Example Notebook.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "from roi import equity, external, utilities, macro, metrics, cost, surveys\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "\n",
    "# make figures readable by humans instead of mice\n",
    "plt.rcParams['figure.figsize'] = [8, 4]\n",
    "plt.rcParams['figure.dpi'] = 100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 1: Setup__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook provides some simple instruction in many of the basic methods offered by the ROI Toolkit. Below, we import two files containing fake simulated data. The code to create these files lives in the `testing` directory under the top level of the `roi-toolkit` repo."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "bls_api_key = \"c8803d0ba66c4592b8b0eff68ac9ebb0\" # BLS API Key - free and easy to get"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"testing/testing-data/test_microdata.csv\") # Fake student-level data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "programs = pd.read_csv(\"testing/testing-data/programs.csv\") # Fake program-level data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we take a look at the sample data. It contains joined educational and wage records, with one row for each individual. This fake data includes only two earnings columns, `earnings_start` and `earnings_end`. In this data, these columns denote annualized individual earnings, denoted in denominal dollars, at some set periods before and after program entry and exit, e.g. the final quarter before entry and the first quarter after.\n",
    "\n",
    "Ideally, in real data you'll have many of these columns—for example, at 1, 5, and 10 years after program completion. All of the steps in this notebook can be repeated with multiple \"after\" columns; in the example, we use only one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Unnamed: 0</th>\n",
       "      <th>race</th>\n",
       "      <th>gender</th>\n",
       "      <th>earnings_start</th>\n",
       "      <th>earnings_end</th>\n",
       "      <th>age</th>\n",
       "      <th>program</th>\n",
       "      <th>program_start</th>\n",
       "      <th>program_end</th>\n",
       "      <th>education_level</th>\n",
       "      <th>...</th>\n",
       "      <th>employed_at_start</th>\n",
       "      <th>state</th>\n",
       "      <th>full_address</th>\n",
       "      <th>Address</th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>Zip</th>\n",
       "      <th>id</th>\n",
       "      <th>start_month</th>\n",
       "      <th>end_month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>46027.000</td>\n",
       "      <td>91781.300</td>\n",
       "      <td>25</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2014</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>1647 COURT STREET REDDING CA 96001</td>\n",
       "      <td>1647 COURT STREET</td>\n",
       "      <td>REDDING</td>\n",
       "      <td>CA</td>\n",
       "      <td>96001.000</td>\n",
       "      <td>5335</td>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>41976.000</td>\n",
       "      <td>66896.700</td>\n",
       "      <td>25</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2017</td>\n",
       "      <td>2025</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>8340</td>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>61396.000</td>\n",
       "      <td>74524.600</td>\n",
       "      <td>51</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2014</td>\n",
       "      <td>2019</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>690 MEDICAL PARK DRIVE AIKEN SC 29801</td>\n",
       "      <td>690 MEDICAL PARK DRIVE</td>\n",
       "      <td>AIKEN</td>\n",
       "      <td>SC</td>\n",
       "      <td>29801.000</td>\n",
       "      <td>13875</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>34089.000</td>\n",
       "      <td>92032.900</td>\n",
       "      <td>18</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2018</td>\n",
       "      <td>2022</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>1201</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>57897.000</td>\n",
       "      <td>79573.600</td>\n",
       "      <td>35</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2011</td>\n",
       "      <td>2013</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>1547 NORTH COAST HWY NEWPORT OR 97365</td>\n",
       "      <td>1547 NORTH COAST HWY</td>\n",
       "      <td>NEWPORT</td>\n",
       "      <td>OR</td>\n",
       "      <td>97365.000</td>\n",
       "      <td>15397</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0   race gender  earnings_start  earnings_end  age  \\\n",
       "0           1  White   Male       46027.000     91781.300   25   \n",
       "1           2  White   Male       41976.000     66896.700   25   \n",
       "2           3  White   Male       61396.000     74524.600   51   \n",
       "3           4  White   Male       34089.000     92032.900   18   \n",
       "4           5  White   Male       57897.000     79573.600   35   \n",
       "\n",
       "                  program  program_start  program_end  education_level  ...  \\\n",
       "0  Electrical Engineering           2014         2016                0  ...   \n",
       "1  Electrical Engineering           2017         2025                0  ...   \n",
       "2  Electrical Engineering           2014         2019                0  ...   \n",
       "3  Electrical Engineering           2018         2022                0  ...   \n",
       "4  Electrical Engineering           2011         2013                0  ...   \n",
       "\n",
       "   employed_at_start  state                           full_address  \\\n",
       "0                  1      8     1647 COURT STREET REDDING CA 96001   \n",
       "1                  1      8                                    NaN   \n",
       "2                  1      8  690 MEDICAL PARK DRIVE AIKEN SC 29801   \n",
       "3                  1      8                                    NaN   \n",
       "4                  1      8  1547 NORTH COAST HWY NEWPORT OR 97365   \n",
       "\n",
       "                  Address     City State       Zip     id  start_month  \\\n",
       "0       1647 COURT STREET  REDDING    CA 96001.000   5335            9   \n",
       "1                     NaN      NaN   NaN       nan   8340            9   \n",
       "2  690 MEDICAL PARK DRIVE    AIKEN    SC 29801.000  13875            8   \n",
       "3                     NaN      NaN   NaN       nan   1201            9   \n",
       "4    1547 NORTH COAST HWY  NEWPORT    OR 97365.000  15397           10   \n",
       "\n",
       "   end_month  \n",
       "0          6  \n",
       "1          6  \n",
       "2          7  \n",
       "3          5  \n",
       "4          6  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 2: Equity analysis__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are four key equity metric classes in the `equity` submodule: `Gini`, `Theil_T`, `Theil_L`, and `Variance_Analysis`. Each method inherits from the `Metric()` parent class. These methods can be supplied with numpy arrays, but the easiest way to use them is with the `.from_dataframe` factory method, demonstrated below with the following syntax:\n",
    "\n",
    "`from_dataframe(frame, group_columns, value_column, sample=False)`\n",
    "\n",
    "The `sample` argument uses only a random subset (without replacement) of the data in order to calculate the equity metric. This is really only necessary with the `Gini()` class, which needs to do pairwise comparisons across all observations, and therefore runs in factorial time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mattlerner/roi/sbir-roi/roi/equity.py:77: UserWarning: When sample == True, Metric() will use the value of sample to create a random subset of values that it will use for all calculations. This number will be stored as attribute .samplesize.\n",
      "  warnings.warn(\"When sample == True, Metric() will use the value of sample to create a random subset of values that it will use for all calculations. This number will be stored as attribute .samplesize.\")\n",
      "/Users/mattlerner/roi/sbir-roi/roi/equity.py:61: UserWarning: Data passed contains 103 NA values. All equity metrics are calculated by ignoring NA values, in effect dropping them. Please ensure that you have a strategy for dealing with these missing values. If values are not missing at random, e.g. if individuals with missing values have common traits, then ALL METRICS calculated will be biased.\n",
      "  warnings.warn(\"Data passed contains {} NA values. All equity metrics are calculated by ignoring NA values, in effect dropping them. Please ensure that you have a strategy for dealing with these missing values. If values are not missing at random, e.g. if individuals with missing values have common traits, then ALL METRICS calculated will be biased.\".format(self.nans))\n"
     ]
    }
   ],
   "source": [
    "gini = equity.Gini.from_dataframe(data, 'gender', 'earnings_end', sample = 2000) # library method\n",
    "gini.calculate()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Above, you'll see the verbose warnings that are common to the `roi-toolkit` module. Users are kept appraised, as much as possible, about what's going on behind the scenes. Warnings should be read, and ignored only after they've been understood."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `.calculate()` method attached to each equity class does the calculation work, as you might expect. It creates several new attributes of the class instance that can be used for equity analysis. The `.viz` attribute, for example, contains a Seaborn box plot allowing the analyst to visualize differences between groups."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gini.viz"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One relatively uncommon feature of the equity classes included in the `equity` submodule is the inclusion of the `.ratio` attribute.\n",
    "\n",
    "Each of the four equity classes included in the submodule is broadly decomposable into \"within\" and \"between\"-group inequality. The values for these subcomponents are accessible (after `.calculate()` is run) at the instance attributes `.within` and `.between`, and their sum--the value of the calculated index--is accessible at `.overall`.\n",
    "\n",
    "The `.ratio` attribute is simply the ratio of `.between` to `.overall`: it takes a value between 0 and 1 and is broadly interpretable as the amount (the percentage) of variation that is accounted for by cross-group--instead of within-group variation.\n",
    "\n",
    "Values closer to 1 indicate that greater variation in, say, wages, is accounted for by someone's group membership. Depending on the grouping variable provided, this may or may not be indicative of an equity issue. Take an extreme case: a value of 1 and a grouping variable of `SES`. In this scenario, all the variation in program graduate wages is accounted for by student socioeconomic status. This would mean that, no matter what classes they take, what their personal aptitude is, or what grades they get, differences in student earnings are wholly accounted for by SES. Many analysts would take this to be a significant equity concern"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3331446175942655"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gini.ratio"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 3: Geocode addresses__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to use information about particular geographic areas in the United States, it may be necessary to u.s. __geocodes__. Geocodes are used by the U.S. Census in order to uniquely identify areas as small as a census tract. The `external.Census.get_batch_geocode()` method takes a dataframe and queries the U.S. Census geocoding API to get geocodes for the addresses included in the frame.\n",
    "\n",
    "This method uses the [geocoding API](https://geocoding.geo.census.gov/geocoder/Geocoding_Services_API.pdf) from the census to get unique twelve-digit IDs that identify specific Census block groups in the U.S. You can find more information on the Census [web site](https://geocoding.geo.census.gov/)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "small_data = data.sample(40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>race</th>\n",
       "      <th>gender</th>\n",
       "      <th>earnings_start</th>\n",
       "      <th>earnings_end</th>\n",
       "      <th>age</th>\n",
       "      <th>program</th>\n",
       "      <th>program_start</th>\n",
       "      <th>program_end</th>\n",
       "      <th>education_level</th>\n",
       "      <th>...</th>\n",
       "      <th>employed_at_start</th>\n",
       "      <th>state</th>\n",
       "      <th>full_address</th>\n",
       "      <th>Address</th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>Zip</th>\n",
       "      <th>id</th>\n",
       "      <th>start_month</th>\n",
       "      <th>end_month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2190</th>\n",
       "      <td>2191</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Male</td>\n",
       "      <td>30353.000</td>\n",
       "      <td>nan</td>\n",
       "      <td>53</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2014</td>\n",
       "      <td>2019</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>620</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3134</th>\n",
       "      <td>3135</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>20183.000</td>\n",
       "      <td>27948.510</td>\n",
       "      <td>18</td>\n",
       "      <td>Biology</td>\n",
       "      <td>2013</td>\n",
       "      <td>2016</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>121 E. CEDAR STREET, PO BOX 100550 FLORENCE SC...</td>\n",
       "      <td>121 E. CEDAR STREET, PO BOX 100550</td>\n",
       "      <td>FLORENCE</td>\n",
       "      <td>SC</td>\n",
       "      <td>29501.000</td>\n",
       "      <td>15210</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9719</th>\n",
       "      <td>9720</td>\n",
       "      <td>Other</td>\n",
       "      <td>Male</td>\n",
       "      <td>26787.000</td>\n",
       "      <td>46582.200</td>\n",
       "      <td>23</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2014</td>\n",
       "      <td>2015</td>\n",
       "      <td>111</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>2029  ASHEVILLE HIGHWAY HENDERSONVILLE NC 28739</td>\n",
       "      <td>2029  ASHEVILLE HIGHWAY</td>\n",
       "      <td>HENDERSONVILLE</td>\n",
       "      <td>NC</td>\n",
       "      <td>28739.000</td>\n",
       "      <td>1784</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9723</th>\n",
       "      <td>9724</td>\n",
       "      <td>Other</td>\n",
       "      <td>Male</td>\n",
       "      <td>27723.000</td>\n",
       "      <td>46797.800</td>\n",
       "      <td>38</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2012</td>\n",
       "      <td>2015</td>\n",
       "      <td>111</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>5003 HORIZONS DRIVE, SUITE 100 COLUMBUS OH 43220</td>\n",
       "      <td>5003 HORIZONS DRIVE, SUITE 100</td>\n",
       "      <td>COLUMBUS</td>\n",
       "      <td>OH</td>\n",
       "      <td>43220.000</td>\n",
       "      <td>14173</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2377</th>\n",
       "      <td>2378</td>\n",
       "      <td>Other</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>27332.000</td>\n",
       "      <td>28858.775</td>\n",
       "      <td>27</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2018</td>\n",
       "      <td>2020</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>700 RIVER DRIVE FORT BRAGG CA 95437</td>\n",
       "      <td>700 RIVER DRIVE</td>\n",
       "      <td>FORT BRAGG</td>\n",
       "      <td>CA</td>\n",
       "      <td>95437.000</td>\n",
       "      <td>3435</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Unnamed: 0                       race             gender  \\\n",
       "2190        2191  Asian or Pacific Islander               Male   \n",
       "3134        3135  Asian or Pacific Islander  Other/unspecified   \n",
       "9719        9720                      Other               Male   \n",
       "9723        9724                      Other               Male   \n",
       "2377        2378                      Other  Other/unspecified   \n",
       "\n",
       "      earnings_start  earnings_end  age      program  program_start  \\\n",
       "2190       30353.000           nan   53  Art History           2014   \n",
       "3134       20183.000     27948.510   18      Biology           2013   \n",
       "9719       26787.000     46582.200   23    Economics           2014   \n",
       "9723       27723.000     46797.800   38    Economics           2012   \n",
       "2377       27332.000     28858.775   27  Art History           2018   \n",
       "\n",
       "      program_end  education_level  ...  employed_at_start  state  \\\n",
       "2190         2019               60  ...                  0      8   \n",
       "3134         2016               60  ...                  1      8   \n",
       "9719         2015              111  ...                  0      8   \n",
       "9723         2015              111  ...                  1      8   \n",
       "2377         2020               60  ...                  0      8   \n",
       "\n",
       "                                           full_address  \\\n",
       "2190                                                NaN   \n",
       "3134  121 E. CEDAR STREET, PO BOX 100550 FLORENCE SC...   \n",
       "9719    2029  ASHEVILLE HIGHWAY HENDERSONVILLE NC 28739   \n",
       "9723   5003 HORIZONS DRIVE, SUITE 100 COLUMBUS OH 43220   \n",
       "2377                700 RIVER DRIVE FORT BRAGG CA 95437   \n",
       "\n",
       "                                 Address            City State       Zip  \\\n",
       "2190                                 NaN             NaN   NaN       nan   \n",
       "3134  121 E. CEDAR STREET, PO BOX 100550        FLORENCE    SC 29501.000   \n",
       "9719             2029  ASHEVILLE HIGHWAY  HENDERSONVILLE    NC 28739.000   \n",
       "9723      5003 HORIZONS DRIVE, SUITE 100        COLUMBUS    OH 43220.000   \n",
       "2377                     700 RIVER DRIVE      FORT BRAGG    CA 95437.000   \n",
       "\n",
       "         id  start_month  end_month  \n",
       "2190    620            8          7  \n",
       "3134  15210            9          5  \n",
       "9719   1784           10          6  \n",
       "9723  14173           10          7  \n",
       "2377   3435            9          5  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "small_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Successfully geocoded 42.5% of 40 passed addresses.\n",
      "Of successfully matched addresses, 88.24% were exact matches\n"
     ]
    }
   ],
   "source": [
    "small_data['geocode'] = external.Census.get_batch_geocode(small_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, the geocoding method is verbose, telling the user exactly how many addresses were matched and what percentage of those were matched exactly. The method returns a Pandas series which is most easily deployed simply by adding it as a column to the original dataframe, which you can see below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>race</th>\n",
       "      <th>gender</th>\n",
       "      <th>earnings_start</th>\n",
       "      <th>earnings_end</th>\n",
       "      <th>age</th>\n",
       "      <th>program</th>\n",
       "      <th>program_start</th>\n",
       "      <th>program_end</th>\n",
       "      <th>education_level</th>\n",
       "      <th>...</th>\n",
       "      <th>state</th>\n",
       "      <th>full_address</th>\n",
       "      <th>Address</th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>Zip</th>\n",
       "      <th>id</th>\n",
       "      <th>start_month</th>\n",
       "      <th>end_month</th>\n",
       "      <th>geocode</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2190</th>\n",
       "      <td>2191</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Male</td>\n",
       "      <td>30353.000</td>\n",
       "      <td>nan</td>\n",
       "      <td>53</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2014</td>\n",
       "      <td>2019</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>620</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3134</th>\n",
       "      <td>3135</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>20183.000</td>\n",
       "      <td>27948.510</td>\n",
       "      <td>18</td>\n",
       "      <td>Biology</td>\n",
       "      <td>2013</td>\n",
       "      <td>2016</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>121 E. CEDAR STREET, PO BOX 100550 FLORENCE SC...</td>\n",
       "      <td>121 E. CEDAR STREET, PO BOX 100550</td>\n",
       "      <td>FLORENCE</td>\n",
       "      <td>SC</td>\n",
       "      <td>29501.000</td>\n",
       "      <td>15210</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9719</th>\n",
       "      <td>9720</td>\n",
       "      <td>Other</td>\n",
       "      <td>Male</td>\n",
       "      <td>26787.000</td>\n",
       "      <td>46582.200</td>\n",
       "      <td>23</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2014</td>\n",
       "      <td>2015</td>\n",
       "      <td>111</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>2029  ASHEVILLE HIGHWAY HENDERSONVILLE NC 28739</td>\n",
       "      <td>2029  ASHEVILLE HIGHWAY</td>\n",
       "      <td>HENDERSONVILLE</td>\n",
       "      <td>NC</td>\n",
       "      <td>28739.000</td>\n",
       "      <td>1784</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>370899309002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9723</th>\n",
       "      <td>9724</td>\n",
       "      <td>Other</td>\n",
       "      <td>Male</td>\n",
       "      <td>27723.000</td>\n",
       "      <td>46797.800</td>\n",
