4. Data Cleaning and Preparation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "#### Large portion of Data Analysis is taken up by preparation: loading, cleaning, transforming and rearranging.\n",
    "#### These tasks take up more than 80% of an analyst's time. This is because the way the data is stored in files or databases is not in the right format.\n",
    "#### Researchers prefer to do ad-hoc processing of data from one form to another using languages like R, Python, etc.\n",
    "#### pandas with built-in Python features provides high-level, flexible abd fast set of tools that enables you to manipulate data into right form."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Handling missing data\n",
    "#### The way missing data is represented in pandas is imperfect but functional for lot of users.\n",
    "#### For numeric data, pandas uses floating-point value NaN. It is called a Sentinel value and can be easily detected."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      aardvak\n",
       "1    artichoke\n",
       "2          NaN\n",
       "3      avocado\n",
       "dtype: object"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "string_data = pd.Series(['aardvak', 'artichoke', np.nan, 'avocado'])\n",
    "string_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    False\n",
       "1    False\n",
       "2     True\n",
       "3    False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "string_data.isnull()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### This convention of NaN was adopted from R, where missing values are refered as NA.\n",
    "#### In statistics, NA may either be data that does not exist or data that was not observed.\n",
    "#### During cleaning data, we should also analyse the missing data to identify data collection problems or potential bias due to missing data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0     True\n",
       "1    False\n",
       "2     True\n",
       "3    False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "string_data[0] = None\n",
    "\n",
    "string_data.isnull()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Filtering 'Out' Missing Data\n",
    "#### We always have the option to filter out missing data by hand using 'isnull' and boolean indexing.\n",
    "#### The 'dropna' function can be pretty useful too. For a Series it returns the Series with only non-null data and index values.\n",
    "#### For DataFrame, it is a bit complex. dropna by default will drop any row that contains even 1 missing value. By passing \"how='all'\" will target rows with all NAs.\n",
    "#### To drop columns, pass 'axis=1'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1.0\n",
       "2    3.5\n",
       "4    7.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from numpy import nan as NA\n",
    "\n",
    "data = pd.Series([1, NA, 3.5, NA, 7])\n",
    "data.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1.0\n",
       "2    3.5\n",
       "4    7.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[data.notnull()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     0    1    2\n",
       "0  1.0  6.5  3.0\n",
       "1  1.0  NaN  NaN\n",
       "2  NaN  NaN  NaN\n",
       "3  NaN  6.5  3.0"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.DataFrame([[1., 6.5, 3.], [1., NA, NA],\n",
    "                   [NA, NA, NA], [NA, 6.5, 3.]])\n",
    "\n",
    "cleaned = data.dropna()\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     0    1    2\n",
       "0  1.0  6.5  3.0"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleaned"
   ]
  },
  {
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     0    1    2\n",
       "0  1.0  6.5  3.0\n",
       "1  1.0  NaN  NaN\n",
       "3  NaN  6.5  3.0"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.dropna(how='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     0    1    2   4\n",
       "0  1.0  6.5  3.0 NaN\n",
       "1  1.0  NaN  NaN NaN\n",
       "2  NaN  NaN  NaN NaN\n",
       "3  NaN  6.5  3.0 NaN"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[4] = NA\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>6.5</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     0    1    2\n",
       "0  1.0  6.5  3.0\n",
       "1  1.0  NaN  NaN\n",
       "2  NaN  NaN  NaN\n",
       "3  NaN  6.5  3.0"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.dropna(axis=1, how='all')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Another DataFrame cleaning method concerns with time series data.\n",
    "#### To keep only rows with certain number of observations, use the 'thresh' argument.\n",
    "#### Any row that contains number of NAs greater than or equal thresh will be eliminated."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.152841</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.392019</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.138522</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.801143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.727443</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.011845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.008346</td>\n",
       "      <td>-0.910039</td>\n",
       "      <td>1.191962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.846771</td>\n",
       "      <td>1.635657</td>\n",
       "      <td>0.192299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.611303</td>\n",
       "      <td>-0.422075</td>\n",
       "      <td>0.599414</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "0 -1.152841       NaN       NaN\n",
       "1  1.392019       NaN       NaN\n",
       "2 -0.138522       NaN -0.801143\n",
       "3  1.727443       NaN  0.011845\n",
       "4 -0.008346 -0.910039  1.191962\n",
       "5  0.846771  1.635657  0.192299\n",
       "6  0.611303 -0.422075  0.599414"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(np.random.randn(7,3))\n",
    "df.iloc[:4, 1] = NA\n",
    "df.iloc[:2, 2] = NA\n",
    "\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.008346</td>\n",
       "      <td>-0.910039</td>\n",
       "      <td>1.191962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.846771</td>\n",
       "      <td>1.635657</td>\n",
       "      <td>0.192299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.611303</td>\n",
       "      <td>-0.422075</td>\n",
       "      <td>0.599414</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "4 -0.008346 -0.910039  1.191962\n",
       "5  0.846771  1.635657  0.192299\n",
       "6  0.611303 -0.422075  0.599414"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.138522</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.801143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.727443</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.011845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.008346</td>\n",
       "      <td>-0.910039</td>\n",
       "      <td>1.191962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.846771</td>\n",
       "      <td>1.635657</td>\n",
       "      <td>0.192299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.611303</td>\n",
       "      <td>-0.422075</td>\n",
       "      <td>0.599414</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "2 -0.138522       NaN -0.801143\n",
       "3  1.727443       NaN  0.011845\n",
       "4 -0.008346 -0.910039  1.191962\n",
       "5  0.846771  1.635657  0.192299\n",
       "6  0.611303 -0.422075  0.599414"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dropna(thresh=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Filling In Missing Data\n",
    "#### Rather than removing NAs and discarding important information in the same rows, we can also fill in the NAs in different ways.\n",
    "#### The 'fillna' is a workhorse function, where the constant we pass replaces missing values.\n",
    "#### If we call fillna with a dict, we can fill different value for each column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.152841</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.392019</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.138522</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.801143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.727443</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.011845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.008346</td>\n",
       "      <td>-0.910039</td>\n",
       "      <td>1.191962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.846771</td>\n",
       "      <td>1.635657</td>\n",
       "      <td>0.192299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.611303</td>\n",
