MPP ROC Curve.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Links to other notebooks in the same folder:\n",
    "<a href='http://pivotal.io/data-science'><img src='https://raw.githubusercontent.com/crawles/Logos/master/Pivotal_TealOnWhite.png' width='200px' align='right'></a>\n",
    "\n",
    "<nav class = \"navbar navbar-light bg-faded\">\n",
    "    <ul class = \"nav navbar-nav\">\n",
    "        <li class = \"\">\n",
    "            <a href = \"MPP Plotting.ipynb\">MPP Plotting</a>\n",
    "        </li>\n",
    "        <li class = \"\">\n",
    "            <a href = \"MPP Decision Tree.ipynb\">MPP Decision Tree</a>\n",
    "        </li>\n",
    "        <li class = \"\">\n",
    "            <a>MPP ROC Curve</a>\n",
    "        </li>\n",
    "        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc": "true"
   },
   "source": [
    " # Table of Contents\n",
    "<div class=\"toc\" style=\"margin-top: 1em;\"><ul class=\"toc-item\" id=\"toc-level0\"><li><span><a href=\"http://localhost:8888/notebooks/mpp-plotting/MPP%20ROC%20Curve.ipynb#Import-useful-libraries\" data-toc-modified-id=\"Import-useful-libraries-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Import useful libraries</a></span></li><li><span><a href=\"http://localhost:8888/notebooks/mpp-plotting/MPP%20ROC%20Curve.ipynb#Connect-to-Database\" data-toc-modified-id=\"Connect-to-Database-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>Connect to Database</a></span></li><li><span><a href=\"http://localhost:8888/notebooks/mpp-plotting/MPP%20ROC%20Curve.ipynb#Autofill-Table-Names\" data-toc-modified-id=\"Autofill-Table-Names-3\"><span class=\"toc-item-num\">3&nbsp;&nbsp;</span>Autofill Table Names</a></span></li><li><span><a href=\"http://localhost:8888/notebooks/mpp-plotting/MPP%20ROC%20Curve.ipynb#Define-magic-functions---helpful-for-interacting-with-cluster\" data-toc-modified-id=\"Define-magic-functions---helpful-for-interacting-with-cluster-4\"><span class=\"toc-item-num\">4&nbsp;&nbsp;</span>Define magic functions - helpful for interacting with cluster</a></span></li><li><span><a href=\"http://localhost:8888/notebooks/mpp-plotting/MPP%20ROC%20Curve.ipynb#Create-Fake-Data\" data-toc-modified-id=\"Create-Fake-Data-5\"><span class=\"toc-item-num\">5&nbsp;&nbsp;</span>Create Fake Data</a></span></li></ul></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import useful libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:47.911088Z",
     "start_time": "2017-11-07T22:24:47.189784Z"
    }
   },
   "outputs": [],
   "source": [
    "from datetime import datetime\n",
    "from dateutil.relativedelta import relativedelta\n",
    "import getopt\n",
    "import os\n",
    "import sys\n",
    "import urllib\n",
    "\n",
    "from IPython.core.display import display, HTML\n",
    "from IPython.core.magic import register_cell_magic, register_line_cell_magic,\\\n",
    "                               register_line_magic\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pandas.io.sql as psql\n",
    "import psycopg2\n",
    "import seaborn as sns\n",
    "\n",
    "# Credentials file to connect to database\n",
    "import credentials\n",
    "from mpp_plotting import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:48.265730Z",
     "start_time": "2017-11-07T22:24:47.914107Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<script>logo = document.getElementById(\"kernel_logo_widget\").getElementsByClassName(\"current_kernel_logo\")[0];\n",
       "logo.src = \"https://github.com/crawles/Logos/blob/master/Pivotal_Python_138x40.png?raw=true\";\n",
       "logo.style = \"display: inline; width:138px; height:40px\";\n",
       "logo.innerHTML = '<a href=\"http://pivotal.io/data-science\">' + logo.innerHTML + '</a>';\n",
       "</script>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Changes logo to a Pivotal logo\n",
    "jPrefs = urllib.urlopen(\"https://raw.githubusercontent.com/crawles/Logos/master/jupyterPrefs.js\").read()\n",
    "HTML('<script>{}</script>'.format(jPrefs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:48.310000Z",
     "start_time": "2017-11-07T22:24:48.268890Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>.container {width:80% !important;}</style>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Set default cell width\n",
    "display(HTML('<style>.container {width:80% !important;}</style>'))\n",
    "\n",
    "# Set default matplotlib settings\n",
    "plt.rcParams['figure.figsize'] = (10, 7)\n",
    "plt.rcParams['lines.linewidth'] = 3\n",
    "plt.rcParams['figure.titlesize'] = 26\n",
    "plt.rcParams['axes.labelsize'] = 18\n",
    "plt.rcParams['axes.titlesize'] = 22\n",
    "plt.rcParams['xtick.labelsize'] = 14\n",
    "plt.rcParams['ytick.labelsize'] = 14\n",
    "plt.rcParams['legend.fontsize'] = 16\n",
    "\n",
    "# Set Seaborn Colours\n",
    "sns.set_palette('colorblind')\n",