       "      <td>38</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2012</td>\n",
       "      <td>2015</td>\n",
       "      <td>111</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>5003 HORIZONS DRIVE, SUITE 100 COLUMBUS OH 43220</td>\n",
       "      <td>5003 HORIZONS DRIVE, SUITE 100</td>\n",
       "      <td>COLUMBUS</td>\n",
       "      <td>OH</td>\n",
       "      <td>43220.000</td>\n",
       "      <td>14173</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2377</th>\n",
       "      <td>2378</td>\n",
       "      <td>Other</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>27332.000</td>\n",
       "      <td>28858.775</td>\n",
       "      <td>27</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2018</td>\n",
       "      <td>2020</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>700 RIVER DRIVE FORT BRAGG CA 95437</td>\n",
       "      <td>700 RIVER DRIVE</td>\n",
       "      <td>FORT BRAGG</td>\n",
       "      <td>CA</td>\n",
       "      <td>95437.000</td>\n",
       "      <td>3435</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "      <td>060450105004</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Unnamed: 0                       race             gender  \\\n",
       "2190        2191  Asian or Pacific Islander               Male   \n",
       "3134        3135  Asian or Pacific Islander  Other/unspecified   \n",
       "9719        9720                      Other               Male   \n",
       "9723        9724                      Other               Male   \n",
       "2377        2378                      Other  Other/unspecified   \n",
       "\n",
       "      earnings_start  earnings_end  age      program  program_start  \\\n",
       "2190       30353.000           nan   53  Art History           2014   \n",
       "3134       20183.000     27948.510   18      Biology           2013   \n",
       "9719       26787.000     46582.200   23    Economics           2014   \n",
       "9723       27723.000     46797.800   38    Economics           2012   \n",
       "2377       27332.000     28858.775   27  Art History           2018   \n",
       "\n",
       "      program_end  education_level  ...  state  \\\n",
       "2190         2019               60  ...      8   \n",
       "3134         2016               60  ...      8   \n",
       "9719         2015              111  ...      8   \n",
       "9723         2015              111  ...      8   \n",
       "2377         2020               60  ...      8   \n",
       "\n",
       "                                           full_address  \\\n",
       "2190                                                NaN   \n",
       "3134  121 E. CEDAR STREET, PO BOX 100550 FLORENCE SC...   \n",
       "9719    2029  ASHEVILLE HIGHWAY HENDERSONVILLE NC 28739   \n",
       "9723   5003 HORIZONS DRIVE, SUITE 100 COLUMBUS OH 43220   \n",
       "2377                700 RIVER DRIVE FORT BRAGG CA 95437   \n",
       "\n",
       "                                 Address            City State       Zip  \\\n",
       "2190                                 NaN             NaN   NaN       nan   \n",
       "3134  121 E. CEDAR STREET, PO BOX 100550        FLORENCE    SC 29501.000   \n",
       "9719             2029  ASHEVILLE HIGHWAY  HENDERSONVILLE    NC 28739.000   \n",
       "9723      5003 HORIZONS DRIVE, SUITE 100        COLUMBUS    OH 43220.000   \n",
       "2377                     700 RIVER DRIVE      FORT BRAGG    CA 95437.000   \n",
       "\n",
       "         id start_month  end_month       geocode  \n",
       "2190    620           8          7                \n",
       "3134  15210           9          5                \n",
       "9719   1784          10          6  370899309002  \n",
       "9723  14173          10          7                \n",
       "2377   3435           9          5  060450105004  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "small_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 4: Get SES for each individual with a geocode__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One useful thing you can do with Census geocodes is to join them with datasets that provide geographically-sourced data that can give a clue as to individuals' socioeconomic status.\n",
    "\n",
    "One such dataset is the [Area Deprivation Index](https://www.neighborhoodatlas.medicine.wisc.edu/) from the University of Wisconsin. The ADI classifies Census block groups on a scale from 1-100, where 100 is the most \"deprived.\" We use ADI here as a proxy for socioeconomic status.\n",
    "\n",
    "The ROI Toolkit is packaged with the latest ADI dataset. When you call the `equity.ADI()` method below, the module loads up the data and then separates the block groups into five quintiles of deprivation, with the highest quintile (\"80-100\") being the most deprived."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "adi = equity.ADI()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On running the `.get_quintile_for_geocodes_frame()` method, you'll get back a Pandas series in the same order as the original dataframe. The method will tell you exactly how many of the rows provided were matched with ADI quintiles. The percentage should match the percentage of addresses that were geocoded above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Geocode merge: Merged 17 of 40 observations in input dataframe (42.5%)\n"
     ]
    }
   ],
   "source": [
    "small_data['ses'] = adi.get_quintile_for_geocodes_frame(small_data, 'geocode') # library method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>Unnamed: 0</th>\n",
       "      <th>race</th>\n",
       "      <th>gender</th>\n",
       "      <th>earnings_start</th>\n",
       "      <th>earnings_end</th>\n",
       "      <th>age</th>\n",
       "      <th>program</th>\n",
       "      <th>program_start</th>\n",
       "      <th>program_end</th>\n",
       "      <th>education_level</th>\n",
       "      <th>...</th>\n",
       "      <th>full_address</th>\n",
       "      <th>Address</th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>Zip</th>\n",
       "      <th>id</th>\n",
       "      <th>start_month</th>\n",
       "      <th>end_month</th>\n",
       "      <th>geocode</th>\n",
       "      <th>ses</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4264</th>\n",
       "      <td>4265</td>\n",
       "      <td>Other</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>32331.000</td>\n",
       "      <td>40334.030</td>\n",
       "      <td>34</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2011</td>\n",
       "      <td>2014</td>\n",
       "      <td>73</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>10003</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2377</th>\n",
       "      <td>2378</td>\n",
       "      <td>Other</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>27332.000</td>\n",
       "      <td>28858.775</td>\n",
       "      <td>27</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2018</td>\n",
       "      <td>2020</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>700 RIVER DRIVE FORT BRAGG CA 95437</td>\n",
       "      <td>700 RIVER DRIVE</td>\n",
       "      <td>FORT BRAGG</td>\n",
       "      <td>CA</td>\n",
       "      <td>95437.000</td>\n",
       "      <td>3435</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "      <td>060450105004</td>\n",
       "      <td>80-100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2570</th>\n",
       "      <td>2571</td>\n",
       "      <td>White</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>30780.000</td>\n",
       "      <td>44509.570</td>\n",
       "      <td>18</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2014</td>\n",
       "      <td>2016</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>11856</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11961</th>\n",
       "      <td>11962</td>\n",
       "      <td>Other</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>19847.000</td>\n",
       "      <td>18384.480</td>\n",
       "      <td>25</td>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2013</td>\n",
       "      <td>2018</td>\n",
       "      <td>123</td>\n",
       "      <td>...</td>\n",
       "      <td>2959 WEST 95TH STREET EVERGREEN PARK IL 60805</td>\n",
       "      <td>2959 WEST 95TH STREET</td>\n",
       "      <td>EVERGREEN PARK</td>\n",
       "      <td>IL</td>\n",
       "      <td>60805.000</td>\n",
       "      <td>1623</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>170318218002</td>\n",
       "      <td>40-60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14108</th>\n",
       "      <td>14109</td>\n",
       "      <td>Black</td>\n",
       "      <td>Male</td>\n",
       "      <td>37930.000</td>\n",
       "      <td>36852.200</td>\n",
       "      <td>35</td>\n",
       "      <td>French Literature</td>\n",
       "      <td>2015</td>\n",
       "      <td>2016</td>\n",
       "      <td>124</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>7927</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2699</th>\n",
       "      <td>2700</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Male</td>\n",
       "      <td>34055.000</td>\n",
       "      <td>35169.600</td>\n",
       "      <td>32</td>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2014</td>\n",
       "      <td>2015</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>8703</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14890</th>\n",
       "      <td>14891</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Male</td>\n",
       "      <td>41165.000</td>\n",
       "      <td>48969.800</td>\n",
       "      <td>29</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2012</td>\n",
       "      <td>2014</td>\n",
       "      <td>125</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>nan</td>\n",
       "      <td>15511</td>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11100</th>\n",
       "      <td>11101</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>20763.000</td>\n",
       "      <td>51185.640</td>\n",
       "      <td>36</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2012</td>\n",
       "      <td>2015</td>\n",
       "      <td>123</td>\n",
       "      <td>...</td>\n",
       "      <td>16600 SHERMAN WAY, STE 200 VAN NUYS CA 91406</td>\n",
       "      <td>16600 SHERMAN WAY, STE 200</td>\n",
       "      <td>VAN NUYS</td>\n",
       "      <td>CA</td>\n",
       "      <td>91406.000</td>\n",
       "      <td>12927</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>060371321023</td>\n",
       "      <td>0-20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>373</th>\n",
       "      <td>374</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>33631.000</td>\n",
       "      <td>41959.170</td>\n",
       "      <td>61</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2016</td>\n",
       "      <td>2020</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>321 FORTUNE BOULEVARD, SUITE 106B MILFORD MA 1757</td>\n",
       "      <td>321 FORTUNE BOULEVARD, SUITE 106B</td>\n",
