02_pandas.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Course Announcements\n",
    "\n",
    "- **Week 1**: weekly quz due *this* Friday (1/17, 11:59 PM) \n",
    "- **A1** due *next* Friday (1/24, 11:59 PM)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Data in Python\n",
    "\n",
    "\n",
    "- data files in Python\n",
    "    - semi-structured files\n",
    "    - Web Scraping & APIs\n",
    "- wrangling data \n",
    "    - `pandas`\n",
    "    - common manipulations\n",
    "- Where to find data?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "# Import standard libraries\n",
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Data 'Friendliness'\n",
    "\n",
    "The degree to which a data filetype easily lends itself to useful analysis."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 'Friendly' File Types:\n",
    "\n",
    "- csv\n",
    "- tsv\n",
    "- json\n",
    "- txt\n",
    "- xml"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 'Unfriendly' File Types:\n",
    "- pdf\n",
    "- docx\n",
    "- html\n",
    "- Anything made to look nice for humans"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### CSV Files\n",
    "\n",
    "- 'Comma Separated Value' files store data, separated by comma's. \n",
    "- Think of them like lists."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1, 2, 3, 4\r\n",
      "5, 6, 7, 8\r\n",
      "9, 10, 11, 12"
     ]
    }
   ],
   "source": [
    "# Note: through this notebook, I will be using '!' to run the shell command 'cat'\n",
    "#  to print out the content of example data files\n",
    "\n",
    "!cat data/dat.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Python has a module devoted to working with csv's\n",
    "import csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "# We can read through our file with the csv module\n",
    "with open('data/dat.csv') as csvfile:\n",
    "    csv_reader = csv.reader(csvfile, delimiter=',')\n",
    "    for row in csv_reader:\n",
    "        print(', '.join(row))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Pandas also has functions to directly load csv data\n",
    "pd.read_csv?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   0   1   2   3\n",
       "0  1   2   3   4\n",
       "1  5   6   7   8\n",
       "2  9  10  11  12"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's read in our csv file\n",
    "pd.read_csv('data/dat.csv', header=None) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## iclicker Question #1\n",
    "\n",
    "What does `pd` in `pd.read_csv()` specify?\n",
    "\n",
    "- A) it's the name of the function\n",
    "- B) that the `read_csv` method is from the pd package\n",
    "- C) that the `read_csv` method is from the pandas package (we're using the shortcut `pd`)\n",
    "- D) to read a csv file into python\n",
    "- E) I'm super lost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### JSON\n",
    "\n",
    "- JavaScript Object Notation files can store hierachical key/value pairings. \n",
    "- Think of them like dictionaries."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\r\n",
      "  \"firstName\": \"John\",\r\n",
      "  \"age\": 53\r\n",
      "}\r\n"
     ]
    }
   ],
   "source": [
    "!cat data/dat.json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dict'> \n",
      " {'firstName': 'John', 'age': '53'}\n"
     ]
    }
   ],
   "source": [
    "# Think of json's as similar to dictionaries\n",
    "d = {'firstName': 'John', 'age': '53'}\n",
    "print(type(d),'\\n',d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Python also has a module for dealing with json\n",
    "import json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "# Load a json file\n",
    "with open('data/dat.json') as dat_file:    \n",
    "    dat = json.load(dat_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dict'>\n"
     ]
    }
   ],
   "source": [
    "# Check what data type this gets loaded as\n",
    "print(type(dat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Pandas also has support for reading in json files\n",
    "pd.read_json?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "first          Alan\n",
       "place    Manchester\n",
       "dtype: object"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# You can read in json formatted strings with pandas\n",
    "pd.read_json('{ \"first\": \"Alan\", \"place\": \"Manchester\"}', typ = 'series')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "firstName    John\n",
       "age            53\n",
       "dtype: object"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read in our json file with pandas\n",
    "pd.read_json('data/dat.json', typ = 'series')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### XML\n",
    "\n",
    "- eXtensible Markup Language files store 'tagged' data. \n",
    "- Think of them like HTML."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<person>\r\n",
      "\t<who>Claude</who>\r\n",
      "\t<what>Info</who>\r\n",
      "\t<when>50s</when>\r\n",
      "</person>"
     ]
    }
   ],
   "source": [
    "!cat data/dat.xml"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# We can read in the XML file with standard python I/O\n",
    "with open('data/dat.xml') as dat_file:\n",
    "    dat = dat_file.read()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'<person>\\n\\t<who>Claude</who>\\n\\t<what>Info</who>\\n\\t<when>50s</when>\\n</person>'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check out the data\n",
    "dat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Beautiful Soup has functions to 'clean up' XML into human-friendlier formats\n",
    "from bs4 import BeautifulSoup\n",
    "nice_dat = BeautifulSoup(dat, 'xml')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n",
      "<person>\n",
      "<who>Claude</who>\n",
      "<what>Info</what>\n",
      "<when>50s</when>\n",
      "</person>\n"
     ]
    }
   ],
   "source": [
    "# Check out the parsed data\n",
    "print(nice_dat)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<center>\n",
    "<img src=\"img/pandas.png\" alt=\"pandas\" width=\"600px\">\n",
    "</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "Pandas is Python library for managing heterogenous data.\n",
    "\n",
    "At it's core, Pandas is used for the **DataFrame** object, which is:\n",
    "- a data structure for labeled rows and columns of data\n",
    "- associated methods and utilities for working with data.\n",
    "- each column contains a `pandas` **Series**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Loading Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "# Load a csv file of data\n",
    "df = pd.read_csv('data/my_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>id</th>\n",
       "      <th>first_name</th>\n",
       "      <th>last_name</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>295</td>\n",
       "      <td>Andrea</td>\n",
       "      <td>Clark</td>\n",
       "      <td>46</td>\n",
       "      <td>-1</td>\n",
       "      <td>24547.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>620</td>\n",
       "      <td>Bill</td>\n",
       "      <td>Woods</td>\n",
       "      <td>46</td>\n",
       "      <td>492</td>\n",
       "      <td>46713.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>891</td>\n",
       "      <td>Alexander</td>\n",
       "      <td>Jacobson</td>\n",
       "      <td>48</td>\n",
       "      <td>489</td>\n",
       "      <td>32071.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>914</td>\n",
       "      <td>Derrick</td>\n",
       "      <td>Bradley</td>\n",
       "      <td>52</td>\n",
       "      <td>-1</td>\n",
       "      <td>30650.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1736</td>\n",
       "      <td>Allison</td>\n",
       "      <td>Thomas</td>\n",
       "      <td>44</td>\n",
       "      <td>-1</td>\n",
