1_EDA.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# You have some data... now what?\n",
    "\n",
    "Usually, you want to spend some time **exploring the data**. This can include a bunch of things:\n",
    "- **Sanity checks**: does the data include the columns you're expecting? How are rows unique? Are there duplicates?\n",
    "- **Hunting for patterns and outliers**: Are there any? What do they look like? Does anything seem correlated?\n",
    "- **Some descriptive statistics** (mean, standard deviation, histograms)\n",
    "And just generally getting to know your data!\n",
    "\n",
    "You can do a lot of this with `pandas`, but pictures contain $\\infty$ words, so now's the time to introduce our visualization tools!\n",
    "\n",
    "## Next tool: `matplotlib`\n",
    "\n",
    "![matplotlib logo](imgs/matplotlib_logo.png)\n",
    "\n",
    "[`matplotlib`](https://matplotlib.org/) is one of the most popular plotting libraries in the Python data science stack. You may find people using [`seaborn`](https://seaborn.pydata.org/) or [`bokeh`](https://bokeh.pydata.org/en/latest/), and definitely explore those and other options, but `matplotlib` is the \"default\" plotting library. It's very powerful and definitely worth investing time in.\n",
    "\n",
    "However, `matplotlib`'s documentation is... not super friendly. (Unless you have experience [MATLAB](https://www.mathworks.com/products/matlab.html), in which case it may be familiar to you.) You can get a handle on it by keeping the following in mind:\n",
    "- The basic call for a `matplotlib` chart is `plt.plot(x, y)`.\n",
    "    - `.plot()`  has lots of additional, handy keyword arguments, like `color='red'`, `alpha=0.2` (20% transparency), and `label=\"Tuition\"`.\n",
    "- `plt.plot()` doesn't always display the plot, though. Instead, it instantiates a matplotlib pyplot object.\n",
    "- You can explicitly display the object with `plt.show()`.\n",
    "- You can modify the object's attributes - e.g. the range of the x- and y-scales, the title, the x- and y-axis labels. For example, to change the title, you can do `plt.plot(x, y)`, then `plt.title(\"My cool chart title\")` and then `plt.show()`!\n",
    "\n",
    "Here's a good [walkthrough of the object-ness of `matplotlib` charts, and how to use it well](http://pbpython.com/effective-matplotlib.html), and [another one](https://realpython.com/blog/python/python-matplotlib-guide/). No need to go through these now, though! You can get pretty far with just the tips above!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Tell Jupyter to display matplotlib plots in your notebook\n",
    "%matplotlib inline\n",
    "\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Let's resize all our charts to be a bit bigger\n",
    "import matplotlib as mpl\n",
    "mpl.rcParams['figure.figsize'] = (12, 8)\n",
    "\n",
    "# matplotlib has certain default styles that you can specify at the top of your notebook\n",
    "# https://matplotlib.org/users/style_sheets.html\n",
    "# Here, I use the 'bmh' style - just a personal preference!\n",
    "plt.style.use('bmh')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Back to the Greeks\n",
    "\n",
    "Let's explore the Duke Greek life dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1739, 16)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Greek Council</th>\n",
       "      <th>Greek Organization</th>\n",
       "      <th>Home City</th>\n",
       "      <th>Home State</th>\n",
       "      <th>Home Country</th>\n",
       "      <th>Latitude of Home City</th>\n",
       "      <th>Longitude of Home City</th>\n",
       "      <th>Free and reduced percentage of  High School</th>\n",
       "      <th>Public or Private High School</th>\n",
       "      <th>Domestic or International High School</th>\n",
       "      <th>Boarding/Day High School</th>\n",
       "      <th>Religous High School?</th>\n",
       "      <th>Gender of High School's Students</th>\n",
       "      <th>Tuition of High School</th>\n",
       "      <th>Sports Team</th>\n",
       "      <th>Merit Scholarship</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Accokeek</td>\n",
       "      <td>Maryland</td>\n",
       "      <td>United States</td>\n",
       "      <td>38.667617</td>\n",
       "      <td>-77.028309</td>\n",
       "      <td>Data not available</td>\n",
       "      <td>private</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Episcopal</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>36610</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Advance</td>\n",
       "      <td>North Carolina</td>\n",
       "      <td>United States</td>\n",
       "      <td>35.941248</td>\n",
       "      <td>-80.409222</td>\n",
       "      <td>0.34741523</td>\n",
       "      <td>public</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>0</td>\n",
       "      <td>Fooball</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Ajman</td>\n",
       "      <td>NaN</td>\n",
       "      <td>United Arab Emirates</td>\n",
       "      <td>25.509129</td>\n",
       "      <td>55.361907</td>\n",
       "      <td>NaN</td>\n",
       "      <td>private</td>\n",
       "      <td>International</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>13093.46</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Alameda</td>\n",
       "      <td>California</td>\n",
       "      <td>United States</td>\n",
       "      <td>37.609029</td>\n",
       "      <td>-121.899142</td>\n",
       "      <td>0.2268</td>\n",
       "      <td>public</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Albuquerque</td>\n",
       "      <td>New Mexico</td>\n",
       "      <td>United States</td>\n",
       "      <td>35.084103</td>\n",
       "      <td>-106.650985</td>\n",
       "      <td>Data not available</td>\n",
       "      <td>private</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>22870</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Greek Council Greek Organization    Home City      Home State  \\\n",
       "0          None               None     Accokeek        Maryland   \n",
       "1          None               None      Advance  North Carolina   \n",
       "2          None               None        Ajman             NaN   \n",
       "3          None               None      Alameda      California   \n",
       "4          None               None  Albuquerque      New Mexico   \n",
       "\n",
       "           Home Country  Latitude of Home City  Longitude of Home City  \\\n",
       "0         United States              38.667617              -77.028309   \n",
       "1         United States              35.941248              -80.409222   \n",
       "2  United Arab Emirates              25.509129               55.361907   \n",