       "      <td>MILFORD</td>\n",
       "      <td>MA</td>\n",
       "      <td>1757.000</td>\n",
       "      <td>13712</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>250277442003</td>\n",
       "      <td>20-40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>283</th>\n",
       "      <td>284</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>40174.000</td>\n",
       "      <td>54447.800</td>\n",
       "      <td>21</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2012</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2101 VISTA PARKWAY SUITE 102 WEST PALM BEACH F...</td>\n",
       "      <td>2101 VISTA PARKWAY SUITE 102</td>\n",
       "      <td>WEST PALM BEACH</td>\n",
       "      <td>FL</td>\n",
       "      <td>33411.000</td>\n",
       "      <td>5450</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>120990078362</td>\n",
       "      <td>40-60</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0                       race             gender  \\\n",
       "4264         4265                      Other  Other/unspecified   \n",
       "2377         2378                      Other  Other/unspecified   \n",
       "2570         2571                      White  Other/unspecified   \n",
       "11961       11962                      Other  Other/unspecified   \n",
       "14108       14109                      Black               Male   \n",
       "2699         2700  Asian or Pacific Islander               Male   \n",
       "14890       14891  Asian or Pacific Islander               Male   \n",
       "11100       11101                      White             Female   \n",
       "373           374                      White             Female   \n",
       "283           284                      White               Male   \n",
       "\n",
       "       earnings_start  earnings_end  age                 program  \\\n",
       "4264        32331.000     40334.030   34             Art History   \n",
       "2377        27332.000     28858.775   27             Art History   \n",
       "2570        30780.000     44509.570   18               Economics   \n",
       "11961       19847.000     18384.480   25            Anthropology   \n",
       "14108       37930.000     36852.200   35       French Literature   \n",
       "2699        34055.000     35169.600   32            Anthropology   \n",
       "14890       41165.000     48969.800   29             Art History   \n",
       "11100       20763.000     51185.640   36  Electrical Engineering   \n",
       "373         33631.000     41959.170   61             Art History   \n",
       "283         40174.000     54447.800   21             Art History   \n",
       "\n",
       "       program_start  program_end  education_level  ...  \\\n",
       "4264            2011         2014               73  ...   \n",
       "2377            2018         2020               60  ...   \n",
       "2570            2014         2016               60  ...   \n",
       "11961           2013         2018              123  ...   \n",
       "14108           2015         2016              124  ...   \n",
       "2699            2014         2015               60  ...   \n",
       "14890           2012         2014              125  ...   \n",
       "11100           2012         2015              123  ...   \n",
       "373             2016         2020                0  ...   \n",
       "283             2012         2018                0  ...   \n",
       "\n",
       "                                            full_address  \\\n",
       "4264                                                 NaN   \n",
       "2377                 700 RIVER DRIVE FORT BRAGG CA 95437   \n",
       "2570                                                 NaN   \n",
       "11961      2959 WEST 95TH STREET EVERGREEN PARK IL 60805   \n",
       "14108                                                NaN   \n",
       "2699                                                 NaN   \n",
       "14890                                                NaN   \n",
       "11100       16600 SHERMAN WAY, STE 200 VAN NUYS CA 91406   \n",
       "373    321 FORTUNE BOULEVARD, SUITE 106B MILFORD MA 1757   \n",
       "283    2101 VISTA PARKWAY SUITE 102 WEST PALM BEACH F...   \n",
       "\n",
       "                                 Address             City  State       Zip  \\\n",
       "4264                                 NaN              NaN    NaN       nan   \n",
       "2377                     700 RIVER DRIVE       FORT BRAGG     CA 95437.000   \n",
       "2570                                 NaN              NaN    NaN       nan   \n",
       "11961              2959 WEST 95TH STREET   EVERGREEN PARK     IL 60805.000   \n",
       "14108                                NaN              NaN    NaN       nan   \n",
       "2699                                 NaN              NaN    NaN       nan   \n",
       "14890                                NaN              NaN    NaN       nan   \n",
       "11100         16600 SHERMAN WAY, STE 200         VAN NUYS     CA 91406.000   \n",
       "373    321 FORTUNE BOULEVARD, SUITE 106B          MILFORD     MA  1757.000   \n",
       "283         2101 VISTA PARKWAY SUITE 102  WEST PALM BEACH     FL 33411.000   \n",
       "\n",
       "          id start_month end_month       geocode     ses  \n",
       "4264   10003          10         7                   NaN  \n",
       "2377    3435           9         5  060450105004  80-100  \n",
       "2570   11856           8         6                   NaN  \n",
       "11961   1623           8         7  170318218002   40-60  \n",
       "14108   7927           8         7                   NaN  \n",
       "2699    8703          10         5                   NaN  \n",
       "14890  15511           9         6                   NaN  \n",
       "11100  12927           9         7  060371321023    0-20  \n",
       "373    13712          10         5  250277442003   20-40  \n",
       "283     5450          10         5  120990078362   40-60  \n",
       "\n",
       "[10 rows x 24 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "small_data.sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 5: Check out SES equity concerns__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once you've classified individuals into SESs, you can use the SES as a grouping variable-- another axis on which to investigate equity concerns. Here we use the `Theil_T()` method from the `equity` submodule. As it is an instance of the `Metric()` parent class, it has all of the same attributes and methods. You can read more about the Theil index on [Wikipedia](https://en.wikipedia.org/wiki/Theil_index)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "theil_t = equity.Theil_T.from_dataframe(small_data, 'ses', 'earnings_end') # library method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "theil_t.calculate()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxMAAAGGCAYAAAAAbtyJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de5RsV10n8O8v9yZdfUMCrBGSQUEZ0MDwEIgPwlJAmNiCYtARZcAlIIowQRFBMMsX4gMRB8NIRNEg/gHqYnCRpRlpDRrCSGQcQCBAQDA8AiQIgSRgqkOSPX9U9U2l031v977dfere/nzWqlWnztl16lf77q6ub599zq3WWgAAALbquKELAAAAjk7CBAAA0EWYAAAAuggTAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHTZP3QB86aqKsndklw/dC0AADCgk5J8uh3if7kWJm7vbkmuHLoIAACYA1+T5FMbbRQmbu/6JPnkJz+Zk08+eehaAABg11133XW5+93vnhxmto4wsYGTTz5ZmAAAgENwAjYAANBFmAAAALoIEwAAQBdhAgAA6OIEbACYQ621jMfjoctgQK21rKysJEkWFhYy+a+whjUajeaiDuaHMAEAc2g8HmdpaWnoMuA2lpeXs7i4OHQZzBHTnAAAgC6OTADAnPvyQ56cHOdX9p5z81dy4rtfnyT58oOflOw7fpg6brkpJ77rdcO8NnPPJxMAzLvj9g/3RZL5sO94Y4C5ZJoTAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOiyf+gCAFhfay3j8ThJMhqNUlUDVwQwHJ+J88mRCYA5NR6Ps7S0lKWlpYO/QAH2Kp+J80mYAAAAuggTAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHQRJgAAgC7CBAAA0EWYAAAAuggTAABAF2ECAADoIkwAAABdhAkAAKDLoGGiqs6uqo9V1biq3lFV33KItl9XVedX1RVVdUNVfbSqfqWqTljT7oFV9bbpPj9ZVS/Y+XcCAAB7z2Bhoqp+KMnLk/xKkockeU+S5aq66wZPuU8m9f5EkvsleW6SZyb5jZl9npzkb5J8PMnpSX42yYuq6hk79DYAAGDP2j/ga/9Mkj9srf1xklTVM5N8d5IfTfKbaxu31t6c5M0zq/61qk5L8qwkz5+ue3KSE5L8aGvtxiTvr6oHTV/r1Tv1RgB2Qmvt4PJ4PB6wEoZwm3/zmbEAu25OPotmX7v5mZgbg4SJ6dSk05O8ZHVda+2WqrooyRlb2NUdk1wz8/iMJJdMg8Sq5SQvrKo7t9a+sE4tC0kWZladtIXXB9gxKysrB5fPOuusASthcLfclMnfymAAt9x0cHFePotWVlZy4MCBocsgw01z+qok+5JcvWb91UlO3cwOqureSX4yyR/MrD51g32ublvPOUmunblduZnXBwCAvW7IaU4bqqrfT/LDq49ba3dYs/2rM5ny9IbW2h8e4cu9JJNzN1adFIECmAMLC7ceNL3gggsyGo0GrIbdNh6Pb/0r8HFz+euavWJm/A35WTT7MzH7+ciwhvp0+lySm5Ocsmb9KUmuSvJLSX57vSdW1d2S/H2StydZe2L1VRvsc3Xb7bTWVpIcnEtQVYevHmAXzH4ejUajLC4uDlgNg/K7iSHN4WeR72vzY5BpTtNzGt6Z5NGr66rquOnjS1trn22tfWT1NtPmq5NcPH3u01prt6zZ9aVJHl5Vx8+sOzPJh9Y7XwIAAOg35P8z8fIkP15VT6mq+yZ5VZITk/zxeo1ngsQnMrl6012q6tSqmj0X4vVJbkxyflXdb3r52efkttOYAACAbTDYJMzW2p9X1V2SvDiTk6P/Ocl3tdbWnkC96swk957e1p7TUNN9XltV35nkvEyOXnwuyYtbay4LCwAA22zQM7paa69M8spNtn1tktduot17k3z7ERUGAAAc1pDTnAAAgKOYMAEAAHQRJgAAgC7CBAAA0EWYAAAAuggTAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHQRJgAAgC7CBAAA0GX/0AUAsL7RaJTl5eWDywB7mc/E+SRMAMypqsri4uLQZQDMBZ+J88k0JwAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQZf/QBQAAh3HLTUNXwBBu/sr6y7vN+OMQhAkAmHMnvut1Q5fAwE589+uHLgHWZZoTAADQxZEJAJhDo9Eoy8vLQ5fBgFprWVlZSZIsLCykqgauaDIuYZYwAQBzqKqyuLg4dBkM7MCBA0OXAIdkmhMAANBFmAAAALoIEwAAQBdhAgAA6CJMAAAAXYQJAACgizABAAB0ESYAAIAuwgQAANBFmAAAALoIEwAAQBdhAgAA6LJ/6AIAgPnTWst4PB66jD2jtZaVlZUkycLCQqqqaz+j0aj7udBDmAAAbmc8HmdpaWnoMtii5eXlLC4uDl0Ge4hpTgAAQBdHJgCAQzrv4V/Mwr42dBnHtJWbk7MvuXOS5LyHfyEL+7by3MrZl9xphyqDQxMmAIBDWtjXMtrCl1uOzMK+bLG/BT2GY5oTAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF32D10AAPOhtZbxeJwkGY1GqaqBKwI2o7XZ5bZxQ9gBjkwAkCQZj8dZWlrK0tLSwVABzL8bb7l1eWVlZbhC2JOECQAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQZdvDRFWdU1X/VFXXV9Vnq+pNVXXamjajqjqvqj5fVV+qqjdW1SlbeI17T/f/xXW2PaGqLq+qcVW9r6oeux3vCwAAuK2dODLxiCTnJXlokjOTHJ/kb6rqxJk2v5PkcUmeMG1/tyR/sZmdV9XxSf40ydvW2faw6bbzkzw4yZuSvKmq7t/7ZgAAgPXt3+4dtta+a/ZxVT01yWeTnJ7kkqq6Y5KnJ3lSa+3vpm2eluSDVfXQ1to/HuYlfi3J5UnekuRha7Y9J8mbW2svmz7+xao6M8mzkzyz/10BAABrbXuYWMcdp/fXTO9Pz+RoxUWrDVprl1fVJ5KckWTDMFFVj8rkaMaDknz/Ok3OSPLyNeuWkzy+q3KAPaS1dnB5PB4PWAnzYHYMzAwNgNvY0TBRVcclOTfJP7TWLpuuPjXJja21tec7XD3dttG+/kOS1yb54dbadVW1XrNTp/vZyn4XkizMrDppo7YAx7KVlZWDy2edddaAlTBvbrwlWRy6CGAu7fTVnM5Lcv8kT9zKk6rq/dMTs79UVX89Xf2HSV7fWrtkm2s8J8m1M7crt3n/AABwTNqxIxNV9cok35Pk4a212S/oVyU5oarutOboxCnTbUny2EymQiXJDdP7RyX53qp6/upLJDmuqm5K8ozW2mumz197VajZ/a7nJbnt1KiTIlAAe9DCwq0HaS+44IKMRqMBq2Fo4/H44BGqE1xIHtjAtoeJmsw/+t0k35fkka21K9Y0eWeSryR5dJI3Tp9zWpJ7JLk0SVprH19n12ck2Tfz+KwkL8zkJOxPTdddOt3vuTPtzlzd73paaytJDh7b32D6FMAxb/bzbzQaZXHRxBYm/GoENrITRybOS/KkTL7sX19Vq+crXNtau6G1dm1VnZ/k5VV1TZLrMgkflx7qSk6ttQ/OPq6qb0pyy8y5GEnyiiRvrarnJbkwk+lV35TkGdv03gAAgKmdOHD5rEyu4HRxks/M3H5ops1zk/xVJkcmLslkGtJ6V2faktba2zMJMs9I8p4kP5Dk8WsCBwAAsA124v+ZOOzB0NbaOMnZ01vv67w2k6s7rV3/hiRv6N0vAACwOU6pAgAAuggTAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHQRJgAAgC7CBAAA0EWYAAAAuggTAABAF2ECAADoIkwAAABd9g9dAADzYTQaZXl5+eAycHQ4YeZPwwsLC8MVwp4kTACQJKmqLC4uDl0GsEVVs8u1cUPYAaY5AQAAXYQJAACgizABAAB0ESYAAIAuwgQAANBFmAAAALoIEwAAQBdhAgAA6CJMAAAAXYQJAACgizABAAB0ESYAAIAu+4cuAACYbys3V5I2dBnHtJWb11/e3HNre4uBLRAmAIBDOvuSOw1dwp5y9iV3HroE2DTTnAAAgC6OTAAAtzMajbK8vDx0GXtGay0rKytJkoWFhVT1TV0ajUbbWRYcljABANxOVWVxcXHoMvaUAwcODF0CbJlpTgAAQBdhAgAA6CJMAAAAXYQJAACgizABAAB0ESYAAIAuwgQAANBFmAAAALoIEwAAQBdhAgAA6CJMAAAAXYQJAACgizABAAB02T90ATCk1lrG4/HQZcCuaq1lZWUlSbKwsJCqGrgimA+j0cjPA2yRMMGeNh6Ps7S0NHQZAMyB5eXlLC4uDl0GHFVMcwIAALo4MgFTNz/uZj8R7A03Jfv+cl8S4x5mfx6ArfMrBFbtj58I9h7jHoAjYJoTAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOiyf+gCuFVrLePxOEkyGo1SVQNXBADAbjravg86MjFHxuNxlpaWsrS0dHAQAQCwdxxt3weFCQAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQRZgAAAC67HiYqKqfq6pWVefOrBtV1XlV9fmq+lJVvbGqTtnEvqqqnl9VH66qlar6VFX9/Jo2j6yqd023f6SqnroDbwsAAPa8HQ0TVfXNSX4iyXvXbPqdJI9L8oQkj0hytyR/sYldviLJjyV5fpL7JPneJP935vXumeTCJH+f5EFJzk3yR1W1dERvBAAAuJ39O7XjqrpDktcl+fEkvzCz/o5Jnp7kSa21v5uue1qSD1bVQ1tr/7jB/u6b5FlJ7t9a+9B09RVrmj0zyRWttedNH3+wqr4tyXOTLG/PO9s5rbWDy+PxeMBK9o7b9HPbuB0Ax6iZz36/e5kHs+Nw9rvhvNqxMJHkvCQXttYuqqpfmFl/epLjk1y0uqK1dnlVfSLJGUnWDROZHMn41yTfU1VvTlLTfbygtXbNtM0Zs/udWs7kCMW6qmohycLMqpMO98Z2ysrKysHls846a6gy9q6bMxmZAOwdN9+66Hcv82ZlZSUHDhwYuoxD2pFpTlX1xCQPSXLOOptPTXJja+2La9ZfPd22kf+U5GszmRr1I0memkkw+V9r9n31Ovs9uaoWN9jvOUmunbldeYgaAACAqW0/MlFVd8/k3IYzW2tdxwur6v2ZBIckeVtr7TGZBJ+FJD/SWvvwtN3Tk7yzqk6bmfq0VS9J8vKZxydloECxsHDrAZILLrggo9FoiDL2lPF4fOtfovYNWwsAA5j57Pe7l3kw+91k9rvhvNqJaU6nJ7lrkndV1eq6fUkeXlXPTrKU5ISqutOaoxOnJLlquvzY3Drh5Ibp/WeS3LQaJKY+OL2/R5IPTZ+/9qpQpyS5rrV2Q9bRWltJcnB+0UzNu272tUejURYXNzqYwo4Y7p8egKHMfPb73cu8GfJ76WbtRJh4S5IHrFn3x0kuT/LSJJ9M8pUkj07yxiSpqtMyCQSXJklr7ePr7Pcfkuyvqnu11j46XfcN0/vV9pdmEkRmnbm6XwAAYPtse5horV2f5LLZdVX15SSfb61dNn18fpKXV9U1Sa5L8rtJLt3oSk5TFyV5V5LXVNVPZzLt6bwkfztztOL3kzy7qn4ryWuSPCrJDyb57u16fwAAwMRQ/wP2c5P8VSZHJi7JZHrS9x/qCa21WzK5otPnps+5MJNpTk+caXNFJsHhzCTvSfK8JD/WWpv7y8ICAMDRZicvDXtQa+2Rax6Pk5w9vW1lP59O8l8P0+biJA/eWoUAAMBWDXVkAgAAOMoJEwAAQBdhAgAA6CJMAAAAXYQJAACgizABAAB0ESYAAIAuwgQAANBFmAAAALoIEwAAQBdhAgAA6CJMAAAAXfYPXQC3Go1GWV5ePrgMAMDecrR9HxQm5khVZXFxcegyAAAYyNH2fdA0JwAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQZf/QBcDcuGnoAmCX3LTBMuxFfgbgiAgTMLXvL/cNXQLsOuMegCNhmhMAANDFkQn2tNFolOXl5aHLgF3VWsvKykqSZGFhIVU1cEUwH0aj0dAlwFFHmGBPq6osLi4OXQbsugMHDgxdAgDHANOcAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAECX/UMXAEy01jIej4cuY1e01rKyspIkWVhYSFUNXBHMt9Fo5OcEmEvCBMyJ8XicpaWlocsA5tDy8nIWFxeHLgPgdkxzAgAAujgyAXPohUlOGLqIHXRjkpdOl4/19wq9Zn9OAOaVMAFz6IQkJ+RYnh/dDi4d++8VerXDNwEYmGlOAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHQRJgAAgC7CBAAA0EWYAAAAuggTAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHTZP3QBsJHWWsbjcZJkNBqlqgauCACYN74vDMuRCebWeDzO0tJSlpaWDn5IAADM8n1hWMIEAADQRZgAAAC6CBMAAEAXYQIAAOgiTAAAAF2ECQAAoIswAQAAdBEmAACALsIEAADQRZgAAAC6CBMAAEAXYQIAAOiypTBRVfuq6ler6oqquqGqPlpVv1hVNdOmqurFVfWZaZuLqurrN7Hv/1lV76yqlar65w3aPLCq3lZV46r6ZFW9YJ02T6iqy6dt3ldVj93KewQAADZnq0cmXpjkWUmeneS+08cvSPKTM21ekOSnkjwzybcm+XKS5aoabWL/r0ny5+ttqKqTk/xNko8nOT3JzyZ5UVU9Y6bNw5L8aZLzkzw4yZuSvKmq7r/5twgAAGzG/i22f1iSC1prF04ff6yq/luSb0kmRyWS/HSSX2utXTBd9yNJrk7y+CR/ttGOW2s/NW1/lyQPXKfJk5OckORHW2s3Jnl/VT0oyc8kefW0zXOSvLm19rLp41+sqjMzCT/P3OJ7BQAADmGrYeLtSZ5RVd/QWvtwVX1jkm/L5At9ktwzyalJLlp9Qmvt2qp6R5IzcogwsQlnJLlkGiRWLSd5YVXdubX2hWmbl6953nImQYajTGvt4PJ4PB6wkt0x+x7bIdoBe8Ps58Be+AyEXrf5/dn8Bt1tWw0Tv5nk5CSXV9XNSfYl+fnW2uum20+d3l+95nlXz2zrdWqSK9bZ7+q2L0zvt/TaVbWQZGFm1UlHVibbZWVl5eDyWWedNWAlu+8rue2gBPaer8ws77XPQOi1srKSAwcODF3GnrLVcyZ+MJPpRk9K8pAkT0ny/Kp6ymZ3UFV/XVVfmt7ev8XX3wnnJLl25nblsOUAAMDRYatHJl6W5Ddba6vTld5XVV+byRfyP0ly1XT9KUk+M/O8U5KsXqHpx5IsTpdn//ByOFdN9zPrlJlth2pzVTb2ktx2atRJESjmwsLCrX+bv+CCCzIabeYc/qPXeDw++NfH4weuBRje7OfAXvgMhF6zvz9nvzuwO7YaJg4kuWXNuptz6xGOKzL54v7oTMPD9CpM35rkVUnSWvtUZ62XJvn1qjq+tbYaQs5M8qHp+RKrbR6d5NyZ5505Xb+u1tpKkoPzaWaucsvAZv8tRqNRFhcXD9H62GIUArOfA3vtMxB6+R63+7YaJv4yyc9X1SeSvD+Ty6/+TCaXdE1rrVXVuUl+oar+JZNw8atJPp3JZVo3VFX3TnKHTM5vWJxeqSlJPjA96fr1SX45yflV9dIk98/k6k3PndnNK5K8taqel+TCJE9M8k1JnhEAAGBbbTVM/GQm4eD3ktw1k5DwB0lePNPmt5KcmMnlWu+U5P8k+a7W2uEuRfFHSR4x8/jd0/t7JvnY9KpQ35nkvCTvTPK5JC9ura1eFjattbdX1ZOS/FqS30jyL0ke31q7bIvvEwAAOIwthYnW2vWZ/D8SP32INi3JL01vW9n3IzfR5r1Jvv0wbd6Q5A1beW0AAGDrtno1JwAAgCTCBAAA0EmYAAAAuggTAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHQRJgAAgC7CBAAA0EWYAAAAuggTAABAl/1DFwAbGY1GWV5ePrgMALCW7wvDEiaYW1WVxcXFocsAAOaY7wvDMs0JAADoIkwAAABdhAkAAKCLMAEAAHQRJgAAgC7CBAAA0EWYAAAAuggTAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHQRJgAAgC77hy4AuL0bkyRt4Cp2zo23Wz523yv0uvHwTQAGJ0zAHHrp0AXsor30XgHgWGOaEwAA0MWRCZgTo9Eoy8vLQ5exK1prWVlZSZIsLCykqgauCObbaDQaugSAdQkTMCeqKouLi0OXsWsOHDgwdAkAwBEyzQkAAOgiTAAAAF2ECQAAoIswAQAAdHEC9gauu+66oUsAAIBBbPa7cLXmf56dVVVfneTKoesAAIA58DWttU9ttFGYWKMmF7y/W5Lrh65lF52USYD6muyt972b9PHO08e7Qz/vPH28O/TzztPHO2+n+/ikJJ9uhwgMpjmtMe2sDdPXsWjmPwy7vrVmftcO0Mc7Tx/vDv288/Tx7tDPO08f77xd6OPD7tMJ2AAAQBdhAgAA6CJMkCQrSX5les/O0Mc7Tx/vDv288/Tx7tDPO08f77zB+9gJ2AAAQBdHJgAAgC7CBAAA0EWYAAAAuggTAABAF2HiKFRV51TVP1XV9VX12ap6U1WdtqbNqKrOq6rPV9WXquqNVXXKmjb3qKoLq+rfp/t5WVXtX9PmkVX1rqpaqaqPVNVT16nn7Kr6WFWNq+odVfUtO/LGd9Em+/jiqmprbr+/po0+PoSqelZVvbeqrpveLq2qx8xsN46P0Cb62DjeZlX1c9N+PHdmnbG8jTboY2P5CFXVi9bpw8tnthvH22AT/Xx0jeXWmttRdkvy5iRPTXK/JN+Y5MIkH09y4kybVyX5RJJHJTk9yaVJ/mFm+74k70vyt0kelOQxSf4tyW/MtLlnki8n+R9J7pvk2UluSrI00+aHMrkc2dOS/Ockr07yhSR3HbqfdqGPL56+31Nnbifr4y318+OSPDbJ1yf5hiS/nuTGJPczjnetj43j7e3vb05yRZL3JDl3Zr2xvPN9bCwfed++KMlla/rwq4zjXe/no2osD96hbtsyKO+SpCV5+PTxHTP5svADM23uM23z0OnjxyS5OckpM22emeTaJCdMH780yWVrXuvPkrx55vE7krxy5vFxST6V5OeG7ped7OPpuosz84tsnefo476+vibJ043jne/j6bJxvH39eockH07yX2b71Vje+T6ebjOWj7x/X5TknzfYZhzvQj9Ptx9VY9k0p2PDHaf310zvT09yfJKLVhu01i7P5K8JZ0xXnZHkfa21q2f2s5zk5Ez+Gr/a5qLc1vLqPqrqhOlrzb7OLdPHZ+TYsraPVz25qj5XVZdV1Uuq6sDMNn28BVW1r6qemOTETP7aZRxvs3X6eJVxvD3OS3Jha21tXxjL22ejPl5lLB+5r6+qT1fVv1bV66rqHtP1xvH22qifVx01Y3n/4Zswz6rquCTnZnKY8bLp6lOT3Nha++Ka5ldPt622uXqd7dlEm5OrajHJnTM51LZem/ts8a3MrQ36OElen8nUp08neWAmfwU4Lcn3T7fr402oqgdk8sV2lORLSb6vtfaBqnpQjONtsVEfTzcbx9tgGtIekskUnLV8Jm+Dw/RxYixvh3dkMsX3Q0n+Y5JfTvK2qrp/jOPttGE/t9auz1E2loWJo995Se6f5NuGLuQYtm4ft9ZePfPwfVX1mSRvqap7tdY+upsFHuU+lMmczzsm+YEkf1JVjxi2pGPOun3cWvuAcXzkquruSV6R5MzW2njoeo5Fm+ljY/nItdb+eubhe6vqHZl8qf3BJDcMU9Wx5zD9fP7RNpZNczqKVdUrk3xPku9orV05s+mqJCdU1Z3WPOWU6bbVNqessz2baHNda+2GJJ/LdM7eIV7nqHaIPl7PO6b3957e6+NNaK3d2Fr7SGvtna21czI5qfI5MY63zSH6eD3G8dadnuSuSd5VVTdV1U1JHpHkp6bLV8dYPlKH7OOq2rfOc4zlIzQ9CvHhTPrQZ/IOWdPP65nrsSxMHIVq4pVJvi/Jo1prV6xp8s4kX0ny6JnnnJbkHrl1nvSlSR5QVXeded6ZSa5L8oGZNo/ObZ25uo/W2o3T15p9neOmjy/NUWwTfbyeB03vPzO918d9jkuyEON4J6328XqM4617S5IHZNJ3q7f/l+R1M8vG8pE5ZB+31m5e5znG8hGqqjskuVcmfegzeYes6ef1zPdY3s6z091255bk95J8MZO/ysxeNmxxps2rMjlk9h2Z/EXn7UnePrN99bJiy5lc+nQpyWez/mXFfiuT+XP/PetfVmyc5CmZXHrsDzK5rNgpO90PQ/ZxJj/0vzjt269L8r1JPprkrfp4S/38kiQPn/bhA6aPb8lkKoNxvMN9bBzvaL9fnNtfGtZY3qE+Npa3rU9/O5Pfe1+X5GGZXHr035LcxTjenX4+Gsfy4B3q1jUI2wa3p860GWUy1/+a6WD6iySnrtnP1yb530n+fTqIfzvJ/jVtHpnk3Zlch/ijs68x0+bZ0w+XlUwOxX3r0H20032c5O5J3prk89MfxH+Z/sCevGY/+vjQ/Xx+ko9N39dnM7mKxJnG8e70sXG8o/1+cW4bJozlHbzSQIAAAABsSURBVOxjY3nb+vTPMjnpdyXJldPH9zKOd6+fj8axXNMdAQAAbIlzJgAAgC7CBAAA0EWYAAAAuggTAABAF2ECAADoIkwAAABdhAkAAKCLMAEAAHQRJgAAgC7CBAAA0EWYAAAAuggTAABAl/8P41loKYM1VGgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "theil_t.viz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.20680992161391945"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "theil_t.ratio"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Again, the ratio attribute here tells the analyst how much variation is attributable to SES."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 6: Calculate graduate wage summary statistics__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another useful think that `roi-toolkit` allows you to do is adjust for inflation. This can be very important when you have multiple years of data. Students' starting earnings will typically be expressed in nominal dollars. Over a long enough timespan, this can make a big difference for equity and return calculations. A student making \\\\$20,000 in 2010 was making almost \\\\$24,000 in 2020 dollars - a 20% difference."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.lines.Line2D at 0x7fc331746110>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Nothing toolkit-specific here. Just getting the distribution for unadjusted wages\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "data['earnings_end'].hist(ax=ax, edgecolor=\"red\", fc=(0, 0, 0, 0.5), bins=30)\n",
    "plt.axvline(x=np.mean(data['earnings_end']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `utilities.multiple_describe()` is a useful wrapper for Pandas grouping and aggregation methods. It takes a dataframe, a grouping column, and a value column, and returns a dataframe containing summary statistics for the provided value at the group level."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>program</th>\n",
       "      <th>n</th>\n",
       "      <th>mean</th>\n",
       "      <th>median</th>\n",
       "      <th>sd</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2288.000</td>\n",
       "      <td>31540.395</td>\n",
       "      <td>31381.600</td>\n",
       "      <td>6542.271</td>\n",
       "      <td>2817.240</td>\n",
       "      <td>63547.440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Art History</td>\n",
       "      <td>2370.000</td>\n",
       "      <td>42312.370</td>\n",
       "      <td>41632.800</td>\n",
       "      <td>9460.368</td>\n",
       "      <td>4287.800</td>\n",
       "      <td>75483.430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Biology</td>\n",
       "      <td>2404.000</td>\n",
       "      <td>35848.828</td>\n",
       "      <td>35303.220</td>\n",
       "      <td>7838.359</td>\n",
       "      <td>6602.715</td>\n",
       "      <td>64873.710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Economics</td>\n",
       "      <td>2327.000</td>\n",
       "      <td>54705.240</td>\n",
       "      <td>54158.720</td>\n",
       "      <td>11048.693</td>\n",
       "      <td>14172.620</td>\n",
       "      <td>102118.940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2386.000</td>\n",
       "      <td>65698.580</td>\n",
       "      <td>64849.900</td>\n",
       "      <td>14206.740</td>\n",
       "      <td>20139.900</td>\n",
       "      <td>135887.290</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>French Literature</td>\n",
       "      <td>2339.000</td>\n",
       "      <td>27875.583</td>\n",
       "      <td>27687.748</td>\n",
       "      <td>5501.041</td>\n",
       "      <td>6088.635</td>\n",
       "      <td>50285.620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Philosophy</td>\n",
       "      <td>2297.000</td>\n",
       "      <td>46128.218</td>\n",
       "      <td>45486.000</td>\n",
       "      <td>10275.570</td>\n",
       "      <td>14191.200</td>\n",
       "      <td>96564.600</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  program        n      mean    median        sd       min  \\\n",
       "0            Anthropology 2288.000 31540.395 31381.600  6542.271  2817.240   \n",
       "1             Art History 2370.000 42312.370 41632.800  9460.368  4287.800   \n",
       "2                 Biology 2404.000 35848.828 35303.220  7838.359  6602.715   \n",
       "3               Economics 2327.000 54705.240 54158.720 11048.693 14172.620   \n",
       "4  Electrical Engineering 2386.000 65698.580 64849.900 14206.740 20139.900   \n",
       "5       French Literature 2339.000 27875.583 27687.748  5501.041  6088.635   \n",