       "      <td>9553.12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     id first_name last_name  age  score     value\n",
       "0   295     Andrea     Clark   46     -1  24547.87\n",
       "1   620       Bill     Woods   46    492  46713.90\n",
       "2   891  Alexander  Jacobson   48    489  32071.74\n",
       "3   914    Derrick   Bradley   52     -1  30650.48\n",
       "4  1736    Allison    Thomas   44     -1   9553.12"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check out a few rows of the dataframe\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "Pandas DataFrame:\n",
    "- Index for each row\n",
    "- Column name for each column\n",
    "- Stores heterogenous types"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Indexing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0          Clark\n",
       "1          Woods\n",
       "2       Jacobson\n",
       "3        Bradley\n",
       "4         Thomas\n",
       "5       Williams\n",
       "6           Wood\n",
       "7         Garcia\n",
       "8        Coleman\n",
       "9         Atkins\n",
       "10       Johnson\n",
       "11        Barnes\n",
       "12        Prince\n",
       "13       Mcmahon\n",
       "14         Jones\n",
       "15         Weber\n",
       "16         Clark\n",
       "17          Wood\n",
       "18       Sanchez\n",
       "19         Adams\n",
       "20      Williams\n",
       "21       Blevins\n",
       "22           Key\n",
       "23        Palmer\n",
       "24         Jones\n",
       "25        Ibarra\n",
       "26       Walters\n",
       "27       Simpson\n",
       "28      Santiago\n",
       "29          Mann\n",
       "         ...    \n",
       "170        Dixon\n",
       "171       Miller\n",
       "172     Matthews\n",
       "173       Brooks\n",
       "174    Wilkinson\n",
       "175        Sharp\n",
       "176     Thompson\n",
       "177        Woods\n",
       "178       Morgan\n",
       "179      Stewart\n",
       "180      Gardner\n",
       "181         Webb\n",
       "182       Martin\n",
       "183       Brooks\n",
       "184       Garcia\n",
       "185        Marks\n",
       "186       Fields\n",
       "187       Adkins\n",
       "188        Reyes\n",
       "189      Andrews\n",
       "190    Dominguez\n",
       "191       Oliver\n",
       "192      Bentley\n",
       "193     Phillips\n",
       "194       Austin\n",
       "195        Ortiz\n",
       "196     Chambers\n",
       "197        Pitts\n",
       "198      Jenkins\n",
       "199        Brown\n",
       "Name: last_name, Length: 200, dtype: object"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Indexing: select a column using its name\n",
    "df['last_name']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pandas.core.series.Series"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(df['last_name'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "500"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Indexing: select a row & column with 'loc'\n",
    "df.loc[10, 'score']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## iclicker Question #2\n",
    "\n",
    "What would be the output of `df['age'] > 10`?\n",
    "\n",
    "- A) subset of `df` including only rows of individuals older than 10\n",
    "- B) a Boolean with `True` for rows where age is greater than 10 and `False` otherwise\n",
    "- C) `id`s of rows where observations are greater than 10 \n",
    "- D) an error\n",
    "- E) I'm super lost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      True\n",
       "1      True\n",
       "2      True\n",
       "3      True\n",
       "4      True\n",
       "5      True\n",
       "6      True\n",
       "7      True\n",
       "8      True\n",
       "9      True\n",
       "10     True\n",
       "11     True\n",
       "12     True\n",
       "13     True\n",
       "14     True\n",
       "15     True\n",
       "16     True\n",
       "17     True\n",
       "18     True\n",
       "19     True\n",
       "20     True\n",
       "21     True\n",
       "22     True\n",
       "23     True\n",
       "24     True\n",
       "25     True\n",
       "26     True\n",
       "27     True\n",
       "28     True\n",
       "29     True\n",
       "       ... \n",
       "170    True\n",
       "171    True\n",
       "172    True\n",
       "173    True\n",
       "174    True\n",
       "175    True\n",
       "176    True\n",
       "177    True\n",
       "178    True\n",
       "179    True\n",
       "180    True\n",
       "181    True\n",
       "182    True\n",
       "183    True\n",
       "184    True\n",
       "185    True\n",
       "186    True\n",
       "187    True\n",
       "188    True\n",
       "189    True\n",
       "190    True\n",
       "191    True\n",
       "192    True\n",
       "193    True\n",
       "194    True\n",
       "195    True\n",
       "196    True\n",
       "197    True\n",
       "198    True\n",
       "199    True\n",
       "Name: age, Length: 200, dtype: bool"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## YOUR CODE HERE\n",
    "df['age'] > 10\n",
    "\n",
    "# to get dataframe \n",
    "# df[df['age'] > 10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Checking out the DataFrame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "scrolled": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(200, 6)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check how large our dataframe is\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['id', 'first_name', 'last_name', 'age', 'score', 'value'], dtype='object')"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check what columns we have in our DataFrame\n",
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "id             object\n",
       "first_name     object\n",
       "last_name      object\n",
       "age             int64\n",
       "score           int64\n",
       "value         float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the datatypes of our variables\n",
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>id</th>\n",
       "      <th>first_name</th>\n",
       "      <th>last_name</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>295</td>\n",
       "      <td>Andrea</td>\n",
       "      <td>Clark</td>\n",
       "      <td>46</td>\n",
       "      <td>-1</td>\n",
       "      <td>24547.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>620</th>\n",
       "      <td>620</td>\n",
       "      <td>Bill</td>\n",
       "      <td>Woods</td>\n",
       "      <td>46</td>\n",
       "      <td>492</td>\n",
       "      <td>46713.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>891</th>\n",
       "      <td>891</td>\n",
       "      <td>Alexander</td>\n",
       "      <td>Jacobson</td>\n",
       "      <td>48</td>\n",
       "      <td>489</td>\n",
       "      <td>32071.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>914</th>\n",
       "      <td>914</td>\n",
       "      <td>Derrick</td>\n",
       "      <td>Bradley</td>\n",
       "      <td>52</td>\n",
       "      <td>-1</td>\n",
       "      <td>30650.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1736</th>\n",
       "      <td>1736</td>\n",
       "      <td>Allison</td>\n",
       "      <td>Thomas</td>\n",
       "      <td>44</td>\n",
       "      <td>-1</td>\n",
       "      <td>9553.12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id first_name last_name  age  score     value\n",
       "id                                                   \n",
       "295    295     Andrea     Clark   46     -1  24547.87\n",
       "620    620       Bill     Woods   46    492  46713.90\n",
       "891    891  Alexander  Jacobson   48    489  32071.74\n",
       "914    914    Derrick   Bradley   52     -1  30650.48\n",
