ClemMLTemplateHR.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Clem's Machine Learning Template\n",
    "## HR Data\n",
    "\n",
    "This is what Clem uses to scout out an initial data set for classification to get a \"lay of the land\".\n",
    "\n",
    "While, it is unwise to favor a particular ML methodology, H2O's GBM has a lot of advantages that makes it really good for a first pass:\n",
    "\n",
    "* Fast\n",
    "* Robust to overfitting over a wide range of hyper parameters\n",
    "* Handles both categorical and continous variables\n",
    "* No need to scale the data\n",
    "\n",
    "Change sections that are labeled:\n",
    "\n",
    "    #@@@@@ CHANGE FOR SPECIFIC PROBLEM\n",
    "\n",
    "This data is from https://www.kaggle.com/ludobenistant/hr-analytics\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Python Libraries\n",
    "from __future__ import division   # So that / does floating point division\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import scipy\n",
    "import commands\n",
    "import math\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from statsmodels.graphics.mosaicplot import mosaic\n",
    "import pylab\n",
    "\n",
    "import h2o\n",
    "\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.metrics import roc_auc_score\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#@@@@@ CHANGE FOR SPECIFIC PROBLEM\n",
    "\n",
    "# Column names in case they aren't in an header\n",
    "\n",
    "column_names =[\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#@@@@@ CHANGE FOR SPECIFIC PROBLEM\n",
    "\n",
    "# http://pandas.pydata.org/pandas-docs/stable/generated/pandas.read_csv.html\n",
    "\n",
    "# Very Problem specific\n",
    "\n",
    "df = pd.read_csv(\n",
    "    \"HR_comma_sep.csv\",\n",
    "    sep=\",\",\n",
    "    header=0,   # Row the header is on or None if no header present\n",
    "    # names=column_names,\n",
    "    # skipinitialspace = True  # Skip spaces after delimiter\n",
    ")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['satisfaction_level', 'last_evaluation', 'number_project',\n",
       "       'average_montly_hours', 'time_spend_company', 'Work_accident',\n",
       "       'left', 'promotion_last_5years', 'sales', 'salary'], dtype=object)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Names of columns - check to see if we got it right\n",
    "df.columns.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>satisfaction_level</th>\n",
       "      <th>last_evaluation</th>\n",
       "      <th>number_project</th>\n",
       "      <th>average_montly_hours</th>\n",
       "      <th>time_spend_company</th>\n",
       "      <th>Work_accident</th>\n",
       "      <th>left</th>\n",
       "      <th>promotion_last_5years</th>\n",
       "      <th>sales</th>\n",
       "      <th>salary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.38</td>\n",
       "      <td>0.53</td>\n",
       "      <td>2</td>\n",
       "      <td>157</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>sales</td>\n",
       "      <td>low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.80</td>\n",
       "      <td>0.86</td>\n",
       "      <td>5</td>\n",
       "      <td>262</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>sales</td>\n",
       "      <td>medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.11</td>\n",
       "      <td>0.88</td>\n",
       "      <td>7</td>\n",
       "      <td>272</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>sales</td>\n",
       "      <td>medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.72</td>\n",
       "      <td>0.87</td>\n",
       "      <td>5</td>\n",
       "      <td>223</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>sales</td>\n",
       "      <td>low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.37</td>\n",
       "      <td>0.52</td>\n",
       "      <td>2</td>\n",
       "      <td>159</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>sales</td>\n",
       "      <td>low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   satisfaction_level  last_evaluation  number_project  average_montly_hours  \\\n",
       "0                0.38             0.53               2                   157   \n",
       "1                0.80             0.86               5                   262   \n",
       "2                0.11             0.88               7                   272   \n",
       "3                0.72             0.87               5                   223   \n",
       "4                0.37             0.52               2                   159   \n",
       "\n",
       "   time_spend_company  Work_accident  left  promotion_last_5years  sales  \\\n",
       "0                   3              0     1                      0  sales   \n",