       "6              Philosophy 2297.000 46128.218 45486.000 10275.570 14191.200   \n",
       "\n",
       "         max  \n",
       "0  63547.440  \n",
       "1  75483.430  \n",
       "2  64873.710  \n",
       "3 102118.940  \n",
       "4 135887.290  \n",
       "5  50285.620  \n",
       "6  96564.600  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utilities.multiple_describe(data, 'program', 'earnings_end') # library method"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Adjust wages for inflation__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Methods to adjust for inflation and fetch historical time series are in the `macro.BLS_Ops()` class, since the data necessary to make these calculations comes originally from the U.S. Bureau of Labor Statistics API. Instantiating the class takes a moment, since it has to read in the BLS output that comes packaged with the module (and that can be refreshed with `setup.py`). \n",
    "\n",
    "The `bls.adjust_to_current_dollars()` method takes a dataframe, a year column in format YYYY (in this case `program_end`) and a value column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Latest CPI year in provided BLS data is 2018.0: All dollars being adjusted to 2018.0 dollars.\n",
      "Years in provided dataframe for which there is no data in the provided CPI frame:\n",
      "\n",
      "{2019, 2020, 2021, 2022, 2023, 2024, 2025, 2026, 2027, 2028}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mattlerner/roi/sbir-roi/roi/macro.py:58: UserWarning: Value column earnings_end contains 821 NA values (5.0%) of total.\n",
      "  warnings.warn(\"Value column {} contains {} NA values ({}%) of total.\".format(value_column_name, value_nas, round(100*value_nas/len(frame_),2)))\n",
      "/Users/mattlerner/roi/sbir-roi/roi/macro.py:71: UserWarning: 5515 rows in column program_end could not be merged with provided CPI data. Please note that (1) the BLS API provides only up to 20 years of data; if you want to use more, you will have to manually combine multiple queries. (2) We do not recommend using more than ten years of historical data in calculations.\n",
      "  warnings.warn(\"{} rows in column {} could not be merged with provided CPI data. Please note that (1) the BLS API provides only up to 20 years of data; if you want to use more, you will have to manually combine multiple queries. (2) We do not recommend using more than ten years of historical data in calculations.\".format(unmerged_len, year_column_name))\n"
     ]
    }
   ],
   "source": [
    "bls = macro.BLS_Ops()\n",
    "data['earnings_end_adjusted'] = bls.adjust_to_current_dollars(data, 'program_end', 'earnings_end') # library method"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once again, the module provides a verbose response, warning the user when the data contains years that can't be adjusted (in this case because they're bizarrely in the future-- this can be a data quality issue!)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>program_end</th>\n",
       "      <th>earnings_end</th>\n",
       "      <th>earnings_end_adjusted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7580</th>\n",
       "      <td>2017</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13231</th>\n",
       "      <td>2017</td>\n",
       "      <td>56959.540</td>\n",
       "      <td>58349.149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12584</th>\n",
       "      <td>2022</td>\n",
       "      <td>62106.000</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7719</th>\n",
       "      <td>2018</td>\n",
       "      <td>39197.950</td>\n",
       "      <td>39197.950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>684</th>\n",
       "      <td>2019</td>\n",
       "      <td>41903.400</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1038</th>\n",
       "      <td>2018</td>\n",
       "      <td>24618.040</td>\n",
       "      <td>24618.040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11612</th>\n",
       "      <td>2015</td>\n",
       "      <td>54890.220</td>\n",
       "      <td>58155.739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16387</th>\n",
       "      <td>2014</td>\n",
       "      <td>41328.360</td>\n",
       "      <td>43840.102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15165</th>\n",
       "      <td>2020</td>\n",
       "      <td>42124.600</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14421</th>\n",
       "      <td>2015</td>\n",
       "      <td>40346.400</td>\n",
       "      <td>42746.681</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       program_end  earnings_end  earnings_end_adjusted\n",
       "7580          2017           nan                    nan\n",
       "13231         2017     56959.540              58349.149\n",
       "12584         2022     62106.000                    nan\n",
       "7719          2018     39197.950              39197.950\n",
       "684           2019     41903.400                    nan\n",
       "1038          2018     24618.040              24618.040\n",
       "11612         2015     54890.220              58155.739\n",
       "16387         2014     41328.360              43840.102\n",
       "15165         2020     42124.600                    nan\n",
       "14421         2015     40346.400              42746.681"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[['program_end','earnings_end','earnings_end_adjusted']].sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Recalculate average graduate wage by program__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we plot the distribution of nominal wages against inflation-adjusted wages. As you can see, the mean wage doesn't change too much -- but the overall distribution does change, which is important. If the composition of programs has changed over time, inflation adjustment can be vital for drawing conclusions about equity. For example, if a program has substantially more low-SES students than it did ten years ago, failing to adjust for inflation could make those students appear better off than they actually are (because they're being compared to nominally lower-earning students many years ago)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.lines.Line2D at 0x7fc3352e79d0>"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "# unadjusted\n",
    "data['earnings_end'].hist(ax=ax, edgecolor=\"red\", fc=(0, 0, 0, 0), bins=30)\n",
    "plt.axvline(x=np.mean(data['earnings_end']), color=\"red\", linestyle=\":\")\n",
    "\n",
    "# adjusted\n",
    "data['earnings_end_adjusted'].hist(ax=ax, edgecolor=\"blue\", fc=(0, 0, 0, 0), bins=30)\n",
    "plt.axvline(x=np.mean(data['earnings_end_adjusted']), color=\"blue\", linestyle=\":\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 7: Earnings premium__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Calculate earnings premium - Simple way__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The most simple earnings-related measurement that analysts might be concerned with is the __earnings premium__, defined here as the additional wages attributable to participation in a given program. Here we do that in the simplest way possible: by subtracting nominal post-program earnings from nominal pre-program earnings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['earnings_change'] = data['earnings_end'] - data['earnings_start']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>program</th>\n",
       "      <th>n</th>\n",
       "      <th>mean</th>\n",
       "      <th>median</th>\n",
       "      <th>sd</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2288.000</td>\n",
       "      <td>1492.413</td>\n",
       "      <td>1748.180</td>\n",
       "      <td>7514.140</td>\n",
       "      <td>-46000.200</td>\n",
       "      <td>41005.440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Art History</td>\n",
       "      <td>2370.000</td>\n",
       "      <td>13117.589</td>\n",
       "      <td>13010.150</td>\n",
       "      <td>8743.259</td>\n",
       "      <td>-49156.200</td>\n",
       "      <td>54181.100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Biology</td>\n",
       "      <td>2404.000</td>\n",
       "      <td>5538.174</td>\n",
       "      <td>5503.500</td>\n",
       "      <td>7822.369</td>\n",
       "      <td>-28674.700</td>\n",
       "      <td>54911.950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Economics</td>\n",
       "      <td>2327.000</td>\n",
       "      <td>24685.001</td>\n",
       "      <td>24170.200</td>\n",
       "      <td>10529.107</td>\n",
       "      <td>-15061.260</td>\n",
       "      <td>77500.780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2386.000</td>\n",
       "      <td>36063.777</td>\n",
       "      <td>35277.400</td>\n",
       "      <td>13514.743</td>\n",
       "      <td>-21673.600</td>\n",
       "      <td>112955.900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>French Literature</td>\n",
       "      <td>2339.000</td>\n",
       "      <td>-2656.493</td>\n",
       "      <td>-2261.950</td>\n",
       "      <td>6824.586</td>\n",
       "      <td>-62843.250</td>\n",
       "      <td>36069.590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Philosophy</td>\n",
       "      <td>2297.000</td>\n",
       "      <td>16914.652</td>\n",
       "      <td>16949.000</td>\n",
       "      <td>9296.036</td>\n",
       "      <td>-22386.120</td>\n",
       "      <td>61526.600</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  program        n      mean    median        sd        min  \\\n",
       "0            Anthropology 2288.000  1492.413  1748.180  7514.140 -46000.200   \n",
       "1             Art History 2370.000 13117.589 13010.150  8743.259 -49156.200   \n",
       "2                 Biology 2404.000  5538.174  5503.500  7822.369 -28674.700   \n",
       "3               Economics 2327.000 24685.001 24170.200 10529.107 -15061.260   \n",
       "4  Electrical Engineering 2386.000 36063.777 35277.400 13514.743 -21673.600   \n",
       "5       French Literature 2339.000 -2656.493 -2261.950  6824.586 -62843.250   \n",
       "6              Philosophy 2297.000 16914.652 16949.000  9296.036 -22386.120   \n",
       "\n",
       "         max  \n",
       "0  41005.440  \n",
       "1  54181.100  \n",
       "2  54911.950  \n",
       "3  77500.780  \n",
       "4 112955.900  \n",
       "5  36069.590  \n",
       "6  61526.600  "
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utilities.multiple_describe(data, 'program', 'earnings_change') # library method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fc335404590>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=\"earnings_change\", y=\"program\", data=data, estimator=np.mean)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Calculate earnings premium - Fancy way__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One of the central concerns about calculating premiums in the above \"simple\" way is the necessity of adjusting for macroeconomic trends. If the average difference between post- and pre-program earnings for participants in a given program is 4%, you might conclude that the program improves earning power by 4%. But if, over the time students spent in a program, overall wages in their region go up by 5%, then in fact the program has actually \"hurt\" students' earning power by 1%.\n",