       "      <td>-0.422075</td>\n",
       "      <td>0.599414</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "0 -1.152841  0.000000  0.000000\n",
       "1  1.392019  0.000000  0.000000\n",
       "2 -0.138522  0.000000 -0.801143\n",
       "3  1.727443  0.000000  0.011845\n",
       "4 -0.008346 -0.910039  1.191962\n",
       "5  0.846771  1.635657  0.192299\n",
       "6  0.611303 -0.422075  0.599414"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.152841</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.392019</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.138522</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>-0.801143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.727443</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.011845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.008346</td>\n",
       "      <td>-0.910039</td>\n",
       "      <td>1.191962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.846771</td>\n",
       "      <td>1.635657</td>\n",
       "      <td>0.192299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.611303</td>\n",
       "      <td>-0.422075</td>\n",
       "      <td>0.599414</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "0 -1.152841  0.500000  0.000000\n",
       "1  1.392019  0.500000  0.000000\n",
       "2 -0.138522  0.500000 -0.801143\n",
       "3  1.727443  0.500000  0.011845\n",
       "4 -0.008346 -0.910039  1.191962\n",
       "5  0.846771  1.635657  0.192299\n",
       "6  0.611303 -0.422075  0.599414"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.fillna({1:0.5, 2: 0})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### By default it returns a new object, but we can modify it to change in-place.\n",
    "#### The interpolation methods used for reindexing like 'ffill' can also be used with fillna.\n",
    "#### It allows you to do lots of creative things, like filling with mean or median values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.152841</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.392019</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.138522</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.801143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.727443</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.011845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.008346</td>\n",
       "      <td>-0.910039</td>\n",
       "      <td>1.191962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.846771</td>\n",
       "      <td>1.635657</td>\n",
       "      <td>0.192299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.611303</td>\n",
       "      <td>-0.422075</td>\n",
       "      <td>0.599414</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "0 -1.152841  0.000000  0.000000\n",
       "1  1.392019  0.000000  0.000000\n",
       "2 -0.138522  0.000000 -0.801143\n",
       "3  1.727443  0.000000  0.011845\n",
       "4 -0.008346 -0.910039  1.191962\n",
       "5  0.846771  1.635657  0.192299\n",
       "6  0.611303 -0.422075  0.599414"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "_ = df.fillna(0, inplace=True)\n",
    "\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.196822</td>\n",
       "      <td>-1.728530</td>\n",
       "      <td>0.011651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.635502</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>1.468730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.357786</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.239709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.456500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.379100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.511740</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-0.005963</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "0 -1.196822 -1.728530  0.011651\n",
       "1  0.635502 -1.557243  1.468730\n",
       "2 -0.357786       NaN  0.239709\n",
       "3 -0.456500       NaN -0.379100\n",
       "4 -0.511740       NaN       NaN\n",
       "5 -0.005963       NaN       NaN"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(np.random.randn(6,3))\n",
    "df.iloc[2:, 1] = NA\n",
    "df.iloc[4:, 2] = NA\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.196822</td>\n",
       "      <td>-1.728530</td>\n",
       "      <td>0.011651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.635502</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>1.468730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.357786</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>0.239709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.456500</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>-0.379100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.511740</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>-0.379100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-0.005963</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>-0.379100</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "0 -1.196822 -1.728530  0.011651\n",
       "1  0.635502 -1.557243  1.468730\n",
       "2 -0.357786 -1.557243  0.239709\n",
       "3 -0.456500 -1.557243 -0.379100\n",
       "4 -0.511740 -1.557243 -0.379100\n",
       "5 -0.005963 -1.557243 -0.379100"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.fillna(method='ffill')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.196822</td>\n",
       "      <td>-1.728530</td>\n",
       "      <td>0.011651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.635502</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>1.468730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.357786</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>0.239709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.456500</td>\n",
       "      <td>-1.557243</td>\n",
       "      <td>-0.379100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.511740</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.379100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-0.005963</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.379100</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2\n",
       "0 -1.196822 -1.728530  0.011651\n",
       "1  0.635502 -1.557243  1.468730\n",
       "2 -0.357786 -1.557243  0.239709\n",
       "3 -0.456500 -1.557243 -0.379100\n",
       "4 -0.511740       NaN -0.379100\n",
       "5 -0.005963       NaN -0.379100"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.fillna(method='ffill', limit=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1.000000\n",
       "1    3.833333\n",
       "2    3.500000\n",
       "3    3.833333\n",
       "4    7.000000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.Series([1., NA, 3.5, NA, 7])\n",
    "\n",
    "data.fillna(data.mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Transformation\n",
    "#### Till now we have seen methods for rearranging data.\n",
    "#### Transformation involves filtering, cleaning and other different functions."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Removing Duplicates\n",
    "#### The DataFrame method 'duplicated' returns boolean Series indicating if each row is a duplicate (i.e. observed in a previous row) or not.\n",
    "#### Similarly, 'drop_duplicates' returns DataFrame where 'duplicated' array is False."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>k1</th>\n",
       "      <th>k2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>one</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>two</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>one</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>two</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>one</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>two</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>two</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    k1  k2\n",
       "0  one   1\n",
       "1  two   1\n",
       "2  one   2\n",
       "3  two   3\n",
       "4  one   3\n",
       "5  two   4\n",
       "6  two   4"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.DataFrame({'k1': ['one', 'two'] * 3 + ['two'],\n",
    "                    'k2': [1,1,2,3,3,4,4]})\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    False\n",
       "1    False\n",
       "2    False\n",
       "3    False\n",
       "4    False\n",
       "5    False\n",
       "6     True\n",
       "dtype: bool"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.duplicated()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>k1</th>\n",
