
       " 15120,\n",
       " 33543,\n",
       " 95032,\n",
       " 72501,\n",
       " 62221,\n",
       " 54025,\n",
       " 78644,\n",
       " 70446,\n",
       " 95008,\n",
       " 33566,\n",
       " 37660,\n",
       " 2831,\n",
       " 29440,\n",
       " 70447,\n",
       " 97058,\n",
       " 37663,\n",
       " 33565,\n",
       " 60176,\n",
       " 2828,\n",
       " 78634,\n",
       " 70448,\n",
       " 95010,\n",
       " 35613,\n",
       " 45850,\n",
       " 62226,\n",
       " 60177,\n",
       " 13066,\n",
       " 97060,\n",
       " 84770,\n",
       " 33563,\n",
       " 13069,\n",
       " 15116,\n",
       " 2830,\n",
       " 60178,\n",
       " 35614,\n",
       " 33567,\n",
       " 23321,\n",
       " 13076,\n",
       " 60175,\n",
       " 62220,\n",
       " 97086,\n",
       " 23322,\n",
       " 15126,\n",
       " 2832,\n",
       " 62223,\n",
       " 60172,\n",
       " 45831,\n",
       " 43780,\n",
       " 70450,\n",
       " 31519,\n",
       " 29470,\n",
       " 23323,\n",
       " 13078,\n",
       " 60173,\n",
       " 45830,\n",
       " 70445,\n",
       " 78633,\n",
       " 84774,\n",
       " 97056,\n",
       " 94930,\n",
       " 13053,\n",
       " 21201,\n",
       " 33523,\n",
       " 94923,\n",
       " 70343,\n",
       " 54014,\n",
       " 60153,\n",
       " 21230,\n",
       " 29418,\n",
       " 25320,\n",
       " 13026,\n",
       " 10977,\n",
       " 82636,\n",
       " 74432,\n",
       " 37617,\n",
       " 13027,\n",
       " 13029,\n",
       " 99027,\n",
       " 94925,\n",
       " 56057,\n",
       " 37616,\n",
       " 29420,\n",
       " 21224,\n",
       " 10983,\n",
       " 13028,\n",
       " 99026,\n",
       " 92879,\n",
       " 15074,\n",
       " 21231,\n",
       " 60148,\n",
       " 33527,\n",
       " 13039,\n",
       " 21219,\n",
       " 74445,\n",
       " 92868,\n",
       " 54002,\n",
       " 10989,\n",
       " 29414,\n",
       " 21218,\n",
       " 92869,\n",
       " 54013,\n",
       " 60154,\n",
       " 43762,\n",
       " 33525,\n",
       " 21229,\n",
       " 27370,\n",
       " 2790,\n",
       " 74435,\n",
       " 60155,\n",
       " 37620,\n",
       " 21228,\n",
       " 94920,\n",
       " 74434,\n",
       " 54015,\n",
       " 60152,\n",
       " 60156,\n",
       " 54011,\n",
       " 56058,\n",
       " 60096,\n",
       " 39629,\n",
       " 53956,\n",
       " 60099,\n",
       " 80634,\n",
       " 2780,\n",
       " 15066,\n",
       " 10968,\n",
       " 21207,\n",
       " 17109,\n",
       " 27344,\n",
       " 53959,\n",
       " 97011,\n",
       " 39666,\n",
       " 21206,\n",
       " 33484,\n",
       " 60097,\n",
       " 76542,\n",
       " 94965,\n",
       " 10974,\n",
       " 13021,\n",
       " 15068,\n",
       " 25303,\n",
       " 27350,\n",
       " 19154,\n",
       " 33486,\n",
       " 21204,\n",
       " 17110,\n",
       " 15065,\n",
       " 21227,\n",
       " 13031,\n",
       " 10980,\n",
       " 99025,\n",
       " 82633,\n",
       " 74437,\n",
       " 60157,\n",
       " 45814,\n",
       " 37618,\n",
       " 21226,\n",
       " 13030,\n",
       " 6883,\n",
       " 74436,\n",
       " 97008,\n",
       " 10970,\n",
       " 17111,\n",
       " 33487,\n",
       " 60098,\n",
       " 76539,\n",
       " 78584,\n",
       " 97009,\n",
       " 94960,\n",
       " 6877,\n",
       " 56050,\n",
       " 29412,\n",
       " 29429,\n",
       " 2804,\n",
       " 72404,\n",
       " 74451,\n",
       " 21244,\n",
       " 2807,\n",
       " 15089,\n",
       " 13040,\n",
       " 60139,\n",
       " 33510,\n",
       " 37604,\n",
       " 23294,\n",
       " 29435,\n",
       " 15090,\n",
       " 33511,\n",
       " 10992,\n",
       " 60136,\n",
       " 39654,\n",
       " 47842,\n",
       " 94939,\n",
       " 70359,\n",
       " 10993,\n",
       " 60137,\n",
       " 84701,\n",
       " 10998,\n",
       " 13045,\n",
       " 39652,\n",
       " 13041,\n",
       " 13036,\n",
       " 15102,\n",
       " 45805,\n",
       " 60134,\n",
       " 92886,\n",
       " 99019,\n",
       " 15101,\n",
       " 13052,\n",
       " 45804,\n",
       " 60135,\n",
       " 74462,\n",
       " 70360,\n",
       " 92887,\n",
       " 11004,\n",
       " 10994,\n",
       " 2808,\n",
       " 29431,\n",
       " 45807,\n",
       " 33513,\n",
       " 70363,\n",
       " 2809,\n",
       " 21234,\n",
       " 45806,\n",
       " 60133,\n",
       " 99016,\n",
       " 2806,\n",
       " 60142,\n",
       " 37601,\n",
       " 39648,\n",
       " 29410,\n",
       " 10984,\n",
       " 25313,\n",
       " 27360,\n",
       " 99037,\n",
       " 78539,\n",
       " 74441,\n",
       " 56055,\n",
       " 60145,\n",
       " 15083,\n",
       " 13034,\n",
       " 21222,\n",
       " 25312,\n",
       " 94941,\n",
       " 92865,\n",
       " 60150,\n",
       " 10990,\n",
       " 13037,\n",
       " 15084,\n",
       " 74447,\n",
       " 78541,\n",
       " 70345,\n",
       " 92870,\n",
       " 60151,\n",
       " 15085,\n",
       " 13035,\n",
       " 54007,\n",
       " 92867,\n",
       " 25314,\n",
       " 74455,\n",
       " 76502,\n",
       " 13044,\n",
       " 6897,\n",
       " 60143,\n",
       " 72401,\n",
       " 29439,\n",
       " 60140,\n",
       " 99009,\n",
       " 29438,\n",
       " 78550,\n",
       " 76501,\n",
       " 45817,\n",
       " 60146,\n",
       " 10986,\n",
       " 13033,\n",
       " 21221,\n",
       " 25315,\n",
       " 92866,\n",
       " 33534,\n",
       " 10987,\n",
       " 13032,\n",
       " 21220,\n",
       " 39565,\n",
       " 64129,\n",
       " 37640,\n",
       " 62037,\n",
       " 98941,\n",
       " 49749,\n",
       " 62035,\n",
       " 64082,\n",
       " 19012,\n",
       " 98940,\n",
       " 90720,\n",
       " 64083,\n",
       " 62034,\n",
       " 23111,\n",
       " 43614,\n",
       " 19010,\n",
       " 21060,\n",
       " 21057,\n",
       " 43615,\n",
       " 2635,\n",
       " 31301,\n",
       " 98937,\n",
       " 47710,\n",
       " 43612,\n",
       " 64086,\n",
       " 2632,\n",
       " 31302,\n",
       " 19008,\n",
       " 27203,\n",
       " 19015,\n",
       " 47711,\n",
       " 64079,\n",
       " 2642,\n",
       " 80501,\n",
       " 27231,\n",
       " 19035,\n",
       " 31326,\n",
       " 19032,\n",
       " 12887,\n",
       " 64078,\n",
       " 84601,\n",
       " 80503,\n",
       " 19033,\n",
       " 45638,\n",
       " 2639,\n",
       " 90723,\n",
       " 49751,\n",
       " 62033,\n",
       " 64080,\n",
       " 2638,\n",
       " 19014,\n",
       " 21061,\n",
       " 98942,\n",
       " 72301,\n",
       " 43611,\n",
       " 64081,\n",
       " 98936,\n",
       " 43613,\n",
       " 84660,\n",
       " 14904,\n",
       " 23112,\n",
       " 98930,\n",
       " 64093,\n",
       " 27215,\n",
       " 23113,\n",
       " 92780,\n",
       " 23114,\n",
       " 12871,\n",
       " 90732,\n",
       " 80487,\n",
       " 39507,\n",
       " 21045,\n",
       " 31308,\n",
       " 33327,\n",
       " 37421,\n",
       " 14905,\n",
       " 16950,\n",
       " 23093,\n",
       " 21044,\n",
       " 33326,\n",
       " 68127,\n",
       " 92691,\n",
       " 12859,\n",
       " 21047,\n",
       " 19018,\n",
       " 27214,\n",
       " 2633,\n",
       " 23117,\n",
       " 27205,\n",
       " 90731,\n",
       " 74339,\n",
       " 23116,\n",
       " 2630,\n",
       " 12865,\n",
       " 80481,\n",
       " 55901,\n",
       " 64089,\n",
       " 62040,\n",
       " 19023,\n",
       " 2631,\n",
       " 43606,\n",
       " 80482,\n",
       " 74337,\n",
       " 76384,\n",
       " 55902,\n",
       " 19020,\n",
       " 27208,\n",
       " 98932,\n",
       " 19021,\n",
       " 12866,\n",
       " 92782,\n",
       " 74343,\n",
       " 12885,\n",
       " ...]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Make a list of all zipcodes appearing at least 60 times (observations are monthly so this translates to five years worth of data)\n",
    "melted.RegionName.value_counts() \\\n",
    "    [melted.RegionName.value_counts() > 60] \\\n",
    "    .index.to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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>value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1996-04-01</th>\n",
       "      <td>358100.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-05-01</th>\n",
       "      <td>356000.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-06-01</th>\n",
       "      <td>353900.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-07-01</th>\n",
       "      <td>351700.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-08-01</th>\n",
       "      <td>349600.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-12-01</th>\n",
       "      <td>1296500.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01</th>\n",
       "      <td>1293000.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02-01</th>\n",
       "      <td>1291200.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-01</th>\n",
       "      <td>1291000.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-04-01</th>\n",
       "      <td>1290000.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>265 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   value\n",
       "time                    \n",
       "1996-04-01  358100.00000\n",
       "1996-05-01  356000.00000\n",
       "1996-06-01  353900.00000\n",
       "1996-07-01  351700.00000\n",
       "1996-08-01  349600.00000\n",
       "...                  ...\n",
       "2017-12-01 1296500.00000\n",
       "2018-01-01 1293000.00000\n",
       "2018-02-01 1291200.00000\n",
       "2018-03-01 1291000.00000\n",
       "2018-04-01 1290000.00000\n",
       "\n",
       "[265 rows x 1 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# MVP: Fit and Predict for One Zipcode, show accuracy over time\n",
    "one_zip = melted[melted['RegionName'] == 20007]\n",
    "\n",
    "one_zip = one_zip[['time', 'value']]\n",
    "one_zip = one_zip.set_index('time')\n",
    "one_zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a18578bd0>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the data over time\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15,10))\n",
    "one_zip.plot(marker='*', ax=ax)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pandas.core.indexes.datetimes.DatetimeIndex"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking the index is datetime format\n",
    "\n",
    "type(one_zip.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking for missing values\n",
    "\n",
    "one_zip.value.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Checking the decomposed seasonal trends:\n",
    "\n",
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "\n",
    "decomposition = seasonal_decompose(one_zip)\n",
    "fig = plt.figure()\n",
    "fig = decomposition.plot()\n",
    "fig.set_size_inches(15, 8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-2.563886562867809,\n",
       " 0.10069322265751685,\n",
       " 15,\n",
       " 249,\n",
       " {'1%': -3.4568881317725864,\n",
       "  '5%': -2.8732185133016057,\n",
       "  '10%': -2.5729936189738876},\n",
       " 4290.592655898536)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Test Stationarity - Appears to not be stationary without any differencing\n",
    "\n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "\n",
    "adfuller(one_zip)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Checking the autocorrelation/partial autocorrelation:\n",
    "import numpy as np\n",
    "from statsmodels.tsa.stattools import acf, pacf\n",
    "from statsmodels.graphics.tsaplots import plot_acf, plot_pacf\n",
    "\n",
    "fig = plt.figure(figsize=(15,10))\n",
    "fig = plt.scatter(x = range(0,61), y=acf(one_zip, nlags=60, fft=True).ravel())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(15,10))\n",
    "fig = plt.scatter(x = range(1,62), y=pacf(one_zip, nlags=60).ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "# Let's try a quick ARIMA model:\n",
    "\n",
    "from statsmodels.tsa.arima_model import ARIMA\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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>RegionName</th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>Metro</th>\n",
       "      <th>CountyName</th>\n",
       "      <th>time</th>\n",
       "      <th>value</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5850</th>\n",
       "      <td>1001</td>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1996-04-01</td>\n",
       "      <td>113100.00000</td>\n",
       "      <td>1996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20573</th>\n",
       "      <td>1001</td>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1996-05-01</td>\n",
       "      <td>112800.00000</td>\n",
       "      <td>1996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35296</th>\n",
       "      <td>1001</td>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1996-06-01</td>\n",
       "      <td>112600.00000</td>\n",
       "      <td>1996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50019</th>\n",
       "      <td>1001</td>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1996-07-01</td>\n",
       "      <td>112300.00000</td>\n",
       "      <td>1996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64742</th>\n",
       "      <td>1001</td>\n",
       "      <td>Agawam</td>\n",
       "      <td>MA</td>\n",
       "      <td>Springfield</td>\n",
       "      <td>Hampden</td>\n",
       "      <td>1996-08-01</td>\n",
