los-prediction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "# pd.set_option('display.max_columns', None)\n",
    "import sys\n",
    "import numpy as np\n",
    "np.set_printoptions(threshold=sys.maxsize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": false
   },
   "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>hadm_id</th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>LOSdays</th>\n",
       "      <th>admit_type</th>\n",
       "      <th>admit_location</th>\n",
       "      <th>AdmitDiagnosis</th>\n",
       "      <th>insurance</th>\n",
       "      <th>religion</th>\n",
       "      <th>marital_status</th>\n",
       "      <th>...</th>\n",
       "      <th>NumMicroLabs</th>\n",
       "      <th>NumNotes</th>\n",
       "      <th>NumOutput</th>\n",
       "      <th>NumRx</th>\n",
       "      <th>NumProcEvents</th>\n",
       "      <th>NumTransfers</th>\n",
       "      <th>NumChartEvents</th>\n",
       "      <th>ExpiredHospital</th>\n",
       "      <th>TotalNumInteract</th>\n",
       "      <th>LOSgroupNum</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100001</td>\n",
       "      <td>F</td>\n",
       "      <td>35</td>\n",
       "      <td>6.17</td>\n",
       "      <td>EMERGENCY</td>\n",
       "      <td>CLINIC REFERRAL/PREMATURE</td>\n",
       "      <td>DIABETIC KETOACIDOSIS</td>\n",
       "      <td>Private</td>\n",
       "      <td>PROTESTANT QUAKER</td>\n",
       "      <td>DIVORCED</td>\n",
       "      <td>...</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.05</td>\n",
       "      <td>5.19</td>\n",
       "      <td>14.91</td>\n",
       "      <td>1.13</td>\n",
       "      <td>0.65</td>\n",
       "      <td>398.70</td>\n",
       "      <td>0</td>\n",
       "      <td>493.89</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100003</td>\n",
       "      <td>M</td>\n",
       "      <td>59</td>\n",
       "      <td>4.04</td>\n",
       "      <td>EMERGENCY</td>\n",
       "      <td>EMERGENCY ROOM ADMIT</td>\n",
       "      <td>UPPER GI BLEED</td>\n",
       "      <td>Private</td>\n",
       "      <td>NOT SPECIFIED</td>\n",
       "      <td>SINGLE</td>\n",
       "      <td>...</td>\n",
       "      <td>1.24</td>\n",
       "      <td>1.59</td>\n",
       "      <td>5.45</td>\n",
       "      <td>7.18</td>\n",
       "      <td>0.99</td>\n",
       "      <td>1.24</td>\n",
       "      <td>373.02</td>\n",
       "      <td>0</td>\n",
       "      <td>465.71</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100006</td>\n",
       "      <td>F</td>\n",
       "      <td>48</td>\n",
       "      <td>12.04</td>\n",
       "      <td>EMERGENCY</td>\n",
       "      <td>EMERGENCY ROOM ADMIT</td>\n",
       "      <td>COPD FLARE</td>\n",
       "      <td>Private</td>\n",
       "      <td>NOT SPECIFIED</td>\n",
       "      <td>SINGLE</td>\n",
       "      <td>...</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.15</td>\n",
       "      <td>4.15</td>\n",
       "      <td>6.23</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.33</td>\n",
       "      <td>286.21</td>\n",
       "      <td>0</td>\n",
       "      <td>344.00</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100007</td>\n",
       "      <td>F</td>\n",
       "      <td>73</td>\n",
       "      <td>7.29</td>\n",
       "      <td>EMERGENCY</td>\n",
       "      <td>EMERGENCY ROOM ADMIT</td>\n",
       "      <td>BOWEL OBSTRUCTION</td>\n",
       "      <td>Private</td>\n",
       "      <td>JEWISH</td>\n",
       "      <td>MARRIED</td>\n",
       "      <td>...</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.17</td>\n",
       "      <td>9.05</td>\n",
       "      <td>11.52</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.96</td>\n",
       "      <td>526.06</td>\n",
       "      <td>0</td>\n",
       "      <td>603.05</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100009</td>\n",
       "      <td>M</td>\n",
       "      <td>60</td>\n",
       "      <td>4.88</td>\n",
       "      <td>EMERGENCY</td>\n",
       "      <td>TRANSFER FROM HOSP/EXTRAM</td>\n",
       "      <td>CORONARY ARTERY DISEASE</td>\n",
       "      <td>Private</td>\n",
       "      <td>CATHOLIC</td>\n",
       "      <td>MARRIED</td>\n",
       "      <td>...</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.34</td>\n",
       "      <td>16.19</td>\n",
       "      <td>25.00</td>\n",
       "      <td>2.87</td>\n",
       "      <td>2.05</td>\n",
       "      <td>554.92</td>\n",
       "      <td>0</td>\n",
       "      <td>679.84</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   hadm_id gender  age  LOSdays admit_type             admit_location  \\\n",
       "0   100001      F   35     6.17  EMERGENCY  CLINIC REFERRAL/PREMATURE   \n",
       "1   100003      M   59     4.04  EMERGENCY       EMERGENCY ROOM ADMIT   \n",
       "2   100006      F   48    12.04  EMERGENCY       EMERGENCY ROOM ADMIT   \n",
       "3   100007      F   73     7.29  EMERGENCY       EMERGENCY ROOM ADMIT   \n",
       "4   100009      M   60     4.88  EMERGENCY  TRANSFER FROM HOSP/EXTRAM   \n",
       "\n",
       "            AdmitDiagnosis insurance           religion marital_status  ...  \\\n",
       "0    DIABETIC KETOACIDOSIS   Private  PROTESTANT QUAKER       DIVORCED  ...   \n",
       "1           UPPER GI BLEED   Private      NOT SPECIFIED         SINGLE  ...   \n",
       "2               COPD FLARE   Private      NOT SPECIFIED         SINGLE  ...   \n",
       "3        BOWEL OBSTRUCTION   Private             JEWISH        MARRIED  ...   \n",
       "4  CORONARY ARTERY DISEASE   Private           CATHOLIC        MARRIED  ...   \n",
       "\n",
       "  NumMicroLabs  NumNotes  NumOutput  NumRx NumProcEvents  NumTransfers  \\\n",
       "0         0.65      0.05       5.19  14.91          1.13          0.65   \n",
       "1         1.24      1.59       5.45   7.18          0.99          1.24   \n",
       "2         0.33      0.15       4.15   6.23          0.00          0.33   \n",
       "3         0.69      0.17       9.05  11.52          0.00          0.96   \n",
       "4         0.61      0.34      16.19  25.00          2.87          2.05   \n",
       "\n",
       "   NumChartEvents  ExpiredHospital  TotalNumInteract  LOSgroupNum  \n",
       "0          398.70                0            493.89            2  \n",
       "1          373.02                0            465.71            2  \n",
       "2          286.21                0            344.00            4  \n",
       "3          526.06                0            603.05            2  \n",
       "4          554.92                0            679.84            2  \n",
       "\n",
       "[5 rows x 28 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw = pd.read_csv('/Users/dereklee/biostat-masters-thesis/mimic3c.csv')\n",
    "raw.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "hadm_id                 0\n",
       "gender                  0\n",
       "age                     0\n",
       "LOSdays                 0\n",
       "admit_type              0\n",
       "admit_location          0\n",
       "AdmitDiagnosis         25\n",
       "insurance               0\n",
       "religion              458\n",
       "marital_status      10128\n",
       "ethnicity               0\n",
       "NumCallouts             0\n",
       "NumDiagnosis            0\n",
       "NumProcs                0\n",
       "AdmitProcedure          0\n",
       "NumCPTevents            0\n",
       "NumInput                0\n",
       "NumLabs                 0\n",
       "NumMicroLabs            0\n",
       "NumNotes                0\n",
       "NumOutput               0\n",
       "NumRx                   0\n",
       "NumProcEvents           0\n",
       "NumTransfers            0\n",
       "NumChartEvents          0\n",
       "ExpiredHospital         0\n",
       "TotalNumInteract        0\n",
       "LOSgroupNum             0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# raw.info()\n",
    "raw.isnull().sum(axis=0)\n",
    "\n",
    "# some na's in AdmitDiagnosis, religion, marital_status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# raw.AdmitDiagnosis.unique() # need to organize these diagnoses again...\n",
    "# raw.AdmitProcedure.unique() # free text\n",
    "# raw.ExpiredHospital.unique() # 0,1\n",
    "# raw.LOSdays.unique() # continuous\n",
    "# (raw.LOSdays < 0).sum() # no negatives\n",
    "# raw[raw.LOSdays==\"na\"] # no NAs\n",
    "\n",
    "# raw.LOSgroupNum.unique() # array([2, 4, 3, 1])\n",
    "# raw.NumCPTevents.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# raw.NumCallouts.unique() # continuous (but why are there decimals?)\n",
    "# (raw.NumChartEvents < 0).sum() # no negatives, continuous (but why are there decimals?)\n",
    "# raw.NumDiagnosis.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# raw.NumInput.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# (raw.NumInput < 0).sum()\n",
    "# raw.NumLabs.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# (raw.NumLabs < 0).sum() \n",
    "# raw.NumMicroLabs.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# (raw.NumMicroLabs < 0).sum()\n",
    "# raw.NumNotes.unique() # no negatives, continuous (but why are there decimals?)\n",
    "\n",
    "\n",
    "# raw.NumOutput.unique() # no negatives, continuous (but why are there decimals?)\n",
    "\n",
    "\n",
    "# raw.NumProcEvents.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# (raw.NumProcEvents < 0).sum() \n",
    "\n",
    "# raw.NumProcs.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# (raw.NumProcs < 0).sum() # no negatives, continuous (but why are there decimals?)\n",
    "\n",
    "# raw.NumRx.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# (raw.NumRx < 0).sum() \n",
    "\n",
    "# raw.NumTransfers.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# (raw.NumTransfers < 0).sum() \n",
    "\n",
    "# raw.TotalNumInteract.unique() # no negatives, continuous (but why are there decimals?)\n",
    "# (raw.TotalNumInteract < 0).sum() \n",
    "\n",
    "# raw.admit_location.unique() # clean up text categories\n",
    "# raw.admit_type.unique() # cleaned!\n",
    "# raw.age.unique() # continuous\n",
    "# raw.ethnicity.unique() # clean up text categories\n",
    "\n",
    "# raw.gender.unique() # ['F', 'M']\n",
    " \n",
    "\n",
    "# raw.insurance.unique() # cleaned unless want to collapse further\n",
    "# raw.marital_status.unique() # missing values\n",
    "# raw.religion.unique() # Check distributions to see if need to collapse religions\n",
    "\n",
    "\n",
    "# max(raw.hadm_id.value_counts()) # only one row per patient"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Missing Values:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['DIVORCED', 'SINGLE', 'MARRIED', nan, 'WIDOWED',\n",
       "       'UNKNOWN (DEFAULT)', 'SEPARATED', 'LIFE PARTNER'], dtype=object)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw.AdmitDiagnosis.isna().sum() # 25 missing\n",
    "raw.religion.isna().sum() # 458 missing\n",
    "raw.marital_status.isna().sum() # 10128 missing\n",
    "\n",
    "raw.religion.unique()\n",
    "\n",
    "raw.marital_status.unique()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(19663, 28)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_er = raw[raw['admit_location']==\"EMERGENCY ROOM ADMIT\"]\n",
    "\n",
    "raw_er_alive = raw_er[raw_er['ExpiredHospital'] == 0]\n",
    "raw_er_alive.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1085, 28)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_er_missing_alive = raw_er_alive[raw_er_alive.isnull().any(axis=1)]\n",
    "raw_er_missing_alive.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "18578"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(raw_er_alive) - len(raw_er_missing_alive)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.05517977928088288"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(raw_er_missing_alive)/len(raw_er_alive)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Normalization:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  after removing the cwd from sys.path.\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:5: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \"\"\"\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:6: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:7: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  import sys\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:8: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:9: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  if __name__ == '__main__':\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:10: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  # Remove the CWD from sys.path while we load stuff.\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:11: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  # This is added back by InteractiveShellApp.init_path()\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:12: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  if sys.path[0] == '':\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:13: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  del sys.path[0]\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:14: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:17: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:18: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:19: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:20: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:21: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:22: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:23: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:24: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:25: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:26: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:27: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:28: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:29: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:30: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n"
     ]
    }
   ],
   "source": [
    "raw_er_alive['NumCalloutsByDay'] = raw_er_alive['NumCallouts']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumDiagnosisByDay'] = raw_er_alive['NumDiagnosis']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumProcsByDay'] = raw_er_alive['NumProcs']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumCPTeventsByDay'] = raw_er_alive['NumCPTevents']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumInputByDay'] = raw_er_alive['NumInput']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumLabsByDay'] = raw_er_alive['NumLabs']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumMicroLabsByDay'] = raw_er_alive['NumMicroLabs']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumNotesByDay'] = raw_er_alive['NumNotes']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumOutputByDay'] = raw_er_alive['NumOutput']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumRxByDay'] = raw_er_alive['NumRx']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumProcEventsByDay'] = raw_er_alive['NumProcEvents']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumTransfersByDay'] = raw_er_alive['NumTransfers']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['NumChartEventsByDay'] = raw_er_alive['NumChartEvents']/raw_er_alive['LOSdays']\n",
    "raw_er_alive['TotalNumInteractByDay'] = raw_er_alive['TotalNumInteract']/raw_er_alive['LOSdays']\n",
    "\n",
    "\n",
    "raw_er_missing_alive['NumCalloutsByDay'] = raw_er_missing_alive['NumCallouts']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumDiagnosisByDay'] = raw_er_missing_alive['NumDiagnosis']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumProcsByDay'] = raw_er_missing_alive['NumProcs']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumCPTeventsByDay'] = raw_er_missing_alive['NumCPTevents']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumInputByDay'] = raw_er_missing_alive['NumInput']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumLabsByDay'] = raw_er_missing_alive['NumLabs']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumMicroLabsByDay'] = raw_er_missing_alive['NumMicroLabs']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumNotesByDay'] = raw_er_missing_alive['NumNotes']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumOutputByDay'] = raw_er_missing_alive['NumOutput']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumRxByDay'] = raw_er_missing_alive['NumRx']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumProcEventsByDay'] = raw_er_missing_alive['NumProcEvents']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['NumTransfersByDay'] = raw_er_missing_alive['NumTransfers']/raw_er_alive['LOSdays']\n",
    "raw_er_missing_alive['NumChartEventsByDay'] = raw_er_missing_alive['NumChartEvents']/raw_er_missing_alive['LOSdays']\n",
    "raw_er_missing_alive['TotalNumInteractByDay'] = raw_er_missing_alive['TotalNumInteract']/raw_er_missing_alive['LOSdays']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['hadm_id', 'gender', 'age', 'LOSdays', 'admit_type', 'admit_location',\n",
       "       'AdmitDiagnosis', 'insurance', 'religion', 'marital_status',\n",
       "       'ethnicity', 'NumCallouts', 'NumDiagnosis', 'NumProcs',\n",
       "       'AdmitProcedure', 'NumCPTevents', 'NumInput', 'NumLabs', 'NumMicroLabs',\n",
       "       'NumNotes', 'NumOutput', 'NumRx', 'NumProcEvents', 'NumTransfers',\n",
       "       'NumChartEvents', 'ExpiredHospital', 'TotalNumInteract', 'LOSgroupNum',\n",
       "       'NumCalloutsByDay', 'NumDiagnosisByDay', 'NumProcsByDay',\n",
       "       'NumCPTeventsByDay', 'NumInputByDay', 'NumLabsByDay',\n",
       "       'NumMicroLabsByDay', 'NumNotesByDay', 'NumOutputByDay', 'NumRxByDay',\n",
       "       'NumProcEventsByDay', 'NumTransfersByDay', 'NumChartEventsByDay',\n",
       "       'TotalNumInteractByDay'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_er_alive.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Drop NA's"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "raw_completecase = raw_er_alive.dropna()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Consolidation (Missing Set):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/pandas/core/indexing.py:494: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self.obj[item] = s\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py:5208: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self[name] = value\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:66: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:67: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n"
     ]
    }
   ],
   "source": [
    "# missing set consolidation:\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['insurance'].isin(['Private', 'Medicaid', 'Medicare', 'Government']), 'insurance'] = \"INSURED\"\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['insurance'].isin(['Self Pay']), 'insurance'] = \"UNINSURED\"\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['marital_status'].isin(['LIFE PARTNER']), 'marital_status'] = \"MARRIED\"\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['marital_status'].isin(['DIVORCED', 'SINGLE','WIDOWED','SEPARATED']), 'marital_status'] = \"UNMARRIED\"\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['marital_status'].isin(['UNKNOWN (DEFAULT)']), 'marital_status'] = \"UNKNOWN\"\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive[\"religion\"].isin(['PROTESTANT QUAKER', 'JEWISH', 'CATHOLIC',\n",
    "       'EPISCOPALIAN', 'OTHER', \"JEHOVAH'S WITNESS\",\n",
    "       'UNITARIAN-UNIVERSALIST', 'LUTHERAN', 'CHRISTIAN SCIENTIST',\n",
    "       'GREEK ORTHODOX', 'BUDDHIST', '7TH DAY ADVENTIST', 'MUSLIM',\n",
    "       'HINDU', 'ROMANIAN EAST. ORTH', 'HEBREW', 'BAPTIST', 'METHODIST']), \"religion\"] = \"RELIGIOUS\"\n",
    "\n",
    "\n",
    "raw_er_missing_alive.religion = raw_er_missing_alive['religion'].replace(['NOT SPECIFIED', 'UNOBTAINABLE'], \"UNKNOWN\")\n",
    "\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['ethnicity'].isin(['ASIAN',\n",
    "                                                        'ASIAN - CHINESE', \n",
    "                                                         'ASIAN - ASIAN INDIAN',\n",
    "                                                        'ASIAN - OTHER',\n",
    "                                                        'ASIAN - FILIPINO',\n",
    "                                                        'ASIAN - CAMBODIAN', \n",
    "                                                        'ASIAN - VIETNAMESE',\n",
    "                                                        'ASIAN - KOREAN',\n",
    "                                                         'ASIAN - THAI', \n",
    "                                                         'ASIAN - JAPANESE',\n",
    "                                                        'NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER']), 'ethnicity'] = \"ASIAN_PACIFICISLANDER\"\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['ethnicity'].isin(['HISPANIC/LATINO - SALVADORAN',\n",
    "                                                        'HISPANIC/LATINO - PUERTO RICAN', \n",
    "                                                         'HISPANIC/LATINO - GUATEMALAN',\n",
    "                                                        'HISPANIC/LATINO - DOMINICAN',\n",
    "                                                        'HISPANIC/LATINO - CUBAN',\n",
    "                                                        'HISPANIC/LATINO - COLOMBIAN',\n",
    "                                                        'HISPANIC/LATINO - HONDURAN', \n",
    "                                                         'HISPANIC/LATINO - MEXICAN',\n",
    "                                                        'HISPANIC/LATINO - CENTRAL AMERICAN (OTHER)', \n",
    "                                                         'SOUTH AMERICAN']), 'ethnicity'] = \"HISPANIC OR LATINO\"\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['ethnicity'].isin(['WHITE - RUSSIAN',\n",
    "                                                        'WHITE - OTHER EUROPEAN',\n",
    "                                                        'WHITE - BRAZILIAN',\n",
    "                                                        'PORTUGUESE',\n",
    "                                                        'WHITE - EASTERN EUROPEAN']), 'ethnicity'] = \"WHITE\"\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['ethnicity'].isin(['BLACK/AFRICAN AMERICAN',\n",
    "                                                         'BLACK/CAPE VERDEAN', \n",
    "                                                         'BLACK/HAITIAN',\n",
    "                                                        'BLACK/AFRICAN']), 'ethnicity'] = \"BLACK\"\n",
    "\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['ethnicity'].isin(['MULTI RACE ETHNICITY',\n",
    "                                                        'AMERICAN INDIAN/ALASKA NATIVE',\n",
    "                                                        'AMERICAN INDIAN/ALASKA NATIVE FEDERALLY RECOGNIZED TRIBE',\n",
    "                                                        'MIDDLE EASTERN',\n",
    "                                                        'CARIBBEAN ISLAND']), 'ethnicity'] = \"OTHER\"\n",
    "\n",
    "raw_er_missing_alive.loc[raw_er_missing_alive['ethnicity'].isin(['UNKNOWN/NOT SPECIFIED',\n",
    "                                                        'UNABLE TO OBTAIN',\n",
    "                                                        'PATIENT DECLINED TO ANSWER']), 'ethnicity'] = \"UNKNOWN\"\n",
    "\n",
    "\n",
    "raw_er_missing_alive['gender'] = raw_er_missing_alive['gender'].replace('F', 0)\n",
    "raw_er_missing_alive['gender'] = raw_er_missing_alive['gender'].replace('M', 1)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary Stats of Missing Set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": false
   },
   "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>age</th>\n",
       "      <th>LOSdays</th>\n",
       "      <th>NumCalloutsByDay</th>\n",
       "      <th>NumDiagnosisByDay</th>\n",
       "      <th>NumProcsByDay</th>\n",
       "      <th>NumCPTeventsByDay</th>\n",
       "      <th>NumInputByDay</th>\n",
       "      <th>NumLabsByDay</th>\n",
       "      <th>NumMicroLabsByDay</th>\n",
       "      <th>NumNotesByDay</th>\n",
       "      <th>NumOutputByDay</th>\n",
       "      <th>NumRxByDay</th>\n",
       "      <th>NumProcEventsByDay</th>\n",
       "      <th>NumTransfersByDay</th>\n",
       "      <th>NumChartEventsByDay</th>\n",
       "      <th>TotalNumInteractByDay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1085.000000</td>\n",
       "      <td>1085.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "      <td>1083.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>55.070046</td>\n",
       "      <td>9.922203</td>\n",
       "      <td>0.045125</td>\n",
       "      <td>1.853373</td>\n",
       "      <td>0.690670</td>\n",
       "      <td>0.334569</td>\n",
       "      <td>11.804612</td>\n",
       "      <td>32.951758</td>\n",
       "      <td>0.363741</td>\n",
       "      <td>7.130839</td>\n",
       "      <td>3.343019</td>\n",
       "      <td>4.099147</td>\n",
       "      <td>0.305980</td>\n",
       "      <td>0.850645</td>\n",
       "      <td>259.819960</td>\n",
       "      <td>323.593436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>20.989118</td>\n",
       "      <td>11.169253</td>\n",
       "      <td>0.258203</td>\n",
       "      <td>8.495302</td>\n",
       "      <td>3.759398</td>\n",
       "      <td>1.205855</td>\n",
       "      <td>42.295168</td>\n",
       "      <td>134.447139</td>\n",
       "      <td>1.401210</td>\n",
       "      <td>56.175758</td>\n",
       "      <td>8.673836</td>\n",
       "      <td>13.694880</td>\n",
       "      <td>2.387928</td>\n",
       "      <td>3.405554</td>\n",
       "      <td>778.623109</td>\n",
       "      <td>1008.492367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000571</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.215814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>39.000000</td>\n",
       "      <td>3.290000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.058752</td>\n",
       "      <td>0.011581</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.343851</td>\n",
       "      <td>2.910684</td>\n",
       "      <td>0.011352</td>\n",
       "      <td>0.016651</td>\n",
       "      <td>0.476390</td>\n",
       "      <td>0.142932</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.028193</td>\n",
       "      <td>33.134198</td>\n",
       "      <td>40.936365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>54.000000</td>\n",
       "      <td>6.460000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.193833</td>\n",
       "      <td>0.037391</td>\n",
       "      <td>0.058013</td>\n",
       "      <td>3.238773</td>\n",
       "      <td>5.914237</td>\n",
       "      <td>0.067010</td>\n",
       "      <td>0.065180</td>\n",
       "      <td>1.003333</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.094139</td>\n",
       "      <td>70.514991</td>\n",
       "      <td>85.072165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>74.000000</td>\n",
       "      <td>12.500000</td>\n",
       "      <td>0.013945</td>\n",
       "      <td>0.665088</td>\n",
       "      <td>0.135881</td>\n",
       "      <td>0.218596</td>\n",
       "      <td>7.442253</td>\n",
       "      <td>13.700893</td>\n",
       "      <td>0.181938</td>\n",
       "      <td>0.337784</td>\n",
       "      <td>2.251264</td>\n",
       "      <td>2.390581</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.322829</td>\n",
       "      <td>148.565213</td>\n",
       "      <td>180.942132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>88.000000</td>\n",
       "      <td>104.170000</td>\n",
       "      <td>4.717391</td>\n",
       "      <td>181.428571</td>\n",
       "      <td>68.047619</td>\n",
       "      <td>22.666667</td>\n",
       "      <td>861.666667</td>\n",
       "      <td>2925.190476</td>\n",
       "      <td>22.666667</td>\n",
       "      <td>1079.809524</td>\n",
       "      <td>90.714286</td>\n",
       "      <td>204.095238</td>\n",
       "      <td>51.023810</td>\n",
       "      <td>45.333333</td>\n",
       "      <td>14807.238095</td>\n",
       "      <td>17973.190476</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               age      LOSdays  NumCalloutsByDay  NumDiagnosisByDay  \\\n",
       "count  1085.000000  1085.000000       1083.000000        1083.000000   \n",
       "mean     55.070046     9.922203          0.045125           1.853373   \n",
       "std      20.989118    11.169253          0.258203           8.495302   \n",
       "min      15.000000     0.000000          0.000000           0.000000   \n",
       "25%      39.000000     3.290000          0.000000           0.058752   \n",
       "50%      54.000000     6.460000          0.000000           0.193833   \n",
       "75%      74.000000    12.500000          0.013945           0.665088   \n",
       "max      88.000000   104.170000          4.717391         181.428571   \n",
       "\n",
       "       NumProcsByDay  NumCPTeventsByDay  NumInputByDay  NumLabsByDay  \\\n",
       "count    1083.000000        1083.000000    1083.000000   1083.000000   \n",
       "mean        0.690670           0.334569      11.804612     32.951758   \n",
       "std         3.759398           1.205855      42.295168    134.447139   \n",
       "min         0.000000           0.000000       0.000000      0.000000   \n",
       "25%         0.011581           0.000000       1.343851      2.910684   \n",
       "50%         0.037391           0.058013       3.238773      5.914237   \n",
       "75%         0.135881           0.218596       7.442253     13.700893   \n",
       "max        68.047619          22.666667     861.666667   2925.190476   \n",
       "\n",
       "       NumMicroLabsByDay  NumNotesByDay  NumOutputByDay   NumRxByDay  \\\n",
       "count        1083.000000    1083.000000     1083.000000  1083.000000   \n",
       "mean            0.363741       7.130839        3.343019     4.099147   \n",
       "std             1.401210      56.175758        8.673836    13.694880   \n",
       "min             0.000000       0.000000        0.000000     0.000000   \n",
       "25%             0.011352       0.016651        0.476390     0.142932   \n",
       "50%             0.067010       0.065180        1.003333     0.812500   \n",
       "75%             0.181938       0.337784        2.251264     2.390581   \n",
       "max            22.666667    1079.809524       90.714286   204.095238   \n",
       "\n",
       "       NumProcEventsByDay  NumTransfersByDay  NumChartEventsByDay  \\\n",
       "count         1083.000000        1083.000000          1083.000000   \n",
       "mean             0.305980           0.850645           259.819960   \n",
       "std              2.387928           3.405554           778.623109   \n",
       "min              0.000000           0.000571             0.000000   \n",
       "25%              0.000000           0.028193            33.134198   \n",
       "50%              0.000000           0.094139            70.514991   \n",
       "75%              0.000000           0.322829           148.565213   \n",
       "max             51.023810          45.333333         14807.238095   \n",
       "\n",
       "       TotalNumInteractByDay  \n",
       "count            1083.000000  \n",
       "mean              323.593436  \n",
       "std              1008.492367  \n",
       "min                 0.215814  \n",
       "25%                40.936365  \n",
       "50%                85.072165  \n",
       "75%               180.942132  \n",
       "max             17973.190476  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_er_missing_alive[['age', 'LOSdays','NumCalloutsByDay', 'NumDiagnosisByDay', 'NumProcsByDay',\n",
    "       'NumCPTeventsByDay', 'NumInputByDay', 'NumLabsByDay',\n",
    "       'NumMicroLabsByDay', 'NumNotesByDay', 'NumOutputByDay', 'NumRxByDay',\n",
    "       'NumProcEventsByDay', 'NumTransfersByDay', 'NumChartEventsByDay',\n",
    "       'TotalNumInteractByDay']].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "hadm_id                    0\n",
       "gender                     0\n",
       "age                        0\n",
       "LOSdays                    0\n",
       "admit_type                 0\n",
       "admit_location             0\n",
       "AdmitDiagnosis             5\n",
       "insurance                  0\n",
       "religion                 188\n",
       "marital_status           989\n",
       "ethnicity                  0\n",
       "NumCallouts                0\n",
       "NumDiagnosis               0\n",
       "NumProcs                   0\n",
