Machine Learning PySpark and MLlib.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "root\n",
      " |-- age: integer (nullable = true)\n",
      " |-- job: string (nullable = true)\n",
      " |-- marital: string (nullable = true)\n",
      " |-- education: string (nullable = true)\n",
      " |-- default: string (nullable = true)\n",
      " |-- balance: integer (nullable = true)\n",
      " |-- housing: string (nullable = true)\n",
      " |-- loan: string (nullable = true)\n",
      " |-- contact: string (nullable = true)\n",
      " |-- day: integer (nullable = true)\n",
      " |-- month: string (nullable = true)\n",
      " |-- duration: integer (nullable = true)\n",
      " |-- campaign: integer (nullable = true)\n",
      " |-- pdays: integer (nullable = true)\n",
      " |-- previous: integer (nullable = true)\n",
      " |-- poutcome: string (nullable = true)\n",
      " |-- deposit: string (nullable = true)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from pyspark.sql import SparkSession\n",
    "spark = SparkSession.builder.appName('ml-bank').getOrCreate()\n",
    "df = spark.read.csv('bank.csv', header = True, inferSchema = True)\n",
    "df.printSchema()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pandas dataframe is prettier than Spark DataFrame.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>59</td>\n",
       "      <td>56</td>\n",
       "      <td>41</td>\n",
       "      <td>55</td>\n",
       "      <td>54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>job</th>\n",
       "      <td>admin.</td>\n",
       "      <td>admin.</td>\n",
       "      <td>technician</td>\n",
       "      <td>services</td>\n",
       "      <td>admin.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>marital</th>\n",
       "      <td>married</td>\n",
       "      <td>married</td>\n",
       "      <td>married</td>\n",
       "      <td>married</td>\n",
       "      <td>married</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>education</th>\n",
       "      <td>secondary</td>\n",
       "      <td>secondary</td>\n",
       "      <td>secondary</td>\n",
       "      <td>secondary</td>\n",
       "      <td>tertiary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>default</th>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>balance</th>\n",
       "      <td>2343</td>\n",
       "      <td>45</td>\n",
       "      <td>1270</td>\n",
       "      <td>2476</td>\n",
       "      <td>184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>housing</th>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan</th>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>contact</th>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month</th>\n",
       "      <td>may</td>\n",
       "      <td>may</td>\n",
       "      <td>may</td>\n",
       "      <td>may</td>\n",
       "      <td>may</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>duration</th>\n",
       "      <td>1042</td>\n",
       "      <td>1467</td>\n",
       "      <td>1389</td>\n",
       "      <td>579</td>\n",
       "      <td>673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>campaign</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pdays</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>previous</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>poutcome</th>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>deposit</th>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0          1           2          3         4\n",
       "age               59         56          41         55        54\n",
       "job           admin.     admin.  technician   services    admin.\n",
       "marital      married    married     married    married   married\n",
       "education  secondary  secondary   secondary  secondary  tertiary\n",
       "default           no         no          no         no        no\n",
       "balance         2343         45        1270       2476       184\n",
       "housing          yes         no         yes        yes        no\n",
       "loan              no         no          no         no        no\n",
       "contact      unknown    unknown     unknown    unknown   unknown\n",
       "day                5          5           5          5         5\n",
       "month            may        may         may        may       may\n",
       "duration        1042       1467        1389        579       673\n",
       "campaign           1          1           1          1         2\n",
       "pdays             -1         -1          -1         -1        -1\n",
       "previous           0          0           0          0         0\n",
       "poutcome     unknown    unknown     unknown    unknown   unknown\n",
       "deposit          yes        yes         yes        yes       yes"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "pd.DataFrame(df.take(5), columns=df.columns).transpose()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Our Classes are perfect balanced"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>deposit</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>no</td>\n",