       "1736  1736    Allison    Thomas   44     -1   9553.12"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Set the index to a string (non-numerical) and use it as index (row labels)\n",
    "df['id'] = df['id'].astype('str')\n",
    "df = df.set_index(df['id'])\n",
    "df.head() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Exploring the data\n",
    "\n",
    "- quantitative (numbers)\n",
    "- qualitative (categorical)\n",
    "- basic descriptive statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "David        6\n",
       "Michael      5\n",
       "Charles      4\n",
       "James        4\n",
       "Eric         4\n",
       "John         3\n",
       "Ashley       3\n",
       "Brian        3\n",
       "Jason        3\n",
       "Elizabeth    3\n",
       "Name: first_name, dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking categorical data\n",
    "df['first_name'].value_counts()[0:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "28730.336296296293"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check a particular descriptive statistic\n",
    "df['value'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    200.000000\n",
       "mean     416.595000\n",
       "std      237.176674\n",
       "min       -1.000000\n",
       "25%      288.750000\n",
       "50%      463.500000\n",
       "75%      596.500000\n",
       "max      942.000000\n",
       "Name: score, dtype: float64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Describe a particular column\n",
    "df['score'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>200.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>189.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>46.020000</td>\n",
       "      <td>416.595000</td>\n",
       "      <td>28730.336296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>10.028582</td>\n",
       "      <td>237.176674</td>\n",
       "      <td>32493.945741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>14.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>39.000000</td>\n",
       "      <td>288.750000</td>\n",
       "      <td>9593.030000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>46.000000</td>\n",
       "      <td>463.500000</td>\n",
       "      <td>17976.510000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>53.000000</td>\n",
       "      <td>596.500000</td>\n",
       "      <td>33163.310000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>69.000000</td>\n",
       "      <td>942.000000</td>\n",
       "      <td>204999.960000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              age       score          value\n",
       "count  200.000000  200.000000     189.000000\n",
       "mean    46.020000  416.595000   28730.336296\n",
       "std     10.028582  237.176674   32493.945741\n",
       "min     14.000000   -1.000000       0.000000\n",
       "25%     39.000000  288.750000    9593.030000\n",
       "50%     46.000000  463.500000   17976.510000\n",
       "75%     53.000000  596.500000   33163.310000\n",
       "max     69.000000  942.000000  204999.960000"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get descriptive statistics of all numerical columns\n",
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "## iclicker Question #3\n",
    "\n",
    "What's the average (mean) age of the individuals in this dataset?\n",
    "\n",
    "- A) 14\n",
    "- B) 46\n",
    "- C) 28730\n",
    "- D) NA\n",
    "- E) I'm super lost/unsure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "46.02"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## YOUR CODE HERE\n",
    "df['age'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "nbgrader": {
     "grade": false,
     "grade_id": "cell-31cc3c7a4ac258dc",
     "locked": true,
     "schema_version": 1,
     "solution": false
    },
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## `pandas`: Common Manipulations\n",
    "\n",
    "You'll want to be *very* familiar with a few common data manipulations when wrangling data, each of which is described below:\n",
    "\n",
    "Manipulation | Description\n",
    "-------|------------\n",
    "**select** | select which columns to include in dataset\n",
    "**filter** | filter dataset to only include specified rows\n",
    "**mutate** | add a new column based on values in other columns\n",
    "**groupby** | group values to apply a function within the specified groups\n",
    "**summarize** | calculate specified summary metric of a specified variable\n",
    "**arrange** | sort rows ascending or descending order of a specified column\n",
    "**merge** | join separate datasets into a single dataset based on a common column\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Selecting & Dropping Columns\n",
    "\n",
    "- include subset of columns of larger data frame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>id</th>\n",
       "      <th>first_name</th>\n",
       "      <th>last_name</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>295</td>\n",
       "      <td>Andrea</td>\n",
       "      <td>Clark</td>\n",
       "      <td>46</td>\n",
       "      <td>-1</td>\n",
       "      <td>24547.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>620</th>\n",
       "      <td>620</td>\n",
       "      <td>Bill</td>\n",
       "      <td>Woods</td>\n",
       "      <td>46</td>\n",
       "      <td>492</td>\n",
       "      <td>46713.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>891</th>\n",
       "      <td>891</td>\n",
       "      <td>Alexander</td>\n",
       "      <td>Jacobson</td>\n",
       "      <td>48</td>\n",
       "      <td>489</td>\n",
       "      <td>32071.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>914</th>\n",
       "      <td>914</td>\n",
       "      <td>Derrick</td>\n",
       "      <td>Bradley</td>\n",
       "      <td>52</td>\n",
       "      <td>-1</td>\n",
       "      <td>30650.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1736</th>\n",
       "      <td>1736</td>\n",
       "      <td>Allison</td>\n",
       "      <td>Thomas</td>\n",
       "      <td>44</td>\n",
       "      <td>-1</td>\n",
       "      <td>9553.12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id first_name last_name  age  score     value\n",
       "id                                                   \n",
       "295    295     Andrea     Clark   46     -1  24547.87\n",
       "620    620       Bill     Woods   46    492  46713.90\n",
       "891    891  Alexander  Jacobson   48    489  32071.74\n",
       "914    914    Derrick   Bradley   52     -1  30650.48\n",
       "1736  1736    Allison    Thomas   44     -1   9553.12"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>295</td>\n",
       "      <td>46</td>\n",
       "      <td>-1</td>\n",
       "      <td>24547.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>620</th>\n",
       "      <td>620</td>\n",
       "      <td>46</td>\n",
       "      <td>492</td>\n",
       "      <td>46713.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>891</th>\n",
       "      <td>891</td>\n",
       "      <td>48</td>\n",
       "      <td>489</td>\n",
       "      <td>32071.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>914</th>\n",
       "      <td>914</td>\n",
       "      <td>52</td>\n",
       "      <td>-1</td>\n",
       "      <td>30650.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1736</th>\n",
       "      <td>1736</td>\n",
       "      <td>44</td>\n",
       "      <td>-1</td>\n",
       "      <td>9553.12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  age  score     value\n",
       "id                              \n",
       "295    295   46     -1  24547.87\n",
       "620    620   46    492  46713.90\n",
       "891    891   48    489  32071.74\n",
       "914    914   52     -1  30650.48\n",
       "1736  1736   44     -1   9553.12"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# specify which columns to include\n",
    "select_df = df[['id', 'age', 'score', 'value']]\n",
    "select_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "# Drop columns we don't want\n",
    "df = df.drop(labels=['first_name', 'last_name'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>295</td>\n",
       "      <td>46</td>\n",
       "      <td>-1</td>\n",
       "      <td>24547.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>620</th>\n",
       "      <td>620</td>\n",
       "      <td>46</td>\n",
       "      <td>492</td>\n",
       "      <td>46713.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>891</th>\n",
       "      <td>891</td>\n",
       "      <td>48</td>\n",
       "      <td>489</td>\n",
       "      <td>32071.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>914</th>\n",
       "      <td>914</td>\n",
       "      <td>52</td>\n",
       "      <td>-1</td>\n",
       "      <td>30650.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1736</th>\n",
       "      <td>1736</td>\n",
       "      <td>44</td>\n",
       "      <td>-1</td>\n",
       "      <td>9553.12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  age  score     value\n",
       "id                              \n",
       "295    295   46     -1  24547.87\n",
       "620    620   46    492  46713.90\n",
       "891    891   48    489  32071.74\n",
       "914    914   52     -1  30650.48\n",
       "1736  1736   44     -1   9553.12"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check out the DataFrame after dropping some columns\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Filtering Data (slicing)\n",
    "\n",
    "- include a subset (slice) of rows from larger data frame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check if we have any data from people below the age of 18\n",
    "sum(df['age'] < 18)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(199, 4)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Select only participants who are 18 or older\n",
    "df = df[df['age'] >= 18]\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Missing Data (NaNs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check for missing values\n",
    "df['value'].hasnans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "11"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check for null values\n",
    "sum(df['value'].isnull())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10753</th>\n",
       "      <td>10753</td>\n",
       "      <td>58</td>\n",
       "      <td>335</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16263</th>\n",
       "      <td>16263</td>\n",
       "      <td>42</td>\n",
       "      <td>323</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24952</th>\n",
       "      <td>24952</td>\n",
       "      <td>51</td>\n",
       "      <td>399</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26091</th>\n",
       "      <td>26091</td>\n",
       "      <td>44</td>\n",
       "      <td>620</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29566</th>\n",
       "      <td>29566</td>\n",
       "      <td>69</td>\n",
       "      <td>224</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35466</th>\n",
       "      <td>35466</td>\n",
       "      <td>55</td>\n",
       "      <td>483</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42138</th>\n",
       "      <td>42138</td>\n",
       "      <td>28</td>\n",
       "      <td>543</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51723</th>\n",
       "      <td>51723</td>\n",
       "      <td>40</td>\n",
       "      <td>443</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55821</th>\n",
       "      <td>55821</td>\n",
       "      <td>41</td>\n",
       "      <td>709</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61534</th>\n",
       "      <td>61534</td>\n",
       "      <td>39</td>\n",
       "      <td>598</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94730</th>\n",
       "      <td>94730</td>\n",
       "      <td>22</td>\n",
       "      <td>-1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          id  age  score  value\n",
       "id                             \n",
       "10753  10753   58    335    NaN\n",
       "16263  16263   42    323    NaN\n",
       "24952  24952   51    399    NaN\n",
       "26091  26091   44    620    NaN\n",
       "29566  29566   69    224    NaN\n",
       "35466  35466   55    483    NaN\n",
       "42138  42138   28    543    NaN\n",
       "51723  51723   40    443    NaN\n",
       "55821  55821   41    709    NaN\n",
       "61534  61534   39    598    NaN\n",
       "94730  94730   22     -1    NaN"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Have a look at the missing values\n",
    "df[df['value'].isnull()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Dealing with Missing Data - NaNs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(188, 4)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dealing with null values: Drop rows with missing data\n",
    "df = df.dropna()\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Finding Missing Data - Bad Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    188.000000\n",
       "mean     414.329787\n",
       "std      238.923293\n",
       "min       -1.000000\n",
       "25%      287.000000\n",
       "50%      463.500000\n",
       "75%      589.250000\n",
       "max      942.000000\n",
       "Name: score, dtype: float64"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check for the properties of specific columns\n",
    "df['score'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check the plot of the data for score to see the distribution\n",
    "df['score'].plot('hist', bins=25);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Dealing with Missing Data - Bad Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "32"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Look for how many values have a -1 value in 'score'\n",
    "sum(df['score'] == -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(156, 4)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Drop any row with -1 value in 'score'\n",
    "df = df[df['score'] != -1]\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Creating new columns (mutating)\n",
    "\n",
    "- `assign` can be very helpful in adding a new column\n",
    "- lambda functions can be used to carry out calculations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "      <th>age_days</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>620</th>\n",
       "      <td>620</td>\n",
       "      <td>46</td>\n",
       "      <td>492</td>\n",
       "      <td>46713.90</td>\n",
       "      <td>16790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>891</th>\n",
       "      <td>891</td>\n",
       "      <td>48</td>\n",
       "      <td>489</td>\n",
       "      <td>32071.74</td>\n",
       "      <td>17520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2049</th>\n",
       "      <td>2049</td>\n",
       "      <td>57</td>\n",
       "      <td>333</td>\n",
       "      <td>138936.92</td>\n",
       "      <td>20805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2607</th>\n",
       "      <td>2607</td>\n",
       "      <td>50</td>\n",
       "      <td>536</td>\n",
       "      <td>9367.27</td>\n",
       "      <td>18250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2635</th>\n",
       "      <td>2635</td>\n",
       "      <td>68</td>\n",
       "      <td>351</td>\n",
       "      <td>66035.28</td>\n",
       "      <td>24820</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  age  score      value  age_days\n",
       "id                                         \n",
       "620    620   46    492   46713.90     16790\n",
       "891    891   48    489   32071.74     17520\n",
       "2049  2049   57    333  138936.92     20805\n",