       "AdmitProcedure             0\n",
       "NumCPTevents               0\n",
       "NumInput                   0\n",
       "NumLabs                    0\n",
       "NumMicroLabs               0\n",
       "NumNotes                   0\n",
       "NumOutput                  0\n",
       "NumRx                      0\n",
       "NumProcEvents              0\n",
       "NumTransfers               0\n",
       "NumChartEvents             0\n",
       "ExpiredHospital            0\n",
       "TotalNumInteract           0\n",
       "LOSgroupNum                0\n",
       "NumCalloutsByDay           2\n",
       "NumDiagnosisByDay          2\n",
       "NumProcsByDay              2\n",
       "NumCPTeventsByDay          2\n",
       "NumInputByDay              2\n",
       "NumLabsByDay               2\n",
       "NumMicroLabsByDay          2\n",
       "NumNotesByDay              2\n",
       "NumOutputByDay             2\n",
       "NumRxByDay                 2\n",
       "NumProcEventsByDay         2\n",
       "NumTransfersByDay          2\n",
       "NumChartEventsByDay        2\n",
       "TotalNumInteractByDay      2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_er_missing_alive.isnull().sum(axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The 2 missings in the continuous predictors all belong to the same 2 patients"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hadm_id</th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>LOSdays</th>\n",
       "      <th>admit_type</th>\n",
       "      <th>admit_location</th>\n",
       "      <th>AdmitDiagnosis</th>\n",
       "      <th>insurance</th>\n",
       "      <th>religion</th>\n",
       "      <th>marital_status</th>\n",
       "      <th>...</th>\n",
       "      <th>NumInputByDay</th>\n",
       "      <th>NumLabsByDay</th>\n",
       "      <th>NumMicroLabsByDay</th>\n",
       "      <th>NumNotesByDay</th>\n",
       "      <th>NumOutputByDay</th>\n",
       "      <th>NumRxByDay</th>\n",
       "      <th>NumProcEventsByDay</th>\n",
       "      <th>NumTransfersByDay</th>\n",
       "      <th>NumChartEventsByDay</th>\n",
       "      <th>TotalNumInteractByDay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10792</th>\n",
       "      <td>118222</td>\n",
       "      <td>0</td>\n",
       "      <td>45</td>\n",
       "      <td>0.0</td>\n",
       "      <td>EMERGENCY</td>\n",
       "      <td>EMERGENCY ROOM ADMIT</td>\n",
       "      <td>HEAD INJURY</td>\n",
       "      <td>INSURED</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56422</th>\n",
       "      <td>195619</td>\n",
       "      <td>1</td>\n",
       "      <td>44</td>\n",
       "      <td>0.0</td>\n",
       "      <td>EMERGENCY</td>\n",
       "      <td>EMERGENCY ROOM ADMIT</td>\n",
       "      <td>ETOH INTOXICATION</td>\n",
       "      <td>UNINSURED</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 42 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       hadm_id  gender  age  LOSdays admit_type        admit_location  \\\n",
       "10792   118222       0   45      0.0  EMERGENCY  EMERGENCY ROOM ADMIT   \n",
       "56422   195619       1   44      0.0  EMERGENCY  EMERGENCY ROOM ADMIT   \n",
       "\n",
       "          AdmitDiagnosis  insurance religion marital_status  ...  \\\n",
       "10792        HEAD INJURY    INSURED  UNKNOWN            NaN  ...   \n",
       "56422  ETOH INTOXICATION  UNINSURED  UNKNOWN            NaN  ...   \n",
       "\n",
       "      NumInputByDay  NumLabsByDay  NumMicroLabsByDay  NumNotesByDay  \\\n",
       "10792           NaN           NaN                NaN            NaN   \n",
       "56422           NaN           NaN                NaN            NaN   \n",
       "\n",
       "      NumOutputByDay  NumRxByDay  NumProcEventsByDay  NumTransfersByDay  \\\n",
       "10792            NaN         NaN                 NaN                NaN   \n",
       "56422            NaN         NaN                 NaN                NaN   \n",
       "\n",
       "       NumChartEventsByDay  TotalNumInteractByDay  \n",
       "10792                  NaN                    NaN  \n",
       "56422                  NaN                    NaN  \n",
       "\n",
       "[2 rows x 42 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_er_missing_alive[raw_er_missing_alive['NumDiagnosisByDay'].isnull()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Outlier removal (Missing Set):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "def log_transform(df,col):\n",
    "    df['log_' + col] = np.log(df[col])\n",
    "\n",
    "log_transform(raw_er_missing_alive, 'age')\n",
    "log_transform(raw_er_missing_alive, 'LOSdays')\n",
    "log_transform(raw_er_missing_alive, 'NumCalloutsByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumDiagnosisByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumProcsByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumCPTeventsByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumInputByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumLabsByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumMicroLabsByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumNotesByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumOutputByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumRxByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumProcEventsByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumTransfersByDay')\n",
    "log_transform(raw_er_missing_alive, 'NumChartEventsByDay')\n",
    "log_transform(raw_er_missing_alive, 'TotalNumInteractByDay')\n",
    "\n",
    "# from pandas.api.types import is_numeric_dtype\n",
    "\n",
    "# def remove_outlier(df):\n",
    "#     low = .05\n",
    "#     high = .95\n",
    "#     quant_df = df.quantile([low, high])\n",
    "#     for name in list(df.columns):\n",
    "#         if is_numeric_dtype(df[name]):\n",
    "#             print(name,\": is numeric\")\n",
    "#             df = df[(df[name] > quant_df.loc[low, name]) \n",
    "#                     & (df[name] < quant_df.loc[high, name])]\n",
    "#     return df\n",
    "\n",
    "\n",
    "\n",
    "# remove_outlier(raw_er_missing_alive).head()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def filter_outlier(df):\n",
    "    q_low = df.quantile(0.003)\n",
    "    q_hi = df.quantile(0.997)\n",
    "    \n",
    "    df_filtered = df[~((df < q_low) | (df > q_hi)).any(axis=1)]\n",
    "#     raw_completecase[~((raw_completecase < (q1 - 1.5 * IQR)) |(raw_completecase > (q3 + 1.5 * IQR))).any(axis=1)]\n",
    "    return df_filtered\n",
    "\n",
    "\n",
    "nooutlier_missing = filter_outlier(raw_er_missing_alive)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1022"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(nooutlier_missing)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "hadm_id                        0\n",
       "gender                         0\n",
       "age                            0\n",
       "LOSdays                        0\n",
       "admit_type                     0\n",
       "admit_location                 0\n",
       "AdmitDiagnosis                 1\n",
       "insurance                      0\n",
       "religion                     183\n",
       "marital_status               932\n",
       "ethnicity                      0\n",
       "NumCallouts                    0\n",
       "NumDiagnosis                   0\n",
       "NumProcs                       0\n",
       "AdmitProcedure                 0\n",
       "NumCPTevents                   0\n",
       "NumInput                       0\n",
       "NumLabs                        0\n",
       "NumMicroLabs                   0\n",
       "NumNotes                       0\n",
       "NumOutput                      0\n",
       "NumRx                          0\n",
       "NumProcEvents                  0\n",
       "NumTransfers                   0\n",
       "NumChartEvents                 0\n",
       "ExpiredHospital                0\n",
       "TotalNumInteract               0\n",
       "LOSgroupNum                    0\n",
       "NumCalloutsByDay               0\n",
       "NumDiagnosisByDay              0\n",
       "NumProcsByDay                  0\n",
       "NumCPTeventsByDay              0\n",
       "NumInputByDay                  0\n",
       "NumLabsByDay                   0\n",
       "NumMicroLabsByDay              0\n",
       "NumNotesByDay                  0\n",
       "NumOutputByDay                 0\n",
       "NumRxByDay                     0\n",
       "NumProcEventsByDay             0\n",
       "NumTransfersByDay              0\n",
       "NumChartEventsByDay            0\n",
       "TotalNumInteractByDay          0\n",
       "log_age                        0\n",
       "log_LOSdays                    0\n",
       "log_NumCalloutsByDay           0\n",
       "log_NumDiagnosisByDay          0\n",
       "log_NumProcsByDay              0\n",
       "log_NumCPTeventsByDay          0\n",
       "log_NumInputByDay              0\n",
       "log_NumLabsByDay               0\n",
       "log_NumMicroLabsByDay          0\n",
       "log_NumNotesByDay              0\n",
       "log_NumOutputByDay             0\n",
       "log_NumRxByDay                 0\n",
       "log_NumProcEventsByDay         0\n",
       "log_NumTransfersByDay          0\n",
       "log_NumChartEventsByDay        0\n",
       "log_TotalNumInteractByDay      0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 254,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nooutlier_missing.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Exploration (Missing Set):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hadm_id</th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>LOSdays</th>\n",
       "      <th>NumCallouts</th>\n",
       "      <th>NumDiagnosis</th>\n",
       "      <th>NumProcs</th>\n",
       "      <th>NumCPTevents</th>\n",
       "      <th>NumInput</th>\n",
       "      <th>NumLabs</th>\n",
       "      <th>NumMicroLabs</th>\n",
       "      <th>NumNotes</th>\n",
       "      <th>NumOutput</th>\n",
       "      <th>NumRx</th>\n",
       "      <th>NumProcEvents</th>\n",
       "      <th>NumTransfers</th>\n",
       "      <th>NumChartEvents</th>\n",
       "      <th>ExpiredHospital</th>\n",
       "      <th>TotalNumInteract</th>\n",
       "      <th>LOSgroupNum</th>\n",
       "      <th>NumCalloutsByDay</th>\n",
       "      <th>NumDiagnosisByDay</th>\n",
       "      <th>NumProcsByDay</th>\n",
       "      <th>NumCPTeventsByDay</th>\n",
       "      <th>NumInputByDay</th>\n",
       "      <th>NumLabsByDay</th>\n",
       "      <th>NumMicroLabsByDay</th>\n",
       "      <th>NumNotesByDay</th>\n",
       "      <th>NumOutputByDay</th>\n",
       "      <th>NumRxByDay</th>\n",
       "      <th>NumProcEventsByDay</th>\n",
       "      <th>NumTransfersByDay</th>\n",
       "      <th>NumChartEventsByDay</th>\n",
       "      <th>TotalNumInteractByDay</th>\n",
       "      <th>log_age</th>\n",
       "      <th>log_LOSdays</th>\n",
       "      <th>log_NumCalloutsByDay</th>\n",
       "      <th>log_NumDiagnosisByDay</th>\n",
       "      <th>log_NumProcsByDay</th>\n",
       "      <th>log_NumCPTeventsByDay</th>\n",
       "      <th>log_NumInputByDay</th>\n",
       "      <th>log_NumLabsByDay</th>\n",
       "      <th>log_NumMicroLabsByDay</th>\n",
       "      <th>log_NumNotesByDay</th>\n",
       "      <th>log_NumOutputByDay</th>\n",
       "      <th>log_NumRxByDay</th>\n",
       "      <th>log_NumProcEventsByDay</th>\n",
       "      <th>log_NumTransfersByDay</th>\n",
       "      <th>log_NumChartEventsByDay</th>\n",
       "      <th>log_TotalNumInteractByDay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.00000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "      <td>1022.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>149354.078278</td>\n",
       "      <td>0.627202</td>\n",
       "      <td>55.575342</td>\n",
       "      <td>9.599080</td>\n",
       "      <td>0.088650</td>\n",
       "      <td>1.855313</td>\n",
       "      <td>0.584276</td>\n",
       "      <td>0.812123</td>\n",
       "      <td>33.028728</td>\n",
       "      <td>48.728483</td>\n",
       "      <td>1.043669</td>\n",
       "      <td>1.892378</td>\n",
       "      <td>8.570470</td>\n",
       "      <td>7.442016</td>\n",
       "      <td>0.305949</td>\n",
       "      <td>0.852573</td>\n",
       "      <td>587.495998</td>\n",
       "      <td>0.0</td>\n",
       "      <td>692.700626</td>\n",
       "      <td>2.364971</td>\n",
       "      <td>0.034787</td>\n",
       "      <td>1.059990</td>\n",
       "      <td>0.306663</td>\n",
       "      <td>0.232529</td>\n",
       "      <td>8.127596</td>\n",
       "      <td>20.014553</td>\n",
       "      <td>0.254198</td>\n",
       "      <td>2.224912</td>\n",
       "      <td>2.531075</td>\n",
       "      <td>2.652917</td>\n",
       "      <td>0.135415</td>\n",
       "      <td>0.510888</td>\n",
       "      <td>181.060446</td>\n",
       "      <td>219.145967</td>\n",
       "      <td>3.929486</td>\n",
       "      <td>1.846137</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-1.612219</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-2.479583</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-2.377210</td>\n",
       "      <td>-inf</td>\n",
       "      <td>4.470502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>28844.193787</td>\n",
       "      <td>0.483786</td>\n",
       "      <td>20.928723</td>\n",
       "      <td>9.366163</td>\n",
       "      <td>0.167846</td>\n",
       "      <td>2.008868</td>\n",
       "      <td>0.757424</td>\n",
       "      <td>0.812692</td>\n",
       "      <td>34.914997</td>\n",
       "      <td>28.326077</td>\n",
       "      <td>1.324976</td>\n",
       "      <td>5.539871</td>\n",
       "      <td>5.911759</td>\n",
       "      <td>5.980348</td>\n",
       "      <td>0.883000</td>\n",
       "      <td>0.879232</td>\n",
       "      <td>383.202993</td>\n",
       "      <td>0.0</td>\n",
       "      <td>431.203866</td>\n",
       "      <td>1.161269</td>\n",
       "      <td>0.156322</td>\n",
       "      <td>3.148238</td>\n",
       "      <td>1.202934</td>\n",
       "      <td>0.639912</td>\n",
       "      <td>20.789748</td>\n",
       "      <td>53.950825</td>\n",
       "      <td>0.757667</td>\n",
       "      <td>11.058828</td>\n",
       "      <td>5.588925</td>\n",
       "      <td>6.643177</td>\n",
       "      <td>0.988487</td>\n",
       "      <td>1.527115</td>\n",
       "      <td>389.608413</td>\n",
       "      <td>480.348907</td>\n",
       "      <td>0.447957</td>\n",
       "      <td>0.963968</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.746586</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.293547</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.787043</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.242439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>100510.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.380000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.001290</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000143</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.001004</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.872247</td>\n",
       "      <td>2.833213</td>\n",
       "      <td>-0.967584</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-6.652863</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-8.849514</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-6.903604</td>\n",
       "      <td>-inf</td>\n",
       "      <td>0.627139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>123497.250000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>39.250000</td>\n",
       "      <td>3.580000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.712500</td>\n",
       "      <td>0.170000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.130000</td>\n",
       "      <td>32.432500</td>\n",
       "      <td>0.150000</td>\n",
       "      <td>0.210000</td>\n",
       "      <td>3.932500</td>\n",
       "      <td>3.370000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.340000</td>\n",
       "      <td>284.602500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>351.600000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.059703</td>\n",
       "      <td>0.011721</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.316651</td>\n",
       "      <td>2.921162</td>\n",
       "      <td>0.013600</td>\n",
       "      <td>0.017314</td>\n",
       "      <td>0.478434</td>\n",
       "      <td>0.160027</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.029023</td>\n",
       "      <td>33.325518</td>\n",
       "      <td>40.918623</td>\n",
       "      <td>3.669891</td>\n",
       "      <td>1.275363</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-2.818385</td>\n",
       "      <td>-4.446360</td>\n",
       "      <td>-inf</td>\n",
       "      <td>0.275086</td>\n",
       "      <td>1.071977</td>\n",
       "      <td>-4.297928</td>\n",
       "      <td>-4.056244</td>\n",
       "      <td>-0.737237</td>\n",
       "      <td>-1.832427</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-3.539677</td>\n",
       "      <td>3.506323</td>\n",
       "      <td>3.711585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>149723.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>55.000000</td>\n",
       "      <td>6.630000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.210000</td>\n",
       "      <td>0.370000</td>\n",
       "      <td>0.615000</td>\n",
       "      <td>19.945000</td>\n",
       "      <td>43.470000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.430000</td>\n",
       "      <td>7.105000</td>\n",
       "      <td>6.935000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.585000</td>\n",
       "      <td>490.405000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>581.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.186949</td>\n",
       "      <td>0.036867</td>\n",
       "      <td>0.057890</td>\n",
       "      <td>3.132771</td>\n",
       "      <td>5.817664</td>\n",
       "      <td>0.067037</td>\n",
       "      <td>0.063710</td>\n",
       "      <td>0.984344</td>\n",
       "      <td>0.792364</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.092820</td>\n",
       "      <td>68.861794</td>\n",
       "      <td>81.697867</td>\n",
       "      <td>4.007333</td>\n",
       "      <td>1.891605</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-1.676920</td>\n",
       "      <td>-3.300446</td>\n",
       "      <td>-2.849218</td>\n",
       "      <td>1.141917</td>\n",
       "      <td>1.760898</td>\n",
       "      <td>-2.702508</td>\n",
       "      <td>-2.753413</td>\n",
       "      <td>-0.015787</td>\n",
       "      <td>-0.232752</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-2.377097</td>\n",
       "      <td>4.232072</td>\n",
       "      <td>4.403028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>173034.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>12.070000</td>\n",
       "      <td>0.130000</td>\n",
       "      <td>2.197500</td>\n",
       "      <td>0.697500</td>\n",
       "      <td>1.345000</td>\n",
       "      <td>44.160000</td>\n",
       "      <td>56.225000</td>\n",
       "      <td>1.380000</td>\n",
       "      <td>1.110000</td>\n",
       "      <td>12.280000</td>\n",
       "      <td>10.817500</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.040000</td>\n",
       "      <td>836.560000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>980.845000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.015919</td>\n",
       "      <td>0.595662</td>\n",
       "      <td>0.126090</td>\n",
       "      <td>0.208041</td>\n",
       "      <td>6.787037</td>\n",
       "      <td>13.054182</td>\n",
       "      <td>0.177337</td>\n",
       "      <td>0.306759</td>\n",
       "      <td>2.122774</td>\n",
       "      <td>2.171924</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.290572</td>\n",
       "      <td>139.520845</td>\n",
       "      <td>167.977947</td>\n",
       "      <td>4.317488</td>\n",
       "      <td>2.490722</td>\n",
       "      <td>-4.14092</td>\n",
       "      <td>-0.518083</td>\n",
       "      <td>-2.070771</td>\n",
       "      <td>-1.570021</td>\n",
       "      <td>1.915009</td>\n",
       "      <td>2.569108</td>\n",
       "      <td>-1.729711</td>\n",
       "      <td>-1.181697</td>\n",
       "      <td>0.752724</td>\n",
       "      <td>0.775613</td>\n",
       "      <td>-inf</td>\n",
       "      <td>-1.235908</td>\n",
       "      <td>4.938214</td>\n",
       "      <td>5.123832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>199765.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>88.000000</td>\n",
       "      <td>69.710000</td>\n",
       "      <td>1.410000</td>\n",
       "      <td>16.900000</td>\n",
       "      <td>7.460000</td>\n",
       "      <td>4.480000</td>\n",
       "      <td>195.360000</td>\n",
       "      <td>263.380000</td>\n",
       "      <td>11.480000</td>\n",
       "      <td>68.020000</td>\n",
       "      <td>30.150000</td>\n",
       "      <td>38.640000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>1838.490000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2105.060000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.985915</td>\n",
       "      <td>33.086957</td>\n",
       "      <td>13.842105</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>242.394737</td>\n",
       "      <td>574.789474</td>\n",
       "      <td>11.333333</td>\n",
       "      <td>161.952381</td>\n",
       "      <td>51.023810</td>\n",
       "      <td>76.184211</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>14.173913</td>\n",
       "      <td>3973.928571</td>\n",
       "      <td>4807.547619</td>\n",
       "      <td>4.477337</td>\n",
       "      <td>4.244344</td>\n",
       "      <td>0.68608</td>\n",
       "      <td>3.499139</td>\n",
       "      <td>2.627715</td>\n",
       "      <td>2.079442</td>\n",
       "      <td>5.490568</td>\n",
       "      <td>6.354004</td>\n",
       "      <td>2.427748</td>\n",
       "      <td>5.087302</td>\n",
       "      <td>3.932292</td>\n",
       "      <td>4.333154</td>\n",
       "      <td>2.995732</td>\n",
       "      <td>2.651403</td>\n",
       "      <td>8.287510</td>\n",
       "      <td>8.477942</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             hadm_id       gender          age      LOSdays  NumCallouts  \\\n",
       "count    1022.000000  1022.000000  1022.000000  1022.000000  1022.000000   \n",
       "mean   149354.078278     0.627202    55.575342     9.599080     0.088650   \n",
       "std     28844.193787     0.483786    20.928723     9.366163     0.167846   \n",
       "min    100510.000000     0.000000    17.000000     0.380000     0.000000   \n",
       "25%    123497.250000     0.000000    39.250000     3.580000     0.000000   \n",
       "50%    149723.500000     1.000000    55.000000     6.630000     0.000000   \n",
       "75%    173034.500000     1.000000    75.000000    12.070000     0.130000   \n",
       "max    199765.000000     1.000000    88.000000    69.710000     1.410000   \n",
       "\n",
       "       NumDiagnosis     NumProcs  NumCPTevents     NumInput      NumLabs  \\\n",
       "count   1022.000000  1022.000000   1022.000000  1022.000000  1022.000000   \n",
       "mean       1.855313     0.584276      0.812123    33.028728    48.728483   \n",
       "std        2.008868     0.757424      0.812692    34.914997    28.326077   \n",
       "min        0.050000     0.000000      0.000000     0.000000     0.000000   \n",
       "25%        0.712500     0.170000      0.000000     9.130000    32.432500   \n",
       "50%        1.210000     0.370000      0.615000    19.945000    43.470000   \n",
       "75%        2.197500     0.697500      1.345000    44.160000    56.225000   \n",
       "max       16.900000     7.460000      4.480000   195.360000   263.380000   \n",
       "\n",
       "       NumMicroLabs     NumNotes    NumOutput        NumRx  NumProcEvents  \\\n",
       "count   1022.000000  1022.000000  1022.000000  1022.000000    1022.000000   \n",
       "mean       1.043669     1.892378     8.570470     7.442016       0.305949   \n",
       "std        1.324976     5.539871     5.911759     5.980348       0.883000   \n",
       "min        0.000000     0.010000     0.000000     0.000000       0.000000   \n",
       "25%        0.150000     0.210000     3.932500     3.370000       0.000000   \n",
       "50%        0.600000     0.430000     7.105000     6.935000       0.000000   \n",
       "75%        1.380000     1.110000    12.280000    10.817500       0.000000   \n",
       "max       11.480000    68.020000    30.150000    38.640000      10.000000   \n",
       "\n",
       "       NumTransfers  NumChartEvents  ExpiredHospital  TotalNumInteract  \\\n",
       "count   1022.000000     1022.000000           1022.0       1022.000000   \n",
       "mean       0.852573      587.495998              0.0        692.700626   \n",
       "std        0.879232      383.202993              0.0        431.203866   \n",
       "min        0.050000        0.000000              0.0          8.500000   \n",
       "25%        0.340000      284.602500              0.0        351.600000   \n",
       "50%        0.585000      490.405000              0.0        581.000000   \n",
       "75%        1.040000      836.560000              0.0        980.845000   \n",
       "max        8.000000     1838.490000              0.0       2105.060000   \n",
       "\n",
       "       LOSgroupNum  NumCalloutsByDay  NumDiagnosisByDay  NumProcsByDay  \\\n",
       "count  1022.000000       1022.000000        1022.000000    1022.000000   \n",
       "mean      2.364971          0.034787           1.059990       0.306663   \n",
       "std       1.161269          0.156322           3.148238       1.202934   \n",
       "min       1.000000          0.000000           0.001290       0.000000   \n",
       "25%       1.000000          0.000000           0.059703       0.011721   \n",
       "50%       2.000000          0.000000           0.186949       0.036867   \n",
       "75%       4.000000          0.015919           0.595662       0.126090   \n",
       "max       4.000000          1.985915          33.086957      13.842105   \n",
       "\n",
       "       NumCPTeventsByDay  NumInputByDay  NumLabsByDay  NumMicroLabsByDay  \\\n",
       "count        1022.000000    1022.000000   1022.000000        1022.000000   \n",
       "mean            0.232529       8.127596     20.014553           0.254198   \n",
       "std             0.639912      20.789748     53.950825           0.757667   \n",
       "min             0.000000       0.000000      0.000000           0.000000   \n",
       "25%             0.000000       1.316651      2.921162           0.013600   \n",
       "50%             0.057890       3.132771      5.817664           0.067037   \n",
       "75%             0.208041       6.787037     13.054182           0.177337   \n",
       "max             8.000000     242.394737    574.789474          11.333333   \n",
       "\n",
       "       NumNotesByDay  NumOutputByDay   NumRxByDay  NumProcEventsByDay  \\\n",
       "count    1022.000000     1022.000000  1022.000000         1022.000000   \n",
       "mean        2.224912        2.531075     2.652917            0.135415   \n",
       "std        11.058828        5.588925     6.643177            0.988487   \n",
       "min         0.000143        0.000000     0.000000            0.000000   \n",
       "25%         0.017314        0.478434     0.160027            0.000000   \n",
       "50%         0.063710        0.984344     0.792364            0.000000   \n",
       "75%         0.306759        2.122774     2.171924            0.000000   \n",
       "max       161.952381       51.023810    76.184211           20.000000   \n",
       "\n",
       "       NumTransfersByDay  NumChartEventsByDay  TotalNumInteractByDay  \\\n",
       "count        1022.000000          1022.000000            1022.000000   \n",
       "mean            0.510888           181.060446             219.145967   \n",
       "std             1.527115           389.608413             480.348907   \n",
       "min             0.001004             0.000000               1.872247   \n",
       "25%             0.029023            33.325518              40.918623   \n",
       "50%             0.092820            68.861794              81.697867   \n",
       "75%             0.290572           139.520845             167.977947   \n",
       "max            14.173913          3973.928571            4807.547619   \n",
       "\n",
       "           log_age  log_LOSdays  log_NumCalloutsByDay  log_NumDiagnosisByDay  \\\n",
       "count  1022.000000  1022.000000            1022.00000            1022.000000   \n",
       "mean      3.929486     1.846137                  -inf              -1.612219   \n",
       "std       0.447957     0.963968                   NaN               1.746586   \n",
       "min       2.833213    -0.967584                  -inf              -6.652863   \n",
       "25%       3.669891     1.275363                  -inf              -2.818385   \n",
       "50%       4.007333     1.891605                  -inf              -1.676920   \n",
       "75%       4.317488     2.490722              -4.14092              -0.518083   \n",
       "max       4.477337     4.244344               0.68608               3.499139   \n",
       "\n",
       "       log_NumProcsByDay  log_NumCPTeventsByDay  log_NumInputByDay  \\\n",
       "count        1022.000000            1022.000000        1022.000000   \n",
       "mean                -inf                   -inf               -inf   \n",
       "std                  NaN                    NaN                NaN   \n",
       "min                 -inf                   -inf               -inf   \n",
       "25%            -4.446360                   -inf           0.275086   \n",
       "50%            -3.300446              -2.849218           1.141917   \n",
       "75%            -2.070771              -1.570021           1.915009   \n",
       "max             2.627715               2.079442           5.490568   \n",
       "\n",
       "       log_NumLabsByDay  log_NumMicroLabsByDay  log_NumNotesByDay  \\\n",
       "count       1022.000000            1022.000000        1022.000000   \n",
       "mean               -inf                   -inf          -2.479583   \n",
       "std                 NaN                    NaN           2.293547   \n",
       "min                -inf                   -inf          -8.849514   \n",
       "25%            1.071977              -4.297928          -4.056244   \n",
       "50%            1.760898              -2.702508          -2.753413   \n",
       "75%            2.569108              -1.729711          -1.181697   \n",
       "max            6.354004               2.427748           5.087302   \n",
       "\n",
       "       log_NumOutputByDay  log_NumRxByDay  log_NumProcEventsByDay  \\\n",
       "count         1022.000000     1022.000000             1022.000000   \n",
       "mean                 -inf            -inf                    -inf   \n",
       "std                   NaN             NaN                     NaN   \n",
       "min                  -inf            -inf                    -inf   \n",
       "25%             -0.737237       -1.832427                    -inf   \n",
       "50%             -0.015787       -0.232752                    -inf   \n",
       "75%              0.752724        0.775613                    -inf   \n",
       "max              3.932292        4.333154                2.995732   \n",
       "\n",
       "       log_NumTransfersByDay  log_NumChartEventsByDay  \\\n",
       "count            1022.000000              1022.000000   \n",
       "mean               -2.377210                     -inf   \n",
       "std                 1.787043                      NaN   \n",
       "min                -6.903604                     -inf   \n",
       "25%                -3.539677                 3.506323   \n",
       "50%                -2.377097                 4.232072   \n",
       "75%                -1.235908                 4.938214   \n",
       "max                 2.651403                 8.287510   \n",
       "\n",
       "       log_TotalNumInteractByDay  \n",
       "count                1022.000000  \n",
       "mean                    4.470502  \n",
       "std                     1.242439  \n",
       "min                     0.627139  \n",
       "25%                     3.711585  \n",
       "50%                     4.403028  \n",
       "75%                     5.123832  \n",
       "max                     8.477942  "
      ]
     },
     "execution_count": 257,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nooutlier_missing.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from scipy import stats\n",