       "      <td>112100.00000</td>\n",
       "      <td>1996</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       RegionName    City State        Metro CountyName       time  \\\n",
       "5850         1001  Agawam    MA  Springfield    Hampden 1996-04-01   \n",
       "20573        1001  Agawam    MA  Springfield    Hampden 1996-05-01   \n",
       "35296        1001  Agawam    MA  Springfield    Hampden 1996-06-01   \n",
       "50019        1001  Agawam    MA  Springfield    Hampden 1996-07-01   \n",
       "64742        1001  Agawam    MA  Springfield    Hampden 1996-08-01   \n",
       "\n",
       "             value  year  \n",
       "5850  113100.00000  1996  \n",
       "20573 112800.00000  1996  \n",
       "35296 112600.00000  1996  \n",
       "50019 112300.00000  1996  \n",
       "64742 112100.00000  1996  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "melted[melted.RegionName == 1001].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "melted.value.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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>RegionName</th>\n",
       "      <th>1001</th>\n",
       "      <th>1002</th>\n",
       "      <th>1005</th>\n",
       "      <th>1007</th>\n",
       "      <th>1008</th>\n",
       "      <th>1010</th>\n",
       "      <th>1011</th>\n",
       "      <th>1013</th>\n",
       "      <th>1020</th>\n",
       "      <th>1026</th>\n",
       "      <th>...</th>\n",
       "      <th>99654</th>\n",
       "      <th>99664</th>\n",
       "      <th>99669</th>\n",
       "      <th>99701</th>\n",
       "      <th>99705</th>\n",
       "      <th>99709</th>\n",
       "      <th>99712</th>\n",
       "      <th>99801</th>\n",
       "      <th>99835</th>\n",
       "      <th>99901</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1996-04-01</th>\n",
       "      <td>113100.00000</td>\n",
       "      <td>161000.00000</td>\n",
       "      <td>103100.00000</td>\n",
       "      <td>133400.00000</td>\n",
       "      <td>117500.00000</td>\n",
       "      <td>115800.00000</td>\n",
       "      <td>87000.00000</td>\n",
       "      <td>88500.00000</td>\n",
       "      <td>97200.00000</td>\n",
       "      <td>91200.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>122700.00000</td>\n",
       "      <td>112200.00000</td>\n",
       "      <td>112500.00000</td>\n",
       "      <td>89400.00000</td>\n",
       "      <td>101900.00000</td>\n",
       "      <td>103000.00000</td>\n",
       "      <td>109200.00000</td>\n",
       "      <td>166800.00000</td>\n",
       "      <td>155500.00000</td>\n",
       "      <td>117100.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-05-01</th>\n",
       "      <td>112800.00000</td>\n",
       "      <td>160100.00000</td>\n",
       "      <td>103400.00000</td>\n",
       "      <td>132700.00000</td>\n",
       "      <td>117300.00000</td>\n",
       "      <td>115700.00000</td>\n",
       "      <td>87000.00000</td>\n",
       "      <td>88500.00000</td>\n",
       "      <td>97100.00000</td>\n",
       "      <td>91100.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>124200.00000</td>\n",
       "      <td>112400.00000</td>\n",
       "      <td>113100.00000</td>\n",
       "      <td>89800.00000</td>\n",
       "      <td>102900.00000</td>\n",
       "      <td>103900.00000</td>\n",
       "      <td>109600.00000</td>\n",
       "      <td>166200.00000</td>\n",
       "      <td>154900.00000</td>\n",
       "      <td>117200.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-06-01</th>\n",
       "      <td>112600.00000</td>\n",
       "      <td>159300.00000</td>\n",
       "      <td>103600.00000</td>\n",
       "      <td>132000.00000</td>\n",
       "      <td>117100.00000</td>\n",
       "      <td>115500.00000</td>\n",
       "      <td>87100.00000</td>\n",
       "      <td>88400.00000</td>\n",
       "      <td>97000.00000</td>\n",
       "      <td>91000.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>125700.00000</td>\n",
       "      <td>112600.00000</td>\n",
       "      <td>113800.00000</td>\n",
       "      <td>90200.00000</td>\n",
       "      <td>103800.00000</td>\n",
       "      <td>104700.00000</td>\n",
       "      <td>110000.00000</td>\n",
       "      <td>165400.00000</td>\n",
       "      <td>154200.00000</td>\n",
       "      <td>117300.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-07-01</th>\n",
       "      <td>112300.00000</td>\n",
       "      <td>158600.00000</td>\n",
       "      <td>103800.00000</td>\n",
       "      <td>131400.00000</td>\n",
       "      <td>117000.00000</td>\n",
       "      <td>115300.00000</td>\n",
       "      <td>87200.00000</td>\n",
       "      <td>88300.00000</td>\n",
       "      <td>96800.00000</td>\n",
       "      <td>90800.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>127200.00000</td>\n",
       "      <td>112800.00000</td>\n",
       "      <td>114400.00000</td>\n",
       "      <td>90600.00000</td>\n",
       "      <td>104700.00000</td>\n",
       "      <td>105500.00000</td>\n",
       "      <td>110400.00000</td>\n",
       "      <td>164400.00000</td>\n",
       "      <td>153500.00000</td>\n",
       "      <td>117400.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-08-01</th>\n",
       "      <td>112100.00000</td>\n",
       "      <td>158000.00000</td>\n",
       "      <td>103900.00000</td>\n",
       "      <td>130800.00000</td>\n",
       "      <td>116800.00000</td>\n",
       "      <td>115100.00000</td>\n",
       "      <td>87400.00000</td>\n",
       "      <td>88200.00000</td>\n",
       "      <td>96600.00000</td>\n",
       "      <td>90700.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>128600.00000</td>\n",
       "      <td>113000.00000</td>\n",
       "      <td>115100.00000</td>\n",
       "      <td>91000.00000</td>\n",
       "      <td>105400.00000</td>\n",
       "      <td>106200.00000</td>\n",
       "      <td>110800.00000</td>\n",
       "      <td>163500.00000</td>\n",
       "      <td>152800.00000</td>\n",
       "      <td>117500.00000</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>2017-12-01</th>\n",
       "      <td>221700.00000</td>\n",
       "      <td>342000.00000</td>\n",
       "      <td>213300.00000</td>\n",
       "      <td>268100.00000</td>\n",
       "      <td>205400.00000</td>\n",
       "      <td>243800.00000</td>\n",
       "      <td>151400.00000</td>\n",
       "      <td>173300.00000</td>\n",
       "      <td>187400.00000</td>\n",
       "      <td>201200.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>258800.00000</td>\n",
       "      <td>219100.00000</td>\n",
       "      <td>250200.00000</td>\n",
       "      <td>206700.00000</td>\n",
       "      <td>261100.00000</td>\n",
       "      <td>262800.00000</td>\n",
       "      <td>273500.00000</td>\n",
       "      <td>402300.00000</td>\n",
       "      <td>396200.00000</td>\n",
       "      <td>291900.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01</th>\n",
       "      <td>221700.00000</td>\n",
       "      <td>344500.00000</td>\n",
       "      <td>215600.00000</td>\n",
       "      <td>268100.00000</td>\n",
       "      <td>204500.00000</td>\n",
       "      <td>244600.00000</td>\n",
       "      <td>150100.00000</td>\n",
       "      <td>174600.00000</td>\n",
       "      <td>188100.00000</td>\n",
       "      <td>204600.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>259400.00000</td>\n",
       "      <td>218900.00000</td>\n",
       "      <td>250600.00000</td>\n",
       "      <td>206100.00000</td>\n",
       "      <td>261300.00000</td>\n",
       "      <td>262800.00000</td>\n",
       "      <td>273800.00000</td>\n",
       "      <td>406200.00000</td>\n",
       "      <td>397500.00000</td>\n",
       "      <td>294200.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02-01</th>\n",
       "      <td>221700.00000</td>\n",
       "      <td>347400.00000</td>\n",
       "      <td>217900.00000</td>\n",
       "      <td>268800.00000</td>\n",
       "      <td>206800.00000</td>\n",
       "      <td>246700.00000</td>\n",
       "      <td>150100.00000</td>\n",
       "      <td>175900.00000</td>\n",
       "      <td>189000.00000</td>\n",
       "      <td>211200.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>259900.00000</td>\n",
       "      <td>218300.00000</td>\n",
       "      <td>250100.00000</td>\n",
       "      <td>205600.00000</td>\n",
       "      <td>261400.00000</td>\n",
       "      <td>262300.00000</td>\n",
       "      <td>273800.00000</td>\n",
       "      <td>409800.00000</td>\n",
       "      <td>398900.00000</td>\n",
       "      <td>297500.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-01</th>\n",
       "      <td>222700.00000</td>\n",
       "      <td>350600.00000</td>\n",
       "      <td>219500.00000</td>\n",
       "      <td>270000.00000</td>\n",
       "      <td>210900.00000</td>\n",
       "      <td>248800.00000</td>\n",
       "      <td>149500.00000</td>\n",
       "      <td>176900.00000</td>\n",
       "      <td>189600.00000</td>\n",
       "      <td>217900.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>260200.00000</td>\n",
       "      <td>218400.00000</td>\n",
       "      <td>249300.00000</td>\n",
       "      <td>205500.00000</td>\n",
       "      <td>261700.00000</td>\n",
       "      <td>262000.00000</td>\n",
       "      <td>273900.00000</td>\n",
       "      <td>412600.00000</td>\n",
       "      <td>402100.00000</td>\n",
       "      <td>302100.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-04-01</th>\n",
       "      <td>223600.00000</td>\n",
       "      <td>353300.00000</td>\n",
       "      <td>220700.00000</td>\n",
       "      <td>270600.00000</td>\n",
       "      <td>214200.00000</td>\n",
       "      <td>250900.00000</td>\n",
       "      <td>149100.00000</td>\n",
       "      <td>177400.00000</td>\n",
       "      <td>189800.00000</td>\n",
       "      <td>220800.00000</td>\n",
       "      <td>...</td>\n",
       "      <td>260500.00000</td>\n",
       "      <td>219100.00000</td>\n",
       "      <td>249300.00000</td>\n",
       "      <td>205700.00000</td>\n",
       "      <td>262000.00000</td>\n",
       "      <td>262400.00000</td>\n",
       "      <td>274600.00000</td>\n",
       "      <td>413300.00000</td>\n",
       "      <td>404800.00000</td>\n",
       "      <td>305100.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>265 rows × 14723 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "RegionName        1001         1002         1005         1007         1008   \\\n",
       "time                                                                          \n",
       "1996-04-01 113100.00000 161000.00000 103100.00000 133400.00000 117500.00000   \n",
       "1996-05-01 112800.00000 160100.00000 103400.00000 132700.00000 117300.00000   \n",
       "1996-06-01 112600.00000 159300.00000 103600.00000 132000.00000 117100.00000   \n",
       "1996-07-01 112300.00000 158600.00000 103800.00000 131400.00000 117000.00000   \n",
       "1996-08-01 112100.00000 158000.00000 103900.00000 130800.00000 116800.00000   \n",
       "...                 ...          ...          ...          ...          ...   \n",
       "2017-12-01 221700.00000 342000.00000 213300.00000 268100.00000 205400.00000   \n",
       "2018-01-01 221700.00000 344500.00000 215600.00000 268100.00000 204500.00000   \n",
       "2018-02-01 221700.00000 347400.00000 217900.00000 268800.00000 206800.00000   \n",
       "2018-03-01 222700.00000 350600.00000 219500.00000 270000.00000 210900.00000   \n",
       "2018-04-01 223600.00000 353300.00000 220700.00000 270600.00000 214200.00000   \n",
       "\n",
       "RegionName        1010         1011         1013         1020         1026   \\\n",
       "time                                                                          \n",
       "1996-04-01 115800.00000  87000.00000  88500.00000  97200.00000  91200.00000   \n",
       "1996-05-01 115700.00000  87000.00000  88500.00000  97100.00000  91100.00000   \n",
       "1996-06-01 115500.00000  87100.00000  88400.00000  97000.00000  91000.00000   \n",
       "1996-07-01 115300.00000  87200.00000  88300.00000  96800.00000  90800.00000   \n",
       "1996-08-01 115100.00000  87400.00000  88200.00000  96600.00000  90700.00000   \n",
       "...                 ...          ...          ...          ...          ...   \n",
       "2017-12-01 243800.00000 151400.00000 173300.00000 187400.00000 201200.00000   \n",
       "2018-01-01 244600.00000 150100.00000 174600.00000 188100.00000 204600.00000   \n",
       "2018-02-01 246700.00000 150100.00000 175900.00000 189000.00000 211200.00000   \n",
       "2018-03-01 248800.00000 149500.00000 176900.00000 189600.00000 217900.00000   \n",
       "2018-04-01 250900.00000 149100.00000 177400.00000 189800.00000 220800.00000   \n",
       "\n",
       "RegionName  ...        99654        99664        99669        99701  \\\n",
       "time        ...                                                       \n",
       "1996-04-01  ... 122700.00000 112200.00000 112500.00000  89400.00000   \n",
       "1996-05-01  ... 124200.00000 112400.00000 113100.00000  89800.00000   \n",
       "1996-06-01  ... 125700.00000 112600.00000 113800.00000  90200.00000   \n",
       "1996-07-01  ... 127200.00000 112800.00000 114400.00000  90600.00000   \n",
       "1996-08-01  ... 128600.00000 113000.00000 115100.00000  91000.00000   \n",
       "...         ...          ...          ...          ...          ...   \n",
       "2017-12-01  ... 258800.00000 219100.00000 250200.00000 206700.00000   \n",
       "2018-01-01  ... 259400.00000 218900.00000 250600.00000 206100.00000   \n",
       "2018-02-01  ... 259900.00000 218300.00000 250100.00000 205600.00000   \n",
       "2018-03-01  ... 260200.00000 218400.00000 249300.00000 205500.00000   \n",
       "2018-04-01  ... 260500.00000 219100.00000 249300.00000 205700.00000   \n",
       "\n",
       "RegionName        99705        99709        99712        99801        99835  \\\n",
       "time                                                                          \n",
       "1996-04-01 101900.00000 103000.00000 109200.00000 166800.00000 155500.00000   \n",
       "1996-05-01 102900.00000 103900.00000 109600.00000 166200.00000 154900.00000   \n",
       "1996-06-01 103800.00000 104700.00000 110000.00000 165400.00000 154200.00000   \n",
       "1996-07-01 104700.00000 105500.00000 110400.00000 164400.00000 153500.00000   \n",
       "1996-08-01 105400.00000 106200.00000 110800.00000 163500.00000 152800.00000   \n",
       "...                 ...          ...          ...          ...          ...   \n",
       "2017-12-01 261100.00000 262800.00000 273500.00000 402300.00000 396200.00000   \n",
       "2018-01-01 261300.00000 262800.00000 273800.00000 406200.00000 397500.00000   \n",
       "2018-02-01 261400.00000 262300.00000 273800.00000 409800.00000 398900.00000   \n",
       "2018-03-01 261700.00000 262000.00000 273900.00000 412600.00000 402100.00000   \n",
       "2018-04-01 262000.00000 262400.00000 274600.00000 413300.00000 404800.00000   \n",
       "\n",
       "RegionName        99901  \n",
       "time                     \n",
       "1996-04-01 117100.00000  \n",
       "1996-05-01 117200.00000  \n",
       "1996-06-01 117300.00000  \n",
       "1996-07-01 117400.00000  \n",
       "1996-08-01 117500.00000  \n",
       "...                 ...  \n",
       "2017-12-01 291900.00000  \n",
       "2018-01-01 294200.00000  \n",
       "2018-02-01 297500.00000  \n",
       "2018-03-01 302100.00000  \n",
       "2018-04-01 305100.00000  \n",
       "\n",
       "[265 rows x 14723 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# zip_df = melted.pivot_table(index='time', columns='RegionName', values='value')\n",
    "# zip_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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>1001</th>\n",
       "      <th>1002</th>\n",
       "      <th>1005</th>\n",
       "      <th>1007</th>\n",
       "      <th>1008</th>\n",
       "      <th>1010</th>\n",
       "      <th>1011</th>\n",
       "      <th>1013</th>\n",
       "      <th>1020</th>\n",
       "      <th>1026</th>\n",
       "      <th>...</th>\n",
       "      <th>99654</th>\n",
       "      <th>99664</th>\n",
       "      <th>99669</th>\n",
       "      <th>99701</th>\n",
       "      <th>99705</th>\n",
       "      <th>99709</th>\n",
       "      <th>99712</th>\n",
       "      <th>99801</th>\n",
       "      <th>99835</th>\n",
       "      <th>99901</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1996-04-01</th>\n",
       "      <td>113100.0</td>\n",
       "      <td>161000.0</td>\n",
       "      <td>103100.0</td>\n",
       "      <td>133400.0</td>\n",
       "      <td>117500.0</td>\n",
       "      <td>115800.0</td>\n",
       "      <td>87000.0</td>\n",
       "      <td>88500.0</td>\n",
       "      <td>97200.0</td>\n",
       "      <td>91200.0</td>\n",
       "      <td>...</td>\n",
       "      <td>122700.0</td>\n",
       "      <td>112200.0</td>\n",
       "      <td>112500.0</td>\n",
       "      <td>89400.0</td>\n",
       "      <td>101900.0</td>\n",
       "      <td>103000.0</td>\n",
       "      <td>109200.0</td>\n",
       "      <td>166800.0</td>\n",
       "      <td>155500.0</td>\n",
       "      <td>117100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-05-01</th>\n",
       "      <td>112800.0</td>\n",
       "      <td>160100.0</td>\n",
       "      <td>103400.0</td>\n",
       "      <td>132700.0</td>\n",
       "      <td>117300.0</td>\n",
       "      <td>115700.0</td>\n",
       "      <td>87000.0</td>\n",
       "      <td>88500.0</td>\n",
       "      <td>97100.0</td>\n",
       "      <td>91100.0</td>\n",
       "      <td>...</td>\n",
       "      <td>124200.0</td>\n",
       "      <td>112400.0</td>\n",
       "      <td>113100.0</td>\n",
       "      <td>89800.0</td>\n",
       "      <td>102900.0</td>\n",
       "      <td>103900.0</td>\n",
       "      <td>109600.0</td>\n",
       "      <td>166200.0</td>\n",
       "      <td>154900.0</td>\n",
       "      <td>117200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-06-01</th>\n",
       "      <td>112600.0</td>\n",
       "      <td>159300.0</td>\n",
       "      <td>103600.0</td>\n",
       "      <td>132000.0</td>\n",
       "      <td>117100.0</td>\n",
       "      <td>115500.0</td>\n",
       "      <td>87100.0</td>\n",
       "      <td>88400.0</td>\n",
       "      <td>97000.0</td>\n",
       "      <td>91000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>125700.0</td>\n",
       "      <td>112600.0</td>\n",
       "      <td>113800.0</td>\n",
       "      <td>90200.0</td>\n",
       "      <td>103800.0</td>\n",
       "      <td>104700.0</td>\n",
       "      <td>110000.0</td>\n",
       "      <td>165400.0</td>\n",
       "      <td>154200.0</td>\n",
       "      <td>117300.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-07-01</th>\n",
       "      <td>112300.0</td>\n",
       "      <td>158600.0</td>\n",
       "      <td>103800.0</td>\n",
       "      <td>131400.0</td>\n",
       "      <td>117000.0</td>\n",
       "      <td>115300.0</td>\n",
       "      <td>87200.0</td>\n",
       "      <td>88300.0</td>\n",
       "      <td>96800.0</td>\n",
       "      <td>90800.0</td>\n",
       "      <td>...</td>\n",
       "      <td>127200.0</td>\n",
       "      <td>112800.0</td>\n",
       "      <td>114400.0</td>\n",
       "      <td>90600.0</td>\n",
       "      <td>104700.0</td>\n",
       "      <td>105500.0</td>\n",
       "      <td>110400.0</td>\n",
       "      <td>164400.0</td>\n",
       "      <td>153500.0</td>\n",
       "      <td>117400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-08-01</th>\n",
       "      <td>112100.0</td>\n",
       "      <td>158000.0</td>\n",
       "      <td>103900.0</td>\n",
       "      <td>130800.0</td>\n",
       "      <td>116800.0</td>\n",
       "      <td>115100.0</td>\n",
       "      <td>87400.0</td>\n",
       "      <td>88200.0</td>\n",
       "      <td>96600.0</td>\n",
       "      <td>90700.0</td>\n",
       "      <td>...</td>\n",
       "      <td>128600.0</td>\n",
       "      <td>113000.0</td>\n",
       "      <td>115100.0</td>\n",
       "      <td>91000.0</td>\n",
       "      <td>105400.0</td>\n",
       "      <td>106200.0</td>\n",
       "      <td>110800.0</td>\n",
       "      <td>163500.0</td>\n",
       "      <td>152800.0</td>\n",
       "      <td>117500.0</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>2017-12-01</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>342000.0</td>\n",
       "      <td>213300.0</td>\n",
       "      <td>268100.0</td>\n",
       "      <td>205400.0</td>\n",
       "      <td>243800.0</td>\n",
       "      <td>151400.0</td>\n",
       "      <td>173300.0</td>\n",
       "      <td>187400.0</td>\n",
       "      <td>201200.0</td>\n",
       "      <td>...</td>\n",
       "      <td>258800.0</td>\n",
       "      <td>219100.0</td>\n",
       "      <td>250200.0</td>\n",
       "      <td>206700.0</td>\n",
       "      <td>261100.0</td>\n",
       "      <td>262800.0</td>\n",
       "      <td>273500.0</td>\n",
       "      <td>402300.0</td>\n",
       "      <td>396200.0</td>\n",
       "      <td>291900.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>344500.0</td>\n",
       "      <td>215600.0</td>\n",
       "      <td>268100.0</td>\n",
       "      <td>204500.0</td>\n",
       "      <td>244600.0</td>\n",
       "      <td>150100.0</td>\n",
       "      <td>174600.0</td>\n",
       "      <td>188100.0</td>\n",
       "      <td>204600.0</td>\n",
       "      <td>...</td>\n",
       "      <td>259400.0</td>\n",
       "      <td>218900.0</td>\n",
       "      <td>250600.0</td>\n",
       "      <td>206100.0</td>\n",
       "      <td>261300.0</td>\n",
       "      <td>262800.0</td>\n",
       "      <td>273800.0</td>\n",
       "      <td>406200.0</td>\n",
       "      <td>397500.0</td>\n",
       "      <td>294200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02-01</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>347400.0</td>\n",
       "      <td>217900.0</td>\n",
       "      <td>268800.0</td>\n",
       "      <td>206800.0</td>\n",
       "      <td>246700.0</td>\n",
       "      <td>150100.0</td>\n",
       "      <td>175900.0</td>\n",
       "      <td>189000.0</td>\n",
       "      <td>211200.0</td>\n",
       "      <td>...</td>\n",
       "      <td>259900.0</td>\n",
       "      <td>218300.0</td>\n",
       "      <td>250100.0</td>\n",
       "      <td>205600.0</td>\n",
       "      <td>261400.0</td>\n",
       "      <td>262300.0</td>\n",
       "      <td>273800.0</td>\n",
       "      <td>409800.0</td>\n",
       "      <td>398900.0</td>\n",
       "      <td>297500.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-01</th>\n",
       "      <td>222700.0</td>\n",
       "      <td>350600.0</td>\n",
       "      <td>219500.0</td>\n",
       "      <td>270000.0</td>\n",
       "      <td>210900.0</td>\n",
       "      <td>248800.0</td>\n",
       "      <td>149500.0</td>\n",
       "      <td>176900.0</td>\n",
       "      <td>189600.0</td>\n",
       "      <td>217900.0</td>\n",
       "      <td>...</td>\n",
       "      <td>260200.0</td>\n",
       "      <td>218400.0</td>\n",
       "      <td>249300.0</td>\n",
       "      <td>205500.0</td>\n",
       "      <td>261700.0</td>\n",
       "      <td>262000.0</td>\n",
       "      <td>273900.0</td>\n",
       "      <td>412600.0</td>\n",
       "      <td>402100.0</td>\n",
       "      <td>302100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-04-01</th>\n",
       "      <td>223600.0</td>\n",
       "      <td>353300.0</td>\n",
       "      <td>220700.0</td>\n",
       "      <td>270600.0</td>\n",
       "      <td>214200.0</td>\n",
       "      <td>250900.0</td>\n",
       "      <td>149100.0</td>\n",
       "      <td>177400.0</td>\n",
       "      <td>189800.0</td>\n",
       "      <td>220800.0</td>\n",
       "      <td>...</td>\n",
       "      <td>260500.0</td>\n",
       "      <td>219100.0</td>\n",
       "      <td>249300.0</td>\n",
       "      <td>205700.0</td>\n",
       "      <td>262000.0</td>\n",
       "      <td>262400.0</td>\n",
       "      <td>274600.0</td>\n",
       "      <td>413300.0</td>\n",
       "      <td>404800.0</td>\n",
       "      <td>305100.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>265 rows × 14723 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                1001      1002      1005      1007      1008      1010  \\\n",
       "time                                                                     \n",
       "1996-04-01  113100.0  161000.0  103100.0  133400.0  117500.0  115800.0   \n",
       "1996-05-01  112800.0  160100.0  103400.0  132700.0  117300.0  115700.0   \n",
       "1996-06-01  112600.0  159300.0  103600.0  132000.0  117100.0  115500.0   \n",
       "1996-07-01  112300.0  158600.0  103800.0  131400.0  117000.0  115300.0   \n",
       "1996-08-01  112100.0  158000.0  103900.0  130800.0  116800.0  115100.0   \n",
       "...              ...       ...       ...       ...       ...       ...   \n",
       "2017-12-01  221700.0  342000.0  213300.0  268100.0  205400.0  243800.0   \n",
       "2018-01-01  221700.0  344500.0  215600.0  268100.0  204500.0  244600.0   \n",
       "2018-02-01  221700.0  347400.0  217900.0  268800.0  206800.0  246700.0   \n",
       "2018-03-01  222700.0  350600.0  219500.0  270000.0  210900.0  248800.0   \n",
       "2018-04-01  223600.0  353300.0  220700.0  270600.0  214200.0  250900.0   \n",
       "\n",
       "                1011      1013      1020      1026  ...     99654     99664  \\\n",
       "time                                                ...                       \n",
       "1996-04-01   87000.0   88500.0   97200.0   91200.0  ...  122700.0  112200.0   \n",
       "1996-05-01   87000.0   88500.0   97100.0   91100.0  ...  124200.0  112400.0   \n",
       "1996-06-01   87100.0   88400.0   97000.0   91000.0  ...  125700.0  112600.0   \n",
       "1996-07-01   87200.0   88300.0   96800.0   90800.0  ...  127200.0  112800.0   \n",
       "1996-08-01   87400.0   88200.0   96600.0   90700.0  ...  128600.0  113000.0   \n",
       "...              ...       ...       ...       ...  ...       ...       ...   \n",
       "2017-12-01  151400.0  173300.0  187400.0  201200.0  ...  258800.0  219100.0   \n",
       "2018-01-01  150100.0  174600.0  188100.0  204600.0  ...  259400.0  218900.0   \n",
       "2018-02-01  150100.0  175900.0  189000.0  211200.0  ...  259900.0  218300.0   \n",
       "2018-03-01  149500.0  176900.0  189600.0  217900.0  ...  260200.0  218400.0   \n",
       "2018-04-01  149100.0  177400.0  189800.0  220800.0  ...  260500.0  219100.0   \n",
       "\n",
       "               99669     99701     99705     99709     99712     99801  \\\n",
       "time                                                                     \n",