    "\n",
    " `roi-toolkit` offers a method to simply correct for trend. The `BLS_Ops.wage_change()` method takes in a dataframe of student states, program start and end dates (in 'YYYY-MM' format), and for each student returns the macroeconomic change in wages for their region over the time period provided.\n",
    " \n",
    "Analysts should be mindful that the `wage_change()` method "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['year_month_entry'] = data['program_start'].astype(str) + '-' + data['start_month'].astype(str).str.zfill(2)\n",
    "data['year_month_exit'] = data['program_end'].astype(str) + '-' + data['end_month'].astype(str).str.zfill(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In many contexts, you'll need state FIPS codes (e.g. \"08\") instead of postal codes (e.g. \"CO\"). This is one of those contexts. Luckily, the toolkit offers the `utilities.State_To_FIPS_series()` method to help you easily translate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['state_fips'] = utilities.State_To_FIPS_series(data['State'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mattlerner/roi/sbir-roi/roi/macro.py:202: UserWarning: Series passed as argument state_code contains invalid values for state codes. Please refer to https://www.bls.gov/respondents/mwr/electronic-data-interchange/appendix-d-usps-state-abbreviations-and-fips-codes.htm for valid codes. Use utilities.State_To_FIPS_series() to convert postal codes to FIPS.\n",
      "  warnings.warn(\"Series passed as argument state_code contains invalid values for state codes. Please refer to https://www.bls.gov/respondents/mwr/electronic-data-interchange/appendix-d-usps-state-abbreviations-and-fips-codes.htm for valid codes. Use utilities.State_To_FIPS_series() to convert postal codes to FIPS.\")\n"
     ]
    }
   ],
   "source": [
    "data['macro_wage_change'] = bls.wage_change(data, 'state_fips','year_month_entry','year_month_exit')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "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>Unnamed: 0</th>\n",
       "      <th>race</th>\n",
       "      <th>gender</th>\n",
       "      <th>earnings_start</th>\n",
       "      <th>earnings_end</th>\n",
       "      <th>age</th>\n",
       "      <th>program</th>\n",
       "      <th>program_start</th>\n",
       "      <th>program_end</th>\n",
       "      <th>education_level</th>\n",
       "      <th>...</th>\n",
       "      <th>Zip</th>\n",
       "      <th>id</th>\n",
       "      <th>start_month</th>\n",
       "      <th>end_month</th>\n",
       "      <th>earnings_change</th>\n",
       "      <th>year_month_entry</th>\n",
       "      <th>year_month_exit</th>\n",
       "      <th>state_fips</th>\n",
       "      <th>earnings_end_adjusted</th>\n",
       "      <th>macro_wage_change</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11452</th>\n",
       "      <td>11453</td>\n",
       "      <td>White</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>36435.000</td>\n",
       "      <td>46395.635</td>\n",
       "      <td>18</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2010</td>\n",
       "      <td>2014</td>\n",
       "      <td>123</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>3096</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>9960.635</td>\n",
       "      <td>2010-10</td>\n",
       "      <td>2014-07</td>\n",
       "      <td>nan</td>\n",
       "      <td>49215.342</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8655</th>\n",
       "      <td>8656</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>26094.000</td>\n",
       "      <td>20468.000</td>\n",
       "      <td>47</td>\n",
       "      <td>French Literature</td>\n",
       "      <td>2018</td>\n",
       "      <td>2023</td>\n",
       "      <td>92</td>\n",
       "      <td>...</td>\n",
       "      <td>92373.000</td>\n",
       "      <td>5731</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>-5626.000</td>\n",
       "      <td>2018-10</td>\n",
       "      <td>2023-06</td>\n",
       "      <td>06</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3283</th>\n",
       "      <td>3284</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>29889.000</td>\n",
       "      <td>30227.890</td>\n",
       "      <td>24</td>\n",
       "      <td>French Literature</td>\n",
       "      <td>2015</td>\n",
       "      <td>2019</td>\n",
       "      <td>60</td>\n",
       "      <td>...</td>\n",
       "      <td>56359.000</td>\n",
       "      <td>11602</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>338.890</td>\n",
       "      <td>2015-08</td>\n",
       "      <td>2019-07</td>\n",
       "      <td>27</td>\n",
       "      <td>nan</td>\n",
       "      <td>6346.600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8406</th>\n",
       "      <td>8407</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>42509.000</td>\n",
       "      <td>37719.000</td>\n",
       "      <td>29</td>\n",
       "      <td>Biology</td>\n",
       "      <td>2013</td>\n",
       "      <td>2016</td>\n",
       "      <td>92</td>\n",
       "      <td>...</td>\n",
       "      <td>34471.000</td>\n",
       "      <td>13898</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>-4790.000</td>\n",
       "      <td>2013-10</td>\n",
       "      <td>2016-05</td>\n",
       "      <td>12</td>\n",
       "      <td>39465.508</td>\n",
       "      <td>2468.960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13165</th>\n",
       "      <td>13166</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>21013.000</td>\n",
       "      <td>35120.250</td>\n",
       "      <td>18</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2017</td>\n",
       "      <td>2020</td>\n",
       "      <td>124</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>12320</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>14107.250</td>\n",
       "      <td>2017-09</td>\n",
       "      <td>2020-07</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14384</th>\n",
       "      <td>14385</td>\n",
       "      <td>Black</td>\n",
       "      <td>Male</td>\n",
       "      <td>39676.000</td>\n",
       "      <td>66086.400</td>\n",
       "      <td>45</td>\n",
       "      <td>Philosophy</td>\n",
       "      <td>2014</td>\n",
       "      <td>2018</td>\n",
       "      <td>124</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>15613</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>26410.400</td>\n",
       "      <td>2014-09</td>\n",
       "      <td>2018-07</td>\n",
       "      <td>nan</td>\n",
       "      <td>66086.400</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1086</th>\n",
       "      <td>1087</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>24942.000</td>\n",
       "      <td>31663.800</td>\n",
       "      <td>32</td>\n",
       "      <td>Biology</td>\n",
       "      <td>2014</td>\n",
       "      <td>2015</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>4888</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>6721.800</td>\n",
       "      <td>2014-08</td>\n",
       "      <td>2015-07</td>\n",
       "      <td>nan</td>\n",
       "      <td>33547.537</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7925</th>\n",
       "      <td>7926</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>41731.000</td>\n",
       "      <td>53892.300</td>\n",
       "      <td>39</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2011</td>\n",
       "      <td>2015</td>\n",
       "      <td>92</td>\n",
       "      <td>...</td>\n",
       "      <td>76017.000</td>\n",
       "      <td>8879</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>12161.300</td>\n",
       "      <td>2011-08</td>\n",
       "      <td>2015-06</td>\n",
       "      <td>48</td>\n",
       "      <td>57098.451</td>\n",
       "      <td>3635.840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4243</th>\n",
       "      <td>4244</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>41654.000</td>\n",
       "      <td>49950.505</td>\n",
       "      <td>18</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2010</td>\n",
       "      <td>2017</td>\n",
       "      <td>73</td>\n",
       "      <td>...</td>\n",
       "      <td>48075.000</td>\n",
       "      <td>2485</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>8296.505</td>\n",
       "      <td>2010-08</td>\n",
       "      <td>2017-05</td>\n",
       "      <td>26</td>\n",
       "      <td>51169.119</td>\n",
       "      <td>4125.680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15520</th>\n",
       "      <td>15521</td>\n",
       "      <td>Black</td>\n",
       "      <td>Female</td>\n",
       "      <td>23696.000</td>\n",
       "      <td>33102.080</td>\n",
       "      <td>61</td>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2017</td>\n",
       "      <td>2022</td>\n",
       "      <td>125</td>\n",
       "      <td>...</td>\n",
       "      <td>78595.000</td>\n",
       "      <td>14317</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "      <td>9406.080</td>\n",
       "      <td>2017-09</td>\n",
       "      <td>2022-05</td>\n",
       "      <td>48</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0                       race             gender  \\\n",
       "11452       11453                      White  Other/unspecified   \n",
       "8655         8656                      White               Male   \n",
       "3283         3284                      White             Female   \n",
       "8406         8407                      White               Male   \n",
       "13165       13166                      White             Female   \n",
       "14384       14385                      Black               Male   \n",
       "1086         1087                      White               Male   \n",
       "7925         7926                      White             Female   \n",
       "4243         4244  Asian or Pacific Islander  Other/unspecified   \n",
       "15520       15521                      Black             Female   \n",
       "\n",
       "       earnings_start  earnings_end  age            program  program_start  \\\n",
       "11452       36435.000     46395.635   18        Art History           2010   \n",
       "8655        26094.000     20468.000   47  French Literature           2018   \n",
       "3283        29889.000     30227.890   24  French Literature           2015   \n",
       "8406        42509.000     37719.000   29            Biology           2013   \n",
       "13165       21013.000     35120.250   18        Art History           2017   \n",
       "14384       39676.000     66086.400   45         Philosophy           2014   \n",
       "1086        24942.000     31663.800   32            Biology           2014   \n",