       "1                   6              0     1                      0  sales   \n",
       "2                   4              0     1                      0  sales   \n",
       "3                   5              0     1                      0  sales   \n",
       "4                   3              0     1                      0  sales   \n",
       "\n",
       "   salary  \n",
       "0     low  \n",
       "1  medium  \n",
       "2  medium  \n",
       "3     low  \n",
       "4     low  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#@@@@@ CHANGE FOR SPECIFIC PROBLEM\n",
    "# Drop any columns that don't make sense, if any\n",
    "if False:\n",
    "    df = df.drop(\"COLUMN-NAME-THAT-IS-USELESS\",  #\n",
    "                 1)  # 1 means drops in column direction\n",
    "    \n",
    "# Fix up any data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get Names of Columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem Specific\n",
    "These are the values that are problem specific.  We use variable names from now on so that the code can be reused."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#@@@@@ CHANGE FOR SPECIFIC PROBLEM\n",
    "YName = 'left'     # Name of Response Variable\n",
    "XUnknowCategoricalValue = None  # or string that represents no categorical\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Automatically convert response variable that are ints\n",
    "# into catgorical values\n",
    "if df[YName].dtype == 'int64':\n",
    "    df[YName] = df[YName].astype(\"category\", ordered=False)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Programmatically get list of names of:\n",
    "* numeric features\n",
    "* non-numeric features\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-- Non-Numeric Input Feature Names:\n",
      "['sales', 'salary']\n",
      "\n",
      "-- Numeric Input Feature Names:\n",
      "['satisfaction_level', 'last_evaluation', 'number_project', 'average_montly_hours', 'time_spend_company', 'Work_accident', 'promotion_last_5years']\n",
      "\n",
      "-- All Input Feature Names:\n",
      "['satisfaction_level', 'last_evaluation', 'number_project', 'average_montly_hours', 'time_spend_company', 'Work_accident', 'promotion_last_5years', 'sales', 'salary']\n",
      "\n",
      "-- Output (Response) Variable name:\n",
      "left\n",
      "\n",
      "-- All possible values of Output (Response Variable):\n",
      "[1, 0]\n"
     ]
    }
   ],
   "source": [
    "# Names of all the input features \n",
    "XNames = [x for x in df.columns.values if x != YName]\n",
    "\n",
    "# Names of all the Non-Numeric input features\n",
    "XNonNumerics = [x for x in XNames \\\n",
    "                if not (df[x].dtype == np.float64 \\\n",
    "                        or df[x].dtype == np.int64)]\n",
    "\n",
    "# Names of all the Non-Numeric input features\n",
    "XNumerics = [x for x in XNames if x not in XNonNumerics]\n",
    "\n",
    "# All the different Classification values that Y can take on\n",
    "YValues = list(df[YName].unique())\n",
    "\n",
    "print(\"-- Non-Numeric Input Feature Names:\")\n",
    "print(XNonNumerics)   # Of type pandas.core.series.Series\n",
    "print(\"\\n-- Numeric Input Feature Names:\")\n",
    "print(XNumerics)\n",
    "print(\"\\n-- All Input Feature Names:\")\n",
    "print(XNames)\n",
    "print(\"\\n-- Output (Response) Variable name:\")\n",
    "print(YName)\n",
    "print(\"\\n-- All possible values of Output (Response Variable):\")\n",
    "print(YValues)   # Classification values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Exploration\n",
    "\n",
    "Here, we plot the data to get a sense of which data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "satisfaction_level       float64\n",
       "last_evaluation          float64\n",
       "number_project             int64\n",
       "average_montly_hours       int64\n",
       "time_spend_company         int64\n",
       "Work_accident              int64\n",
       "promotion_last_5years      int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[XNumerics].dtypes "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 is 23.8%\n",
      "0 is 76.2%\n"
     ]
    }
   ],
   "source": [
    "# how balanced is the data\n",
    "for w in YValues:\n",
    "    s = sum(list(df[YName] == w))\n",
    "    print(\"{} is {:.1f}%\".format(w,\n",
    "                             100*s/len(df)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Numeric Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>satisfaction_level</th>\n",
       "      <td>14999.0</td>\n",
       "      <td>0.612834</td>\n",
       "      <td>0.248631</td>\n",
       "      <td>0.09</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.64</td>\n",