       "      <th>k2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>one</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>two</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>one</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>two</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>one</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>two</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    k1  k2\n",
       "0  one   1\n",
       "1  two   1\n",
       "2  one   2\n",
       "3  two   3\n",
       "4  one   3\n",
       "5  two   4"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.drop_duplicates()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Both the above methods by default consider all of the columns. You can also specify any subset of the DataFrame to detect duplicates.\n",
    "#### By default, both keep the first observation in case of duplicates. We can specify \"keep='last'\" to instead keep the last observation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>k1</th>\n",
       "      <th>k2</th>\n",
       "      <th>v1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>one</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>two</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    k1  k2  v1\n",
       "0  one   1   0\n",
       "1  two   1   1"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['v1'] = range(7)\n",
    "data.drop_duplicates(['k1'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>k1</th>\n",
       "      <th>k2</th>\n",
       "      <th>v1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>one</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>two</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>one</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>two</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>one</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>two</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    k1  k2  v1\n",
       "0  one   1   0\n",
       "1  two   1   1\n",
       "2  one   2   2\n",
       "3  two   3   3\n",
       "4  one   3   4\n",
       "6  two   4   6"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.drop_duplicates(['k1', 'k2'], keep='last')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Transforming Data Using a Function or Mapping\n",
    "#### We sometimes need to make transformations based on the values present in an array, Series or column in a DataFrame.\n",
    "#### We can use the map method with a function or dict-like object having the mapping to add or change a column.\n",
    "#### Sometimes the column that we base our mapping on may have varying case from our map. In such a case, we can convert all the values to lowercase. Or just pass a function that does it for us."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>food</th>\n",
       "      <th>ounces</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>bacon</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>pulled pork</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bacon</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Pastrami</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>corned beef</td>\n",
       "      <td>7.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Bacon</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>pastrami</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>honey ham</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>nova lox</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          food  ounces\n",
       "0        bacon     4.0\n",
       "1  pulled pork     3.0\n",
       "2        bacon    12.0\n",
       "3     Pastrami     6.0\n",
       "4  corned beef     7.5\n",
       "5        Bacon     8.0\n",
       "6     pastrami     3.0\n",
       "7    honey ham     5.0\n",
       "8     nova lox     6.0"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.DataFrame({'food': ['bacon', 'pulled pork', 'bacon',\n",
    "                             'Pastrami', 'corned beef', 'Bacon',\n",
    "                            'pastrami', 'honey ham', 'nova lox'],\n",
    "                    'ounces': [4, 3, 12, 6, 7.5, 8, 3, 5, 6]})\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "meat_to_animal = {\n",
    "    'bacon': 'pig',\n",
    "    'pulled pork': 'pig',\n",
    "    'pastrami': 'cow',\n",
    "    'corned beef': 'cow',\n",
    "    'honey ham': 'pig',\n",
    "    'nova lox': 'salmon'\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0          bacon\n",
       "1    pulled pork\n",
       "2          bacon\n",
       "3       pastrami\n",
       "4    corned beef\n",
       "5          bacon\n",
       "6       pastrami\n",
       "7      honey ham\n",
       "8       nova lox\n",
       "Name: food, dtype: object"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lowercased = data['food'].str.lower()\n",
    "lowercased"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>food</th>\n",
       "      <th>ounces</th>\n",
       "      <th>animal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>bacon</td>\n",
       "      <td>4.0</td>\n",
       "      <td>pig</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>pulled pork</td>\n",
       "      <td>3.0</td>\n",
       "      <td>pig</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bacon</td>\n",
       "      <td>12.0</td>\n",
       "      <td>pig</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Pastrami</td>\n",
       "      <td>6.0</td>\n",
       "      <td>cow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>corned beef</td>\n",
       "      <td>7.5</td>\n",
       "      <td>cow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Bacon</td>\n",
       "      <td>8.0</td>\n",
       "      <td>pig</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>pastrami</td>\n",
       "      <td>3.0</td>\n",
       "      <td>cow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>honey ham</td>\n",
       "      <td>5.0</td>\n",
       "      <td>pig</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>nova lox</td>\n",
       "      <td>6.0</td>\n",
       "      <td>salmon</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          food  ounces  animal\n",
       "0        bacon     4.0     pig\n",
       "1  pulled pork     3.0     pig\n",
       "2        bacon    12.0     pig\n",
       "3     Pastrami     6.0     cow\n",
       "4  corned beef     7.5     cow\n",
       "5        Bacon     8.0     pig\n",
       "6     pastrami     3.0     cow\n",
       "7    honey ham     5.0     pig\n",
       "8     nova lox     6.0  salmon"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['animal'] = lowercased.map(meat_to_animal)\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       pig\n",
       "1       pig\n",
       "2       pig\n",
       "3       cow\n",
       "4       cow\n",
       "5       pig\n",
       "6       cow\n",
       "7       pig\n",
       "8    salmon\n",
       "Name: food, dtype: object"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['food'].map(lambda x: meat_to_animal[x.lower()])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Replacing Values\n",
    "#### The fillna method is a special case of more general values replacement.\n",
    "#### The map function modifies a subset of values but, 'replace' provides simpler and more flexible way to do so.\n",
    "#### Passing the sentinel (or garbage) value followed by the replcae value will create a new object with the values replaced.\n",
    "#### If we want in-place replacement, use \"inplace=True\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      1.0\n",
       "1   -999.0\n",
       "2      2.0\n",
       "3   -999.0\n",
       "4      3.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.Series([1., -999., 2., -999., 3.])\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1.0\n",
       "1    NaN\n",
       "2    2.0\n",
       "3    NaN\n",
       "4    3.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.replace(-999, np.nan)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### To replace multiple values with a single value, pass a list followed by substitute value.\n",
    "#### To have different replacements for different values, pass list of substitutes.\n",
    "#### You can also pass a dict as argument to replace multiple substitutes.\n",
    "#### NOTE - 'data.replace' is different from 'data.str.replace'. The latter is for element-wise string substitution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1.0\n",
       "1    NaN\n",
       "2    2.0\n",
       "3    NaN\n",
       "4    3.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.replace([-999, -1000], np.nan)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1.0\n",