    "import matplotlib.pylab as plt\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "\n",
    "(nooutlier_missing.gender).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_missing.gender.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Missing Gender\")\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Histogram of Missing Patient Age when Admitted')"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(nooutlier_missing.age)\n",
    "plt.title(\"Histogram of Missing Patient Age when Admitted\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'LOS in Days')"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(nooutlier_missing.LOSdays, bins = 50)\n",
    "plt.title(\"LOS in Days\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAE9CAYAAADnIbI9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAVW0lEQVR4nO3dfdScdX3n8fcHUlCRhwRDRIgELbZQXVbMKquLdU23BewaBEUobaKGzZ4ubnFxj8U9q671lIW21IdTlnNyiBopVXPUmiy1uhbBNbsLbfABEOshy8Y8NIQgD4rUVuS7f8wVHW7uyMw9NzNJfu/XOXPu6/dwzfWdP/KZK7+55ppUFZKkNhww6QIkSeNj6EtSQwx9SWqIoS9JDTH0Jakhhr4kNcTQ134vyZuSbJjhvqcl+fZs1yRNiqGvfVaSm5I8kOTgp+oYVfWVqvqFvmNuTvIre6jngiQPd4+/T/JYX/vhp6pGaRiGvvZJSRYBpwEFvHaixXSq6rqqemZVPRM4A/i73e2uT5o4Q1/7qmXAzcBHgeX9A0mOTLI+yfeS/DXw/CnjleTfJbkryfeTvC/J85P8726ftUkO6ua+Ksm2bvta4LnAf+/O3t8xTMFJ3pnkk1P6/luSK7vtDUl+P8nGJA8l+fMkc/vmviLJzUkeTPL1JK8c5vgSGPrady0Drusev5ZkQd/YVcAPgaOBt3SPqX4NeAlwKvAOYBXwm8BC4IXA+VN3qKrfArYA/7o7e/+DIWu+FnhNksMAujeWNwIfm/K6lgHPAQK8v5u7EFgPvAeYB1wKfCbJkUPWoMYZ+trnJPkXwHHA2qq6Ffi/wG90YwcC5wDvrqofVNUdwJppnuYPqup7VfVN4A7gf1TV3VX1EPCXwItnu+6q2gb8n64+gDOB7VX1jb5pa6rqzqr6AfBu4LwkofdGsL6qvlBVj1XV54FvAKfPdp3avxn62hctpxfS93XtP+OnSzzzgTnA1r7535nmOXb2bf/9NO2nag1+Db3/UdD9vXbK+NS6D6Z3Zn8ccH63tPNgkgfp/S/lOU9RndpPzZl0AdIwkjwdOBc4MMk9XffBwBFJTqZ31v4ovWWav+3GnzuLJYx6W9rPAH+S5Jfofdj7O1PGF/ZtPxf4B+B+em8GH6mq3x7x+GqcZ/ra15wF/Bg4Cfin3eNE4CvAsqr6Mb1g/S9JnpHkJKZ80DuincDzZrpzVT0C/DnwceB/VdXfTZmyLMkvJjkEeC+9Jayi9z+C1yX5V0kOTPK0JP8yiWf6Goqhr33NcnpnvFuq6p7dD+BPgAuSzAHeSm955h56V/d8ZBaP/1+B/9wtsfzHGT7HGuBFPHFph67vT4EdwIHA2wCqajPwOuBdwC56Hyi/Hf8Na0jxR1Sk8UryPOA2YEH3ge3u/g3ANVX10UnVpv2fZwnSGCU5ALgE+LP+wJfGxQ9ypTFJcjiwHdhM73sC0ti5vCNJDXF5R5IaYuhLUkP26jX9Zz3rWbVo0aJJlyFJ+5Rbb731vqqaP93YXh36ixYtYuPGjZMuQ9qjt7zlLVx//fUcddRR3HHHHQDcf//9vPGNb2Tz5s0sWrSItWvXMnfuXK677jquuOIKqopDDz2Uq6++mpNPPpmtW7eybNkydu7cSRJWrlzJxRdfPOFXpn1ZkuluPQK4vCON5E1vehOf//znH9d3+eWXs2TJEu666y6WLFnC5ZdfDsDxxx/Pl7/8ZW6//Xbe9a53sXLlSgDmzJnDlVdeyZ133snNN9/MVVddxZ133jn216I2GPrSCF75ylcyb968x/WtW7eO5ct7d35Yvnw5n/3sZwF4+ctfzty5vdvjn3rqqWzbtg2Ao48+mlNOOQWAQw89lBNPPJHt27eP6yWoMYa+NMt27tzJ0UcfDcCzn/1sdu7c+YQ5q1ev5owzznhC/+bNm/na177Gy172sqe8TrVpr17Tl/Z1SejdDv+nbrzxRlavXs2GDY//rfaHH36Yc845hw984AMcdthh4yxTDfFMX5plCxYsYMeOHQDs2LGDo4466idjt912GxdeeCHr1q3jyCN/+qNXP/rRjzjnnHO44IILOPvss8des9ph6Euz7LWvfS1r1vR+rGvNmjUsXboUgC1btnD22Wdz7bXX8oIXvOAn86uKFStWcOKJJ3LJJZdMpGa1Y6++DcPixYvLSza1Nzv//PO56aabuO+++1iwYAHvfe97Oeusszj33HPZsmULxx13HGvXrmXevHlceOGFfPrTn+a4444DelftbNy4kQ0bNnDaaafxohe9iAMO6J2HXXbZZZx55pmTfGnahyW5taoWTztm6EvS/uVnhb4f5DZk0aV/MekSpGltvvw1ky6hGU+6pp/kw0nuTXJHX9+8JF9Mclf3d27XnyQfSrIpyW1JTunbZ3k3/64ks/nzdZKkAQ3yQe5HgdOn9F0K3FBVJwA3dG3o/dDzCd1jJXA19N4kgPcALwNeCrxn9xuFJGl8njT0q+p/AvdP6V5K73c+6f6e1df/seq5GTgiydH0fjDii1V1f1U9AHyRJ76RSJKeYjO9ZHNBVe3otu8BFnTbxwBb++Zt6/r21C9JGqORr9Ov3uU/s3YJUJKVSTYm2bhr167ZelpJEjMP/Z3dsg3d33u7/u3Awr55x3Z9e+p/gqpaVVWLq2rx/PnT3g5akjRDMw399cDuK3CWA+v6+pd1V/GcCjzULQN9AfjVJHO7D3B/teuTJI3Rk16nn+TjwKuAZyXZRu8qnMuBtUlWAN8Bzu2mfw44E9gEPAK8GaCq7k/yPuBvunm/V1VTPxyWJD3FnjT0q+r8PQwtmWZuARft4Xk+DHx4qOokSbPKG65JUkMMfUlqiKEvSQ0x9CWpIYa+JDXE0Jekhhj6ktQQQ1+SGmLoS1JDDH1JaoihL0kNMfQlqSGGviQ1xNCXpIYY+pLUEENfkhpi6EtSQwx9SWqIoS9JDTH0Jakhhr4kNcTQl6SGGPqS1BBDX5IaYuhLUkMMfUlqiKEvSQ0x9CWpIYa+JDXE0Jekhhj6ktQQQ1+SGjJS6Cf5D0m+meSOJB9P8rQkxye5JcmmJJ9MclA39+CuvakbXzQbL0CSNLgZh36SY4DfARZX1QuBA4HzgCuA91fVzwMPACu6XVYAD3T97+/mSZLGaNTlnTnA05PMAZ4B7ABeDXyqG18DnNVtL+3adONLkmTE40uShjDj0K+q7cAfAVvohf1DwK3Ag1X1aDdtG3BMt30MsLXb99Fu/pFTnzfJyiQbk2zctWvXTMuTJE1jlOWdufTO3o8HngMcApw+akFVtaqqFlfV4vnz54/6dJKkPqMs7/wK8P+qaldV/Qj4DPAK4IhuuQfgWGB7t70dWAjQjR8OfHeE40uShjRK6G8BTk3yjG5tfglwJ3Aj8PpuznJgXbe9vmvTjX+pqmqE40uShjTKmv4t9D6Q/Spwe/dcq4DfBS5Jsonemv3qbpfVwJFd/yXApSPULUmagTlPPmXPquo9wHumdN8NvHSauT8E3jDK8SRJo/EbuZLUEENfkhpi6EtSQwx9SWqIoS9JDTH0Jakhhr4kNcTQl6SGGPqS1BBDX5IaYuhLUkMMfUlqiKEvSQ0x9CWpIYa+JDXE0Jekhhj6ktQQQ1+SGmLoS1JDDH1JaoihL0kNMfQlqSGGviQ1xNCXpIYY+pLUEENfkhpi6EtSQwx9SWqIoS9JDTH0Jakhhr4kNWSk0E9yRJJPJfnbJN9K8s+TzEvyxSR3dX/ndnOT5ENJNiW5Lckps/MSJEmDGvVM/4PA56vqF4GTgW8BlwI3VNUJwA1dG+AM4ITusRK4esRjS5KGNOPQT3I48EpgNUBV/WNVPQgsBdZ009YAZ3XbS4GPVc/NwBFJjp5x5ZKkoY1ypn88sAv4SJKvJbkmySHAgqra0c25B1jQbR8DbO3bf1vXJ0kak1FCfw5wCnB1Vb0Y+AE/XcoBoKoKqGGeNMnKJBuTbNy1a9cI5UmSphol9LcB26rqlq79KXpvAjt3L9t0f+/txrcDC/v2P7bre5yqWlVVi6tq8fz580coT5I01YxDv6ruAbYm+YWuawlwJ7AeWN71LQfWddvrgWXdVTynAg/1LQNJksZgzoj7/3vguiQHAXcDb6b3RrI2yQrgO8C53dzPAWcCm4BHurmSpDEaKfSr6uvA4mmGlkwzt4CLRjmeJGk0fiNXkhpi6EtSQwx9SWqIoS9JDTH0Jakhhr4kNcTQl6SGGPqS1BBDX5IaYuhLUkMMfUlqiKEvSQ0x9CWpIYa+JDXE0Jekhhj6ktQQQ1+SGmLoS1JDDH1JaoihL0kNMfQlqSGGviQ1xNCXpIYY+pLUEENfkhpi6EtSQwx9SWqIoS9JDTH0Jakhhr4kNcTQl6SGGPqS1JCRQz/JgUm+luT6rn18kluSbEryySQHdf0Hd+1N3fiiUY8tSRrObJzpXwx8q699BfD+qvp54AFgRde/Anig639/N0+SNEYjhX6SY4HXANd07QCvBj7VTVkDnNVtL+3adONLuvmSpDEZ9Uz/A8A7gMe69pHAg1X1aNfeBhzTbR8DbAXoxh/q5j9OkpVJNibZuGvXrhHLkyT1m3HoJ/l14N6qunUW66GqVlXV4qpaPH/+/Nl8aklq3pwR9n0F8NokZwJPAw4DPggckWROdzZ/LLC9m78dWAhsSzIHOBz47gjHlyQNacZn+lX1zqo6tqoWAecBX6qqC4Abgdd305YD67rt9V2bbvxLVVUzPb4kaXhPxXX6vwtckmQTvTX71V3/auDIrv8S4NKn4NiSpJ9hlOWdn6iqm4Cbuu27gZdOM+eHwBtm43iSpJnxG7mS1BBDX5IaYuhLUkMMfUlqiKEvSQ0x9CWpIYa+JDXE0Jekhhj6ktQQQ1+SGmLoS1JDDH1JaoihL0kNMfQlqSGGviQ1xNCXpIYY+pLUEENfkhpi6EtSQwx9SWqIoS9JDTH0Jakhhr4kNcTQl6SGGPqS1BBDX5IaYuhLUkMMfUlqiKEvSQ0x9CWpIYa+JDVkxqGfZGGSG5PcmeSbSS7u+ucl+WKSu7q/c7v+JPlQkk1Jbktyymy9CEnSYEY5038UeHtVnQScClyU5CTgUuCGqjoBuKFrA5wBnNA9VgJXj3BsSdIMzDj0q2pHVX212/4+8C3gGGApsKabtgY4q9teCnysem4Gjkhy9IwrlyQNbVbW9JMsAl4M3AIsqKod3dA9wIJu+xhga99u27o+SdKYjBz6SZ4JfBp4W1V9r3+sqgqoIZ9vZZKNSTbu2rVr1PIkSX1GCv0kP0cv8K+rqs903Tt3L9t0f+/t+rcDC/t2P7bre5yqWlVVi6tq8fz580cpT5I0xShX7wRYDXyrqv64b2g9sLzbXg6s6+tf1l3FcyrwUN8ykCRpDOaMsO8rgN8Cbk/y9a7vPwGXA2uTrAC+A5zbjX0OOBPYBDwCvHmEY0uSZmDGoV9VG4DsYXjJNPMLuGimx5Mkjc5v5EpSQwx9SWqIoS9JDTH0Jakhhr4kNcTQl6SGGPqS1BBDX5IaYuhLUkMMfUlqiKEvSQ0x9CWpIYa+JDXE0Jekhhj6ktQQQ1+SGmLoS1JDDH1JaoihL0kNMfQlqSGGviQ1xNCXpIYY+pLUEENfkhpi6EtSQwx9SWqIoS9JDTH0Jakhhr4kNcTQl6SGGPqS1BBDX5IaMvbQT3J6km8n2ZTk0nEfX5JaNtbQT3IgcBVwBnAScH6Sk8ZZgyS1bNxn+i8FNlXV3VX1j8AngKVjrkGSmjVnzMc7Btja194GvKx/QpKVwMqu+XCSb4+pNmlYzwLum3QR+4NcMekK9jvH7Wlg3KH/pKpqFbBq0nVITybJxqpaPOk6pGGMe3lnO7Cwr31s1ydJGoNxh/7fACckOT7JQcB5wPox1yBJzRrr8k5VPZrkrcAXgAOBD1fVN8dZgzSLXIbUPidVNekaJElj4jdyJakhhr4kNcTQl6SGGPrSAJIsmnQN0mww9KXB/FWSS5PsdV9olIZh6EuDeTGwALg1yWmTLkaaKS/ZlIaQ5CXADfTuG/UYEKCq6p9MtDBpQP5XVRpQklcDHwSuoXeL8McmW5E0PENfGkCST9C7V9RvVNXtk65HmilDXxrMX1XVNZMuQhqVH+RKgzksyYqpnUlWJHnbJAqSZsIPcqUBJLkVOLWqfjSl/yBgox/kal/hmb40mDlTAx+g+9nPTKAeaUYMfWkwByRZMLVzuj5pb2boS4P5Q+AvkvxykkO7x6uA64E/mmxp0uBc05cGlOQM4FLghV3XHcDlVfWXk6tKGo6hL0kN8Tp9aQBJ3v0zhquq3je2YqQReKYvDSDJ26fpPgRYARxZVc8cc0nSjBj60pCSHApcTC/w1wJXVtW9k61KGozLO9KAkswDLgEuANYAp1TVA5OtShqOoS8NIMkfAmcDq4AXVdXDEy5JmhGXd6QBJHkM+AfgUaD/H83u++kfNpHCpCEZ+pLUEL+RKw2g+wGV3dvHTxk7e/wVSTPjmb40gCRfrapTpm5P15b2Zp7pS4PJHrana0t7LUNfGkztYXu6trTX8pJNaTDPS7Ke3ln97m269vF73k3au7imLw0gyS//rPGq+vK4apFGYehLUkNc05cGkGRpkov62rckubt7vH6StUnDMPSlwbwDWN/XPhj4Z8CrgN+eREHSTPhBrjSYg6pqa197Q1V9F/hukkMmVZQ0LM/0pcHM7W9U1Vv7mvPHXIs0Y4a+NJhbkvybqZ1J/i3w1xOoR5oRr96RBpDkKOCz9O60+dWu+yX01vbPqqqdk6pNGoahLw2hu/HaL3XNb1bVlyZZjzQsQ18aQZIjgIuq6vcnXYs0CNf0pQEkWZhkVZLrk1yY5JAkVwJ3AUdNuj5pUF6yKQ3mY8CXgU8DpwMbga/T++nEeyZZmDQMl3ekAST5RlWd3NfeBjy3qh6bYFnS0DzTlwaUZC4/vXf+d4HDkwSgqu6fWGHSEDzTlwaQZDPwGNP/YEpV1fPGW5E0M4a+JDXEq3ekAST5zb7tV0wZe+sT95D2Tp7pSwPwh9G1v/BMXxqMP4yu/YKhLw3GH0bXfsHlHWkASR4BNtE7q39+t03Xfl5VeU997RO8Tl8azImTLkCaDZ7pS1JDPNOXBpDk+zx+7T5dO/S+nHXYRAqThmToS4O5AXg28BngE1W1ZcL1SDPi8o40oCSHA2cD5wFPAz5J7w3A++5on2HoS0NKcgC94P8QcFlV/fGES5IG5vKONKAkLwfOB04DNgCvq6qvTLYqaTie6UsD6O6y+SDwCeBLwKP941X11Wl2k/Y6hr40gCQ3sedv3lZVvXqM5UgzZuhLUkO89440gCTv6Nt+w5Sxy8ZfkTQzhr40mPP6tt85Zez0cRYijcLQlwbjrZW1XzD0pcF4a2XtF/wgVxpAkh8DP6B3Vv904JHdQ8DTqurnJlWbNAxDX5Ia4vKOJDXE0Jekhhj6ktQQQ1+SGmLoS1JD/j9BQJau0EpswQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_missing.admit_type).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_missing.admit_type.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Admit Type\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_missing.ethnicity).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_missing.ethnicity.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Ethnicity\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_missing.insurance).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_missing.insurance.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Insurance\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MARRIED      45\n",
       "UNMARRIED    44\n",
       "UNKNOWN       1\n",
       "Name: marital_status, dtype: int64"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nooutlier_missing.marital_status.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_missing.religion).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_missing.religion.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Religion\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Non-missing data exploration:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "log_transform(raw_completecase, 'age')\n",
    "log_transform(raw_completecase, 'LOSdays')\n",
    "log_transform(raw_completecase, 'NumCalloutsByDay')\n",
    "log_transform(raw_completecase, 'NumDiagnosisByDay')\n",
    "log_transform(raw_completecase, 'NumProcsByDay')\n",
    "log_transform(raw_completecase, 'NumCPTeventsByDay')\n",
    "log_transform(raw_completecase, 'NumInputByDay')\n",
    "log_transform(raw_completecase, 'NumLabsByDay')\n",
    "log_transform(raw_completecase, 'NumMicroLabsByDay')\n",
    "log_transform(raw_completecase, 'NumNotesByDay')\n",
    "log_transform(raw_completecase, 'NumOutputByDay')\n",
    "log_transform(raw_completecase, 'NumRxByDay')\n",
    "log_transform(raw_completecase, 'NumProcEventsByDay')\n",
    "log_transform(raw_completecase, 'NumTransfersByDay')\n",
    "log_transform(raw_completecase, 'NumChartEventsByDay')\n",
    "log_transform(raw_completecase, 'TotalNumInteractByDay')\n",
    "\n",
    "# from pandas.api.types import is_numeric_dtype\n",
    "\n",
    "# def remove_outlier(df):\n",
    "#     low = .05\n",
    "#     high = .95\n",
    "#     quant_df = df.quantile([low, high])\n",
    "#     for name in list(df.columns):\n",
    "#         if is_numeric_dtype(df[name]):\n",
    "#             print(name,\": is numeric\")\n",
    "#             df = df[(df[name] > quant_df.loc[low, name]) \n",
    "#                     & (df[name] < quant_df.loc[high, name])]\n",
    "#     return df\n",
    "\n",
    "\n",
    "\n",
    "# remove_outlier(raw_completecase).head()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['hadm_id', 'gender', 'age', 'LOSdays', 'admit_type', 'admit_location',\n",
       "       'AdmitDiagnosis', 'insurance', 'religion', 'marital_status',\n",
       "       'ethnicity', 'NumCallouts', 'NumDiagnosis', 'NumProcs',\n",
       "       'AdmitProcedure', 'NumCPTevents', 'NumInput', 'NumLabs', 'NumMicroLabs',\n",
       "       'NumNotes', 'NumOutput', 'NumRx', 'NumProcEvents', 'NumTransfers',\n",
       "       'NumChartEvents', 'ExpiredHospital', 'TotalNumInteract', 'LOSgroupNum',\n",
       "       'NumCalloutsByDay', 'NumDiagnosisByDay', 'NumProcsByDay',\n",
       "       'NumCPTeventsByDay', 'NumInputByDay', 'NumLabsByDay',\n",
       "       'NumMicroLabsByDay', 'NumNotesByDay', 'NumOutputByDay', 'NumRxByDay',\n",
       "       'NumProcEventsByDay', 'NumTransfersByDay', 'NumChartEventsByDay',\n",
       "       'TotalNumInteractByDay', 'log_age', 'log_LOSdays',\n",
       "       'log_NumCalloutsByDay', 'log_NumDiagnosisByDay', 'log_NumProcsByDay',\n",
       "       'log_NumCPTeventsByDay', 'log_NumInputByDay', 'log_NumLabsByDay',\n",
       "       'log_NumMicroLabsByDay', 'log_NumNotesByDay', 'log_NumOutputByDay',\n",
       "       'log_NumRxByDay', 'log_NumProcEventsByDay', 'log_NumTransfersByDay',\n",
       "       'log_NumChartEventsByDay', 'log_TotalNumInteractByDay'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_completecase.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Consolidation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/pandas/core/indexing.py:494: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self.obj[item] = s\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py:5208: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self[name] = value\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:66: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:67: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n"
     ]
    }
   ],
   "source": [
    "# included patients consolidation:\n",
    "\n",
    "raw_completecase.loc[raw_completecase['insurance'].isin(['Private', 'Medicaid', 'Medicare', 'Government']), 'insurance'] = \"INSURED\"\n",
    "raw_completecase.loc[raw_completecase['insurance'].isin(['Self Pay']), 'insurance'] = \"UNINSURED\"\n",
    "\n",
    "raw_completecase.loc[raw_completecase['marital_status'].isin(['LIFE PARTNER']), 'marital_status'] = \"MARRIED\"\n",
    "raw_completecase.loc[raw_completecase['marital_status'].isin(['DIVORCED', 'SINGLE','WIDOWED','SEPARATED']), 'marital_status'] = \"UNMARRIED\"\n",
    "raw_completecase.loc[raw_completecase['marital_status'].isin(['UNKNOWN (DEFAULT)']), 'marital_status'] = \"UNKNOWN\"\n",
    "\n",
    "raw_completecase.loc[raw_completecase[\"religion\"].isin(['PROTESTANT QUAKER', 'JEWISH', 'CATHOLIC',\n",
    "       'EPISCOPALIAN', 'OTHER', \"JEHOVAH'S WITNESS\",\n",
    "       'UNITARIAN-UNIVERSALIST', 'LUTHERAN', 'CHRISTIAN SCIENTIST',\n",
    "       'GREEK ORTHODOX', 'BUDDHIST', '7TH DAY ADVENTIST', 'MUSLIM',\n",
    "       'HINDU', 'ROMANIAN EAST. ORTH', 'HEBREW', 'BAPTIST', 'METHODIST']), \"religion\"] = \"RELIGIOUS\"\n",
    "\n",
    "\n",
    "raw_completecase.religion = raw_completecase['religion'].replace(['NOT SPECIFIED', 'UNOBTAINABLE'], \"UNKNOWN\")\n",
    "\n",
    "\n",
    "raw_completecase.loc[raw_completecase['ethnicity'].isin(['ASIAN',\n",
    "                                                        'ASIAN - CHINESE', \n",
    "                                                         'ASIAN - ASIAN INDIAN',\n",
    "                                                        'ASIAN - OTHER',\n",
    "                                                        'ASIAN - FILIPINO',\n",
    "                                                        'ASIAN - CAMBODIAN', \n",
    "                                                        'ASIAN - VIETNAMESE',\n",
    "                                                        'ASIAN - KOREAN',\n",
    "                                                         'ASIAN - THAI', \n",
    "                                                         'ASIAN - JAPANESE',\n",
    "                                                        'NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER']), 'ethnicity'] = \"ASIAN_PACIFICISLANDER\"\n",
    "\n",
    "raw_completecase.loc[raw_completecase['ethnicity'].isin(['HISPANIC/LATINO - SALVADORAN',\n",
    "                                                        'HISPANIC/LATINO - PUERTO RICAN', \n",
    "                                                         'HISPANIC/LATINO - GUATEMALAN',\n",
    "                                                        'HISPANIC/LATINO - DOMINICAN',\n",
    "                                                        'HISPANIC/LATINO - CUBAN',\n",
    "                                                        'HISPANIC/LATINO - COLOMBIAN',\n",
    "                                                        'HISPANIC/LATINO - HONDURAN', \n",
    "                                                         'HISPANIC/LATINO - MEXICAN',\n",
    "                                                        'HISPANIC/LATINO - CENTRAL AMERICAN (OTHER)', \n",
    "                                                         'SOUTH AMERICAN']), 'ethnicity'] = \"HISPANIC OR LATINO\"\n",
    "\n",
    "raw_completecase.loc[raw_completecase['ethnicity'].isin(['WHITE - RUSSIAN',\n",
    "                                                        'WHITE - OTHER EUROPEAN',\n",
    "                                                        'WHITE - BRAZILIAN',\n",
    "                                                        'PORTUGUESE',\n",
    "                                                        'WHITE - EASTERN EUROPEAN']), 'ethnicity'] = \"WHITE\"\n",
    "\n",
    "raw_completecase.loc[raw_completecase['ethnicity'].isin(['BLACK/AFRICAN AMERICAN',\n",
    "                                                         'BLACK/CAPE VERDEAN', \n",
    "                                                         'BLACK/HAITIAN',\n",
    "                                                        'BLACK/AFRICAN']), 'ethnicity'] = \"BLACK\"\n",
    "\n",
    "\n",
    "raw_completecase.loc[raw_completecase['ethnicity'].isin(['MULTI RACE ETHNICITY',\n",
    "                                                        'AMERICAN INDIAN/ALASKA NATIVE',\n",
    "                                                        'AMERICAN INDIAN/ALASKA NATIVE FEDERALLY RECOGNIZED TRIBE',\n",
    "                                                        'MIDDLE EASTERN',\n",
    "                                                        'CARIBBEAN ISLAND']), 'ethnicity'] = \"OTHER\"\n",
    "\n",
    "raw_completecase.loc[raw_completecase['ethnicity'].isin(['UNKNOWN/NOT SPECIFIED',\n",
    "                                                        'UNABLE TO OBTAIN',\n",
    "                                                        'PATIENT DECLINED TO ANSWER']), 'ethnicity'] = \"UNKNOWN\"\n",
    "\n",
    "\n",
    "raw_completecase['gender'] = raw_completecase['gender'].replace('F', 0)\n",
    "raw_completecase['gender'] = raw_completecase['gender'].replace('M', 1)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Outlier Removal:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "nooutlier_completecase = filter_outlier(raw_completecase)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17790"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(nooutlier_completecase)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Graphs:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_completecase.gender).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_completecase.gender.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Gender\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have 6000 more males than females"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Histogram of Patient Age when Admitted')"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(nooutlier_completecase.age)\n",
    "plt.title(\"Histogram of Patient Age when Admitted\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_completecase.ExpiredHospital).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_completecase.ExpiredHospital.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Died at Hospital?\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We should probably get rid of those who died since we want Length of Stay, which will be quite different for those who are going to die?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'LOS in Days')"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(nooutlier_completecase.LOSdays, bins = 50)\n",
    "plt.title(\"LOS in Days\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Total Number of CPT Events by Day')"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(nooutlier_completecase.NumCPTeventsByDay)\n",
    "plt.title(\"Total Number of CPT Events by Day\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Total Number of Chart Events')"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(nooutlier_completecase.NumChartEventsByDay)\n",
    "plt.title(\"Total Number of Chart Events\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_completecase.admit_location).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_completecase.admit_location.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Admit Location\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Maybe we need to perform ANOVA to find out which categories can be collapsed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_completecase.admit_type).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_completecase.admit_type.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Admit Type\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Imbalance heavily favoring emergency."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_completecase.ethnicity).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_completecase.ethnicity.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Ethnicity\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Definitely consolidate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_completecase.insurance).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_completecase.insurance.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Insurance\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Insurance is worth consolidating."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAE7CAYAAAA//e0KAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAfZ0lEQVR4nO3de5xXdb3v8debixdEbgqkgIwZCcygI0xgesItBqJ2QkTydpKSs+liZXml9tlb011abQMtt23ybimpm47sRIpEU88jhUHZgqJJcldx5KaOIox+zh9rzfhjnIG5wG/9ht/7+XjMg7W+6/aZfjnv3/qu71pLEYGZmRW3dlkXYGZm2XMYmJmZw8DMzBwGZmaGw8DMzHAYmJkZDgMzACQdJukdSe1buP1Vkn6zu+syyxeHgbUpklZK2ibp4Hrtz0oKSSUt2W9ErI6IzhHxQbq/xyT979ZXXFffDyStSANnraTf5Sxr1rEcPLYnOAysLVoBnFM7I2kI0KmlO5PUYXcUtZP9TwK+DHw+IjoDFcAje/KYZs3lMLC26G7g/Jz5ScBduStIOi09W3hL0hpJV+UsK0nPIiZLWg3Mz2nrIOlHwOeAX6bf5H+ZbndDuq+3JC2S9Lkm1vsZ4I8R8XeAiHg9Imak+2zWsSSNBX4AnJWu/99p+0pJn8/5HevOHiTtJ+k3kjZI2ixpoaTeTazdioTDwNqip4AukgalffxnA/W7TapJAqMbcBrwDUmn11vnBGAQcHJuY0T8E/AE8K206+hb6aKFQDnQA7gHuF/Sfk2s93xJl0mqyL0u0dxjRcRc4MfA79L1j27C8ScBXYF+wEHA14H3mrCdFRGHgbVVtWcHo4FlwLrchRHxWEQsiYgPI+I54F6SP/65roqI6oho0h/GiPhNRGyIiJqIuB7YFziyKdsB3yYJnb8Ab0i6Yk8cqxHbSULgUxHxQUQsioi3Wrgv20s5DKytuhs4F/gK9bqIACSNkPSopCpJW0i+DR9cb7U1zTmgpEslLZO0RdJmkm/b9ffZoIj4bUR8nuRM5evANZJObmz91hyrAXcDfwRmSnpV0k8ldWzhvmwv5TCwNikiVpFcSD4VmNXAKvcAs4F+EdEV+BWg+rvZ2SFyZ9I++8uBLwHdI6IbsKWBfe6q7u0RcT/wHFDWwmM1VHc1O15E/0S9Y/4wIgYDxwFfYMdrLmYOA2vTJgOjIqK6gWUHAhsjYquk4SRnEc2xHvhkvf3VAFVAB0n/AnRpyo4kfSW9oH2gpHaSTgFKgadbeKz1QImk3P9+FwNnS+ooqQI4M+f4J0oakl6reIuk2+jDptRuxcNhYG1WRPw9IiobWfxN4GpJbwP/AtzXzN3fAJwpaZOkG0m6WeYCfwNWAVtpejfTWyQjgFYDm4GfAt+IiCdbeKz70383SHomnf5n4AhgE/BDkjOjWp8AHkjrWEZy3eLuJtZuRUJ+uY2ZmfnMwMzMHAZmZuYwMDMzHAZmZgbs0Qd07UkHH3xwlJSUZF2GmVmbsWjRojcjomdDy9psGJSUlFBZ2dioQjMzq0/SqsaWuZvIzMwcBoXshhtuoKysjNLSUqZPnw7A4sWLOfbYYykvL6eiooIFCxYAsGnTJsaPH89RRx3F8OHDWbp0ad1+SkpKGDJkSN02ZmYfExFt8mfYsGGxN1uyZEmUlpZGdXV1bN++PU466aR4+eWXY/To0TFnzpyIiHjooYfihBNOiIiISy+9NK666qqIiFi2bFmMGjWqbl/9+/ePqqqqvP8OZlZYgMpo5G/qLs8MJN0m6Q1JS3PaekiaJ+nl9N/uabsk3ShpuaTnJA3N2WZSuv7L6ZufatuHSVqSbnOjpGY9+GtvtWzZMkaMGEGnTp3o0KEDJ5xwArNmzUISb72VPH14y5YtHHrooQC88MILjBo1CoCBAweycuVK1q9fn1n9Zta2NKWb6A5gbL22qcAjETGA5PV9U9P2U4AB6c8U4GZIwgO4EhgBDAeurA2QdJ1/zNmu/rGKUllZGU888QQbNmzg3XffZc6cOaxZs4bp06dz2WWX0a9fPy699FKuvfZaAI4++mhmzUoe3rlgwQJWrVrF2rVrAZDEmDFjGDZsGDNmzMjsdzKzwrXLMIiIx4GN9ZrHAXem03cCp+e035WekTwFdJN0CMlLPeZFxMaI2ATMA8amy7pExFPpKcxdOfsqaoMGDeKKK65gzJgxjB07lvLyctq3b8/NN9/MtGnTWLNmDdOmTWPy5MkATJ06lc2bN1NeXs4vfvELjjnmGNq3T16o9eSTT/LMM8/w8MMPc9NNN/H4449n+auZWQFq6QXk3hHxWjr9OlD7PtU+7Ph0xbVp287a1zbQbsDkyZNZtGgRjz/+ON27d+fTn/40d955J2eccQYAEydOrLuA3KVLF26//XYWL17MXXfdRVVVFZ/8ZPJU5D59kv9Je/Xqxfjx4+u2MTOr1erRROk3+rw8+lTSFEmVkiqrqqrycchMvfHGGwCsXr2aWbNmce6553LooYfyl7/8BYD58+czYMAAADZv3sy2bdsAuOWWWxg5ciRdunShurqat99+G4Dq6mr+9Kc/UVZW1sDRzKyYtfSms/WSDomI19KunjfS9nUkL92u1TdtWwf8Q732x9L2vg2s36CImAHMAKioqNjrn709YcIENmzYQMeOHbnpppvo1q0bv/71r7nooouoqalhv/32q7sGsGzZMiZNmoQkSktLufXWWwFYv34948ePB6CmpoZzzz2XsWN9WcbMdtSk9xlIKgH+EBFl6fzPgA0RcZ2kqUCPiLhc0mnAt0heRTgCuDEihqcXkBcBtaOLngGGRcRGSQuA75C89WkO8IuImLOrmioqKiJfdyCXTH0oL8fJysrrTsu6BDPLA0mLIqLBm412eWYg6V6Sb/UHS1pLMiroOuA+SZNJ3sT0pXT1OSRBsBx4F/gqQPpH/xpgYbre1RFRe1H6myQjlvYHHk5/zMwsj3YZBhFxTiOLTmpg3QAubGQ/twG3NdBeyUcvBjczswz4cRRmZuYwMDMzh4GZmeEwMDMzHAZmZobDwMzMcBiYmRkOAzMzw2FgZmY4DMzMDIeBmZnhMDAzMxwGZmaGw8DMzHAYmJkZDgMzM8NhYGZmOAzMzAyHgZmZ4TAwMzMcBmZmhsPAzMxwGJiZGQ4DMzPDYWBmZjgMzMwMh4GZmeEwMDMzHAZmZobDwMzMcBiYmRkOAzMzw2FgZma0MgwkfU/S85KWSrpX0n6SDpf0tKTlkn4naZ903X3T+eXp8pKc/Xw/bX9J0smt+5XMzKy5WhwGkvoA3wEqIqIMaA+cDfwEmBYRnwI2AZPTTSYDm9L2ael6SBqcblcKjAX+XVL7ltZlVgheeuklysvL6366dOnC9OnTOeuss+raSkpKKC8vB2D79u1MmjSJIUOGMGjQIK699loA1qxZw4knnsjgwYMpLS3lhhtuyPLXsr1Yh92w/f6StgOdgNeAUcC56fI7gauAm4Fx6TTAA8AvJSltnxkR7wMrJC0HhgN/bWVtZpk58sgjWbx4MQAffPABffr0Yfz48Xz3u9+tW+eSSy6ha9euANx///28//77LFmyhHfffZfBgwdzzjnnsO+++3L99dczdOhQ3n77bYYNG8bo0aMZPHhwJr+X7b1afGYQEeuAfwNWk4TAFmARsDkiatLV1gJ90uk+wJp025p0/YNy2xvYZgeSpkiqlFRZVVXV0tLN8uqRRx7hiCOOoH///nVtEcF9993HOeecA4Akqqurqamp4b333mOfffahS5cuHHLIIQwdOhSAAw88kEGDBrFu3bpMfg/bu7Wmm6g7ybf6w4FDgQNIunn2mIiYEREVEVHRs2fPPXkos91m5syZdX/0az3xxBP07t2bAQMGAHDmmWdywAEHcMghh3DYYYdx6aWX0qNHjx22WblyJc8++ywjRozIW+1WPFpzAfnzwIqIqIqI7cAs4Higm6Ta7qe+QO3XmHVAP4B0eVdgQ257A9uYtWnbtm1j9uzZTJw4cYf2e++9d4eAWLBgAe3bt+fVV19lxYoVXH/99bzyyit1y9955x0mTJjA9OnT6dKlS97qt+LRmjBYDRwrqVPa938S8ALwKHBmus4k4MF0enY6T7p8fkRE2n52OtrocGAAsKAVdZkVjIcffpihQ4fSu3fvuraamhpmzZrFWWedVdd2zz33MHbsWDp27EivXr04/vjjqaysBJKLyxMmTOC8887jjDPOyPvvYMWhNdcMnia5EPwMsCTd1wzgCuDi9ELwQcCt6Sa3Agel7RcDU9P9PA/cRxIkc4ELI+KDltZlVkjqnwEA/PnPf2bgwIH07du3ru2www5j/vz5AFRXV/PUU08xcOBAIoLJkyczaNAgLr744rzWbsVFyZfztqeioiJqvzntaSVTH8rLcbKy8rrTsi5hj8rq8/tw21bW3fxV+nz9Ftrte0Bd+5sPTWPfQ4/kwGNOzVn3PTbMmc72N9cAwQFDPk/XERPYuvZ51v/2Cjr2LAEJgO4jz2f/Iz5Tt+3e/vnZ7iNpUURUNLSstUNLzawR7fbZj34X3fux9oNP+14D6+5Pz9O//7H2/fqW0v+KP+yR+sxy+XEUZmbmMDAzM4eBmZnhMDAzMxwGZmaGw8DMzHAYmJkZDgMzM8NhYGZmOAzMzAyHgZmZ4TAwMzMcBmZmhsPAzMxwGJiZGQ4DMzPDYWBmZjgMzMwMh4GZmeEwMDMzHAZmZobDwMzMcBiYmRkOAzMzw2FgZmY4DMzMDIeBmZnhMDAzMxwGZmaGw8DMzHAYmJkZDgMzM6OVYSCpm6QHJL0oaZmkz0rqIWmepJfTf7un60rSjZKWS3pO0tCc/UxK139Z0qTW/lJmZtY8rT0zuAGYGxEDgaOBZcBU4JGIGAA8ks4DnAIMSH+mADcDSOoBXAmMAIYDV9YGiJmZ5UeLw0BSV2AkcCtARGyLiM3AOODOdLU7gdPT6XHAXZF4Cugm6RDgZGBeRGyMiE3APGBsS+syM7Pma82ZweFAFXC7pGcl3SLpAKB3RLyWrvM60Dud7gOsydl+bdrWWPvHSJoiqVJSZVVVVStKNzOzXK0Jgw7AUODmiDgGqOajLiEAIiKAaMUxdhARMyKiIiIqevbsubt2a2ZW9FoTBmuBtRHxdDr/AEk4rE+7f0j/fSNdvg7ol7N937StsXYzM8uTFodBRLwOrJF0ZNp0EvACMBuoHRE0CXgwnZ4NnJ+OKjoW2JJ2J/0RGCOpe3rheEzaZmZmedKhldt/G/itpH2AV4CvkgTMfZImA6uAL6XrzgFOBZYD76brEhEbJV0DLEzXuzoiNrayLjMza4ZWhUFELAYqGlh0UgPrBnBhI/u5DbitNbWYmVnL+Q5kMzNzGJiZmcPAzMxwGJiZGQ4DMzPDYWBmZjgMzMwMh4GZmeEwMDMzHAZmZobDwMzMcBiYmRkOAzMzw2FgZmY4DMzMDIeBmZnhMDAzMxwGZmaGw8DMzHAYmJkZDgMzM8NhYGZmOAzMzAyHgZmZ4TAwMzMcBmZmhsPAzMxwGJiZGQ4DMzPDYWBmZjgMzMwMh4GZmeEwMDMzdkMYSGov6VlJf0jnD5f0tKTlkn4naZ+0fd90fnm6vCRnH99P21+SdHJrazIzs+bZHWcGFwHLcuZ/AkyLiE8Bm4DJaftkYFPaPi1dD0mDgbOBUmAs8O+S2u+GuszMrIlaFQaS+gKnAbek8wJGAQ+kq9wJnJ5Oj0vnSZeflK4/DpgZEe9HxApgOTC8NXWZmVnztPbMYDpwOfBhOn8QsDkiatL5tUCfdLoPsAYgXb4lXb+uvYFtdiBpiqRKSZVVVVWtLN3MzGq1OAwkfQF4IyIW7cZ6dioiZkRERURU9OzZM1+HNTPb63VoxbbHA1+UdCqwH9AFuAHoJqlD+u2/L7AuXX8d0A9YK6kD0BXYkNNeK3cbMzPLgxafGUTE9yOib0SUkFwAnh8R5wGPAmemq00CHkynZ6fzpMvnR0Sk7Weno40OBwYAC1pal5mZNV9rzgwacwUwU9K/As8Ct6bttwJ3S1oObCQJECLieUn3AS8ANcCFEfHBHqjLzMwasVvCICIeAx5Lp1+hgdFAEbEVmNjI9j8CfrQ7ajEzs+bzHchmZuYwMDMzh4GZmeEwMDMzHAZmZobDwMzMcBiYmRkOAzMzw2FgZmY4DMzMDIeBmZnhMDAzMxwGZmaGw8DMzHAYmJkZDgMzM8NhYGZmOAzMzAyHgZmZ4TAwMzMcBmZmhsPAzMxwGJiZGQ4DMzPDYWBmZjgMzMwMh4GZmeEwMDMzHAZmZobDwMzMcBiYmRkOAzMzw2FgZma0Igwk9ZP0qKQXJD0v6aK0vYekeZJeTv/tnrZL0o2Slkt6TtLQnH1NStd/WdKk1v9aZmbWHK05M6gBLomIwcCxwIWSBgNTgUciYgDwSDoPcAowIP2ZAtwMSXgAVwIjgOHAlbUBYmZm+dHiMIiI1yLimXT6bWAZ0AcYB9yZrnYncHo6PQ64KxJPAd0kHQKcDMyLiI0RsQmYB4xtaV1mZtZ8u+WagaQS4BjgaaB3RLyWLnod6J1O9wHW5Gy2Nm1rrL2h40yRVCmpsqqqaneUbmZm7IYwkNQZ+E/guxHxVu6yiAggWnuMnP3NiIiKiKjo2bPn7tqtmVnRa1UYSOpIEgS/jYhZafP6tPuH9N830vZ1QL+czfumbY21m5lZnrRmNJGAW4FlEfHznEWzgdoRQZOAB3Paz09HFR0LbEm7k/4IjJHUPb1wPCZtMzOzPOnQim2PB74MLJG0OG37AXAdcJ+kycAq4EvpsjnAqcBy4F3gqwARsVHSNcDCdL2rI2JjK+oyM7NmanEYRMSTgBpZfFID6wdwYSP7ug24raW1mJlZ6/gOZDMzcxiYmZnDwMzMcBiYmRkOAzMzw2FgZmY4DMzMDIeBmZnhMDAzMxwGZmaGw8DMzHAYmJkZDgMzM8NhYGZmOAzMzAyHgZmZ4TAwMzMcBmZmhsPAzMxwGJiZGQ4DMzPDYWBmZjgMzMwMh4GZmeEwMDMzHAZmZobDwMzsYy644AJ69epFWVlZXdv9999PaWkp7dq1o7Kysq59wYIFlJeXU15eztFHH83vf//7LEpuNYeBmVk9X/nKV5g7d+4ObWVlZcyaNYuRI0d+rL2yspLFixczd+5cvva1r1FTU5PPcneLDlkXYGZWaEaOHMnKlSt3aBs0aFCD63bq1KlueuvWrUjak6XtMT4zMDNrpaeffprS0lKGDBnCr371Kzp0aHvfsx0GZmatNGLECJ5//nkWLlzItddey9atW7MuqdkcBmZmu8mgQYPo3LkzS5cuzbqUZnMYmJm1wooVK+ouGK9atYoXX3yRkpKSbItqgYLp2JI0FrgBaA/cEhHXZVySmRWAkqkP5f2YVbN/yvurl/DBe2/R4cCD6fo/zqP9/p3ZOO8/+OC9LYw4YTT79Dqc3mddwztL5/PWUw9A+/ZI7eh63AVU/NvTTTrOyutO28O/SdMVRBhIag/cBIwG1gILJc2OiBeyrczMilHPL17eYHunTx/3sbbOZaPoXDZqT5e0xxVKN9FwYHlEvBIR24CZwLiMazIzKxoFcWYA9AHW5MyvBUbUX0nSFGBKOvuOpJfyUFsWDgbezNfB9JN8Halo+PNr2/L2+WXw2fVvbEGhhEGTRMQMYEbWdexpkiojoiLrOqxl/Pm1bcX6+RVKN9E6oF/OfN+0zczM8qBQwmAhMEDS4ZL2Ac4GZmdck5lZ0SiIbqKIqJH0LeCPJENLb4uI5zMuK0t7fVfYXs6fX9tWlJ+fIiLrGszMLGOF0k1kZmYZchiYmZnDwMzMHAZmZkaBjCYykDQEGJjOLouItvcM3CIlqQNwCjmfHzA3Itreuw+taHk0UcYkdQUeJLnp7jlAwBBgNTAuIt7KsDzbBUl9gPnAa8CzJJ/fMcAngBMj4tUMy7OdkPQ2UPsHsPZdlUHyJXmfiCiqL8sOg4xJuhHYBlweER+mbe2A64D9I+LbWdZnOyfpDmBxREyv1/4dYFhETMqkMGs2SZ2BC4GvAb+PiEsyLimvHAYZk/QCcFT9LoW062FJRDT8Fm4rCJJejIiBjSx7KSKOzHdN1jySugHfBc4H7gGmRcSGbKvKv6I6DSpQ2xrqW07vyn4/i4KsWd7bybJ381aFNZukg4FLgLOA24BjImJLtlVlx2GQvf0kHcNHfZa1BOybQT3WPF0lndFAu4Au+S7GmmUVUAXcThLck6WP/jOMiJ9nVFcmHAbZex1o7P90r+ezEGuRx4H/uZNlVrh+xkcXkA/MspBC4GsGZlaUJHWPiE1Z11EofNNZxiRdnjM9sd6yH+e/ImsOSdNzpi+qt+yOvBdkzfGSpBck/VrSVyV9OuuCsuQwyN7ZOdPfr7dsbD4LsRYZmTNdfxjpUfksxJonInoBpwP/D/gsMEvSekkP5n5JKxa+ZpA9NTLd0LwVnp19flbgIuJvwN+AOyQdAZwKXASMAX6aZW355jDIXjQy3dC8FZ52krqTnGXXTteGQvvsyrJdkXQccBzJWUE/4BXgKeB/Ac9kWFomfAE5Y5I+AKpJ/oDsz0dj0wXsFxEds6rNdk3SSuBDGj4riIj4ZH4rsqaS9CHJH/1pJHccF/V9IQ4DMytKkj5BcmZwHDCcpKfkGeCvwF8j4pUMy8s7h0HGJPXY2fKI2JivWqz5JA3d2fKIKLruhrZKUifgApJHUxweEUXVzedrBtlbRHJtoMFuBsDdDIXt+p0sC2BUvgqx5kmfGPxZPjo7OAZ4GfgvkhFGRcVnBmZWlCRVkXYJkfzxXxgRO3vW1F7NYVCg0htgLouIf8y6Fms+SaNJHks+OutabNfSx1cTEe9kXUtWfNNZxiQdJelPkpZK+ldJh0j6T5IXpryQdX22c5JGSfqbpHck/UbSEEmVJO+juDnr+mznJH1D0mqSh9atlrRK0jezrisLDoPs/ZrkGeoTSJ6guBj4O/CpiJiWZWHWJNcDU4CDgAdIuhzuiIhhETEr08pspyT9H5KHDP5DRBwUET2AE4FT0mVFxd1EGZO0OCLKc+Zf8dj0tkPSMxExNGfeL7RpIyS9BBwdEVvrte8P/HdEFNWzijyaKHv132fwfu68hyYWvG713mfQIXfeZwcFLeoHQdr4XnpDWlHxmUHGJD26k8URER6aWMAk3b6TxRERF+StGGsWSY8AP46IR+q1jwL+OSJOzKaybDgMzKwoSSoFHgSeJLnfB6ACOB4YFxHPZ1VbFnwBuUBJGi1pXtZ12K5Jap++T7d2fh9JUyQty7Iu27n0j30ZyRvpStKfx4GyYgsC8DWDzKWnpL8CDgX+L/ATkneyCvhRhqVZE0g6G/gPoFrSyySf2W3AQuC8LGuzXUuvGdyWdR2FwN1EGZP0LPA9kiGJpwC/AaZGxC8zLcyaRNJS4PSIWJ4+p+ivwJkR8V8Zl2a7IGkFjT8mPiLiiHzWkzWHQcY8NLFta+DzWxoRZVnWZE0j6aB6Te2ALwGXAs9ExIT8V5UddxNlz0MT27Zeki7Ome+WOx8RP8+gJmuCiNgAIKkd8GXgMpKbPk+LiKK7+99nBhnz0MS2TdKVO1kcEXF13oqxZpHUkeSR1d8jGVF0XUQsz7aq7DgMCpik3hGxPus6rGUkfSYiFmZdhzVM0lqgBpgOrK6/vNjOyh0GBUZSN5LnFJ0LDIqIQzMuyZpB0mDgnPRnc0RUZFySNULSHez8AnJRnZU7DApA+iyUcSQBcAxwIHA68HhEFN1t8W2NpBI+CoDtQH+gIiJWZleV7YqkrhGxpZFlFRFRme+asuSbzjIm6R7gb8Bo4BckN75siojHHASFT9JfgYdIBmNMiIhhwNsOgjbhz5K6129M30Xx+wzqyZTDIHuDgU3AMmBZRHxA46euVnjWk5zJ9QZ6pm3+/NqGGcCjkmo/NySdm7aflllVGXE3UQGQNJCki+Es4E3gSJJb4n3xuA1I36V7BslnOADoBpwcEQsyLcx2SdKXgcuBMST//X0dGFuMZ3YOgwIjaRjJtYOJwNqIOC7jkqwZJPUmuXHpbOCwiOiXcUm2C5ImknTRrgZOjYg3My4pEw6DAiVJwOci4vGsa7GWkdQ/IlZlXYc1TNISki49kVz0rwKq0/mIiKMyLC/vfAdyxiTduItVHAYFTNLsXazyxbwUYi3xhawLKCQOg+x9HVgK3Ae8ykdvPLO24bPAGuBe4Gn8+bUZPmvbkbuJMpY+LGsiycWrGuB3wAMRsTnTwqxJJLUnGRZ8DnAUyTDTe4vxefhtTQNPLVXOvJ9aatmR1JfkwuPFwBURcXfGJVkzSNqXJBR+BvzQjyEvbH5q6Y7cTVQg0mfhn0PyLfNhPnoNnxW4NAROI/n8SoAbKcKbltoaP7V0Rz4zyJikq0n+kCwDZgJzI6Im26qsqSTdRfLqxDnAzIhYmnFJ1kR+aumOHAYZk/QhsAJ4N22q/UCKcnhbW5N+ftXp7Mf6nyOiS/6rsqbwU0t35DDImKT+O1vuEQ9me0bOu0TqhzgU4VNLfc0gY/5jb5aZpXx00xnpdBXwZESsyKyqjDgMMubhbWaZ6dxAW3/gnyRdFREz811QltxNlDEPbzMrLJJ6AH+OiKFZ15JPPjPImIe3mRWWiNiYPhusqDgMMtbA8LbTi3l4m1nWJJ1I8o6RouJuoox5eJtZNnKeWpqrB8kzws6PiBfzX1V2HAYZ8/A2s2w0MKw7gA0RUd3Q+ns7dxNlz8PbzDLgYd078juQs9eZ5B26nXOmK4CHJZ2dZWFmVjzcTVSginV4m5llw2cGBSoiNuIXpZhZnjgMClSxDm8zs2z4AnLGdjW8Lf8VmVkx8jWDjHl4m5kVAoeBmZn5moGZmTkMzMwMh4GZmeEwMDMz4P8DoZKnh91jxSoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_completecase.marital_status).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_completecase.marital_status.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Marital Status\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also worth consolidating."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(nooutlier_completecase.religion).value_counts().plot(kind='bar')\n",
    "for index, value in enumerate(nooutlier_completecase.religion.value_counts()):\n",
    "    plt.text(index, value+10, str(value))\n",
    "    plt.title(\"Religion\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Correlations:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import seaborn as sns\n",
    "continuous = nooutlier_completecase[['age', 'LOSdays', 'NumCalloutsByDay', 'NumDiagnosisByDay', 'NumProcsByDay',\n",
    "                              'NumCPTeventsByDay', 'NumInputByDay', 'NumLabsByDay', 'NumMicroLabsByDay',\n",
    "       'NumNotesByDay', 'NumOutputByDay', 'NumRxByDay', 'NumProcEventsByDay', 'NumTransfersByDay',\n",