       "      <td>5873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>yes</td>\n",
       "      <td>5289</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  deposit  count\n",
       "0      no   5873\n",
       "1     yes   5289"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby('deposit').count().toPandas()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary statistics for numeric variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>summary</th>\n",
       "      <td>count</td>\n",
       "      <td>mean</td>\n",
       "      <td>stddev</td>\n",
       "      <td>min</td>\n",
       "      <td>max</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>11162</td>\n",
       "      <td>41.231947679627304</td>\n",
       "      <td>11.913369192215518</td>\n",
       "      <td>18</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>balance</th>\n",
       "      <td>11162</td>\n",
       "      <td>1528.5385235620856</td>\n",
       "      <td>3225.413325946149</td>\n",
       "      <td>-6847</td>\n",
       "      <td>81204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <td>11162</td>\n",
       "      <td>15.658036194230425</td>\n",
       "      <td>8.420739541006462</td>\n",
       "      <td>1</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>duration</th>\n",
       "      <td>11162</td>\n",
       "      <td>371.99381831213043</td>\n",
       "      <td>347.12838571630687</td>\n",
       "      <td>2</td>\n",
       "      <td>3881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>campaign</th>\n",
       "      <td>11162</td>\n",
       "      <td>2.508421429851281</td>\n",
       "      <td>2.7220771816614824</td>\n",
       "      <td>1</td>\n",
       "      <td>63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pdays</th>\n",
       "      <td>11162</td>\n",
       "      <td>51.33040673714388</td>\n",
       "      <td>108.75828197197717</td>\n",
       "      <td>-1</td>\n",
       "      <td>854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>previous</th>\n",
       "      <td>11162</td>\n",
       "      <td>0.8325568894463358</td>\n",
       "      <td>2.292007218670508</td>\n",
       "      <td>0</td>\n",
       "      <td>58</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              0                   1                   2      3      4\n",
       "summary   count                mean              stddev    min    max\n",
       "age       11162  41.231947679627304  11.913369192215518     18     95\n",
       "balance   11162  1528.5385235620856   3225.413325946149  -6847  81204\n",
       "day       11162  15.658036194230425   8.420739541006462      1     31\n",
       "duration  11162  371.99381831213043  347.12838571630687      2   3881\n",
       "campaign  11162   2.508421429851281  2.7220771816614824      1     63\n",
       "pdays     11162   51.33040673714388  108.75828197197717     -1    854\n",
       "previous  11162  0.8325568894463358   2.292007218670508      0     58"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numeric_features = [t[0] for t in df.dtypes if t[1] == 'int']\n",
    "df.select(numeric_features).describe().toPandas().transpose()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Correlations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/python3/lib/python3.6/site-packages/ipykernel/__main__.py:3: FutureWarning: pandas.scatter_matrix is deprecated. Use pandas.plotting.scatter_matrix instead\n",
      "  app.launch_new_instance()\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe345cdbf28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "numeric_data = df.select(numeric_features).toPandas()\n",
    "\n",
    "axs = pd.scatter_matrix(numeric_data, figsize=(8, 8));\n",
    "\n",
    "# Rotate axis labels and remove axis ticks\n",
    "n = len(numeric_data.columns)\n",
    "for i in range(n):\n",
    "    v = axs[i, 0]\n",
    "    v.yaxis.label.set_rotation(0)\n",
    "    v.yaxis.label.set_ha('right')\n",
    "    v.set_yticks(())\n",
    "    h = axs[n-1, i]\n",
    "    h.xaxis.label.set_rotation(90)\n",
    "    h.set_xticks(())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It's obvious that there aren't highly correlated independent variables. Therefore, we will keep all of them for the model. However, day and month columns are not really useful, we will remove these two columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "root\n",
      " |-- age: integer (nullable = true)\n",
      " |-- job: string (nullable = true)\n",
      " |-- marital: string (nullable = true)\n",
      " |-- education: string (nullable = true)\n",
      " |-- default: string (nullable = true)\n",
      " |-- balance: integer (nullable = true)\n",
      " |-- housing: string (nullable = true)\n",
      " |-- loan: string (nullable = true)\n",
      " |-- contact: string (nullable = true)\n",
      " |-- duration: integer (nullable = true)\n",
      " |-- campaign: integer (nullable = true)\n",
      " |-- pdays: integer (nullable = true)\n",
      " |-- previous: integer (nullable = true)\n",
      " |-- poutcome: string (nullable = true)\n",