       "3         United States              37.609029             -121.899142   \n",
       "4         United States              35.084103             -106.650985   \n",
       "\n",
       "  Free and reduced percentage of  High School Public or Private High School  \\\n",
       "0                          Data not available                       private   \n",
       "1                                  0.34741523                        public   \n",
       "2                                         NaN                       private   \n",
       "3                                      0.2268                        public   \n",
       "4                          Data not available                       private   \n",
       "\n",
       "  Domestic or International High School Boarding/Day High School  \\\n",
       "0                              Domestic                      Day   \n",
       "1                              Domestic                      Day   \n",
       "2                         International                      Day   \n",
       "3                              Domestic                      Day   \n",
       "4                              Domestic                      Day   \n",
       "\n",
       "  Religous High School? Gender of High School's Students  \\\n",
       "0             Episcopal                            Co-Ed   \n",
       "1          Unafilliated                            Co-Ed   \n",
       "2          Unafilliated                            Co-Ed   \n",
       "3          Unafilliated                            Co-Ed   \n",
       "4          Unafilliated                            Co-Ed   \n",
       "\n",
       "  Tuition of High School Sports Team Merit Scholarship  \n",
       "0                  36610        None              None  \n",
       "1                      0     Fooball              None  \n",
       "2               13093.46        None              None  \n",
       "3                      0        None              None  \n",
       "4                  22870        None              None  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Loading the Greek life data\n",
    "df = pd.read_csv('data/duke-greek-life.csv')\n",
    "print(df.shape)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " I don't really know anything about Duke or its Greek life, so I have a bunch of questions:\n",
    "- How many students are there? (I'm assuming the level of observation is a student.)\n",
    "- How many students are in a Greek organization?\n",
    "- What's the distribution of high school tuitions? (Do Duke students go to fancy high schools?)\n",
    "- How many had a merit scholarship?\n",
    "- Where are students from?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1739"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many students are there?\n",
    "len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "None                   1150\n",
       "Gamma Phi Beta           39\n",
       "Kappa Alpha Theta        37\n",
       "Alpha Delta Pi           34\n",
       "Zeta Tau Alpha           33\n",
       "Pi Beta Phi              33\n",
       "Delta Delta Delta        32\n",
       "Kappa Kappa Gamma        32\n",
       "Alpha Phi                30\n",
       "Delta Sigma Phi          29\n",
       "Sigma Phi Epsilon        27\n",
       "Pi Kappa Phi             26\n",
       "Alpha Tau Omega          25\n",
       "Delta Gamma              25\n",
       "Chi Omega                25\n",
       "Delta Tau Delta          24\n",
       "Pi Kappa Alpha           21\n",
       "Kappa Alpha Order        20\n",
       "Alpha Epsilon Pi         20\n",
       "Sigma Nu                 17\n",
       "Sigma Alpha Epsilon      16\n",
       "Kappa Sigma              11\n",
       "Sigma Chi                11\n",
       "Chi Psi                  10\n",
       "Alpha Delta Phi           6\n",
       "Delta Kappa Epsilon       3\n",
       "Psi Upsilon               3\n",
       "Name: Greek Organization, dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many students are in a Greek organization?\n",
    "df['Greek Organization'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "26"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many Greek orgs are there?\n",
    "df['Greek Organization'].nunique() - 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.33870040253018974"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# What percentage of students are in a Greek org?\n",
    "len(df[df['Greek Organization'] != \"None\"]) / len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count     1695\n",
       "unique     394\n",
       "top          0\n",
       "freq      1059\n",
       "Name: Tuition of High School, dtype: object"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# What's the distribution of high school tuitions?\n",
    "df['Tuition of High School'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ah ha! Our first data cleaning\n",
    "\n",
    "High school tuition sounds like it should be a **numeric** column, but the `dtype` above is `object`. (That is, `pandas` thinks that column is a string.) Why? \n",
    "\n",
    "Let's try to make it numeric with `pandas`'s method, `to_numeric()`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Unable to parse string \"16,600\" at position 451",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32mpandas/_libs/src/inference.pyx\u001b[0m in \u001b[0;36mpandas._libs.lib.maybe_convert_numeric\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Unable to parse string \"16,600\"",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-7-a925eb2ccede>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto_numeric\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'Tuition of High School'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/.pyenv/versions/3.6.1/envs/pyladies/lib/python3.6/site-packages/pandas/core/tools/numeric.py\u001b[0m in \u001b[0;36mto_numeric\u001b[0;34m(arg, errors, downcast)\u001b[0m\n\u001b[1;32m    131\u001b[0m             \u001b[0mcoerce_numeric\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0merrors\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m'ignore'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'raise'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    132\u001b[0m             values = lib.maybe_convert_numeric(values, set(),\n\u001b[0;32m--> 133\u001b[0;31m                                                coerce_numeric=coerce_numeric)\n\u001b[0m\u001b[1;32m    134\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    135\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/src/inference.pyx\u001b[0m in \u001b[0;36mpandas._libs.lib.maybe_convert_numeric\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Unable to parse string \"16,600\" at position 451"