    "       'NumChartEventsByDay','TotalNumInteractByDay']]\n",
    "a4_dims = (20, 10)\n",
    "corrMatrix=continuous.corr()\n",
    "mask_ut=np.triu(np.ones(corrMatrix.shape)).astype(np.bool)\n",
    "fig, ax = plt.subplots(figsize=a4_dims)\n",
    "sns.heatmap(corrMatrix,mask=mask_ut,annot=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://datascience.stackexchange.com/questions/12554/does-xgboost-handle-multicollinearity-by-itself"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we don't use XGBoost, need to consider the following:\n",
    "\n",
    "+ `NumProcsEventsByDay` and `NumDiagnosisByDay` = 0.8  \n",
    "+ `NumNotesByDay` and `NumDiagnosisByDay` = 0.74  \n",
    "+ `NumTransfersByDay` and `NumDiagnosisByDay` = 0.82  \n",
    "+ `NumLabsByDay` and `NumProcsEventsByDay` = 0.72   \n",
    "+ `NumTransfersByDay` and `NumProcsByDay` = 0.71    \n",
    "+ `NumMicroLabsByDay` and `NumInputByDay` = 0.79   \n",
    "+ `TotalNumInteractByDay` and `NumInputByDay` = 0.94    \n",
    "+ `NumChartEventsByDay` and `NumInputByDay` = 0.94   \n",
    "+ `NumMicroLabsByDay` and `NumLabsByDay` = 0.71  \n",
    "+ `NumNotesByDay` and `NumLabsByDay` = 0.72  \n",
    "+ `TotalNumInteractByDay` and `NumLabsByDay` = 0.7    \n",
    "+ `TotalNumInteractByDay` and `NumNotesByDay` = 0.78  \n",
    "+ `NumChartEventsByDay` and `NumOutputByDay` = 0.77  \n",
    "+ `TotalNumInteractByDay` and `NumOutputByDay` = 0.73  \n",
    "+ `NumChartEventsByDay` and `TotalNumInteractByDay` = 0.98  \n",
    "\n",
    "To drop: `NumOutputByDay` and `TotalNumInteractByDay` and `NumTransfersByDay` and `NumMicroLabsByDay` and `NumDiagnosisByDay` and `NumProcsEventsByDay` and `NumInputByDay`\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Final Dataset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['hadm_id', 'gender', 'age', 'LOSdays', 'admit_type', 'admit_location',\n",
       "       'AdmitDiagnosis', 'insurance', 'religion', 'marital_status',\n",
       "       'ethnicity', 'NumCallouts', 'NumDiagnosis', 'NumProcs',\n",
       "       'AdmitProcedure', 'NumCPTevents', 'NumInput', 'NumLabs', 'NumMicroLabs',\n",
       "       'NumNotes', 'NumOutput', 'NumRx', 'NumProcEvents', 'NumTransfers',\n",
       "       'NumChartEvents', 'ExpiredHospital', 'TotalNumInteract', 'LOSgroupNum',\n",
       "       'NumCalloutsByDay', 'NumDiagnosisByDay', 'NumProcsByDay',\n",
       "       'NumCPTeventsByDay', 'NumInputByDay', 'NumLabsByDay',\n",
       "       'NumMicroLabsByDay', 'NumNotesByDay', 'NumOutputByDay', 'NumRxByDay',\n",
       "       'NumProcEventsByDay', 'NumTransfersByDay', 'NumChartEventsByDay',\n",
       "       'TotalNumInteractByDay', 'log_age', 'log_LOSdays',\n",
       "       'log_NumCalloutsByDay', 'log_NumDiagnosisByDay', 'log_NumProcsByDay',\n",
       "       'log_NumCPTeventsByDay', 'log_NumInputByDay', 'log_NumLabsByDay',\n",
       "       'log_NumMicroLabsByDay', 'log_NumNotesByDay', 'log_NumOutputByDay',\n",
       "       'log_NumRxByDay', 'log_NumProcEventsByDay', 'log_NumTransfersByDay',\n",
       "       'log_NumChartEventsByDay', 'log_TotalNumInteractByDay'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nooutlier_completecase.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Not using AdmitDiagnosis nor AdmitProcedure nor LOSgroupNum nor admit_type (all are EMERGENCY) nor ExpiredHospital (all are 0)\n",
    "# nor Transfers nor Callouts\n",
    "to_dummify_simple = ['insurance', 'religion', 'marital_status','ethnicity']\n",
    "\n",
    "# simple_data = raw_completecase[['gender', 'age', 'LOSdays', 'admit_location', 'admit_type',\n",
    "#         'insurance', 'religion', 'marital_status',\n",
    "#        'ethnicity', 'NumCalloutsByDay','NumCPTeventsByDay', 'NumLabsByDay', \n",
    "#                                 'NumNotesByDay', 'NumRxByDay', 'NumChartEventsByDay', 'ExpiredHospital']]\n",
    "\n",
    "# not using NumCallouts b/c usually only 1 callout, also don't use NumProcEvents b/c already have NumProcs\n",
    "\n",
    "simple_data = nooutlier_completecase[['gender', 'age', 'LOSdays',\n",
    "        'insurance', 'religion', 'marital_status',\n",
    "       'ethnicity', 'NumDiagnosisByDay', 'NumProcsByDay',\n",
    "                              'NumCPTeventsByDay', 'NumInputByDay', 'NumLabsByDay', 'NumMicroLabsByDay',\n",
    "       'NumNotesByDay', 'NumOutputByDay', 'NumRxByDay',\n",
    "       'NumChartEventsByDay','TotalNumInteractByDay']]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "simplecontinuous = simple_data[['age', 'LOSdays','NumDiagnosisByDay', 'NumProcsByDay',\n",
    "                              'NumCPTeventsByDay', 'NumInputByDay', 'NumLabsByDay', 'NumMicroLabsByDay',\n",
    "       'NumNotesByDay', 'NumOutputByDay', 'NumRxByDay', 'NumTransfersByDay',\n",
    "       'NumChartEventsByDay','TotalNumInteractByDay']]\n",
    "a4_dims = (20, 10)\n",
    "corrMatrix=simplecontinuous.corr()\n",
    "mask_ut=np.triu(np.ones(corrMatrix.shape)).astype(np.bool)\n",
    "fig, ax = plt.subplots(figsize=a4_dims)\n",
    "sns.heatmap(corrMatrix,mask=mask_ut,annot=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Need to further remove `NumNotesByDay` and `NumCPTeventsByDay`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "simplecontinuous_corrected = simple_data[['age', 'LOSdays',\n",
    "                                'NumLabsByDay','NumRxByDay', 'NumChartEventsByDay']]\n",
    "a4_dims = (20, 10)\n",
    "corrMatrix=simplecontinuous_corrected.corr()\n",
    "mask_ut=np.triu(np.ones(corrMatrix.shape)).astype(np.bool)\n",
    "fig, ax = plt.subplots(figsize=a4_dims)\n",
    "sns.heatmap(corrMatrix,mask=mask_ut,annot=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "simple_data_corrected = simple_data[['gender', 'age', 'LOSdays', 'admit_location', 'admit_type',\n",
    "        'insurance', 'religion', 'marital_status',\n",
    "       'ethnicity','NumLabsByDay', \n",
    "                                'NumRxByDay', 'NumChartEventsByDay', 'ExpiredHospital']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17790"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(simple_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [],
   "source": [
    "# a = sorted(nooutlier_simple_data.AdmitDiagnosis.unique())\n",
    "# a_df = pd.DataFrame(a)\n",
    "# a_df.to_csv(\"/Users/dereklee/biostat-masters-thesis/admit_diagnoses.csv\")"
   ]
  },
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Example of how diffficult to look at text data: ABSCESS;RULE OUT TUBERCULOSIS![image.png](attachment:image.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4898"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(nooutlier_completecase.AdmitDiagnosis.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# simple_data.loc[nooutlier_simple_data['AdmitDiagnosis'].isin(['02-239','14-AAA']), 'AdmitDiagnosis'] = \"UNKNOWN\"\n",
    "\n",
    "\n",
    "# nooutlier_simple_data.loc[nooutlier_simple_data['AdmitDiagnosis'].isin([' CHEST PAIN',\n",
    "# '2ND DEGREE HEART BLOCK',\n",
    "# '3RD DEGREE HEART BLOCK',\n",
    "# '? AORTIC FILLING DEFECT',\n",
    "# '?? CVA',\n",
    "# 'A- FLUTTER',\n",
    "# 'AAA REPAIR',\n",
    "# 'AAA;TELEMETRY',\n",
    "# 'ABDOMINAL AORTIC ANEURYSM',\n",
    "# 'ABDOMINAL AORTIC ANEURYSM LEAK',\n",
    "# 'ABDOMINAL AORTIC ANEURYSM THROMBOSIS',\n",
    "# 'ABDOMINAL AORTIC ANEURYSM;',\n",
    "# 'ABDOMINAL AORTIC ANEURYSM;? ULCERATIVE COLITIS',\n",
    "# 'ABDOMINAL AORTIC ANEURYSM;TELEMETRY',\n",
    "# 'ABNORMAL AORTIC ANEURYSM',\n",
    "# 'ACUTE COROMARY SYNDROME;PEUMONIA',\n",
    "# 'ACUTE CORONARUY SYNDROME',\n",
    "# 'ACUTE CORONARY DISEASE',\n",
    "# 'ACUTE CORONARY SYNDROME',\n",
    "# 'ACUTE CORONARY SYNDROME;CHEST PAIN',\n",
    "# 'ACUTE CORONARY SYNDROME;CONGESTIVE HEART FAILURE',\n",
    "# 'ACUTE CORONARY SYNDROME;ST ELEVATION MI',\n",
    "# 'ACUTE CORONARY SYNDROME;TELEMETRY',\n",
    "# 'ACUTE CVA;TELEMETRY',\n",
    "# 'ACUTE MI,SEPSIS',\n",
    "# 'ACUTE MYOCARDIAL INFARCTION',\n",
    "# 'ACUTE MYOCARDIAL INFARCTION;PNEUMONIA;CONGESTIVE HEART FAILURE',\n",
    "# 'ACUTE MYOCARDIAL INFARCTION;TELEMETRY',\n",
    "# 'ACUTE MYOCARDIAL INFRACTION',\n",
    "# 'ACUTE PULMONARY EMBOLISM',\n",
    "# 'ACUTE STROKE',\n",
    "# 'ACUTE STROKE;TELEMETRY','ACUTE VASCULAR OCCLUSION','AF/HYPOTENSION',\n",
    "# 'AFIB',\n",
    "# 'AFIB WITH RAPID VENTRICULAR RESPONSE',\n",
    "# 'AFIB WITH RVR',\n",
    "# AFIB-CONGESTIVE HEART FAILURE,\n",
    "# AFIB/AFLUTTER,\n",
    "# AFIB/PNA,\n",
    "# AFLUTTER,\n",
    "# AGITATION AND GIB,\n",
    "# AICD FIRING,\n",
    "# ANEMIA,\n",
    "# ANEMIA BLEED;TELEMETRY,\n",
    "# ANEMIA, PNEUMONIA,\n",
    "# ANEMIA,SHORTNESS OF BREATH,RENAL INSUFFICIENCY,\n",
    "# ANEMIA,THROMBOCYTOPENIA,ACUTE RENAL FAILURE,\n",
    "# ANEMIA-RENAL FAILURE,\n",
    "# ANEMIA-THROMBOCYTOPENIA-COAGULOPATHY,\n",
    "# ANEMIA; UNABLE TO AMBULATE,\n",
    "# ANEMIA;ATRIAL FIBRILLATION;ETOH WITHDRAWAL,\n",
    "# ANEMIA;CONGESTIVE HEART FAILURE,\n",
    "# ANEMIA;FEVER,\n",
    "# ANEMIA;GASTROINTESTINAL BLEED,\n",
    "# ANEMIA;GI BLEED;RESP ARREST,\n",
    "# ANEMIA;METHICILLIN RESISTANT STAPH AUREUS,\n",
    "# ANEMIA;RULE OUT MYOCARDIAL INFARCTION,\n",
    "# ANEMIA;SVT,\n",
    "# ANEMIA;TELEMETRY,\n",
    "# ANEURYSM,\n",
    "# ANEURYSM;TELEMETRY,\n",
    "# ANGINA,\n",
    "# ANGIOEDEMA,\n",
    "# AORTIC ANEURYSM,\n",
    "# AORTIC DESSCTION,\n",
    "# AORTIC DILATATION,\n",
    "# AORTIC DISECTION,\n",
    "# AORTIC DISSECT,\n",
    "# AORTIC DISSECTION,\n",
    "# AORTIC DISSECTION;TELEMETRY,\n",
    "# AORTIC HEMATOMA,\n",
    "# AORTIC INJURY,\n",
    "# AORTIC STENOSIS,\n",
    "# AORTIC ULCERATION,\n",
    "# AORTIC VALVE ABSCESS,\n",
    "# AORTIC VALVE MALFUNCTIONING,\n",
    "# 'A,FLUTTER W,RVR,? NEW BRAIN MASS',\n",
    "# ARRHYTHMIA\n",
    "# ARRYTHMIA\n",
    "# ARTERIAL INSUFFICIENCY\n",
    "# ARTERIAL OCCLUSION\n",
    "# ARTERIAL THROMBOSIS;TELEMETRY\n",
    "# ARTERIOVEOUS MALFORMATION,\n",
    "# ATERIAL INSUFFICIENCY,\n",
    "# ATRIAL FIBRILLATION,\n",
    "# ATRIAL FIBRILLATION W/ RVR,\n",
    "# ATRIAL FIBRILLATION,PNEUMONIA,\n",
    "# ATRIAL FIBRILLATION; RVR,\n",
    "# ATRIAL FIBRILLATION;(+) TRAPONIN,\n",
    "# ATRIAL FIBRILLATION;CHRONIC OBST PULM DISEASE,\n",
    "# ATRIAL FIBRILLATION;DYSPNEA,\n",
    "# ATRIAL FIBRILLATION;PERICARDIAL EFFUSION,\n",
    "# ATRIAL FIBRILLATION;TELEMETRY,\n",
    "# ATRIAL FIBRILLATION;TELEMETRY;HYPOTENSION,\n",
    "# ATRIAL FIBRILLATION;TELEMETRY;RVR;TELEMETRY,\n",
    "# ATRIAL FIBRILLATION;TELEMETRY;TELEMETRY,\n",
    "# ATRIAL FIBRILLATION;TELEMETRY;URINARY TRACT INFECTION,\n",
    "# ATRIAL FLUTTER;TELEMTRY,\n",
    "# 'ATYPICAL CHEST PAIN',BISCUPID VALVE;DILATED AORTIC ROOT,\n",
    "# 'BIVENTRICULAR HEART FAILURE','BLEEDING AV FISTULA','BLEEDING FISTUA',BRADYCARDIA,\n",
    "# BRADYCARDIA SHOCK,\n",
    "# BRADYCARDIA-ACUTE RENAL FAILURE,\n",
    "# BRADYCARDIA-NAUSEA-VOMITINH,\n",
    "# BRADYCARDIA-RENAL FAILURE-HYPERKALEMIA,\n",
    "# BRADYCARDIA-SYNCOPE,\n",
    "# BRADYCARDIA;ACUTE RENAL FAILURE,\n",
    "# BRADYCARDIA;CARDIAC ARREST,\n",
    "# BRADYCARDIA;HYPOTENSION;SEPSIS,\n",
    "# BRADYCARDIA;RENAL FAILURE;HYPERKALEMIA,\n",
    "# 'BRADYCARDIA;SYNCOPE','BRUGADA SYNDROME;SYNCOPE;TELEMETRY;DEHYDRATION',CARDIAC  ARREST,\n",
    "# CARDIAC ARREST,\n",
    "# CARDIAC ARREST;GASTROINTESTINAL BLEED,\n",
    "# CARDIAC ARREST;TELEMETRY,\n",
    "# CARDIAC ARREST;URINARY TRACT INFECTION,\n",
    "# CARDIAC CATH,\n",
    "# CARDIAC CATH S/P CARDIAC ARREST,\n",
    "# CARDIAC TAMPIONADE,\n",
    "# CARDIAC TAMPONADE,\n",
    "# CARDIOGENIC SHOCK,\n",
    "# CARDIOMELOPATHY;CONGESTIVE HEART FAILURE,\n",
    "# CARDIOMYOPATHY,\n",
    "# CAROTID INJURIES,\n",
    "# 'CATHETER INFECTION','CHF EXACERBATION',\n",
    "# 'CHF, PNEUMONIA',\n",
    "# 'CHF, R/O MI, ARF',\n",
    "# 'CHF,ACUTE CORONARY SYNDROME',\n",
    "# 'CHF,ARF',\n",
    "# 'CHF,PNEUMONIA',\n",
    "# 'CHF,R/O MI',\n",
    "# 'CHF/HYPONATREMIA','CHR;AFIB WITH RVR',CONGESTIVE HEART DISEASE,\n",
    "# 'CONGESTIVE HEART FAILUE, PNEUMONIA',\n",
    "# CONGESTIVE HEART FAILURE,\n",
    "# CONGESTIVE HEART FAILURE EXACERBATION,\n",
    "# CONGESTIVE HEART FAILURE EXACERBATION-R/O MI,\n",
    "# 'CONGESTIVE HEART FAILURE,ACUTE RENAL FAILOURE',\n",
    "# 'CONGESTIVE HEART FAILURE,COPD FLARE',\n",
    "# 'CONGESTIVE HEART FAILURE,ENDOCARDITIS',\n",
    "# 'CONGESTIVE HEART FAILURE,REPIRATORY FAILURE',\n",
    "# 'CONGESTIVE HEART FAILURE,SEPSIS',\n",
    "# 'CONGESTIVE HEART FAILURE,SHORTNESS OF BREATH',\n",
    "# CONGESTIVE HEART FAILURE-PNEUMONIA,\n",
    "# CONGESTIVE HEART FAILURE-R/O MYOCARDIAL INFARCTION,\n",
    "# CONGESTIVE HEART FAILURE-SUBDURAL HEMATOMA,\n",
    "# CONGESTIVE HEART FAILURE/PNEUMONIA,\n",
    "# CONGESTIVE HEART FAILURE/V TACH,\n",
    "# CONGESTIVE HEART FAILURE; HYPERCARBIC RESP FAILURE,\n",
    "# CONGESTIVE HEART FAILURE; SHORTNESS OF BREATH,\n",
    "# CONGESTIVE HEART FAILURE;A FIB VS ATRIAL TACHUCARDIA,\n",
    "# CONGESTIVE HEART FAILURE;A-FIB,\n",
    "# CONGESTIVE HEART FAILURE;ABDOMINAL PAIN,\n",
    "# CONGESTIVE HEART FAILURE;ACUTE RENAL FAILURE,\n",
    "# CONGESTIVE HEART FAILURE;APPENDICITIS,\n",
    "# CONGESTIVE HEART FAILURE;CHEST PAIN,\n",
    "# CONGESTIVE HEART FAILURE;CHRONIC OBST PULM DISEASE,\n",
    "# CONGESTIVE HEART FAILURE;COPD EXACERBATION,\n",
    "# CONGESTIVE HEART FAILURE;DELTA MS,\n",
    "# CONGESTIVE HEART FAILURE;EXACERBATION,\n",
    "# CONGESTIVE HEART FAILURE;HIP FRACTURE,\n",
    "# CONGESTIVE HEART FAILURE;HYPERKALEMIA;ALTERED MENTAL STATUS;TELEMETRY,\n",
    "# CONGESTIVE HEART FAILURE;HYPOTENSION,\n",
    "# CONGESTIVE HEART FAILURE;METHICILLIN RESISTANT STAPH AUREUS,\n",
    "# CONGESTIVE HEART FAILURE;NON-ST SEGMENT ELEVATION MYOCARDIAL INFARCTION,\n",
    "# CONGESTIVE HEART FAILURE;PNEUMONIA,\n",
    "# CONGESTIVE HEART FAILURE;PNEUMONIA;RULE OUT MYOCARDIAL INFARCTION,\n",
    "# CONGESTIVE HEART FAILURE;RENAL FAILURE,\n",
    "# CONGESTIVE HEART FAILURE;RESPIRATORY DISTRESS,\n",
    "# CONGESTIVE HEART FAILURE;RESPIRATORY FAILURE,\n",
    "# CONGESTIVE HEART FAILURE;RULE-OUT MYOCARDIAL INFARCTION,\n",
    "# CONGESTIVE HEART FAILURE;RULE-OUT MYOCARDIAL INFARCTION;TELEMETRY,\n",
    "# CONGESTIVE HEART FAILURE;S/P RENAL TRANSPLANT;HYPERKALEMIA,\n",
    "# CONGESTIVE HEART FAILURE;SEPSIS,\n",
    "# CONGESTIVE HEART FAILURE;SHORTNESS OF BREATH,\n",
    "# CONGESTIVE HEART FAILURE;TELEMETRY,\n",
    "# CONGESTIVE HEART FAILURE;URINARY TRACT INFECTION,\n",
    "# 'CONGESTIVE HEART FAILURE;URINARY TRACT INFECTION;PNEUMONIA',CORONARY ARTERY DISEASE,\n",
    "# CORONARY ARTERY DISEASE;SOB;TELEMETRY,\n",
    "# 'CORONARY ARTERY DISEASE\\CORONARY ARTERY BYPASS GRAFT WITH AVR ','CVA, BRADYCARDIA',\n",
    "# 'CVA;HYPOTENSION',DEEP VEIN THROMBOSIS,\n",
    "# DEEP VEIN THROMBOSIS;HEMOCULT POSITIVE,\n",
    "# \"DEEP VEIN THROMBOSIS;PULMONARY EMBOLIS\",\"DILATED CARDIOMYOPATHY\",'DVT, ANEMIA','FEMORAL ARTERY THROMBOSIS',\n",
    "# HEART BLOCK,\n",
    "# HEART BLOCK,NON STE MI,\n",
    "# HEART FAILURE,\n",
    "# HEART FAILURE,CHOLECYSTITIS,\n",
    "# 'HEART MASS','HEMORRAGIC STROKE;TELEMETRY',HYPERTENBSIVE EMERGENCY,\n",
    "# HYPERTENSION,\n",
    "# HYPERTENSION,CONGESTIVE HEART FAILURE,\n",
    "# HYPERTENSION-ACUTE RENAL FAILURE,\n",
    "# HYPERTENSION/VOMITING/HYPERGLYCEMIA,\n",
    "# HYPERTENSION;RENAL ARTERY ARTHEROSCLEROSIS\\AORTIC FENESTRATION & ANGIOGRAM W/BILATERAL RENAL STENT,\n",
    "# HYPERTENSION;SEPSIS,\n",
    "# HYPERTENSIVE,\n",
    "# HYPERTENSIVE CRISIS,\n",
    "# HYPERTENSIVE EMERGENCY,\n",
    "# HYPERTENSIVE EMERGENCY;TELEMETRY,\n",
    "# HYPERTENSIVE HEMORRHAGE,\n",
    "# HYPERTENSIVE URENCY,\n",
    "# HYPERTENSIVE URGENCY,\n",
    "# 'HYPERTENSIVE URGENCY-R/O MYOCARDIAL INFARCTION','HYPOTENSION',\n",
    "# HYPOTENSION ABDOMINAL PAIN,\n",
    "# HYPOTENSION FEVER,\n",
    "# HYPOTENSION WITH HEMODIALYSIS,\n",
    "# HYPOTENSION, PNEUMONIA, URINARY TRACT INFECTION,\n",
    "# HYPOTENSION, RENAL FAILURE,\n",
    "# HYPOTENSION,BRADYCARDIA,\n",
    "# HYPOTENSION,DIARRHEA,\n",
    "# HYPOTENSION,MENTAL STATUS CHANGES,\n",
    "# HYPOTENSION,SEPSIS,\n",
    "# HYPOTENSION,UREMIA,\n",
    "# HYPOTENSION-ANEMIA,\n",
    "# HYPOTENSION-BLEEDING CATH SITE,\n",
    "# HYPOTENSION-BRADYCARDIA,\n",
    "# HYPOTENSION.GASTRIC ILEAL BYPASS,\n",
    "# HYPOTENSION; BRADYCARDIA,\n",
    "# HYPOTENSION;AAA REPAIR,\n",
    "# HYPOTENSION;ABDOMINAL PAIN,\n",
    "# HYPOTENSION;ACUTE CORONARY SYNDROME,\n",
    "# HYPOTENSION;ACUTE RENAL FAILURE,\n",
    "# HYPOTENSION;ATRIAL FIBRILLATION;TELEMETRY,\n",
    "# HYPOTENSION;BRADYCARDIA,\n",
    "# HYPOTENSION;BRADYCARDIA;TELEMETRY,\n",
    "# HYPOTENSION;CHEST PAIN,\n",
    "# HYPOTENSION;DIARRHEA,\n",
    "# HYPOTENSION;DISSEMINATED ZOSTER,\n",
    "# HYPOTENSION;ETOH WITHDRAWEL,\n",
    "# HYPOTENSION;FEMUR FX,\n",
    "# HYPOTENSION;FEVER,\n",
    "# HYPOTENSION;FEVER,ETIOLOGY UNKNOWN,\n",
    "# HYPOTENSION;HYPERGLYCEMIA;HYPERKALEMIA,\n",
    "# HYPOTENSION;HYPOGLYCEMIA;SEIZURE,\n",
    "# HYPOTENSION;HYPOTHERMIA,\n",
    "# HYPOTENSION;HYPOTHERMIA;HYPOXIA;MENTAL STATUS CHANGES;BRADYCARDIA,\n",
    "# HYPOTENSION;INFECTION,\n",
    "# HYPOTENSION;INSULIN DEPENDENT DIABETES MELLITUS;ESRD,\n",
    "# HYPOTENSION;PRE-SYNCOPE,\n",
    "# HYPOTENSION;RENAL FAILURE,\n",
    "# HYPOTENSION;RESPIRATORY DISTRESS,\n",
    "# HYPOTENSION;RULE OUT SEPSIS,\n",
    "# HYPOTENSION;SEPSIS,\n",
    "# HYPOTENSION;SEPSIS;TELEMETRY,\n",
    "# HYPOTENSION;SHORTNESS OF BREATH,\n",
    "# HYPOTENSION;SUBSTANCE ABUSE;PANCREATITIS,\n",
    "# HYPOTENSION;TELEMETRY,\n",
    "# HYPOTENSION;THROMOCYTOPENIA,\n",
    "# HYPOTENSION;URINARY TRACT INFECTION,\n",
    "# 'HYPOTENSION;UROSEPSIS;ACUTE RENAL FAILURE,'\n",
    "# 'HYPOTENSIVE;SEPSIS','ICD FIRING',\n",
    "# 'ICD FIRING;SOB','ILIAC ANEURYSM','INFERIOR MYOCARDIAL INFARCTION','INTRAVENTRICULAR BLEED',\n",
    "# 'INTRAVENTRICULAR HEMORRHAGE','ISCHEMIC BOWEL','ISCHEMIC COLITIS','ISCHEMIC FOOT',\n",
    "# 'ISCHEMIC FOOT;ABDOMINAL PAIN;TELEMETRY',\n",
    "# 'ISCHEMIC LEG',\n",
    "# 'ISCHEMIC LEGS','LARGE PERICARDIAL EFFUSION,MEDIASTINAL MASS','LEAKING ANEURYSM',\n",
    "# 'LEAKING THORACIC ANEURYSM','LEFT BKA ISCHEMIA','LEFT BUNDLE BRANCH BLOCK/BRADYCARDIA',\n",
    "# 'LEFT IJ CLOT','LEFT LEG HEMATOMA-CELLULITIS','LEFT LOWER EXTREMITY POPITEAL ANEURYSM',\n",
    "# 'LEFT SIDED HEMATOMA;STATUS POST MOTOR VEHICLE ACCIDENT WITH INJURIES',\n",
    "# 'LEFT SUBDURAL HEMATOMA','LEFT THALMIC BLEED','LT ATRIAL THROMBOSIS','MALIGNANT HYPERTENSION',\n",
    "# 'MCA SROKE','MASSIVE PULMONARY EMBOLUS',    \n",
    "# ]), 'AdmitDiagnosis'] = \"CARDIOVASCULAR\"\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# nooutlier_simple_data.loc[nooutlier_simple_data['AdmitDiagnosis'].isin(['? ISCHEMIC BOWEL',\n",
    "#                                                                        '? PERITONITIS',\n",
    "#                                                                        '?PANCREATITIS',\n",
    "#                                                                        'ABD PAIN',\n",
    "#                                                                         'ABD PAIN, LIKELY LGIB',\n",
    "#                                                                         'ABD TRAUMA,FREE FLUID IN ABD',\n",
    "#                                                                         'ABDOMINAL ABSCESS',\n",
    "#                                                                        'ABDOMINAL DISTENSION',\n",
    "#                                                                         'ABDOMINAL INFECTION',\n",
    "#                                                                         'ABDOMINAL MASS',\n",
    "#                                                                         'ABDOMINAL METS',\n",
    "#                                                                         'ABDOMINAL PAIN',\n",
    "#                                                                         'ABDOMINAL PAIN,R/O PANCREATIC CANCER IN COMMON BILE DUCT',\n",
    "#                                                                         'ABDOMINAL PAIN-DIARRHEA',\n",
    "#                                                                         'ABDOMINAL PAIN-ELECTROLYTE ABNORMALITIES',\n",
    "#                                                                         'ABDOMINAL PAIN-ETOH WITHDRAWAL',\n",
    "#                                                                         'ABDOMINAL PAIN-FEVER',\n",
    "#                                                                         'ABDOMINAL PAIN-RENAL INSUFFICIENCY',\n",
    "#                                                                         'ABDOMINAL PAIN/ CIRRHOSIS',\n",
    "#                                                                         'ABDOMINAL PAIN/?CHOLANGITIS',\n",
    "#                                                                         'ABDOMINAL PAIN;AORTIC DISSECTION;TELEMETRY',\n",
    "#                                                                         'ABDOMINAL PAIN;ASCITES;TAHYCARDIA',\n",
    "#                                                                         'ABDOMINAL PAIN;CELLULITIS',\n",
    "#                                                                         'ABDOMINAL PAIN;EVALUATION FOR LYMPHOMA',\n",
    "#                                                                         'ABDOMINAL PAIN;EXPLORATORY LAPAROSCOPY;BILE DUCT OBSTRUCTION',\n",
    "#                                                                         'ABDOMINAL PAIN;FEVER',\n",
    "#                                                                         'ABDOMINAL PAIN;INABILITY TO TOLERATE PO',\n",
    "#                                                                         'ABDOMINAL PAIN;MALFUCNTION STENT GRAFT',\n",
    "#                                                                         'ABDOMINAL PAIN;NAUSEA;VOMITING',\n",
    "#                                                                         'ABDOMINAL PAIN;PERFORATION',\n",
    "#                                                                         'ABDOMINAL PAIN;SHORTNESS OF BREATH;ELEVATED LIVER FUNCTION TEST',\n",
    "#                                                                         'ABDOMINAL PAIN;STENT PLACEMENT',\n",
    "#                                                                         'ABDOMINAL PAIN;TELEMETRY',\n",
    "#                                                                         'ABDOMINAL PERFORATION',\n",
    "#                                                                         'ABDOMINAL SEPSIS',\n",
    "#                                                                         'ABDOMINAL TRAUMA',\n",
    "#                                                                         'ABDOMINAL WALL HEMATOMA',\n",
    "#                                                                         'ABDOMONAL PAIN',\n",
    "#                                                                         'ABSCESS;RULE OUT TUBERCULOSIS',\n",
    "#                                                                         'ACUTE CHOLECYSTITIS',\n",
    "#                                                                         'ACUTE CHOLECYSTITIS/ACS',\n",
    "#                                                                         'ACUTE CHOLEYSTITIS,LLL PNEUMONIA,URINARY TRACT INFECTION',\n",
    "#                                                                        'ACUTE LIVER FAILURE',\n",
    "#                                                                        'ANURIC RENAL FAILURE; GI BLEED',\n",
    "#                                                                        'APPENDICITIS',\n",
    "#                                                                        'ASCENDING CHOLANGITIS',\n",
    "#                                                                         'ASCENDING CHOLANGITIS & PANCREATIC MASS',\n",
    "#                                                                        'ASCITES',\n",
    "# 'ASCITES, HYPOGLYCEMIA',\n",
    "# 'ASCITES,ABDOMINAL PAIN',\n",
    "# 'ASCITES;HYPERKALEMIA',\n",
    "# 'ASCITES;PLEURAL EFFUSION',\n",
    "# 'ASCITES;SHORTNESS OF BREATH',\n",
    "# 'ASCITIS','BILE LEAK FROM BILIARY DRAIN',\n",
    "# 'BILE OBSTRUCTION;TELEMETRY',\n",
    "# 'BILIARY DUCT DILATION',\n",
    "# 'BILIARY OBSTRUCTION',\n",
    "# 'BILIARY STENT OBSTRUCTION','BOWEL ISCHEMIA',\n",
    "# 'BOWEL OBSTRUCTION',\n",
    "# 'BOWEL OBSTRUCTION;TELEMETRY',\n",
    "# 'BOWEL OBTRUCTION',\n",
    "# 'BOWEL PERFORATION',\n",
    "# \"BOWEL PERFORATION;SEPSIS\",\"BRBPR\"\n",
    "# \"BRBPR, DIARRHEA\",\n",
    "# \"BRBPR/BLOOD LOSS ANEMIA\",'BRIGHT RED BLOOD BANITING/HEMIPTHSIS',\n",
    "# 'BRIGHT RED BLOOD PER RECTUM',\n",
    "# 'BRIGHT RED BLOOD PER RECTUM;METHICILLIN RESISTANT STAPH AUREUS',\n",
    "# \"BRIGHT RED BLOOD PER RECTUM;URINARY TRACT INFECTION\",'CCY','CDIFF',\n",
    "# 'CECAL VOLVULUS','CHOLANGIOCARCINOMA,DEHYDRATION',\n",
    "# 'CHOLANGIOLIOLITIS',\n",
    "# 'CHOLANGITIS',\n",
    "# 'CHOLANGITIS ABSCESS',\n",
    "# 'CHOLANGITIS;CHOLECYSTITIS',\n",
    "# 'CHOLANGITIS;SEPSIS',\n",
    "# 'CHOLANGITIS;TELEMETRY',\n",
    "# 'CHOLANGITIS\\ERCP',\n",
    "# 'CHOLANIOCARCINOMA - PAIN CONTROL',\n",
    "# 'CHOLECITHIASIS',\n",
    "# 'CHOLECYSITIS',\n",
    "# 'CHOLECYSITITS',\n",
    "# 'CHOLECYSTITIS',\n",
    "# 'CHOLEDOCHOLITHIADSIS',\n",
    "# 'CHOLEDOCHOLITHIASIS',\n",
    "# 'CHOLEYCYSTITIS',\n",
    "# 'CHOLYCYSTITIS','CHRONIC PANCREATITIS','CIRRHOSIS',\n",
    "# 'CIRRHOSIS;VOMITING',\n",
    "# 'CLAUDICATION','CLOGGED G-TUBE',\n",
    "# 'CLOSE LOOP OBSTRUCTON','COLITIS',\n",
    "# 'COLITIS-GASTROENTERITIS',\n",
    "# 'COLON CANCER',\n",
    "# 'COMMON BILE DUCT DILATION',\n",
    "# 'COMMON BILE DUCT OBSTRUCTION;PNEUMONIA',\n",
    "# 'COMMON BILE DUCT STONE',\n",
    "# 'COMMON BILE DUCT STONES','CONSTIPATION','DECOMPRESSED CIRRHOSIS;HYPONATREMIA',\n",
    "# 'DECONGESTINAL LIVER FAILURE',\n",
    "# 'DECUBITIS ULCER',\n",
    "# 'DECUBITUS ULCER',\"DIAPHRAGMATIC HERIA\",\"DIVERTICUL;ITIS\",\n",
    "# \"DIVERTICULITIS\",\n",
    "# \"DIVERTICULITIS;FEVER\",\"DUODENAL PERFORATION\",'G-TUBE FELL OUT',\n",
    "# 'G.I. BLEED,GALLBLADDER MASS',\n",
    "# 'GALLSTONE',\n",
    "# 'GALLSTONE PANCREATITIS',\n",
    "# 'GALLSTONE PANCREATITIS;CHOLECYSTITIS',\n",
    "# 'GALLSTONE,PANCREATITIS',\n",
    "# 'GALLSTONE;PANCREATITIS',\n",
    "# 'GALLSTONES',\n",
    "# 'GALLSTONES/PANCRATITIS','GASSTROINTESTINAL BLEED',\n",
    "# 'GASTORINTESTINAL BLEED',\n",
    "# 'GASTRIC OUTLET OBSTRUCTION',\n",
    "# 'GASTROENTERITIS',\n",
    "# 'GASTROENTERITIS,CHEST PAIN',\n",
    "# 'GASTROINTESINAL BLEED',\n",
    "# 'GASTROINTESTINAL BLEED',\n",
    "# 'GASTROINTESTINAL BLEED-ANEMIA',\n",
    "# 'GASTROINTESTINAL BLEED-R/O MYOCARDIAL INFARCTION',\n",
    "# 'GASTROINTESTINAL BLEED-RLL PNEUMONIA-R/O MI',\n",
    "# 'GASTROINTESTINAL BLEED/JAUNDICE',\n",
    "# 'GASTROINTESTINAL BLEED; R/O MI',\n",
    "# 'GASTROINTESTINAL BLEED;ALTERED MENTAL STATUS',\n",
    "# 'GASTROINTESTINAL BLEED;CHEST PAIN;TELEMETRY;SHORTNESS OF BREATH',\n",
    "# 'GASTROINTESTINAL BLEED;CHRONIC OBST PULM DISEASE',\n",
    "# 'GASTROINTESTINAL BLEED;CIRRHOSIS',\n",
    "# 'GASTROINTESTINAL BLEED;COPD EXACERBATION',\n",
    "# 'GASTROINTESTINAL BLEED;DIABETIC KETOACIDOSIS;S/P CARDIAC ARREST',\n",
    "# 'GASTROINTESTINAL BLEED;HYPERKALEMIA',\n",
    "# 'GASTROINTESTINAL BLEED;HYPERTENSION;RENAL FAILURE',\n",
    "# 'GASTROINTESTINAL BLEED;HYPOTENSION',\n",
    "# 'GASTROINTESTINAL BLEED;RULE-OUT MYOCARDIAL INFARCTION;TELEMETRY',\n",
    "# 'GASTROINTESTINAL BLEED;TELEMETRY',\n",
    "# 'GASTROINTESTINAL INFECTION',\n",
    "# 'GASTROINTESTINAL PERFORATION',\n",
    "# 'GASTROINTESTIONAL BLEED',\n",
    "# 'GASTROINTETINAL BLEED',\n",
    "# 'GASTROPARESIS',\n",
    "# 'GASTROPARESIS,HYPERTENSION'\n",
    "# 'GASTROPASESIS',\n",
    "# 'GI BLEED',\n",
    "# 'GI BLEED,ABDOMINAL PAIN',\n",
    "# 'GI BLEED,CHEST PAIN,HYPOTENSION',\n",
    "# 'GI BLEED,DIABETIC KETOACIDOSIS',\n",
    "# 'GI BLEED/ CHEST PAIN',\n",
    "# 'GI BLEED/NSTEMI',\n",
    "# 'GI BLEED;ASPIRATION PNEUMONIA;URINARY TRACT INFECTION',\n",
    "# 'GI BLEED;TELEMETRY',\n",
    "# 'GI BLEEDING','HEPARIN INFUSIN, S/P IVC FILTER PLACEMENT',\n",
    "# 'HEPATIC BLEED',\n",
    "# 'HEPATIC ENCEPHALOPATHY',\n",
    "# 'HEPATIC FAILURE',\n",
    "# 'HEPATIC INSUFFIENCY',\n",
    "# 'HEPATITIS',\n",
    "# \"HEPATITIS, ACIDOSIS, ETOH WITHDRAWAL, ELEVATED ACETAMINOPHEN LEVEL\",\n",
    "# \"HEPATORENAL SYNDROME HEMETEMSIS\",'ICSHEMIC BOWEL','ILEUS','INTESTINAL LYMPHADENOPATHY',\n",
    "# 'INTRA-ABDOMINAL ABSCESS',\n",
    "# 'INTRAABDOMINAL BLEEDING',\n",
    "# 'INTRABDOMINAL HEMORRHAGE','INTUSSUCEPTION','J-TUBE FAILURE',\n",
    "# 'JAUNDICE','JUNCTIONAL RHYTHM-GASTROINTESTINAL BLEED','LARGE BOWEL OBSTRUCTION',\n",
    "# 'LIVER ABSCESS',\n",
    "# 'LIVER BLEED & HEMATOMA, HEMOPERITON',\n",
    "# 'LIVER DISEASE',\n",
    "# 'LIVER FAILURE',\n",
    "# 'LIVER FAILURE;ACUTE RENAL FAILURE;ALTERED MENTAL STATUS',\n",
    "# 'LIVER FAILURE;SUBDURAL HEMATOMA',\n",
    "# 'LIVER FAILURE;TELEMETRY',\n",
    "# 'LIVER LACERATION',\n",
    "# 'LIVER LACERATION;S/P MOTOR VEHICLE ACCIDEMT',\n",
    "# 'LIVER MASS',\n",
    "# 'LIVER TRANSPLANT','LOWER GASTRINTESTINAL BLEED',\n",
    "# 'LOWER GASTRO INTESTINAL BLEED',\n",
    "# 'LOWER GASTROINTESTINAL BLEED',\n",
    "# 'LOWER GASTROINTESTINAL BLEED;A FIB',\n",
    "# 'LOWER GASTROINTESTINAL BLEED;ANEMIA',\n",
    "# 'LOWER GI  BLEED',\n",
    "# 'LOWER GI BLEED',\n",
    "# 'LOWER GI BLEED;TELEMETRY',\n",
    "# 'LOWER GI BLEEDING',\n",
    "# 'LOWER GI BLEEED',\n",
    "# 'LOWER GIB',\n",
    "# 'LOWER GIB;ATRIAL FIBRILLATION;TELEMETRY',\n",
    "# 'LOWER GSTROINTESTINAL BLEED','MESENTERIC BLEEDING',\n",
    "# 'MESENTERIC ISCHEMIA',\n",
    "# 'MESENTERIC VASCULAR INJURY',\n",
    "# 'MESTERIC ISCHEMIA;ABDOMINAL PAIN',\n",
    "# ]), 'AdmitDiagnosis'] = \"GI\"\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# nooutlier_simple_data.loc[nooutlier_simple_data['AdmitDiagnosis'].isin([' HYPOXIA','ACUTE RESPIRATORY FAILURE/AMR',\n",
    "# 'ACUTE SHORTNESS OF BREATH;CONFUSION','ADULT RESPIRATORY DISTRESS SYNDROME','AIRWAY MASS',\n",
    "# 'AIRWAY OBSTRUCTION','ARDS VS SEPSIS',ASPIRATION FOREIGN BODY,\n",
    "# ASPIRATION PNA,\n",
    "# ASPIRATION PNEUMONIA,\n",
    "# ASPIRATION PNEUMONIA,SEPSIS,\n",
    "# ASPIRATION/RESPIRATORY FAILURE,ASTHMA,\n",
    "# ASTHMA & COPD EXACERBATION,\n",
    "# ASTHMA /COPD EXACERBATION,\n",
    "# ASTHMA EXACERBATION,\n",
    "# ASTHMA EXACERBATION;COPD EXACERBATION,\n",
    "# ASTHMA EXACERBATION;TELEMETRY,\n",
    "# ASTHMA FLARE;PNEUMONIA,\n",
    "# ASTHMA,COPD EXACERBATION,\n",
    "# ASTHMA-COPD EXACERBATION,\n",
    "# ASTHMA/COPD EXACERBATION,\n",
    "# ASTHMA; COPD EXACERBATION,\n",
    "# ASTHMA;CHRONIC OBST PULM DISEASE,\n",
    "# ASTHMA;COPD EXACERBATION,\n",
    "# ASTHMA;COPD EXACERBATION;TELEMETRY,\n",
    "# ASTHMA;COPDEXACERBATION,\n",
    "# 'ASTHMA;TROP LEAK',BILATERAL PLEURAL EFFUSIONS;PNEUMONIA,\n",
    "# BILATERAL PULMONARY EMBOLI,\n",
    "# 'BILATERAL PULMONARY EMBOLIS','BIPOLAR PHEUMONIA','BRONCHITIS',\n",
    "# 'BRONCHOSPASM',CHRONIC OBST PULM DISEASE,\n",
    "# CHRONIC OBST PULM DISEASE;CONGESTIVE HEART FAILURE,\n",
    "# CHRONIC OBST PULM DISEASE;PNEUMONIA;METHICILLIN RESISTANT STAPH AUREUS;VANCOMYCIN RESISTANT ENTEROCOCCUS,\n",
    "# CHRONIC OBST PULM DISEASE;TELEMETRY,\n",
    "# CHRONIC OBSTRUCTIVE PULMONARY DISEASE,\n",
    "# CHRONIC OBSTRUCTIVE PULMONARY DISEASE EXACERBATION,\n",
    "# CHRONIC OBSTRUCTIVE PULMONARY DISEASE-EXACERBATION,\n",
    "# CHRONIC OBSTRUCTIVE PULMONARY DISEASE;CHOLANGITIS,\n",
    "# \"CHRONIC OBSTRUCTIVE PULMONARY DISEASE;PNEUMONIA\",COPD EXACERBATION,\n",
    "# COPD EXACERBATION,PNEUMONIA,\n",
    "# COPD EXACERBATION; PNEUMONIA,\n",
    "# COPD EXACERBATION;RIB FRACTURE,\n",
    "# COPD EXACERBATION;TELEMETRY,\n",
    "# COPD FLARE,\n",
    "# COPD FLARE-HYPOTENSION,\n",
    "# COPD FLARE-RESPIRATORY DISTRESS,\n",
    "# COPD FLARE;SHORTNESS OF BREATH;RULE OUT MYOCARDIAL INFARCTION,\n",
    "# 'COPD, PNEUMONIA, ACUTE CORONARY SYNDROME','CSF RHINORRHEA',DYSNEA,\n",
    "# DYSPHAGIA,\n",
    "# DYSPNEA,\n",
    "# DYSPNEA ON EXERTION-CHROMIC RENAL INSUFFICIENCY,\n",
    "# DYSPNEA, RESPIRATORY DISTRESS,\n",
    "# \"DYSPNEA,FEVER\",\n",
    "# DYSPNEA,HYPOTENSION,\n",
    "# \"DYSPNEA,HYPOXIA\",\n",
    "# DYSPNEA-COPD EXACERBATION,\n",
    "# DYSPNEA;ACUTE RENAL FAILURE,\n",
    "# DYSPNEA;COPD EXACERBATION,\n",
    "# DYSPNEA;HYPOTENSION,\n",
    "# DYSPNEA;HYPOXIA,\n",
    "# DYSPNEA;NAUSEA W\\VOMITING;TELEMETRY,\n",
    "# DYSPNEA;PULMONARY EDEMA,\n",
    "# DYSPNEA;R/O FLU,\n",
    "# DYSPNEA;RULE-OUT MYOCARDIAL INFARCTION;TELEMETRY,\n",
    "# DYSPNEA;TACHYCARDIA,\n",
    "# DYSPNEA;TELEMETRY,\n",
    "# \"DYSPNEA;TRACHIAL STENOSIS\",'EMPHYSEMA',\n",
    "# 'EMPYEMA','EPIGLOTITIS',\n",
    "# 'EPIGLOTTITIS',\n",
    "# \"EPIGLOTTITIS;FEVER\",'HYPERCARBIA;PNEUMONIA',\n",
    "# 'HYPERCARBIC RESPIRATORY FAILURE',\n",
    "# 'HYPERCARBIC RESPIRATORY FAILURE-PULSELESS ELECTRICAL','HYPOVENTILATION','HYPOXIA',\n",