      " |-- deposit: string (nullable = true)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df = df.select('age', 'job', 'marital', 'education', 'default', 'balance', 'housing', 'loan', 'contact', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'deposit')\n",
    "cols = df.columns\n",
    "df.printSchema()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preparing Data for Machine Learning\n",
    "\n",
    "Category Indexing, One-Hot Encoding and VectorAssembler - a feature transformer that merges multiple columns into a vector column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pyspark.ml.feature import OneHotEncoderEstimator, StringIndexer, VectorAssembler\n",
    "\n",
    "categoricalColumns = ['job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'poutcome']\n",
    "stages = []\n",
    "\n",
    "for categoricalCol in categoricalColumns:\n",
    "    stringIndexer = StringIndexer(inputCol = categoricalCol, outputCol = categoricalCol + 'Index')\n",
    "    encoder = OneHotEncoderEstimator(inputCols=[stringIndexer.getOutputCol()], outputCols=[categoricalCol + \"classVec\"])\n",
    "    stages += [stringIndexer, encoder]\n",
    "\n",
    "label_stringIdx = StringIndexer(inputCol = 'deposit', outputCol = 'label')\n",
    "stages += [label_stringIdx]\n",
    "\n",
    "numericCols = ['age', 'balance', 'duration', 'campaign', 'pdays', 'previous']\n",
    "assemblerInputs = [c + \"classVec\" for c in categoricalColumns] + numericCols\n",
    "assembler = VectorAssembler(inputCols=assemblerInputs, outputCol=\"features\")\n",
    "stages += [assembler]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The above code are taken from databricks' official site and it indexes each categorical column using the StringIndexer, then converts the indexed categories into one-hot encoded variables. The resulting output has the binary vectors appended to the end of each row. We use the StringIndexer again to encode our labels to label indices.\n",
    "\n",
    "Next, we use the VectorAssembler to combine all the feature columns into a single vector column."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pipeline\n",
    "\n",
    "We use Pipeline to chain multiple Transformers and Estimators together to specify our machine learning workflow. A Pipeline’s stages are specified as an ordered array."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "root\n",
      " |-- label: double (nullable = false)\n",
      " |-- features: vector (nullable = true)\n",
      " |-- age: integer (nullable = true)\n",
      " |-- job: string (nullable = true)\n",
      " |-- marital: string (nullable = true)\n",
      " |-- education: string (nullable = true)\n",
      " |-- default: string (nullable = true)\n",
      " |-- balance: integer (nullable = true)\n",
      " |-- housing: string (nullable = true)\n",
      " |-- loan: string (nullable = true)\n",
      " |-- contact: string (nullable = true)\n",
      " |-- duration: integer (nullable = true)\n",
      " |-- campaign: integer (nullable = true)\n",
      " |-- pdays: integer (nullable = true)\n",
      " |-- previous: integer (nullable = true)\n",
      " |-- poutcome: string (nullable = true)\n",
      " |-- deposit: string (nullable = true)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from pyspark.ml import Pipeline\n",
    "pipeline = Pipeline(stages = stages)\n",
    "pipelineModel = pipeline.fit(df)\n",
    "df = pipelineModel.transform(df)\n",
    "selectedCols = ['label', 'features'] + cols\n",
    "df = df.select(selectedCols)\n",
    "df.printSchema()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>label</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>features</th>\n",
       "      <td>(0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>(0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>(0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>(0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>(0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>59</td>\n",
       "      <td>56</td>\n",
       "      <td>41</td>\n",
       "      <td>55</td>\n",
       "      <td>54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>job</th>\n",
       "      <td>admin.</td>\n",
       "      <td>admin.</td>\n",
       "      <td>technician</td>\n",
       "      <td>services</td>\n",
       "      <td>admin.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>marital</th>\n",
       "      <td>married</td>\n",
       "      <td>married</td>\n",
       "      <td>married</td>\n",
       "      <td>married</td>\n",
       "      <td>married</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>education</th>\n",
       "      <td>secondary</td>\n",
       "      <td>secondary</td>\n",
       "      <td>secondary</td>\n",
       "      <td>secondary</td>\n",
       "      <td>tertiary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>default</th>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>balance</th>\n",
       "      <td>2343</td>\n",
       "      <td>45</td>\n",
       "      <td>1270</td>\n",
       "      <td>2476</td>\n",
       "      <td>184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>housing</th>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan</th>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>contact</th>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>duration</th>\n",