       "1996-04-01  112500.0   89400.0  101900.0  103000.0  109200.0  166800.0   \n",
       "1996-05-01  113100.0   89800.0  102900.0  103900.0  109600.0  166200.0   \n",
       "1996-06-01  113800.0   90200.0  103800.0  104700.0  110000.0  165400.0   \n",
       "1996-07-01  114400.0   90600.0  104700.0  105500.0  110400.0  164400.0   \n",
       "1996-08-01  115100.0   91000.0  105400.0  106200.0  110800.0  163500.0   \n",
       "...              ...       ...       ...       ...       ...       ...   \n",
       "2017-12-01  250200.0  206700.0  261100.0  262800.0  273500.0  402300.0   \n",
       "2018-01-01  250600.0  206100.0  261300.0  262800.0  273800.0  406200.0   \n",
       "2018-02-01  250100.0  205600.0  261400.0  262300.0  273800.0  409800.0   \n",
       "2018-03-01  249300.0  205500.0  261700.0  262000.0  273900.0  412600.0   \n",
       "2018-04-01  249300.0  205700.0  262000.0  262400.0  274600.0  413300.0   \n",
       "\n",
       "               99835     99901  \n",
       "time                            \n",
       "1996-04-01  155500.0  117100.0  \n",
       "1996-05-01  154900.0  117200.0  \n",
       "1996-06-01  154200.0  117300.0  \n",
       "1996-07-01  153500.0  117400.0  \n",
       "1996-08-01  152800.0  117500.0  \n",
       "...              ...       ...  \n",
       "2017-12-01  396200.0  291900.0  \n",
       "2018-01-01  397500.0  294200.0  \n",
       "2018-02-01  398900.0  297500.0  \n",
       "2018-03-01  402100.0  302100.0  \n",
       "2018-04-01  404800.0  305100.0  \n",
       "\n",
       "[265 rows x 14723 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zip_df = pd.read_pickle('../data/wip/zip_df.pkl')\n",
    "zip_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      13684\n",
       "15         1\n",
       "21         2\n",
       "87       118\n",
       "93        31\n",
       "99        21\n",
       "105       52\n",
       "107       13\n",
       "111       78\n",
       "117       56\n",
       "123       27\n",
       "129       20\n",
       "135        7\n",
       "141        6\n",
       "147        9\n",
       "153       11\n",
       "159        7\n",
       "165        1\n",
       "167      230\n",
       "171       41\n",
       "177       33\n",
       "183       24\n",
       "189       27\n",
       "195       18\n",
       "201       55\n",
       "207       42\n",
       "213       53\n",
       "219       56\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many rows have x number of missing values (number of rows is the value column of output, number of missing is index)\n",
    "missing_count = zip_df.isna().sum().value_counts().sort_index()\n",
    "missing_count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1039"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many rows are missing values? (sum of counts in missing_count minus the number of entries with zero missing: 13684)\n",
    "missing_count.sum() - 13684"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.07056985668681655"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# About Seven Percent of zipcodes are missing 15 or more entries\n",
    "1039/len(zip_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2018-04-01 00:00:00')"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zip_df.index.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    14723\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Are there any missing values recently-- last six months-- (no), if so discard those\n",
    "zip_df.iloc[zip_df.index > '2017-10-01'].isna().sum().value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Project Plan\n",
    "1) Shape data by zipcode(columns), indexed by date\n",
    "\n",
    "    - Plan to deal with missing values, how many months missing is okay? Imputation potentially?\n",
    "\n",
    "2) Test models on 1 zipcode, outputting best model (accuracy measure?) and best differencing params\n",
    "\n",
    "    - Difference/decompose each column before fitting models, use adfuller as stationarity test\n",
    "    - Use a 75-25 Train/Test Split\n",
    "\n",
    "3) Create decision function based off ROI adjusted for loss scenarios/conf int.\n",
    "\n",
    "4) Test models on 10 zipcodes, outputting best model (accuracy measure?), best differencing params for each zipcode, as well as largest predicted ROI\n",
    "\n",
    "5) Test models on all zipcodes as in step 4\n",
    "\n",
    "6) Modular add-ons (rent offset, performance over time, etc.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['1001', '1002', '1005', '1007', '1008', '1010', '1011', '1013', '1020',\n",
       "       '1026',\n",
       "       ...\n",
       "       '99654', '99664', '99669', '99701', '99705', '99709', '99712', '99801',\n",
       "       '99835', '99901'],\n",
       "      dtype='object', length=14723)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert columns to string for easier index access, move this step higher up the workflow\n",
    "zip_df.columns = [str(x) for x in zip_df.columns]\n",
    "zip_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "zip_df.to_pickle('../data/wip/zip_df.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "# Notes from checkin:\n",
    "\n",
    "# Results by client class, who can take advantage of x? Do you need minimum amount of capital etc.?\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time\n",
       "1996-04-01    113100.0\n",
       "1996-05-01    112800.0\n",
       "1996-06-01    112600.0\n",
       "1996-07-01    112300.0\n",
       "1996-08-01    112100.0\n",
       "                ...   \n",
       "2017-12-01    221700.0\n",
       "2018-01-01    221700.0\n",
       "2018-02-01    221700.0\n",
       "2018-03-01    222700.0\n",
       "2018-04-01    223600.0\n",
       "Name: 1001, Length: 265, dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zip_df['1001']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-1.7467926412738275,\n",
       " 0.40717889973199395,\n",
       " 13,\n",
       " 251,\n",
       " {'1%': -3.4566744514553016,\n",
       "  '5%': -2.8731248767783426,\n",
       "  '10%': -2.5729436702592023},\n",
       " 3563.9412458378274)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from statsmodels.tsa.stattools import adfuller\n",
    "adfuller(zip_df['1001'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "from statsmodels.tsa.seasonal import seasonal_decompose"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "decomposition = seasonal_decompose(zip_df['1001'])\n",
    "fig = plt.figure()\n",
    "fig = decomposition.plot()\n",
    "fig.set_size_inches(15, 8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "one_train = zip_df['1001']\n",
    "\n",
    "fig, axs = plt.subplots(2, figsize=(15,10))\n",
    "axs[0] = one_train.plot()\n",
    "axs[1] = one_train.diff().plot()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-1.7467926412738275,\n",
       " 0.40717889973199395,\n",
       " 13,\n",
       " 251,\n",
       " {'1%': -3.4566744514553016,\n",
       "  '5%': -2.8731248767783426,\n",
       "  '10%': -2.5729436702592023},\n",
       " 3563.9412458378274)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Stationarity test with no lag\n",
    "adfuller(one_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-1.9453525470037503,\n",
       " 0.3110356104138362,\n",
       " 12,\n",
       " 251,\n",
       " {'1%': -3.4566744514553016,\n",
       "  '5%': -2.8731248767783426,\n",
       "  '10%': -2.5729436702592023},\n",
       " 3552.048382789642)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Stationarity with 1 lag\n",
    "adfuller(one_train.diff()[1:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-6.627156477050341,\n",
       " 5.8372030590914575e-09,\n",
       " 11,\n",
       " 251,\n",
       " {'1%': -3.4566744514553016,\n",
       "  '5%': -2.8731248767783426,\n",
       "  '10%': -2.5729436702592023},\n",
       " 3540.671779819917)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Stationarity with 2 lags\n",
    "adfuller(one_train.diff()[1:].diff()[1:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time\n",
       "1996-06-01     100.0\n",
       "1996-07-01    -100.0\n",
       "1996-08-01     100.0\n",
       "1996-09-01       0.0\n",
       "1996-10-01     100.0\n",
       "               ...  \n",
       "2017-12-01    -400.0\n",
       "2018-01-01    -600.0\n",
       "2018-02-01       0.0\n",
       "2018-03-01    1000.0\n",
       "2018-04-01    -100.0\n",
       "Name: 1001, Length: 263, dtype: float64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "one_train.diff()[1:].diff()[1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time\n",
       "2012-11-01    182100.0\n",
       "2012-12-01    182800.0\n",
       "2013-01-01    183400.0\n",
       "2013-02-01    183700.0\n",
       "2013-03-01    183300.0\n",
       "                ...   \n",
       "2017-12-01    221700.0\n",
       "2018-01-01    221700.0\n",
       "2018-02-01    221700.0\n",
       "2018-03-01    222700.0\n",
       "2018-04-01    223600.0\n",
       "Name: 1001, Length: 66, dtype: float64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_ind = one_train.index[0:round(len(one_train)*.75)]\n",
    "test_ind = [i for i in one_train.index if not i in train_ind]\n",
    "test_ind\n",
    "one_test = one_train[test_ind]\n",
    "one_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time\n",
       "1996-04-01    113100.0\n",
       "1996-05-01    112800.0\n",
       "1996-06-01    112600.0\n",
       "1996-07-01    112300.0\n",
       "1996-08-01    112100.0\n",
       "                ...   \n",
       "2012-06-01    182400.0\n",
       "2012-07-01    181200.0\n",
       "2012-08-01    180400.0\n",
       "2012-09-01    180400.0\n",
       "2012-10-01    181200.0\n",
       "Name: 1001, Length: 199, dtype: float64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "one_train = one_train[train_ind]\n",
    "one_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time\n",
       "1996-06-01    100.0\n",
       "1996-07-01   -100.0\n",
       "1996-08-01    100.0\n",
       "1996-09-01      0.0\n",
       "1996-10-01    100.0\n",
       "              ...  \n",
       "2012-06-01   -300.0\n",
       "2012-07-01    400.0\n",
       "2012-08-01    400.0\n",
       "2012-09-01    800.0\n",
       "2012-10-01    800.0\n",
       "Name: 1001, Length: 197, dtype: float64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "one_train.diff()[1:].diff()[1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series([], Name: 1001, dtype: float64)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "one_train[one_train.index == '2018-04-01']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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>1001</th>\n",
       "      <th>1002</th>\n",
       "      <th>1005</th>\n",
       "      <th>1007</th>\n",
       "      <th>1008</th>\n",
       "      <th>1010</th>\n",
       "      <th>1011</th>\n",
       "      <th>1013</th>\n",
       "      <th>1020</th>\n",
       "      <th>1026</th>\n",
       "      <th>...</th>\n",
       "      <th>99654</th>\n",
       "      <th>99664</th>\n",
       "      <th>99669</th>\n",
       "      <th>99701</th>\n",
       "      <th>99705</th>\n",
       "      <th>99709</th>\n",
       "      <th>99712</th>\n",
       "      <th>99801</th>\n",
       "      <th>99835</th>\n",
       "      <th>99901</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1996-04-01</th>\n",
       "      <td>113100.0</td>\n",
       "      <td>161000.0</td>\n",
       "      <td>103100.0</td>\n",
       "      <td>133400.0</td>\n",
       "      <td>117500.0</td>\n",
       "      <td>115800.0</td>\n",
       "      <td>87000.0</td>\n",
       "      <td>88500.0</td>\n",
       "      <td>97200.0</td>\n",
       "      <td>91200.0</td>\n",
       "      <td>...</td>\n",
       "      <td>122700.0</td>\n",
       "      <td>112200.0</td>\n",
       "      <td>112500.0</td>\n",
       "      <td>89400.0</td>\n",
       "      <td>101900.0</td>\n",
       "      <td>103000.0</td>\n",
       "      <td>109200.0</td>\n",
       "      <td>166800.0</td>\n",
       "      <td>155500.0</td>\n",
       "      <td>117100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-05-01</th>\n",
       "      <td>112800.0</td>\n",
       "      <td>160100.0</td>\n",
       "      <td>103400.0</td>\n",
       "      <td>132700.0</td>\n",
       "      <td>117300.0</td>\n",
       "      <td>115700.0</td>\n",
       "      <td>87000.0</td>\n",
       "      <td>88500.0</td>\n",
       "      <td>97100.0</td>\n",
       "      <td>91100.0</td>\n",
       "      <td>...</td>\n",
       "      <td>124200.0</td>\n",
       "      <td>112400.0</td>\n",
       "      <td>113100.0</td>\n",
       "      <td>89800.0</td>\n",
       "      <td>102900.0</td>\n",
       "      <td>103900.0</td>\n",
       "      <td>109600.0</td>\n",
       "      <td>166200.0</td>\n",
       "      <td>154900.0</td>\n",
       "      <td>117200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-06-01</th>\n",
       "      <td>112600.0</td>\n",
       "      <td>159300.0</td>\n",
       "      <td>103600.0</td>\n",
       "      <td>132000.0</td>\n",
       "      <td>117100.0</td>\n",
       "      <td>115500.0</td>\n",
       "      <td>87100.0</td>\n",
       "      <td>88400.0</td>\n",
       "      <td>97000.0</td>\n",
       "      <td>91000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>125700.0</td>\n",
       "      <td>112600.0</td>\n",
       "      <td>113800.0</td>\n",
       "      <td>90200.0</td>\n",