     ]
    }
   ],
   "source": [
    "pd.to_numeric(df['Tuition of High School'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The comma in \"16,600\" is the problem. Let's try to correct it and see if that helps."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Unable to parse string \"1,094.79\" at position 829",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32mpandas/_libs/src/inference.pyx\u001b[0m in \u001b[0;36mpandas._libs.lib.maybe_convert_numeric\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Unable to parse string \"1,094.79\"",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-8-9828410c5d6f>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'Tuition of High School'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreplace\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mto_replace\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"16,600\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalue\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"16600\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minplace\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto_numeric\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'Tuition of High School'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/.pyenv/versions/3.6.1/envs/pyladies/lib/python3.6/site-packages/pandas/core/tools/numeric.py\u001b[0m in \u001b[0;36mto_numeric\u001b[0;34m(arg, errors, downcast)\u001b[0m\n\u001b[1;32m    131\u001b[0m             \u001b[0mcoerce_numeric\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0merrors\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m'ignore'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'raise'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    132\u001b[0m             values = lib.maybe_convert_numeric(values, set(),\n\u001b[0;32m--> 133\u001b[0;31m                                                coerce_numeric=coerce_numeric)\n\u001b[0m\u001b[1;32m    134\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    135\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/src/inference.pyx\u001b[0m in \u001b[0;36mpandas._libs.lib.maybe_convert_numeric\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Unable to parse string \"1,094.79\" at position 829"
     ]
    }
   ],
   "source": [
    "df['Tuition of High School'].replace(to_replace=\"16,600\", value=\"16600\", inplace=True)\n",
    "pd.to_numeric(df['Tuition of High School'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hmmm. Still not finished. Also, it's probably not efficient to manually correct for every comma in every column. We can use a shortcut - Python's string methods! These are also available to us in `pandas`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       36610.00\n",
       "1           0.00\n",
       "2       13093.46\n",
       "3           0.00\n",
       "4       22870.00\n",
       "5       22870.00\n",
       "6       22870.00\n",
       "7           0.00\n",
       "8           0.00\n",
       "9           0.00\n",
       "10          0.00\n",
       "11          0.00\n",
       "12          0.00\n",
       "13      14170.00\n",
       "14          0.00\n",
       "15          0.00\n",
       "16          0.00\n",
       "17          0.00\n",
       "18          0.00\n",
       "19      26670.00\n",
       "20          0.00\n",
       "21          0.00\n",
       "22          0.00\n",
       "23       5450.21\n",
       "24      20000.00\n",
       "25      20000.00\n",
       "26          0.00\n",
       "27          0.00\n",
       "28       9494.00\n",
       "29      15000.00\n",
       "          ...   \n",
       "1709        0.00\n",
       "1710    21518.00\n",
       "1711        0.00\n",
       "1712        0.00\n",
       "1713        0.00\n",
       "1714    21190.00\n",
       "1715        0.00\n",
       "1716        0.00\n",
       "1717        0.00\n",
       "1718        0.00\n",
       "1719        0.00\n",
       "1720    42830.00\n",
       "1721    21350.00\n",
       "1722     8300.00\n",
       "1723        0.00\n",
       "1724    13135.00\n",
       "1725     3165.54\n",
       "1726    36100.00\n",
       "1727         NaN\n",
       "1728         NaN\n",
       "1729     5508.00\n",
       "1730    23717.50\n",
       "1731         NaN\n",
       "1732        0.00\n",
       "1733        0.00\n",
       "1734     8950.00\n",
       "1735         NaN\n",
       "1736    58300.00\n",
       "1737        0.00\n",
       "1738        0.00\n",
       "Name: Tuition of High School, Length: 1739, dtype: float64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.to_numeric(df['Tuition of High School'].str.replace(',',''))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That worked! But we didn't actually change the column. Instead, `pandas` just returned a new Series for us (a Series is a column in a pandas DataFrame). Let's replace the column itself:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count      1695.000000\n",
       "mean      10555.198232\n",
       "std       16650.980438\n",
       "min           0.000000\n",
       "25%           0.000000\n",
       "50%           0.000000\n",
       "75%       20012.500000\n",
       "max      118660.780000\n",
       "Name: Tuition of High School, dtype: float64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Tuition of High School'] = pd.to_numeric(df['Tuition of High School'].str.replace(',',''))\n",
    "df['Tuition of High School'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect! Now we can use one of `pandas`'s built-in plotting methods (yes, it has a few! they're built on top of `matplotlib`) to see a frequency histogram of this