       "      <td>1042</td>\n",
       "      <td>1467</td>\n",
       "      <td>1389</td>\n",
       "      <td>579</td>\n",
       "      <td>673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>campaign</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pdays</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>previous</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>poutcome</th>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>deposit</th>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                           0  \\\n",
       "label                                                      1   \n",
       "features   (0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "age                                                       59   \n",
       "job                                                   admin.   \n",
       "marital                                              married   \n",
       "education                                          secondary   \n",
       "default                                                   no   \n",
       "balance                                                 2343   \n",
       "housing                                                  yes   \n",
       "loan                                                      no   \n",
       "contact                                              unknown   \n",
       "duration                                                1042   \n",
       "campaign                                                   1   \n",
       "pdays                                                     -1   \n",
       "previous                                                   0   \n",
       "poutcome                                             unknown   \n",
       "deposit                                                  yes   \n",
       "\n",
       "                                                           1  \\\n",
       "label                                                      1   \n",
       "features   (0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "age                                                       56   \n",
       "job                                                   admin.   \n",
       "marital                                              married   \n",
       "education                                          secondary   \n",
       "default                                                   no   \n",
       "balance                                                   45   \n",
       "housing                                                   no   \n",
       "loan                                                      no   \n",
       "contact                                              unknown   \n",
       "duration                                                1467   \n",
       "campaign                                                   1   \n",
       "pdays                                                     -1   \n",
       "previous                                                   0   \n",
       "poutcome                                             unknown   \n",
       "deposit                                                  yes   \n",
       "\n",
       "                                                           2  \\\n",
       "label                                                      1   \n",
       "features   (0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "age                                                       41   \n",
       "job                                               technician   \n",
       "marital                                              married   \n",
       "education                                          secondary   \n",
       "default                                                   no   \n",
       "balance                                                 1270   \n",
       "housing                                                  yes   \n",
       "loan                                                      no   \n",
       "contact                                              unknown   \n",
       "duration                                                1389   \n",
       "campaign                                                   1   \n",
       "pdays                                                     -1   \n",
       "previous                                                   0   \n",
       "poutcome                                             unknown   \n",
       "deposit                                                  yes   \n",
       "\n",
       "                                                           3  \\\n",
       "label                                                      1   \n",
       "features   (0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "age                                                       55   \n",
       "job                                                 services   \n",
       "marital                                              married   \n",