       "      <td>0.82</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_evaluation</th>\n",
       "      <td>14999.0</td>\n",
       "      <td>0.716102</td>\n",
       "      <td>0.171169</td>\n",
       "      <td>0.36</td>\n",
       "      <td>0.56</td>\n",
       "      <td>0.72</td>\n",
       "      <td>0.87</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>number_project</th>\n",
       "      <td>14999.0</td>\n",
       "      <td>3.803054</td>\n",
       "      <td>1.232592</td>\n",
       "      <td>2.00</td>\n",
       "      <td>3.00</td>\n",
       "      <td>4.00</td>\n",
       "      <td>5.00</td>\n",
       "      <td>7.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average_montly_hours</th>\n",
       "      <td>14999.0</td>\n",
       "      <td>201.050337</td>\n",
       "      <td>49.943099</td>\n",
       "      <td>96.00</td>\n",
       "      <td>156.00</td>\n",
       "      <td>200.00</td>\n",
       "      <td>245.00</td>\n",
       "      <td>310.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time_spend_company</th>\n",
       "      <td>14999.0</td>\n",
       "      <td>3.498233</td>\n",
       "      <td>1.460136</td>\n",
       "      <td>2.00</td>\n",
       "      <td>3.00</td>\n",
       "      <td>3.00</td>\n",
       "      <td>4.00</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Work_accident</th>\n",
       "      <td>14999.0</td>\n",
       "      <td>0.144610</td>\n",
       "      <td>0.351719</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>promotion_last_5years</th>\n",
       "      <td>14999.0</td>\n",
       "      <td>0.021268</td>\n",
       "      <td>0.144281</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         count        mean        std    min     25%     50%  \\\n",
       "satisfaction_level     14999.0    0.612834   0.248631   0.09    0.44    0.64   \n",
       "last_evaluation        14999.0    0.716102   0.171169   0.36    0.56    0.72   \n",
       "number_project         14999.0    3.803054   1.232592   2.00    3.00    4.00   \n",
       "average_montly_hours   14999.0  201.050337  49.943099  96.00  156.00  200.00   \n",
       "time_spend_company     14999.0    3.498233   1.460136   2.00    3.00    3.00   \n",
       "Work_accident          14999.0    0.144610   0.351719   0.00    0.00    0.00   \n",
       "promotion_last_5years  14999.0    0.021268   0.144281   0.00    0.00    0.00   \n",
       "\n",
       "                          75%    max  \n",
       "satisfaction_level       0.82    1.0  \n",
       "last_evaluation          0.87    1.0  \n",
       "number_project           5.00    7.0  \n",
       "average_montly_hours   245.00  310.0  \n",
       "time_spend_company       4.00   10.0  \n",
       "Work_accident            0.00    1.0  \n",
       "promotion_last_5years    0.00    1.0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[XNumerics].describe().transpose()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "satisfaction_level has 92 float64 uniques in a range 0.9\n",
      "last_evaluation has 65 float64 uniques in a range 0.6\n",
      "number_project has 6 int uniques with a density of 100.0% in range 6\n",
      "average_montly_hours has 215 int uniques with a density of 100.0% in range 215\n",
      "time_spend_company has 8 int uniques with a density of 88.9% in range 9\n",
      "Work_accident has 2 int uniques with a density of 100.0% in range 2\n",
      "promotion_last_5years has 2 int uniques with a density of 100.0% in range 2\n"
     ]
    }
   ],
   "source": [
    "for xname in XNumerics:\n",
    "    uniques = len(df[xname].unique())\n",
    "    dtype = df[xname].dtypes\n",
    "    if dtype == 'int64':\n",
    "        range_x = df[xname].max() - df[xname].min() + 1\n",
    "        print(\n",
    "            \"{} has {} int uniques with a density of {:.1f}% in range {}\".format(\n",
    "            xname, uniques, 100.0*uniques/range_x, range_x))\n",
    "    else:\n",
    "        range_x = df[xname].max() - df[xname].min()\n",
    "        print(\"{} has {} float64 uniques in a range {:.1f}\".format(\n",
    "            xname, uniques, range_x))\n",
    "       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Number of unknown Numeric Values\n",
    "nrow = len(df)\n",
    "for xname in XNumerics:\n",
    "    count =  (df[xname].apply(math.isnan)).sum()\n",
    "    if count > 0:\n",
    "        print (\"{} has {} '{}'s which is {:.1f}%\".format(\n",
    "            xname,\n",
    "            count,\n",
    "            XUnknowCategoricalValue,\n",