       "      <td>103800.0</td>\n",
       "      <td>104700.0</td>\n",
       "      <td>110000.0</td>\n",
       "      <td>165400.0</td>\n",
       "      <td>154200.0</td>\n",
       "      <td>117300.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-07-01</th>\n",
       "      <td>112300.0</td>\n",
       "      <td>158600.0</td>\n",
       "      <td>103800.0</td>\n",
       "      <td>131400.0</td>\n",
       "      <td>117000.0</td>\n",
       "      <td>115300.0</td>\n",
       "      <td>87200.0</td>\n",
       "      <td>88300.0</td>\n",
       "      <td>96800.0</td>\n",
       "      <td>90800.0</td>\n",
       "      <td>...</td>\n",
       "      <td>127200.0</td>\n",
       "      <td>112800.0</td>\n",
       "      <td>114400.0</td>\n",
       "      <td>90600.0</td>\n",
       "      <td>104700.0</td>\n",
       "      <td>105500.0</td>\n",
       "      <td>110400.0</td>\n",
       "      <td>164400.0</td>\n",
       "      <td>153500.0</td>\n",
       "      <td>117400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996-08-01</th>\n",
       "      <td>112100.0</td>\n",
       "      <td>158000.0</td>\n",
       "      <td>103900.0</td>\n",
       "      <td>130800.0</td>\n",
       "      <td>116800.0</td>\n",
       "      <td>115100.0</td>\n",
       "      <td>87400.0</td>\n",
       "      <td>88200.0</td>\n",
       "      <td>96600.0</td>\n",
       "      <td>90700.0</td>\n",
       "      <td>...</td>\n",
       "      <td>128600.0</td>\n",
       "      <td>113000.0</td>\n",
       "      <td>115100.0</td>\n",
       "      <td>91000.0</td>\n",
       "      <td>105400.0</td>\n",
       "      <td>106200.0</td>\n",
       "      <td>110800.0</td>\n",
       "      <td>163500.0</td>\n",
       "      <td>152800.0</td>\n",
       "      <td>117500.0</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>2017-12-01</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>342000.0</td>\n",
       "      <td>213300.0</td>\n",
       "      <td>268100.0</td>\n",
       "      <td>205400.0</td>\n",
       "      <td>243800.0</td>\n",
       "      <td>151400.0</td>\n",
       "      <td>173300.0</td>\n",
       "      <td>187400.0</td>\n",
       "      <td>201200.0</td>\n",
       "      <td>...</td>\n",
       "      <td>258800.0</td>\n",
       "      <td>219100.0</td>\n",
       "      <td>250200.0</td>\n",
       "      <td>206700.0</td>\n",
       "      <td>261100.0</td>\n",
       "      <td>262800.0</td>\n",
       "      <td>273500.0</td>\n",
       "      <td>402300.0</td>\n",
       "      <td>396200.0</td>\n",
       "      <td>291900.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>344500.0</td>\n",
       "      <td>215600.0</td>\n",
       "      <td>268100.0</td>\n",
       "      <td>204500.0</td>\n",
       "      <td>244600.0</td>\n",
       "      <td>150100.0</td>\n",
       "      <td>174600.0</td>\n",
       "      <td>188100.0</td>\n",
       "      <td>204600.0</td>\n",
       "      <td>...</td>\n",
       "      <td>259400.0</td>\n",
       "      <td>218900.0</td>\n",
       "      <td>250600.0</td>\n",
       "      <td>206100.0</td>\n",
       "      <td>261300.0</td>\n",
       "      <td>262800.0</td>\n",
       "      <td>273800.0</td>\n",
       "      <td>406200.0</td>\n",
       "      <td>397500.0</td>\n",
       "      <td>294200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02-01</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>347400.0</td>\n",
       "      <td>217900.0</td>\n",
       "      <td>268800.0</td>\n",
       "      <td>206800.0</td>\n",
       "      <td>246700.0</td>\n",
       "      <td>150100.0</td>\n",
       "      <td>175900.0</td>\n",
       "      <td>189000.0</td>\n",
       "      <td>211200.0</td>\n",
       "      <td>...</td>\n",
       "      <td>259900.0</td>\n",
       "      <td>218300.0</td>\n",
       "      <td>250100.0</td>\n",
       "      <td>205600.0</td>\n",
       "      <td>261400.0</td>\n",
       "      <td>262300.0</td>\n",
       "      <td>273800.0</td>\n",
       "      <td>409800.0</td>\n",
       "      <td>398900.0</td>\n",
       "      <td>297500.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-01</th>\n",
       "      <td>222700.0</td>\n",
       "      <td>350600.0</td>\n",
       "      <td>219500.0</td>\n",
       "      <td>270000.0</td>\n",
       "      <td>210900.0</td>\n",
       "      <td>248800.0</td>\n",
       "      <td>149500.0</td>\n",
       "      <td>176900.0</td>\n",
       "      <td>189600.0</td>\n",
       "      <td>217900.0</td>\n",
       "      <td>...</td>\n",
       "      <td>260200.0</td>\n",
       "      <td>218400.0</td>\n",
       "      <td>249300.0</td>\n",
       "      <td>205500.0</td>\n",
       "      <td>261700.0</td>\n",
       "      <td>262000.0</td>\n",
       "      <td>273900.0</td>\n",
       "      <td>412600.0</td>\n",
       "      <td>402100.0</td>\n",
       "      <td>302100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-04-01</th>\n",
       "      <td>223600.0</td>\n",
       "      <td>353300.0</td>\n",
       "      <td>220700.0</td>\n",
       "      <td>270600.0</td>\n",
       "      <td>214200.0</td>\n",
       "      <td>250900.0</td>\n",
       "      <td>149100.0</td>\n",
       "      <td>177400.0</td>\n",
       "      <td>189800.0</td>\n",
       "      <td>220800.0</td>\n",
       "      <td>...</td>\n",
       "      <td>260500.0</td>\n",
       "      <td>219100.0</td>\n",
       "      <td>249300.0</td>\n",
       "      <td>205700.0</td>\n",
       "      <td>262000.0</td>\n",
       "      <td>262400.0</td>\n",
       "      <td>274600.0</td>\n",
       "      <td>413300.0</td>\n",
       "      <td>404800.0</td>\n",
       "      <td>305100.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>265 rows × 14723 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                1001      1002      1005      1007      1008      1010  \\\n",
       "time                                                                     \n",
       "1996-04-01  113100.0  161000.0  103100.0  133400.0  117500.0  115800.0   \n",
       "1996-05-01  112800.0  160100.0  103400.0  132700.0  117300.0  115700.0   \n",
       "1996-06-01  112600.0  159300.0  103600.0  132000.0  117100.0  115500.0   \n",
       "1996-07-01  112300.0  158600.0  103800.0  131400.0  117000.0  115300.0   \n",
       "1996-08-01  112100.0  158000.0  103900.0  130800.0  116800.0  115100.0   \n",
       "...              ...       ...       ...       ...       ...       ...   \n",
       "2017-12-01  221700.0  342000.0  213300.0  268100.0  205400.0  243800.0   \n",
       "2018-01-01  221700.0  344500.0  215600.0  268100.0  204500.0  244600.0   \n",
       "2018-02-01  221700.0  347400.0  217900.0  268800.0  206800.0  246700.0   \n",
       "2018-03-01  222700.0  350600.0  219500.0  270000.0  210900.0  248800.0   \n",
       "2018-04-01  223600.0  353300.0  220700.0  270600.0  214200.0  250900.0   \n",
       "\n",
       "                1011      1013      1020      1026  ...     99654     99664  \\\n",
       "time                                                ...                       \n",
       "1996-04-01   87000.0   88500.0   97200.0   91200.0  ...  122700.0  112200.0   \n",
       "1996-05-01   87000.0   88500.0   97100.0   91100.0  ...  124200.0  112400.0   \n",
       "1996-06-01   87100.0   88400.0   97000.0   91000.0  ...  125700.0  112600.0   \n",
       "1996-07-01   87200.0   88300.0   96800.0   90800.0  ...  127200.0  112800.0   \n",
       "1996-08-01   87400.0   88200.0   96600.0   90700.0  ...  128600.0  113000.0   \n",
       "...              ...       ...       ...       ...  ...       ...       ...   \n",
       "2017-12-01  151400.0  173300.0  187400.0  201200.0  ...  258800.0  219100.0   \n",
       "2018-01-01  150100.0  174600.0  188100.0  204600.0  ...  259400.0  218900.0   \n",
       "2018-02-01  150100.0  175900.0  189000.0  211200.0  ...  259900.0  218300.0   \n",
       "2018-03-01  149500.0  176900.0  189600.0  217900.0  ...  260200.0  218400.0   \n",
       "2018-04-01  149100.0  177400.0  189800.0  220800.0  ...  260500.0  219100.0   \n",
       "\n",
       "               99669     99701     99705     99709     99712     99801  \\\n",
       "time                                                                     \n",
       "1996-04-01  112500.0   89400.0  101900.0  103000.0  109200.0  166800.0   \n",
       "1996-05-01  113100.0   89800.0  102900.0  103900.0  109600.0  166200.0   \n",
       "1996-06-01  113800.0   90200.0  103800.0  104700.0  110000.0  165400.0   \n",
       "1996-07-01  114400.0   90600.0  104700.0  105500.0  110400.0  164400.0   \n",
       "1996-08-01  115100.0   91000.0  105400.0  106200.0  110800.0  163500.0   \n",
       "...              ...       ...       ...       ...       ...       ...   \n",
       "2017-12-01  250200.0  206700.0  261100.0  262800.0  273500.0  402300.0   \n",
       "2018-01-01  250600.0  206100.0  261300.0  262800.0  273800.0  406200.0   \n",
       "2018-02-01  250100.0  205600.0  261400.0  262300.0  273800.0  409800.0   \n",
       "2018-03-01  249300.0  205500.0  261700.0  262000.0  273900.0  412600.0   \n",
       "2018-04-01  249300.0  205700.0  262000.0  262400.0  274600.0  413300.0   \n",
       "\n",
       "               99835     99901  \n",
       "time                            \n",
       "1996-04-01  155500.0  117100.0  \n",
       "1996-05-01  154900.0  117200.0  \n",
       "1996-06-01  154200.0  117300.0  \n",
       "1996-07-01  153500.0  117400.0  \n",
       "1996-08-01  152800.0  117500.0  \n",
       "...              ...       ...  \n",
       "2017-12-01  396200.0  291900.0  \n",
       "2018-01-01  397500.0  294200.0  \n",
       "2018-02-01  398900.0  297500.0  \n",
       "2018-03-01  402100.0  302100.0  \n",
       "2018-04-01  404800.0  305100.0  \n",
       "\n",
       "[265 rows x 14723 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zip_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9770114942528736"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ROI from beginning to end of table\n",
    "\n",
    "(zip_df['1001'][-1]/zip_df['1001'][0]) - 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Lag: 14\n",
      "Coefficients: [-4.76619252  0.60141155 -0.38384097  0.0622626  -0.00643859 -0.21530151\n",
      "  0.26172035 -0.16637076  0.08781042 -0.31303174  0.26210386  0.03789678\n",
      " -0.29948158 -0.1878581   0.250815  ]\n",
      "72\n",
      "263\n",
      "predicted=184694.000000, expected=184000.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=184378.000000, expected=182400.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=183827.000000, expected=181200.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=182352.000000, expected=180400.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=181068.000000, expected=180400.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=180890.000000, expected=181200.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=180796.000000, expected=182100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=179758.000000, expected=182800.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=178851.000000, expected=183400.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=178553.000000, expected=183700.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=178000.000000, expected=183300.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=177190.000000, expected=183500.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=176853.000000, expected=184800.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=176280.000000, expected=187000.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=175090.000000, expected=188600.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=174420.000000, expected=189100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=174591.000000, expected=188700.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=174071.000000, expected=188100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=172661.000000, expected=187100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=171785.000000, expected=185900.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=171521.000000, expected=185100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=170733.000000, expected=184800.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=169840.000000, expected=184900.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=169459.000000, expected=184500.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=168734.000000, expected=184500.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=167460.000000, expected=185100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=166927.000000, expected=185800.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=166881.000000, expected=186400.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=165879.000000, expected=186900.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=164433.000000, expected=187400.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=163995.000000, expected=187900.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=163817.000000, expected=188800.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=162818.000000, expected=189800.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=161862.000000, expected=190800.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=161466.000000, expected=191500.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=160547.000000, expected=192200.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=159301.000000, expected=192400.