       "education                                          secondary   \n",
       "default                                                   no   \n",
       "balance                                                 2476   \n",
       "housing                                                  yes   \n",
       "loan                                                      no   \n",
       "contact                                              unknown   \n",
       "duration                                                 579   \n",
       "campaign                                                   1   \n",
       "pdays                                                     -1   \n",
       "previous                                                   0   \n",
       "poutcome                                             unknown   \n",
       "deposit                                                  yes   \n",
       "\n",
       "                                                           4  \n",
       "label                                                      1  \n",
       "features   (0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "age                                                       54  \n",
       "job                                                   admin.  \n",
       "marital                                              married  \n",
       "education                                           tertiary  \n",
       "default                                                   no  \n",
       "balance                                                  184  \n",
       "housing                                                   no  \n",
       "loan                                                      no  \n",
       "contact                                              unknown  \n",
       "duration                                                 673  \n",
       "campaign                                                   2  \n",
       "pdays                                                     -1  \n",
       "previous                                                   0  \n",
       "poutcome                                             unknown  \n",
       "deposit                                                  yes  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(df.take(5), columns=df.columns).transpose()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, we now have features column and label column\n",
    "\n",
    "Randomly split data into train and test sets. set seed for reproducibility"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Dataset Count: 7764\n",
      "Test Dataset Count: 3398\n"
     ]
    }
   ],
   "source": [
    "train, test = df.randomSplit([0.7, 0.3], seed = 2018)\n",
    "print(\"Training Dataset Count: \" + str(train.count()))\n",
    "print(\"Test Dataset Count: \" + str(test.count()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic Regression Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pyspark.ml.classification import LogisticRegression\n",
    "\n",
    "lr = LogisticRegression(featuresCol = 'features', labelCol = 'label', maxIter=10)\n",
    "lrModel = lr.fit(train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can obtain the coefficients by using LogisticRegressionModel's attributes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe3400e3b38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "beta = np.sort(lrModel.coefficients)\n",
    "\n",
    "plt.plot(beta)\n",
    "plt.ylabel('Beta Coefficients')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Obtain the receiver-operating characteristic and areaUnderROC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe3400a1cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training set areaUnderROC: 0.8849092421146739\n"
     ]
    }
   ],
   "source": [
    "trainingSummary = lrModel.summary\n",
    "\n",
    "roc = trainingSummary.roc.toPandas()\n",
    "plt.plot(roc['FPR'],roc['TPR'])\n",
    "plt.ylabel('False Positive Rate')\n",
    "plt.xlabel('True Positive Rate')\n",
    "plt.title('ROC Curve')\n",
    "plt.show()\n",
    "\n",
    "print('Training set areaUnderROC: ' + str(trainingSummary.areaUnderROC))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Precision and Recall"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe34006f6a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pr = trainingSummary.pr.toPandas()\n",
    "plt.plot(pr['recall'],pr['precision'])\n",
    "plt.ylabel('Precision')\n",
    "plt.xlabel('Recall')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set the model threshold to maximize F-Measure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe33af04320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f = trainingSummary.fMeasureByThreshold.toPandas()\n",
    "plt.plot(f['threshold'],f['F-Measure'])\n",
    "plt.ylabel('F-Measure')\n",
    "plt.xlabel('Threshold')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make predictions on the test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "|age|       job|label|       rawPrediction|prediction|         probability|\n",
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "| 37|management|  0.0|[1.19871810716723...|       0.0|[0.76829666339830...|\n",