       "2607  2607   50    536    9367.27     18250\n",
       "2635  2635   68    351   66035.28     24820"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# convert age in years to age in (approximate) days\n",
    "df = df.assign(age_days = df['age'] * 365)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Grouping & summarizing\n",
    "\n",
    "- group by a particular variable\n",
    "- calculate summary statistics/metrics within group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "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>score</th>\n",
       "      <th>value</th>\n",
       "      <th>age_days</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>775.000000</td>\n",
       "      <td>9116.030000</td>\n",
       "      <td>8395.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>674.500000</td>\n",
       "      <td>10920.255000</td>\n",
       "      <td>9855.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>820.000000</td>\n",
       "      <td>16190.663333</td>\n",
       "      <td>10220.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>539.000000</td>\n",
       "      <td>5366.800000</td>\n",
       "      <td>10585.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>335.666667</td>\n",
       "      <td>5485.036667</td>\n",
       "      <td>11315.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>570.000000</td>\n",
       "      <td>11230.650000</td>\n",
       "      <td>11680.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>750.000000</td>\n",
       "      <td>6984.920000</td>\n",
       "      <td>12045.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>621.500000</td>\n",
       "      <td>20480.336667</td>\n",
       "      <td>12410.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>614.666667</td>\n",
       "      <td>27419.950000</td>\n",
       "      <td>12775.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>715.333333</td>\n",
       "      <td>15719.360000</td>\n",
       "      <td>13140.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>477.000000</td>\n",
       "      <td>12581.357500</td>\n",
       "      <td>13505.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>488.000000</td>\n",
       "      <td>12339.096667</td>\n",
       "      <td>13870.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>456.833333</td>\n",
       "      <td>11370.056667</td>\n",
       "      <td>14235.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>458.500000</td>\n",
       "      <td>12317.655000</td>\n",
       "      <td>14600.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>504.250000</td>\n",
       "      <td>14609.110000</td>\n",
       "      <td>14965.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>597.400000</td>\n",
       "      <td>58733.974000</td>\n",
       "      <td>15330.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>480.600000</td>\n",
       "      <td>9479.648000</td>\n",
       "      <td>15695.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>457.625000</td>\n",
       "      <td>12186.182500</td>\n",
       "      <td>16060.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>494.000000</td>\n",
       "      <td>12808.595000</td>\n",
       "      <td>16425.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>494.428571</td>\n",
       "      <td>23698.015714</td>\n",
       "      <td>16790.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>531.571429</td>\n",
       "      <td>39429.118571</td>\n",
       "      <td>17155.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>531.000000</td>\n",
       "      <td>34962.375000</td>\n",
       "      <td>17520.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>514.600000</td>\n",
       "      <td>18375.030000</td>\n",
       "      <td>17885.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>512.250000</td>\n",
       "      <td>17888.720000</td>\n",
       "      <td>18250.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>464.000000</td>\n",
       "      <td>19805.317500</td>\n",
       "      <td>18615.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>501.000000</td>\n",
       "      <td>49976.079000</td>\n",
       "      <td>18980.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>523.111111</td>\n",
       "      <td>30803.248889</td>\n",
       "      <td>19345.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>472.200000</td>\n",
       "      <td>62484.416000</td>\n",
       "      <td>19710.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>149.000000</td>\n",
       "      <td>12632.320000</td>\n",
       "      <td>20075.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>531.250000</td>\n",
       "      <td>66800.067500</td>\n",
       "      <td>20440.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>313.833333</td>\n",
       "      <td>44684.031667</td>\n",
       "      <td>20805.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>400.500000</td>\n",
       "      <td>22387.640000</td>\n",
       "      <td>21170.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>500.000000</td>\n",
       "      <td>34938.080000</td>\n",
       "      <td>21535.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>300.500000</td>\n",
       "      <td>38482.447500</td>\n",
       "      <td>21900.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>497.000000</td>\n",
       "      <td>34068.430000</td>\n",
       "      <td>22265.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>622.000000</td>\n",
       "      <td>105449.110000</td>\n",
       "      <td>22630.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>347.000000</td>\n",
       "      <td>105598.790000</td>\n",
       "      <td>22995.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>297.333333</td>\n",
       "      <td>79593.543333</td>\n",
       "      <td>23360.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>347.000000</td>\n",
       "      <td>48514.625000</td>\n",
       "      <td>24820.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>343.000000</td>\n",
       "      <td>60605.080000</td>\n",
       "      <td>25185.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          score          value  age_days\n",
       "age                                     \n",
       "23   775.000000    9116.030000    8395.0\n",
       "27   674.500000   10920.255000    9855.0\n",
       "28   820.000000   16190.663333   10220.0\n",
       "29   539.000000    5366.800000   10585.0\n",
       "31   335.666667    5485.036667   11315.0\n",
       "32   570.000000   11230.650000   11680.0\n",
       "33   750.000000    6984.920000   12045.0\n",
       "34   621.500000   20480.336667   12410.0\n",
       "35   614.666667   27419.950000   12775.0\n",
       "36   715.333333   15719.360000   13140.0\n",
       "37   477.000000   12581.357500   13505.0\n",
       "38   488.000000   12339.096667   13870.0\n",
       "39   456.833333   11370.056667   14235.0\n",
       "40   458.500000   12317.655000   14600.0\n",
       "41   504.250000   14609.110000   14965.0\n",
       "42   597.400000   58733.974000   15330.0\n",
       "43   480.600000    9479.648000   15695.0\n",
       "44   457.625000   12186.182500   16060.0\n",
       "45   494.000000   12808.595000   16425.0\n",
       "46   494.428571   23698.015714   16790.0\n",
       "47   531.571429   39429.118571   17155.0\n",
       "48   531.000000   34962.375000   17520.0\n",
       "49   514.600000   18375.030000   17885.0\n",
       "50   512.250000   17888.720000   18250.0\n",
       "51   464.000000   19805.317500   18615.0\n",
       "52   501.000000   49976.079000   18980.0\n",
       "53   523.111111   30803.248889   19345.0\n",
       "54   472.200000   62484.416000   19710.0\n",
       "55   149.000000   12632.320000   20075.0\n",
       "56   531.250000   66800.067500   20440.0\n",
       "57   313.833333   44684.031667   20805.0\n",