data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1124a0f98>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAD5CAYAAADBX4k8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvFvnyVgAAIABJREFUeJztnX1wHGeZ4H+P5sMaS6PR15Tsi1N2AjZOcA7yUSQUd2wKseRjqQ3ngr2ktiCAU9wdhssuV8UmxR9wd0XdsmwdLHV7kDuS3XDFBlgISSoXSHKBsLdVjmHt+EJCSCyHDFFKkkdfo69R5kPv/TEtWTZSFI1a3a8eP7+qKfe83ep+fu6Zfqa736dfcc5hGIZhnJ+0xB2AYRiGER+WBAzDMM5jLAkYhmGcx1gSMAzDOI+xJGAYhnEeY0nAMAzjPGbNJCAi94jIaRF5dlnbl0Xk1yLyjIj8UEQ6l827U0QGROQFEbluWfv1QduAiNwRvophGIaxXt7ImcDfAtef0/Y4cMA598+BF4E7AUTkUuBm4K3B3/x3EUmISAL4a+AG4FLglmBZwzAMI0bWTALOuX8Axs9pe8w5VwvePgXsCqZvAr7jnHvNOfcbYAB4R/AacM695JyrAN8JljUMwzBiJIx7Ah8HfhRMXwC8smzeYNC2WrthGIYRI8mN/LGIfA6oAd8OJxz48Y9/7IaGhhARnHN0dXWRz+epVqskEgkA6vU6qVSKWq1xMpJMJpuaX61WERESiQS1Wo1EIoFzjoWFhaX5LS0ttLS0UKvVSCaTLCwsrHu+iFCv10kmk9TrdZxzS/PNyZzMyZzCcKpUKqP9/f35NQ6xv0PTSUBEPgq8H+h3Zx5A9Cpw4bLFdgVtvE77WeRyOa6++uqmYioUCuzevbupv/UNTS6gy8dc/EWTz3pdjh8/XmhmO01dDhKR64HPAn/onJtbNush4GYR2SYiFwF7gZ8DvwD2ishFIpKmcfP4oWa2/XqkUqmwVxkbmlxAl4+5+Ismn6hc1jwTEJH7gGuBXhEZBD5PozfQNuBxEQF4yjn3b51zz4nI94Bf0bhMdNg5Vw/W8yngUSAB3OOcey5smVwuF/YqY0OTC+jyMRd/0eQTlcuaScA5d8sKzXe/zvJfBL64QvsjwCPrim6djI6O0tbWtpmbiAxNLqDLx1z8RZNPVC6qKobtV4C/aPIxF3/R5BOVi6okUKlU4g4hNDS5gC4fc/EXTT5RuahKAuVyOe4QQkOTC+jyMRd/0eQTlYuqJLBjx464QwgNTS6gy8dc/EWTT1QuqpLA8PBw3CGEhiYX0OVjLv6iyScqlw1VDPvG138xwvHHx9decBN47LbLQ11fOp0OdX1xo8nHXPxFk09ULqrOBIbKenSy2WzcIYSKJh9z8RdNPlG56DlqAvuy9bhDCI2xsbG4QwgVTT7m4i+afKJyUZUETs0m4g4hNLq6uuIOIVQ0+ZiLv2jyicpFVRLoSS/EHUJoaOrqBrp8zMVfNPlYF9Em6Ey5tRfaIszPz8cdQqho8jEXf9HkE5WLqiRwbFJPZydN/Z1Bl4+5+IsmH6sTaIIrO2trL7RF0NTfGXT5mIu/aPKJykVVEpisStwhhEZra2vcIYSKJh9z8RdNPlG5qEoCYxU9OplMJu4QQkWTj7n4iyafqFz0HDWBN7XpqROYmJiIO4RQ0eRjLv6iyScqF1VJ4MVpPXUCPT09cYcQKpp8zMVfNPlE5aIqCezM6KkTmJ6ejjuEUNHkYy7+osknKhdVSSCb1FMnoGlwDNDlYy7+osnHBpVpAqsT8BdNPubiL5p8rE6gCaxOwF80+ZiLv2jysTqBJrAuov6iycdc/EWTj3URbYLpmp5iMU2DY4AuH3PxF00+NqhME+zZrqdOoFQqxR1CqGjyMRd/0eQTlYuqJPD8tJ4bw729vXGHECqafMzFXzT5ROWiKgnstjMBb9HkYy7+osnHzgSaYHtCT51AtVqNO4RQ0eRjLv6iyScqlzWTgIjcIyKnReTZZW3dIvK4iJwM/u0K2kVEviYiAyLyjIhcsexvbg2WPykit26GjNUJ+IsmH3PxF00+PtUJ/C1w/TltdwBPOOf2Ak8E7wFuAPYGr08AX4dG0gA+D1wNvAP4/GLiCBOrE/AXTT7m4i+afLypE3DO/QMwfk7zTcC9wfS9wAeWtX/LNXgK6BSRncB1wOPOuXHn3ATwOL+bWDbMyGt6rm61tbXFHUKoaPIxF3/R5BOVS7NHzT7n3FAwPQz0BdMXAK8sW24waFutPVSqep4fRyKh54mooMvHXPxFk09ULhu+iO6ccyIS2h3Z06dPc+jQIZLJJPV6nYMHD3L48GGGh4dpa2sjkUgwNTVFPp9nfHwc5xz5fJ6RkRHelquRENjZusCJUpLLOmrUnHByJsGBjhqD5RZSLdC3bYFjk0mu7KwxVxcKcwkuydZ4eS5BNunoSZ+ZP10Thsot7MvWOTWboCe9QGfKLc2frApjlRYKhQI9PT1MT09TqVTYsWMHw8PDZDIZ0uk0pVKJ3t5eSqUS1Wp1af5qTpVKhampKdrb2wGYmZmhr6+PYrGIiNDd3U2xWKSjo4N6vc7s7OzSOlOpFLlcjtHRUXK5HJVKhXK5vDQ/nU6TzWYZGxujq6uLcrnM/Pz80vzW1lYymQwTExOhOU1MTJDJZBgZGdnyTtVqldnZ2bM+e1vVaXR0dNXv01Z0mpubA1DhVK/XmZqaWvO4t+jULOLc2sdvEdkDPOycOxC8fwG41jk3FFzuedI59xYRuSuYvm/5cosv59y/CdrPWm45R44ccfv3729K5pZvHYvt0RGP3XZ5qOubm5tj+/btoa4zTjT5mIu/aPJZr8vx48eP9ff3X7Xe7TR7xHwIWOzhcyvw4LL2jwS9hK4BSsFlo0eB94lIV3BD+H1BW6jsbddTJzA+fu5tmK2NJh9z8RdNPlG5rHk5SETuo/FLvldEBmn08vlz4HsicggoAH8ULP4IcCMwAMwBHwNwzo2LyH8GfhEs95+cc6EbJsO7KhU7b+QMbSuhycdc/EWTT1QuayYB59wtq8zqX2FZBxxeZT33APesK7p18sspPXUC+Xw+7hBCRZOPufiLJp+oXPT0qQTentNTJzAyMhJ3CKGiycdc/EWTT1QuqpLA0LwenY3e8fcNTT7m4i+afKJy0XPUNAzDMNaNqiSws1VPtdjMzEzcIYSKJh9z8RdNPlG5qEoCJ0p6bgz39fWtvdAWQpOPufiLJp+oXFQlgcs69NwYLhaLcYcQKpp8zMVfNPlE5aIqCdScnjGGRfS4gC4fc/EXTT5RuahKAidn9Dw8qru7O+4QQkWTj7n4iyafqFxUJYEDdjnIWzT5mIu/aPKxy0FNMFjWo9PR0RF3CKGiycdc/EWTT1Queo6aQEqRTb2u52F4oMvHXPxFk09ULooOm41xArQwOzsbdwihosnHXPxFk09ULqqSgA007y+afMzFXzT5+DTQ/JbBBpr3F00+5uIvmny8GWh+KzFX19NHOJVKxR1CqGjyMRd/0eQTlYuqJFCY01MnkMvl4g4hVDT5mIu/aPKJykVVErgkq+dy0OjoaNwhhIomH3PxF00+UbmoSgIv25mAt2jyMRd/0eRjZwJNkE3qGV+0UqnEHUKoaPIxF3/R5BOVi6ok0JPWUydQLpfjDiFUNPmYi79o8onKRVUSsDoBf9HkYy7+osnH6gSawOoE/EWTj7n4iyYfqxNogumanjqBdDoddwihosnHXPxFk09ULqqSwJCip4hms9m4QwgVTT7m4i+afKJy0XPUBPZl9TxBcGxsLO4QQkWTj7n4iyafqFxUJYFTs3rqBLq6uuIOIVQ0+ZiLv2jyicpFVRKwLqL+osnHXPxFk491EW2CzpSeYrH5+fm4QwgVTT7m4i+afKJy2VASEJE/FZHnRORZEblPRFpF5CIROSoiAyLyXRFJB8tuC94PBPP3hCGwHKsT8BdNPubiL5p8vK8TEJELgH8PXOWcOwAkgJuBLwFfcc69GZgADgV/cgiYCNq/EiwXKlYn4C+afMzFXzT5bJU6gSSQEZEksB0YAt4DfD+Yfy/wgWD6puA9wfx+EQm1Y/9kVU+dQGtra9whhIomH3PxF00+Ubk0ff3EOfeqiPwl8FugDDwGHAMmnXOLP8kHgQuC6QuAV4K/rYlICegBznpe6unTpzl06BDJZJJ6vc7Bgwc5fPgww8PDtLW1kUgkmJqaIp/PMz4+jnOOfD7PyMgI2xOOfe01drYucKKU5LKOGjUnnJxJcKCjxmC5hVRLYyziY5NJruysMVcXCnMJLsnWeHkuQTbp6EmfmT9dE4bKLezL1jk1m6AnvUBnyi3Nn6wKY5UWCoUCPT09TE9PU6lU2LFjB8PDw2QyGdLpNKVSid7eXkqlEtVqdWn+ak6ZTIZCoUB7ezsAMzMz9PX1USwWERG6u7spFot0dHRQr9eZnZ1dWmcqlSKXyzE6Okoul6NSqVAul5fmp9NpstksY2NjdHV1US6XmZ+fX5rf2tpKJpNhYmIiNKf5+XlyuRwjIyNb3imTyTA4OHjWZ2+rOtVqNQqFworfp63o1NLSwsTExKrHiK3k1N7eTqFQWPO4t+jULOJcczdTRaQL+AHwr4FJ4O9p/ML/QnDJBxG5EPiRc+6AiDwLXO+cGwzmnQKuds6dlQSOHDni9u/f31RMX3zgKD8bjadi8LHbLg91fYVCgd27d4e6zjjR5GMu/qLJZ70ux48fP9bf33/VerezkctB7wV+45wrOueqwP3Au4DO4PIQwC7g1WD6VeBCgGB+Dgi1GuLFaT11Aj09PXGHECqafMzFXzT5ROWykSTwW+AaEdkeXNvvB34F/BT4YLDMrcCDwfRDwXuC+T9xzZ6GrMLOjJ46genp6bhDCBVNPubiL5p8onJpOgk4547SuPxzHPhlsK7/AfwZ8BkRGaBxzf/u4E/uBnqC9s8Ad2wg7hWxQWX8RZOPufiLJp+oXDbUsd4593ng8+c0vwS8Y4Vl54EPbWR7a2F1Av6iycdc/EWTj/d1Aj5idQL+osnHXPxFk89WqRPwirGKHp1MJhN3CKGiycdc/EWTT1Queo6a2KAyPqPJx1z8RZOPDSrTBHu26xlPoFQqxR1CqGjyMRd/0eQTlYuqJPD8tJ4bw729vXGHECqafMzFXzT5ROWiKgnstjMBb9HkYy7+osnHzgSaYHtCT51AtVqNO4RQ0eRjLv6iyScqF1VJwOoE/EWTj7n4iyYfqxNoAqsT8BdNPubiL5p8rE6gCUZe06PT1tYWdwihosnHXPxFk09ULnqOmkBVz/PjSCT0PBEVdPmYi79o8onKRVUS2KXoKaJTU1NxhxAqmnzMxV80+UTloioJPDul58ZwPp+PO4RQ0eRjLv6iyScqF1VJYG+7njqB8fHxuEMIFU0+5uIvmnyiclGVBJKip04g5PF2YkeTj7n4iyafqFxUJYFf2uUgb9HkYy7+osnHLgc1wdtzeuoERkZG4g4hVDT5mIu/aPKJykVVEhia16PT3t4edwihosnHXPxFk09ULnqOmoZhGMa6UZUEdrbqqROYmZmJO4RQ0eRjLv6iyScqF1VJ4ERJz43hvr6+uEMIFU0+5uIvmnyiclGVBC7r0HNjuFgsxh1CqGjyMRd/0eQTlYuqJFBzesYYFtHjArp8zMVfNPlE5aIqCZyc0fPwqO7u7rhDCBVNPubiL5p8onJRlQQO2OUgb9HkYy7+osnHLgc1wWBZj05HR0fcIYSKJh9z8RdNPlG56DlqAilFNvW6nofhgS4fc/EXTT5RuWzosCkinSLyfRH5tYg8LyLvFJFuEXlcRE4G/3YFy4qIfE1EBkTkGRG5IhyFM/Rt01MnMDs7G3cIoaLJx1z8RZNPVC4b/e38V8CPnXP7gbcBzwN3AE845/YCTwTvAW4A9gavTwBf3+C2fwcbaN5fNPmYi79o8vF+oHkRyQHvBu4GcM5VnHOTwE3AvcFi9wIfCKZvAr7lGjwFdIrIzqYjXwEbaN5fNPmYi79o8tkKA81fBBSBvxGRp0XkmyLSBvQ554aCZYaBxbK3C4BXlv39YNAWGnN1PX2EU6lU3CGEiiYfc/EXTT5RuWzk+kkSuAL4tHPuqIj8FWcu/QDgnHMi6xvp5fTp0xw6dIhkMkm9XufgwYMcPnyY4eFh2traSCQSTE1Nkc/nGR8fxzlHPp9vPHbVwb72GjtbFzhRSnJZR42aE07OJDjQUWOw3EKqpXHv4Nhkkis7a8zVhcJcgkuyNV6eS5BNOnrSZ+ZP14Shcgv7snVOzSboSS/QmXJL8yerwlilhUKhQE9PD9PT01QqFXbs2MHw8DCZTIZ0Ok2pVKK3t5dSqUS1Wl2av5pTW1sbhUJh6UmCMzMz9PX1USwWERG6u7spFot0dHRQr9eZnZ1dWmcqlSKXyzE6Okoul6NSqVAul5fmp9NpstksY2NjdHV1US6XmZ+fX5rf2tpKJpNhYmIiNKfXXnuN+fl5RkZGtrxTe3s7g4ODZ332tqqTiFAoFFb8Pm1Fp2QyycTExKrHiK3klMvlKBQKax73Nvq0UWl29BoR2QE85ZzbE7z/lzSSwJuBa51zQ8Hlniedc28RkbuC6fuC5V9YXG75eo8cOeL279/fVExffOAoPxtNN/W3G+Wx2y4PdX2FQoHdu3eHus440eRjLv6iyWe9LsePHz/W399/1Xq30/TlIOfcMPCKiLwlaOoHfgU8BNwatN0KPBhMPwR8JOgldA1QOjcBbJSX5/RUDOdyubhDCBVNPubiL5p8onLZaHeaTwPfFpE08BLwMRqJ5XsicggoAH8ULPsIcCMwAMwFy4ZKNqlnfNFKpRJ3CKGiycdc/EWTT1QuG0oCzrkTwEqnH/0rLOuAwxvZ3lr0pPXUCZTL5bhDCBVNPubiL5p8onJRVGNrdQI+o8nHXPxFk4/3dQI+YnUC/qLJx1z8RZPPVqgT8I7pmp46gXQ6nl5Om4UmH3PxF00+UbmoSgJDip4ims1m4w4hVDT5mIu/aPKJykXPURPYl9XzBMGxsbG4QwgVTT7m4i+afKJyUZUETs3qqRPo6uqKO4RQ0eRjLv6iyScqF1VJwLqI+osmH3PxF00+1kW0CTpTeorF5ufn4w4hVDT5mIu/aPKJykVVErA6AX/R5GMu/qLJx+oEmsDqBPxFk4+5+IsmH6sTaILJqp46gdbW1rhDCBVNPubiL5p8onJRlQTGKnp0MplM3CGEiiYfc/EXTT5Rueg5agJvatNTJzAxMRF3CKGiycdc/EWTT1QuqpLAi9N66gR6enriDiFUNPmYi79o8onKRVUS2JnRUycwPT0ddwihosnHXPxFk09ULqqSgA0q4y+afMzFXzT5ROWiKglYnYC/aPIxF3/R5GN1Ak1gdQL+osnHXPxFk4/VCTSBdRH1F00+5uIvmnysi2gT2KAy/qLJx1z8RZOPDSrTBHu266kTKJVKcYcQKpp8zMVfNPlE5aIqCTw/refGcG9vb9whhIomH3PxF00+UbmoSgK77UzAWzT5mIu/aPKxM4Em2J7QUydQrVbjDiFUNPmYi79o8onKRVUSsDoBf9HkYy7+osnH6gSawOoE/EWTj7n4iyYfqxNogpHX9Oi0tbXFHUKoaPIxF3/R5BOVy4aPmiKSEJGnReTh4P1FInJURAZE5Lsikg7atwXvB4L5eza67XOp6nl+HImEnieigi4fc/EXTT5RuYTx0/l24Pll778EfMU592ZgAjgUtB8CJoL2rwTLhcouRU8RnZqaijuEUNHkYy7+osknKpcNJQER2QX8AfDN4L0A7wG+HyxyL/CBYPqm4D3B/P5g+dB4dkrPjeF8Ph93CKGiycdc/EWTT1QuGz0T+CrwWWDxJ3gPMOmcW7xDOwhcEExfALwCEMwvBcuHxt52PXUC4+PjcYcQKpp8zMVfNPlE5dL0T2cReT9w2jl3TESuDSug06dPc+jQIZLJJPV6nYMHD3L48GGGh4dpa2sjkUgwNTVFPp9nfHwc5xz5fJ6RkRH6ttXZ1y7sbF3gRCnJZR01ak44OZPgQEeNwXILqRbo27bAsckkV3bWmKsLhbkEl2RrvDyXIJt09KTPzJ+uCUPlFvZl65yaTdCTXqAz5ZbmT1aFsUoLhUKBnp4epqenqVQq7Nixg+HhYTKZDOl0mlKpRG9vL6VSiWq1ujR/NadarUahUKC9vR2AmZkZ+vr6KBaLiAjd3d0Ui0U6Ojqo1+vMzs4urTOVSpHL5RgdHSWXy1GpVCiXy0vz0+k02WyWsbExurq6KJfLzM/PL81vbW0lk8kwMTERmtPk5CS9vb2MjIxsead6vc7g4OBZn72t6jQ3N0ehUFjx+7QVncrlMhMTE6seI7aS08LCAoVCYc3j3qJTs4hzzRVYich/AT4M1IBWoAP4IXAdsMM5VxORdwJfcM5dJyKPBtNHRCQJDAN5d04AR44ccfv3728qpg/de4xSNZ4eQo/ddnmo65ufn6e1tTXUdcaJJh9z8RdNPut1OX78+LH+/v6r1rudpo+Yzrk7nXO7nHN7gJuBnzjn/hj4KfDBYLFbgQeD6YeC9wTzf3JuAtgob8/pqRMYGRmJO4RQ0eRjLv6iyScql8342fxnwGdEZIDGNf+7g/a7gZ6g/TPAHWFveGheT53ARk/xfEOTj7n4iyafqFxC6U7jnHsSeDKYfgl4xwrLzAMfCmN7hmEYRjjo+ekM7GzVUycwMzMTdwihosnHXPxFk09ULqqSwImSnjqBvr6+uEMIFU0+5uIvmnyiclGVBC7r0HNjuFgsxh1CqGjyMRd/0eQTlYuqJFBzesYYDrmYOnY0+ZiLv2jyicpFVRI4OaPn4VHd3d1xhxAqmnzMxV80+UTloioJHLDLQd6iycdc/EWTj10OaoLBsh6djo6OuEMIFU0+5uIvmnyictFz1ARSimzqdT0PwwNdPubiL5p8onJRdNhsPBhOC7Ozs3GHECqafMzFXzT5ROWiKgnYQPP+osnHXPxFk48NNN8ENtC8v2jyMRd/0eRjA803wVxdTx/hVCoVdwihosnHXPxFk09ULqqSQGFOT51ALpeLO4RQ0eRjLv6iyScqF1VJ4JKsnstBo6OjcYcQKpp8zMVfNPlE5aIqCbxsZwLeosnHXPxFk4+dCTRBNhnqQGWxUqlU4g4hVDT5mIu/aPKJykVVEuhJ66kTKJfLcYcQKpp8zMVfNPlE5aIqCVidgL9o8jEXf9HkY3UCTWB1Av6iycdc/EWTj9UJNMF0TU+dQDqdjjuEUNHkYy7+osknKhdVSWBI0VNEs9ls3CGEiiYfc/EXTT5Rueg5agL7snqeIDg2NhZ3CKGiycdc/EWTT1QuqpLAqVk9dQJdXV1xhxAqmnzMxV80+UTloioJWBdRf9HkYy7+osnHuog2QWdKT7HY/Px83CGEiiYfc/EXTT5RuahKAlYn4C+afMzFXzT5WJ1AE1idgL9o8jEXf9Hk432dgIhcKCI/FZFfichzInJ70N4tIo+LyMng366gXUTkayIyICLPiMgVYUksMlnVUyfQ2toadwihosnHXPxFk09ULhs5E6gB/8E5dylwDXBYRC4F7gCecM7tBZ4I3gPcAOwNXp8Avr6Bba/IWEXPiU0mk4k7hFDR5GMu/qLJJyqXpo+azrkh59zxYHoaeB64ALgJuDdY7F7gA8H0TcC3XIOngE4R2dl05CvwpjY9dQITExNxhxAqmnzMxV80+UTlEspPZxHZA1wOHAX6nHNDwaxhoC+YvgB4ZdmfDQZtofHitJ46gZ6enrhDCBVNPubiL5p8onLZcHcaEWkHfgD8iXNuSuTMdXnnnBORdfXbPH36NIcOHSKZTFKv1zl48CCHDx9meHiYtrY2EokEU1NT5PN5xsfHcc6Rz+cZGRnhbZ01srOOna0LnCgluayjRs0JJ2cSHOioMVhuIdUCfdsWODaZ5MrOGnN1oTCX4JJsjZfnEmSTjp70mfnTNWGo3MK+bJ1Tswl60gt0ptzS/MmqMFZp4YsPHOXF6QQ7Mwtkk2fmj1VamK4Je7bXeX46ye7tdbYnzswfea2F6gLsyizw7FSSve11kuJwQFJgaL6Rp1dz+uTv7aNerzM7O8uOHTsYHh4mlUqRy+UYHR0ll8tRqVQol8tL89PpNNlslrGxMbq6uiiXy8zPzy/Nb21tJZPJMDExQU9PD9PT01QqlaX5mUyGdDpNqVSit7eXUqlEtVpdmr/SfpqamuLiiy9mZGSE9vZ2AGZmZujr66NYLCIidHd3UywW6ejo8NoJYHJy8qzP3lZ1KpVKjI2Nrfh92opOlUqF3t7eVY8RW8mppaWFsbGxNY97i07NIs4137deRFLAw8Cjzrn/GrS9AFzrnBsKLvc86Zx7i4jcFUzfd+5yy9d55MgRt3///qbi+eIDR/nZqI4HSP1eb+UNuTx22+URRLNxCoUCu3fvjjuMUDAXf9Hks16X48ePH+vv779qvdvZSO8gAe4Gnl9MAAEPAbcG07cCDy5r/0jQS+gaoHRuAtgomuoENLmA9d/2FU0uoMtnK9QJvAv4MPAeETkRvG4E/hz4fRE5Cbw3eA/wCPASMAD8T+CTG9j2imiqE9DkAtZ/21c0uYAun6hcmv656Zz7R2C1jvn9KyzvgMPNbu+NoKmLqCYXsK57vqLJBXT5eN9F1Ec0DSqjyQVssA9f0eQCunxsUJkm2LNdT52AJheAUqkUdwihYS7+osknKhdVdx+fn9aj80Zd3vfNpzc5kpVZb6+k3t7eTYokeszFXzT5ROWi6kxgt6Jfz5pcwH6h+YomF9DlE5WLqiSwPaFnPAFNLgDVajXuEELDXPxFk09ULqqSgKa+9ZpcwPpv+4omF9DlsxXqBLxDU996TS5g/bd9RZML6PLxfjwBHxl5TY+OJheAtra2uEMIDXPxF00+UbmoOtJU9Ywzr8oFIJHQ84RXc/EXTT5RuahKArsyeo6cmlyApadvasBc/EWTT1QuqpLAs1N6bqZqcgHI5/NxhxAa5uIvmnyiclF1pNnbXmdsXEde891lvUVq13RXeWo8Fcq243589vj4ONu3b481hrDQ5AK6fKJy8fco0wTJ9Y1f4zU6NrdwAAAHhklEQVSaXECXz0bG4PANTS6gyycqF1VJ4JeKLqFocgFdPnbJwV80+UTloioJvD2np2+9JhfQ5TMyMhJ3CKGhyQV0+UTloioJLI7HqwFNLqDLZ6NjuvqEJhfQ5ROVi55vpmEYhrFuVCWBna16+tZrcgFdPjMzM3GHEBqaXECXT1QuqpLAiZKem4+aXECXT19fX9whhIYmF9DlE5WLnm8mcFlHjX8c0zG8nCYXCNcn7oF0isUiF154YSwxhI0mF9DlE5WLqjOBmtMzLq8mF9DlI2IuvqLJJyoXVUng5Iyeh0dpcgFdPt3d3XGHEBqaXECXT1QuqpLAgQ49fdE1uYAun2KxGHcIoaHJBXT5ROWiKgkMlvXoaHIBXT4dHR1xhxAamlxAl09ULnq+mUBKkY0mF9DlU6/X4w4hNDS5gC6fqFwUfTWhb5uevuiaXECXz+zsbNwhhIYmF9DlE5WLqiSgaXB2TS6gy8cGM/cXTT5qB5oXketF5AURGRCRO8Jct6bB2TW5gC4fG8zcXzT5qBxoXkQSwF8DNwCXAreIyKVhrf+ffvqjsFYVO5pcQJfPAw88EHcIoaHJBXT5ROUS9Tn6O4AB59xLACLyHeAm4FdhrPzpJ3/EvstuCmNVsaPJBXT4LFYqv3jP3/G/294d6bY3azS1+++/n9tvv31T1h0Hmnyicon6ctAFwCvL3g8GbaGQUXSHQ5ML6PLR5FKr6blMB7p8onKRKIdjE5EPAtc7524L3n8YuNo596nFZR555JHpoaGhpa9ZR0dHsbu7e/SNrH98fLz3jS7rO5pcQJePufiLJp8mXHb39/eveziyqC8HvQosfyLSrqBtiRtvvDEbaUSGYRjnMVGf2P4C2CsiF4lIGrgZeCjiGAzDMIyASM8EnHM1EfkU8CiQAO5xzj0XZQyGYRjGMpxzKl7A9cALwABwR9zxLIvrQuCnNHpAPQfcHrR3A48DJ4N/u4J2Ab4WeDwDXLFsXbcGy58Ebl3WfiXwy+BvvkZwr2cTnRLA08DDwfuLgKPB9r8LpIP2bcH7gWD+nmXruDNofwG4Lo79CHQC3wd+DTwPvHOr7hfgT4PP17PAfUDrVtovwD3AaeDZZW2bvi9W28YmuHw5+Jw9A/wQ6Gz2/7yZ/fq68W7mlyyqF42D0ingYiAN/D/g0rjjCmLbufghBbLAizRqJP5icccCdwBfCqZvBH4UfNCvAY4u+7C+FPzbFUwvfil+Hiwrwd/esMlOnwH+jjNJ4HvAzcH0N4B/F0x/EvhGMH0z8N1g+tJgH20LPtCngn0Y6X4E7gVuC6bTNJLCltsvNHrY/QbILNsfH91K+wV4N3AFZx84N31frLaNTXB5H5AMpr+0zGXd/+fr3a9rxrtZX7AoXzR+wT267P2dwJ1xx7VKrA8Cv08jw+8M2nYCLwTTdwG3LFv+hWD+LcBdy9rvCtp2Ar9e1n7WcpsQ/y7gCeA9wMPBl2p02Qd8aV/QuOz3zmA6GSwn5+6fxeWi3I9AjsaBU85p33L7hTNdr7uD/+eHgeu22n4B9nD2gXPT98Vq2wjb5Zx5/wr49kr/l2v9nzfzfVsrVi09nje1/iAsRGQPcDmNU7U+59xQMGsYWBxQdDWX12sfXKF9s/gq8Flg8YlwPcCkc26xU/Py7S/FHMwvBcuv13EzuAgoAn8jIk+LyDdFpI0tuF+cc68Cfwn8Fhii8f98jK25X5YTxb5YbRubycdpnI3A+l2a+b69LlqSgPeISDvwA+BPnHNTy+e5RuqOrmCjSUTk/cBp59yxuGMJgSSNU/avO+cuB2ZpXA5YYgvtly4alfcXAf8MaKNxPVkNUeyLKLYhIp8DasC3N3M760FLEliz/iBORCRFIwF82zl3f9A8IiI7g/k7adxIgtVdXq991wrtm8G7gD8UkZeB79C4JPRXQKeILPY0W779pZiD+TlgjPU7bgaDwKBz7mjw/vs0ksJW3C/vBX7jnCs656rA/TT21VbcL8uJYl+sto3QEZGPAu8H/jhIOKwR80rtY6x/v74+m3GNMuoXjV91L9H4JbR4E+WtcccVxCbAt4CvntP+Zc6+IfUXwfQfcPZNr58H7d00rmF3Ba/fAN3BvHNvet0Ygde1nLkx/PecfaPqk8H0Yc6+UfW9YPqtnH0z7CUaN8Ii3Y/A/wXeEkx/IdgnW26/AFfT6Bm0PdjWvcCnt9p+4XfvCWz6vlhtG5vgcj2NHoL5c5Zb9//5evfrmrFu1hcs6heNHgMv0rij/rm441kW17+gcYr5DHAieN1I41rdEzS6pv2fZR9WofGk1VM0urRdtWxdH6fR/WsA+Niy9qtodA08Bfw3NrmLaLDNazmTBC4OvmQDwQd0W9DeGrwfCOZfvOzvPxfE+wLLes1EuR+BtwP/FOybB4IDx5bcL8B/pNEF8VngfwUHlS2zX2h0ax0CqjTO0g5FsS9W28YmuAzQuF6/eAz4RrP/583s19d7RfrsIMMwDMMvtNwTMAzDMJrAkoBhGMZ5jCUBwzCM8xhLAoZhGOcxlgQMwzDOYywJGIZhnMdYEjAMwziPsSRgGIZxHvP/Afc2xVEXyOO+AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['Tuition of High School'].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    1695\n",
       "True       44\n",
       "Name: Tuition of High School, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Tuition of High School'].isnull().value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's modify this. Let's add:\n",
    "- More bins! I like more granular histograms.\n",
    "- A vertical line for the average tuition."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['Tuition of High School'].hist(bins=20)\n",
    "plt.axvline(df['Tuition of High School'].mean(), \n",
    "            color='yellow', \n",
    "            label=f\"Mean: ${df['Tuition of High School'].mean():,.2f}\")\n",
    "plt.legend()\n",
    "plt.title('Distribution of high school tuitions\\n(Duke University students)')\n",
    "plt.xlabel('USD ($)')\n",