    "            100.0*count/nrow))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEGCAYAAABsLkJ6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf2ca6d9d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEGCAYAAABsLkJ6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf27176a50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEJCAYAAABv6GdPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCyFEAZLkLoQQBej/ByJDFp4TtRBLAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf270154d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAEGCAYAAACHGfl5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nzIAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf26f70b50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEGCAYAAACJnEVTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf26f02f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEGCAYAAABsLkJ6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf26e41ad0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEGCAYAAABsLkJ6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRBLq/wN0ZAfEaLr7dgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf26d3d5d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Violin Plot of numeric data\n",
    "for x in XNumerics:\n",
    "    sns.violinplot(y = df[x],\n",
    "                   x = df[YName],\n",
    "                   # Width indicates relative count.  >50K is less frequent\n",
    "                   scale=\"count\" , \n",
    "                   inner=\"box\",\n",
    "                   palette=\"muted\",\n",
    "                   split=True\n",
    "                  )\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Non-Numeric Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sales has 10 uniques\n",
      "salary has 3 uniques\n"
     ]
    }
   ],
   "source": [
    "for xname in XNonNumerics:\n",
    "    uniques = len(df[xname].unique())\n",
    "    print(\"{} has {} uniques\".format(\n",
    "        xname, uniques))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Number of unknown NonNumeric Values\n",
    "nrow = len(df)\n",
    "for xname in XNonNumerics:\n",
    "    count =  (df[xname] == XUnknowCategoricalValue).sum()\n",
    "    if count > 0:\n",
    "        print (\"{} has {} '{}'s which is {:.1f}%\".format(\n",
    "            xname,\n",
    "            count,\n",
    "            XUnknowCategoricalValue,\n",
    "            100.0*count/nrow))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>unique</th>\n",
       "      <th>top</th>\n",
       "      <th>freq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>sales</th>\n",
       "      <td>14999</td>\n",
       "      <td>10</td>\n",
       "      <td>sales</td>\n",
       "      <td>4140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>salary</th>\n",
       "      <td>14999</td>\n",
       "      <td>3</td>\n",
       "      <td>low</td>\n",
       "      <td>7316</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        count unique    top  freq\n",
       "sales   14999     10  sales  4140\n",
       "salary  14999      3    low  7316"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[XNonNumerics].describe().transpose()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Number of unknown categorical values\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "nrow = len(df)\n",
    "for xname in XNonNumerics:\n",
    "    count =  (df[xname] == XUnknowCategoricalValue).sum()\n",
    "    if count > 0:\n",
    "        print (\"{} has {} Unknowns which is {:.1f}%\".format(\n",
    "            xname, count, 100.0*count/nrow))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZ4AAAEACAYAAACDEBA8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\neIQQQmSVBB4hhBBZ9f8BGuAFJasV66cAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf26ce2650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAEACAYAAABYq7oeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAEXFJREFUeJzt3WtsnfV9wPHvOce32HEckzjYIZUSSPInhAIB2hJ6AUZZ\nL2+qTnuxrtrWdlMrOqpu06bdXiBeTJqmdW3XlmlTtfXFJk1tp9LbNjqmptWgg41wWSj6kzShpMQY\nJyS28S3nnOfZi3MIbv52cYnt58T+fqQozzk+l9+Bv/318zzHTinPcyRJmqtc9ACSpNZjHCRJCeMg\nSUoYB0lSwjhIkhLGQZKUMA7SMggh3BpCuKnoOaTXq+TPOUgXLoRwF1ADNgE9wHeAGeD2OZd/EXgJ\n+ALwwRjjF4uZVnpt7jlIS+MZ4FIgB4aa13UAp+dc/mbzdr/R3JZalnGQlsZW4OXm9izwBhp7Ehvn\nXM6BfwP2xRhHihhSWiwPK0krKITwPmAqxvgfRc8i/SzGQZKU8LCSJClhHCRJibaiB1hNQghXA18H\nPh1j/HzR80ivCCF8GriJxknxT8YY/6fgkdTi3HNYIiGEHuBzwH8WPYs0VwjhFmBXjHE/8JvAXxc8\nki4CxmHpzALvBU4UPYh0ntuB+wBijE8D/SGEDcWOpFZnHJZIjLEWY5wueg5pHoPA6JzLo83rpAUZ\nB2ntKRU9gFqfcZBWvxP89J7CVmC4oFl0kTAO0ur3HeCXAUII1wMnYowTxY6kVudPSC+REMINwKeA\n7UAVeB74pRjjS0XOJQGEEP4ceAeQAb8dY3yi4JHU4oyDJCnhYSVJUsI4SJISxkGSlDAOkqSEcZAk\nJYyDJClhHCRJiVXz7zncds89LfEDG39/55185G/+pugx+PLHf4833/tXRY+hFvPvd97Ju1tgfao1\nHLv77gV/z5Z7Dktsx5YtRY8gLSi4PrVIxkGSlDAOkqSEcZAkJYyDJClhHCRJCeMgSUoYB0lSwjhI\nkhLGQZKUMA6SpIRxkCQljIMkKWEcJEkJ4yBJShgHSVLCOEiSEsZBkpQwDpKkhHGQJCWMgyQpYRwk\nSQnjIElKGAdJUsI4SJISxkGSlDAOkqSEcZAkJYyDJClhHCRJCeMgSUoYB0lSwjhIkhLGQZKUMA6S\npIRxkCQljIMkKWEcJEkJ4yBJShiHJTR6//0ATB4+XPAkUqqtuT7Lrk8tQlvRA7SiqaNHmX3xRdr7\n+jh76hSdW7Zw9uRJunfuZOb4ccpdXdQnJ+m74QZefvppsmqVyrp15+4/OzJC24YNzA4Ps+G66xh/\n8kk6Nm/m7MmT5PU6pUqF+tQUHZs2MfvCC2y49lraNmwo8BXrYlI+epTSiy+S9/VROnWKfMsWSidP\nku3cSfn4cfKuLkqTk9RvuIHy009TqlbJ56zP0sgIpQ0bKA8PU7/uOipPPkm2eTPlkyehXodKBaam\nyDdtovzCC9SuvRZcn2uOew4LaN+4kfV79lCfmKBn927W7djBqQceIK/XG39qNWpjY8ycOEHfvn10\nX375oh63Y2CA3r17qY2N0bNrF51DQ5w9dWqZX41Wm3zjRrI9eyhNTJDt3k22YwftDzwA9Tqleh1q\nNUpjY5RPnKC+bx/ZItdnNjBAfe9eSmNjZLt2kQ0NUXZ9rknLvucQQrga+HyM8dYQwtdjjO9b7udc\nCqVKpfF3W+M/UalUomfPHrqGhugcHKQ+NUW5qwvyHIA8y857gNK5j1Gvv3p1uQyl0rnHb9w5X74X\notXplfXTXJ+UStT37CEbGiIfHISpKZizPlnk+qS5PnF9rnkruudwsYRhIdnMDJNHjjD++OOMP/EE\n5DmdQ0OMPfooU0ePNr7wN7X39VEdH2fy8GFq4+MFTq01Y2aG8pEjVB5/nEpzfeZDQ1QefZTy0aON\nL/xNeV8fpfFxyocPU3J9ah6vuecQQvgQcAuwGdgL/CnwAeAq4IPAjcCvAhlwX4zxUyGEbcBXgFng\niTmPdTLGuDmEcAC4K8Z4KIRwV/OxDwCfBGrA9cCfAe8G9gF/EGO8bwle76LMPUS0+Y47AOgcHKRz\ncDC57YZrrjm33XvVVQBc8ra3AbDpllsA6Nm1C4CurVtffdx3vrPxsZ07l3J0rQFzDxHVmuszHxyk\nPs/6rM9Zn1lzfdab67PWXJ9Zc33W56zPWnN9Zq7PNWuxh5V2AW8Hfgv4YxpfsD8E/AmwAXhb83YP\nhhC+AtwF/HOM8bMhhD8Erl3k81wHXAm8A/gnYAdwE/AJYMXiIElr3WIPK/1vjDEHhoEnY4x1YAS4\nhkY4vtv80wtsp7FX8VDzvgd+jnmeiDHONp/nmRjjZPN5+n6OxxAwMNBb9AiSLmKL3XOoLbB9CY09\nhI/NvXFzb+GVM2DzBWjuGa72RTxPaZFzFiar1Xj5hz8seoxzRkcnih5BrWRmhsrBgwBUHnmE+pve\n1DjxLC3gQt+t9ChwWwihG5gGPgP8ERBpnIt4FLhtnvuNA0PAIeCtzb8vbllGz+7dRU8hzavyzDNk\n27c3LnR1URoeJp9zjkE634W+W+k5GkH4PvDfwAsxxmngs8BHQgj3A/3z3O/vgC+EEL4NnLjAGVpC\nuaODSldX0WNI85uehu5uAPLubkrT0wUPpFb3mnsOMcYvzdn+FvCt87eBe8+7z4+Bt8zzWJubf38b\n+PY8T3eg+fFDwK3nb0t6ffL16xs/+wCUJifJhoYKnkitzp+QXiLV06cZax7THTt4kOrYWMETSa/K\nrriC8rPPNi5Uq+RbthQ6j1qfv1tpibT399PX3ziC1nf99QVPI52nq4v6zTcDUL/xxoKH0cXAPQdJ\nUsI4SJISxkGSlDAOkqSEcZAkJYyDJClhHCRJCeMgSUoYB0lSwjhIkhLGQZKUMA6SpIRxkCQljIMk\nKWEcJEkJ4yBJShgHSVLCOEiSEsZBkpQwDpKkhHGQJCWMgyQpYRwkSQnjIElKGAdJUsI4SJISxkGS\nlDAOkqSEcZAkJYyDJClhHCRJCeMgSUoYB0lSwjhIkhLGQZKUMA6SpIRxkCQljIMkKWEcJEmJUp7n\nRc+wJEZHJ1rihQwM9DI6OlH0GNK8XJ+aa2Cgt7TQx1ZNHD58z20t8UL+4e7v8uF7bit6jEL9xce/\nwUfvvaboMTSPr919jPffs6PoMdQivnb3sQXj4GElSVLCOEiSEsZBkpQwDpKkhHGQJCWMgyQpYRwk\nSQnjIElKGAdJUsI4SJISxkGSlDAOkqSEcZAkJYyDJClhHCRJCeMgSUoYB0lSwjhIkhLGQZKUMA6S\npIRxkCQljIMkKWEcJEkJ4yBJShgHSVLCOEiSEsZBkpQwDpKkhHGQJCWMgyQpYRwkSQnjIElKGAdJ\nUsI4SJISxkGSlDAOkqSEcZAkJYyDJClhHCRJCeMgSUoYB0lSwjhIkhLGQZKUMA6SpIRxkCQljEMB\nXjw6xQuHJ4seQ5rXmaNlXjrsl4a1zhVQkDyDwz84zfFDExx5+AyzkzV+9PAZAH544BSnjk9TO5tx\n+KHTBU+qNSmD539QYfRQmRMPV6hOwomHKwA8d6CN8eMl6mfh+YcqBQ+q5WIcCnL80ASX7uzhDVf3\n0tVbYfJMjerZjNrZjHW9bYy/eJYzw7P0X9ZV9Khag0YPVejfmTFwdUZHb87MmRL1s40gdPTmTL1Y\n5uXhMusvy4seVctkxeMQQvhQCOEvV/p5W1Gp9NOXe/rbGY6TXLKti6yeMzYyS//WzmKGk85bn139\nGS/FMr3bMvI6TI2UWL81K2Y2LTv3HAryhjf2MnJkip8cmmB2os7GoU4u2dbFyJFJegc6WLehjXo1\np9Lu/yKtvIE31jlzpMzooTJnJ0qsH8rp3ZZz+kiF7oGcjg059SpU2oueVMulragnDiF8EviV5sX7\ngH8BPhdjfE8I4WbgX4FLaATs8Rjj1cVMuvS2XN497/U9G9u58f2DAGy9cv1KjiSds/Hy+fcGujbm\n7H5/FYBNV7rHsNoVFYcdwO3Am5qXHwG+CmwLIZSAtwKPAXuBzubHdZEYGOgtegRJF6ioOOwD7o8x\n1gBCCA8C1wL/B+wG3gzcC+wH1gEHihlTr8fo6ETRI0i6QEUd0M756dNdHUBGIwI3Ad3Ad5vbb21u\nr1r1WsZzT45z8JsjRY8iJWozjbesjjxWYfiRCrlvUFoTiorDY8D+EEJbCKENeEvzuu8BvwYciTGe\nBAaAgRjj8YLmXBF5BkO7e2jr8OSzWs9Lz1TYsD3j0n11Kl0wOVx67TvpolfUYaVnaewNfI9GoL4Y\nY/wxQAjhKuCLzdudBl4oYsCVZBTUymrT0N58D0V7d05tukRj51+r2YrHIcb4pTkXvzDPx7fO2f71\nlZhJ0sI61udUp6BrI1QnS/QM+U6ltcBvWVvA5Okqzx4cY3aqzrMHx5gaqxY9knTOxisyxp8tM/JY\nhawKPVvca1gLCvs5B72qp7+dnv4+tl/fV/QoUqKtCy67uV70GFph7jlIkhLGQZKUMA6SpIRxkCQl\njIMkKWEcJEkJ4yBJShgHSVLCOEiSEsZBkpQwDpKkhHGQJCWMgyQpYRwkSQnjIElKGAdJUsI4SJIS\nxkGSlDAOkqSEcZAkJYyDJClhHCRJCeMgSUoYB0lSwjhIkhLGQZKUMA6SpIRxkCQljIMkKWEcJEkJ\n4yBJShgHSVLCOEiSEsZBkpQwDpKkhHGQJCWMgyQpYRwkSQnjIElKGAdJUsI4SJISxkGSlDAOkqSE\ncZAkJYyDJClRyvO86BmWxOjoREu8kIGBXkZHJ4oeQ5qX61NzDQz0lhb62KqJwz233bYsL+TjX/4G\n9775mkXf/u5jx7hnx47lGEW6YK5PzXX3sWMLxsHDSpKkxKqMw9GpKQ5PTp67/NTEBNkCe0j3j46u\n1FgSAEfLZQ6XX/3Ue6pcJlvgtve3ta3MUNJ5Vu3KG5mdZbpe51S1SjXLuHL9ep6amOBslrGuUuH4\nzAx3bN7MTJbx1MQEJ6tVQk8Pg52dRY+uNWCkVGK6UuFUqUQVuDLLeKpc5mypxLo853i5zB21GjOl\nEk+Vy5wslwn1OoOr5DCwWt+q3HMA2NTRwTUbNtBZLjNZrwPwk5kZ9vX1sX3dunN7Em2lEnt7ewk9\nPYzMzhY5staQTXnONfU6nXnOZKlx2Pcn5TL76nW2Z9m5PYm2PGdvlhHqdUbKq/bTVS1o1a62tuYn\nXAn4Wd9rLfZ20lJ6ZZf9NdfnIm8nLbVlj0MI4dYQwlfPu+4zIYQF3zIRQng2hLB+qWcZ7OzksbEx\njk1PL/VDSxdsMM95rFLhmHsIagGFnHOIMf7Ocj7+5d3d57b39/ef297W1cUl7e1U85zhmRkA3jUw\nAMClnZ1c6vkGrYDLs1dPP++v16F52HNblnFJnlMFhpuBeFetBsClec6lzdtJK2Gl4rA+hPCPwLXA\nV4BfAO4CzjQvnwW+D7w9xnhr8z53hRDe25zxXTHGC/7JnTPVKocnJ8mBsH7Jd0ykC3KmVOJwudxY\nn4ZABVup/dergI8C+4FPzLn+d4EvxxhvAc7/tv1QjPEdwI+B25diiCt6eripv5/9/f1c1tW1qPsM\nDPQuxVNLr+mKLOOmep399TqX+a4kFWyl4nAwxjgVY3yZxrm1V+wBHmxuf+O8+/xX8+/ngb6lGKKW\nZTw5Ps43R0YWfR9/1YBWygzwUKXCY5UKj1QqnoBWoVYqDrUFri/BuXftnf+5UDvvdhcsA3b39NDh\nCT+1oGcqFbZnGfvqdbqA4dKSLHvpdSn6q+SPgBub2+9Z7ifrKJfpqlSW+2mk12UaeOWtFN15zrRx\nUIGKjsNngY+FEB6gsXfgWTitWevznKnm9mSpRK/nHVSgZX+3UozxAHBgzuXNr2yHEPYCd8UYHwwh\nfAAYaN5m+5zb//5SzXK6WuXY1BRT9ToHx8a4orubvvb2pXp46YJckWUcrFQYofFd0hbjoAIV/buV\nJoC/DSHkNE4JfHg5n6y/vZ3+vj6u71uS89vSkuoCbvYtrGoRhcYhxvgc8LYiZ5AkpYo+5yBJakGr\n5l+CkyQtHfccJEkJ4yBJShgHSVLCOEiSEsZBkpQwDpKkxP8D3b2iNS7b7Z0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf269d1810>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['font.size'] = 10.0\n",