       "1    NaN\n",
       "2    2.0\n",
       "3    NaN\n",
       "4    3.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.replace([-999, -1000], [np.nan, 0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1.0\n",
       "1    NaN\n",
       "2    2.0\n",
       "3    NaN\n",
       "4    3.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.replace({-999: np.nan, -1000: 0})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Renaming Axis Indexes\n",
    "#### Just like values, axis labels can also be transformed by a function or mapping to produce differntly labeled objects.\n",
    "#### We can also modify axes in-place without any new data structure."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.DataFrame(np.arange(12).reshape((3,4)),\n",
    "                   index = ['Ohio', 'Colorado', 'New York'],\n",
    "                   columns = ['one', 'two', 'three', 'four'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['OHIO', 'COLO', 'NEW '], dtype='object')"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transform = lambda x: x[:4].upper()\n",
    "\n",
    "data.index.map(transform)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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>one</th>\n",
       "      <th>two</th>\n",
       "      <th>three</th>\n",
       "      <th>four</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>OHIO</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>COLO</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NEW</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      one  two  three  four\n",
       "OHIO    0    1      2     3\n",
       "COLO    4    5      6     7\n",
       "NEW     8    9     10    11"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.index = data.index.map(transform)\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### To get a transformed version of a dataset without modifying the original, use 'rename'.\n",
    "#### It can also be used in conjunction with a dict-like object providing new values for subset of the axis labels.\n",
    "#### It saves you from copying DataFrame manuallyand then assigning it index and columns. To modify in-place, use parameter 'inplace=True'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>ONE</th>\n",
       "      <th>TWO</th>\n",
       "      <th>THREE</th>\n",
       "      <th>FOUR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Ohio</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Colo</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>New</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      ONE  TWO  THREE  FOUR\n",
       "Ohio    0    1      2     3\n",
       "Colo    4    5      6     7\n",
       "New     8    9     10    11"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.rename(index=str.title, columns=str.upper)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "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>one</th>\n",
       "      <th>two</th>\n",
       "      <th>peekaboo</th>\n",
       "      <th>four</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>INDIANA</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>COLO</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NEW</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         one  two  peekaboo  four\n",
       "INDIANA    0    1         2     3\n",
       "COLO       4    5         6     7\n",
       "NEW        8    9        10    11"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.rename(index={'OHIO':\"INDIANA\"},\n",
    "           columns = {'three':'peekaboo'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>one</th>\n",
       "      <th>two</th>\n",
       "      <th>three</th>\n",
       "      <th>four</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>INDIANA</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>COLO</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NEW</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         one  two  three  four\n",
       "INDIANA    0    1      2     3\n",
       "COLO       4    5      6     7\n",
       "NEW        8    9     10    11"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.rename(index={'OHIO': 'INDIANA'}, inplace=True)\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Discretization and Binning\n",
    "#### Continuous data is always discretized or seperated into 'bins' for analysis.\n",
    "#### To bin a set of continuous data, use the 'cut' method from pandas.\n",
    "#### In below example, we are binnning set of gaes into groups 18 to 25, 26 to 35, 36 to 60 and 61 nd older."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(18, 25], (18, 25], (18, 25], (25, 35], (18, 25], ..., (25, 35], (60, 100], (35, 60], (35, 60], (25, 35]]\n",
       "Length: 12\n",
       "Categories (4, interval[int64]): [(18, 25] < (25, 35] < (35, 60] < (60, 100]]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ages = [20, 22, 25, 27, 21, 23, 37, 31, 61, 45, 41, 32]\n",
    "bins = [18, 25, 35, 60, 100]\n",
    "\n",
    "cats = pd.cut(ages, bins)\n",
    "cats"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### pandas returns a special Categorical object from cuts function.\n",
    "#### The output describes the bins that each of the element is in. You can treat it like a bin name for each element.\n",
    "#### Internally, the output contains a categories array specifying distinct category names along with a labeling for the 'ages' data in the 'codes' attribute."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, 1, 0, 0, 2, 1, 3, 2, 2, 1], dtype=int8)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cats.codes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "IntervalIndex([(18, 25], (25, 35], (35, 60], (60, 100]]\n",
       "              closed='right',\n",
       "              dtype='interval[int64]')"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cats.categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(18, 25]     5\n",
       "(35, 60]     3\n",
       "(25, 35]     3\n",
       "(60, 100]    1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.value_counts(cats)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The interval system for cut is consistent with the mathematical notation. A parenthesis means that the side is open and a square bracket means that it is closed (inclusive).\n",
    "#### We can changes which side is closed by passing 'right=False'.\n",
    "#### We can have our own bin names by passing a list or array to the labels option."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[18, 26), [18, 26), [18, 26), [26, 36), [18, 26), ..., [26, 36), [61, 100), [36, 61), [36, 61), [26, 36)]\n",
       "Length: 12\n",
       "Categories (4, interval[int64]): [[18, 26) < [26, 36) < [36, 61) < [61, 100)]"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.cut(ages, [18, 26, 36, 61, 100], right=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Youth, Youth, Youth, YoungAdult, Youth, ..., YoungAdult, Senior, MiddleAges, MiddleAges, YoungAdult]\n",
       "Length: 12\n",
       "Categories (4, object): [Youth < YoungAdult < MiddleAges < Senior]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "group_names = ['Youth', 'YoungAdult', 'MiddleAges', 'Senior']\n",
    "pd.cut(ages, bins, labels=group_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Instead of specifying specific intervals, we can just pass an integer to get equal length bins of the same number based on the max and min value of the data.\n",
    "#### The 'precision' parameter limits decimal precision in the values. 'precision=2' limits decimal precision to 2 digits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0.51, 0.74], (0.28, 0.51], (0.51, 0.74], (0.055, 0.28], (0.51, 0.74], ..., (0.28, 0.51], (0.055, 0.28], (0.055, 0.28], (0.28, 0.51], (0.055, 0.28]]\n",
       "Length: 20\n",
       "Categories (4, interval[float64]): [(0.055, 0.28] < (0.28, 0.51] < (0.51, 0.74] < (0.74, 0.97]]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = np.random.rand(20)\n",
    "\n",
    "pd.cut(data, 4, precision=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### cut has a closely related function - 'qcut' that bins data based on sample quantiles.\n",
    "#### Based on distribution, using cut will not usually result in each bin have the same number of data points.\n",
    "#### But as qcut uses sample quantiles, you wil rougjly obtain equal-size bins.\n",
    "#### We can even pass our own quantiles to qcut."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(-3.413, -0.659], (0.0656, 0.778], (0.0656, 0.778], (0.778, 3.118], (-0.659, 0.0656], ..., (0.0656, 0.778], (-0.659, 0.0656], (-0.659, 0.0656], (0.0656, 0.778], (0.778, 3.118]]\n",