    "plt.ylabel('Frequency')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "public                1096\n",
       "private                625\n",
       "Data not available      10\n",
       "Home School              8\n",
       "Name: Public or Private High School, dtype: int64"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's one hot encode (AKA make dummy variables) the public or private high school column\n",
    "df['Public or Private High School'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school_type_Data not available</th>\n",
       "      <th>school_type_Home School</th>\n",
       "      <th>school_type_private</th>\n",
       "      <th>school_type_public</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   school_type_Data not available  school_type_Home School  \\\n",
       "0                               0                        0   \n",
       "1                               0                        0   \n",
       "2                               0                        0   \n",
       "3                               0                        0   \n",
       "4                               0                        0   \n",
       "\n",
       "   school_type_private  school_type_public  \n",
       "0                    1                   0  \n",
       "1                    0                   1  \n",
       "2                    1                   0  \n",
       "3                    0                   1  \n",
       "4                    1                   0  "
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "school_ohe = pd.get_dummies(df['Public or Private High School'], prefix='school_type')\n",
    "school_ohe.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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>Greek Council</th>\n",
       "      <th>Greek Organization</th>\n",
       "      <th>Home City</th>\n",
       "      <th>Home State</th>\n",
       "      <th>Home Country</th>\n",
       "      <th>Latitude of Home City</th>\n",
       "      <th>Longitude of Home City</th>\n",
       "      <th>Free and reduced percentage of  High School</th>\n",
       "      <th>Public or Private High School</th>\n",
       "      <th>Domestic or International High School</th>\n",
       "      <th>Boarding/Day High School</th>\n",
       "      <th>Religous High School?</th>\n",
       "      <th>Gender of High School's Students</th>\n",
       "      <th>Tuition of High School</th>\n",
       "      <th>Sports Team</th>\n",
       "      <th>Merit Scholarship</th>\n",
       "      <th>school_type_Data not available</th>\n",
       "      <th>school_type_Home School</th>\n",
       "      <th>school_type_private</th>\n",
       "      <th>school_type_public</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Accokeek</td>\n",
       "      <td>Maryland</td>\n",
       "      <td>United States</td>\n",
       "      <td>38.667617</td>\n",
       "      <td>-77.028309</td>\n",
       "      <td>Data not available</td>\n",
       "      <td>private</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Episcopal</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>36610.00</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Advance</td>\n",
       "      <td>North Carolina</td>\n",
       "      <td>United States</td>\n",
       "      <td>35.941248</td>\n",
       "      <td>-80.409222</td>\n",
       "      <td>0.34741523</td>\n",
       "      <td>public</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Fooball</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Ajman</td>\n",
       "      <td>NaN</td>\n",
       "      <td>United Arab Emirates</td>\n",
       "      <td>25.509129</td>\n",
       "      <td>55.361907</td>\n",
       "      <td>NaN</td>\n",
       "      <td>private</td>\n",
       "      <td>International</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>13093.46</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Alameda</td>\n",
       "      <td>California</td>\n",
       "      <td>United States</td>\n",
       "      <td>37.609029</td>\n",
       "      <td>-121.899142</td>\n",
       "      <td>0.2268</td>\n",
       "      <td>public</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>0.00</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Albuquerque</td>\n",
       "      <td>New Mexico</td>\n",
       "      <td>United States</td>\n",
       "      <td>35.084103</td>\n",
       "      <td>-106.650985</td>\n",
       "      <td>Data not available</td>\n",
       "      <td>private</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>22870.00</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Greek Council Greek Organization    Home City      Home State  \\\n",
       "0          None               None     Accokeek        Maryland   \n",
       "1          None               None      Advance  North Carolina   \n",
       "2          None               None        Ajman             NaN   \n",
       "3          None               None      Alameda      California   \n",
       "4          None               None  Albuquerque      New Mexico   \n",
       "\n",
       "           Home Country  Latitude of Home City  Longitude of Home City  \\\n",
       "0         United States              38.667617              -77.028309   \n",
       "1         United States              35.941248              -80.409222   \n",
       "2  United Arab Emirates              25.509129               55.361907   \n",
       "3         United States              37.609029             -121.899142   \n",
       "4         United States              35.084103             -106.650985   \n",
       "\n",
       "  Free and reduced percentage of  High School Public or Private High School  \\\n",
       "0                          Data not available                       private   \n",
       "1                                  0.34741523                        public   \n",
       "2                                         NaN                       private   \n",
       "3                                      0.2268                        public   \n",
       "4                          Data not available                       private   \n",
       "\n",
       "  Domestic or International High School Boarding/Day High School  \\\n",
       "0                              Domestic                      Day   \n",
       "1                              Domestic                      Day   \n",
       "2                         International                      Day   \n",
       "3                              Domestic                      Day   \n",
       "4                              Domestic                      Day   \n",
       "\n",
       "  Religous High School? Gender of High School's Students  \\\n",