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=158939.000000, expected=192100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=158717.000000, expected=191500.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=157377.000000, expected=191000.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=156090.000000, expected=191200.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=155904.000000, expected=192300.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=155468.000000, expected=193600.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=154101.000000, expected=194600.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=153197.000000, expected=195300.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=152844.000000, expected=195900.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=151796.000000, expected=195900.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=150620.000000, expected=196100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=150374.000000, expected=197100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=149841.000000, expected=198000.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=148242.000000, expected=199300.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=147201.000000, expected=201100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=147167.000000, expected=202100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=146368.000000, expected=202300.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=144818.000000, expected=202600.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=144136.000000, expected=202700.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=143787.000000, expected=202600.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=142504.000000, expected=203900.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=141376.000000, expected=206000.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=141155.000000, expected=208200.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=140292.000000, expected=210500.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=138590.000000, expected=212500.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=137889.000000, expected=213900.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=137838.000000, expected=215700.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=136600.000000, expected=218200.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=135004.000000, expected=220100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=134586.000000, expected=221100.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=134091.000000, expected=221700.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=132554.000000, expected=221700.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=131555.000000, expected=221700.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=131351.000000, expected=222700.000000\n",
      "Test RMSE: 45681.431\n",
      "predicted=130163.000000, expected=223600.000000\n",
      "Test RMSE: 45681.431\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.7/site-packages/statsmodels/tsa/ar_model.py:691: FutureWarning: \n",
      "statsmodels.tsa.AR has been deprecated in favor of statsmodels.tsa.AutoReg and\n",
      "statsmodels.tsa.SARIMAX.\n",
      "\n",
      "AutoReg adds the ability to specify exogenous variables, include time trends,\n",
      "and add seasonal dummies. The AutoReg API differs from AR since the model is\n",
      "treated as immutable, and so the entire specification including the lag\n",
      "length must be specified when creating the model. This change is too\n",
      "substantial to incorporate into the existing AR api. The function\n",
      "ar_select_order performs lag length selection for AutoReg models.\n",
      "\n",
      "AutoReg only estimates parameters using conditional MLE (OLS). Use SARIMAX to\n",
      "estimate ARX and related models using full MLE via the Kalman Filter.\n",
      "\n",
      "To silence this warning and continue using AR until it is removed, use:\n",
      "\n",
      "import warnings\n",
      "warnings.filterwarnings('ignore', 'statsmodels.tsa.ar_model.AR', FutureWarning)\n",
      "\n",
      "  warnings.warn(AR_DEPRECATION_WARN, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAD4CAYAAAAZ1BptAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deXiU1dnH8e8NIQkEAkkgbAFCMOyI7CiuiIpWcW+1VhFR6vrqq61LrVqXvrXaVsUdV3DfC1oVEbS4IBCQfZFIEAIJYQ1hy3reP86DRkyAhISZJL/Pdc01M2eeeeYODLl5znIfc84hIiJSlnqhDkBERMKXkoSIiJRLSUJERMqlJCEiIuVSkhARkXJFhDqAqta8eXOXnJwc6jBERGqUOXPmbHTOtdi7vdYlieTkZNLS0kIdhohIjWJmP5TVru4mEREpl5KEiIiUS0lCRETKpSQhIiLlUpIQEZFyKUmIiEi5lCRERKRcShIiIjXc6k07ufeDJWzdWVDl5651i+lEROoC5xxpP2zhuS8ymLwkm/pmHJmSwLDuLav0c5QkRERqkJISx+TF2Tz13++Zn5lL04YNuPr4TlxyZDItY6Or/POUJEREaoDiEsdHi7J4dGo6y9fnkZzQiHvP6sm5fdvSKLL6fpUrSYiIhEhBUQmzV21m3pqtFJeUv5V0UYnjw4VZpOdsp1OLGB654AhOP7wN9etZtceoJCEicgjl7ixkytL1TFu2nunfbWR7ftEBva9LyyY8emEfTuvV+pAkhz32myTMrB0wAWgFlADjnHOPmNmDwBlAAfA9MMo5tzV4z23AaKAY+B/n3OSgfTjwCFAfeNY5d3/Q3hF4HYgH5gIXO+cKzCwq+Ox+wCbgN865VVX0s4uIHDKbdxTw3JcrGf/1D2zPLyKxSRRn9G7N0K4tObJTAg0b1N/n+w9lYijtQK4kioCbnHNzzawJMMfMpgBTgNucc0Vm9nfgNuAWM+sOXAD0ANoAn5pZ5+BcjwMnAZnAbDOb5JxbAvwdeMg597qZPYVPME8G91ucc4eZ2QXBcb+pop9dRKTabdqezzNfZDBhxip2FRZzWq/WXHFMCr2TmmIWml/8FbHfJOGcywKygsd5ZrYUaOuc+6TUYd8A5wWPzwRed87lAxlmlg4MDF5Ld86tBDCz14Ezg/MNBX4bHDMe+As+SZwZPAZ4G3jMzMw5V37nnYhImJi/ZiuXvjCL3F2FnNG7DdeecBipLZuEOqwKqdCYhJklA32AmXu9dBnwRvC4LT5p7JEZtAGs2at9EJAAbHXOFZVxfNs97wmuWHKD4zfuFdcYYAxA+/btK/IjiYhUi6+/38gV49NIaBzFG78/ks41LDnsccArrs2sMfAOcINzblup9tvxXVKv7Gkq4+2uEu37OtfPG5wb55zr75zr36LFL3bfExE5pD5dsp5LX5hN27iGvHVlzU0QcIBXEmbWAJ8gXnHOvVuqfSRwOnBiqS6gTKBdqbcnAeuCx2W1bwSamVlEcDVR+vg958o0swigKbD5wH88EZFDa+K8tdz45nx6tonlxVEDiYuJDHVIB2W/VxLmR1aeA5Y65/5Vqn04cAswwjm3s9RbJgEXmFlUMGspFZgFzAZSzayjmUXiB7cnBcnlM34a0xgJTCx1rpHB4/OAaRqPEJFw9fGibG54Yx79O8TxyhWDa3yCgAO7khgCXAwsNLN5QdufgLFAFDAlGKH/xjl3pXNusZm9CSzBd0Nd45wrBjCza4HJ+CmwzzvnFgfnuwV43czuA77FJyWC+5eCwe/N+MQiIhJ2lmfnceOb8zg8qRnjLxtI9H6mtNYUVtv+Y96/f3+XlpYW6jBEpA7ZurOAEY99xa7CYj647uhqqaFU3cxsjnOu/97tKhUuInIQiopLuPbVb8nO3c1Tv+tXIxPEvqgsh4jIQbj/o2V8mb6RB849nH4d4kIdTpXTlYSISCW9NGMVz36ZwcgjO/DrAe32e3xNpCsJEZEKcs7x0JTvGDstnaFdE/nz6d1DHVK1UZIQEamAwuIS/vTuQt6ak8lv+rfjr2f3JKJ+7e2UUZIQETlAO/KLuPqVufz3uw1cf2IqNwxLrRFF+g6GkoSIyAFYlr2NG16fx4qc7dx/Ti8uGFg36sQpSYiI7ENxiePZL1byz0++I7ZhBM9fOoDjOtedGnFKEiIi5VizeSc3vTmfWas2M7xHK/56dk8SGkeFOqxDSklCRKQMs1dt5rIXZgPwz/N7c07ftrV+/KEsShIiInuZuXITo16cTavYaCaMHkhSXKNQhxQyShIiIqXM+H4Tl704mzbNonntisEk1rIyGxWlJCEiEvgqfSOjx8+mXVwjXr1iMC2a1K3xh7LU3hUgIiIVsOcKIjkhhtfGKEHsoSsJEanzlmfnMealNNrFN+KVywfVuRlM+6IrCRGp07JzdzPqhVk0bFCfF0cNUILYi5KEiNRZebsLufSFWeTuKuSFUQPq9Cym8qi7SUTqpIKiEq56eS7pOdt5/tIB9GjTNNQhhSUlCRGpc0pKHLe8s4Av0zfy4HmHc2wdKrNRUepuEpE6xTnHff9ZynvfruWmkzpzfv/auVlQVVGSEJE65YnPv+f5rzIYNSSZa4ceFupwwp6ShIjUGa/M/IEHJy/n7D5tueNX3etkLaaKUpIQkTrhw4VZ/PnfixjaNZEHzjucevWUIA6EkoSI1HqfLcvh+te/pV/7OB7/bV8a1OLtRqua/qREpFb7Kn0jv395Dl1aNeG5SwfQMLJ+qEOqUZQkRKTWmpWxmcvHp5HSPIaXLhtE04YNQh1SjaMkISK10rw1W38s+f3S6EHExUSGOqQaSUlCRGqdjxdlcclzM4mPieSVy1XR9WBoxbWI1Brrt+3mzomLmLx4PT3axPL0xf1o1bRubxp0sJQkRCTsbdtdSO7OQnYWFLM9v4gd+UXUM6NRVH0aR0UQExXB9O828H8fLqWgqIRbT+3K5Ud3JEKzmA6akoSIhJRzjtxdheTk5ZOzLZ+cvN1k5e4mY+OOH2+bdxQc0LmOTEngb+f0Irl5TDVHXXcoSYhISJSUOD5alM2j01awLDvvF6+3jI2iY/MYTunRiuSERsTFRBITGUFMVH1ioiJwDnYU+KuKnfnFxMVEMqxbolZRVzElCRE5pIpLHP9ZmMWjU1ewImc7KS1iuO3UrrRp1pDEJlEkxkbTMjaKRpH69RQO9vu3YGbtgAlAK6AEGOece8TM4oE3gGRgFfBr59wW82n8EeA0YCdwqXNubnCukcCfg1Pf55wbH7T3A14EGgIfAtc751x5n3HQP7WIhMTarbu4YnwaS7K2kZrYmLEX9uFXvVpTXyUywtaBjOoUATc557oBg4FrzKw7cCsw1TmXCkwNngOcCqQGtzHAkwDBL/y7gEHAQOAuM4sL3vNkcOye9w0P2sv7DBGpYZas28Y5T3zFmi07GXthHybfcCwjerdRgghz+00SzrmsPVcCzrk8YCnQFjgTGB8cNh44K3h8JjDBed8AzcysNXAKMMU5tzm4GpgCDA9ei3XOzXDOOfxVS+lzlfUZIlKDfJW+kV8/PQPDePvKoxjRu40K7NUQFZofZmbJQB9gJtDSOZcFPpEAicFhbYE1pd6WGbTtqz2zjHb28Rl7xzXGzNLMLG3Dhg0V+ZFEpJpNnLeWS1+YRdtmDXnvmqPo0qpJqEOSCjjgkSEzawy8A9zgnNu2jxkEZb3gKtF+wJxz44BxAP3796/Qe0WkemzbXcjfPlzGa7NWMzglnqcv7q/aSTXQASUJM2uATxCvOOfeDZrXm1lr51xW0GWUE7RnAqX3A0wC1gXtx+/V/nnQnlTG8fv6DBEJYx8vyubOiYvYuD2fy4/uyB+HdyEqQtVXa6L9djcFs5WeA5Y65/5V6qVJwMjg8UhgYqn2S8wbDOQGXUWTgZPNLC4YsD4ZmBy8lmdmg4PPumSvc5X1GSIShjK37OTKl+Zw5ctziI+J5N/XDOHPp3dXgqjBDuRKYghwMbDQzOYFbX8C7gfeNLPRwGrg/OC1D/HTX9PxU2BHATjnNpvZvcDs4Lh7nHObg8dX8dMU2I+CG/v4DBEJI/PWbOXZL1by0aJs6tcz/nhKF8Ycm6LNfWoB8xOKao/+/fu7tLS0UIchUuuVlDimLF3PM9NXkvbDFppER/Dbge0ZeVQybZo1DHV4UkFmNsc513/vdi1pFJEK2bucRrv4htx1RnfO79+OxlH6lVLb6G9URPYpv6g4KLyXz/c523nmi5WsyNlOpxYxPPybIzhDC+JqNSUJkTrMOcey7DymLcth2rIcfti082evFxQVs2130c/aOrdszKMX9uE0ldOoE5QkROqYkhJH2g9b+GDBOj5dsp51ubsBODypKSd1/3kV1Yh6RovGUSTGRpHYJJqWsdF0bdVEq6XrECUJkTpi0dpcJs5bywcLssjK3U10g3ock9qC64elckKXRBJjtYOb/JKShEgt5pxjxvebeHjqCmZlbKZBfeO4zi249dSuDOvWkhgNNMt+6BsiUgs55/gyfSNjp65g9qottIyN4o7Tu3Nu37Y0axQZ6vCkBlGSEKll1mzeyZ/eW8gXKzbSumk0957Zg/P7tyO6gVY9S8UpSYjUEsUljhe+yuCfn3xH/XrGXWd057eD2qskhhwUJQmRauScIyt3Nzl5+azf5u+37SqkddNoOjaPIaV5Y5o2OvjKqEvWbeO2dxcwPzOXE7smcu9ZPbXqWaqEkoRIFXPOsSRrG5Pmr+OD+Vms3bprn8fHx0TSr0McJ3ZN5ISuibSswCyjdVt38a8p3/HO3EziG0Xy6IV9OP3w1uyjlL9IhShJiFSR7zds5/3565g0fx0rN+wgop5xTGpzrjwuhTbNGpLYJJrE2CiaNmzA2q27yNiwg4yNO1iRk8dX6ZuYsmQ9AL3aNuW4zi0YlBJPvw5xNIr85T/T3F2FPPF5Oi9+tQrn4IpjUrj6+E4alJYqpwJ/Igchc8tOPliQxaR561iStQ0zGNQxnhG92zK8ZyviYw7sl7ZzjuXr85i61K98nrdmK8Uljoh6xuFJTeneJpatOwuD8hi7ycrdTUFxCWcf0ZYbT+5MUlyjav5JpbYrr8CfkoRIJaxYn8ej09J5f8E6nIMj2jXjjN5tOP3w1hXqLirP9vwi0lZtZlbGZmZmbGbF+jwSGkfRokkUiU2iaBUbzdl929KjTdMq+GlEVAVWpEosz85j7LQVfLgwi4YN6jPmmBQuGtSB9glV+z/5xlERHN8lkeO7lLmtu8ghoyQhcgCWrNvGo9NW8NGibGIi63PVcZ24/JiUA+5OEqmplCSE3J2FzFq1me/W55GcEEPPtrG0j2+kGTL4ekdjp67gkyXraRIVwXVDD2P00R01QCx1hpJEHbVifR6vzFzNzIzNLMvext5DU02iI+jZpilHdUrghK6J9GgTW2eShnOOr9I38eyXK/l8+QZioyO4YVgqo47qWCVrGkRqEg1c10GT5q/jlrcX4HD06xDHoI4JDOoYT/c2sazauJNF63JZtDaX+ZlbWbzOJ5BWsdGc0DWRM3q35siUhFqZMPKLipk4bx3Pf5nBsuw8mjeO4tKjOnDJUcnERis5SO2mgWuhqLiE+z9axrNfZjAgOY7HL+pLYpOfz8TpldSUXkk/zZjZkJfP58tz+Gx5Du/PX8drs1bTr0Mc/3NiKsemNq/RyaK4xLFk3TZmZmzim5WbmZWxiW27i+jaqgkPnnc4I45oo5IWUufpSqKO2Lg9n2tfncs3Kzdz6VHJ/Om0bkRG1KvQOXYXFvP2nEye/Px71m7dRe92zbjhxFSO79KiRiSLwuISFq3NZWbGZmau3ETaqi3k5ftd15ITGjGoYwJn9G7DkMNq55WSyL5onUQdtn7bbs576mtytuXzt3N6cU7fpIM6X0FRCe/MzeTxz9LJ3LKLYzu34K4zutOpReMqirhqFBSVsHDtVr5Z6dcapK3azM6CYgBSWsQwqGMCg1PiGdQxgVZNteGO1G1KEnVU7s5Cfv30DDK37OTlywfRp31clZ27sLiEl2b8wENTvmN3UTGXHd2R64am0jiEG9ms3rSTacvWM235BmZnbGZXoU8KqYmNGZQSz+CUBAZ2jP9FN5tIXackUQftKijmd8/NZGFmLi+MGsCQw5pXy+dsyMvngY+X8dacTFrGRjFqSEfO65dE88ZR1fJ5pTnnWLR2Gx8sXMe0pTmsyNkOQErzGI5Jbf5jUkg4BLGI1GRKEnVMYXEJV0xIY/p3G3jior4M79m62j9z7uot3P/RMmZlbCainnFS95ZcOLA9Qw5rTv16VdvHn56Tx6T5Wbw/fx0ZG30xvUEp8Qzt2pKhXRPp2DymSj9PpLZTktiflSuhoAA6dYIGNXu6Y0FRCX98ez4T563jb+f04sKB7Q/p56fn5PH6rDW8MzeTLTsLiYyoR4f4RnRsHkPHFjG0aBzFjvxidhYUsT2/iMLiEg5LbEzPtk3p0aYpTRv+8s+/sLiE2as289myHKYuy2Hlhh2YwZEpCYzo3YbhPVtpgZvIQVCS2J+rr4Ynn4SICJ8ounaF7t3h+OPhmGOgYc3YwGXOD1u47d0FfLd+OzcP78LVxx8Wsljyi4qZujSH+ZlbfyyL/cOmnRQUlwAQFVGPxlERmMHG7QU/vq99fCPiSi1ac0DGhh3k5RcRWb8eg1LiGdatJaf2bEViFRTTExElif366K3PyJ72JRc13UVk+newdCl89x0UFUFUlE8UJ58Ml14KLVpUfeD7kLe7kEaREfvsstmeX8Q/Ji9n/IxVtI6N5r6zezK0a8tDF+QBKi5xbM8volFkfRrU/2kK7sbt+Sxam8viddtYsm4bOwqKfva+lk38Yr6jU5uHdGBcpLZSktiP57/M4L7/LCGlRWOevrifn865YwdMnw5TpsAnn8DixdCyJbz0Epx0UpXHXlLiWLA2l6/SN7Jyww4yNm4nY+MOtuwspGGD+nRvE0uvtk3p0SaWZo0iycnbHewvkM9/l+eQtW03I49M5g+ndNEvUhGpECWJA/B1+kaufe1bCopK+Oeve3NKj1Y/P2D+fPjtb2HJErj5ZrjvvoMev9iRX8T07zYwbZlf1byn26VVbPSPffjt4xuRnbubRWtzWZK17ce5/nskxESS0iKGW0/tRr8OVTfFVUTqDiWJA7R26y6ufnkO8zNzueaETtwwrPPPukXYuRNuvBGefhoGDoRXX/VjGBWwu7CYz5dv4P0F65i6dD27C0uIjY7guC6JnNg1keM6tyCunBLUxSWOjI3b2ZFfTGJsFM0bR/08PhGRSlCSqIDdhcXcNXExb6StITWxMXeP6MFRe68xePttuOIKPyPqnnvg+uv9oPc+LM/O44WvMvjPgizy8ouIj4nktF6t+FWvNgxIjiNCv+xFJESUJCphypL13PPBYtZs3sWverXmT7/qRttmpWY5rVkD11wD778PffvCuHHQr9/PzuGcY/qKjTz7xUq+WLGR6Ab1+FWvNow4og1DOiUoMYhIWKh0kjCz54HTgRznXM+g7QjgKSAaKAKuds7NMl8V7RHgNGAncKlzbm7wnpHAn4PT3uecGx+09wNeBBoCHwLXO+ecmcUDbwDJwCrg1865Lfv7Qat6Md3uwmLGTV/JE5+nA3DBgPZcOLA9XVo18Qc4B++8A9ddBzk5MHo09OzJjmYJTM813llTwKLtRqP4ppx/XBcuOKpTuV1JIiKhcjBJ4lhgOzChVJL4BHjIOfeRmZ0G3OycOz54fB0+SQwCHnHODQp+4acB/fHT3ucA/ZxzW8xsFnA98A0+SYwNzvsAsNk5d7+Z3QrEOedu2d8PWl0rrjO37OQfk5fz4cJsCopL6NO+GRcOaM9RhyVQzwzL3UrsX+6g4SsTqFdYWP6JoqLg2GPh7rvhyCOrPE4Rkcqo9H4SzrnpZpa8dzMQGzxuCqwLHp+JTyYO+MbMmplZa+B4YIpzbnMQzBRguJl9DsQ652YE7ROAs4CPgnMdH5x3PPA5sN8kUV2S4hrx8AV9uPOMAt6dm8lrs1Zz8zsLfn5Qy7Ox/z2TJLeLX7eL5IxWESSX7IC8PNi+3U+p3bQJXnkFjjoKTj0V7r33F11UIiLhorKT6W8AJpvZP4B6wFFBe1tgTanjMoO2fbVnltEO0NI5lwXgnMsys8TygjGzMcAYgPbtq7cERXxMJJcfk8Loozsyd/UW0oOCcnvERjfg+C6JNIzcx2Y1994Ljz0GDzwA/fvDGWfAH/8IRx8N2sdARMJIZZPEVcD/OufeMbNfA88Bw4CyfsO5SrRXiHNuHDAOfHdTRd9fGWZGvw7x9OsQX/E3x8TALbfAVVfBww/DI4/4we8BA+Cmm+Dcc/c7U0pE5FCo7NSakcC7weO3gIHB40ygXanjkvBdUftqTyqjHWB90FVFcJ9TyVjDV2ws3HmnnyX1xBOwdStccAF06AAjRsAf/uDXY3z2me+yEhE5xCqbJNYBxwWPhwIrgseTgEvMGwzkBl1Gk4GTzSzOzOKAk4HJwWt5ZjY4mBl1CTCx1LlGBo9HlmqvfRo18lcVy5bBxIl+vCIjAx5/HK68EoYOhfh4P+B9330wcyYUF+//vCIiB+lAZje9hh9Abg6sB+4CluOnukYAu/FTYOcEv+gfA4bjp8COcs6lBee5DPhTcNq/OudeCNr789MU2I+A64IpsAnAm0B7YDVw/p6B732pVftJlJRAZqZPHp9/7utHzZ3rp922bu2n244eDcnJoY5URGo4LaarLTZsgE8/hZdfho8+8m2nnOKr0/brBx07Qv19DJqLiJRBSaI2+uEHeO45f1sXDOVER/u9MHr2hOOOg+HDISlp3+cRkTpPSaI2KyqCtDRfnXbxYn+bPx+ys/3rPXv6ZDFypH8sIrIXJYm6xjmfLD7+2N+mT/eD3Vdf7Vd7x1di6q6I1FrlJQlVl6utzPxVwx/+4McwsrN9gnjiCejc2U+t1QwpEdkPJYm6Ij4eHn0Uvv3WJ48rr/T3997rZ0+JiJRB3U11kXPw1lswdix89ZVv69kTTjsNCgt9NdsNG2DjRjjsML9OY+hQ/1hlQ0RqJY1JSNnWroV33/VJ48sv/cK+xERo0QLi4mDRIn8MQNu2fvvWv/zFHycitYaShOxfUdEva0Y5B+npMG0aTJ4M773np9i+/LKq14rUIhq4lv0rq6igGaSmwu9/7684pkzxdaQGD/YlQoqKDn2cInLIKElIxQwbBgsXwvnnwx13+DpTkyZpppRILaUkIRUXFwevvupvWVlw5pl+Wu1DD0FubqijE5EqpDEJOThFRX6cYuzYnwa+e/XyNaQ6doSUFDjmGOjSJdSRisg+aOBaqt+cOfDCC37dRUYGrF7905jFkCFw+eW+myomJrRxisgvaOBaql+/fn5b1k8/he+/h127/MyoBx7way5GjfIlzq+6CpYuDXW0InIAlCSk+kREQKdOfv/upUvhiy/gnHP81Ub37r7o4OTJfpqtiIQldTfJobdhg68d9fjjvqZUly6+rPkRR0CfPn5MQ11SIoeUxiQk/OTnwxtvwPjxvqbUli2+3cwPdv/ud3DeeX42lYhUK41JSPiJioJLLoGpU2HTJr+J0sSJcPvtsH49jBkDrVr5LqopU0IdrUidpCQh4cEM2reHESN8ZdqlS/1GSldfDV9/DSefDBdc8NNGSiJySChJSHgy87OlHnrIX2Hccw/8+9++btTTT0NJSagjFKkTlCQk/EVF+RIgCxZA375+L4xBg/zU2jlzVBJEpBqVUdFNJEx17uzHL8aPhwcfhFtu8e1xcX52VMuWEBn5023wYDjjDO2BIXIQNLtJaq6sLF/CfOpUvwZj2zYoKPAbJ+3e7a8weveGO++Es86CerpwFimPZjdJ7dO6NVx0ETz/PKxY4WdEbdkC27f7JDFhgl/1fe65Plm8916oIxapcZQkpHaKiICLL4YlS/wGSYWFfirtxRerUq1IBShJSO1Wv76/2li0yG+7+uqrfmX311+HOjKRGkFJQuqGiAi46y4/drFnRfcdd/hChLVsXE6kKmngWuqebdvg2mvhpZf88+bN/ZTaQYN8QcKWLSEx0d83b64Bb6kTVLtJZG8LF/pup5kz/W3p0l9eVbRvD7feCpdd5tdriNRSShIi+5OXB2vX+llSOTm+BMhrr8GMGdC2rU8Wl18O0dGhjlSkyilJiFSGc34dxt13++1ZW7WCG2/0q76bNAl1dCJVRuskRCrDDIYNg+nT4bPPoEcPuPlm3w11xx1+bwyRWkxXEiIVNXs23H+/X5wXHe3Lf/Tt+9Otc2cNdkuNoysJkaoyYAC88w4sXuzHKHbs8Ht7X3QRdOsGqal+170dO0IdqchB22+SMLPnzSzHzBbt1X6dmS03s8Vm9kCp9tvMLD147ZRS7cODtnQzu7VUe0czm2lmK8zsDTOLDNqjgufpwevJVfEDi1SZbt1g7Fg/Myovz1epffZZP3322mt9l9Sdd/qBcJEa6kCuJF4EhpduMLMTgDOBw51zPYB/BO3dgQuAHsF7njCz+mZWH3gcOBXoDlwYHAvwd+Ah51wqsAUYHbSPBrY45w4DHgqOEwlPDRr4vblHj/bTar/8Eo49Fu67z6+9+Mc/fGkQkRpmv0nCOTcd2LxX81XA/c65/OCYnKD9TOB151y+cy4DSAcGBrd059xK51wB8DpwppkZMBR4O3j/eOCsUucaHzx+GzgxOF4kvJnBkCF+zGLpUjjhBPjjH/0mSioHIjVMZcckOgPHBN1A/zWzAUF7W2BNqeMyg7by2hOArc65or3af3au4PXc4PhfMLMxZpZmZmkbNNtEwkmXLjBpkk8YW7f65DFqlN9lb+VK7bAnYa+ymw5FAHHAYGAA8KaZpQBl/U/fUXYycvs4nv289vNG58YB48DPbtpn5CKHmpnfz2LYML/e4uGH4cUX/WuNG0PPnnDYYZCU5BftJSVB//7+XiTEKpskMoF3nZ8/O8vMSoDmQXu7UsclAeuCx2W1bwSamVlEcLVQ+vg958o0swigKb/s9hKpORo39jvq3XWXnxm1cKEf7F6wwI9hrF3707hFdDT89a9w/fW+kq1IiFS2u+nf+AIskzsAAA4zSURBVLEEzKwzEIn/hT8JuCCYmdQRSAVmAbOB1GAmUyR+cHtSkGQ+A84LzjsSmBg8nhQ8J3h9mqttizqkbmrc2BcTvPxyPzvq888hI8NvlJSd7WdLnXQS3HSTr1a7bFmoI5Y67ECmwL4GzAC6mFmmmY0GngdSgmmxrwMjnbcYeBNYAnwMXOOcKw6uEq4FJgNLgTeDYwFuAW40s3T8mMNzQftzQELQfiPw47RZkVqpXj1feXbgQJg40W+WtHy53//igQegqGj/5xCpYlpxLRLOsrPh6qv9wHefPvDMM36WlEgV04prkZqoVSu/uvvttyEry19l3HSTVnPLIVPZgWsROVTM4Nxz4cQTfbnyf/3LJ44TT/RJpGVLf9+5s58pFaF/1lJ19G0SqSmaNYOnnvI1om6/HT76yO97UVz80zGNGvnuqEGD4JRT/LRbkYOgMQmRmqykBDZt8mMXCxf+tMvet99CQQH85je+2GBCmetQRX6kTYdE6pL8fF8v6u67IT4enn4azjwz1FFJGNPAtUhdEhXlu6TS0qB1a7/i++KLYfXqUEcmNYyShEhtdvjhvvvprrvg9dchJQUuuABmzQp1ZFJDKEmI1HaRkfCXv/iCgjfeCB9/7Ae2jznGd0lNnuxLgtSyrmepGkoSInVFu3Z+5faaNb7IYHa2L2E+fLgvJpiQ4Kfazp8f6kgljChJiNQ1TZr4woErVsDGjb521GOPwXnnwdSpvgzIb36jmlECKEmI1G0JCXDccXDNNTBunC80ePvt8J//QI8ecOmlvitK6iwlCRH5SVyc33I1IwNuuMEPdnfpAv/3f75KrdQ5ShIi8kstWsA//wlLlviV27ffDt27w7vv+kV6UmeoLIeIlC8lxdeJmjrVX1mce66vJdW6NXTo4G+9e/s6Un37aoOkWkhJQkT278QTfamP997zVxc//ACrVvk1GK+/7o9p1gxOOAHOPht+9zufTKTGU5IQkQMTEQHnn//L9vXrYdo0f7Xx6ac+kXzwATz3nN+FT2o0jUmIyMFp2RIuvBCefdYPeN9/v9//YvBgP81WajQlCRGpOmZwyy1+FXd2NvTvD++/H+qo5CAoSYhI1Rs2DObMgcMOgxEjfDfVkiWhjkoqQUlCRKpHhw7w5Zdwxx2+XlTPnn5AW11QNYr2kxCR6rdxIzz4IDz6qF9n0b69Hwjfc0tJgcsug9NO0/arIaJNh0Qk9LKzfaJYswaKivytsNBPpc3KgjZtYNQoGD0aOnYMdbR1ipKEiISvoiJfL+qZZ/ze3fXqwd//Dv/7v1pvcYhoZzoRCV8REX571Q8+8Iv0RoyAm27yK7xzc0MdXZ2mJCEi4aVdO7/O4p//9NNn+/WDefNCHVWdpSQhIuHHzO+i9/nnvvrskUfC73/vV3YXFYU6ujpFSUJEwteQITB3rl9n8corvoZU27Zw1VW+/Ed6OhQXhzrKWk1JQkTCW2IiTJgAOTm+G+r44/3zc86B1FS/096AAX6Qe9OmUEdb62h2k4jUPDt3wqJFP90WLvRdU82a+f27f/tbzYqqoPJmN2nViojUPI0awcCB/rbHwoVwxRV+VfeECfDkk36RnhwUdTeJSO3Qqxd89RU89hjMmOHLgNx1F2zfHurIajQlCRGpPerXh2uu8cUER4yAe+7x4xbPPqsB7kpSkhCR2icpye+YN2OGL+9xxRVwxBFw773w4ot+c6Rly/z0Wtmn/SYJM3vezHLMbFEZr/3BzJyZNQ+em5mNNbN0M1tgZn1LHTvSzFYEt5Gl2vuZ2cLgPWPN/GiTmcWb2ZTg+ClmFlc1P7KI1BmDB/suqLfe8usr7rzT14Y66STo1g1atfLrMb7/PtSRhq0DuZJ4ERi+d6OZtQNOAlaXaj4VSA1uY4Ang2PjgbuAQcBA4K5Sv/SfDI7d8749n3UrMNU5lwpMDZ6LiFSMGZx3HixdCrt2+YTw3//Cyy/Dqaf6goOpqXD66f4KQ35mv0nCOTcd2FzGSw8BNwOl59CeCUxw3jdAMzNrDZwCTHHObXbObQGmAMOD12KdczOcn4s7ATir1LnGB4/Hl2oXEamc6Gg/4+nYY+Gii+C11+CHH/yeF2lp/grjD3/Q+EUplRqTMLMRwFrn3Py9XmoLrCn1PDNo21d7ZhntAC2dc1kAwX3iPuIZY2ZpZpa2YcOGSvxEIlJntWkDd9/tk8V11/maUWedBXl5oY4sLFQ4SZhZI+B24M6yXi6jzVWivUKcc+Occ/2dc/1btGhR0beLiEBUFIwdC48/7suVDxniE0cdV5kriU5AR2C+ma0CkoC5ZtYKfyXQrtSxScC6/bQnldEOsD7ojiK4z6lErCIiFXP11T5JrF7tF+s9/rgvCVJHVThJOOcWOucSnXPJzrlk/C/6vs65bGAScEkwy2kwkBt0FU0GTjazuGDA+mRgcvBanpkNDmY1XQJMDD5qErBnFtTIUu0iItXrpJP89Nm2beHaa6F1azj5ZHj+eV8GZMUKn0Sys2v9NNoDmQL7GjAD6GJmmWY2eh+HfwisBNKBZ4CrAZxzm4F7gdnB7Z6gDeAq4NngPd8DHwXt9wMnmdkK/Cyq+yv2o4mIHIRu3XwF2oUL4dZb/ayo0aP9yu7OnaFDB588Wrb03VS1tIS5CvyJiBwI52DOHFi5EgoKID/f3//73/DJJ9Cnj68XNWhQqCOtFO1xLSJSHZzzJcxvuAGysmDMGHjwQV/CvAbRHtciItXBzG+KtGyZ39PimWf8gPeyZaGOrEooSYiIVIUmTfwai08/9ZsfDRzod8+r4ZQkRESq0gkn+LGLbt387nm33VajB7WVJEREqlq7djB9uh+fuP9+iI/3yePmm32xwTVr9n+OMKGd6UREqkNUFDz9NPzqVzB5MsyeDY884mdEASQn+xpSxx0HQ4f652FIs5tERA6VggK/7uLrr30l2unTYcMGP/j9l7/An/8M9ULTwaPZTSIioRYZCf36+UKCb78N69f7XfQuushvtXrqqT5phBElCRGRUDHzA9wTJsC4cf7qok8fv1FSmFCSEBEJNTO/xeo33/g9L447Dv7nfyAzc//vrWZKEiIi4eKII/z02csu8yU+OnXyVWlXr97/e6uJBq5FRMLRqlXwt7/BCy/458ccA0lJvqhgmzbQs6efVmtlbctTcardJCJSE61eDf/4B8ya5WtDZWVBYaF/bdgwP622e/eD/hjNbhIRqYnat/elyL/5xu+Ut3u33wTp0Uf9vtyHH+5rRm3dWi0fryQhIlKT1KsHLVr4zZBWrIDLL/dXE507w2efVf3HVfkZRUTk0GjeHJ56yg929+njE0UVU1kOEZGark8fX/qjGuhKQkREyqUkISIi5VKSEBGRcilJiIhIuZQkRESkXEoSIiJSLiUJEREpl5KEiIiUq9YV+DOzDcAPlXx7c2BjFYZTnRRr9VCs1aOmxFpT4oSqj7WDc67F3o21LkkcDDNLK6sKYjhSrNVDsVaPmhJrTYkTDl2s6m4SEZFyKUmIiEi5lCR+blyoA6gAxVo9FGv1qCmx1pQ44RDFqjEJEREpl64kRESkXEoSIiJSLiWJgJkNN7PlZpZuZreGOp7SzOx5M8sxs0Wl2uLNbIqZrQju40IZYxBTOzP7zMyWmtliM7s+jGONNrNZZjY/iPXuoL2jmc0MYn3DzCJDHeseZlbfzL41sw+C52EZq5mtMrOFZjbPzNKCtrD7DgCYWTMze9vMlgXf2yPDMVYz6xL8ee65bTOzGw5FrEoS+H98wOPAqUB34EIz6x7aqH7mRWD4Xm23AlOdc6nA1OB5qBUBNznnugGDgWuCP8dwjDUfGOqc6w0cAQw3s8HA34GHgli3AKNDGOPergeWlnoezrGe4Jw7otQ8/nD8DgA8AnzsnOsK9Mb/+YZdrM655cGf5xFAP2An8B6HIlbnXJ2/AUcCk0s9vw24LdRx7RVjMrCo1PPlQOvgcWtgeahjLCPmicBJ4R4r0AiYCwzCr2CNKOt7EeIYk4JfAkOBDwAL41hXAc33agu77wAQC2QQTOAJ51j3iu9k4KtDFauuJLy2wJpSzzODtnDW0jmXBRDcJ4Y4np8xs2SgDzCTMI016L6ZB+QAU4Dvga3OuaLgkHD6HjwM3AyUBM8TCN9YHfCJmc0xszFBWzh+B1KADcALQTfes2YWQ3jGWtoFwGvB42qPVUnCszLaNDe4ksysMfAOcINzbluo4ymPc67Y+cv3JGAg0K2sww5tVL9kZqcDOc65OaWbyzg05LEGhjjn+uK7b68xs2NDHVA5IoC+wJPOuT7ADsKga2lfgnGnEcBbh+ozlSS8TKBdqedJwLoQxXKg1ptZa4DgPifE8QBgZg3wCeIV59y7QXNYxrqHc24r8Dl+HKWZmUUEL4XL92AIMMLMVgGv47ucHiY8Y8U5ty64z8H3mw8kPL8DmUCmc25m8PxtfNIIx1j3OBWY65xbHzyv9liVJLzZQGowWyQSfzk3KcQx7c8kYGTweCS+/z+kzMyA54Clzrl/lXopHGNtYWbNgscNgWH4QcvPgPOCw8IiVufcbc65JOdcMv67Oc05dxFhGKuZxZhZkz2P8f3niwjD74BzLhtYY2ZdgqYTgSWEYaylXMhPXU1wKGIN9SBMuNyA04Dv8P3St4c6nr1iew3IAgrx//sZje+TngqsCO7jwyDOo/FdHguAecHttDCN9XDg2yDWRcCdQXsKMAtIx1/SR4U61r3iPh74IFxjDWKaH9wW7/m3FI7fgSCuI4C04HvwbyAujGNtBGwCmpZqq/ZYVZZDRETKpe4mEREpl5KEiIiUS0lCRETKpSQhIiLlUpIQEZFyKUmIiEi5lCRERKRc/w8J0ZXFudPk9AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from statsmodels.tsa.ar_model import AR\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from math import sqrt\n",
    "\n",
    "# split dataset using second order differenced data\n",
    "diff_1 = zip_df['1001'].diff()[1:]\n",
    "diff_2 = diff_1.diff()[1:]\n",
    "test_undiff = zip_df['1001'][-72:].values\n",
    "X = zip_df['1001'].diff()[1:].diff()[1:].values\n",
    "train, test = X[0:len(X)-72], X[len(X)-72:]\n",
    "# train autoregression\n",
    "model = AR(train)\n",
    "model_fit = model.fit()\n",
    "print('Lag: %s' % model_fit.k_ar)\n",
    "print('Coefficients: %s' % model_fit.params)\n",
    "# make predictions on the differenced data\n",
    "predictions = model_fit.predict(start=len(train), end=len(train)+len(test)-1, dynamic=False)\n",
    "print(len(predictions))\n",
    "\n",
    "print(X.size)\n",
    "np.r_[train, predictions].size\n",
    "#convert predictions back to values\n",
    "\n",
    "predicted_values = transform_diff2s(zip_df['1001'][0], diff_1[0], np.r_[train, predictions])\n",
    "\n",
    "model_predictions = predicted_values[-72:]\n",
    "# print(X[-72], zip_df['1001'][[-74, -72]])\n",
    "for i in range(len(model_predictions)):\n",
    "    print('predicted=%f, expected=%f' % (model_predictions[i], test_undiff[i]))\n",
    "    rmse = sqrt(mean_squared_error(test_undiff, model_predictions))\n",
    "    print('Test RMSE: %.3f' % rmse)\n",
    "# plot results\n",
    "plt.plot(test_undiff)\n",
    "plt.plot(model_predictions, color='red')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([113100, 112800, 112600, 112300, 112100, 111900, 111800, 111700,\n",
       "       111700, 111900, 112100, 112300, 112500, 112800, 113000, 113300,\n",
       "       113600, 113900, 114200, 114500, 114800, 115200, 115600, 115900,\n",
       "       116200, 116400, 116600, 116700, 116800, 116900, 117100, 117300,\n",
       "       117500, 117800, 118300, 118700, 119100, 119500, 120000, 120400,\n",
       "       120900, 121300, 121700, 122100, 122600, 123100, 123800, 124400,\n",
       "       124900, 125400, 125900, 126400, 127000, 127500, 128100, 128800,\n",
       "       129400, 130200, 131100, 131900, 132700, 133400, 134100, 134800,\n",
       "       135400, 136100, 136800, 137500, 138400, 139400, 140600, 141800,\n",
       "       143000, 144200, 145400, 146800, 148200, 149500, 150900, 152600,\n",
       "       154500, 155900, 156800, 157700, 158700, 159900, 161100, 162200,\n",
       "       163300, 164300, 165500, 166800, 168100, 169300, 170600, 172000,\n",
       "       173600, 175300, 177200, 179200, 181300, 183500, 185700, 187600,\n",
       "       189500, 191600, 193900, 196400, 199300, 202300, 205500, 208600,\n",
       "       211500, 214200, 216800, 219000, 220500, 221600, 222200, 222500,\n",
       "       222600, 222400, 222100, 221600, 221100, 220700, 220600, 220600,\n",
       "       220700, 221100, 221600, 222000, 222200, 221900, 221300, 220300,\n",
       "       219200, 218100, 217100, 216200, 215200, 213900, 212400, 210800,\n",
       "       209100, 207800, 206800, 205900, 205100, 204300, 203700, 203200,\n",
       "       202500, 201700, 201000, 200300, 199500, 198800, 197900, 197200,\n",
       "       197100, 197500, 198100, 198600, 198300, 197900, 197800, 198300,\n",
       "       199100, 199700, 199500, 198900, 197500, 195900, 193500, 190900,\n",
       "       189600, 190100, 190200, 189700, 188900, 188100, 187600, 187900,\n",
       "       188500, 188300, 187500, 187500, 187800, 187200, 186200, 185800,\n",
       "       185300, 184000, 182400, 181200, 180400, 180400, 181200, 182100,\n",
       "       182800, 183400, 183700, 183300, 183500, 184800, 187000, 188600,\n",
       "       189100, 188700, 188100, 187100, 185900, 185100, 184800, 184900,\n",
       "       184500, 184500, 185100, 185800, 186400, 186900, 187400, 187900,\n",
       "       188800, 189800, 190800, 191500, 192200, 192400, 192100, 191500,\n",
       "       191000, 191200, 192300, 193600, 194600, 195300, 195900, 195900,\n",
       "       196100, 197100, 198000, 199300, 201100, 202100, 202300, 202600,\n",
       "       202700, 202600, 203900, 206000, 208200, 210500, 212500, 213900,\n",
       "       215700, 218200, 220100, 221100, 221700, 221700, 221700, 222700,\n",
       "       223600])"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df = pd.DataFrame({'A': zip_df['1001'].values})\n",
    "df['diff1'] = df['A'].diff()\n",
    "df['diff2'] = df['diff1'].diff()\n",
    "\n",
    "def transform_diff2s(x_start, diff_start, pred_diff2s):\n",
    "    \"\"\"Transform second order differenced predictions to the values of the original formatted data\n",
    "    \n",
    "    Params:\n",
    "        x_start (numeric) the first value in the original series x\n",
    "        diff_start (numeric) the first non-null value in the 1st order difference, ie. x.diff().iloc[1]\n",
    "        pred_diff2s (series-like object) the predicted second order differenced values from some model\n",
    "        \n",
    "    Returns:\n",
    "        (series) undifferenced values in original format of x\"\"\"\n",
    "    import numpy as np\n",
    "    \n",
    "    x = x_start\n",
    "\n",
    "    x2 = diff_start\n",
    "    \n",
    "    x_diff2 = pred_diff2s\n",
    "    x_diff = np.r_[x2, x_diff2].cumsum().astype(int)\n",
    "    transformed_preds = np.r_[x, x_diff].cumsum().astype(int)\n",
    "    \n",
    "    return transformed_preds\n",
    "\n",
    "# x, x_diff = df['A'].iloc[0], df['B'].iloc[1:]\n",
    "# df['D'] = np.r_[x, x_diff].cumsum().astype(int)\n",
    "# x2, x_diff2 = df['B'].iloc[1], df['C'].iloc[2:]\n",
    "# x_diff_test = np.r_[x2, x_diff2].cumsum().astype(int)\n",
    "# x_diff_test == x_diff\n",
    "# df['E'] = np.r_[x, np.r_[x2, x_diff2].cumsum().astype(int)].astype(int)\n",
    "# print(x.size, x_diff.size, x_diff2.size, x2.size, len(df['A']))\n",
    "transform_diff2s(df['A'][0], df['diff1'][1], df['diff2'][2:])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": 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>t</th>\n",
       "      <th>t+1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>113100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>113100.0</td>\n",
       "      <td>112800.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>112800.0</td>\n",
       "      <td>112600.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>112600.0</td>\n",
       "      <td>112300.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>112300.0</td>\n",
       "      <td>112100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>260</th>\n",
       "      <td>221100.0</td>\n",
       "      <td>221700.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>261</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>221700.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>262</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>221700.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>263</th>\n",
       "      <td>221700.0</td>\n",
       "      <td>222700.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264</th>\n",
       "      <td>222700.0</td>\n",