       "Length: 1000\n",
       "Categories (4, interval[float64]): [(-3.413, -0.659] < (-0.659, 0.0656] < (0.0656, 0.778] < (0.778, 3.118]]"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = np.random.randn(1000)\n",
    "cats = pd.qcut(data, 4)\n",
    "cats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.778, 3.118]      250\n",
       "(0.0656, 0.778]     250\n",
       "(-0.659, 0.0656]    250\n",
       "(-3.413, -0.659]    250\n",
       "dtype: int64"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.value_counts(cats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(-3.413, -1.172], (0.0656, 1.378], (0.0656, 1.378], (0.0656, 1.378], (-1.172, 0.0656], ..., (0.0656, 1.378], (-1.172, 0.0656], (-1.172, 0.0656], (0.0656, 1.378], (1.378, 3.118]]\n",
       "Length: 1000\n",
       "Categories (4, interval[float64]): [(-3.413, -1.172] < (-1.172, 0.0656] < (0.0656, 1.378] < (1.378, 3.118]]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.qcut(data, [0, 0.1, 0.5, 0.9, 1.])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Detecting and Filtering Outliers\n",
    "#### Filtering and Transforming outliers is mostly a matter of applying array operations.\n",
    "#### To find values exceeding a threshold, just use boolean indexing with other functions like 'abs()' based on requirement.\n",
    "#### To get all rows having at least one value exceed a threshold, use the 'any(1)' method.\n",
    "#### Values can also be set based on these criteria. So you can cap values based on an interval or threshold.\n",
    "#### You can also use the 'np.sign()' function to get 1 and -1 where the data is positive or negative respectively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.014491</td>\n",
       "      <td>0.014646</td>\n",
       "      <td>-0.012298</td>\n",
       "      <td>0.039840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.010034</td>\n",
       "      <td>1.010854</td>\n",
       "      <td>1.019774</td>\n",
       "      <td>0.967757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-3.701866</td>\n",
       "      <td>-3.387856</td>\n",
       "      <td>-3.039110</td>\n",
       "      <td>-3.390256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.657859</td>\n",
       "      <td>-0.654879</td>\n",
       "      <td>-0.685004</td>\n",
       "      <td>-0.619386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.066684</td>\n",
       "      <td>-0.001560</td>\n",
       "      <td>0.020953</td>\n",
       "      <td>0.058701</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.692534</td>\n",
       "      <td>0.713840</td>\n",
       "      <td>0.628787</td>\n",
       "      <td>0.696540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.058465</td>\n",
       "      <td>3.226376</td>\n",
       "      <td>4.065409</td>\n",
       "      <td>3.088712</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 0            1            2            3\n",
       "count  1000.000000  1000.000000  1000.000000  1000.000000\n",
       "mean      0.014491     0.014646    -0.012298     0.039840\n",
       "std       1.010034     1.010854     1.019774     0.967757\n",
       "min      -3.701866    -3.387856    -3.039110    -3.390256\n",
       "25%      -0.657859    -0.654879    -0.685004    -0.619386\n",
       "50%       0.066684    -0.001560     0.020953     0.058701\n",
       "75%       0.692534     0.713840     0.628787     0.696540\n",
       "max       3.058465     3.226376     4.065409     3.088712"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.DataFrame(np.random.randn(1000, 4))\n",
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "360    3.099117\n",
       "559    4.065409\n",
       "583   -3.039110\n",
       "976    3.180584\n",
       "Name: 2, dtype: float64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "col = data[2]\n",
    "col[np.abs(col) > 3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>3.058465</td>\n",
       "      <td>0.016649</td>\n",
       "      <td>-1.722464</td>\n",
       "      <td>1.367502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>228</th>\n",
       "      <td>-2.056428</td>\n",
       "      <td>-1.475965</td>\n",
       "      <td>-0.003494</td>\n",
       "      <td>-3.390256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>360</th>\n",
       "      <td>0.124732</td>\n",
       "      <td>-0.404802</td>\n",
       "      <td>3.099117</td>\n",
       "      <td>-0.237581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>388</th>\n",
       "      <td>-0.253419</td>\n",
       "      <td>3.105240</td>\n",
       "      <td>0.078645</td>\n",
       "      <td>-0.148384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>470</th>\n",
       "      <td>-3.701866</td>\n",
       "      <td>-2.025745</td>\n",
       "      <td>0.727206</td>\n",
       "      <td>-1.468939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>472</th>\n",
       "      <td>-0.484421</td>\n",
       "      <td>-3.387856</td>\n",
       "      <td>-0.112233</td>\n",
       "      <td>-0.068768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>508</th>\n",
       "      <td>0.323240</td>\n",
       "      <td>3.226376</td>\n",
       "      <td>-1.808320</td>\n",
       "      <td>-0.094250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>559</th>\n",
       "      <td>-0.579666</td>\n",
       "      <td>1.194872</td>\n",
       "      <td>4.065409</td>\n",
       "      <td>-1.209842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>583</th>\n",
       "      <td>0.658055</td>\n",
       "      <td>1.219842</td>\n",
       "      <td>-3.039110</td>\n",
       "      <td>0.021802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>636</th>\n",
       "      <td>-3.087210</td>\n",
       "      <td>1.114326</td>\n",
       "      <td>-0.585877</td>\n",
       "      <td>0.080978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>664</th>\n",
       "      <td>-1.837414</td>\n",
       "      <td>-0.281346</td>\n",
       "      <td>-0.871586</td>\n",
       "      <td>3.088712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>678</th>\n",
       "      <td>-3.232977</td>\n",
       "      <td>0.249556</td>\n",
       "      <td>2.571467</td>\n",
       "      <td>1.558432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>735</th>\n",
       "      <td>0.080877</td>\n",
       "      <td>-0.120672</td>\n",
       "      <td>-1.181248</td>\n",
       "      <td>-3.254930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>976</th>\n",
       "      <td>0.713734</td>\n",
       "      <td>0.905299</td>\n",
       "      <td>3.180584</td>\n",
       "      <td>0.013758</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            0         1         2         3\n",
       "186  3.058465  0.016649 -1.722464  1.367502\n",
       "228 -2.056428 -1.475965 -0.003494 -3.390256\n",
       "360  0.124732 -0.404802  3.099117 -0.237581\n",
       "388 -0.253419  3.105240  0.078645 -0.148384\n",
       "470 -3.701866 -2.025745  0.727206 -1.468939\n",
       "472 -0.484421 -3.387856 -0.112233 -0.068768\n",
       "508  0.323240  3.226376 -1.808320 -0.094250\n",
       "559 -0.579666  1.194872  4.065409 -1.209842\n",
       "583  0.658055  1.219842 -3.039110  0.021802\n",
       "636 -3.087210  1.114326 -0.585877  0.080978\n",
       "664 -1.837414 -0.281346 -0.871586  3.088712\n",
       "678 -3.232977  0.249556  2.571467  1.558432\n",
       "735  0.080877 -0.120672 -1.181248 -3.254930\n",
       "976  0.713734  0.905299  3.180584  0.013758"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[(np.abs(data) > 3).any(1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.015455</td>\n",
       "      <td>0.014702</td>\n",
       "      <td>-0.013604</td>\n",
       "      <td>0.040396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.006524</td>\n",
       "      <td>1.008608</td>\n",
       "      <td>1.015092</td>\n",
       "      <td>0.965337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-3.000000</td>\n",
       "      <td>-3.000000</td>\n",
       "      <td>-3.000000</td>\n",
       "      <td>-3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.657859</td>\n",
       "      <td>-0.654879</td>\n",
       "      <td>-0.685004</td>\n",
       "      <td>-0.619386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.066684</td>\n",
       "      <td>-0.001560</td>\n",
       "      <td>0.020953</td>\n",
       "      <td>0.058701</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.692534</td>\n",