      "| 40|management|  0.0|[2.20534940465796...|       0.0|[0.90072886169926...|\n",
      "| 53|management|  0.0|[1.02590348276690...|       0.0|[0.73612093009497...|\n",
      "| 32|management|  0.0|[1.25795481657702...|       0.0|[0.77867383994058...|\n",
      "| 54|management|  0.0|[1.33232096924268...|       0.0|[0.79122429116078...|\n",
      "| 40|management|  0.0|[1.57095096412779...|       0.0|[0.82791913346617...|\n",
      "| 56|management|  0.0|[3.06095963426752...|       0.0|[0.95525333386804...|\n",
      "| 50|management|  0.0|[-0.8102603273804...|       1.0|[0.30783502428597...|\n",
      "| 47|management|  0.0|[0.67024288891379...|       0.0|[0.66155754396054...|\n",
      "| 44|management|  0.0|[1.29756265761715...|       0.0|[0.78542449653716...|\n",
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "predictions = lrModel.transform(test)\n",
    "predictions.select('age', 'job', 'label', 'rawPrediction', 'prediction', 'probability').show(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Evaluate our Logistic Regression model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Area Under ROC 0.8858324614449619\n"
     ]
    }
   ],
   "source": [
    "from pyspark.ml.evaluation import BinaryClassificationEvaluator\n",
    "\n",
    "evaluator = BinaryClassificationEvaluator()\n",
    "print('Test Area Under ROC', evaluator.evaluate(predictions))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'areaUnderROC'"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluator.getMetricName()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pretty good."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Try tuning the model with the ParamGridBuilder and the CrossValidator."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Area Under ROC 0.8847737068068359\n"
     ]
    }
   ],
   "source": [
    "from pyspark.ml.tuning import ParamGridBuilder, CrossValidator\n",
    "\n",
    "# Create ParamGrid for Cross Validation\n",
    "paramGrid = (ParamGridBuilder()\n",
    "             .addGrid(lr.regParam, [0.01, 0.5, 2.0])\n",
    "             .addGrid(lr.elasticNetParam, [0.0, 0.5, 1.0])\n",
    "             .addGrid(lr.maxIter, [1, 5, 10])\n",
    "             .build())\n",
    "\n",
    "cv = CrossValidator(estimator=lr, estimatorParamMaps=paramGrid, evaluator=evaluator, numFolds=5)\n",
    "\n",
    "cvModel = cv.fit(train)\n",
    "predictions = cvModel.transform(test)\n",
    "print('Test Area Under ROC', evaluator.evaluate(predictions))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'areaUnderROC'"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluator.getMetricName()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision Tree Classifier\n",
    "\n",
    "Decision trees are widely used since they are easy to interpret, handle categorical features, extend to the multiclass classification setting, do not require feature scaling, and are able to capture non-linearities and feature interactions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---+----------+-----+--------------+----------+--------------------+\n",
      "|age|       job|label| rawPrediction|prediction|         probability|\n",
      "+---+----------+-----+--------------+----------+--------------------+\n",
      "| 37|management|  0.0|[1130.0,387.0]|       0.0|[0.74489123269611...|\n",
      "| 40|management|  0.0| [1333.0,86.0]|       0.0|[0.93939393939393...|\n",
      "| 53|management|  0.0|[1130.0,387.0]|       0.0|[0.74489123269611...|\n",
      "| 32|management|  0.0|[1130.0,387.0]|       0.0|[0.74489123269611...|\n",
      "| 54|management|  0.0| [1333.0,86.0]|       0.0|[0.93939393939393...|\n",
      "| 40|management|  0.0|  [373.0,30.0]|       0.0|[0.92555831265508...|\n",
      "| 56|management|  0.0| [1333.0,86.0]|       0.0|[0.93939393939393...|\n",
      "| 50|management|  0.0|[788.0,1230.0]|       1.0|[0.39048562933597...|\n",
      "| 47|management|  0.0|[788.0,1230.0]|       1.0|[0.39048562933597...|\n",
      "| 44|management|  0.0|[1130.0,387.0]|       0.0|[0.74489123269611...|\n",
      "+---+----------+-----+--------------+----------+--------------------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from pyspark.ml.classification import DecisionTreeClassifier\n",
    "\n",
    "dt = DecisionTreeClassifier(featuresCol = 'features', labelCol = 'label', maxDepth = 3)\n",
    "dtModel = dt.fit(train)\n",
    "predictions = dtModel.transform(test)\n",
    "predictions.select('age', 'job', 'label', 'rawPrediction', 'prediction', 'probability').show(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Evaluate our Decision Tree model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Area Under ROC: 0.7807240050065357\n"
     ]
    }
   ],
   "source": [
    "evaluator = BinaryClassificationEvaluator()\n",
    "print(\"Test Area Under ROC: \" + str(evaluator.evaluate(predictions, {evaluator.metricName: \"areaUnderROC\"})))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'areaUnderROC'"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluator.getMetricName()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One simple decision tree performed poorly because it is too weak given the range of different features. The prediction accuracy of decision trees can be improved by Ensemble methods."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "|age|       job|label|       rawPrediction|prediction|         probability|\n",