       "      <td>223600.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>265 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            t       t+1\n",
       "0         NaN  113100.0\n",
       "1    113100.0  112800.0\n",
       "2    112800.0  112600.0\n",
       "3    112600.0  112300.0\n",
       "4    112300.0  112100.0\n",
       "..        ...       ...\n",
       "260  221100.0  221700.0\n",
       "261  221700.0  221700.0\n",
       "262  221700.0  221700.0\n",
       "263  221700.0  222700.0\n",
       "264  222700.0  223600.0\n",
       "\n",
       "[265 rows x 2 columns]"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's Try a Naive Persistence Model as a baseline:\n",
    "\n",
    "# Create lagged dataset\n",
    "values = pd.DataFrame(zip_df['1001'].values)\n",
    "df = pd.concat([values.shift(1), values], axis=1)\n",
    "df.columns = ['t', 't+1']\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "# split into train and test sets\n",
    "\n",
    "X = df.values\n",
    "train_size = int(len(X) * 0.75)\n",
    "train, test = X[1:train_size], X[train_size:]\n",
    "train_X, train_y = train[:,0], train[:,1]\n",
    "test_X, test_y = test[:,0], test[:,1]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE: 1047.100\n"
     ]
    },
    {
     "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>67.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>644.776119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>831.261105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-1200.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>600.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2500.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 0\n",
       "count    67.000000\n",
       "mean    644.776119\n",
       "std     831.261105\n",
       "min   -1200.000000\n",
       "25%     100.000000\n",
       "50%     600.000000\n",
       "75%    1000.000000\n",
       "max    2500.000000"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# persistence model\n",
    "\n",
    "predictions = [x for x in test_X]\n",
    "\n",
    "\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from math import sqrt\n",
    "# load data\n",
    "series = zip_df['1001']\n",
    "# create lagged dataset\n",
    "values = pd.DataFrame(series.values)\n",
    "dataframe = pd.concat([values.shift(1), values], axis=1)\n",
    "dataframe.columns = ['t', 't+1']\n",
    "# split into train and test sets\n",
    "X = dataframe.values\n",
    "train_size = int(len(X) * 0.75)\n",
    "train, test = X[1:train_size], X[train_size:]\n",
    "train_X, train_y = train[:,0], train[:,1]\n",
    "test_X, test_y = test[:,0], test[:,1]\n",
    "# persistence model\n",
    "predictions = [x for x in test_X]\n",
    "# skill of persistence model\n",
    "rmse = sqrt(mean_squared_error(test_y, predictions))\n",
    "print('Test RMSE: %.3f' % rmse)\n",
    "# calculate residuals\n",
    "residuals = [test_y[i]-predictions[i] for i in range(len(predictions))]\n",
    "residuals = pd.DataFrame(residuals)\n",
    "residuals.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "205 60\n"
     ]
    }
   ],
   "source": [
    "# Test Harness - We will be making predictions based off of a five year holding period, lets withhold 60 months as a validation set\n",
    "split_point = len(zip_df['1001']) - 60\n",
    "train, test = zip_df['1001'][:split_point], zip_df['1001'][split_point:]\n",
    "\n",
    "print(len(train), len(test))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 1.0,\n",
       " 2.104417123236605,\n",
       " 2.6457513110645907,\n",
       " 3.0,\n",
       " 3.255764119219941,\n",
       " 3.4509550720186013]"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Performance Measure-- We will be using RMSE initially\n",
    "\n",
    "def RMSE(y_true, y_pred):\n",
    "    \"\"\"Simple wrapper function on mean_squared_error to return RMSE\n",
    "    \n",
    "    Params:\n",
    "        predictions (series or array like object), the predicted values from model\n",
    "        test_data (series or array like object), the true target values\n",
    "        \n",
    "    Returns:\n",
    "        RMSE (list) list of accumulated RMSE values for each observation in consecutive time order\n",
    "        i.e. the first return value will be just the error of first prediction, second the sqrt of mean squared error for first 2 predictions, etc.\"\"\"\n",
    "    # Import necessary function\n",
    "    from sklearn.metrics import mean_squared_error\n",
    "    \n",
    "    # Ensure predictions and test_data are same size\n",
    "    if len(y_pred) != len(y_true):\n",
    "        return \"Test data and predictions must have equal length\"\n",
    "    \n",
    "    else:\n",
    "        rmse = [mean_squared_error(y_true[:i+1], y_pred[:i+1], squared=False) for i, _ in enumerate(y_pred)]\n",
    "        return rmse\n",
    "    \n",
    "test_a = [i for i in range(11)]\n",
    "test_preds = [i + 1 if i < 6 else i + 5 for i in test_a]\n",
    "\n",
    "RMSE(test_a, test_preds)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ADF Statistic: -6.627156\n",
      "p-value: 0.000000\n",
      "Critical Values:\n",
      "\t1%: -3.457\n",
      "\t5%: -2.873\n",
      "\t10%: -2.573\n"
     ]
    }
   ],
   "source": [
    "# Stationarity Test Wrapper:\n",
    "def test_stationarity(series):\n",
    "    \"\"\"Simple wrapper around adfuller that prints in more readable format\n",
    "    \n",
    "    Params:\n",
    "        series (series) a timeseries\n",
    "    Returns:\n",
    "        None\"\"\"\n",
    "    \n",
    "    from statsmodels.tsa.stattools import adfuller \n",
    "    \n",
    "    result = adfuller(series)\n",
    "    print('ADF Statistic: %f' % result[0])\n",
    "    print('p-value: %f' % result[1])\n",
    "    print('Critical Values:')\n",
    "    \n",
    "    for key, value in result[4].items():\n",
    "        print('\\t%s: %.3f' % (key, value))\n",
    "    return None\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ADF Statistic: -1.746793\n",
      "p-value: 0.407179\n",
      "Critical Values:\n",
      "\t1%: -3.457\n",
      "\t5%: -2.873\n",
      "\t10%: -2.573\n",
      "ADF Statistic: -1.945353\n",
      "p-value: 0.311036\n",
      "Critical Values:\n",
      "\t1%: -3.457\n",
      "\t5%: -2.873\n",
      "\t10%: -2.573\n",
      "ADF Statistic: -6.627156\n",
      "p-value: 0.000000\n",
      "Critical Values:\n",
      "\t1%: -3.457\n",
      "\t5%: -2.873\n",
      "\t10%: -2.573\n"
     ]
    }
   ],
   "source": [
    "# Example of stationarity test, this suggests that d param in ARIMA model should be at least 2 for the 1001 zipcode series\n",
    "# because the ADF statistic is not less than critical value at 5% significance level until second order differencing (with p value of 0)\n",
    "\n",
    "test_stationarity(zip_df['1001'])\n",
    "test_stationarity(zip_df['1001'].diff()[1:])\n",
    "test_stationarity(zip_df['1001'].diff()[1:].diff()[1:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This suggests that at least two levels of differencing is required. The d parameter in our\n",
    "# ARIMA model should at least be a value of 2. The next step is to select the lag values for\n",
    "# the Autoregression (AR) and Moving Average (MA) parameters, p and q respectively. We can\n",
    "# do this by reviewing Autocorrelation Function (ACF) and Partial Autocorrelation Function\n",
    "# (PACF) plots. The example below creates ACF and PACF plots for the series. (TEXT OF THIS APPROACH TAKEN FROM INTRO TO TIMESERIES FORECASTING w/ PYTHON\n",
    "# BROWNLEE)\n",
    "\n",
    "from statsmodels.graphics.tsaplots import plot_acf\n",
    "from statsmodels.graphics.tsaplots import plot_pacf\n",
    "\n",
    "series = zip_df['1001']\n",
    "plt.figure(figsize=(15,10))\n",
    "plt.subplot(211)\n",
    "plot_acf(series, lags=60, ax=plt.gca())\n",
    "plt.subplot(212)\n",
    "plot_pacf(series, lags=60, ax=plt.gca())\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Predicted=183768.513, Expected=184800.000\n",
      ">Predicted=186189.249, Expected=187000.000\n",
      ">Predicted=189212.764, Expected=188600.000\n",
      ">Predicted=190096.867, Expected=189100.000\n",
      ">Predicted=189479.643, Expected=188700.000\n",
      ">Predicted=188271.599, Expected=188100.000\n",
      ">Predicted=187561.946, Expected=187100.000\n",
      ">Predicted=186156.530, Expected=185900.000\n",
      ">Predicted=184750.228, Expected=185100.000\n",
      ">Predicted=184396.494, Expected=184800.000\n",
      ">Predicted=184550.653, Expected=184900.000\n",
      ">Predicted=185010.334, Expected=184500.000\n",
      ">Predicted=184028.019, Expected=184500.000\n",
      ">Predicted=184505.170, Expected=185100.000\n",
      ">Predicted=185735.701, Expected=185800.000\n",
      ">Predicted=186446.998, Expected=186400.000\n",
      ">Predicted=186968.813, Expected=186900.000\n",
      ">Predicted=187379.042, Expected=187400.000\n",
      ">Predicted=187881.308, Expected=187900.000\n",
      ">Predicted=188398.861, Expected=188800.000\n",
      ">Predicted=189734.123, Expected=189800.000\n",
      ">Predicted=190796.787, Expected=190800.000\n",
      ">Predicted=191774.959, Expected=191500.000\n",
      ">Predicted=192168.311, Expected=192200.000\n",
      ">Predicted=192896.828, Expected=192400.000\n",
      ">Predicted=192571.586, Expected=192100.000\n",
      ">Predicted=191778.988, Expected=191500.000\n",
      ">Predicted=190933.670, Expected=191000.000\n",
      ">Predicted=190552.429, Expected=191200.000\n",
      ">Predicted=191491.484, Expected=192300.000\n",
      ">Predicted=193469.864, Expected=193600.000\n",
      ">Predicted=194858.471, Expected=194600.000\n",
      ">Predicted=195506.383, Expected=195300.000\n",
      ">Predicted=195935.613, Expected=195900.000\n",
      ">Predicted=196476.799, Expected=195900.000\n",
      ">Predicted=195855.684, Expected=196100.000\n",
      ">Predicted=196358.269, Expected=197100.000\n",
      ">Predicted=198212.342, Expected=198000.000\n",
      ">Predicted=198862.686, Expected=199300.000\n",
      ">Predicted=200610.633, Expected=201100.000\n",
      ">Predicted=202917.760, Expected=202100.000\n",
      ">Predicted=202964.194, Expected=202300.000\n",
      ">Predicted=202435.471, Expected=202600.000\n",
      ">Predicted=202958.818, Expected=202700.000\n",
      ">Predicted=202808.920, Expected=202600.000\n",
      ">Predicted=202538.191, Expected=203900.000\n",
      ">Predicted=205352.713, Expected=206000.000\n",
      ">Predicted=208114.862, Expected=208200.000\n",
      ">Predicted=210316.608, Expected=210500.000\n",
      ">Predicted=212747.059, Expected=212500.000\n",
      ">Predicted=214370.712, Expected=213900.000\n",
      ">Predicted=215205.917, Expected=215700.000\n",
      ">Predicted=217559.202, Expected=218200.000\n",
      ">Predicted=220766.053, Expected=220100.000\n",
      ">Predicted=221941.897, Expected=221100.000\n",
      ">Predicted=222054.675, Expected=221700.000\n",
      ">Predicted=222302.418, Expected=221700.000\n",
      ">Predicted=221690.805, Expected=221700.000\n",
      ">Predicted=221801.032, Expected=222700.000\n",
      ">Predicted=223858.206, Expected=223600.000\n",
      "[1031.4866985405097, 927.7075211919339, 836.0148683973058, 878.9919076108116, 860.0406929017748, 788.2254670816855, 750.351372695476, 707.7250427900551, 677.3590908778957, 655.1454165106089, 633.4751764319269, 624.1420234719915, 613.778028443555, 612.4440656062894, 591.9100519069308, 573.234868526809, 556.3698898873722, 540.7169094214676, 526.312662547552, 520.7690319722064, 508.4218087352363, 496.732860903015, 489.1856690362478, 478.92955486107326, 479.6582981396266, 479.350348604886, 473.444034141473, 465.0817707650818, 472.5491117453241, 487.4927564475448, 480.1347591377336, 474.7768555800487, 468.9062666770184, 461.99949619774713, 465.67238183727994, 460.9611693030205, 470.75650882832963, 465.796480617483, 465.0879418474773, 465.71035111011054, 477.3957945270853, 482.6842112720771, 477.69796409849044, 473.84756484168565, 469.5869176012263, 505.99791778340904, 509.41225852705725, 504.22771825986723, 499.7432198455681, 495.9527991100473, 495.47023456359386, 495.4435991371659, 498.5785309563195, 502.1877358706926, 510.3864274330783, 508.0245512364417, 509.83120600976747, 505.41844206699636, 514.6022925083654, 511.38351393675936]\n"
     ]
    }
   ],
   "source": [
    "# ACF shows significant lag for about 20-22 months\n",
    "# PACF shows a significant lag for 2 months\n",
    "# A good starting point for p and q values is 20, 2 >> this does not converge\n",
    "\n",
    "\n",
    "from statsmodels.tsa.arima.model import ARIMA\n",
    "\n",
    "# load data\n",
    "series = zip_df['1001']\n",
    "# prepare data\n",
    "X = series.values\n",
    "X = X.astype('float32')\n",
    "\n",
    "# set trainset to include all but last 60 months (5 years)\n",
    "train_size = int(len(X) - 60)\n",
    "train, test = X[0:train_size], X[train_size:]\n",
    "# walk-forward validation\n",
    "history = [x for x in train]\n",
    "predictions = list()\n",
    "for i in range(len(test)):\n",
    "    # predict\n",
    "    model = ARIMA(history, order=(7,2,1))\n",
    "    model_fit = model.fit()\n",
    "    yhat = model_fit.forecast()[0]\n",
    "    predictions.append(yhat)\n",
    "    # observation\n",
    "    obs = test[i]\n",
    "    history.append(obs)\n",
    "    print('>Predicted=%.3f, Expected=%.3f' % (yhat, obs))\n",
    "# report performance\n",
    "rmse = RMSE(test, predictions)\n",
    "print(rmse)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}