       "      <td>0.713840</td>\n",
       "      <td>0.628787</td>\n",
       "      <td>0.696540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 0            1            2            3\n",
       "count  1000.000000  1000.000000  1000.000000  1000.000000\n",
       "mean      0.015455     0.014702    -0.013604     0.040396\n",
       "std       1.006524     1.008608     1.015092     0.965337\n",
       "min      -3.000000    -3.000000    -3.000000    -3.000000\n",
       "25%      -0.657859    -0.654879    -0.685004    -0.619386\n",
       "50%       0.066684    -0.001560     0.020953     0.058701\n",
       "75%       0.692534     0.713840     0.628787     0.696540\n",
       "max       3.000000     3.000000     3.000000     3.000000"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Capping outside -3 to 3\n",
    "data[np.abs(data) > 3] = np.sign(data) * 3 \n",
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     0    1    2    3\n",
       "0  1.0  1.0 -1.0 -1.0\n",
       "1 -1.0 -1.0  1.0 -1.0\n",
       "2 -1.0  1.0  1.0 -1.0\n",
       "3  1.0  1.0  1.0  1.0\n",
       "4  1.0  1.0 -1.0  1.0"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sign(data).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Permutation and Random Sampling\n",
    "#### We can easily perform Permutation (randomly reordering) on a Series or rows of a DataFrame using 'numpy.random.permutation'.\n",
    "#### Calling it with length of the axis you want to permute cerates an array of integers indicating the new ordering.\n",
    "#### we can then use the same array in an iloc-based indexing or with an equivalent 'take' function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 0, 4, 2, 3])"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(np.arange(5*4).reshape((5,4)))\n",
    "sampler = np.random.permutation(5)\n",
    "sampler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12</td>\n",
       "      <td>13</td>\n",
       "      <td>14</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>17</td>\n",
       "      <td>18</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    0   1   2   3\n",
       "0   0   1   2   3\n",
       "1   4   5   6   7\n",
       "2   8   9  10  11\n",
       "3  12  13  14  15\n",
       "4  16  17  18  19"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>17</td>\n",
       "      <td>18</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12</td>\n",
       "      <td>13</td>\n",
       "      <td>14</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    0   1   2   3\n",
       "1   4   5   6   7\n",
       "0   0   1   2   3\n",
       "4  16  17  18  19\n",
       "2   8   9  10  11\n",
       "3  12  13  14  15"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.take(sampler)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Use the 'sample' method to select random subset without replacement.\n",
    "#### To generate subset with replacement (i.e. repeat choices) pass 'replace=True'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>17</td>\n",
       "      <td>18</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    0   1   2   3\n",
       "1   4   5   6   7\n",
       "2   8   9  10  11\n",
       "4  16  17  18  19"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(n=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    5\n",
       "1    7\n",
       "0    5\n",
       "4    4\n",
       "4    4\n",
       "4    4\n",
       "4    4\n",
       "3    6\n",
       "0    5\n",
       "2   -1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "choices = pd.Series([5, 7, -1, 6, 4])\n",
    "draws = choices.sample(n=10, replace=True)\n",
    "draws"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Computing Indicator / Dummy Variables\n",
    "#### Another transformation is to convert a categorical variable into a 'dummy' or 'indicator' matrix.\n",
    "#### If a column has k distinct values, we can derive a matrix or a DataFrame with k columns all containing 1s and 0s.\n",
    "#### pandas has the 'get_dummies' function to do this.\n",
    "#### You may want to add a prefix to the columns in the indicator DataFrame. This DataFrame can be merged with other data. get_dummies has the 'prefix' argument to do this."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>a</th>\n",
       "      <th>b</th>\n",
       "      <th>c</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   a  b  c\n",
       "0  0  1  0\n",
       "1  0  1  0\n",
       "2  1  0  0\n",
       "3  0  0  1\n",
       "4  1  0  0\n",
       "5  0  1  0"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({'key': ['b', 'b', 'a', 'c', 'a', 'b'],\n",
    "                  'data1': range(6)})\n",
    "pd.get_dummies(df['key'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "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>data1</th>\n",
       "      <th>key_a</th>\n",
       "      <th>key_b</th>\n",
       "      <th>key_c</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   data1  key_a  key_b  key_c\n",
       "0      0      0      1      0\n",
       "1      1      0      1      0\n",
       "2      2      1      0      0\n",
       "3      3      0      0      1\n",
       "4      4      1      0      0\n",
       "5      5      0      1      0"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dummies = pd.get_dummies(df['key'], prefix='key')\n",
    "df_with_dummy = df[['data1']].join(dummies)\n",
    "df_with_dummy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### But if a row belongs to multiple categories, then dummying it becomes complicated.\n",
    "#### Adding indicator variable to a column like genre involves a bit of data wrangling.\n",
    "#### First we get all the unique categorical values in the DataFrame."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\adity\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:4: ParserWarning: Falling back to the 'python' engine because the 'c' engine does not support regex separators (separators > 1 char and different from '\\s+' are interpreted as regex); you can avoid this warning by specifying engine='python'.\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>movie_id</th>\n",
       "      <th>title</th>\n",
       "      <th>genres</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>Toy Story (1995)</td>\n",
       "      <td>Animation|Children's|Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>Jumanji (1995)</td>\n",
       "      <td>Adventure|Children's|Fantasy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>Grumpier Old Men (1995)</td>\n",
       "      <td>Comedy|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>Waiting to Exhale (1995)</td>\n",
       "      <td>Comedy|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>Father of the Bride Part II (1995)</td>\n",
       "      <td>Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>Heat (1995)</td>\n",
       "      <td>Action|Crime|Thriller</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>Sabrina (1995)</td>\n",
       "      <td>Comedy|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>Tom and Huck (1995)</td>\n",
       "      <td>Adventure|Children's</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>Sudden Death (1995)</td>\n",
       "      <td>Action</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>GoldenEye (1995)</td>\n",
       "      <td>Action|Adventure|Thriller</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   movie_id                               title                        genres\n",
       "0         1                    Toy Story (1995)   Animation|Children's|Comedy\n",
       "1         2                      Jumanji (1995)  Adventure|Children's|Fantasy\n",
       "2         3             Grumpier Old Men (1995)                Comedy|Romance\n",
       "3         4            Waiting to Exhale (1995)                  Comedy|Drama\n",
       "4         5  Father of the Bride Part II (1995)                        Comedy\n",
       "5         6                         Heat (1995)         Action|Crime|Thriller\n",
       "6         7                      Sabrina (1995)                Comedy|Romance\n",
       "7         8                 Tom and Huck (1995)          Adventure|Children's\n",
       "8         9                 Sudden Death (1995)                        Action\n",
       "9        10                    GoldenEye (1995)     Action|Adventure|Thriller"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnames = ['movie_id', 'title', 'genres']\n",
    "\n",
    "movies = pd.read_table('datasets/movielens/movies.dat', sep='::',\n",
    "                      header=None, names=mnames)\n",