    "# HYPOXIA RESPIRATORY FAILURE,\n",
    "# 'HYPOXIA, SEIZURES',\n",
    "# HYPOXIA,ACUTE CORONARY SYNDROME,HYPOTENSION,ACUTE RENAL FAILURE,\n",
    "# HYPOXIA,PNEUMONIA,\n",
    "# HYPOXIA,SHOCK,\n",
    "# HYPOXIA-INTRACRANIAL HEMORRHAGE,\n",
    "# HYPOXIA-MULTIPLE SCLEROSIS,\n",
    "# HYPOXIA-PANCYTOPENIA,\n",
    "# HYPOXIA/HYPOTENSION/NAUSEA/VOMITING,\n",
    "# HYPOXIA;ACUTE RENAL FAILURE;TELEMETRY,\n",
    "# HYPOXIA;ASPIRATION PNEUMONIA;ATRIAL FIBRILLATION,\n",
    "# HYPOXIA;BENZO OVERDOSE,\n",
    "# HYPOXIA;CHANGE IN MENTAL STATUS,\n",
    "# HYPOXIA;DIARRHEA,\n",
    "# HYPOXIA;HYPOTENSION,\n",
    "# HYPOXIA;MORBID OBESITY,\n",
    "# HYPOXIA;NSTEMI;PNEUMONIA,\n",
    "# HYPOXIA;PLEURAL EFFUSIONS,\n",
    "# HYPOXIA;PULMONARY EMNBOLUS,\n",
    "# HYPOXIA;R/O PNA,\n",
    "# HYPOXIA;SHORTNESS OF BREATH,\n",
    "# HYPOXIA;SHORTNESS OF BREATH;TELEMETRY,\n",
    "# 'HYPOXIA;TELEMETRY','LEFT LUNG EFFUSION','LEFT PLEURO EFFUSION','LEFT PNEUMOTHORAX',\n",
    "# 'LOBAR PNEUMONIA','LUNG ABCESS;PAINLESS JAUNDICE',\n",
    "# LUNG CA,\n",
    "# LUNG CANCER,\n",
    "# 'LUNG CANCER,HYPERKALEMIA',\n",
    "# 'LUNG CANCER,NAUSEA,VOMITING',\n",
    "# LUNG CANCER;PNEUMONIA,\n",
    "# LUNG MASS,\n",
    "# 'LUNG MASS, DIVERTICULAR ABSCESS','LYMPHOMA;SHORTNESS OF BREATH',\n",
    "# 'LYNPHOMA;TUBERCULOSIS','MEDIASTINAL MASS',\n",
    "# 'MEDIASTINITIS',                                                                      \n",
    "# ]), 'AdmitDiagnosis'] = \"RESPIRATORY\"\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# nooutlier_simple_data.loc[nooutlier_simple_data['AdmitDiagnosis'].isin(['ACUTE CEREBRAL VASCULAR ACCIDENT',\n",
    "# ACUTE ON CHRONIC SDH,\n",
    "# ACUTE OR CHRONIC SUBDURAL HEMATOMA,ACUTE SUBDURAL HEMATOMA,\n",
    "# ACUTE SUBDURAL HEMATOMA;FACIAL FRACTURE,\n",
    "# 'ACUTE SUBDURAL HEMATOMA;TELEMETRY',\n",
    "# BILATERAL SAH,\n",
    "# BILATERAL SUBARACHNOID HEMORRHAGE,\n",
    "# 'BILATERAL SUBDURAL HEMATOMAS','BLEEDING BRAIN MASS',BRAIN ABSCESS,\n",
    "# BRAIN ANEURYSM,\n",
    "# BRAIN CANCER,\n",
    "# BRAIN EDEMA,\n",
    "# BRAIN HERNIATION,\n",
    "# BRAIN MASS,\n",
    "# BRAIN MASS WITH 12 MM SHIFT,\n",
    "# BRAIN MASS, BLEED,\n",
    "# BRAIN MASS/SDH,\n",
    "# BRAIN MASS/SZ,\n",
    "# BRAIN MASS; SEIZURES,\n",
    "# BRAIN MASS;TELEMETRY,\n",
    "# BRAIN METASTASIS,\n",
    "# BRAIN METS,\n",
    "# 'BRAIN TUMOR',CERBRAL VASCULAR ACCIDENT,\n",
    "# CEREBELLAR BLEED,\n",
    "# CEREBELLAR HEMORRHAGE,\n",
    "# CEREBELLAR INFACTS,\n",
    "# CEREBELLAR LESIONS,\n",
    "# CEREBELLAR MASS,\n",
    "# CEREBRAL ANEURYSM,\n",
    "# CEREBRAL BLEED,\n",
    "# CEREBRAL CONTUSION,\n",
    "# CEREBRAL EDEMA,\n",
    "# CEREBRAL HEMORRHAGE,\n",
    "# CEREBRAL MASS,\n",
    "# CEREBRAL VASCULAR ACCIDENT,\n",
    "# CEREBROVASCULA ACCIDENT,\n",
    "# CEREBROVASCULAR ACCIDENT,\n",
    "# CEREBROVASCULAR ACCIDENT;FEV,\n",
    "# 'CEREBROVASCULAR ACCIDENT;TELEMETRY','CONCUSSION',\n",
    "# 'CONFUSION','CYST IN BRAIN','HYDROCEPHALUS',\n",
    "# 'HYDROCEPHALUS/VP MALFUNTION','ICB',ICH,\n",
    "# ICH/C2 FX,\n",
    "# ICH/SAH,\n",
    "# 'ICH;TELEMETRY','ICP','INTER CRANIAL HEMORRHAGE',\n",
    "# INTERCRANIAL BLEED,\n",
    "# INTERCRANIAL HEMORRAGE,\n",
    "# INTERCRANIAL HEMORRHAGE,\n",
    "# 'INTERCRANIAL HEMORRHAGE;TELEMETRY','INTRA CRANIAL HEMORRHAGE','INTRACANIAL HEMORRHAGE',\n",
    "# INTRACEREBRAL HEMORRHAGE,\n",
    "# INTRACRANIAL,\n",
    "# INTRACRANIAL BLEED,\n",
    "# INTRACRANIAL CONTUSION;TELEMETRY,\n",
    "# INTRACRANIAL HEMORRAGE,\n",
    "# INTRACRANIAL HEMORRHAGE,\n",
    "# INTRACRANIAL HEMORRHAGE/HEMOTHORAX/SPINAL FX/S/P FALL,\n",
    "# INTRACRANIAL HEMORRHAGE; FEMUR FX; S/P ATV CRASH,\n",
    "# INTRACRANIAL HEMORRHAGE;FACIAL FRACTURES,\n",
    "# INTRACRANIAL HEMORRHAGE;FINGER DISLOCATION,\n",
    "# INTRACRANIAL HEMORRHAGE;SZ,\n",
    "# INTRACRANIAL HEMORRHAGE;TELEMETRY,\n",
    "# INTRACRANIAL MASS,\n",
    "# INTRACRANIAL SWELLING,\n",
    "# 'INTRACRANIAL TUMOR','INTRAPARANCHYMAL HEMORRHAGE',\n",
    "# INTRAPARENCHYMAL,\n",
    "# INTRAPARENCHYMAL BLEED,\n",
    "# INTRAPARENCHYMAL BRAIN HEMORRHAGE,\n",
    "# INTRAPARENCHYMAL HEMORRAGE,\n",
    "# INTRAPARENCHYMAL HEMORRHAGE,\n",
    "# INTRAPARENCHYMAL HEMORRHAGE;S/P MOTOR VEHICLE ACCIDENT;?BRAIN MASS,\n",
    "# INTRAPARENCHYMAL HEMORRHAGE;TELEMETRY,\n",
    "# INTRAPARENCHYMEL BLEED,\n",
    "# INTRAPARENCYMAL HEMORRHAGE,\n",
    "# INTRAPRANCHYMAL BLEED,\n",
    "# INTRAPRANCHYMECAL BLEED,\n",
    "# 'INTRATHECAL MASS','IPH','LEFT CEREBRAL VASCULAR ACCIDENT',\n",
    "# LEFT TEMPORAL LOBE LESION-S/P SEIZURE,\n",
    "# 'LEFT TEMPORAL MASS','MASS LESION IN BRAIN','MENIGITIS',\n",
    "# 'MENINGIOMA',\n",
    "# 'MENINGITIS',\n",
    "# 'MENINGOENCEPHALITIS',\n",
    "# ]), 'AdmitDiagnosis'] = \"NEUROLOGY\"\n",
    "\n",
    "\n",
    "# nooutlier_simple_data.loc[nooutlier_simple_data['AdmitDiagnosis'].isin([ACUTE RENAL, \n",
    "# ACUTE RENAL FAILRUE,\n",
    "# ACUTE RENAL FAILURE,\n",
    "# 'ACUTE RENAL FAILURE, CELLULITIS',\n",
    "# 'ACUTE RENAL FAILURE, UREMIC SYNDROME'\n",
    "# 'ACUTE RENAL FAILURE,HYPOXIA'\n",
    "# 'ACUTE RENAL FAILURE,UROSEPSIS,HYPOTENSION,HYPOXIA'\n",
    "# ACUTE RENAL FAILURE-ACIDOSIS,\n",
    "# ACUTE RENAL FAILURE-BACTEREMIA,\n",
    "# ACUTE RENAL FAILURE-CHF EXACERBATION,\n",
    "# ACUTE RENAL FAILURE-UTI-?FASCITIS/HEMATOMA OF THIGH,\n",
    "# ACUTE RENAL FAILURE/CHOLE/R/O MYOCARDIAL INFARCTION,\n",
    "# ACUTE RENAL FAILURE:? LINE INFECTION,\n",
    "# ACUTE RENAL FAILURE; HYPOGLYCEMIA,\n",
    "# ACUTE RENAL FAILURE;CHANGE IN MENTAL STATUS,\n",
    "# ACUTE RENAL FAILURE;CHEST PAIN;TELEMETRY,\n",
    "# ACUTE RENAL FAILURE;CONGESTIVE HEART FAILURE;PERICARDIAL EFFUSION,\n",
    "# ACUTE RENAL FAILURE;DIABETIC KETOACIDOSIS,\n",
    "# ACUTE RENAL FAILURE;NAUSIA;VOMOTTING;DIARRHEA,\n",
    "# ACUTE RENAL FAILURE;SEPSIS,\n",
    "# ACUTE RENAL FAILURE;SEPSIS;TELEMETRY,\n",
    "# ACUTE RENAL FAILURE;SEPTIC ARTHRITIS,\n",
    "# ACUTE RENAL FAILURE;TACHYCARDIA,\n",
    "# ACUTE RENAL FAILURE;TELEMETRY,\n",
    "# ACUTE RENAL FAILURE;VOMITING;DIARRHEA;TELEMETRY,\n",
    "# ARF AND RESPIRATORY DISTRESS,\n",
    "# ARF, BACK PAIN,\n",
    "# ARF,GIB,\n",
    "# 'ARF/ABD PAIN',BLADDER BLEED;ANEMIA,\n",
    "# BLADDER CANCER,\n",
    "# BLADDER PERFORATION,\n",
    "# BLADDER RUPTURE,FRACTURES,\n",
    "# 'BLADDER RUPTURE;SEPSIS','CHRONIC RENAL FAILURE','END STAGE RENAL DISEASE',\n",
    "# 'END STAGE RENAL DISEASE-URTICARIA-FEVER',\n",
    "# 'END STAGE RENAL DISEASE; RASH',\n",
    "# 'END STAGE RENAL DISEASE;HYPOGLYCEMIA;HYPERTENSION',\"ESRD-FEVER\",'KIDNEY ABSCESS',\n",
    "# KIDNEY BLEED,\n",
    "# KIDNEY STONES,\n",
    "# 'KIDNEY THROMBOEMBOLISM','LACTIC ACIDOSIS','LEFT RENAL FRACTURE',\n",
    "# ]), 'AdmitDiagnosis'] = \"UROLOGY\"\n",
    "  \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# nooutlier_simple_data.loc[nooutlier_simple_data['AdmitDiagnosis'].isin(['? BRAIN TUMOR','? OVERDOSE',\n",
    "# '?SEPSIS',\n",
    "# 'ACETABULAR FRACTURE',\n",
    "# 'ACETOMENGOHEN OVERDOSE','ACIDOSIS',\n",
    "# 'ACUTE FEVER;SEPSIS;TELEMETRY',\n",
    "# 'ADENEXAL MASS',\n",
    "# AKTERED MENTAL STATUS,\n",
    "# ALCAHOL WITHDRAWAL,\n",
    "# ALCOHOL KETOACIDOSIS,\n",
    "# ALCOHOL WITHDRAWAL,\n",
    "# ALCOHOL WITHDRAWAL/UTI,\n",
    "# ALCOHOL WITHDRAWAL; ? UPPER GI BLEED,\n",
    "# ALLERGIC REACTON,\n",
    "# ALTERED MENTAL STATU,\n",
    "# ALTERED MENTAL STATUA,\n",
    "# ALTERED MENTAL STATUS,\n",
    "# ALTERED MENTAL STATUS CHANGE;TELEMETRY,\n",
    "# ALTERED MENTAL STATUS CHANGES,\n",
    "# ALTERED MENTAL STATUS CHANGES;METHICILLIN RESISTANT STAPH AUREUS,\n",
    "# ALTERED MENTAL STATUS CHANGES;URINARY TRACT INFECTION;PNEUMONIA;TELEMETRY,\n",
    "# ALTERED MENTAL STATUS,FEVER,HYPOTENSION,\n",
    "# ALTERED MENTAL STATUS-HYPOGLYCEMIA,\n",
    "# ALTERED MENTAL STATUS-HYPOTHERMIA,\n",
    "# ALTERED MENTAL STATUS-NARCOTIC OD-HYPOTENSION-HYPERGLYCENIA,\n",
    "# ALTERED MENTAL STATUS/ GI BLEED,\n",
    "# ALTERED MENTAL STATUS;CONGESTIVE HEART FAILURE;PNEUMONIA,\n",
    "# ALTERED MENTAL STATUS;FEBRILE,\n",
    "# ALTERED MENTAL STATUS;PNEUMONIA,\n",
    "# ALTERED MENTAL STATUS;PNEUMONIA;DEHYDRATION,\n",
    "# ALTERED MENTAL STATUS;R/O MENINGITIS,\n",
    "# ALTERED MENTAL STATUS;RESPIRATORY FAILURE,\n",
    "# ALTERED MENTAL STATUS;TELEMETRY,\n",
    "# ALTERED;RLL;PNEUMONIA,\n",
    "# AMANITA INTOXICATION,\n",
    "# AMI;DYSPNEA,\n",
    "# AMPUTATION,\n",
    "# AMS,\n",
    "# ANAPHYLAXIS,\n",
    "# ANASARCA,\n",
    "# ANKLE FRACTURE,\n",
    "# ANKLE FRACTURE;HYPOXIA,\n",
    "# ANTERIOR CERVICAL COLLECTION,\n",
    "# ANTERIOR CORD SYNDROME,\n",
    "# ANTICHOLINERGIC TOXICITY,\n",
    "# ANTIFREEZE INGESTION,\n",
    "# ASA TOXICITY,\n",
    "# ASCENDING PARALYSIS;TELEMETRY,\n",
    "# 'ASSAULT',\n",
    "# 'B-BLOCKER;BENZO OD',BACK PAIN,\n",
    "# BACK PAIN R/O CORD COMPRESSION,\n",
    "# BACK PAIN,CANNOT AMBULATE,\n",
    "# BACK PAIN-ACUTE PAIN,\n",
    "# BACK PAIN-DISSECTION,\n",
    "# BACK PAIN-SACRAL FRACTURE,\n",
    "# BACK PAIN-WEAKNESS,\n",
    "# BACTEREMIA,\n",
    "# BACTEREMIA,DIABETIC KETOACIDOSIS,\n",
    "# BACTERIAL ENDOCARDITIS;TELEMETRY,\n",
    "# BACTERIAL MENINGITIS,\n",
    "# BACTERIAL,MENINGITIS,\n",
    "# BACTERMIA,\n",
    "# BACTREMIA,\n",
    "# BASAL ANEURYSM,\n",
    "# BASAL GANGLIA BLEED,\n",
    "# BASAL GANGLIA HEMORRHAGE,\n",
    "# BASAL GANGLIA INTRACRANIAL HEMORRHAGE,\n",
    "# BEHYDRATION,\n",
    "# BENIGN PNEUMATOSIS,\n",
    "# BILATERAL ANKLE FX; LACERATED PERINEUM,\n",
    "# BILATERAL DVT,\n",
    "# BILATERAL FEMORAL  FX,\n",
    "# 'BILATERAL HIP FRACTURE','BLAST CRISIS',BLEEDING HEPATOMA,\n",
    "# BLEEDING LIVER METASTATIC,\n",
    "# BLEEDING MASS,\n",
    "# BLEEDING TRACHEOSTOMY,\n",
    "# \"BLLEDING;D&C\",\n",
    "# BLUNT TAUMA,\n",
    "# BLUNT TRAUMA,\n",
    "# \"BLUNT TRAUMA;S/P MOTOR VEHICLE ACCIDENT\",\n",
    "# BLUNT TRAUMA;TELEMETRY,\n",
    "# BOERHAAVE SYNDROME,\n",
    "# BONY PAIN,\n",
    "# \"BOORHAAVE'S\",BREAST CANCER-RESPIRATORY FAILURE,\n",
    "# BREAST CANCER/FEVER, BACK PAIN,\n",
    "# BREAST CANCER;BRAIN METS;SEIZURE,\n",
    "# BREAST CANCER;DEHYDRATION,\n",
    "# BREAST CANCER;PNEUMONIA,\n",
    "# 'BREAST CANCER;RENAL FAILURE;HYDRONEPHROSIS',BULEMIA,HYPOKALEMIA,\n",
    "# C SPINE FRACTURE,\n",
    "# C SPINE FRACTURE,FRONTAL SINUS FRACTURE,\n",
    "# C SPINE FX,\n",
    "# C-SPINE FRACTURE,\n",
    "# C-SPINE FRACTURE;S/P FALL,\n",
    "# C-SPINE FX/ PARALYSIS S/P FALL,\n",
    "# C-SPINE INJURY,\n",
    "# C1 FRACTURE,\n",
    "# C1 FRACTURE/ANKLE FRACTURE/S/P MVC,\n",
    "# C2 FRACTURE,\n",
    "# C2 FRACTURE, S/P FALL,\n",
    "# C2 FX,\n",
    "# C2 FXS;RIB FXS,\n",
    "# C4-C5 SUBLUXATION,\n",
    "# C5 BURST FRACTURE,\n",
    "# C5-6 FRACTURE,\n",
    "# C6 FRACTURE,\n",
    "# C6 SPINOUS FRACTURE,\n",
    "# C7 FRACTURE,\n",
    "# 'C7 SPINAL FRACTURE',CELLULIIS,\n",
    "# CELLULIITIS-SEPSIS,\n",
    "# CELLULITIS,\n",
    "# CELLULITIS LEFT LOWER EXTREMITY,\n",
    "# CELLULITIS SEPSIS,\n",
    "# CELLULITIS,INABILITY TO URINATE,\n",
    "# CELLULITIS,MENTAL STATUS CHANGE,\n",
    "# CELLULITIS,SEPSIS,\n",
    "# CELLULITIS/SEPSIS,\n",
    "# CELLULITIS; MENTAL STATUS CHANGES,\n",
    "# CELLULITIS;DEEP VEIN THROMBOSIS,\n",
    "# CELLULITIS;SEPSIS,\n",
    "# 'CELLULITIS;TELEMETRY',CERVICAL EPIDURAL ABCESS,\n",
    "# CERVICAL EPIDURAL HEMATOMA,\n",
    "# CERVICAL FRACTURE,\n",
    "# CERVICAL HEMATOMA,\n",
    "# CERVICAL SPINE FRACTURE,\n",
    "# CERVICAL SPINE INJURY,\n",
    "# CERVICAL STENOSIS WITH MYELOPATHY,\n",
    "# CHANGE IN MENTAL STATUS,\n",
    "# 'CHANGE IN MENTAL STATUS, HYPOTENSION,BRADYCARDIA',\n",
    "# 'CHANGE IN MENTAL STATUS,HYPERTENSIVE URGENCY',\n",
    "# 'CHANGE IN MENTAL STATUS,SEIZURE',\n",
    "# CHANGE IN MENTAL STATUS;GASTROINTESTINAL BLEED;TELEMETRY,\n",
    "# CHANGES IN MENTAL STATUS;S/P FALLS,\n",
    "# CHARCOAL ASPIRATION,\n",
    "# CHESST PAIN,\n",
    "# CHEST,\n",
    "# CHEST ABSCESS,\n",
    "# CHEST DISCOMFORT,\n",
    "# CHEST PAI N,\n",
    "# CHEST PAIN,\n",
    "# CHEST PAIN R/O MI,\n",
    "# CHEST PAIN WITH LEFT ARM NUMBNESS,\n",
    "# CHEST PAIN, AORITIC ULCERATION,\n",
    "# CHEST PAIN,GIB,\n",
    "# CHEST PAIN,HYPERTENSION,\n",
    "# CHEST PAIN,HYPOTENSION,\n",
    "# CHEST PAIN,RENAL FAILURE,\n",
    "# CHEST PAIN,STE MI,\n",
    "# CHEST PAIN-CONGESTIVE HEART FAILURE,\n",
    "# CHEST PAIN-LEUKEMIA,\n",
    "# CHEST PAIN-PNEUMONIA,\n",
    "# CHEST PAIN-R/O MYOCARDIAL INFARCTION,\n",
    "# CHEST PAIN/ CHANGE IN MENTAL STATUS,\n",
    "# CHEST PAIN/ SHOTNESS OF BREATH,\n",
    "# CHEST PAIN/SEPSIS,\n",
    "# CHEST PAIN/SHORTNESS OF BREATH,\n",
    "# CHEST PAIN; MENTAL STATUS CHANGE,\n",
    "# CHEST PAIN; R/O PULMONARY EMBOLUS,\n",
    "# CHEST PAIN; WOUND INFECTION,\n",
    "# CHEST PAIN;?TIA,\n",
    "# CHEST PAIN;CATH,\n",
    "# CHEST PAIN;HEART FAILURE,\n",
    "# CHEST PAIN;HYPERGLYCEMIA,\n",
    "# CHEST PAIN;HYPOTENSION,\n",
    "# CHEST PAIN;MYOCARDIAL INFARCTION,\n",
    "# CHEST PAIN;PNA,\n",
    "# CHEST PAIN;RULE OUT TUBERCULOSIS,\n",
    "# CHEST PAIN;RULE-OUT MYOCARDIAL INFARCTION,\n",
    "# CHEST PAIN;RULE-OUT MYOCARDIAL INFARCTION;TELEMETRY,\n",
    "# CHEST PAIN;TELEMETRY,\n",
    "# CHEST PAIN;TELEMETRY,HYPOPHOSPHATEMIA,\n",
    "# CHEST PAIN;TELEMETRY; GI BLEED,\n",
    "# CHEST PAIN;TELEMETRY;? PARTIAL SBO,\n",
    "# CHEST PAIN;TELEMETRY;GI BLEED,\n",
    "# CHEST PAIN;TELEMETRY;HYPOXIA,\n",
    "# CHEST PAIN;TELEMETRY;MELENA,\n",
    "# CHEST PAIN;TELEMETRY;PNEUMONIA,\n",
    "# CHEST PAIN;TELEMETRY;THROAT PAIN,\n",
    "# CHEST STAB WOUND,\n",
    "# CHEST STAB WOUNDS,\n",
    "# 'CHEST WALL HEMATOMA',\"CHRON'S DISEASE;THROMBOCTOPENIA;CVID\",\n",
    "# \"CHRONIC LYMPHOCYTIC LEUKEMIA;SEIZURE\",CLOTTED AV FISTULA,\n",
    "# CLOTTED GRAFT,\n",
    "# COCAINE OD,\n",
    "# COCAINE OVERDOSE,\n",
    "# CODE SEPISI,\n",
    "# CODE STROKE,\n",
    "# COGIATED DELIRUM,\n",
    "# 'COLD LEG','COMPARTMENT SYNDROME','COMPLICATION FEMUR HARDWARE',\n",
    "# 'COMPOUND FRACTURE;URINARY TRACT INFECTION',\n",
    "# 'COMPRESSION FRACTURE','CORD COMPRESSION',\n",
    "# 'CORD INJURY',COUGH,\n",
    "# COUGH;FEVER,\n",
    "# \"CROHN'S FLARE\",\n",
    "# \"CROHN'S; PERFORATION\",\n",
    "# CROHNS FLAIR,\n",
    "# CRUSH INJURY TO ABDOMEN,\n",
    "# CRUSH TRAUMA,\n",
    "# \"CRUSHING INJURY /RIGHT LEG\",DEHISENCE OF WOUND,\n",
    "# DEHYDRATION,\n",
    "# DEHYDRATION-NAUSEA-VOMITING,\n",
    "# DEHYDRATION/INCREASE BS,\n",
    "# DEHYDRATION/UTI,\n",
    "# DEHYDRATION;ABDOMINAL PAIN,\n",
    "# DEHYDRATION;ELEVATED INR,\n",
    "# DEHYDRATION;HYPERNATREMIA,\n",
    "# DEHYDRATION;JAUNDICE,\n",
    "# DEHYDRATION;MALNUTRITION,\n",
    "# DEHYDRATION;NAUSEA;VOMITING;MENTAL STATUS CHANGES,\n",
    "# DEHYDRATION;TELEMETRY,\n",
    "# DEHYDRATION;THROAT CANCER,\n",
    "# DELERIUM,\n",
    "# DELTA MS,\n",
    "# DELTA MS/HYPERTENSION,\n",
    "# DEMENTIA;FALLS,\n",
    "# DENS FRACTURE,\n",
    "# DEPRESSED SKULL FRACTURE,\n",
    "# DIABETIC FOOT,\n",
    "# DIABETIC FOOT ULCER,\n",
    "# DIABETIC KETO ACIDOSIS,\n",
    "# \"DIABETIC KETO ACIDOSIS,SEIZURES\",\n",
    "# DIABETIC KETOACIDOSIS,\n",
    "# \"DIABETIC KETOACIDOSIS,R/O MI\",\n",
    "# DIABETIC KETOACIDOSIS-CONGESTIVE HEART FAILURE,\n",
    "# DIABETIC KETOACIDOSIS;,\n",
    "# DIABETIC KETOACIDOSIS;GASTROINTESTINAL BLEED,\n",
    "# DIABETIC KETOACIDOSIS;HYPOTENTION,\n",
    "# DIABETIC KETOACIDOSIS;SUBDURAL HEMATOMA,\n",
    "# DIABETIC KETOACIDOSIS;TELEMETRY,\n",
    "# \"DIABETID KETOACIDOSIS\",\"DIARRHEA\",\n",
    "# DIARRHEA AND BRBPR,\n",
    "# DIARRHEA-HYPOTENSION,\n",
    "# DIARRHEA;DEHYDRATION,\n",
    "# \"DIARRHEA;DEHYDRATION;CHEST PAIN\",\"DICA-HEMATEMESIS\",\n",
    "# \"DIG TOXICITY\",DIPLOPIA,\n",
    "# DISPLACED TRACH TUBE,\n",
    "# DISSECTING AORTIC ANEURYSM,\n",
    "# DISSECTING ASCENDING AORTIC ANEURYSM,\n",
    "# DISSECTION,\n",
    "# DISSEMINATED ZOSTER,\n",
    "# \"DISTAL FEMUR FRACTURE\",\"DIZZINESS\",DIZZINESS,\n",
    "# \"DKA, ARF, CHOLECYSTITIS\",\n",
    "# \"DKA, DEHYDRATION, HEMATORIA\",\n",
    "# \"DKA, SEPSIS\",\n",
    "# 'DRAIN LEAKAGE',\n",
    "# 'DRUG OVERDOSE',DYSTONIA;S/P G-TUBE PLACEMENT,\n",
    "# DYSTONIC RXN,\n",
    "# EARLY SEPSIS;PNEUMONIA,\n",
    "# EDEMA,\n",
    "# EKG CHANGES,\n",
    "# EKG CHANGES;R/O ISCHEMIA,\n",
    "# ELEVATED BILIRUBIN,\n",
    "# ELEVATED INR,\n",
    "# ELEVATED INR;COUMADIN TOXICITY;LOWER GI BLEED,\n",
    "# \"ELEVATED LFTS\",ENCEPHALITIS,\n",
    "# ENCEPHALOPATHY,\n",
    "# 'ENCEPHALOPATHY;TELEMETRY',\"ENDOCARDITIS\",\n",
    "# \"ENDOMETRITIS\",EPIDIRAL ABSCESS,\n",
    "# EPIDURAL,\n",
    "# EPIDURAL ABSCESS,\n",
    "# EPIDURAL HEMATOMA,\n",
    "# EPIDURAL HEMATOMA;TELEMETRY,\n",
    "# EPIDURAL HEMORRHAGE,\n",
    "# EPIDURAL NECK HEMATOMA,\n",
    "# \"EPIGASTRIC PAIN\",'EPISTAXIS',\n",
    "# 'EPISTAXIS;THROMBOCYTOPENIA',ESOPHAGEAL OBSTRUCTION,\n",
    "# ESOPHAGEAL PERFORATION,\n",
    "# ESOPHAGEAL RESECTION,\n",
    "# ESOPHAGEAL RUPTURE,\n",
    "# 'ESOPHAGEAL TEAR',ETHYLENE GLYCOL INTOXICATION,\n",
    "# ETOH CIRRHOSIS;TELEMETRY,\n",
    "# ETOH IMTOXICATION,\n",
    "# ETOH INTOXICATION,\n",
    "# ETOH WITHDRAWAL,\n",
    "# ETOH WITHDRAWAL/ PANCREATITIS,\n",
    "# ETOH WITHDRAWAL/DETOX,\n",
    "# ETOH WITHDRAWAL;FAILURE TO THRIVE,\n",
    "# ETOH WITHDRAWAL;GASTROINTESTINAL BLEED;TELEMETRY,\n",
    "# ETOH WITHDRAWAL;SEIZURES,\n",
    "# 'ETOH-UNRESPONSIVE',EXP LAP,\n",
    "# EXPLORATORY LAPAROTOMY,\n",
    "# EXPLORATORY LAPAROTOMY FOR SMALL BOWEL OBSTRUCTION,\n",
    "# EXPLORATORY LAPORATOMY,\n",
    "# EXPLORATORY LAPORATOMY/? BOWEL PROLAPSE,\n",
    "# EXPLORTARY LAP/? PERFORATED BOWEL,\n",
    "# EXPOSURE,\n",
    "# FACIAL CELLULITIS,\n",
    "# FACIAL FRACTURE,\n",
    "# FACIAL FRACTURE/PELVIC FRACTURE,\n",
    "# FACIAL FRACTURE;ORBITAL FRACTURE,\n",
    "# FACIAL FRACTURE;TRACH;S/P MOTOR VEHICLE ACCIDENT,\n",
    "# FACIAL FRACTURES,\n",
    "# FACIAL FRACTURES/WRIST FRACTURES/INTRACRANIAL HEMORRHAGE,\n",
    "# FACIAL FRACTURES;RADIUS FRACTURE,\n",
    "# FACIAL FX;MOTOR VEHICLE ACCIDENT,\n",
    "# FACIAL SWELLING,\n",
    "# FACIAL TRAUMA,\n",
    "# FACIAL TRAUMA;S/P FALL,\n",
    "# FAILURE TO THIRVE,\n",
    "# FAILURE TO THRIVE,\n",
    "# FAILURE TO THRIVE;HYPOTENSION,\n",
    "# FALL,\n",
    "# FALL, ANKLE FX,\n",
    "# FALL,INTRACRANIAL HEMORRHAGE,\n",
    "# FALL/HEAD INJURY,\n",
    "# FALL; ? OD,\n",
    "# FALL;FEMUR FRACTURE,\n",
    "# FALL;HYPONATREMIS;HYPERKALEMIA,\n",
    "# FALL;RIB FX;UTI,\n",
    "# FALLS,\n",
    "# FATIGUE;MALAISE;POSTIVE TROPONIN,\n",
    "# FEBRILE NEUTROPENIA,\n",
    "# FEBRILE SYNDROME;DIALYSIS,\n",
    "# 'FEER;NAUSEA;VOMITING','FEMORAL FRACTURE',\n",
    "# FEMORAL NECK FRACTURE,\n",
    "# FEMUR FRACTURE,\n",
    "# FEMUR FRACTURE;S/P FALL,\n",
    "# FEMUR FRACTURE;TACHYCARDIA,\n",
    "# FEMUR FX,\n",
    "# FEVER,\n",
    "# FEVER OF AN UNKNOWN ORIGIN,\n",
    "# FEVER OF UNKNOWN ORIGIN,\n",
    "# FEVER OF UNKNOWN ORIGIN;DEHYDRATION,\n",
    "# FEVER UNKNOWN ORIGIN,\n",
    "# FEVER, A-FIB,\n",
    "# FEVER, ABDOMINAL PAIN,\n",
    "# FEVER, LIKELY LINE SEPSIS,\n",
    "# FEVER, MENTAL STATUS CHANGE,\n",
    "# FEVER,ALTERED MENTAL STATUS,\n",
    "# FEVER,DEHYDRATION,\n",
    "# FEVER,DEHYDRATION,FAILURE TO THRIVE,\n",
    "# FEVER,DIARRHEA,DEHYDRATION,\n",
    "# FEVER,ETIOLOGY,\n",
    "# FEVER,ETIOLOGY UNKNOWN,\n",
    "# FEVER,HIV,HEMATURIA,\n",
    "# FEVER,HYPOTENSION,\n",
    "# FEVER,LETHARGY,\n",
    "# FEVER,ORIGIN UNKNOWN,\n",
    "# FEVER,ORIGIN UNKNOWN;HYPOTENSION,\n",
    "# FEVER,PNEUMONIA,\n",
    "# FEVER,R/O MENINGITIS,\n",
    "# FEVER,R/O MI,\n",
    "# FEVER,SHAKING CHILLS OF UNKNOWN ETIOLOGY,\n",
    "# FEVER,VOMITING,\n",
    "# FEVER-ALTERED MENTAL STATUS,\n",
    "# FEVER-FAILURE TO THRIVE,\n",
    "# FEVER-FLU,\n",
    "# FEVER-HYPOTENSION,\n",
    "# FEVER-LINE SEPSIS,\n",
    "# FEVER-MENTAL STATUS CHANGES,\n",
    "# FEVER-NEUTROPENIA,\n",
    "# FEVER-SEPSIS,\n",
    "# FEVER/HYPER K/ AFIB RVR,\n",
    "# FEVER/HYPOTENSION,\n",
    "# FEVER/HYPOXIA,\n",
    "# FEVER/SYNCOPE,\n",
    "# FEVER; DEHYDRATION; EVAL FOR C-DIFF,\n",
    "# FEVER; MENTAL STATUS CHANGES,\n",
    "# FEVER; N/V,\n",
    "# FEVER; NEUTROPENIA,\n",
    "# FEVER;ALTERED MENTAL STATUS,\n",
    "# FEVER;BILIARY SLUDGE,\n",
    "# FEVER;CHANGE IN MENTAL STATUS,\n",
    "# FEVER;HEADACHE,\n",
    "# FEVER;HICKMAN LINE INFECTION,\n",
    "# FEVER;HYPERKALEMIA;CONGESTIVE HEART FAILURE,\n",
    "# FEVER;HYPOTENSION,\n",
    "# FEVER;HYPOTENSIVE URGENCY;METHICILLIN RESISTANT STAPH AUREUS,\n",
    "# FEVER;HYPOTENTION;TELEMETRY,\n",
    "# FEVER;LINE SEPSIS,\n",
    "# FEVER;MENINGITIS,\n",
    "# FEVER;MENTAL STATUS CHANGES,\n",
    "# FEVER;MYELOMA,\n",
    "# FEVER;NEUTROPENIA,\n",
    "# FEVER;PARALYSIS,\n",
    "# FEVER;R/O ENDOCARDITIS,\n",
    "# FEVER;R/O LINE SEPSIS,\n",
    "# FEVER;SEIZURE,\n",
    "# FEVER;SEPSIS,\n",
    "# FEVER;SEPSIS;HYPOXIA,\n",
    "# FEVER;TELEMETRY,\n",
    "# FEVER;URINARY TRACT INFECTION,\n",
    "# FEVER;WOUND DRAINAGE,\n",
    "# FEVERE,\n",
    "# FEVERS,\n",
    "# FINGER AMPUTATION,\n",
    "# FLAIL CHEST,\n",
    "# FOOT INFECTION;HYPERGLYCEMIA,\n",
    "# FOOT PAIN,\n",
    "# FOUND DOWN,\n",
    "# FOUND DOWN, S/P FALL,\n",
    "# FOUNIER GANFRENE,\n",
    "# \"FOURNIER'S GANGRENE;ANEURYSM\",\n",
    "# FRACTURED RIBS,\n",
    "# FREE ABDOMINAL AIR,\n",
    "# FREE AIR ABDOMEN,\n",
    "# 'FREQUENT FALLS-HYPOTENSION',FRONTAL BRAIN TUMOR,\n",
    "# FRONTAL MASS,\n",
    "# FUNGAL MENINGITIS,\n",
    "# 'FUNGEMIA','GAIT CHANGES','GANGRENE',\n",
    "# 'GANGRENE LEFT FOOT',GLOBE RUPTURE,\n",
    "# GRAFT INFECTION,\n",
    "# GRAM NEGATIVE RODS;BACTEREMIA,\n",
    "# GROIN CELLULITIS,\n",
    "# GUIANNE BARRE SYNDROME,\n",
    "# GUILLAIN BARRE SYNDROME,\n",
    "# GULLIAN BARRE,\n",
    "# GUN SHOT,\n",
    "# GUN SHOT WOUND,\n",
    "# GUN SHOT WOUND BACK,\n",
    "# GUN SHOT WOUND TO FACE,\n",
    "# GUN SHOT WOUND TO HEAD,\n",
    "# GUNSHOT WOUND;TELEMETRY,\n",
    "# HAND AMPUTATION,\n",
    "# HAND AMPUTATION;TELEMETRY,\n",
    "# HAND LACERATION,\n",
    "# HARD PALATE HEMATOMA,\n",
    "# HARDWARE INFECTION,\n",
    "# HD CATHETER LINE INFECTION,\n",
    "# HEAD BLEED,\n",
    "# HEAD BLEED WITH SHIFT,\n",
    "# HEAD INJURY,\n",
    "# HEAD INJURY-RESPIRATORY FAILURE,\n",
    "# HEAD INJURY;S/P FALL,\n",
    "# HEAD INJURY;SUBDURAL HEMORRHAGE;SUBARACHNOID HEMATOMA;PI BLEED,\n",
    "# HEAD MASS,\n",
    "# HEAD MASS WITH SHIFT,\n",
    "# HEAD MASS;UTI,\n",
    "# HEAD TRAUMA,\n",
    "# HEAD TRAUMA;SCALP LACERATION,\n",
    "# HEADACH,\n",
    "# HEADACHE,\n",
    "# HEADACHE,SINUSITIS,\n",
    "# HEADACHE-SEIZURES-R/O SHEEHAN SYNDROME,\n",
    "# 'HEADACHE;FEVER;ACUTE RENAL FAILURE',HEMANGIOMA,\n",
    "# HEMATEMESIS,\n",
    "# HEMATEMSIS,\n",
    "# HEMATOCHEZIA,\n",
    "# HEMATOMA,\n",
    "# HEMATURIA,\n",
    "# \"HEMATURIA,HYPOTENSION\",\n",
    "# HEMOLYTIC ANEMIA,\n",
    "# HEMOPERITENEUM,\n",
    "# HEMOPERITONEUM,\n",
    "# HEMOPNEUMOTHORAX,\n",
    "# HEMOPTOSIS,\n",
    "# \"HEMOPTYSIS\",HEMORRHAGIC CONTUSION,\n",
    "# HEMOTHORAX,\n",
    "# HEMOTHORAX/RIB FX,\n",
    "# HEMOTHORAX;FEVER;CHEST PAIN;TELEMETRY,\n",
    "# HEPARIN INFUSIN, S/P IVC FILTER PLACEMENT,\n",
    "# 'HEPATIC BLEED',HERNIA,\n",
    "# HERNIATED DISK,\n",
    "# HEROIN OVERDOSE-ARDS,\n",
    "# HIGH INR,\n",
    "# HIP FRACTURE,\n",
    "# 'HIP FRACTURE,R/O MI',\n",
    "# HIP FRACTURE/ESRD,\n",
    "# HIP FRACTURE;PNEUMONIA,\n",
    "# HIP FX,\n",
    "# HIP PAIN,\n",
    "# 'HIV,ESOPHAGITIS',\n",
    "# HTN URGENCY;GASTROPARESIS,\n",
    "# HUMERAL FRACTURE,\n",
    "# 'HUMERUS FRACTURE','HYPERCALCEMIA','HYPERGLYCEMIA'\n",
    "# HYPERGLYCEMIA,SEIZURES,\n",
    "# HYPERGLYCEMIA-DIABETES,\n",
    "# HYPERGLYCEMIA-DIABETIC KETOACIDOSIS,\n",
    "# HYPERGLYCEMIA;HEADACHE;? SEIZURE,\n",
    "# HYPERGLYCEMIA;SUBDURAL HEMATOMA,\n",
    "# HYPERGLYCEMIA;TELEMETRY,\n",
    "# HYPERGLYCEMIC COMA,\n",
    "# HYPERKALEMIA,\n",
    "# 'HYPERKALEMIA, ANEMIA, NAUSEA, ARF',\n",
    "# HYPERKALEMIA,? DIABETIC KETOACIDOSIS,\n",
    "# HYPERKALEMIA,HYPERGLYCEMIA,UNRESPONSIVE,\n",
    "# HYPERKALEMIA-RENAL FAILURE-ALTERED MENTAL STATUS,\n",
    "# HYPERKALEMIA; ACUTE RENAL FAILURE,\n",
    "# HYPERKALEMIA;BRADYCARDIA,\n",
    "# 'HYPERKALEMIA;MYOCARDIAL INFARCT;CHRONIC OBST PULM DISEASE;RENAL FAILURE',\n",
    "# HYPERKALEMIA;PULMONARY EDMA,\n",
    "# HYPERKALEMIA;RENAL FAILURE,\n",
    "# HYPERKALEMIA;TELEMETRU,\n",
    "# HYPERKALEMIA;TELEMETRY,\n",
    "# 'HYPERNATREMIA',\n",
    "# 'HYPERNATREMIA, UROSEPSIS',\n",
    "# HYPERNATREMIA,MENTAL STATUS CHANGE,\n",
    "# HYPERNATREMIA-ALTERED MENTAL STATUS,\n",
    "# HYPERNATREMIA;HEMORRHAGIC STROKE,\n",
    "# 'HYPEROSMOLAR NON-KETOTIC HYPERGLYCEMIA',HYPERTHERMIA,\n",
    "# HYPERTRNSIVE EMERGENCY.SEIZURE,\n",
    "# HYPOENSIA;SEPSIS,\n",
    "# HYPOGLYCEMIA,\n",
    "# \"HYPOGLYCEMIA, \",\n",
    "# HYPOGLYCEMIA-RESPIRATORY DISTRESS,\n",
    "# HYPOGLYCEMIA/HYPOTHERMIA/?BACTEREMIA,\n",
    "# HYPOGLYCEMIA;ACUTE RENAL FAILURE,\n",
    "# HYPOGLYCEMIA;CHRONIC RENAL INSUFFICIENCY,\n",
    "# HYPOGLYCEMIA;DELTA MS;TELEMETRY,\n",
    "# HYPOKALEMIA,\n",
    "# HYPOKALEMIA SHOCK;PELVIC FRACTURE;COMATOSE;STATUS POST MOTOR VEHICLE ACCIDENT WITH INJURIES,\n",
    "# HYPOKALEMIA-URINARY TRACT INFECTION,\n",
    "# HYPOKALEMIA/HEPATITIS,\n",
    "# HYPONATREMIA,\n",
    "# \"HYPONATREMIA, LOW BICARB, DVT\",\n",
    "# HYPONATREMIA,SEIZURE,\n",
    "# HYPONATREMIA;BRAIN MASS,\n",
    "# HYPONATREMIA;GENERAL WEAKNESS,\n",
    "# HYPONATREMIA;HYPOTENSION,\n",
    "# HYPONATREMIA;LE EDEMA,\n",
    "# HYPONATREMIA;NAUSEA;VOMITING,\n",
    "# HYPONATREMIA;RENAL FAILURE,\n",
    "# HYPONATREMIA;RULE-OUT MYOCARDIAL INFARCTION;TELEMETRY,\n",
    "# 'HYPONATREMIA;URINARY TRACT INFECTION','HYPOTHERMIA',\n",
    "# 'HYPOTHERMIA;HYPOTENSION;TELEMETRY',\n",
    "# 'HYPOTHERMIC;HYPOTENSIVE','HYPOVOLEMIA',\n",
    "# 'HYPOVOLEMIC SHOCK','IMI',\n",
    "# INABILITY TO WALK;VANCOMYCIN RESISTANT ENTEROCOCCUS,\n",
    "# INCARCERATED HERNIA,\n",
    "# INCARCERATED PARASTOMAL HERNIA,\n",
    "# INFECTED AICD,\n",
    "# INFECTED KIDNEY STONE,\n",
    "# INFECTED KIDNEY STONE;PYELONEPHRITIS,\n",
    "# INFECTED OBSTRUCTED KIDNEY STONE,\n",
    "# INFECTED ORTHOPEDIC HARDWARE,\n",
    "# INFECTED PANCREATIC PSEUDOCYST,\n",
    "# INFECTED STENT,\n",
    "# 'INFECTED STONE,ACUTE RENAL FAILURE,SEPSIS',\n",
    "# INFECTIN,\n",
    "# INFECTION,\n",
    "# INFECTION OF THE SPINAL STIMULATOR;CELLULITIS,\n",
    "# 'INFECTION;GASTROINTESTINAL BLEED','INFLUENZA LIKE ILLNESS',\n",
    "# INFLUENZA;PNEUMONIA,\n",
    "# INGESTION,\n",
    "# INGESTION OF FOREIGN BODY,\n",
    "# INGESTION OVERDOSE,\n",
    "# 'INGUINAL HERNIA REPAIR','INTERMITTENT CHEST PAIN','INTOXICATION','INTRACTABLE NAUSEA,VOMITING',\n",
    "# 'INTRACTABLE PAIN',JAW ABSCESS,\n",
    "# JAW PAIN,\n",
    "# 'JOINT EFFUSION','KNEE PAIN',\n",
    "# L 1 FRACTURE,\n",
    "# L ARM INFECTION,\n",
    "# L SIDED PNEUMOTHORAX,\n",
    "# L THIGH HEMATOMA;FACTOR V111 INHIBITOR,\n",
    "# L. HIP FRACTURE; RF; ANEMIA,\n",
    "# L. SIDED EFFUSION,\n",
    "# L1 FX,\n",
    "# 'L2 COMPRESSION FRACTURE;HYPOTENSION','LE WEAKNESS','LEFT ANKLE FRACTURE',\n",
    "# LEFT COMPRESSION FRACTURE,\n",
    "# 'LEFT DISTAL RADIAL FRACTURE, RIGHT TIBIAL PLATEAU FRACTURE',\n",
    "# LEFT FEMORAL FRACTURE,\n",
    "# LEFT FEMORAL PSEUDOANEURYSM,\n",
    "# LEFT FEMUR FRACTURE,\n",
    "# LEFT FINGER CELLULITIS, INFECTION,\n",
    "# LEFT FOOT ULCER,\n",
    "# LEFT FRONTAL HEMORRHAGE,\n",
    "# LEFT HIP DISLOCATION,\n",
    "# LEFT HIP FRACTURE,\n",
    "# LEFT HIP FRACTURE/R/O MI,\n",
    "# LEFT HIP FRACTURE;LUNG CANCER,\n",
    "# 'LEFT HUMERUS FRACTURE',LEFT KNEE PAIN;WEAKNESS,\n",
    "# LEFT LEG AMPUTATION,\n",
    "# LEFT LEG CELLULIIS;SEPSIS,\n",
    "# 'LEFT LEG CELLULITIS',\n",
    "# LEFT LEG INFECTION;S/P CABG;EKG CHANGES,\n",
    "# LEFT LEG NUMBNESS,\n",
    "# 'LEFT LEG WEAKNESS','LEFT PARALYSIS','LEFT RADIUS FRACTURE',\n",
    "# LEFT SIDED WEAKNESS,\n",
    "# 'LEFT SIDED WEAKNESS;2/2 BRAIN MASS','LEFT TALUS FRACTURE',\n",
    "# LEFT TIBIA AND FIBULA FRACTURE,\n",
    "# LEFT TIBIAL FRACTURE,\n",
    "# LEG CELLILITIS,\n",
    "# LEG PAIN,\n",
    "# LEG REPAIR S/P MVA,\n",
    "# LEG SWELLING,\n",
    "# LEG WEAKNESS,\n",
    "# LEG WEAKNESS AND SHORTNESS OF BREATH,\n",
    "# LEG WEAKNESS;URINARY TRACT INFECTION,\n",
    "# LETHARGY,\n",
    "# LEUKEMIA,\n",
    "# LEUKOCYTOSIS,\n",
    "# LGIB,\n",
    "# LINE INFECTION,\n",
    "# LINE SEPSIS,\n",
    "# LITHIUM OVERDOSE,\n",
    "# 'LITHIUM TOXICITY ',\n",
    "# LOW BACK PAIN,\n",
    "# LOW HCT,\n",
    "# LOW URINE OUTPUT,\n",
    "# LOWER,\n",
    "# LOWER BACK PAIN,\n",
    "# LOWER EXTREMETY EDEMA;CONGESTIVE HEART FAILURE,\n",
    "# LOWER EXTREMITY  PARALYSIS,\n",
    "# LOWER EXTREMITY EDEMA,\n",
    "# 'LOWER EXTREMITY WEAKNESS','LRG PAIN','LT FEMUR FRACTURE',\n",
    "# LT FEMUR FRACTURE;PULMONARY EDEMA,\n",
    "# LT FINGER AMPUTATION,\n",
    "# LT HIP FRACTURE,\n",
    "# LT HIP PAIN,\n",
    "# LT LOWER LOBE PNEUMONIA,\n",
    "# 'LUMBAR ',\n",
    "# LUMBAR DISC HERNIATION,\n",
    "# LUMBAR FRACTURE,\n",
    "# 'LUMBAR SPINE COMPRESSION','LYME CARDITIS,CONGESTIVE HEART FAILURE',\n",
    "# LYME RIDICULITIS,\n",
    "# LYMPHADENOPATHY;NECK MASS,\n",
    "# 'LYMPHOMA','MALARIA','MANDIBLE FX;HIP FX; WRIST FX','MANDIBULAR FRACTURE','MAST CELL DEGRANULATION',\n",
    "# 'MASTOCYTOSIS','MELENA',\n",
    "# 'MELENA;TELEMETRY','MENTAL STATUS CHANGE',\n",
    "# 'MENTAL STATUS CHANGES',\n",
    "# 'MENTAL STATUS CHANGES, RHABDO, ARF',\n",
    "# 'MENTAL STATUS CHANGES-HYPOXIA',\n",
    "# 'MENTAL STATUS CHANGES;HYPOGLYCEMIA',\n",
    "# ]), 'AdmitDiagnosis'] = \"OTHER\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## One Hot Encode"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [],
   "source": [
    "simple_data_dummified = pd.get_dummies(simple_data, prefix=to_dummify_simple, columns=to_dummify_simple)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>LOSdays</th>\n",
       "      <th>NumDiagnosisByDay</th>\n",
       "      <th>NumProcsByDay</th>\n",
       "      <th>NumCPTeventsByDay</th>\n",
       "      <th>NumInputByDay</th>\n",
       "      <th>NumLabsByDay</th>\n",
       "      <th>NumMicroLabsByDay</th>\n",
       "      <th>NumNotesByDay</th>\n",
       "      <th>NumOutputByDay</th>\n",
       "      <th>NumRxByDay</th>\n",
       "      <th>NumChartEventsByDay</th>\n",
       "      <th>TotalNumInteractByDay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "      <td>17790.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.542721</td>\n",
       "      <td>59.845868</td>\n",
       "      <td>9.573891</td>\n",
       "      <td>0.584353</td>\n",
       "      <td>0.097386</td>\n",
       "      <td>0.225033</td>\n",
       "      <td>3.778828</td>\n",
       "      <td>9.849649</td>\n",
       "      <td>0.225873</td>\n",
       "      <td>0.473518</td>\n",
       "      <td>1.233655</td>\n",
       "      <td>2.007243</td>\n",
       "      <td>100.193063</td>\n",
       "      <td>119.030870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.498186</td>\n",
       "      <td>17.237928</td>\n",
       "      <td>8.540678</td>\n",
       "      <td>1.399089</td>\n",
       "      <td>0.249859</td>\n",
       "      <td>0.353661</td>\n",
       "      <td>6.087294</td>\n",
       "      <td>16.006624</td>\n",
       "      <td>0.475779</td>\n",
       "      <td>2.475450</td>\n",
       "      <td>1.857088</td>\n",
       "      <td>3.368053</td>\n",
       "      <td>160.594029</td>\n",
       "      <td>186.675588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.002557</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000144</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.213242</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>49.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.075092</td>\n",
       "      <td>0.009785</td>\n",
       "      <td>0.039351</td>\n",
       "      <td>0.700198</td>\n",
       "      <td>2.922694</td>\n",
       "      <td>0.026101</td>\n",
       "      <td>0.015060</td>\n",
       "      <td>0.279771</td>\n",
       "      <td>0.456800</td>\n",
       "      <td>24.242433</td>\n",
       "      <td>30.809805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>60.000000</td>\n",
       "      <td>6.880000</td>\n",
       "      <td>0.209538</td>\n",
       "      <td>0.033206</td>\n",
       "      <td>0.127986</td>\n",
       "      <td>1.895960</td>\n",
       "      <td>5.583901</td>\n",
       "      <td>0.083815</td>\n",
       "      <td>0.056842</td>\n",
       "      <td>0.716837</td>\n",
       "      <td>1.063492</td>\n",
       "      <td>54.771185</td>\n",
       "      <td>66.180242</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>74.000000</td>\n",
       "      <td>11.880000</td>\n",
       "      <td>0.555000</td>\n",
       "      <td>0.088421</td>\n",
       "      <td>0.276952</td>\n",
       "      <td>4.405984</td>\n",
       "      <td>10.765582</td>\n",
       "      <td>0.223333</td>\n",
       "      <td>0.214099</td>\n",