       "0             Episcopal                            Co-Ed   \n",
       "1          Unafilliated                            Co-Ed   \n",
       "2          Unafilliated                            Co-Ed   \n",
       "3          Unafilliated                            Co-Ed   \n",
       "4          Unafilliated                            Co-Ed   \n",
       "\n",
       "   Tuition of High School Sports Team Merit Scholarship  \\\n",
       "0                36610.00        None              None   \n",
       "1                    0.00     Fooball              None   \n",
       "2                13093.46        None              None   \n",
       "3                    0.00        None              None   \n",
       "4                22870.00        None              None   \n",
       "\n",
       "   school_type_Data not available  school_type_Home School  \\\n",
       "0                               0                        0   \n",
       "1                               0                        0   \n",
       "2                               0                        0   \n",
       "3                               0                        0   \n",
       "4                               0                        0   \n",
       "\n",
       "   school_type_private  school_type_public  \n",
       "0                    1                   0  \n",
       "1                    0                   1  \n",
       "2                    1                   0  \n",
       "3                    0                   1  \n",
       "4                    1                   0  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's add those dummies back into the original dataframe via merge!\n",
    "df = pd.merge(df, school_ohe, left_index=True, right_index=True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Greek Council', 'Greek Organization', 'Home City', 'Home State',\n",
       "       'Home Country', 'Latitude of Home City', 'Longitude of Home City',\n",
       "       'Free and reduced percentage of  High School',\n",
       "       'Public or Private High School',\n",
       "       'Domestic or International High School', 'Boarding/Day High School',\n",
       "       'Religous High School?', 'Gender of High School's Students',\n",
       "       'Tuition of High School', 'Sports Team', 'Merit Scholarship',\n",
       "       'school_type_Data not available', 'school_type_Home School',\n",
       "       'school_type_private', 'school_type_public'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# I've forgotten - which columns do we have?\n",
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "None          0.661875\n",
       "Sorority      0.184014\n",
       "Fraternity    0.154112\n",
       "Name: Greek Council, dtype: float64"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Greek Council'].value_counts() / len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "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>greek_council</th>\n",
       "      <th>greek_organization</th>\n",
       "      <th>home_city</th>\n",
       "      <th>home_state</th>\n",
       "      <th>home_country</th>\n",
       "      <th>latitude_of_home_city</th>\n",
       "      <th>longitude_of_home_city</th>\n",
       "      <th>free_and_reduced_percentage_of__high_school</th>\n",
       "      <th>public_or_private_high_school</th>\n",
       "      <th>domestic_or_international_high_school</th>\n",
       "      <th>boarding/day_high_school</th>\n",
       "      <th>religous_high_school?</th>\n",
       "      <th>gender_of_high_school's_students</th>\n",
       "      <th>tuition_of_high_school</th>\n",
       "      <th>sports_team</th>\n",
       "      <th>merit_scholarship</th>\n",
       "      <th>school_type_data_not_available</th>\n",
       "      <th>school_type_home_school</th>\n",
       "      <th>school_type_private</th>\n",
       "      <th>school_type_public</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Accokeek</td>\n",
       "      <td>Maryland</td>\n",
       "      <td>United States</td>\n",
       "      <td>38.667617</td>\n",
       "      <td>-77.028309</td>\n",
       "      <td>Data not available</td>\n",
       "      <td>private</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Episcopal</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>36610.00</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Advance</td>\n",
       "      <td>North Carolina</td>\n",
       "      <td>United States</td>\n",
       "      <td>35.941248</td>\n",
       "      <td>-80.409222</td>\n",
       "      <td>0.34741523</td>\n",
       "      <td>public</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Fooball</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Ajman</td>\n",
       "      <td>NaN</td>\n",
       "      <td>United Arab Emirates</td>\n",
       "      <td>25.509129</td>\n",
       "      <td>55.361907</td>\n",
       "      <td>NaN</td>\n",
       "      <td>private</td>\n",
       "      <td>International</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>13093.46</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Alameda</td>\n",
       "      <td>California</td>\n",
       "      <td>United States</td>\n",
       "      <td>37.609029</td>\n",
       "      <td>-121.899142</td>\n",
       "      <td>0.2268</td>\n",
       "      <td>public</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>0.00</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>Albuquerque</td>\n",
       "      <td>New Mexico</td>\n",
       "      <td>United States</td>\n",
       "      <td>35.084103</td>\n",
       "      <td>-106.650985</td>\n",
       "      <td>Data not available</td>\n",
       "      <td>private</td>\n",
       "      <td>Domestic</td>\n",
       "      <td>Day</td>\n",
       "      <td>Unafilliated</td>\n",
       "      <td>Co-Ed</td>\n",
       "      <td>22870.00</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  greek_council greek_organization    home_city      home_state  \\\n",
       "0          None               None     Accokeek        Maryland   \n",
       "1          None               None      Advance  North Carolina   \n",
       "2          None               None        Ajman             NaN   \n",
       "3          None               None      Alameda      California   \n",
       "4          None               None  Albuquerque      New Mexico   \n",
       "\n",
       "           home_country  latitude_of_home_city  longitude_of_home_city  \\\n",
       "0         United States              38.667617              -77.028309   \n",
       "1         United States              35.941248              -80.409222   \n",
       "2  United Arab Emirates              25.509129               55.361907   \n",