    "movies[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Animation', \"Children's\", 'Comedy', 'Adventure', 'Fantasy',\n",
       "       'Romance', 'Drama', 'Action', 'Crime', 'Thriller', 'Horror',\n",
       "       'Sci-Fi', 'Documentary', 'War', 'Musical', 'Mystery', 'Film-Noir',\n",
       "       'Western'], dtype=object)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_genres = []\n",
    "for x in movies.genres:\n",
    "    all_genres.extend(x.split('|'))\n",
    "    \n",
    "genres = pd.unique(all_genres)\n",
    "genres"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### One way to get this would be to start with a DataFrame of all zeros.\n",
    "#### The iterate through each row and set the entry in each row of 'dummies' to 1. For this we use 'dummies.columns' to compute column indices for each category.\n",
    "#### The we use iloc to set values based on those indices. After that we combine this indicator DataFrame with original DataFrame."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "zero_matrix = np.zeros((len(movies), len(genres)))\n",
    "dummies = pd.DataFrame(zero_matrix, columns=genres)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Animation', \"Children's\", 'Comedy']"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gen = movies.genres[0]\n",
    "gen.split('|')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 2], dtype=int64)"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dummies.columns.get_indexer(gen.split('|'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i, gen in enumerate(movies.genres):\n",
    "    indices = dummies.columns.get_indexer(gen.split('|'))\n",
    "    dummies.iloc[i, indices] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "movie_id                                       1\n",
       "title                           Toy Story (1995)\n",
       "genres               Animation|Children's|Comedy\n",
       "Genre_Animation                                1\n",
       "Genre_Children's                               1\n",
       "Genre_Comedy                                   1\n",
       "Genre_Adventure                                0\n",
       "Genre_Fantasy                                  0\n",
       "Genre_Romance                                  0\n",
       "Genre_Drama                                    0\n",
       "Genre_Action                                   0\n",
       "Genre_Crime                                    0\n",
       "Genre_Thriller                                 0\n",
       "Genre_Horror                                   0\n",
       "Genre_Sci-Fi                                   0\n",
       "Genre_Documentary                              0\n",
       "Genre_War                                      0\n",
       "Genre_Musical                                  0\n",
       "Genre_Mystery                                  0\n",
       "Genre_Film-Noir                                0\n",
       "Genre_Western                                  0\n",
       "Name: 0, dtype: object"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "movies_windic = movies.join(dummies.add_prefix('Genre_'))\n",
    "movies_windic.iloc[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### NOTE - For larger data, the above method for creating indicator method is not speedy. It would be better to write a lower-level function that writes directly to NumPy array and then wraps the result in a DataFrame.\n",
    "#### Another useful method in statistical applications is to combine the dummy function with discretization function like cut. This shows if a value is present in a bin or not.\n",
    "#### For below example we will use random seed to make the example more deterministic."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.92961609, 0.31637555, 0.18391881, 0.20456028, 0.56772503,\n",
       "       0.5955447 , 0.96451452, 0.6531771 , 0.74890664, 0.65356987])"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.random.seed(12345)\n",
    "values = np.random.rand(10)\n",
    "values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "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>(0.0, 0.2]</th>\n",
       "      <th>(0.2, 0.4]</th>\n",
       "      <th>(0.4, 0.6]</th>\n",
       "      <th>(0.6, 0.8]</th>\n",
       "      <th>(0.8, 1.0]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   (0.0, 0.2]  (0.2, 0.4]  (0.4, 0.6]  (0.6, 0.8]  (0.8, 1.0]\n",
       "0           0           0           0           0           1\n",
       "1           0           1           0           0           0\n",
       "2           1           0           0           0           0\n",
       "3           0           1           0           0           0\n",
       "4           0           0           1           0           0\n",
       "5           0           0           1           0           0\n",
       "6           0           0           0           0           1\n",
       "7           0           0           0           1           0\n",
       "8           0           0           0           1           0\n",
       "9           0           0           0           1           0"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bins = [0, 0.2, 0.4, 0.6, 0.8, 1]\n",
    "pd.get_dummies(pd.cut(values, bins))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## String Manipulation\n",
    "#### Python is a popular raw data manipulation language due to its ease of use for string and text processing.\n",
    "#### Simple text operations can be done using String object's built-in methods.\n",
    "#### For more complex pattern matching and text manipulations, we can use regular expressions.\n",
    "#### pandas enable us to apply both string and regex functions on whole arrays of data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### String Object Methods\n",
    "#### In most string manipulation scenarios, inbuilt string methods are mostly sufficient.\n",
    "#### A string can be broken based on a seperator using 'split'.\n",
    "#### It is often combined with 'strip' to trim out whitespace, including line breaks.\n",
    "#### Substrings can be concatenated together using the '+' operator.\n",
    "#### A more faster and pythonic way to do so is to pass a list or tuple of substrings to the 'join' method on the 'stitching' string."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['a', 'b', '   guido']"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val = 'a,b,   guido'\n",
    "\n",
    "val.split(',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['a', 'b', 'guido']"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pieces = [x.strip() for x in val.split(',')]\n",
    "pieces"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'a::b::guido'"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first, second, third = pieces\n",
    "first + '::' + second + '::' + third"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'a::b::guido'"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'::'.join(pieces)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The 'in' keyword is the best way to detect substring. Although, 'index' and 'find' can also be used.\n",
    "#### There is 1 major difference between 'find' and 'index'. 'index' returns an Exception if substring is not found. 'find' returns a -1.\n",
    "#### 'count' returns number of occurences of a particular substring.\n",
    "#### 'replace' substitutes occurence of one pattern for another. It is commonly used to delete patterns by passing an empty string as replacement."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'guido' in val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val.index(',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val.find(':')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "substring not found",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-86-2c016e7367ac>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mval\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m':'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;31mValueError\u001b[0m: substring not found"
     ]
    }
   ],
   "source": [
    "# val.index(':')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val.count(',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'a::b::   guido'"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val.replace(',', '::')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'ab   guido'"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val.replace(',', '')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Regular Expression\n",