       "58   400.500000   22387.640000   21170.0\n",
       "59   500.000000   34938.080000   21535.0\n",
       "60   300.500000   38482.447500   21900.0\n",
       "61   497.000000   34068.430000   22265.0\n",
       "62   622.000000  105449.110000   22630.0\n",
       "63   347.000000  105598.790000   22995.0\n",
       "64   297.333333   79593.543333   23360.0\n",
       "68   347.000000   48514.625000   24820.0\n",
       "69   343.000000   60605.080000   25185.0"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# caclculate average within each age\n",
    "df.groupby('age').agg('mean')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Sorting Rows (arrange)\n",
    "\n",
    "- specify order in which to display rows"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "scrolled": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "      <th>age_days</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>620</th>\n",
       "      <td>620</td>\n",
       "      <td>46</td>\n",
       "      <td>492</td>\n",
       "      <td>46713.90</td>\n",
       "      <td>16790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>891</th>\n",
       "      <td>891</td>\n",
       "      <td>48</td>\n",
       "      <td>489</td>\n",
       "      <td>32071.74</td>\n",
       "      <td>17520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2049</th>\n",
       "      <td>2049</td>\n",
       "      <td>57</td>\n",
       "      <td>333</td>\n",
       "      <td>138936.92</td>\n",
       "      <td>20805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2607</th>\n",
       "      <td>2607</td>\n",
       "      <td>50</td>\n",
       "      <td>536</td>\n",
       "      <td>9367.27</td>\n",
       "      <td>18250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2635</th>\n",
       "      <td>2635</td>\n",
       "      <td>68</td>\n",
       "      <td>351</td>\n",
       "      <td>66035.28</td>\n",
       "      <td>24820</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  age  score      value  age_days\n",
       "id                                         \n",
       "620    620   46    492   46713.90     16790\n",
       "891    891   48    489   32071.74     17520\n",
       "2049  2049   57    333  138936.92     20805\n",
       "2607  2607   50    536    9367.27     18250\n",
       "2635  2635   68    351   66035.28     24820"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "scrolled": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>score</th>\n",
       "      <th>value</th>\n",
       "      <th>age_days</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>81282</th>\n",
       "      <td>81282</td>\n",
       "      <td>23</td>\n",
       "      <td>775</td>\n",
       "      <td>9116.03</td>\n",
       "      <td>8395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70735</th>\n",
       "      <td>70735</td>\n",
       "      <td>27</td>\n",
       "      <td>665</td>\n",
       "      <td>3975.59</td>\n",
       "      <td>9855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15887</th>\n",
       "      <td>15887</td>\n",
       "      <td>27</td>\n",
       "      <td>684</td>\n",
       "      <td>17864.92</td>\n",
       "      <td>9855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96101</th>\n",
       "      <td>96101</td>\n",
       "      <td>28</td>\n",
       "      <td>918</td>\n",
       "      <td>18303.20</td>\n",
       "      <td>10220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60177</th>\n",
       "      <td>60177</td>\n",
       "      <td>28</td>\n",
       "      <td>748</td>\n",
       "      <td>25035.69</td>\n",
       "      <td>10220</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          id  age  score     value  age_days\n",
       "id                                          \n",
       "81282  81282   23    775   9116.03      8395\n",
       "70735  70735   27    665   3975.59      9855\n",
       "15887  15887   27    684  17864.92      9855\n",
       "96101  96101   28    918  18303.20     10220\n",
       "60177  60177   28    748  25035.69     10220"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sort by values in age\n",
    "df = df.sort_values(by = ['age'])\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Combining datasets\n",
    "![](img/join.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "## Create two DataFrames\n",
    "left = pd.DataFrame({'key': ['A', 'B', 'C', 'D'], 'value': np.random.randn(4)})    \n",
    "right = pd.DataFrame({'key': ['B', 'D', 'E', 'F'], 'value': np.random.randn(4)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>key</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>-0.035785</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>B</td>\n",
       "      <td>0.222361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C</td>\n",
       "      <td>-0.760123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>D</td>\n",
       "      <td>-0.048016</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key     value\n",
       "0   A -0.035785\n",
       "1   B  0.222361\n",
       "2   C -0.760123\n",
       "3   D -0.048016"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "left"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>key</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>B</td>\n",
       "      <td>0.713750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>D</td>\n",
       "      <td>1.268733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>E</td>\n",
       "      <td>0.652800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>F</td>\n",
       "      <td>1.110714</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key     value\n",
       "0   B  0.713750\n",
       "1   D  1.268733\n",
       "2   E  0.652800\n",
       "3   F  1.110714"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "right"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "pd.merge? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "scrolled": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>key</th>\n",
       "      <th>value_x</th>\n",
       "      <th>value_y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>B</td>\n",
       "      <td>0.222361</td>\n",
       "      <td>0.713750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>D</td>\n",
       "      <td>-0.048016</td>\n",
       "      <td>1.268733</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key   value_x   value_y\n",
       "0   B  0.222361  0.713750\n",
       "1   D -0.048016  1.268733"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# inner merge by default\n",
    "pd.merge(left, right, on='key')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>key</th>\n",
       "      <th>value_x</th>\n",
       "      <th>value_y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>B</td>\n",
       "      <td>0.222361</td>\n",
       "      <td>0.713750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>D</td>\n",
       "      <td>-0.048016</td>\n",
       "      <td>1.268733</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key   value_x   value_y\n",
       "0   B  0.222361  0.713750\n",
       "1   D -0.048016  1.268733"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# same as above\n",
    "pd.merge(left, right, on='key', how='inner')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>key</th>\n",
       "      <th>value_x</th>\n",
       "      <th>value_y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>B</td>\n",
       "      <td>0.222361</td>\n",
       "      <td>0.713750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>D</td>\n",
       "      <td>-0.048016</td>\n",
       "      <td>1.268733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>E</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.652800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>F</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.110714</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key   value_x   value_y\n",