    "# http://www.statsmodels.org/0.6.1/generated/statsmodels.graphics.mosaicplot.mosaic.html\n",
    "    \n",
    "for x in XNonNumerics:\n",
    "    m = mosaic(df[[x, YName]],\n",
    "           [x, YName],\n",
    "           horizontal=False\n",
    "          )\n",
    "    pylab.xlabel(x, fontsize =5)\n",
    "    pylab.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using H2O to get a first impression of variable importance\n",
    "\n",
    "While the assignment prefers, a more \"fundamental\" approach, it has been my experience that running an initial H2O GBM (Gradient Boosted Machine) model has a number of advantages which are useful for gaining a first impression of the data.\n",
    "\n",
    "H2O GBM\n",
    "\n",
    "* has no need to scale the data\n",
    "* can handle both categorical and continous features\n",
    "* is robust against overfitting\n",
    "* general builds a very good model\n",
    "* is fast\n",
    "* can show you variable importance in the model it builds.\n",
    "\n",
    "However, the \"No Free Lunch Theorem\", does caution us from depending blindly on any one model.  https://en.wikipedia.org/wiki/No_free_lunch_in_search_and_optimization\n",
    "\n",
    "George Box also says, \"All models are wrong, but some are useful.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Checking whether there is an H2O instance running at http://localhost:54321..... not found.\n",
      "Attempting to start a local H2O server...\n",
      "  Java Version: openjdk version \"1.8.0_121\"; OpenJDK Runtime Environment (Zulu 8.20.0.5-linux64) (build 1.8.0_121-b15); OpenJDK 64-Bit Server VM (Zulu 8.20.0.5-linux64) (build 25.121-b15, mixed mode)\n",
      "  Starting server from /usr/local/anaconda2/lib/python2.7/site-packages/h2o/backend/bin/h2o.jar\n",
      "  Ice root: /tmp/tmpd1PEp9\n",
      "  JVM stdout: /tmp/tmpd1PEp9/h2o_clem_started_from_python.out\n",
      "  JVM stderr: /tmp/tmpd1PEp9/h2o_clem_started_from_python.err\n",
      "  Server is running at http://127.0.0.1:54321\n",
      "Connecting to H2O server at http://127.0.0.1:54321... successful.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td>H2O cluster uptime:</td>\n",
       "<td>01 secs</td></tr>\n",
       "<tr><td>H2O cluster version:</td>\n",
       "<td>3.10.4.8</td></tr>\n",
       "<tr><td>H2O cluster version age:</td>\n",
       "<td>8 days </td></tr>\n",
       "<tr><td>H2O cluster name:</td>\n",
       "<td>H2O_from_python_clem_1ln428</td></tr>\n",
       "<tr><td>H2O cluster total nodes:</td>\n",
       "<td>1</td></tr>\n",
       "<tr><td>H2O cluster free memory:</td>\n",
       "<td>6.973 Gb</td></tr>\n",
       "<tr><td>H2O cluster total cores:</td>\n",
       "<td>8</td></tr>\n",
       "<tr><td>H2O cluster allowed cores:</td>\n",
       "<td>8</td></tr>\n",
       "<tr><td>H2O cluster status:</td>\n",
       "<td>accepting new members, healthy</td></tr>\n",
       "<tr><td>H2O connection url:</td>\n",
       "<td>http://127.0.0.1:54321</td></tr>\n",
       "<tr><td>H2O connection proxy:</td>\n",
       "<td>None</td></tr>\n",
       "<tr><td>H2O internal security:</td>\n",
       "<td>False</td></tr>\n",
       "<tr><td>Python version:</td>\n",
       "<td>2.7.13 final</td></tr></table></div>"
      ],
      "text/plain": [
       "--------------------------  ------------------------------\n",
       "H2O cluster uptime:         01 secs\n",
       "H2O cluster version:        3.10.4.8\n",
       "H2O cluster version age:    8 days\n",
       "H2O cluster name:           H2O_from_python_clem_1ln428\n",
       "H2O cluster total nodes:    1\n",
       "H2O cluster free memory:    6.973 Gb\n",
       "H2O cluster total cores:    8\n",
       "H2O cluster allowed cores:  8\n",
       "H2O cluster status:         accepting new members, healthy\n",
       "H2O connection url:         http://127.0.0.1:54321\n",
       "H2O connection proxy:\n",
       "H2O internal security:      False\n",
       "Python version:             2.7.13 final\n",
       "--------------------------  ------------------------------"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "PName = \"predict\"   # The name of the predict field for H2O output\n",
    "h2o.init()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |█████████████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "# Make the H2O version of the data frame\n",
    "dfh = h2o.H2OFrame(df)\n",