    "sns.set_style('darkgrid')\n",
    "blue, green, red, purple, yellow, cyan = sns.color_palette('colorblind')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Connect to Database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:48.402878Z",
     "start_time": "2017-11-07T22:24:48.312571Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<cursor object at 0x113478a50; closed: 0>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conn = psycopg2.connect(**credentials.login_info_dict)\n",
    "conn.autocommit = True\n",
    "\n",
    "schema_name = 'template'\n",
    "psql.execute('SET search_path TO {}'.format(schema_name), conn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Autofill Table Names\n",
    "One downside of interacting with a remote database in Python is that table names will not be imported in. When this is run, it will import all of the schema names as classes and their respective table names as variables. That way, when we type a schema name, we can use tab completion to list out all of its columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:48.442745Z",
     "start_time": "2017-11-07T22:24:48.405577Z"
    }
   },
   "outputs": [],
   "source": [
    "class Schema:\n",
    "    def __init__(self, tables):\n",
    "        for t in tables:\n",
    "            exec('self.{t} = \"{t}\"'.format(t=t))\n",
    "\n",
    "def refresh_tables(conn):\n",
    "    \"\"\"Refreshes the auto-fill tables.\"\"\"\n",
    "    sql = '''\n",
    "    SELECT table_schema, array_agg(table_name::TEXT) AS tables\n",
    "      FROM information_schema.tables\n",
    "     GROUP BY table_schema;\n",
    "    '''\n",
    "    info_df = psql.read_sql(sql, conn)\n",
    "\n",
    "    for row in info_df.iterrows():\n",
    "        command = '''\n",
    "        global {s}\n",
    "        {s} = Schema({tables})\n",
    "        '''.format(s=row[1][0], tables=row[1][1])\n",
    "        exec(dedent(command))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Define magic functions - helpful for interacting with cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:48.537204Z",
     "start_time": "2017-11-07T22:24:48.446744Z"
    }
   },
   "outputs": [],
   "source": [
    "@register_cell_magic\n",
    "def readsql(line, cell):\n",
    "    \"\"\"\n",
    "    Extract the code in the specific cell (should be valid SQL), \n",
    "    and execute it using the connection object to the backend \n",
    "    database. The resulting pandas DataFrame is rendered inline \n",
    "    below the cell using IPython.display. You'd use this for SELECT.\n",
    "    \n",
    "    Returns a DataFrame with the name specified in the magic\n",
    "    function. If this is not specified, then the DataFrame\n",
    "    is called _df. This also takes in an option \"-h\", followed \n",
    "    by a number. This will show only the specified number of rows\n",
    "    in the DataFrame.\n",
    "    \"\"\"\n",
    "    # Use the global connection object defined above.\n",
    "    global conn\n",
    "    optlist, args = getopt.getopt(line.split(), 'ih:')\n",
    "    optdict = dict(optlist) \n",
    "    # If '-h' tag is specified, set the number of rows to display\n",
    "    if '-h' in optdict:\n",
    "        head_num = int(optdict['-h'])\n",
    "     \n",
    "    split_cell = cell.split('$')\n",
    "    if '-i' not in optdict:\n",
    "        if len(split_cell) > 1:\n",
    "            split_cell[0] = split_cell[0].format(**globals())\n",
    "            split_cell[-1] = split_cell[-1].format(**globals())\n",
    "            cell = '$'.join(split_cell)\n",
    "        elif len(split_cell) == 1:\n",
    "            cell = cell.format(**globals()) \n",
    "    \n",
    "    # If there is more than one table name specified,\n",
    "    # throw an exception.\n",
    "    if len(args) > 1:\n",
    "        raise Exception('More than one table name specified')\n",
    "\n",
    "    elif len(args) == 1:\n",
    "        # If a table name is specified, store it as that\n",
    "        table_name = args[0]\n",
    "        globals()[table_name] = psql.read_sql(cell, conn)\n",
    "        if '-h' in optdict:\n",
    "            # If head_num is not 0, then display rows\n",
    "            if head_num != 0:\n",
    "                display(globals()[table_name].head(int(optdict['-h'])))\n",
    "        else:\n",
    "            display(globals()[table_name])\n",
    "\n",
    "    else:\n",
    "        # Otherwise, call it _df\n",
    "        global _df\n",
    "        _df = psql.read_sql(cell, conn)\n",
    "        if '-h' in optdict:\n",
    "            # If head_num is not 0, then display rows\n",
    "            if head_num != 0:\n",
    "                display(_df.head(head_num))\n",
    "        else:\n",
    "            display(_df)\n",
    "\n",
    "@register_cell_magic\n",
    "def execsql(line, cell):\n",