       "0   B  0.222361  0.713750\n",
       "1   D -0.048016  1.268733\n",
       "2   E       NaN  0.652800\n",
       "3   F       NaN  1.110714"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# right merge\n",
    "pd.merge(left, right, on='key', how='right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>key</th>\n",
       "      <th>value_x</th>\n",
       "      <th>value_y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>-0.035785</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>B</td>\n",
       "      <td>0.222361</td>\n",
       "      <td>0.713750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C</td>\n",
       "      <td>-0.760123</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>D</td>\n",
       "      <td>-0.048016</td>\n",
       "      <td>1.268733</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key   value_x   value_y\n",
       "0   A -0.035785       NaN\n",
       "1   B  0.222361  0.713750\n",
       "2   C -0.760123       NaN\n",
       "3   D -0.048016  1.268733"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# right merge\n",
    "pd.merge(left, right, on='key', how='left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "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>key</th>\n",
       "      <th>value_x</th>\n",
       "      <th>value_y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>-0.035785</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>B</td>\n",
       "      <td>0.222361</td>\n",
       "      <td>0.713750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C</td>\n",
       "      <td>-0.760123</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>D</td>\n",
       "      <td>-0.048016</td>\n",
       "      <td>1.268733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>E</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.652800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>F</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.110714</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key   value_x   value_y\n",
       "0   A -0.035785       NaN\n",
       "1   B  0.222361  0.713750\n",
       "2   C -0.760123       NaN\n",
       "3   D -0.048016  1.268733\n",
       "4   E       NaN  0.652800\n",
       "5   F       NaN  1.110714"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# outer join\n",
    "pd.merge(left, right, on='key', how='outer')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## iclicker Question #4\n",
    "\n",
    "If table A had 5 rows and table B had 5 rows and 3 of those rows in each table were from the same observations present in the *other* table, how many rows would be present if an **inner merge** were carried out?\n",
    "\n",
    "- A) 3\n",
    "- B) 5\n",
    "- C) 10\n",
    "- D) 13\n",
    "- E) Totally unsure"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## iclicker Question #5\n",
    "\n",
    "If table A had 5 rows and table B had 5 rows and 3 of those rows in each table were from the same observations present in the *other* table, how many rows would be present if a **left merge** were carried out?\n",
    "\n",
    "- A) 3\n",
    "- B) 5\n",
    "- C) 10\n",
    "- D) 13\n",
    "- E) Totally unsure"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Visualizing Data\n",
    "\n",
    "- We'll have a whole lecture (or two) on visualization\n",
    "- For now, we'll just look at one uniquely-pandas approach"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot all numerical columns, and their interactions\n",
    "pd.plotting.scatter_matrix(df[['age', 'score', 'value']], figsize=[12, 12], marker=12);  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Check for correlations between variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "# Take the correlations between all numerical columns\n",
    "df.corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Application Program Interface (APIs)\n",
    "\n",
    "- APIs are basically a way for software to talk to software \n",
    "    - They are an interface into an application / website / database designed for computers / software."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Notes on APIs:\n",
    "- Follow API guidelines! \n",
    "- These guidelines typically specify the number / rate / size of requests"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Github API\n",
    "\n",
    "You can access the github api with the following API. Just added specifiers for what you are looking for. \n",
    "\n",
    "https://api.github.com/\n",
    "\n",
    "For example, the following URL will search for the user 'ShanEllis'\n",
    "\n",
    "https://api.github.com/users/shanellis\n",
    "\n",
    "<center>\n",
    "<img src=\"img/github.png\" alt=\"sql\" height=\"100\" width=\"100\">\n",
    "</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Requesting Web Pages from Python"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "# The requests module allows you to send URL requests from python\n",
    "import requests  \n",
    "from bs4 import BeautifulSoup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "# Request data from the Github API on a particular user\n",
    "page = requests.get('https://api.github.com/users/shanellis')  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "b'{\"login\":\"ShanEllis\",\"id\":6606571,\"node_id\":\"MDQ6VXNlcjY2MDY1NzE=\",\"avatar_url\":\"https://avatars3.githubusercontent.com/u/6606571?v=4\",\"gravatar_id\":\"\",\"url\":\"https://api.github.com/users/ShanEllis\",\"html_url\":\"https://github.com/ShanEllis\",\"followers_url\":\"https://api.github.com/users/ShanEllis/followers\",\"following_url\":\"https://api.github.com/users/ShanEllis/following{/other_user}\",\"gists_url\":\"https://api.github.com/users/ShanEllis/gists{/gist_id}\",\"starred_url\":\"https://api.github.com/users/ShanEllis/starred{/owner}{/repo}\",\"subscriptions_url\":\"https://api.github.com/users/ShanEllis/subscriptions\",\"organizations_url\":\"https://api.github.com/users/ShanEllis/orgs\",\"repos_url\":\"https://api.github.com/users/ShanEllis/repos\",\"events_url\":\"https://api.github.com/users/ShanEllis/events{/privacy}\",\"received_events_url\":\"https://api.github.com/users/ShanEllis/received_events\",\"type\":\"User\",\"site_admin\":false,\"name\":\"Shannon Ellis\",\"company\":null,\"blog\":\"shanellis.com\",\"location\":\"San Diego, CA\",\"email\":null,\"hireable\":null,\"bio\":null,\"public_repos\":37,\"public_gists\":1,\"followers\":37,\"following\":0,\"created_at\":\"2014-02-06T16:09:58Z\",\"updated_at\":\"2020-01-23T15:25:59Z\"}'"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The content we get back is a messily organized json file\n",
    "page.content"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "## iclicker Question #6\n",
    "\n",
    "What type/format of output is this?\n",
    "\n",
    "- A) CSV\n",
    "- B) XML\n",
    "- C) JSON\n",
    "- D) API\n",
    "- E) I'm super lost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "# We can read in the json data with pandas\n",
    "import pandas as pd\n",
    "git_data = pd.read_json(page.content, typ='series')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "login                                                          ShanEllis\n",
       "id                                                               6606571\n",
       "node_id                                             MDQ6VXNlcjY2MDY1NzE=\n",
       "avatar_url             https://avatars3.githubusercontent.com/u/66065...\n",
       "gravatar_id                                                             \n",
       "url                               https://api.github.com/users/ShanEllis\n",
       "html_url                                    https://github.com/ShanEllis\n",
       "followers_url           https://api.github.com/users/ShanEllis/followers\n",