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "| 37|management|  0.0|[14.0335043427458...|       0.0|[0.70167521713729...|\n",
      "| 40|management|  0.0|[15.5696428848403...|       0.0|[0.77848214424201...|\n",
      "| 53|management|  0.0|[14.0604626680319...|       0.0|[0.70302313340159...|\n",
      "| 32|management|  0.0|[14.9300431640802...|       0.0|[0.74650215820401...|\n",
      "| 54|management|  0.0|[14.4483343905905...|       0.0|[0.72241671952952...|\n",
      "| 40|management|  0.0|[15.0798534256099...|       0.0|[0.75399267128049...|\n",
      "| 56|management|  0.0|[18.2682164556330...|       0.0|[0.91341082278165...|\n",
      "| 50|management|  0.0|[5.99972486043756...|       1.0|[0.29998624302187...|\n",
      "| 47|management|  0.0|[10.8585049161417...|       0.0|[0.54292524580708...|\n",
      "| 44|management|  0.0|[10.6040194546245...|       0.0|[0.53020097273122...|\n",
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from pyspark.ml.classification import RandomForestClassifier\n",
    "\n",
    "rf = RandomForestClassifier(featuresCol = 'features', labelCol = 'label')\n",
    "rfModel = rf.fit(train)\n",
    "predictions = rfModel.transform(test)\n",
    "predictions.select('age', 'job', 'label', 'rawPrediction', 'prediction', 'probability').show(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Area Under ROC: 0.8846453518867426\n"
     ]
    }
   ],
   "source": [
    "evaluator = BinaryClassificationEvaluator()\n",
    "print(\"Test Area Under ROC: \" + str(evaluator.evaluate(predictions, {evaluator.metricName: \"areaUnderROC\"})))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'areaUnderROC'"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluator.getMetricName()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cacheNodeIds: If false, the algorithm will pass trees to executors to match instances with nodes. If true, the algorithm will cache node IDs for each instance. Caching can speed up training of deeper trees. Users can set how often should the cache be checkpointed or disable it by setting checkpointInterval. (default: False)\n",
      "checkpointInterval: set checkpoint interval (>= 1) or disable checkpoint (-1). E.g. 10 means that the cache will get checkpointed every 10 iterations. Note: this setting will be ignored if the checkpoint directory is not set in the SparkContext. (default: 10)\n",
      "featureSubsetStrategy: The number of features to consider for splits at each tree node. Supported options: auto, all, onethird, sqrt, log2, (0.0-1.0], [1-n]. (default: auto)\n",
      "featuresCol: features column name. (default: features, current: features)\n",
      "impurity: Criterion used for information gain calculation (case-insensitive). Supported options: entropy, gini (default: gini)\n",
      "labelCol: label column name. (default: label, current: label)\n",
      "maxBins: Max number of bins for discretizing continuous features.  Must be >=2 and >= number of categories for any categorical feature. (default: 32)\n",
      "maxDepth: Maximum depth of the tree. (>= 0) E.g., depth 0 means 1 leaf node; depth 1 means 1 internal node + 2 leaf nodes. (default: 5)\n",
      "maxMemoryInMB: Maximum memory in MB allocated to histogram aggregation. If too small, then 1 node will be split per iteration, and its aggregates may exceed this size. (default: 256)\n",
      "minInfoGain: Minimum information gain for a split to be considered at a tree node. (default: 0.0)\n",
      "minInstancesPerNode: Minimum number of instances each child must have after split. If a split causes the left or right child to have fewer than minInstancesPerNode, the split will be discarded as invalid. Should be >= 1. (default: 1)\n",
      "numTrees: Number of trees to train (>= 1). (default: 20)\n",
      "predictionCol: prediction column name. (default: prediction)\n",
      "probabilityCol: Column name for predicted class conditional probabilities. Note: Not all models output well-calibrated probability estimates! These probabilities should be treated as confidences, not precise probabilities. (default: probability)\n",
      "rawPredictionCol: raw prediction (a.k.a. confidence) column name. (default: rawPrediction)\n",
      "seed: random seed. (default: -8158597925214029612)\n",
      "subsamplingRate: Fraction of the training data used for learning each decision tree, in range (0, 1]. (default: 1.0)\n"
     ]
    }
   ],
   "source": [
    "print(rf.explainParams())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gradient-boosted Tree Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "|age|       job|label|       rawPrediction|prediction|         probability|\n",
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "| 37|management|  0.0|[0.57808138910181...|       0.0|[0.76063477260811...|\n",
      "| 40|management|  0.0|[1.37467582901950...|       0.0|[0.93987672346171...|\n",