       "3         United States              37.609029             -121.899142   \n",
       "4         United States              35.084103             -106.650985   \n",
       "\n",
       "  free_and_reduced_percentage_of__high_school public_or_private_high_school  \\\n",
       "0                          Data not available                       private   \n",
       "1                                  0.34741523                        public   \n",
       "2                                         NaN                       private   \n",
       "3                                      0.2268                        public   \n",
       "4                          Data not available                       private   \n",
       "\n",
       "  domestic_or_international_high_school boarding/day_high_school  \\\n",
       "0                              Domestic                      Day   \n",
       "1                              Domestic                      Day   \n",
       "2                         International                      Day   \n",
       "3                              Domestic                      Day   \n",
       "4                              Domestic                      Day   \n",
       "\n",
       "  religous_high_school? gender_of_high_school's_students  \\\n",
       "0             Episcopal                            Co-Ed   \n",
       "1          Unafilliated                            Co-Ed   \n",
       "2          Unafilliated                            Co-Ed   \n",
       "3          Unafilliated                            Co-Ed   \n",
       "4          Unafilliated                            Co-Ed   \n",
       "\n",
       "   tuition_of_high_school sports_team merit_scholarship  \\\n",
       "0                36610.00        None              None   \n",
       "1                    0.00     Fooball              None   \n",
       "2                13093.46        None              None   \n",
       "3                    0.00        None              None   \n",
       "4                22870.00        None              None   \n",
       "\n",
       "   school_type_data_not_available  school_type_home_school  \\\n",
       "0                               0                        0   \n",
       "1                               0                        0   \n",
       "2                               0                        0   \n",
       "3                               0                        0   \n",
       "4                               0                        0   \n",
       "\n",
       "   school_type_private  school_type_public  \n",
       "0                    1                   0  \n",
       "1                    0                   1  \n",
       "2                    1                   0  \n",
       "3                    0                   1  \n",
       "4                    1                   0  "
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# I don't really like typing out the long Proper English Words as column names,\n",
    "# let's make them lowercase snake_case!\n",
    "for col in df.columns:\n",
    "    df.rename(columns={col: col.lower().replace(' ', '_') }, inplace=True)\n",
    "    \n",
    "# Check to make sure it worked    \n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One of the handiest `pandas` methods: `.apply()`\n",
    "\n",
    "`pandas` has lots of built-in things you can do: summing across rows or columns, taking averages, changing the data types, and so on. But sometimes you want to do something specific and a bit more complicated. For example, you want to use multiple columns to calculate a value. In those cases, `.apply()` is helpful.\n",
    "\n",
    "I highly recommend watching the [Data School](http://www.dataschool.io/) tutorial on `.apply()`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/jpeg": "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\n",
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"400\"\n",
       "            height=\"300\"\n",
       "            src=\"https://www.youtube.com/embed/P_q0tkYqvSk\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.YouTubeVideo at 0x109507c18>"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import YouTubeVideo\n",
    "YouTubeVideo('P_q0tkYqvSk')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I actually had trouble thinking of a good `.apply()` usage with this dataset! So here's a kinda crappy one: let's infer the biological sex of each student observation. We can do this by:\n",
    "1. Assuming that `Fraternity == male` and `Sorority == female`.\n",
    "2. Assuming that, if they aren't a member of a Greek organization, they have a 50% likelihood of being male/female.\n",
    "\n",
    "This will be an imperfect measure, for a bunch of reasons. And we won't correctly assign sex to each observation, we'll just get the right \"averages\" out. \n",
    "\n",
    "Anyway, here's how `.apply()` works! (Side note: We're going to use [`scipy`](https://www.scipy.org/) - another **great data science tool** - to randomly sample from a [Bernoulli distribution](https://en.wikipedia.org/wiki/Bernoulli_trial)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.531915\n",
       "1    0.468085\n",
       "Name: guessed_sex, dtype: float64"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from scipy.stats import bernoulli\n",
    "\n",
    "# This function will take each row as the input\n",
    "def guessing_sex(row):\n",
    "    \"\"\"Male = 1, female = 0\n",
    "    \"\"\"\n",
    "    if row['greek_council'] == 'Fraternity':\n",
    "        return 1\n",
    "    elif row['greek_council'] == 'Sorority':\n",
    "        return 0\n",
    "    else:\n",
    "        return bernoulli.rvs(p=0.5)\n",
    "    \n",
    "# Applying that function to the dataframe\n",
    "df['guessed_sex'] = df.apply(guessing_sex, axis=1)\n",
    "df['guessed_sex'].value_counts() / len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Now checking: None\n",
      "0    0.347901\n",
      "1    0.313974\n",
      "Name: guessed_sex, dtype: float64\n",
      "\n",
      "Now checking: Fraternity\n",
      "1    0.154112\n",
      "Name: guessed_sex, dtype: float64\n",
      "\n",
      "Now checking: Sorority\n",
      "0    0.184014\n",
      "Name: guessed_sex, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# Sanity check\n",
    "for membership in df['greek_council'].unique():\n",
    "    print(f\"\\nNow checking: {membership}\")\n",
    "    print(df[df['greek_council']==membership]['guessed_sex'].value_counts() / len(df))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Onto [statistics and machine learning $\\rightarrow$](2_StatsML.ipynb)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}