       "following_url          https://api.github.com/users/ShanEllis/followi...\n",
       "gists_url              https://api.github.com/users/ShanEllis/gists{/...\n",
       "starred_url            https://api.github.com/users/ShanEllis/starred...\n",
       "subscriptions_url      https://api.github.com/users/ShanEllis/subscri...\n",
       "organizations_url            https://api.github.com/users/ShanEllis/orgs\n",
       "repos_url                   https://api.github.com/users/ShanEllis/repos\n",
       "events_url             https://api.github.com/users/ShanEllis/events{...\n",
       "received_events_url    https://api.github.com/users/ShanEllis/receive...\n",
       "type                                                                User\n",
       "site_admin                                                         False\n",
       "name                                                       Shannon Ellis\n",
       "company                                                             None\n",
       "blog                                                       shanellis.com\n",
       "location                                                   San Diego, CA\n",
       "email                                                               None\n",
       "hireable                                                            None\n",
       "bio                                                                 None\n",
       "public_repos                                                          37\n",
       "public_gists                                                           1\n",
       "followers                                                             37\n",
       "following                                                              0\n",
       "created_at                                          2014-02-06T16:09:58Z\n",
       "updated_at                                          2020-01-23T15:25:59Z\n",
       "dtype: object"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check out the pandas series object full of data\n",
    "git_data  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Authorized Access - OAuth\n",
    "\n",
    "Open Authorization is a protocol to authorize access (of a user / application) to an API."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "OAuth provides a secure way to 'log-in' without using account names and passwords. \n",
    "\n",
    "It is effectively a set of keys, and passwords you can use to access APIs. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Web Scraping vs. APIs\n",
    "\n",
    "Web scraping and APIs are different approaches:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "- APIs are an interface to interact with an application, designed for programmatic use\n",
    "    - They allow systematic, controlled access to (for example) and applications database\n",
    "    - They typically return structured (friendly) data "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "- Web scraping (typically) involves navigating through the internet, programmatically following an architecture built for humans\n",
    "    - This can be hard to systematize, being dependent on the idiosyncracies of a web page, at the time you request it\n",
    "    - This typically returns relatively unstructured data\n",
    "    - This entails much more wrangling of the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Where to Find Data?\n",
    "\n",
    "* [Awesome Public Datasets](https://github.com/awesomedata/awesome-public-datasets/blob/master/README.rst)\n",
    "* [Data.gov](https://catalog.data.gov/dataset)\n",
    "* [Data Is Plural](https://docs.google.com/spreadsheets/d/1wZhPLMCHKJvwOkP4juclhjFgqIY8fQFMemwKL2c64vk/edit#gid=0)\n",
    "* [UCSD Datasets](https://ucsd.libguides.com/data-statistics/home)\n",
    "* [Datasets | Deep Learning](http://deeplearning.net/datasets/)\n",
    "* [Stanford | Social Science Data Collection](https://data.stanford.edu/)\n",
    "* [Eviction Lab (email required)](https://evictionlab.org/get-the-data/)\n",
    "* [San Diego Data](https://data.sandiego.gov/)\n",
    "* [US Census](https://www.census.gov/)\n",
    "* [Open Climate Data](http://openclimatedata.net/)\n",
    "* [Data and Story Library](https://dasl.datadescription.com/datafiles/)\n",
    "* [UCSD behavioral mobile data](http://extrasensory.ucsd.edu/)\n",
    "* [Kaggle](https://www.kaggle.com/)\n",
    "* [FiveThirtyEight](https://data.fivethirtyeight.com/)\n",
    "* [data.world](https://data.world/)\n",
    "* [Free Datasets - R and Data Mining ](http://www.rdatamining.com/resources/data)\n",
    "* [Data Sources for Cool Data Science Projects](https://blog.thedataincubator.com/2014/10/data-sources-for-cool-data-science-projects-part-1/)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Notes on Working with Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Data Science is Ad-Hoc\n",
    "\n",
    "- It is part of the job description to put things together that were not designed to go together.\n",
    "- We do not have universal solutions, but haphazard, idiosyncratic systems, for data collection, storage and analysis.\n",
    "- Data is everywhere. But relatively little of it was collected *as data*."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "notes"
    }
   },
   "source": [
    "### Data Collection, Curation, and Storage are Difficult\n",
    "\n",
    "- It can be difficult to choose broadly useful standards\n",
    "- Take time to think about your data, and how you will load, store, organize and save it"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "notes"
    }
   },
   "source": [
    "### Data is Inherently Noisy\n",
    "\n",
    "- We live in a messy, noisy, world, with messy, noisy, people, using messy, noisy instruments.\n",
    "- There is no perfect data. \n",
    "    - There is better / or worse data, given the context."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "notes"
    }
   },
   "source": [
    "### Different Objectives\n",
    "\n",
    "- Humans and computers are different.\n",
    "- We interact with '*data*' in different ways.\n",
    "- This underlies many aspects of data wrangling\n",
    "    - The 'friendliness' of data types / files\n",
    "    - The difference between web scraping and APIs\n",
    "    - A disconnect between data in the real world, and data we want to use"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## So... What to do?\n",
    "\n",
    "- Think about how your data are stored & its structure?\n",
    "- Look at your data before you anayze it\n",
    "    - are there missing values? \n",
    "    - outlier values? \n",
    "- Are your data trustworthy? \n",
    "    - source?\n",
    "    - how was it generated?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Specific Recommendations\n",
    "\n",
    "- Prioritize using well structured, common, open file types\n",
    "    - Take advantage of existing tools to deal with these files (numpy, pandas, etc.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "- Look into, and then follow, common conventions\n",
    "    - Minimize custom objects, workflows and data files \n",
    "- Look for APIs. Ask if they are available.\n",
    "    - Acknowledge that web scraping and/or wrangling unstructured data are complex / long tasks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "- Think about data flow from the beginning. Organize your data pipeline, consider the 'wrangling' aspects throughout\n",
    "    - Set yourself up with well organized, labelled approach to your data\n",
    "    - Think about when and how you might want/need to save out intermediate results."
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "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": "3.6.7"
  },
  "rise": {
   "scroll": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}