    "#### It is a flexible way to search or match (often complex) string patterns in text.\n",
    "#### The expression 'regex' is a string formed by 'regular expression'.\n",
    "#### Python has built-in 're' module for applying regular expressions to strings.\n",
    "#### This module works in 3 categories, all are related : 'pattern matching', 'substitution' and 'splitting'.\n",
    "#### We first describe a pattern that we want to locate in text. This can be used for many purposes.    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Once we define a regex, we can use it to split a text using 'split'.\n",
    "#### In this command, the regex is first compiled and then its split method is called on to the passed text.\n",
    "#### We can compile regex on our own using 're.compile' which will form a reusable regex object.\n",
    "#### Compiling regex beforehand is highly recommended if we have to apply it to many strings. This will save CPU cycles.\n",
    "#### NOTE - To avoid using escape sequences in regex, use raw string literals. eg - r'C:\\x' instead of 'c:\\\\x'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['foo', 'bar', 'baz', 'qux']"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import re\n",
    "\n",
    "text = \"foo bar\\t baz   \\tqux\"\n",
    "re.split('\\s+', text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['foo', 'bar', 'baz', 'qux']"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regex = re.compile('\\s+')\n",
    "regex.split(text)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We can use the 'findall' function to list all the matching patterns to the regex.\n",
    "#### To get only the first match in a string, we can use 'search' method.\n",
    "#### 'match' only matches at the beginning of the string i.e. it checks if the string startes with the regex.\n",
    "#### Using 'IGNORECASE'  makes the regex case-insensitive."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[' ', '\\t ', '   \\t']"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regex.findall(text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "text = \"\"\"Dave dave@google.com\n",
    "Steve steve@gmail.com\n",
    "Rob rob@gmail.com\n",
    "Ryan ryan@yahoo.com\n",
    "\"\"\"\n",
    "\n",
    "pattern = r'[A-Z0-9._%+-]+@[A-Z0-9.-]+\\.[A-Z]{2,4}'\n",
    "regex = re.compile(pattern, flags=re.IGNORECASE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 'findall' returns a list of matching pattern.\n",
    "#### 'search' returns a special match object for the first matching pattern. It can only tell us the start and end positions of the pattern in a string.\n",
    "#### 'match' will return None if the pattern does not exist in the string.\n",
    "#### 'sub' will return a new string with occurences of the pattern replaced by a new pattern."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['dave@google.com', 'steve@gmail.com', 'rob@gmail.com', 'ryan@yahoo.com']"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regex.findall(text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<_sre.SRE_Match object; span=(5, 20), match='dave@google.com'>"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m = regex.search(text)\n",
    "m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'dave@google.com'"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "text[m.start():m.end()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "None\n"
     ]
    }
   ],
   "source": [
    "print(regex.match(text))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dave REDACTED\n",
      "Steve REDACTED\n",
      "Rob REDACTED\n",
      "Ryan REDACTED\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(regex.sub('REDACTED', text))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### To find the patterns and at the same time segment it into different parts, put parentheses around the parts of the pattern to segment.\n",
    "#### In this case, 'findall' will return a list of tuples, when the pattern has groups.\n",
    "#### 'match' will return a tuple of the pattern components with its 'groups' method.\n",
    "#### 'sub' can also be used to access the groups in each match with special symbols '\\1', '\\2'. Here '\\1' corresponds to first matched group."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "pattern = r'([A-Z0-9._%+-]+)@([A-Z0-9.-]+)\\.([A-Z]{2,4})'\n",
    "regex = re.compile(pattern, flags=re.IGNORECASE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('dave', 'google', 'com'),\n",
       " ('steve', 'gmail', 'com'),\n",
       " ('rob', 'gmail', 'com'),\n",
       " ('ryan', 'yahoo', 'com')]"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regex.findall(text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('wes', 'bright', 'net')"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m = regex.match('wes@bright.net')\n",
    "m.groups()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dave Username: dave, Domain: google, Suffix: com\n",
      "Steve Username: steve, Domain: gmail, Suffix: com\n",
      "Rob Username: rob, Domain: gmail, Suffix: com\n",
      "Ryan Username: ryan, Domain: yahoo, Suffix: com\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(regex.sub(r'Username: \\1, Domain: \\2, Suffix: \\3', text))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Vectorized String Functions in pandas\n",
    "#### The presence of missing data in columns makes data cleaning in datasets very difficult.\n",
    "#### We could apply string or regex methods using lambda or other functions to each value using data.map, but this will fail on the NA values.\n",
    "#### To avoid this Series has array-oriented methods for string operations that skip NA values.\n",
    "#### They can be accessed using the 'str' attribute."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dave     dave@google.com\n",
       "Steve    steve@gmail.com\n",
       "Rob        rob@gmail.com\n",
       "Wes                  NaN\n",
       "dtype: object"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = {'Dave': 'dave@google.com', 'Steve': 'steve@gmail.com',\n",
    "        'Rob': 'rob@gmail.com', 'Wes': np.nan}\n",
    "data = pd.Series(data)\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dave     False\n",
       "Steve    False\n",
       "Rob      False\n",
       "Wes       True\n",
       "dtype: bool"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isnull()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We can check if a string contains a pattern using the 'str.contains' function.\n",
    "#### We can use regex too, with options like 'IGNORECASE'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dave     False\n",
       "Steve     True\n",
       "Rob       True\n",
       "Wes        NaN\n",
       "dtype: object"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.str.contains('gmail')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'([A-Z0-9._%+-]+)@([A-Z0-9.-]+)\\\\.([A-Z]{2,4})'"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pattern"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dave     [(dave, google, com)]\n",
       "Steve    [(steve, gmail, com)]\n",
       "Rob        [(rob, gmail, com)]\n",
       "Wes                        NaN\n",
       "dtype: object"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.str.findall(pattern, flags=re.IGNORECASE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### To retrieve elements from Vector, we can either use 'str.get' or 'index'.\n",
    "#### To access elements in embedded list, we pass an index to either of these functions.\n",
    "#### We can also slice strings with this syntax."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dave     True\n",
       "Steve    True\n",
       "Rob      True\n",
       "Wes       NaN\n",
       "dtype: object"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches = data.str.match(pattern, flags=re.IGNORECASE)\n",
    "matches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dave    NaN\n",
       "Steve   NaN\n",
       "Rob     NaN\n",
       "Wes     NaN\n",
       "dtype: float64"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches.str.get(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dave    NaN\n",
       "Steve   NaN\n",
       "Rob     NaN\n",
       "Wes     NaN\n",
       "dtype: float64"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches.str[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dave     dave@\n",
       "Steve    steve\n",
       "Rob      rob@g\n",
       "Wes        NaN\n",
       "dtype: object"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.str[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}