       "      <td>1.457137</td>\n",
       "      <td>2.272208</td>\n",
       "      <td>110.875430</td>\n",
       "      <td>131.951395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>88.000000</td>\n",
       "      <td>69.790000</td>\n",
       "      <td>34.023810</td>\n",
       "      <td>6.851852</td>\n",
       "      <td>6.851852</td>\n",
       "      <td>97.253012</td>\n",
       "      <td>408.166667</td>\n",
       "      <td>8.198347</td>\n",
       "      <td>88.925926</td>\n",
       "      <td>27.425926</td>\n",
       "      <td>74.310345</td>\n",
       "      <td>2835.543478</td>\n",
       "      <td>3234.173913</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             gender           age       LOSdays  NumDiagnosisByDay  \\\n",
       "count  17790.000000  17790.000000  17790.000000       17790.000000   \n",
       "mean       0.542721     59.845868      9.573891           0.584353   \n",
       "std        0.498186     17.237928      8.540678           1.399089   \n",
       "min        0.000000     18.000000      0.420000           0.002557   \n",
       "25%        0.000000     49.000000      4.000000           0.075092   \n",
       "50%        1.000000     60.000000      6.880000           0.209538   \n",
       "75%        1.000000     74.000000     11.880000           0.555000   \n",
       "max        1.000000     88.000000     69.790000          34.023810   \n",
       "\n",
       "       NumProcsByDay  NumCPTeventsByDay  NumInputByDay  NumLabsByDay  \\\n",
       "count   17790.000000       17790.000000   17790.000000  17790.000000   \n",
       "mean        0.097386           0.225033       3.778828      9.849649   \n",
       "std         0.249859           0.353661       6.087294     16.006624   \n",
       "min         0.000000           0.000000       0.000000      0.000000   \n",
       "25%         0.009785           0.039351       0.700198      2.922694   \n",
       "50%         0.033206           0.127986       1.895960      5.583901   \n",
       "75%         0.088421           0.276952       4.405984     10.765582   \n",
       "max         6.851852           6.851852      97.253012    408.166667   \n",
       "\n",
       "       NumMicroLabsByDay  NumNotesByDay  NumOutputByDay    NumRxByDay  \\\n",
       "count       17790.000000   17790.000000    17790.000000  17790.000000   \n",
       "mean            0.225873       0.473518        1.233655      2.007243   \n",
       "std             0.475779       2.475450        1.857088      3.368053   \n",
       "min             0.000000       0.000144        0.000000      0.000000   \n",
       "25%             0.026101       0.015060        0.279771      0.456800   \n",
       "50%             0.083815       0.056842        0.716837      1.063492   \n",
       "75%             0.223333       0.214099        1.457137      2.272208   \n",
       "max             8.198347      88.925926       27.425926     74.310345   \n",
       "\n",
       "       NumChartEventsByDay  TotalNumInteractByDay  \n",
       "count         17790.000000           17790.000000  \n",
       "mean            100.193063             119.030870  \n",
       "std             160.594029             186.675588  \n",
       "min               0.000000               0.213242  \n",
       "25%              24.242433              30.809805  \n",
       "50%              54.771185              66.180242  \n",
       "75%             110.875430             131.951395  \n",
       "max            2835.543478            3234.173913  "
      ]
     },
     "execution_count": 217,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('max_columns', None)\n",
    "\n",
    "# 'age', 'LOSdays', 'NumDiagnosisByDay','NumProcsByDay',\n",
    "#                               'NumCPTeventsByDay', 'NumInputByDay', 'NumLabsByDay', 'NumMicroLabsByDay',\n",
    "# 'NumNotesByDay', 'NumOutputByDay', 'NumRxByDay', 'NumTransfersByDay',\n",
    "# simple_data_continuous = simple_data_dummified[[ \n",
    "       \n",
    "#        'NumChartEventsByDay','TotalNumInteractByDay']]\n",
    "simple_data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "test differences between population"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "age:  RanksumsResult(statistic=-5.911792511687827, pvalue=3.3840465270801e-09)\n",
      "LOSdays:  RanksumsResult(statistic=-2.3038649771312776, pvalue=0.021230224247779872)\n",
      "NumCPTeventsByDay:  RanksumsResult(statistic=-11.1632851140381, pvalue=6.166980181005684e-29)\n",
      "NumInputByDay:  RanksumsResult(statistic=11.960097390372642, pvalue=5.749445045381141e-33)\n",
      "NumLabsByDay:  RanksumsResult(statistic=2.316377500193698, pvalue=0.020537666146626426)\n",
      "NumMicroLabsByDay:  RanksumsResult(statistic=-4.725246978556025, pvalue=2.298358458567412e-06)\n",
      "NumNotesByDay:  RanksumsResult(statistic=3.535796170838605, pvalue=0.00040654824340618747)\n",
      "NumOutputByDay:  RanksumsResult(statistic=10.530767448215, pvalue=6.23242649576322e-26)\n",
      "NumRxByDay:  RanksumsResult(statistic=-6.433441819318081, pvalue=1.2474635798874134e-10)\n",
      "NumChartEventsByDay:  RanksumsResult(statistic=7.4454724538701775, pvalue=9.659808654104399e-14)\n",
      "TotalNumInteractByDay:  RanksumsResult(statistic=7.168180135087336, pvalue=7.600126041331155e-13)\n",
      "NumProcsByDay:  RanksumsResult(statistic=3.6820312865252802, pvalue=0.00023138300672409306)\n",
      "NumDiagnosisByDay:  RanksumsResult(statistic=-1.4320723318276547, pvalue=0.1521231218263168)\n"
     ]
    }
   ],
   "source": [
    "from scipy.stats import ranksums\n",
    "\n",
    "print(\"age: \", ranksums(nooutlier_missing.age,simple_data.age))\n",
    "print(\"LOSdays: \", ranksums(nooutlier_missing.LOSdays,simple_data.LOSdays))\n",
    "# print(\"callout: \", ranksums(raw_er_missing_alive.NumCalloutsByDay,simple_data_dummified.NumCalloutsByDay))\n",
    "\n",
    "print(\"NumCPTeventsByDay: \", ranksums(nooutlier_missing.NumCPTeventsByDay,simple_data.NumCPTeventsByDay))\n",
    "print(\"NumInputByDay: \", ranksums(nooutlier_missing.NumInputByDay,simple_data.NumInputByDay))\n",
    "print(\"NumLabsByDay: \", ranksums(nooutlier_missing.NumLabsByDay,simple_data.NumLabsByDay))\n",
    "print(\"NumMicroLabsByDay: \", ranksums(nooutlier_missing.NumMicroLabsByDay,simple_data.NumMicroLabsByDay))\n",
    "print(\"NumNotesByDay: \", ranksums(nooutlier_missing.NumNotesByDay,simple_data.NumNotesByDay))\n",
    "print(\"NumOutputByDay: \", ranksums(nooutlier_missing.NumOutputByDay,simple_data.NumOutputByDay))\n",
    "print(\"NumRxByDay: \", ranksums(nooutlier_missing.NumRxByDay,simple_data.NumRxByDay))\n",
    "# print(\"procevents: \", ranksums(raw_er_missing_alive.NumProcEventsByDay,simple_data_dummified.NumProcEventsByDay))\n",
    "print(\"NumChartEventsByDay: \", ranksums(nooutlier_missing.NumChartEventsByDay,simple_data.NumChartEventsByDay))\n",
    "print(\"TotalNumInteractByDay: \", ranksums(nooutlier_missing.TotalNumInteractByDay,simple_data.TotalNumInteractByDay))\n",
    "print(\"NumProcsByDay: \", ranksums(nooutlier_missing.NumProcsByDay,simple_data.NumProcsByDay))\n",
    "\n",
    "\n",
    "# Nonsig:\n",
    "print(\"NumDiagnosisByDay: \", ranksums(nooutlier_missing.NumDiagnosisByDay,simple_data.NumDiagnosisByDay))\n",
    "# print(\"NumTransfersByDay: \", ranksums(nooutlier_missing.NumTransfersByDay,simple_data.NumTransfersByDay))\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing:  55.0\n",
      "Final data:  60.0\n"
     ]
    }
   ],
   "source": [
    "print(\"Missing: \", np.median(nooutlier_missing.age))\n",
    "print(\"Final data: \", np.median(simple_data.age))\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ 87., 132., 127., 130.,  83.,  71.,  72.,  41.,  38.,  33.,  30.,\n",
       "         14.,  18.,  22.,  17.,  14.,  16.,   7.,  13.,   8.,   7.,   7.,\n",
       "          3.,   1.,   8.,   5.,   0.,   1.,   2.,   1.,   3.,   0.,   0.,\n",
       "          1.,   2.,   0.,   0.,   0.,   0.,   1.,   1.,   1.,   1.,   1.,\n",
       "          1.,   0.,   0.,   1.,   0.,   1.]),\n",
       " array([ 0.38  ,  1.7666,  3.1532,  4.5398,  5.9264,  7.313 ,  8.6996,\n",
       "        10.0862, 11.4728, 12.8594, 14.246 , 15.6326, 17.0192, 18.4058,\n",
       "        19.7924, 21.179 , 22.5656, 23.9522, 25.3388, 26.7254, 28.112 ,\n",
       "        29.4986, 30.8852, 32.2718, 33.6584, 35.045 , 36.4316, 37.8182,\n",
       "        39.2048, 40.5914, 41.978 , 43.3646, 44.7512, 46.1378, 47.5244,\n",
       "        48.911 , 50.2976, 51.6842, 53.0708, 54.4574, 55.844 , 57.2306,\n",
       "        58.6172, 60.0038, 61.3904, 62.777 , 64.1636, 65.5502, 66.9368,\n",
       "        68.3234, 69.71  ]),\n",
       " <a list of 50 Patch objects>)"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plt.hist(raw_er_missing_alive.age, bins=50)\n",
    "# plt.title(\"Missing Data Age\")\n",
    "\n",
    "plt.hist(nooutlier_missing.LOSdays, bins=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([7.540e+02, 2.092e+03, 2.396e+03, 2.483e+03, 1.711e+03, 1.513e+03,\n",
       "        1.304e+03, 8.200e+02, 7.690e+02, 5.770e+02, 4.990e+02, 4.330e+02,\n",
       "        3.320e+02, 3.230e+02, 2.500e+02, 1.800e+02, 2.040e+02, 1.460e+02,\n",
       "        1.310e+02, 1.200e+02, 1.050e+02, 7.600e+01, 6.800e+01, 5.800e+01,\n",
       "        5.600e+01, 4.100e+01, 3.600e+01, 3.800e+01, 2.500e+01, 4.100e+01,\n",
       "        2.500e+01, 2.400e+01, 1.300e+01, 1.400e+01, 2.400e+01, 1.200e+01,\n",
       "        1.200e+01, 1.000e+01, 1.300e+01, 7.000e+00, 1.300e+01, 1.200e+01,\n",
       "        6.000e+00, 5.000e+00, 3.000e+00, 3.000e+00, 4.000e+00, 1.000e+00,\n",
       "        3.000e+00, 5.000e+00]),\n",
       " array([ 0.42  ,  1.8074,  3.1948,  4.5822,  5.9696,  7.357 ,  8.7444,\n",
       "        10.1318, 11.5192, 12.9066, 14.294 , 15.6814, 17.0688, 18.4562,\n",
       "        19.8436, 21.231 , 22.6184, 24.0058, 25.3932, 26.7806, 28.168 ,\n",
       "        29.5554, 30.9428, 32.3302, 33.7176, 35.105 , 36.4924, 37.8798,\n",
       "        39.2672, 40.6546, 42.042 , 43.4294, 44.8168, 46.2042, 47.5916,\n",
       "        48.979 , 50.3664, 51.7538, 53.1412, 54.5286, 55.916 , 57.3034,\n",
       "        58.6908, 60.0782, 61.4656, 62.853 , 64.2404, 65.6278, 67.0152,\n",
       "        68.4026, 69.79  ]),\n",
       " <a list of 50 Patch objects>)"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plt.hist(simple_data.age, bins=50)\n",
    "# plt.title(\"Age\")\n",
    "\n",
    "plt.hist(simple_data.LOSdays, bins=50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Categorical (used chi-square test in excel):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# a = pd.Categorical(nooutlier_raw_missing_alive.gender, categories=[0,1])\n",
    "# b = pd.Categorical(nooutlier_simple_data.gender, categories=[0,1])\n",
    "# from scipy.stats import chi2_contingency\n",
    "# obs = np.array([[10, 10, 20], [20, 20, 20]])\n",
    "# chi2_contingency(obs)\n",
    "\n",
    "# pd.crosstab(nooutlier_raw_missing_alive['insurance'], nooutlier_simple_data['insurance'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gender                               int64\n",
       "age                                  int64\n",
       "LOSdays                            float64\n",
       "NumDiagnosisByDay                  float64\n",
       "NumProcsByDay                      float64\n",
       "NumCPTeventsByDay                  float64\n",
       "NumInputByDay                      float64\n",
       "NumLabsByDay                       float64\n",
       "NumMicroLabsByDay                  float64\n",
       "NumNotesByDay                      float64\n",
       "NumOutputByDay                     float64\n",
       "NumRxByDay                         float64\n",
       "NumChartEventsByDay                float64\n",
       "TotalNumInteractByDay              float64\n",
       "insurance_INSURED                    uint8\n",
       "insurance_UNINSURED                  uint8\n",
       "religion_RELIGIOUS                   uint8\n",
       "religion_UNKNOWN                     uint8\n",
       "marital_status_MARRIED               uint8\n",
       "marital_status_UNKNOWN               uint8\n",
       "marital_status_UNMARRIED             uint8\n",
       "ethnicity_ASIAN_PACIFICISLANDER      uint8\n",
       "ethnicity_BLACK                      uint8\n",
       "ethnicity_HISPANIC OR LATINO         uint8\n",
       "ethnicity_OTHER                      uint8\n",
       "ethnicity_UNKNOWN                    uint8\n",
       "ethnicity_WHITE                      uint8\n",
       "dtype: object"
      ]
     },
     "execution_count": 223,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "simple_data_dummified.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Modeling:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = simple_data_dummified.LOSdays\n",
    "X = simple_data_dummified.loc[:, simple_data_dummified.columns != 'LOSdays']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn import metrics\n",
    "import seaborn as sns\n",
    "\n",
    "def plot_feature_importance(importance,names,model_type):\n",
    "\n",
    "#Create arrays from feature importance and feature names\n",
    "    feature_importance = np.array(importance)\n",
    "    feature_names = np.array(names)\n",
    "\n",
    "    #Create a DataFrame using a Dictionary\n",
    "    data={'feature_names':feature_names,'feature_importance':feature_importance}\n",
    "    fi_df = pd.DataFrame(data)\n",
    "\n",
    "    #Sort the DataFrame in order decreasing feature importance\n",
    "    fi_df.sort_values(by=['feature_importance'], ascending=False,inplace=True)\n",
    "\n",
    "    #Define size of bar plot\n",
    "    plt.figure(figsize=(10,8))\n",
    "    #Plot Searborn bar chart\n",
    "    sns.barplot(x=fi_df['feature_importance'], y=fi_df['feature_names'])\n",
    "    #Add chart labels\n",
    "    plt.title(model_type + 'FEATURE IMPORTANCE')\n",
    "    plt.xlabel('FEATURE IMPORTANCE')\n",
    "    plt.ylabel('FEATURE NAMES')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# XGBoost:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- 1.2815141677856445 seconds ---\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start_time = time.time()\n",
    "\n",
    "from xgboost import XGBRegressor, plot_importance\n",
    "\n",
    "xgbmodel = XGBRegressor()\n",
    "xgbmodel.fit(X_train, y_train)\n",
    "\n",
    "print(\"--- %s seconds ---\" % (time.time() - start_time))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error XGBoost:  1.2596824058459808\n",
      "Mean Squared Error XGBoost:  4.409099803260259\n",
      "Root Mean Squared Error XGBoost: 2.099785656504077\n",
      "Explained Variance Score XGBoost: 0.9368008313513144\n"
     ]
    }
   ],
   "source": [
    "# make predictions for test data\n",
    "y_pred_xgb = xgbmodel.predict(X_test)\n",
    "predictions_xgb = [round(value) for value in y_pred_xgb ]\n",
    "\n",
    "mae_xgb = metrics.mean_absolute_error(y_test, predictions_xgb)\n",
    "mse_xgb = metrics.mean_squared_error(y_test, predictions_xgb)\n",
    "print(\"Mean Absolute Error XGBoost: \", mae_xgb)\n",
    "print(\"Mean Squared Error XGBoost: \", mse_xgb)\n",
    "print('Root Mean Squared Error XGBoost:', np.sqrt(metrics.mean_squared_error(y_test, predictions_xgb)))\n",
    "print(\"Explained Variance Score XGBoost:\", metrics.explained_variance_score(predictions_xgb, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "88.92592592592592"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(simple_data_dummified.NumNotesByDay)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "features_xgb = simple_data_dummified.loc[:, simple_data_dummified.columns != 'LOSdays'].columns\n",
    "\n",
    "plot_feature_importance(xgbmodel.feature_importances_,features_xgb,'XG BOOST ')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[7.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 17.0,\n",
       " 6.0,\n",
       " 8.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 25.0,\n",
       " 4.0,\n",
       " 8.0,\n",
       " 8.0,\n",
       " 13.0,\n",
       " 6.0,\n",
       " 5.0,\n",
       " 38.0,\n",
       " 62.0,\n",
       " 7.0,\n",
       " 22.0,\n",
       " 6.0,\n",
       " 8.0,\n",
       " 7.0,\n",
       " 4.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 38.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 10.0,\n",
       " 10.0,\n",
       " 2.0,\n",
       " 17.0,\n",
       " 3.0,\n",
       " 8.0,\n",
       " 6.0,\n",
       " 10.0,\n",
       " 39.0,\n",
       " 3.0,\n",
       " 6.0,\n",
       " 19.0,\n",
       " 10.0,\n",
       " 27.0,\n",
       " 10.0,\n",
       " 7.0,\n",
       " 8.0,\n",
       " 10.0,\n",
       " 7.0,\n",
       " 27.0,\n",
       " 6.0,\n",
       " 2.0,\n",
       " 15.0,\n",
       " 3.0,\n",
       " 15.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 17.0,\n",
       " 7.0,\n",
       " 10.0,\n",
       " 14.0,\n",
       " 4.0,\n",
       " 8.0,\n",
       " 14.0,\n",
       " 7.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 12.0,\n",
       " 16.0,\n",
       " 6.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 12.0,\n",
       " 8.0,\n",
       " 24.0,\n",
       " 17.0,\n",
       " 13.0,\n",
       " 4.0,\n",
       " 13.0,\n",
       " 21.0,\n",
       " 12.0,\n",
       " 6.0,\n",
       " 18.0,\n",
       " 2.0,\n",
       " 16.0,\n",
       " 11.0,\n",
       " 9.0,\n",
       " 46.0,\n",
       " 4.0,\n",
       " 14.0,\n",
       " 15.0,\n",
       " 4.0,\n",
       " 6.0,\n",
       " 5.0,\n",
       " 6.0,\n",
       " 9.0,\n",
       " 7.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 7.0,\n",
       " 7.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 3.0,\n",
       " 11.0,\n",
       " 8.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 11.0,\n",
       " 8.0,\n",
       " 28.0,\n",
       " 9.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 6.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 33.0,\n",
       " 28.0,\n",
       " 2.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 11.0,\n",
       " 3.0,\n",
       " 16.0,\n",
       " 10.0,\n",
       " 12.0,\n",
       " 1.0,\n",
       " 11.0,\n",
       " 8.0,\n",
       " 20.0,\n",
       " 14.0,\n",
       " 2.0,\n",
       " 5.0,\n",
       " 1.0,\n",
       " 9.0,\n",
       " 19.0,\n",
       " 16.0,\n",
       " 12.0,\n",
       " 4.0,\n",
       " 25.0,\n",
       " 4.0,\n",
       " 12.0,\n",
       " 1.0,\n",
       " 5.0,\n",
       " 7.0,\n",
       " 16.0,\n",
       " 9.0,\n",
       " 11.0,\n",
       " 21.0,\n",
       " 26.0,\n",
       " 2.0,\n",
       " 2.0,\n",
       " 10.0,\n",
       " 10.0,\n",
       " 10.0,\n",
       " 15.0,\n",
       " 1.0,\n",
       " 2.0,\n",
       " 15.0,\n",
       " 2.0,\n",
       " 10.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 10.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 21.0,\n",
       " 10.0,\n",
       " 9.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 9.0,\n",
       " 10.0,\n",
       " 13.0,\n",
       " 7.0,\n",
       " 11.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 18.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 2.0,\n",
       " 11.0,\n",
       " 44.0,\n",
       " 8.0,\n",
       " 21.0,\n",
       " 17.0,\n",
       " 23.0,\n",
       " 7.0,\n",
       " 5.0,\n",
       " 19.0,\n",
       " 11.0,\n",
       " 19.0,\n",
       " 8.0,\n",
       " 7.0,\n",
       " 25.0,\n",
       " 2.0,\n",
       " 11.0,\n",
       " 21.0,\n",
       " 12.0,\n",
       " 3.0,\n",
       " 44.0,\n",
       " 5.0,\n",
       " 9.0,\n",
       " 7.0,\n",
       " 12.0,\n",
       " 10.0,\n",
       " 1.0,\n",
       " 17.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 11.0,\n",
       " 5.0,\n",
       " 9.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 9.0,\n",
       " 5.0,\n",
       " 3.0,\n",
       " 11.0,\n",
       " 11.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 14.0,\n",
       " 4.0,\n",
       " 7.0,\n",
       " 12.0,\n",
       " 11.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 12.0,\n",
       " 10.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 6.0,\n",
       " 2.0,\n",
       " 8.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 7.0,\n",
       " 14.0,\n",
       " 13.0,\n",
       " 6.0,\n",
       " 5.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 10.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 7.0,\n",
       " 13.0,\n",
       " 6.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 19.0,\n",
       " 2.0,\n",
       " 9.0,\n",
       " 8.0,\n",
       " 2.0,\n",
       " 12.0,\n",
       " 43.0,\n",
       " 9.0,\n",
       " 7.0,\n",
       " 2.0,\n",
       " 10.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 8.0,\n",
       " 5.0,\n",
       " 16.0,\n",
       " 32.0,\n",
       " 6.0,\n",
       " 2.0,\n",
       " 26.0,\n",
       " 18.0,\n",
       " 8.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 21.0,\n",
       " 11.0,\n",
       " 6.0,\n",
       " 14.0,\n",
       " 3.0,\n",
       " 8.0,\n",
       " 1.0,\n",
       " 10.0,\n",
       " 4.0,\n",
       " 8.0,\n",
       " 13.0,\n",
       " 5.0,\n",
       " 2.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 2.0,\n",
       " 8.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 6.0,\n",
       " 23.0,\n",
       " 22.0,\n",
       " 29.0,\n",
       " 13.0,\n",
       " 4.0,\n",
       " 13.0,\n",
       " 4.0,\n",
       " 35.0,\n",
       " 4.0,\n",
       " 11.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 8.0,\n",
       " 6.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 9.0,\n",
       " 15.0,\n",
       " 5.0,\n",
       " 26.0,\n",
       " 6.0,\n",
       " 8.0,\n",
       " 3.0,\n",
       " 9.0,\n",
       " 17.0,\n",
       " 32.0,\n",
       " 2.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 4.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 20.0,\n",
       " 4.0,\n",
       " 8.0,\n",
       " 8.0,\n",
       " 28.0,\n",
       " 5.0,\n",
       " 15.0,\n",
       " 7.0,\n",
       " 3.0,\n",
       " 13.0,\n",
       " 3.0,\n",
       " 13.0,\n",
       " 7.0,\n",
       " 15.0,\n",
       " 6.0,\n",
       " 8.0,\n",
       " 7.0,\n",
       " 8.0,\n",
       " 30.0,\n",
       " 34.0,\n",
       " 19.0,\n",
       " 9.0,\n",
       " 9.0,\n",
       " 2.0,\n",
       " 10.0,\n",
       " 4.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 16.0,\n",
       " 5.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 12.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 6.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 22.0,\n",
       " 14.0,\n",
       " 5.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 21.0,\n",
       " 9.0,\n",
       " 5.0,\n",
       " 21.0,\n",
       " 6.0,\n",
       " 10.0,\n",
       " 2.0,\n",
       " 6.0,\n",
       " 2.0,\n",
       " 6.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 32.0,\n",
       " 11.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 7.0,\n",
       " 7.0,\n",
       " 2.0,\n",
       " 7.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 11.0,\n",
       " 24.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 34.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 15.0,\n",
       " 29.0,\n",
       " 7.0,\n",
       " 7.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 15.0,\n",
       " 11.0,\n",
       " 14.0,\n",
       " 4.0,\n",
       " 7.0,\n",
       " 36.0,\n",
       " 38.0,\n",
       " 8.0,\n",
       " 10.0,\n",
       " 20.0,\n",
       " 7.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 10.0,\n",
       " 43.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 1.0,\n",
       " 3.0,\n",
       " 13.0,\n",
       " 5.0,\n",
       " 6.0,\n",
       " 6.0,\n",
       " 6.0,\n",
       " 16.0,\n",
       " 13.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 12.0,\n",
       " 5.0,\n",
       " 10.0,\n",
       " 5.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 9.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 10.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 7.0,\n",
       " 8.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 4.0,\n",
       " 32.0,\n",
       " 14.0,\n",
       " 10.0,\n",
       " 6.0,\n",
       " 10.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 3.0,\n",
       " 20.0,\n",
       " 9.0,\n",
       " 10.0,\n",
       " 41.0,\n",
       " 14.0,\n",
       " 8.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 10.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 17.0,\n",
       " 6.0,\n",
       " 7.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 14.0,\n",
       " 3.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 2.0,\n",
       " 6.0,\n",
       " 31.0,\n",
       " 4.0,\n",
       " 14.0,\n",
       " 6.0,\n",
       " 1.0,\n",
       " 10.0,\n",
       " 15.0,\n",
       " 9.0,\n",
       " 6.0,\n",
       " 14.0,\n",
       " 5.0,\n",
       " 18.0,\n",
       " 3.0,\n",
       " 32.0,\n",
       " 17.0,\n",
       " 11.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 10.0,\n",
       " 15.0,\n",
       " 11.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 7.0,\n",
       " 7.0,\n",
       " 11.0,\n",
       " 13.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 6.0,\n",
       " 6.0,\n",
       " 12.0,\n",
       " 4.0,\n",
       " 37.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 8.0,\n",
       " 18.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 9.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 6.0,\n",
       " 11.0,\n",
       " 14.0,\n",
       " 14.0,\n",
       " 9.0,\n",
       " 4.0,\n",
       " 6.0,\n",
       " 8.0,\n",
       " 7.0,\n",
       " 2.0,\n",
       " 19.0,\n",
       " 28.0,\n",
       " 15.0,\n",
       " 12.0,\n",
       " 4.0,\n",
       " 10.0,\n",
       " 11.0,\n",
       " 4.0,\n",
       " 11.0,\n",
       " 13.0,\n",
       " 2.0,\n",
       " 7.0,\n",
       " 27.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 16.0,\n",
       " 2.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 6.0,\n",
       " 0.0,\n",
       " 6.0,\n",
       " 6.0,\n",
       " 10.0,\n",
       " 11.0,\n",
       " 8.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 8.0,\n",
       " 7.0,\n",
       " 27.0,\n",
       " 3.0,\n",
       " 9.0,\n",
       " 10.0,\n",
       " 25.0,\n",
       " 10.0,\n",
       " 19.0,\n",
       " 6.0,\n",
       " 7.0,\n",
       " 33.0,\n",
       " 6.0,\n",
       " 7.0,\n",
       " 5.0,\n",
       " 17.0,\n",
       " 3.0,\n",
       " 13.0,\n",
       " 5.0,\n",
       " 21.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 10.0,\n",
       " 6.0,\n",
       " 27.0,\n",
       " 9.0,\n",
       " 17.0,\n",
       " 12.0,\n",
       " 6.0,\n",
       " 7.0,\n",
       " 11.0,\n",
       " 14.0,\n",
       " 7.0,\n",
       " 11.0,\n",
       " 24.0,\n",
       " 5.0,\n",
       " 8.0,\n",
       " 8.0,\n",
       " 14.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 13.0,\n",
       " 3.0,\n",
       " 9.0,\n",
       " 3.0,\n",
       " 2.0,\n",
       " 13.0,\n",
       " 12.0,\n",
       " 27.0,\n",
       " 1.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 3.0,\n",
       " 9.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 13.0,\n",
       " 12.0,\n",
       " 6.0,\n",
       " 15.0,\n",
       " 5.0,\n",
       " 25.0,\n",
       " 2.0,\n",
       " 11.0,\n",
       " 10.0,\n",
       " 16.0,\n",
       " 30.0,\n",
       " 10.0,\n",
       " 3.0,\n",
       " 6.0,\n",
       " 10.0,\n",
       " 12.0,\n",
       " 15.0,\n",
       " 7.0,\n",
       " 2.0,\n",
       " 16.0,\n",
       " 7.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 1.0,\n",
       " 6.0,\n",
       " 1.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 23.0,\n",
       " 5.0,\n",
       " 20.0,\n",
       " 36.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 24.0,\n",
       " 5.0,\n",
       " 9.0,\n",
       " 1.0,\n",
       " 14.0,\n",
       " 6.0,\n",
       " 5.0,\n",
       " 9.0,\n",
       " 12.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 6.0,\n",
       " 17.0,\n",
       " 5.0,\n",
       " 8.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 10.0,\n",
       " 2.0,\n",
       " 21.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 29.0,\n",
       " 15.0,\n",
       " 5.0,\n",
       " 11.0,\n",
       " 20.0,\n",
       " 12.0,\n",
       " 12.0,\n",
       " 9.0,\n",
       " 9.0,\n",
       " 5.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 2.0,\n",
       " 18.0,\n",
       " 5.0,\n",
       " 16.0,\n",
       " 23.0,\n",
       " 12.0,\n",
       " 15.0,\n",
       " 31.0,\n",
       " 2.0,\n",
       " 12.0,\n",
       " 4.0,\n",
       " 15.0,\n",
       " 7.0,\n",
       " 11.0,\n",
       " 13.0,\n",
       " 14.0,\n",
       " 18.0,\n",
       " 3.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 11.0,\n",
       " 2.0,\n",
       " 10.0,\n",
       " 4.0,\n",
       " 11.0,\n",
       " 11.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 4.0,\n",
       " 7.0,\n",
       " 38.0,\n",
       " 8.0,\n",
       " 7.0,\n",
       " 9.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 9.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 14.0,\n",
       " 11.0,\n",
       " 17.0,\n",
       " 4.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 12.0,\n",
       " 18.0,\n",
       " 17.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 8.0,\n",
       " 8.0,\n",
       " 4.0,\n",
       " 2.0,\n",
       " 39.0,\n",
       " 8.0,\n",
       " 16.0,\n",
       " 21.0,\n",
       " 19.0,\n",
       " 4.0,\n",
       " 1.0,\n",
       " 8.0,\n",
       " 7.0,\n",
       " 7.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 6.0,\n",
       " 14.0,\n",
       " 15.0,\n",
       " 11.0,\n",
       " 31.0,\n",
       " 16.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 17.0,\n",
       " 13.0,\n",
       " 1.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 2.0,\n",
       " 12.0,\n",
       " 11.0,\n",
       " 5.0,\n",
       " 29.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 12.0,\n",
       " 4.0,\n",
       " 22.0,\n",
       " 3.0,\n",
       " 4.0,\n",
       " 14.0,\n",
       " 6.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 15.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 2.0,\n",
       " 24.0,\n",
       " 12.0,\n",
       " 32.0,\n",
       " 10.0,\n",
       " 7.0,\n",
       " 5.0,\n",
       " 5.0,\n",
       " 16.0,\n",
       " 5.0,\n",
       " 19.0,\n",
       " 3.0,\n",
       " 43.0,\n",
       " 7.0,\n",
       " 5.0,\n",
       " 23.0,\n",
       " 4.0,\n",
       " 5.0,\n",
       " 9.0,\n",
       " 9.0,\n",
       " 18.0,\n",
       " 14.0,\n",
       " 8.0,\n",
       " 33.0,\n",
       " 8.0,\n",
       " 8.0,\n",
       " 10.0,\n",
       " 2.0,\n",
       " 7.0,\n",
       " 7.0,\n",
       " 6.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 8.0,\n",
       " 6.0,\n",
       " 6.0,\n",
       " 1.0,\n",
       " 4.0,\n",
       " 7.0,\n",
       " 3.0,\n",
       " 16.0,\n",
       " 10.0,\n",
       " 6.0,\n",
       " 22.0,\n",
       " 8.0,\n",
       " 7.0,\n",
       " 7.0,\n",
       " 5.0,\n",
       " 4.0,\n",
       " 12.0,\n",
       " 6.0,\n",
       " 3.0,\n",
       " 9.0,\n",
       " 16.0,\n",
       " 14.0,\n",
       " 28.0,\n",
       " 8.0,\n",
       " 3.0,\n",
       " 9.0,\n",
       " 4.0,\n",
       " 11.0,\n",
       " 26.0,\n",
       " 4.0,\n",
       " 6.0,\n",
       " 4.0,\n",
       " 14.0,\n",
       " 8.0,\n",
       " 6.0,\n",
       " 6.0,\n",
       " 14.0,\n",
       " 13.0,\n",
       " 14.0,\n",
       " 10.0,\n",
       " 29.0,\n",
       " 6.0,\n",
       " 3.0,\n",
       " 16.0,\n",
       " 34.0,\n",
       " 5.0,\n",
       " 8.0,\n",
       " 24.0,\n",
       " 3.0,\n",
       " 11.0,\n",
       " 11.0,\n",
       " 26.0,\n",
       " 19.0,\n",
       " 5.0,\n",
       " 6.0,\n",
       " 10.0,\n",
       " 7.0,\n",
       " 15.0,\n",
       " 5.0,\n",
       " 14.0,\n",
       " 10.0,\n",
       " 8.0,\n",
       " 18.0,\n",
       " 3.0,\n",
       " 5.0,\n",
       " 3.0,\n",
       " 2.0,\n",
       " 4.0,\n",
       " 33.0,\n",
       " 47.0,\n",
       " 12.0,\n",
       " 6.0,\n",
       " 54.0,\n",
       " 5.0,\n",
       " 0.0,\n",
       " 13.0,\n",
       " 7.0,\n",
       " 14.0,\n",
       " 10.0,\n",
       " 4.0,\n",
       " 4.0,\n",
       " 9.0,\n",
       " 4.0,\n",
       " 11.0,\n",
       " 17.0,\n",
       " 8.0,\n",
       " 6.0,\n",
       " 6.0,\n",
       " 13.0,\n",
       " 9.0,\n",
       " 5.0,\n",
       " 3.0,\n",
       " 7.0,\n",
       " 44.0,\n",
       " 3.0,\n",
       " 8.0,\n",
       " 5.0,\n",
       " 6.0,\n",
       " 2.0,\n",
       " 12.0,\n",
       " 7.0,\n",
       " 5.0,\n",
       " 12.0,\n",
       " 31.0,\n",
       " 9.0,\n",
       " 18.0,\n",
       " 13.0,\n",
       " 9.0,\n",
       " 10.0,\n",
       " 33.0,\n",
       " 2.0,\n",
       " 14.0,\n",
       " 8.0,\n",
       " 31.0,\n",
       " 8.0,\n",
       " 6.0,\n",
       " 3.0,\n",
       " ...]"