    "dfh[YName] = dfh[YName].asfactor()   # Force response"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfh[YName].anyfactor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# http://docs.h2o.ai/h2o/latest-stable/h2o-docs/data-science/gbm.html\n",
    "\n",
    "# For doing a grid search, see here:\n",
    "# https://github.com/h2oai/h2o-3/blob/master/h2o-docs/src/product/tutorials/gbm/gbmTuning.ipynb\n",
    "    \n",
    "    \n",
    "gbm = h2o.estimators.gbm.H2OGradientBoostingEstimator(\n",
    "   # nfolds=4,\n",
    "   #     ntrees=5,\n",
    "   #     max_depth=5,    # Of a tree\n",
    ")\n",
    "x = \"\"\"\n",
    "    # DEFAULTS:\n",
    "    nfolds = 0,    # No cross validation folds\n",
    "    ntrees=50,\n",
    "    learn_rate=0.1,\n",
    "    max_depth=5,    # Of a tree\n",
    "    min_row=10,     # Fewest allowed observations in a leaf (weighted)\n",
    "    nbins=20,   # For numeric cols, build histogram of at most this many bins\n",
    "    stopping_tolerance=0.01, #10-fold increase in threshold as defined in rf_v1\n",
    "    stopping_rounds=2,\n",
    "    score_each_iteration=True,\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gbm Model Build progress: |███████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "gbm.train(x=XNames, y=YName, training_frame=dfh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gbm prediction progress: |████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "pred_train = gbm.predict(dfh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training error = 0.019\n",
      "Training accuracy = 0.981\n"
     ]
    }
   ],
   "source": [
    "print (\"Training error = {:.3f}\".format(\n",
    "    1 - (pred_train['predict'] == dfh[YName]).sum()/dfh.nrows))\n",
    "\n",
    "print (\"Training accuracy = {:.3f}\".format(\n",
    "    (pred_train['predict'] == dfh[YName]).sum()/dfh.nrows))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead>\n",
       "<tr><th style=\"text-align: right;\">  predict</th><th style=\"text-align: right;\">       p0</th><th style=\"text-align: right;\">      p1</th></tr>\n",
       "</thead>\n",
       "<tbody>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.0231102</td><td style=\"text-align: right;\">0.97689 </td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.180919 </td><td style=\"text-align: right;\">0.819081</td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.0118534</td><td style=\"text-align: right;\">0.988147</td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.049089 </td><td style=\"text-align: right;\">0.950911</td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.0231102</td><td style=\"text-align: right;\">0.97689 </td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.0231102</td><td style=\"text-align: right;\">0.97689 </td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.013452 </td><td style=\"text-align: right;\">0.986548</td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.048097 </td><td style=\"text-align: right;\">0.951903</td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.0336916</td><td style=\"text-align: right;\">0.966308</td></tr>\n",
       "<tr><td style=\"text-align: right;\">        1</td><td style=\"text-align: right;\">0.0231102</td><td style=\"text-align: right;\">0.97689 </td></tr>\n",
       "</tbody>\n",
       "</table>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "category"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[YName].dtype"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### F1 Score\n",
    "\n",
    "\n",
    "$$F1 = 2 * \\frac{precision * recall}{precision + recall}$$\n",
    "\n",
    "See:\n",
    "\n",
    "https://en.wikipedia.org/wiki/F1_score\n",
    "\n",
    "http://scikit-learn.org/stable/modules/generated/sklearn.metrics.f1_score.html\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score for y = '1' is 0.959\n",
      "F1 Score for y = '0' is 0.988\n"
     ]
    }
   ],
   "source": [
    "# Multinominal classifier\n",
    "f1 = f1_score(dfh[YName].as_data_frame()[YName].tolist(),\n",
    "     pred_train[PName].as_data_frame()[PName].tolist(),\n",
    "     average=None,         \n",
    "     labels=YValues\n",
    "     )    \n",
    "for i, yvalue in enumerate(YValues):\n",
    "    print(\"F1 Score for y = '{}' is {:.3f}\".format(yvalue, f1[i]))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ROC\n",
    "\n",
    "http://scikit-learn.org/stable/modules/generated/sklearn.metrics.roc_curve.html\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAD4CAYAAADxeG0DAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBbmIiOcU5CIinlOQi4h47v8DCFuzrCEyPzgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf26ce2250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-- AUC is 0.993\n"
     ]
    }
   ],
   "source": [
    "for i, yvalue in enumerate(YValues):\n",