      "| 53|management|  0.0|[-0.0012929624008...|       1.0|[0.49935351915983...|\n",
      "| 32|management|  0.0|[0.61900313605401...|       0.0|[0.77521678642033...|\n",
      "| 54|management|  0.0|[0.98157815641818...|       0.0|[0.87687413211579...|\n",
      "| 40|management|  0.0|[0.96138354833170...|       0.0|[0.87244668327834...|\n",
      "| 56|management|  0.0|[1.39120025731353...|       0.0|[0.94171733839668...|\n",
      "| 50|management|  0.0|[-0.6141629093446...|       1.0|[0.22647458093662...|\n",
      "| 47|management|  0.0|[-0.0439971283470...|       1.0|[0.47801561939801...|\n",
      "| 44|management|  0.0|[0.26452511568224...|       0.0|[0.62926156628314...|\n",
      "+---+----------+-----+--------------------+----------+--------------------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from pyspark.ml.classification import GBTClassifier\n",
    "\n",
    "gbt = GBTClassifier(maxIter=10)\n",
    "gbtModel = gbt.fit(train)\n",
    "predictions = gbtModel.transform(test)\n",
    "predictions.select('age', 'job', 'label', 'rawPrediction', 'prediction', 'probability').show(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Area Under ROC: 0.8940728473145346\n"
     ]
    }
   ],
   "source": [
    "evaluator = BinaryClassificationEvaluator()\n",
    "print(\"Test Area Under ROC: \" + str(evaluator.evaluate(predictions, {evaluator.metricName: \"areaUnderROC\"})))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'areaUnderROC'"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluator.getMetricName()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Gradient-boosted Tree achieved the best results, we will try tuning this model with the ParamGridBuilder and the CrossValidator. Before that we can use explainParams() to print a list of all params and their definitions to understand what params available for tuning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cacheNodeIds: If false, the algorithm will pass trees to executors to match instances with nodes. If true, the algorithm will cache node IDs for each instance. Caching can speed up training of deeper trees. Users can set how often should the cache be checkpointed or disable it by setting checkpointInterval. (default: False)\n",
      "checkpointInterval: set checkpoint interval (>= 1) or disable checkpoint (-1). E.g. 10 means that the cache will get checkpointed every 10 iterations. Note: this setting will be ignored if the checkpoint directory is not set in the SparkContext. (default: 10)\n",
      "featuresCol: features column name. (default: features)\n",
      "labelCol: label column name. (default: label)\n",
      "lossType: Loss function which GBT tries to minimize (case-insensitive). Supported options: logistic (default: logistic)\n",
      "maxBins: Max number of bins for discretizing continuous features.  Must be >=2 and >= number of categories for any categorical feature. (default: 32)\n",
      "maxDepth: Maximum depth of the tree. (>= 0) E.g., depth 0 means 1 leaf node; depth 1 means 1 internal node + 2 leaf nodes. (default: 5)\n",
      "maxIter: max number of iterations (>= 0). (default: 20, current: 10)\n",
      "maxMemoryInMB: Maximum memory in MB allocated to histogram aggregation. If too small, then 1 node will be split per iteration, and its aggregates may exceed this size. (default: 256)\n",
      "minInfoGain: Minimum information gain for a split to be considered at a tree node. (default: 0.0)\n",
      "minInstancesPerNode: Minimum number of instances each child must have after split. If a split causes the left or right child to have fewer than minInstancesPerNode, the split will be discarded as invalid. Should be >= 1. (default: 1)\n",
      "predictionCol: prediction column name. (default: prediction)\n",
      "seed: random seed. (default: -7674267899484452859)\n",
      "stepSize: Step size (a.k.a. learning rate) in interval (0, 1] for shrinking the contribution of each estimator. (default: 0.1)\n",
      "subsamplingRate: Fraction of the training data used for learning each decision tree, in range (0, 1]. (default: 1.0)\n"
     ]
    }
   ],
   "source": [
    "print(gbt.explainParams())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8981050997838095"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pyspark.ml.tuning import ParamGridBuilder, CrossValidator\n",
    "\n",
    "paramGrid = (ParamGridBuilder()\n",
    "             .addGrid(gbt.maxDepth, [2, 4, 6])\n",
    "             .addGrid(gbt.maxBins, [20, 60])\n",
    "             .addGrid(gbt.maxIter, [10, 20])\n",
    "             .build())\n",
    "\n",
    "cv = CrossValidator(estimator=gbt, estimatorParamMaps=paramGrid, evaluator=evaluator, numFolds=5)\n",
    "\n",
    "# Run cross validations.  This can take about 6 minutes since it is training over 20 trees!\n",
    "cvModel = cv.fit(train)\n",
    "predictions = cvModel.transform(test)\n",
    "evaluator.evaluate(predictions)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "conda_python3",
   "language": "python",
   "name": "conda_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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}