      ]
     },
     "execution_count": 229,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions_xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Baseline MAE is 5.57\n"
     ]
    }
   ],
   "source": [
    "# \"Learn\" the median from the training data\n",
    "median_train = np.median(y_train)\n",
    "# Get predictions on the test set\n",
    "baseline_predictions = np.ones(y_test.shape) * median_train\n",
    "# Compute MAE\n",
    "mae_baseline = metrics.mean_absolute_error(y_test, baseline_predictions)\n",
    "print(\"Baseline MAE is {:.2f}\".format(mae_baseline))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hyperparameter tuning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://blog.cambridgespark.com/hyperparameter-tuning-in-xgboost-4ff9100a3b2f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {\n",
    "    # Parameters that we are going to tune.\n",
    "    'max_depth':6,\n",
    "    'min_child_weight': 1,\n",
    "    'eta':.3,\n",
    "    'subsample': 1,\n",
    "    'colsample_bytree': 1,\n",
    "    # Other parameters\n",
    "    'objective':'reg:linear',\n",
    "}\n",
    "\n",
    "params['eval_metric'] = \"mae\"\n",
    "\n",
    "num_boost_round = 999"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xgb\n",
    "dtrain = xgb.DMatrix(X_train, label=y_train)\n",
    "dtest = xgb.DMatrix(X_test, label=y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[23:06:55] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[0]\tTest-mae:6.45095\n",
      "[1]\tTest-mae:4.51872\n",
      "[2]\tTest-mae:3.21318\n",
      "[3]\tTest-mae:2.37486\n",
      "[4]\tTest-mae:1.86827\n",
      "[5]\tTest-mae:1.57913\n",
      "[6]\tTest-mae:1.42726\n",
      "[7]\tTest-mae:1.34279\n",
      "[8]\tTest-mae:1.29949\n",
      "[9]\tTest-mae:1.27483\n",
      "[10]\tTest-mae:1.26107\n",
      "[11]\tTest-mae:1.25183\n",
      "[12]\tTest-mae:1.24859\n",
      "[13]\tTest-mae:1.24134\n",
      "[14]\tTest-mae:1.23852\n",
      "[15]\tTest-mae:1.23932\n",
      "[16]\tTest-mae:1.23587\n",
      "[17]\tTest-mae:1.23515\n",
      "[18]\tTest-mae:1.23538\n",
      "[19]\tTest-mae:1.23410\n",
      "[20]\tTest-mae:1.23373\n",
      "[21]\tTest-mae:1.23148\n",
      "[22]\tTest-mae:1.22970\n",
      "[23]\tTest-mae:1.23151\n",
      "[24]\tTest-mae:1.23172\n",
      "[25]\tTest-mae:1.22812\n",
      "[26]\tTest-mae:1.22753\n",
      "[27]\tTest-mae:1.22588\n",
      "[28]\tTest-mae:1.22571\n",
      "[29]\tTest-mae:1.22476\n",
      "[30]\tTest-mae:1.22483\n",
      "[31]\tTest-mae:1.22292\n",
      "[32]\tTest-mae:1.22430\n",
      "[33]\tTest-mae:1.22360\n",
      "[34]\tTest-mae:1.22348\n",
      "[35]\tTest-mae:1.22225\n",
      "[36]\tTest-mae:1.22475\n",
      "[37]\tTest-mae:1.22556\n",
      "[38]\tTest-mae:1.22662\n",
      "[39]\tTest-mae:1.22825\n",
      "[40]\tTest-mae:1.22761\n",
      "[41]\tTest-mae:1.22664\n",
      "[42]\tTest-mae:1.22775\n",
      "[43]\tTest-mae:1.22855\n",
      "[44]\tTest-mae:1.22754\n",
      "[45]\tTest-mae:1.22712\n",
      "Best MAE: 1.22 with 36 rounds\n"
     ]
    }
   ],
   "source": [
    "xgbmodel_tuned = xgb.train(\n",
    "    params,\n",
    "    dtrain,\n",
    "    num_boost_round=num_boost_round,\n",
    "    evals=[(dtest, \"Test\")],\n",
    "    early_stopping_rounds=10\n",
    ")\n",
    "\n",
    "print(\"Best MAE: {:.2f} with {} rounds\".format(\n",
    "                 xgbmodel_tuned.best_score,\n",
    "                 xgbmodel_tuned.best_iteration+1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[23:06:56] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:06:56] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:06:56] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:06:56] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:06:56] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\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>train-mae-mean</th>\n",
       "      <th>train-mae-std</th>\n",
       "      <th>test-mae-mean</th>\n",
       "      <th>test-mae-std</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6.346214</td>\n",
       "      <td>0.034727</td>\n",
       "      <td>6.340220</td>\n",
       "      <td>0.146609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.459052</td>\n",
       "      <td>0.024914</td>\n",
       "      <td>4.464006</td>\n",
       "      <td>0.110257</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.154328</td>\n",
       "      <td>0.018296</td>\n",
       "      <td>3.170774</td>\n",
       "      <td>0.084777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2.288008</td>\n",
       "      <td>0.014536</td>\n",
       "      <td>2.333439</td>\n",
       "      <td>0.063921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.752826</td>\n",
       "      <td>0.012413</td>\n",
       "      <td>1.834495</td>\n",
       "      <td>0.052558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.434811</td>\n",
       "      <td>0.011454</td>\n",
       "      <td>1.553012</td>\n",
       "      <td>0.051337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.253621</td>\n",
       "      <td>0.008667</td>\n",
       "      <td>1.401426</td>\n",
       "      <td>0.047942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.151001</td>\n",
       "      <td>0.007782</td>\n",
       "      <td>1.325911</td>\n",
       "      <td>0.048060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.090303</td>\n",
       "      <td>0.006617</td>\n",
       "      <td>1.282442</td>\n",
       "      <td>0.046659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.051155</td>\n",
       "      <td>0.005701</td>\n",
       "      <td>1.262041</td>\n",
       "      <td>0.046579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.023911</td>\n",
       "      <td>0.007312</td>\n",
       "      <td>1.247609</td>\n",
       "      <td>0.045372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.003503</td>\n",
       "      <td>0.007257</td>\n",
       "      <td>1.241344</td>\n",
       "      <td>0.045991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.988047</td>\n",
       "      <td>0.007363</td>\n",
       "      <td>1.237711</td>\n",
       "      <td>0.044729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.973342</td>\n",
       "      <td>0.009031</td>\n",
       "      <td>1.234657</td>\n",
       "      <td>0.042415</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.964257</td>\n",
       "      <td>0.010250</td>\n",
       "      <td>1.233669</td>\n",
       "      <td>0.041619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.958159</td>\n",
       "      <td>0.010405</td>\n",
       "      <td>1.234180</td>\n",
       "      <td>0.041479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.948910</td>\n",
       "      <td>0.009228</td>\n",
       "      <td>1.233096</td>\n",
       "      <td>0.042646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.941599</td>\n",
       "      <td>0.008379</td>\n",
       "      <td>1.231458</td>\n",
       "      <td>0.042347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.933508</td>\n",
       "      <td>0.009956</td>\n",
       "      <td>1.230309</td>\n",
       "      <td>0.043167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.925629</td>\n",
       "      <td>0.010313</td>\n",
       "      <td>1.227850</td>\n",
       "      <td>0.042707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.916991</td>\n",
       "      <td>0.010470</td>\n",
       "      <td>1.226426</td>\n",
       "      <td>0.041590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.908275</td>\n",
       "      <td>0.010390</td>\n",
       "      <td>1.225448</td>\n",
       "      <td>0.040208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.899096</td>\n",
       "      <td>0.011133</td>\n",
       "      <td>1.224315</td>\n",
       "      <td>0.037572</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.890262</td>\n",
       "      <td>0.010121</td>\n",
       "      <td>1.224777</td>\n",
       "      <td>0.036481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.882774</td>\n",
       "      <td>0.009812</td>\n",
       "      <td>1.223708</td>\n",
       "      <td>0.035675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.872621</td>\n",
       "      <td>0.010319</td>\n",
       "      <td>1.223796</td>\n",
       "      <td>0.035022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.865180</td>\n",
       "      <td>0.010921</td>\n",
       "      <td>1.224790</td>\n",
       "      <td>0.034728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.857937</td>\n",
       "      <td>0.010204</td>\n",
       "      <td>1.224541</td>\n",
       "      <td>0.034001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.850340</td>\n",
       "      <td>0.009990</td>\n",
       "      <td>1.224230</td>\n",
       "      <td>0.033599</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0.845109</td>\n",
       "      <td>0.009162</td>\n",
       "      <td>1.223701</td>\n",
       "      <td>0.033822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.839880</td>\n",
       "      <td>0.010046</td>\n",
       "      <td>1.223003</td>\n",
       "      <td>0.034010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>0.833752</td>\n",
       "      <td>0.008985</td>\n",
       "      <td>1.222602</td>\n",
       "      <td>0.034492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.826344</td>\n",
       "      <td>0.009579</td>\n",
       "      <td>1.222833</td>\n",
       "      <td>0.034488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>0.821536</td>\n",
       "      <td>0.009268</td>\n",
       "      <td>1.222894</td>\n",
       "      <td>0.033605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>0.816293</td>\n",
       "      <td>0.008982</td>\n",
       "      <td>1.223043</td>\n",
       "      <td>0.034148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>0.810904</td>\n",
       "      <td>0.007572</td>\n",
       "      <td>1.223215</td>\n",
       "      <td>0.033927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>0.805303</td>\n",
       "      <td>0.007306</td>\n",
       "      <td>1.222243</td>\n",
       "      <td>0.034856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>0.798979</td>\n",
       "      <td>0.007965</td>\n",
       "      <td>1.222773</td>\n",
       "      <td>0.034737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>0.793694</td>\n",
       "      <td>0.006958</td>\n",
       "      <td>1.221859</td>\n",
       "      <td>0.035446</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>0.787756</td>\n",
       "      <td>0.006874</td>\n",
       "      <td>1.220722</td>\n",
       "      <td>0.035545</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>0.782284</td>\n",
       "      <td>0.007611</td>\n",
       "      <td>1.220396</td>\n",
       "      <td>0.035843</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>0.776633</td>\n",
       "      <td>0.007947</td>\n",
       "      <td>1.219764</td>\n",
       "      <td>0.036291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>0.773141</td>\n",
       "      <td>0.006820</td>\n",
       "      <td>1.219938</td>\n",
       "      <td>0.035637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>0.767519</td>\n",
       "      <td>0.006879</td>\n",
       "      <td>1.219951</td>\n",
       "      <td>0.035566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>0.761734</td>\n",
       "      <td>0.007287</td>\n",
       "      <td>1.219448</td>\n",
       "      <td>0.035354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>0.757455</td>\n",
       "      <td>0.007716</td>\n",
       "      <td>1.219641</td>\n",
       "      <td>0.035196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>0.752529</td>\n",
       "      <td>0.008340</td>\n",
       "      <td>1.218971</td>\n",
       "      <td>0.034997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>0.747729</td>\n",
       "      <td>0.008109</td>\n",
       "      <td>1.218733</td>\n",
       "      <td>0.035131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>0.743224</td>\n",
       "      <td>0.007688</td>\n",
       "      <td>1.218696</td>\n",
       "      <td>0.033787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>0.739033</td>\n",
       "      <td>0.006872</td>\n",
       "      <td>1.219522</td>\n",
       "      <td>0.033394</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>0.735268</td>\n",
       "      <td>0.007100</td>\n",
       "      <td>1.218782</td>\n",
       "      <td>0.032992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>0.730659</td>\n",
       "      <td>0.007393</td>\n",
       "      <td>1.218279</td>\n",
       "      <td>0.032610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>0.726264</td>\n",
       "      <td>0.007202</td>\n",
       "      <td>1.218404</td>\n",
       "      <td>0.032343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>0.721372</td>\n",
       "      <td>0.007475</td>\n",
       "      <td>1.218123</td>\n",
       "      <td>0.031868</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    train-mae-mean  train-mae-std  test-mae-mean  test-mae-std\n",
       "0         6.346214       0.034727       6.340220      0.146609\n",
       "1         4.459052       0.024914       4.464006      0.110257\n",
       "2         3.154328       0.018296       3.170774      0.084777\n",
       "3         2.288008       0.014536       2.333439      0.063921\n",
       "4         1.752826       0.012413       1.834495      0.052558\n",
       "5         1.434811       0.011454       1.553012      0.051337\n",
       "6         1.253621       0.008667       1.401426      0.047942\n",
       "7         1.151001       0.007782       1.325911      0.048060\n",
       "8         1.090303       0.006617       1.282442      0.046659\n",
       "9         1.051155       0.005701       1.262041      0.046579\n",
       "10        1.023911       0.007312       1.247609      0.045372\n",
       "11        1.003503       0.007257       1.241344      0.045991\n",
       "12        0.988047       0.007363       1.237711      0.044729\n",
       "13        0.973342       0.009031       1.234657      0.042415\n",
       "14        0.964257       0.010250       1.233669      0.041619\n",
       "15        0.958159       0.010405       1.234180      0.041479\n",
       "16        0.948910       0.009228       1.233096      0.042646\n",
       "17        0.941599       0.008379       1.231458      0.042347\n",
       "18        0.933508       0.009956       1.230309      0.043167\n",
       "19        0.925629       0.010313       1.227850      0.042707\n",
       "20        0.916991       0.010470       1.226426      0.041590\n",
       "21        0.908275       0.010390       1.225448      0.040208\n",
       "22        0.899096       0.011133       1.224315      0.037572\n",
       "23        0.890262       0.010121       1.224777      0.036481\n",
       "24        0.882774       0.009812       1.223708      0.035675\n",
       "25        0.872621       0.010319       1.223796      0.035022\n",
       "26        0.865180       0.010921       1.224790      0.034728\n",
       "27        0.857937       0.010204       1.224541      0.034001\n",
       "28        0.850340       0.009990       1.224230      0.033599\n",
       "29        0.845109       0.009162       1.223701      0.033822\n",
       "30        0.839880       0.010046       1.223003      0.034010\n",
       "31        0.833752       0.008985       1.222602      0.034492\n",
       "32        0.826344       0.009579       1.222833      0.034488\n",
       "33        0.821536       0.009268       1.222894      0.033605\n",
       "34        0.816293       0.008982       1.223043      0.034148\n",
       "35        0.810904       0.007572       1.223215      0.033927\n",
       "36        0.805303       0.007306       1.222243      0.034856\n",
       "37        0.798979       0.007965       1.222773      0.034737\n",
       "38        0.793694       0.006958       1.221859      0.035446\n",
       "39        0.787756       0.006874       1.220722      0.035545\n",
       "40        0.782284       0.007611       1.220396      0.035843\n",
       "41        0.776633       0.007947       1.219764      0.036291\n",
       "42        0.773141       0.006820       1.219938      0.035637\n",
       "43        0.767519       0.006879       1.219951      0.035566\n",
       "44        0.761734       0.007287       1.219448      0.035354\n",
       "45        0.757455       0.007716       1.219641      0.035196\n",
       "46        0.752529       0.008340       1.218971      0.034997\n",
       "47        0.747729       0.008109       1.218733      0.035131\n",
       "48        0.743224       0.007688       1.218696      0.033787\n",
       "49        0.739033       0.006872       1.219522      0.033394\n",
       "50        0.735268       0.007100       1.218782      0.032992\n",
       "51        0.730659       0.007393       1.218279      0.032610\n",
       "52        0.726264       0.007202       1.218404      0.032343\n",
       "53        0.721372       0.007475       1.218123      0.031868"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results = xgb.cv(\n",
    "    params,\n",
    "    dtrain,\n",
    "    num_boost_round=num_boost_round,\n",
    "    seed=42,\n",
    "    nfold=5,\n",
    "    metrics={'mae'},\n",
    "    early_stopping_rounds=10\n",
    ")\n",
    "cv_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.2181234"
      ]
     },
     "execution_count": 235,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results['test-mae-mean'].min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {},
   "outputs": [],
   "source": [
    "gridsearch_params = [\n",
    "    (max_depth, min_child_weight)\n",
    "    for max_depth in range(9,12)\n",
    "    for min_child_weight in range(5,8)\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CV with max_depth=9, min_child_weight=5\n",
      "[23:07:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:23: FutureWarning: \n",
      "The current behaviour of 'Series.argmin' is deprecated, use 'idxmin'\n",
      "instead.\n",
      "The behavior of 'argmin' will be corrected to return the positional\n",
      "minimum in the future. For now, use 'series.values.argmin' or\n",
      "'np.argmin(np.array(values))' to get the position of the minimum\n",
      "row.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tMAE 1.2006320000000001 for 22 rounds\n",
      "CV with max_depth=9, min_child_weight=6\n",
      "[23:07:10] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:10] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:10] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:10] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:10] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.217154 for 13 rounds\n",
      "CV with max_depth=9, min_child_weight=7\n",
      "[23:07:12] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:12] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:12] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:12] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:12] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2194276 for 13 rounds\n",
      "CV with max_depth=10, min_child_weight=5\n",
      "[23:07:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2061366 for 14 rounds\n",
      "CV with max_depth=10, min_child_weight=6\n",
      "[23:07:17] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:17] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:17] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:17] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:17] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2166902 for 15 rounds\n",
      "CV with max_depth=10, min_child_weight=7\n",
      "[23:07:19] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:19] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:19] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:19] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:19] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2098136 for 13 rounds\n",
      "CV with max_depth=11, min_child_weight=5\n",
      "[23:07:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2132852 for 12 rounds\n",
      "CV with max_depth=11, min_child_weight=6\n",
      "[23:07:24] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:24] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:24] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:24] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:24] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2189066 for 14 rounds\n",
      "CV with max_depth=11, min_child_weight=7\n",
      "[23:07:27] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:27] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:27] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:27] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:07:27] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2194916 for 26 rounds\n",
      "Best params: 9, 5, MAE: 1.2006320000000001\n"
     ]
    }
   ],
   "source": [
    "# Define initial best params and MAE\n",
    "min_mae = float(\"Inf\")\n",
    "best_params = None\n",
    "for max_depth, min_child_weight in gridsearch_params:\n",
    "    print(\"CV with max_depth={}, min_child_weight={}\".format(\n",
    "                             max_depth,\n",
    "                             min_child_weight))\n",
    "    # Update our parameters\n",
    "    params['max_depth'] = max_depth\n",
    "    params['min_child_weight'] = min_child_weight\n",
    "    # Run CV\n",
    "    cv_results = xgb.cv(\n",
    "        params,\n",
    "        dtrain,\n",
    "        num_boost_round=num_boost_round,\n",
    "        seed=42,\n",
    "        nfold=5,\n",
    "        metrics={'mae'},\n",
    "        early_stopping_rounds=10\n",
    "    )\n",
    "    # Update best MAE\n",
    "    mean_mae = cv_results['test-mae-mean'].min()\n",
    "    boost_rounds = cv_results['test-mae-mean'].argmin()\n",
    "    print(\"\\tMAE {} for {} rounds\".format(mean_mae, boost_rounds))\n",
    "    if mean_mae < min_mae:\n",
    "        min_mae = mean_mae\n",
    "        best_params = (max_depth,min_child_weight)\n",
    "print(\"Best params: {}, {}, MAE: {}\".format(best_params[0], best_params[1], min_mae))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [],
   "source": [
    "params['max_depth'] = 9\n",
    "params['min_child_weight'] = 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CV with subsample=1.0, colsample=1.0\n",
      "[23:08:04] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:04] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:04] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:04] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:04] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:35: FutureWarning: \n",
      "The current behaviour of 'Series.argmin' is deprecated, use 'idxmin'\n",
      "instead.\n",
      "The behavior of 'argmin' will be corrected to return the positional\n",
      "minimum in the future. For now, use 'series.values.argmin' or\n",
      "'np.argmin(np.array(values))' to get the position of the minimum\n",
      "row.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tMAE 1.2006320000000001 for 22 rounds\n",
      "CV with subsample=1.0, colsample=0.9\n",
      "[23:08:06] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:06] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:06] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:06] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.21156 for 14 rounds\n",
      "CV with subsample=1.0, colsample=0.8\n",
      "[23:08:09] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:09] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:09] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:09] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:09] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2113176 for 16 rounds\n",
      "CV with subsample=1.0, colsample=0.7\n",
      "[23:08:11] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:11] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:11] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:11] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:11] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2155626000000002 for 14 rounds\n",
      "CV with subsample=0.9, colsample=1.0\n",
      "[23:08:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:14] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2116978 for 13 rounds\n",
      "CV with subsample=0.9, colsample=0.9\n",
      "[23:08:16] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:16] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:16] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:16] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:16] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2101418 for 15 rounds\n",
      "CV with subsample=0.9, colsample=0.8\n",
      "[23:08:20] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:20] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:20] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:20] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:20] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2159499999999999 for 15 rounds\n",
      "CV with subsample=0.9, colsample=0.7\n",
      "[23:08:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2076688 for 17 rounds\n",
      "CV with subsample=0.8, colsample=1.0\n",
      "[23:08:23] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:23] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:24] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:24] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:24] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2165966 for 14 rounds\n",
      "CV with subsample=0.8, colsample=0.9\n",
      "[23:08:25] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:25] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:25] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:25] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:25] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2206272 for 15 rounds\n",
      "CV with subsample=0.8, colsample=0.8\n",
      "[23:08:28] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:28] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:28] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:28] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:28] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2212124 for 13 rounds\n",
      "CV with subsample=0.8, colsample=0.7\n",
      "[23:08:29] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:29] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:29] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:29] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:29] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tMAE 1.2210687999999998 for 18 rounds\n",
      "CV with subsample=0.7, colsample=1.0\n",
      "[23:08:31] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:31] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:31] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:31] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:31] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2294278 for 13 rounds\n",
      "CV with subsample=0.7, colsample=0.9\n",
      "[23:08:33] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:33] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:33] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:33] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:33] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.218609 for 13 rounds\n",
      "CV with subsample=0.7, colsample=0.8\n",
      "[23:08:35] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:35] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:35] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:35] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:35] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2251218 for 13 rounds\n",
      "CV with subsample=0.7, colsample=0.7\n",
      "[23:08:37] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:37] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:37] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:37] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:08:37] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "\tMAE 1.2322886 for 12 rounds\n",
      "Best params: 1.0, 1.0, MAE: 1.2006320000000001\n"
     ]
    }
   ],
   "source": [
    "# Tune subsample: corresponds to the fraction of observations (the rows) to subsample at each step. \n",
    "# By default it is set to 1 meaning that we use all rows.\n",
    "\n",
    "# Tune colsample_bytree: corresponds to the fraction of features (the columns) to use. \n",
    "# By default it is set to 1 meaning that we will use all features. \n",
    "\n",
    "gridsearch_params = [\n",
    "    (subsample, colsample)\n",
    "    for subsample in [i/10. for i in range(7,11)]\n",
    "    for colsample in [i/10. for i in range(7,11)]\n",
    "]\n",
    "\n",
    "min_mae = float(\"Inf\")\n",
    "best_params = None\n",
    "# We start by the largest values and go down to the smallest\n",
    "for subsample, colsample in reversed(gridsearch_params):\n",
    "    print(\"CV with subsample={}, colsample={}\".format(\n",
    "                             subsample,\n",
    "                             colsample))\n",
    "    # We update our parameters\n",
    "    params['subsample'] = subsample\n",
    "    params['colsample_bytree'] = colsample\n",
    "    # Run CV\n",
    "    cv_results = xgb.cv(\n",
    "        params,\n",
    "        dtrain,\n",
    "        num_boost_round=num_boost_round,\n",
    "        seed=42,\n",
    "        nfold=5,\n",
    "        metrics={'mae'},\n",
    "        early_stopping_rounds=10\n",
    "    )\n",
    "    # Update best score\n",
    "    mean_mae = cv_results['test-mae-mean'].min()\n",
    "    boost_rounds = cv_results['test-mae-mean'].argmin()\n",
    "    print(\"\\tMAE {} for {} rounds\".format(mean_mae, boost_rounds))\n",
    "    if mean_mae < min_mae:\n",
    "        min_mae = mean_mae\n",
    "        best_params = (subsample,colsample)\n",
    "print(\"Best params: {}, {}, MAE: {}\".format(best_params[0], best_params[1], min_mae))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {},
   "outputs": [],
   "source": [
    "params['subsample'] = 1.0\n",
    "params['colsample_bytree'] = 1.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 4 µs, sys: 2 µs, total: 6 µs\n",
      "Wall time: 485 µs\n",
      "CV with eta=0.3\n",
      "[23:09:05] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:05] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:05] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:05] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:05] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "CPU times: user 17.1 s, sys: 263 ms, total: 17.4 s\n",
      "Wall time: 2.38 s\n",
      "\tMAE 1.2006320000000001 for 22 rounds\n",
      "\n",
      "CV with eta=0.2\n",
      "[23:09:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:07] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dereklee/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:15: FutureWarning: \n",
      "The current behaviour of 'Series.argmin' is deprecated, use 'idxmin'\n",
      "instead.\n",
      "The behavior of 'argmin' will be corrected to return the positional\n",
      "minimum in the future. For now, use 'series.values.argmin' or\n",
      "'np.argmin(np.array(values))' to get the position of the minimum\n",
      "row.\n",
      "  from ipykernel import kernelapp as app\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 35.5 s, sys: 1.22 s, total: 36.7 s\n",
      "Wall time: 5.59 s\n",
      "\tMAE 1.1736537999999999 for 48 rounds\n",
      "\n",
      "CV with eta=0.1\n",
      "[23:09:13] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:13] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:13] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:13] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:13] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "CPU times: user 1min 10s, sys: 1.15 s, total: 1min 11s\n",
      "Wall time: 9.69 s\n",
      "\tMAE 1.1447942000000002 for 92 rounds\n",
      "\n",
      "CV with eta=0.05\n",
      "[23:09:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:22] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "CPU times: user 1min 26s, sys: 1.03 s, total: 1min 27s\n",
      "Wall time: 11.5 s\n",
      "\tMAE 1.1398798 for 110 rounds\n",
      "\n",
      "CV with eta=0.01\n",
      "[23:09:34] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:34] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:34] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:34] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:09:34] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "CPU times: user 8min 41s, sys: 12 s, total: 8min 53s\n",
      "Wall time: 1min 21s\n",
      "\tMAE 1.1326964 for 640 rounds\n",
      "\n",
      "CV with eta=0.005\n",
      "[23:10:55] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:10:55] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:10:55] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:10:55] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[23:10:55] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "CPU times: user 15min 45s, sys: 24.1 s, total: 16min 9s\n",
      "Wall time: 2min 28s\n",
      "\tMAE 1.1372584 for 998 rounds\n",
      "\n",
      "Best params: 0.01, MAE: 1.1326964\n"
     ]
    }
   ],
   "source": [
    "# Tune eta - learning rate\n",
    "\n",
    "%time\n",
    "# This can take some time…\n",
    "min_mae = float(\"Inf\")\n",
    "best_params = None\n",
    "for eta in [.3, .2, .1, .05, .01, .005]:\n",
    "    print(\"CV with eta={}\".format(eta))\n",
    "    # We update our parameters\n",
    "    params['eta'] = eta\n",
    "    # Run and time CV\n",
    "    %time cv_results = xgb.cv(params, dtrain, num_boost_round=num_boost_round, seed=42, nfold=5, metrics=['mae'], early_stopping_rounds=10)\n",
    "    # Update best score\n",
    "    mean_mae = cv_results['test-mae-mean'].min()\n",
    "    boost_rounds = cv_results['test-mae-mean'].argmin()\n",
    "    print(\"\\tMAE {} for {} rounds\\n\".format(mean_mae, boost_rounds))\n",
    "    if mean_mae < min_mae:\n",
    "        min_mae = mean_mae\n",
    "        best_params = eta\n",
    "print(\"Best params: {}, MAE: {}\".format(best_params, min_mae))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [],
   "source": [
    "params['eta'] = 0.01"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_depth': 9,\n",
       " 'min_child_weight': 5,\n",
       " 'eta': 0.01,\n",
       " 'subsample': 1.0,\n",
       " 'colsample_bytree': 1.0,\n",
       " 'objective': 'reg:linear',\n",
       " 'eval_metric': 'mae'}"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[23:13:50] WARNING: /Users/travis/build/dmlc/xgboost/src/objective/regression_obj.cu:170: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[0]\tTest-mae:9.12460\n",
      "[1]\tTest-mae:9.03382\n",
      "[2]\tTest-mae:8.94400\n",
      "[3]\tTest-mae:8.85503\n",
      "[4]\tTest-mae:8.76694\n",
      "[5]\tTest-mae:8.67966\n",
      "[6]\tTest-mae:8.59353\n",
      "[7]\tTest-mae:8.50835\n",
      "[8]\tTest-mae:8.42404\n",
      "[9]\tTest-mae:8.34029\n",
      "[10]\tTest-mae:8.25781\n",
      "[11]\tTest-mae:8.17594\n",
      "[12]\tTest-mae:8.09526\n",
      "[13]\tTest-mae:8.01507\n",