    "    # Get the True and Predicted values\n",
    "    y_true = list(df[YName])\n",
    "    y_true = map(lambda x : 1 if x == yvalue else 0, y_true)\n",
    "\n",
    "    # These are the predicted probabilities\n",
    "    if type(yvalue) is str:\n",
    "        yvalue_column_name = yvalue\n",
    "        y_pred = pred_train[yvalue].as_data_frame()[yvalue].tolist()\n",
    "    else:\n",
    "        yvalue_column_name = \"p\" + str(yvalue)\n",
    "        y_pred = (pred_train.as_data_frame()[yvalue_column_name].tolist())\n",
    "\n",
    "    # Area Under the ROC Curve (AUC)\n",
    "    auc = roc_auc_score(y_true, y_pred)\n",
    "    FalsePositiveRate, TruePositiveRate, thresholds = roc_curve(\n",
    "        y_true,\n",
    "        y_pred)\n",
    "    pylab.plot(FalsePositiveRate,\n",
    "                           TruePositiveRate,\n",
    "                           label=yvalue)\n",
    "pylab.legend(loc='lower right', shadow=True)\n",
    "pylab.show()\n",
    "\n",
    "print(\"-- AUC is {:.3f}\".format(auc))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4gAAAJTCAYAAABKPjiWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLP8XBYBeUcFGkfIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdf26f70a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "gbm.varimp_plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>variable</th>\n",
       "      <th>relative_importance</th>\n",
       "      <th>scaled_importance</th>\n",
       "      <th>percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>satisfaction_level</td>\n",
       "      <td>6767.692383</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.526354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>time_spend_company</td>\n",
       "      <td>2118.022949</td>\n",
       "      <td>0.312961</td>\n",
       "      <td>0.164728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>last_evaluation</td>\n",
       "      <td>1721.129028</td>\n",
       "      <td>0.254315</td>\n",
       "      <td>0.133860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>number_project</td>\n",
       "      <td>1409.998047</td>\n",
       "      <td>0.208343</td>\n",
       "      <td>0.109662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>average_montly_hours</td>\n",
       "      <td>759.435059</td>\n",
       "      <td>0.112215</td>\n",
       "      <td>0.059065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>sales</td>\n",
       "      <td>52.549866</td>\n",
       "      <td>0.007765</td>\n",
       "      <td>0.004087</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>salary</td>\n",
       "      <td>18.600296</td>\n",
       "      <td>0.002748</td>\n",
       "      <td>0.001447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Work_accident</td>\n",
       "      <td>10.255219</td>\n",
       "      <td>0.001515</td>\n",
       "      <td>0.000798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>promotion_last_5years</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                variable  relative_importance  scaled_importance  percentage\n",
       "0     satisfaction_level          6767.692383           1.000000    0.526354\n",
       "1     time_spend_company          2118.022949           0.312961    0.164728\n",
       "2        last_evaluation          1721.129028           0.254315    0.133860\n",
       "3         number_project          1409.998047           0.208343    0.109662\n",
       "4   average_montly_hours           759.435059           0.112215    0.059065\n",
       "5                  sales            52.549866           0.007765    0.004087\n",
       "6                 salary            18.600296           0.002748    0.001447\n",
       "7          Work_accident            10.255219           0.001515    0.000798\n",
       "8  promotion_last_5years             0.000000           0.000000    0.000000"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "varimp = pd.DataFrame(gbm.varimp(),\n",
    "             columns=\n",
    "             [\"variable\",\n",
    "              \"relative_importance\",\n",
    "              \"scaled_importance\",\n",
    "              \"percentage\"]\n",
    "            )\n",
    "top5 = varimp['variable'][0:5].tolist() + [YName]\n",
    "varimp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# TBD - figure out a pairs plot for top 5 most important variables\n",
    "#  -- tricky - how to do it with both categorical and numeric values\n",
    "\n",
    "# from pandas.tools.plotting import scatter_matrix\n",
    "# scatter_matrix(df[top5],\n",
    "\n",
    "#### df[top5].plot(subplots=True, layout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "if False:\n",
    "    h2o.cluster().shutdown(prompt = False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}