    "    \"\"\"\n",
    "    Extract the code in the specific cell (should be valid SQL),\n",
    "    and execute it using the connection object to the backend \n",
    "    database. You'd use this for CREATE/UPDATE/DELETE.\n",
    "    \"\"\"\n",
    "    # Use the global connection object defined above.\n",
    "    global conn\n",
    "    optlist, args = getopt.getopt(line.split(), 'ih:')\n",
    "    optdict = dict(optlist)\n",
    "\n",
    "    split_cell = cell.split('$')\n",
    "    if '-i' not in optdict:\n",
    "        if len(split_cell) > 1:\n",
    "            split_cell[0] = split_cell[0].format(**globals())\n",
    "            split_cell[-1] = split_cell[-1].format(**globals())\n",
    "            cell = '$'.join(split_cell)\n",
    "        elif len(split_cell) == 1:\n",
    "            cell = cell.format(**globals())\n",
    "    psql.execute(cell, conn)\n",
    "\n",
    "@register_cell_magic\n",
    "def printsql(line, cell):\n",
    "    \"\"\"\n",
    "    Show the SQL query that will be run.\n",
    "    \"\"\"\n",
    "    optlist, args = getopt.getopt(line.split(), 'ih:')\n",
    "    optdict = dict(optlist)\n",
    "    \n",
    "    split_cell = cell.split('$')\n",
    "    if '-i' not in optdict:\n",
    "        if len(split_cell) > 1:\n",
    "            split_cell[0] = split_cell[0].format(**globals())\n",
    "            split_cell[-1] = split_cell[-1].format(**globals())\n",
    "            cell = '$'.join(split_cell)\n",
    "        elif len(split_cell) == 1:\n",
    "            cell = cell.format(**globals())\n",
    "    print cell\n",
    "\n",
    "# We delete these to avoid name conflicts for automagic to work\n",
    "del execsql, readsql, printsql"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create Fake Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:48.604651Z",
     "start_time": "2017-11-07T22:24:48.539310Z"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%%execsql\n",
    "DROP TABLE IF EXISTS roc_data;\n",
    "CREATE TABLE roc_data\n",
    "   AS SELECT random()::INT AS y_true,\n",
    "             random() AS y_score\n",
    "        FROM generate_series(1, 2000);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:48.655824Z",
     "start_time": "2017-11-07T22:24:48.607674Z"
    }
   },
   "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>y_true</th>\n",
       "      <th>y_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.000286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0.000787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0.001300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0.002896</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   y_true   y_score\n",
       "0       0  0.000142\n",
       "1       1  0.000286\n",
       "2       1  0.000787\n",
       "3       0  0.001300\n",
       "4       1  0.002896"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%readsql -h 5 df\n",
    "SELECT *\n",
    "  FROM roc_data\n",
    " ORDER BY y_score;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:49.048482Z",
     "start_time": "2017-11-07T22:24:48.658178Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAHwCAYAAAC7apkrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1135e4ed0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_curve, roc_auc_score\n",
    "\n",
    "fpr, tpr, thresh = roc_curve(df.y_true, df.y_score)\n",
    "auc_score = roc_auc_score(df.y_true, df.y_score)\n",
    "plt.plot(fpr, tpr)\n",
    "plt.title('ROC Curve (AUC = {})'.format(auc_score))\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We compare this to the ROC curve computed in database."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:49.155046Z",
     "start_time": "2017-11-07T22:24:49.050672Z"
    }
   },
   "outputs": [],
   "source": [
    "roc_df = get_roc_values('roc_data', 'y_true', 'y_score', conn)\n",
    "# Compute the AUC in database\n",
    "db_auc_score = get_roc_auc_score(roc_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-11-07T22:24:49.374223Z",
     "start_time": "2017-11-07T22:24:49.158126Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAHwCAYAAAC7apkrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11795c2d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(roc_df.fpr, roc_df.tpr)\n",
    "plt.title('ROC Curve (AUC = {})'.format(db_auc_score))\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.tight_layout()"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "hide_input": false,
  "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"
  },
  "nav_menu": {},
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "toc_cell": true,
   "toc_position": {},
   "toc_section_display": "block",
   "toc_window_display": true
  },
  "toc_position": {
   "height": "982px",
   "left": "0px",
   "right": "auto",
   "top": "106px",
   "width": "212px"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}