      "[14]\tTest-mae:7.93581\n",
      "[15]\tTest-mae:7.85656\n",
      "[16]\tTest-mae:7.77915\n",
      "[17]\tTest-mae:7.70249\n",
      "[18]\tTest-mae:7.62595\n",
      "[19]\tTest-mae:7.55048\n",
      "[20]\tTest-mae:7.47544\n",
      "[21]\tTest-mae:7.40108\n",
      "[22]\tTest-mae:7.32813\n",
      "[23]\tTest-mae:7.25514\n",
      "[24]\tTest-mae:7.18292\n",
      "[25]\tTest-mae:7.11171\n",
      "[26]\tTest-mae:7.04104\n",
      "[27]\tTest-mae:6.97107\n",
      "[28]\tTest-mae:6.90202\n",
      "[29]\tTest-mae:6.83336\n",
      "[30]\tTest-mae:6.76563\n",
      "[31]\tTest-mae:6.69874\n",
      "[32]\tTest-mae:6.63186\n",
      "[33]\tTest-mae:6.56590\n",
      "[34]\tTest-mae:6.50091\n",
      "[35]\tTest-mae:6.43642\n",
      "[36]\tTest-mae:6.37296\n",
      "[37]\tTest-mae:6.30965\n",
      "[38]\tTest-mae:6.24759\n",
      "[39]\tTest-mae:6.18584\n",
      "[40]\tTest-mae:6.12439\n",
      "[41]\tTest-mae:6.06409\n",
      "[42]\tTest-mae:6.00398\n",
      "[43]\tTest-mae:5.94509\n",
      "[44]\tTest-mae:5.88628\n",
      "[45]\tTest-mae:5.82838\n",
      "[46]\tTest-mae:5.77133\n",
      "[47]\tTest-mae:5.71457\n",
      "[48]\tTest-mae:5.65863\n",
      "[49]\tTest-mae:5.60311\n",
      "[50]\tTest-mae:5.54751\n",
      "[51]\tTest-mae:5.49264\n",
      "[52]\tTest-mae:5.43801\n",
      "[53]\tTest-mae:5.38478\n",
      "[54]\tTest-mae:5.33142\n",
      "[55]\tTest-mae:5.27912\n",
      "[56]\tTest-mae:5.22702\n",
      "[57]\tTest-mae:5.17581\n",
      "[58]\tTest-mae:5.12454\n",
      "[59]\tTest-mae:5.07429\n",
      "[60]\tTest-mae:5.02414\n",
      "[61]\tTest-mae:4.97469\n",
      "[62]\tTest-mae:4.92570\n",
      "[63]\tTest-mae:4.87778\n",
      "[64]\tTest-mae:4.82953\n",
      "[65]\tTest-mae:4.78266\n",
      "[66]\tTest-mae:4.73562\n",
      "[67]\tTest-mae:4.68950\n",
      "[68]\tTest-mae:4.64357\n",
      "[69]\tTest-mae:4.59811\n",
      "[70]\tTest-mae:4.55308\n",
      "[71]\tTest-mae:4.50885\n",
      "[72]\tTest-mae:4.46453\n",
      "[73]\tTest-mae:4.42082\n",
      "[74]\tTest-mae:4.37754\n",
      "[75]\tTest-mae:4.33521\n",
      "[76]\tTest-mae:4.29290\n",
      "[77]\tTest-mae:4.25117\n",
      "[78]\tTest-mae:4.20992\n",
      "[79]\tTest-mae:4.16907\n",
      "[80]\tTest-mae:4.12860\n",
      "[81]\tTest-mae:4.08867\n",
      "[82]\tTest-mae:4.04927\n",
      "[83]\tTest-mae:4.01054\n",
      "[84]\tTest-mae:3.97167\n",
      "[85]\tTest-mae:3.93346\n",
      "[86]\tTest-mae:3.89603\n",
      "[87]\tTest-mae:3.85852\n",
      "[88]\tTest-mae:3.82100\n",
      "[89]\tTest-mae:3.78432\n",
      "[90]\tTest-mae:3.74807\n",
      "[91]\tTest-mae:3.71203\n",
      "[92]\tTest-mae:3.67667\n",
      "[93]\tTest-mae:3.64179\n",
      "[94]\tTest-mae:3.60701\n",
      "[95]\tTest-mae:3.57245\n",
      "[96]\tTest-mae:3.53866\n",
      "[97]\tTest-mae:3.50513\n",
      "[98]\tTest-mae:3.47176\n",
      "[99]\tTest-mae:3.43882\n",
      "[100]\tTest-mae:3.40589\n",
      "[101]\tTest-mae:3.37380\n",
      "[102]\tTest-mae:3.34198\n",
      "[103]\tTest-mae:3.31061\n",
      "[104]\tTest-mae:3.27957\n",
      "[105]\tTest-mae:3.24880\n",
      "[106]\tTest-mae:3.21854\n",
      "[107]\tTest-mae:3.18861\n",
      "[108]\tTest-mae:3.15871\n",
      "[109]\tTest-mae:3.12923\n",
      "[110]\tTest-mae:3.10020\n",
      "[111]\tTest-mae:3.07165\n",
      "[112]\tTest-mae:3.04340\n",
      "[113]\tTest-mae:3.01548\n",
      "[114]\tTest-mae:2.98779\n",
      "[115]\tTest-mae:2.96064\n",
      "[116]\tTest-mae:2.93368\n",
      "[117]\tTest-mae:2.90701\n",
      "[118]\tTest-mae:2.88058\n",
      "[119]\tTest-mae:2.85430\n",
      "[120]\tTest-mae:2.82837\n",
      "[121]\tTest-mae:2.80298\n",
      "[122]\tTest-mae:2.77777\n",
      "[123]\tTest-mae:2.75309\n",
      "[124]\tTest-mae:2.72833\n",
      "[125]\tTest-mae:2.70434\n",
      "[126]\tTest-mae:2.68060\n",
      "[127]\tTest-mae:2.65735\n",
      "[128]\tTest-mae:2.63427\n",
      "[129]\tTest-mae:2.61169\n",
      "[130]\tTest-mae:2.58917\n",
      "[131]\tTest-mae:2.56686\n",
      "[132]\tTest-mae:2.54496\n",
      "[133]\tTest-mae:2.52344\n",
      "[134]\tTest-mae:2.50206\n",
      "[135]\tTest-mae:2.48095\n",
      "[136]\tTest-mae:2.46016\n",
      "[137]\tTest-mae:2.43965\n",
      "[138]\tTest-mae:2.41948\n",
      "[139]\tTest-mae:2.39961\n",
      "[140]\tTest-mae:2.37967\n",
      "[141]\tTest-mae:2.36014\n",
      "[142]\tTest-mae:2.34089\n",
      "[143]\tTest-mae:2.32201\n",
      "[144]\tTest-mae:2.30298\n",
      "[145]\tTest-mae:2.28441\n",
      "[146]\tTest-mae:2.26611\n",
      "[147]\tTest-mae:2.24821\n",
      "[148]\tTest-mae:2.23030\n",
      "[149]\tTest-mae:2.21279\n",
      "[150]\tTest-mae:2.19556\n",
      "[151]\tTest-mae:2.17864\n",
      "[152]\tTest-mae:2.16172\n",
      "[153]\tTest-mae:2.14510\n",
      "[154]\tTest-mae:2.12866\n",
      "[155]\tTest-mae:2.11228\n",
      "[156]\tTest-mae:2.09640\n",
      "[157]\tTest-mae:2.08072\n",
      "[158]\tTest-mae:2.06538\n",
      "[159]\tTest-mae:2.05006\n",
      "[160]\tTest-mae:2.03494\n",
      "[161]\tTest-mae:2.02002\n",
      "[162]\tTest-mae:2.00550\n",
      "[163]\tTest-mae:1.99123\n",
      "[164]\tTest-mae:1.97697\n",
      "[165]\tTest-mae:1.96310\n",
      "[166]\tTest-mae:1.94941\n",
      "[167]\tTest-mae:1.93604\n",
      "[168]\tTest-mae:1.92269\n",
      "[169]\tTest-mae:1.90963\n",
      "[170]\tTest-mae:1.89686\n",
      "[171]\tTest-mae:1.88419\n",
      "[172]\tTest-mae:1.87190\n",
      "[173]\tTest-mae:1.85952\n",
      "[174]\tTest-mae:1.84737\n",
      "[175]\tTest-mae:1.83545\n",
      "[176]\tTest-mae:1.82399\n",
      "[177]\tTest-mae:1.81238\n",
      "[178]\tTest-mae:1.80086\n",
      "[179]\tTest-mae:1.78935\n",
      "[180]\tTest-mae:1.77805\n",
      "[181]\tTest-mae:1.76724\n",
      "[182]\tTest-mae:1.75640\n",
      "[183]\tTest-mae:1.74586\n",
      "[184]\tTest-mae:1.73539\n",
      "[185]\tTest-mae:1.72538\n",
      "[186]\tTest-mae:1.71507\n",
      "[187]\tTest-mae:1.70530\n",
      "[188]\tTest-mae:1.69571\n",
      "[189]\tTest-mae:1.68580\n",
      "[190]\tTest-mae:1.67627\n",
      "[191]\tTest-mae:1.66674\n",
      "[192]\tTest-mae:1.65751\n",
      "[193]\tTest-mae:1.64857\n",
      "[194]\tTest-mae:1.63974\n",
      "[195]\tTest-mae:1.63099\n",
      "[196]\tTest-mae:1.62229\n",
      "[197]\tTest-mae:1.61391\n",
      "[198]\tTest-mae:1.60531\n",
      "[199]\tTest-mae:1.59670\n",
      "[200]\tTest-mae:1.58831\n",
      "[201]\tTest-mae:1.58023\n",
      "[202]\tTest-mae:1.57218\n",
      "[203]\tTest-mae:1.56447\n",
      "[204]\tTest-mae:1.55679\n",
      "[205]\tTest-mae:1.54906\n",
      "[206]\tTest-mae:1.54161\n",
      "[207]\tTest-mae:1.53436\n",
      "[208]\tTest-mae:1.52713\n",
      "[209]\tTest-mae:1.52008\n",
      "[210]\tTest-mae:1.51312\n",
      "[211]\tTest-mae:1.50614\n",
      "[212]\tTest-mae:1.49936\n",
      "[213]\tTest-mae:1.49276\n",
      "[214]\tTest-mae:1.48633\n",
      "[215]\tTest-mae:1.47989\n",
      "[216]\tTest-mae:1.47366\n",
      "[217]\tTest-mae:1.46743\n",
      "[218]\tTest-mae:1.46150\n",
      "[219]\tTest-mae:1.45551\n",
      "[220]\tTest-mae:1.44969\n",
      "[221]\tTest-mae:1.44419\n",
      "[222]\tTest-mae:1.43868\n",
      "[223]\tTest-mae:1.43326\n",
      "[224]\tTest-mae:1.42769\n",
      "[225]\tTest-mae:1.42243\n",
      "[226]\tTest-mae:1.41726\n",
      "[227]\tTest-mae:1.41212\n",
      "[228]\tTest-mae:1.40699\n",
      "[229]\tTest-mae:1.40216\n",
      "[230]\tTest-mae:1.39727\n",
      "[231]\tTest-mae:1.39260\n",
      "[232]\tTest-mae:1.38791\n",
      "[233]\tTest-mae:1.38329\n",
      "[234]\tTest-mae:1.37887\n",
      "[235]\tTest-mae:1.37446\n",
      "[236]\tTest-mae:1.37016\n",
      "[237]\tTest-mae:1.36598\n",
      "[238]\tTest-mae:1.36186\n",
      "[239]\tTest-mae:1.35765\n",
      "[240]\tTest-mae:1.35376\n",
      "[241]\tTest-mae:1.34977\n",
      "[242]\tTest-mae:1.34581\n",
      "[243]\tTest-mae:1.34203\n",
      "[244]\tTest-mae:1.33835\n",
      "[245]\tTest-mae:1.33452\n",
      "[246]\tTest-mae:1.33104\n",
      "[247]\tTest-mae:1.32736\n",
      "[248]\tTest-mae:1.32384\n",
      "[249]\tTest-mae:1.32042\n",
      "[250]\tTest-mae:1.31700\n",
      "[251]\tTest-mae:1.31368\n",
      "[252]\tTest-mae:1.31038\n",
      "[253]\tTest-mae:1.30727\n",
      "[254]\tTest-mae:1.30416\n",
      "[255]\tTest-mae:1.30112\n",
      "[256]\tTest-mae:1.29801\n",
      "[257]\tTest-mae:1.29492\n",
      "[258]\tTest-mae:1.29198\n",
      "[259]\tTest-mae:1.28901\n",
      "[260]\tTest-mae:1.28607\n",
      "[261]\tTest-mae:1.28320\n",
      "[262]\tTest-mae:1.28038\n",
      "[263]\tTest-mae:1.27764\n",
      "[264]\tTest-mae:1.27478\n",
      "[265]\tTest-mae:1.27212\n",
      "[266]\tTest-mae:1.26946\n",
      "[267]\tTest-mae:1.26689\n",
      "[268]\tTest-mae:1.26434\n",
      "[269]\tTest-mae:1.26181\n",
      "[270]\tTest-mae:1.25943\n",
      "[271]\tTest-mae:1.25703\n",
      "[272]\tTest-mae:1.25466\n",
      "[273]\tTest-mae:1.25235\n",
      "[274]\tTest-mae:1.25006\n",
      "[275]\tTest-mae:1.24779\n",
      "[276]\tTest-mae:1.24569\n",
      "[277]\tTest-mae:1.24354\n",
      "[278]\tTest-mae:1.24141\n",
      "[279]\tTest-mae:1.23937\n",
      "[280]\tTest-mae:1.23735\n",
      "[281]\tTest-mae:1.23535\n",
      "[282]\tTest-mae:1.23350\n",
      "[283]\tTest-mae:1.23173\n",
      "[284]\tTest-mae:1.22983\n",
      "[285]\tTest-mae:1.22812\n",
      "[286]\tTest-mae:1.22646\n",
      "[287]\tTest-mae:1.22463\n",
      "[288]\tTest-mae:1.22279\n",
      "[289]\tTest-mae:1.22104\n",
      "[290]\tTest-mae:1.21924\n",
      "[291]\tTest-mae:1.21757\n",
      "[292]\tTest-mae:1.21601\n",
      "[293]\tTest-mae:1.21444\n",
      "[294]\tTest-mae:1.21286\n",
      "[295]\tTest-mae:1.21134\n",
      "[296]\tTest-mae:1.20964\n",
      "[297]\tTest-mae:1.20823\n",
      "[298]\tTest-mae:1.20689\n",
      "[299]\tTest-mae:1.20554\n",
      "[300]\tTest-mae:1.20418\n",
      "[301]\tTest-mae:1.20289\n",
      "[302]\tTest-mae:1.20165\n",
      "[303]\tTest-mae:1.20044\n",
      "[304]\tTest-mae:1.19911\n",
      "[305]\tTest-mae:1.19776\n",
      "[306]\tTest-mae:1.19634\n",
      "[307]\tTest-mae:1.19507\n",
      "[308]\tTest-mae:1.19389\n",
      "[309]\tTest-mae:1.19270\n",
      "[310]\tTest-mae:1.19165\n",
      "[311]\tTest-mae:1.19054\n",
      "[312]\tTest-mae:1.18940\n",
      "[313]\tTest-mae:1.18838\n",
      "[314]\tTest-mae:1.18726\n",
      "[315]\tTest-mae:1.18627\n",
      "[316]\tTest-mae:1.18527\n",
      "[317]\tTest-mae:1.18422\n",
      "[318]\tTest-mae:1.18325\n",
      "[319]\tTest-mae:1.18241\n",
      "[320]\tTest-mae:1.18151\n",
      "[321]\tTest-mae:1.18072\n",
      "[322]\tTest-mae:1.17990\n",
      "[323]\tTest-mae:1.17905\n",
      "[324]\tTest-mae:1.17815\n",
      "[325]\tTest-mae:1.17738\n",
      "[326]\tTest-mae:1.17664\n",
      "[327]\tTest-mae:1.17595\n",
      "[328]\tTest-mae:1.17535\n",
      "[329]\tTest-mae:1.17465\n",
      "[330]\tTest-mae:1.17392\n",
      "[331]\tTest-mae:1.17328\n",
      "[332]\tTest-mae:1.17252\n",
      "[333]\tTest-mae:1.17199\n",
      "[334]\tTest-mae:1.17138\n",
      "[335]\tTest-mae:1.17069\n",
      "[336]\tTest-mae:1.17013\n",
      "[337]\tTest-mae:1.16946\n",
      "[338]\tTest-mae:1.16883\n",
      "[339]\tTest-mae:1.16826\n",
      "[340]\tTest-mae:1.16767\n",
      "[341]\tTest-mae:1.16703\n",
      "[342]\tTest-mae:1.16651\n",
      "[343]\tTest-mae:1.16609\n",
      "[344]\tTest-mae:1.16549\n",
      "[345]\tTest-mae:1.16509\n",
      "[346]\tTest-mae:1.16464\n",
      "[347]\tTest-mae:1.16401\n",
      "[348]\tTest-mae:1.16347\n",
      "[349]\tTest-mae:1.16300\n",
      "[350]\tTest-mae:1.16249\n",
      "[351]\tTest-mae:1.16209\n",
      "[352]\tTest-mae:1.16166\n",
      "[353]\tTest-mae:1.16123\n",
      "[354]\tTest-mae:1.16085\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[355]\tTest-mae:1.16051\n",
      "[356]\tTest-mae:1.16016\n",
      "[357]\tTest-mae:1.15981\n",
      "[358]\tTest-mae:1.15940\n",
      "[359]\tTest-mae:1.15899\n",
      "[360]\tTest-mae:1.15868\n",
      "[361]\tTest-mae:1.15823\n",
      "[362]\tTest-mae:1.15784\n",
      "[363]\tTest-mae:1.15753\n",
      "[364]\tTest-mae:1.15721\n",
      "[365]\tTest-mae:1.15695\n",
      "[366]\tTest-mae:1.15643\n",
      "[367]\tTest-mae:1.15603\n",
      "[368]\tTest-mae:1.15563\n",
      "[369]\tTest-mae:1.15528\n",
      "[370]\tTest-mae:1.15492\n",
      "[371]\tTest-mae:1.15459\n",
      "[372]\tTest-mae:1.15427\n",
      "[373]\tTest-mae:1.15388\n",
      "[374]\tTest-mae:1.15352\n",
      "[375]\tTest-mae:1.15319\n",
      "[376]\tTest-mae:1.15288\n",
      "[377]\tTest-mae:1.15257\n",
      "[378]\tTest-mae:1.15231\n",
      "[379]\tTest-mae:1.15202\n",
      "[380]\tTest-mae:1.15170\n",
      "[381]\tTest-mae:1.15143\n",
      "[382]\tTest-mae:1.15114\n",
      "[383]\tTest-mae:1.15081\n",
      "[384]\tTest-mae:1.15052\n",
      "[385]\tTest-mae:1.15019\n",
      "[386]\tTest-mae:1.14986\n",
      "[387]\tTest-mae:1.14974\n",
      "[388]\tTest-mae:1.14961\n",
      "[389]\tTest-mae:1.14933\n",
      "[390]\tTest-mae:1.14907\n",
      "[391]\tTest-mae:1.14883\n",
      "[392]\tTest-mae:1.14865\n",
      "[393]\tTest-mae:1.14837\n",
      "[394]\tTest-mae:1.14813\n",
      "[395]\tTest-mae:1.14787\n",
      "[396]\tTest-mae:1.14776\n",
      "[397]\tTest-mae:1.14759\n",
      "[398]\tTest-mae:1.14732\n",
      "[399]\tTest-mae:1.14715\n",
      "[400]\tTest-mae:1.14681\n",
      "[401]\tTest-mae:1.14656\n",
      "[402]\tTest-mae:1.14639\n",
      "[403]\tTest-mae:1.14606\n",
      "[404]\tTest-mae:1.14590\n",
      "[405]\tTest-mae:1.14567\n",
      "[406]\tTest-mae:1.14543\n",
      "[407]\tTest-mae:1.14528\n",
      "[408]\tTest-mae:1.14512\n",
      "[409]\tTest-mae:1.14489\n",
      "[410]\tTest-mae:1.14478\n",
      "[411]\tTest-mae:1.14456\n",
      "[412]\tTest-mae:1.14435\n",
      "[413]\tTest-mae:1.14417\n",
      "[414]\tTest-mae:1.14398\n",
      "[415]\tTest-mae:1.14389\n",
      "[416]\tTest-mae:1.14377\n",
      "[417]\tTest-mae:1.14357\n",
      "[418]\tTest-mae:1.14339\n",
      "[419]\tTest-mae:1.14314\n",
      "[420]\tTest-mae:1.14302\n",
      "[421]\tTest-mae:1.14280\n",
      "[422]\tTest-mae:1.14260\n",
      "[423]\tTest-mae:1.14246\n",
      "[424]\tTest-mae:1.14233\n",
      "[425]\tTest-mae:1.14209\n",
      "[426]\tTest-mae:1.14200\n",
      "[427]\tTest-mae:1.14177\n",
      "[428]\tTest-mae:1.14169\n",
      "[429]\tTest-mae:1.14154\n",
      "[430]\tTest-mae:1.14144\n",
      "[431]\tTest-mae:1.14132\n",
      "[432]\tTest-mae:1.14114\n",
      "[433]\tTest-mae:1.14096\n",
      "[434]\tTest-mae:1.14077\n",
      "[435]\tTest-mae:1.14071\n",
      "[436]\tTest-mae:1.14053\n",
      "[437]\tTest-mae:1.14049\n",
      "[438]\tTest-mae:1.14030\n",
      "[439]\tTest-mae:1.14014\n",
      "[440]\tTest-mae:1.13989\n",
      "[441]\tTest-mae:1.13968\n",
      "[442]\tTest-mae:1.13954\n",
      "[443]\tTest-mae:1.13945\n",
      "[444]\tTest-mae:1.13922\n",
      "[445]\tTest-mae:1.13911\n",
      "[446]\tTest-mae:1.13900\n",
      "[447]\tTest-mae:1.13887\n",
      "[448]\tTest-mae:1.13877\n",
      "[449]\tTest-mae:1.13868\n",
      "[450]\tTest-mae:1.13854\n",
      "[451]\tTest-mae:1.13840\n",
      "[452]\tTest-mae:1.13823\n",
      "[453]\tTest-mae:1.13803\n",
      "[454]\tTest-mae:1.13787\n",
      "[455]\tTest-mae:1.13778\n",
      "[456]\tTest-mae:1.13771\n",
      "[457]\tTest-mae:1.13764\n",
      "[458]\tTest-mae:1.13756\n",
      "[459]\tTest-mae:1.13744\n",
      "[460]\tTest-mae:1.13725\n",
      "[461]\tTest-mae:1.13717\n",
      "[462]\tTest-mae:1.13710\n",
      "[463]\tTest-mae:1.13698\n",
      "[464]\tTest-mae:1.13691\n",
      "[465]\tTest-mae:1.13681\n",
      "[466]\tTest-mae:1.13670\n",
      "[467]\tTest-mae:1.13668\n",
      "[468]\tTest-mae:1.13655\n",
      "[469]\tTest-mae:1.13653\n",
      "[470]\tTest-mae:1.13655\n",
      "[471]\tTest-mae:1.13649\n",
      "[472]\tTest-mae:1.13638\n",
      "[473]\tTest-mae:1.13623\n",
      "[474]\tTest-mae:1.13621\n",
      "[475]\tTest-mae:1.13620\n",
      "[476]\tTest-mae:1.13606\n",
      "[477]\tTest-mae:1.13597\n",
      "[478]\tTest-mae:1.13592\n",
      "[479]\tTest-mae:1.13577\n",
      "[480]\tTest-mae:1.13562\n",
      "[481]\tTest-mae:1.13559\n",
      "[482]\tTest-mae:1.13545\n",
      "[483]\tTest-mae:1.13531\n",
      "[484]\tTest-mae:1.13516\n",
      "[485]\tTest-mae:1.13510\n",
      "[486]\tTest-mae:1.13510\n",
      "[487]\tTest-mae:1.13499\n",
      "[488]\tTest-mae:1.13489\n",
      "[489]\tTest-mae:1.13476\n",
      "[490]\tTest-mae:1.13474\n",
      "[491]\tTest-mae:1.13464\n",
      "[492]\tTest-mae:1.13455\n",
      "[493]\tTest-mae:1.13453\n",
      "[494]\tTest-mae:1.13450\n",
      "[495]\tTest-mae:1.13436\n",
      "[496]\tTest-mae:1.13434\n",
      "[497]\tTest-mae:1.13428\n",
      "[498]\tTest-mae:1.13426\n",
      "[499]\tTest-mae:1.13420\n",
      "[500]\tTest-mae:1.13421\n",
      "[501]\tTest-mae:1.13413\n",
      "[502]\tTest-mae:1.13413\n",
      "[503]\tTest-mae:1.13413\n",
      "[504]\tTest-mae:1.13411\n",
      "[505]\tTest-mae:1.13402\n",
      "[506]\tTest-mae:1.13402\n",
      "[507]\tTest-mae:1.13402\n",
      "[508]\tTest-mae:1.13402\n",
      "[509]\tTest-mae:1.13397\n",
      "[510]\tTest-mae:1.13380\n",
      "[511]\tTest-mae:1.13368\n",
      "[512]\tTest-mae:1.13356\n",
      "[513]\tTest-mae:1.13344\n",
      "[514]\tTest-mae:1.13332\n",
      "[515]\tTest-mae:1.13323\n",
      "[516]\tTest-mae:1.13311\n",
      "[517]\tTest-mae:1.13304\n",
      "[518]\tTest-mae:1.13296\n",
      "[519]\tTest-mae:1.13291\n",
      "[520]\tTest-mae:1.13284\n",
      "[521]\tTest-mae:1.13266\n",
      "[522]\tTest-mae:1.13265\n",
      "[523]\tTest-mae:1.13259\n",
      "[524]\tTest-mae:1.13258\n",
      "[525]\tTest-mae:1.13257\n",
      "[526]\tTest-mae:1.13246\n",
      "[527]\tTest-mae:1.13246\n",
      "[528]\tTest-mae:1.13234\n",
      "[529]\tTest-mae:1.13230\n",
      "[530]\tTest-mae:1.13226\n",
      "[531]\tTest-mae:1.13212\n",
      "[532]\tTest-mae:1.13211\n",
      "[533]\tTest-mae:1.13211\n",
      "[534]\tTest-mae:1.13208\n",
      "[535]\tTest-mae:1.13198\n",
      "[536]\tTest-mae:1.13199\n",
      "[537]\tTest-mae:1.13192\n",
      "[538]\tTest-mae:1.13192\n",
      "[539]\tTest-mae:1.13191\n",
      "[540]\tTest-mae:1.13191\n",
      "[541]\tTest-mae:1.13180\n",
      "[542]\tTest-mae:1.13172\n",
      "[543]\tTest-mae:1.13165\n",
      "[544]\tTest-mae:1.13166\n",
      "[545]\tTest-mae:1.13159\n",
      "[546]\tTest-mae:1.13159\n",
      "[547]\tTest-mae:1.13149\n",
      "[548]\tTest-mae:1.13145\n",
      "[549]\tTest-mae:1.13133\n",
      "[550]\tTest-mae:1.13120\n",
      "[551]\tTest-mae:1.13119\n",
      "[552]\tTest-mae:1.13118\n",
      "[553]\tTest-mae:1.13111\n",
      "[554]\tTest-mae:1.13110\n",
      "[555]\tTest-mae:1.13100\n",
      "[556]\tTest-mae:1.13100\n",
      "[557]\tTest-mae:1.13085\n",
      "[558]\tTest-mae:1.13087\n",
      "[559]\tTest-mae:1.13087\n",
      "[560]\tTest-mae:1.13091\n",
      "[561]\tTest-mae:1.13086\n",
      "[562]\tTest-mae:1.13088\n",
      "[563]\tTest-mae:1.13091\n",
      "[564]\tTest-mae:1.13094\n",
      "[565]\tTest-mae:1.13087\n",
      "[566]\tTest-mae:1.13085\n",
      "[567]\tTest-mae:1.13080\n",
      "[568]\tTest-mae:1.13078\n",
      "[569]\tTest-mae:1.13081\n",
      "[570]\tTest-mae:1.13085\n",
      "[571]\tTest-mae:1.13086\n",
      "[572]\tTest-mae:1.13082\n",
      "[573]\tTest-mae:1.13076\n",
      "[574]\tTest-mae:1.13074\n",
      "[575]\tTest-mae:1.13072\n",
      "[576]\tTest-mae:1.13068\n",
      "[577]\tTest-mae:1.13066\n",
      "[578]\tTest-mae:1.13062\n",
      "[579]\tTest-mae:1.13057\n",
      "[580]\tTest-mae:1.13060\n",
      "[581]\tTest-mae:1.13057\n",
      "[582]\tTest-mae:1.13052\n",
      "[583]\tTest-mae:1.13053\n",
      "[584]\tTest-mae:1.13051\n",
      "[585]\tTest-mae:1.13047\n",
      "[586]\tTest-mae:1.13043\n",
      "[587]\tTest-mae:1.13037\n",
      "[588]\tTest-mae:1.13032\n",
      "[589]\tTest-mae:1.13030\n",
      "[590]\tTest-mae:1.13025\n",
      "[591]\tTest-mae:1.13024\n",
      "[592]\tTest-mae:1.13021\n",
      "[593]\tTest-mae:1.13020\n",
      "[594]\tTest-mae:1.13020\n",
      "[595]\tTest-mae:1.13015\n",
      "[596]\tTest-mae:1.13013\n",
      "[597]\tTest-mae:1.13008\n",
      "[598]\tTest-mae:1.13008\n",
      "[599]\tTest-mae:1.13005\n",
      "[600]\tTest-mae:1.13006\n",
      "[601]\tTest-mae:1.13001\n",
      "[602]\tTest-mae:1.12999\n",
      "[603]\tTest-mae:1.12993\n",
      "[604]\tTest-mae:1.12985\n",
      "[605]\tTest-mae:1.12982\n",
      "[606]\tTest-mae:1.12981\n",
      "[607]\tTest-mae:1.12976\n",
      "[608]\tTest-mae:1.12981\n",
      "[609]\tTest-mae:1.12978\n",
      "[610]\tTest-mae:1.12975\n",
      "[611]\tTest-mae:1.12981\n",
      "[612]\tTest-mae:1.12976\n",
      "[613]\tTest-mae:1.12972\n",
      "[614]\tTest-mae:1.12968\n",
      "[615]\tTest-mae:1.12966\n",
      "[616]\tTest-mae:1.12962\n",
      "[617]\tTest-mae:1.12959\n",
      "[618]\tTest-mae:1.12958\n",
      "[619]\tTest-mae:1.12964\n",
      "[620]\tTest-mae:1.12960\n",
      "[621]\tTest-mae:1.12959\n",
      "[622]\tTest-mae:1.12960\n",
      "[623]\tTest-mae:1.12961\n",
      "[624]\tTest-mae:1.12958\n",
      "[625]\tTest-mae:1.12956\n",
      "[626]\tTest-mae:1.12963\n",
      "[627]\tTest-mae:1.12965\n",
      "[628]\tTest-mae:1.12960\n",
      "[629]\tTest-mae:1.12957\n",
      "[630]\tTest-mae:1.12960\n",
      "[631]\tTest-mae:1.12953\n",
      "[632]\tTest-mae:1.12953\n",
      "[633]\tTest-mae:1.12950\n",
      "[634]\tTest-mae:1.12945\n",
      "[635]\tTest-mae:1.12943\n",
      "[636]\tTest-mae:1.12941\n",
      "[637]\tTest-mae:1.12937\n",
      "[638]\tTest-mae:1.12932\n",
      "[639]\tTest-mae:1.12929\n",
      "[640]\tTest-mae:1.12927\n",
      "[641]\tTest-mae:1.12927\n",
      "[642]\tTest-mae:1.12927\n",
      "[643]\tTest-mae:1.12927\n",
      "[644]\tTest-mae:1.12925\n",
      "[645]\tTest-mae:1.12930\n",
      "[646]\tTest-mae:1.12928\n",
      "[647]\tTest-mae:1.12929\n",
      "[648]\tTest-mae:1.12929\n",
      "[649]\tTest-mae:1.12926\n",
      "[650]\tTest-mae:1.12926\n",
      "[651]\tTest-mae:1.12922\n",
      "[652]\tTest-mae:1.12922\n",
      "[653]\tTest-mae:1.12924\n",
      "[654]\tTest-mae:1.12919\n",
      "[655]\tTest-mae:1.12918\n",
      "[656]\tTest-mae:1.12919\n",
      "[657]\tTest-mae:1.12915\n",
      "[658]\tTest-mae:1.12912\n",
      "[659]\tTest-mae:1.12913\n",
      "[660]\tTest-mae:1.12913\n",
      "[661]\tTest-mae:1.12912\n",
      "[662]\tTest-mae:1.12910\n",
      "[663]\tTest-mae:1.12907\n",
      "[664]\tTest-mae:1.12909\n",
      "[665]\tTest-mae:1.12911\n",
      "[666]\tTest-mae:1.12911\n",
      "[667]\tTest-mae:1.12909\n",
      "[668]\tTest-mae:1.12911\n",
      "[669]\tTest-mae:1.12914\n",
      "[670]\tTest-mae:1.12910\n",
      "[671]\tTest-mae:1.12911\n",
      "[672]\tTest-mae:1.12906\n",
      "[673]\tTest-mae:1.12896\n",
      "[674]\tTest-mae:1.12902\n",
      "[675]\tTest-mae:1.12909\n",
      "[676]\tTest-mae:1.12916\n",
      "[677]\tTest-mae:1.12915\n",
      "[678]\tTest-mae:1.12911\n",
      "[679]\tTest-mae:1.12908\n",
      "[680]\tTest-mae:1.12908\n",
      "[681]\tTest-mae:1.12898\n",
      "[682]\tTest-mae:1.12899\n",
      "[683]\tTest-mae:1.12900\n",
      "[684]\tTest-mae:1.12896\n",
      "[685]\tTest-mae:1.12890\n",
      "[686]\tTest-mae:1.12887\n",
      "[687]\tTest-mae:1.12887\n",
      "[688]\tTest-mae:1.12882\n",
      "[689]\tTest-mae:1.12876\n",
      "[690]\tTest-mae:1.12872\n",
      "[691]\tTest-mae:1.12867\n",
      "[692]\tTest-mae:1.12861\n",
      "[693]\tTest-mae:1.12859\n",
      "[694]\tTest-mae:1.12851\n",
      "[695]\tTest-mae:1.12849\n",
      "[696]\tTest-mae:1.12847\n",
      "[697]\tTest-mae:1.12842\n",
      "[698]\tTest-mae:1.12839\n",
      "[699]\tTest-mae:1.12841\n",
      "[700]\tTest-mae:1.12837\n",
      "[701]\tTest-mae:1.12831\n",
      "[702]\tTest-mae:1.12827\n",
      "[703]\tTest-mae:1.12819\n",
      "[704]\tTest-mae:1.12815\n",
      "[705]\tTest-mae:1.12813\n",
      "[706]\tTest-mae:1.12807\n",
      "[707]\tTest-mae:1.12803\n",
      "[708]\tTest-mae:1.12796\n",
      "[709]\tTest-mae:1.12792\n",
      "[710]\tTest-mae:1.12787\n",
      "[711]\tTest-mae:1.12786\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[712]\tTest-mae:1.12784\n",
      "[713]\tTest-mae:1.12779\n",
      "[714]\tTest-mae:1.12773\n",
      "[715]\tTest-mae:1.12778\n",
      "[716]\tTest-mae:1.12782\n",
      "[717]\tTest-mae:1.12777\n",
      "[718]\tTest-mae:1.12778\n",
      "[719]\tTest-mae:1.12769\n",
      "[720]\tTest-mae:1.12767\n",
      "[721]\tTest-mae:1.12768\n",
      "[722]\tTest-mae:1.12771\n",
      "[723]\tTest-mae:1.12768\n",
      "[724]\tTest-mae:1.12762\n",
      "[725]\tTest-mae:1.12755\n",
      "[726]\tTest-mae:1.12756\n",
      "[727]\tTest-mae:1.12754\n",
      "[728]\tTest-mae:1.12749\n",
      "[729]\tTest-mae:1.12752\n",
      "[730]\tTest-mae:1.12747\n",
      "[731]\tTest-mae:1.12751\n",
      "[732]\tTest-mae:1.12748\n",
      "[733]\tTest-mae:1.12743\n",
      "[734]\tTest-mae:1.12744\n",
      "[735]\tTest-mae:1.12743\n",
      "[736]\tTest-mae:1.12745\n",
      "[737]\tTest-mae:1.12738\n",
      "[738]\tTest-mae:1.12737\n",
      "[739]\tTest-mae:1.12733\n",
      "[740]\tTest-mae:1.12734\n",
      "[741]\tTest-mae:1.12727\n",
      "[742]\tTest-mae:1.12731\n",
      "[743]\tTest-mae:1.12730\n",
      "[744]\tTest-mae:1.12722\n",
      "[745]\tTest-mae:1.12712\n",
      "[746]\tTest-mae:1.12712\n",
      "[747]\tTest-mae:1.12708\n",
      "[748]\tTest-mae:1.12704\n",
      "[749]\tTest-mae:1.12698\n",
      "[750]\tTest-mae:1.12695\n",
      "[751]\tTest-mae:1.12693\n",
      "[752]\tTest-mae:1.12685\n",
      "[753]\tTest-mae:1.12680\n",
      "[754]\tTest-mae:1.12673\n",
      "[755]\tTest-mae:1.12669\n",
      "[756]\tTest-mae:1.12666\n",
      "[757]\tTest-mae:1.12667\n",
      "[758]\tTest-mae:1.12660\n",
      "[759]\tTest-mae:1.12657\n",
      "[760]\tTest-mae:1.12657\n",
      "[761]\tTest-mae:1.12652\n",
      "[762]\tTest-mae:1.12652\n",
      "[763]\tTest-mae:1.12647\n",
      "[764]\tTest-mae:1.12643\n",
      "[765]\tTest-mae:1.12644\n",
      "[766]\tTest-mae:1.12639\n",
      "[767]\tTest-mae:1.12637\n",
      "[768]\tTest-mae:1.12635\n",
      "[769]\tTest-mae:1.12634\n",
      "[770]\tTest-mae:1.12635\n",
      "[771]\tTest-mae:1.12633\n",
      "[772]\tTest-mae:1.12630\n",
      "[773]\tTest-mae:1.12629\n",
      "[774]\tTest-mae:1.12629\n",
      "[775]\tTest-mae:1.12629\n",
      "[776]\tTest-mae:1.12627\n",
      "[777]\tTest-mae:1.12621\n",
      "[778]\tTest-mae:1.12620\n",
      "[779]\tTest-mae:1.12616\n",
      "[780]\tTest-mae:1.12617\n",
      "[781]\tTest-mae:1.12618\n",
      "[782]\tTest-mae:1.12618\n",
      "[783]\tTest-mae:1.12616\n",
      "[784]\tTest-mae:1.12610\n",
      "[785]\tTest-mae:1.12604\n",
      "[786]\tTest-mae:1.12610\n",
      "[787]\tTest-mae:1.12608\n",
      "[788]\tTest-mae:1.12608\n",
      "[789]\tTest-mae:1.12607\n",
      "[790]\tTest-mae:1.12605\n",
      "[791]\tTest-mae:1.12602\n",
      "[792]\tTest-mae:1.12599\n",
      "[793]\tTest-mae:1.12598\n",
      "[794]\tTest-mae:1.12603\n",
      "[795]\tTest-mae:1.12601\n",
      "[796]\tTest-mae:1.12597\n",
      "[797]\tTest-mae:1.12595\n",
      "[798]\tTest-mae:1.12597\n",
      "[799]\tTest-mae:1.12595\n",
      "[800]\tTest-mae:1.12590\n",
      "[801]\tTest-mae:1.12592\n",
      "[802]\tTest-mae:1.12588\n",
      "[803]\tTest-mae:1.12586\n",
      "[804]\tTest-mae:1.12591\n",
      "[805]\tTest-mae:1.12588\n",
      "[806]\tTest-mae:1.12588\n",
      "[807]\tTest-mae:1.12586\n",
      "[808]\tTest-mae:1.12584\n",
      "[809]\tTest-mae:1.12583\n",
      "[810]\tTest-mae:1.12581\n",
      "[811]\tTest-mae:1.12580\n",
      "[812]\tTest-mae:1.12579\n",
      "[813]\tTest-mae:1.12582\n",
      "[814]\tTest-mae:1.12580\n",
      "[815]\tTest-mae:1.12576\n",
      "[816]\tTest-mae:1.12574\n",
      "[817]\tTest-mae:1.12573\n",
      "[818]\tTest-mae:1.12572\n",
      "[819]\tTest-mae:1.12569\n",
      "[820]\tTest-mae:1.12571\n",
      "[821]\tTest-mae:1.12570\n",
      "[822]\tTest-mae:1.12568\n",
      "[823]\tTest-mae:1.12565\n",
      "[824]\tTest-mae:1.12564\n",
      "[825]\tTest-mae:1.12562\n",
      "[826]\tTest-mae:1.12566\n",
      "[827]\tTest-mae:1.12565\n",
      "[828]\tTest-mae:1.12562\n",
      "[829]\tTest-mae:1.12557\n",
      "[830]\tTest-mae:1.12558\n",
      "[831]\tTest-mae:1.12557\n",
      "[832]\tTest-mae:1.12556\n",
      "[833]\tTest-mae:1.12558\n",
      "[834]\tTest-mae:1.12557\n",
      "[835]\tTest-mae:1.12564\n",
      "[836]\tTest-mae:1.12563\n",
      "[837]\tTest-mae:1.12561\n",
      "[838]\tTest-mae:1.12561\n",
      "[839]\tTest-mae:1.12560\n",
      "[840]\tTest-mae:1.12558\n",
      "[841]\tTest-mae:1.12551\n",
      "[842]\tTest-mae:1.12551\n",
      "[843]\tTest-mae:1.12548\n",
      "[844]\tTest-mae:1.12546\n",
      "[845]\tTest-mae:1.12545\n",
      "[846]\tTest-mae:1.12543\n",
      "[847]\tTest-mae:1.12542\n",
      "[848]\tTest-mae:1.12537\n",
      "[849]\tTest-mae:1.12546\n",
      "[850]\tTest-mae:1.12549\n",
      "[851]\tTest-mae:1.12548\n",
      "[852]\tTest-mae:1.12557\n",
      "[853]\tTest-mae:1.12556\n",
      "[854]\tTest-mae:1.12555\n",
      "[855]\tTest-mae:1.12552\n",
      "[856]\tTest-mae:1.12552\n",
      "[857]\tTest-mae:1.12552\n",
      "[858]\tTest-mae:1.12556\n",
      "[859]\tTest-mae:1.12556\n",
      "[860]\tTest-mae:1.12559\n",
      "[861]\tTest-mae:1.12558\n",
      "[862]\tTest-mae:1.12562\n",
      "[863]\tTest-mae:1.12563\n",
      "[864]\tTest-mae:1.12562\n",
      "[865]\tTest-mae:1.12561\n",
      "[866]\tTest-mae:1.12559\n",
      "[867]\tTest-mae:1.12557\n",
      "[868]\tTest-mae:1.12563\n",
      "[869]\tTest-mae:1.12562\n",
      "[870]\tTest-mae:1.12561\n",
      "[871]\tTest-mae:1.12561\n",
      "[872]\tTest-mae:1.12560\n",
      "[873]\tTest-mae:1.12560\n",
      "[874]\tTest-mae:1.12559\n",
      "[875]\tTest-mae:1.12560\n",
      "[876]\tTest-mae:1.12561\n",
      "[877]\tTest-mae:1.12562\n",
      "[878]\tTest-mae:1.12562\n",
      "[879]\tTest-mae:1.12560\n",
      "[880]\tTest-mae:1.12560\n",
      "[881]\tTest-mae:1.12560\n",
      "[882]\tTest-mae:1.12554\n",
      "[883]\tTest-mae:1.12554\n",
      "[884]\tTest-mae:1.12550\n",
      "[885]\tTest-mae:1.12552\n",
      "[886]\tTest-mae:1.12550\n",
      "[887]\tTest-mae:1.12549\n",
      "[888]\tTest-mae:1.12551\n",
      "[889]\tTest-mae:1.12550\n",
      "[890]\tTest-mae:1.12551\n",
      "[891]\tTest-mae:1.12554\n",
      "[892]\tTest-mae:1.12552\n",
      "[893]\tTest-mae:1.12551\n",
      "[894]\tTest-mae:1.12552\n",
      "[895]\tTest-mae:1.12552\n",
      "[896]\tTest-mae:1.12551\n",
      "[897]\tTest-mae:1.12551\n",
      "[898]\tTest-mae:1.12550\n",
      "[899]\tTest-mae:1.12550\n",
      "[900]\tTest-mae:1.12550\n",
      "[901]\tTest-mae:1.12549\n",
      "[902]\tTest-mae:1.12550\n",
      "[903]\tTest-mae:1.12550\n",
      "[904]\tTest-mae:1.12553\n",
      "[905]\tTest-mae:1.12551\n",
      "[906]\tTest-mae:1.12553\n",
      "[907]\tTest-mae:1.12553\n",
      "[908]\tTest-mae:1.12551\n",
      "[909]\tTest-mae:1.12554\n",
      "[910]\tTest-mae:1.12556\n",
      "[911]\tTest-mae:1.12557\n",
      "[912]\tTest-mae:1.12559\n",
      "[913]\tTest-mae:1.12561\n",
      "[914]\tTest-mae:1.12560\n",
      "[915]\tTest-mae:1.12564\n",
      "[916]\tTest-mae:1.12562\n",
      "[917]\tTest-mae:1.12564\n",
      "[918]\tTest-mae:1.12564\n",
      "[919]\tTest-mae:1.12563\n",
      "[920]\tTest-mae:1.12563\n",
      "[921]\tTest-mae:1.12566\n",
      "[922]\tTest-mae:1.12560\n",
      "[923]\tTest-mae:1.12559\n",
      "[924]\tTest-mae:1.12559\n",
      "[925]\tTest-mae:1.12561\n",
      "[926]\tTest-mae:1.12563\n",
      "[927]\tTest-mae:1.12560\n",
      "[928]\tTest-mae:1.12561\n",
      "[929]\tTest-mae:1.12562\n",
      "[930]\tTest-mae:1.12562\n",
      "[931]\tTest-mae:1.12566\n",
      "[932]\tTest-mae:1.12566\n",
      "[933]\tTest-mae:1.12565\n",
      "[934]\tTest-mae:1.12562\n",
      "[935]\tTest-mae:1.12558\n",
      "[936]\tTest-mae:1.12557\n",
      "[937]\tTest-mae:1.12558\n",
      "[938]\tTest-mae:1.12556\n",
      "[939]\tTest-mae:1.12557\n",
      "[940]\tTest-mae:1.12555\n",
      "[941]\tTest-mae:1.12551\n",
      "[942]\tTest-mae:1.12549\n",
      "[943]\tTest-mae:1.12548\n",
      "[944]\tTest-mae:1.12549\n",
      "[945]\tTest-mae:1.12549\n",
      "[946]\tTest-mae:1.12549\n",
      "[947]\tTest-mae:1.12544\n",
      "[948]\tTest-mae:1.12542\n",
      "[949]\tTest-mae:1.12541\n",
      "[950]\tTest-mae:1.12537\n",
      "[951]\tTest-mae:1.12535\n",
      "[952]\tTest-mae:1.12532\n",
      "[953]\tTest-mae:1.12535\n",
      "[954]\tTest-mae:1.12536\n",
      "[955]\tTest-mae:1.12531\n",
      "[956]\tTest-mae:1.12530\n",
      "[957]\tTest-mae:1.12530\n",
      "[958]\tTest-mae:1.12527\n",
      "[959]\tTest-mae:1.12529\n",
      "[960]\tTest-mae:1.12529\n",
      "[961]\tTest-mae:1.12530\n",
      "[962]\tTest-mae:1.12525\n",
      "[963]\tTest-mae:1.12523\n",
      "[964]\tTest-mae:1.12516\n",
      "[965]\tTest-mae:1.12514\n",
      "[966]\tTest-mae:1.12509\n",
      "[967]\tTest-mae:1.12508\n",
      "[968]\tTest-mae:1.12503\n",
      "[969]\tTest-mae:1.12502\n",
      "[970]\tTest-mae:1.12501\n",
      "[971]\tTest-mae:1.12501\n",
      "[972]\tTest-mae:1.12502\n",
      "[973]\tTest-mae:1.12501\n",
      "[974]\tTest-mae:1.12500\n",
      "[975]\tTest-mae:1.12499\n",
      "[976]\tTest-mae:1.12495\n",
      "[977]\tTest-mae:1.12492\n",
      "[978]\tTest-mae:1.12488\n",
      "[979]\tTest-mae:1.12487\n",
      "[980]\tTest-mae:1.12482\n",
      "[981]\tTest-mae:1.12478\n",
      "[982]\tTest-mae:1.12478\n",
      "[983]\tTest-mae:1.12473\n",
      "[984]\tTest-mae:1.12469\n",
      "[985]\tTest-mae:1.12468\n",
      "[986]\tTest-mae:1.12461\n",
      "[987]\tTest-mae:1.12457\n",
      "[988]\tTest-mae:1.12454\n",
      "[989]\tTest-mae:1.12458\n",
      "[990]\tTest-mae:1.12453\n",
      "[991]\tTest-mae:1.12452\n",
      "[992]\tTest-mae:1.12449\n",
      "[993]\tTest-mae:1.12444\n",
      "[994]\tTest-mae:1.12439\n",
      "[995]\tTest-mae:1.12439\n",
      "[996]\tTest-mae:1.12434\n",
      "[997]\tTest-mae:1.12435\n",
      "[998]\tTest-mae:1.12431\n"
     ]
    }
   ],
   "source": [
    "best_model = xgb.train(\n",
    "    params,\n",
    "    dtrain,\n",
    "    num_boost_round=num_boost_round,\n",
    "    evals=[(dtest, \"Test\")]\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.1243123665942847"
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics.mean_absolute_error(best_model.predict(dtest), y_test)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score XGBoost: 0.9448749658031844\n"
     ]
    }
   ],
   "source": [
    "print(\"Explained Variance Score XGBoost:\", metrics.explained_variance_score(best_model.predict(dtest), y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fdba4ecab70>"
      ]
     },
     "execution_count": 247,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from xgboost import plot_importance\n",
    "fig, ax = plt.subplots(figsize=(14, 12))\n",
    "plot_importance(best_model, ax=ax, importance_type='gain')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_model.save_model(\"xgb_best.model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "loaded_model = xgb.Booster()\n",
    "loaded_model.load_model(\"xgb_best.model\")\n",
    "# And use it for predictions.\n",
    "loaded_model.predict(dtest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# !pip install graphviz\n",
    "from xgboost import plot_tree\n",
    "plot_tree(best_model)\n",
    "fig = plt.gcf()\n",
    "fig.set_size_inches(300, 200)\n",
    "# plt.figure(figsize=(10,8))\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}