       "7925        41731.000     53892.300   39          Economics           2011   \n",
       "4243        41654.000     49950.505   18        Art History           2010   \n",
       "15520       23696.000     33102.080   61       Anthropology           2017   \n",
       "\n",
       "       program_end  education_level  ...       Zip     id  start_month  \\\n",
       "11452         2014              123  ...       nan   3096           10   \n",
       "8655          2023               92  ... 92373.000   5731           10   \n",
       "3283          2019               60  ... 56359.000  11602            8   \n",
       "8406          2016               92  ... 34471.000  13898           10   \n",
       "13165         2020              124  ...       nan  12320            9   \n",
       "14384         2018              124  ...       nan  15613            9   \n",
       "1086          2015                0  ...       nan   4888            8   \n",
       "7925          2015               92  ... 76017.000   8879            8   \n",
       "4243          2017               73  ... 48075.000   2485            8   \n",
       "15520         2022              125  ... 78595.000  14317            9   \n",
       "\n",
       "       end_month earnings_change year_month_entry year_month_exit state_fips  \\\n",
       "11452          7        9960.635          2010-10         2014-07        nan   \n",
       "8655           6       -5626.000          2018-10         2023-06         06   \n",
       "3283           7         338.890          2015-08         2019-07         27   \n",
       "8406           5       -4790.000          2013-10         2016-05         12   \n",
       "13165          7       14107.250          2017-09         2020-07        nan   \n",
       "14384          7       26410.400          2014-09         2018-07        nan   \n",
       "1086           7        6721.800          2014-08         2015-07        nan   \n",
       "7925           6       12161.300          2011-08         2015-06         48   \n",
       "4243           5        8296.505          2010-08         2017-05         26   \n",
       "15520          5        9406.080          2017-09         2022-05         48   \n",
       "\n",
       "       earnings_end_adjusted  macro_wage_change  \n",
       "11452              49215.342                nan  \n",
       "8655                     nan                nan  \n",
       "3283                     nan           6346.600  \n",
       "8406               39465.508           2468.960  \n",
       "13165                    nan                nan  \n",
       "14384              66086.400                nan  \n",
       "1086               33547.537                nan  \n",
       "7925               57098.451           3635.840  \n",
       "4243               51169.119           4125.680  \n",
       "15520                    nan                nan  \n",
       "\n",
       "[10 rows x 28 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once you've gotten the macro-level wage change, you can simply subtract it from the nominal earnings change in order to get a trend-adjusted wage change."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['trend_adjusted_change'] = data['earnings_change'] - data['macro_wage_change']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fc335d2ee90>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "original_summary = utilities.multiple_describe(data, 'program', 'earnings_change')\n",
    "adjusted_summary = utilities.multiple_describe(data, 'program', 'trend_adjusted_change')\n",
    "original_summary['calculation'] = \"original\"\n",
    "adjusted_summary['calculation'] = \"adjusted\"\n",
    "\n",
    "sns.barplot(x=\"mean\", y=\"program\", hue=\"calculation\", data=pd.concat([original_summary, adjusted_summary]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Calculate earnings premium - VERY fancy way__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perhaps the most novel part of the `roi-toolkit` package is the `metrics.Earnings_Premium()` class. This class requires a dataframe and several columns. Most of these columns are self-explanatory (please see the [docstrings](https://github.com/brighthive/roi-toolkit/blob/master/roi/metrics.py)), but the user also needs to supply an `education_level` column containining individuals' education levels __prior__ to program entry. This column needs to match up with [CPS EDUC codes](https://cps.ipums.org/cps-action/variables/EDUC#codes_section).\n",
    "\n",
    "The backbone of this method is a modified [Mincer earnings function](https://en.wikipedia.org/wiki/Mincer_earnings_function) that is fit on ten years of national-level Current Population Survey data. The complete model is stored in the `roi-toolkit` repo, while its coefficients alone are packaged with the module itself.\n",
    "\n",
    "For comprehensive background on the Mincer model, please read James Heckman's magisterial [NBER paper](https://www.nber.org/system/files/working_papers/w9732/w9732.pdf) on the topic.\n",
    "\n",
    "The Mincer model fits a basic log-linear model to estimate the approximate impacts of education and work experience on wages. It provides broad numerical answers to questions such as \"For an additional X years of work experience, what percentage raise can a person expect who has Y education and Z previous years in the labor force?\"\n",
    "\n",
    "The Mincer model is not particularly effective at forecasting overall earnings using this data alone: individuals' earnings vary dramatically based on numerous other factors. However, the values of the coefficients on __additional__ education are fairly precisely estimated. In other words, though the Mincer model can't very effectively predict someone's wages if all you know about them is their education and work experience, it can be very useful in helping forecast their wages if you __already know their historical earnings__.\n",
    "\n",
    "That's how it's used here: for every individual in a dataset, the `metrics.Earnings_Premium()` method uses the Mincer model to predict a counterfactual wage: if, instead of participating in their educational or training program, they'd stayed in the labor force, what would their wage be? The earnings premium is calculated as the difference between their observed post-program wage and their predicted counterfactual wage."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mattlerner/miniconda3/lib/python3.7/site-packages/pandas/core/indexing.py:670: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  iloc._setitem_with_indexer(indexer, value)\n"
     ]
    }
   ],
   "source": [
    "# library method\n",
    "premium = metrics.Earnings_Premium(data, 'state_fips', 'education_level', 'earnings_start', 'earnings_end', 'program_start', 'program_end', 'age')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['predicted_wage'] = premium.predicted_wage\n",
    "data['earnings_premium'] = premium.full_premium"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>race</th>\n",
       "      <th>gender</th>\n",
       "      <th>earnings_start</th>\n",
       "      <th>earnings_end</th>\n",
       "      <th>age</th>\n",
       "      <th>program</th>\n",
       "      <th>program_start</th>\n",
       "      <th>program_end</th>\n",
       "      <th>education_level</th>\n",
       "      <th>...</th>\n",
       "      <th>end_month</th>\n",
       "      <th>earnings_change</th>\n",
       "      <th>year_month_entry</th>\n",
       "      <th>year_month_exit</th>\n",
       "      <th>state_fips</th>\n",
       "      <th>earnings_end_adjusted</th>\n",
       "      <th>macro_wage_change</th>\n",
       "      <th>trend_adjusted_change</th>\n",
       "      <th>predicted_wage</th>\n",
       "      <th>earnings_premium</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>13134</th>\n",
       "      <td>13135</td>\n",
       "      <td>Other</td>\n",
       "      <td>Male</td>\n",
       "      <td>34179.000</td>\n",
       "      <td>42249.900</td>\n",
       "      <td>33</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2012</td>\n",
       "      <td>2015</td>\n",
       "      <td>124</td>\n",
       "      <td>...</td>\n",
       "      <td>6</td>\n",
       "      <td>8070.900</td>\n",
       "      <td>2012-08</td>\n",
       "      <td>2015-06</td>\n",
       "      <td>26</td>\n",
       "      <td>44763.424</td>\n",
       "      <td>2792.920</td>\n",
       "      <td>5277.980</td>\n",
       "      <td>39463.487</td>\n",
       "      <td>2786.413</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10111</th>\n",
       "      <td>10112</td>\n",
       "      <td>Black</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>23225.000</td>\n",
       "      <td>22868.400</td>\n",
       "      <td>61</td>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2013</td>\n",
       "      <td>2016</td>\n",
       "      <td>111</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>-356.600</td>\n",
       "      <td>2013-09</td>\n",
       "      <td>2016-07</td>\n",
       "      <td>nan</td>\n",
       "      <td>23927.279</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>19485.773</td>\n",
       "      <td>3382.627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15202</th>\n",
       "      <td>15203</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>22376.000</td>\n",
       "      <td>46803.680</td>\n",
       "      <td>54</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2016</td>\n",
       "      <td>2018</td>\n",
       "      <td>125</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>24427.680</td>\n",
       "      <td>2016-10</td>\n",
       "      <td>2018-07</td>\n",
       "      <td>48</td>\n",
       "      <td>46803.680</td>\n",
       "      <td>2241.720</td>\n",
       "      <td>22185.960</td>\n",
       "      <td>20746.926</td>\n",
       "      <td>26056.754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10827</th>\n",
       "      <td>10828</td>\n",
       "      <td>Black</td>\n",
       "      <td>Female</td>\n",
       "      <td>34726.000</td>\n",
       "      <td>37235.880</td>\n",
       "      <td>40</td>\n",
       "      <td>Philosophy</td>\n",
       "      <td>2015</td>\n",
       "      <td>2021</td>\n",
       "      <td>111</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>2509.880</td>\n",
       "      <td>2015-08</td>\n",
       "      <td>2021-05</td>\n",
       "      <td>06</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>42246.742</td>\n",
       "      <td>-5010.862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8320</th>\n",
       "      <td>8321</td>\n",
       "      <td>White</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>17815.000</td>\n",
       "      <td>19084.200</td>\n",
       "      <td>18</td>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2014</td>\n",
       "      <td>2018</td>\n",
       "      <td>92</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>1269.200</td>\n",
       "      <td>2014-08</td>\n",
       "      <td>2018-07</td>\n",
       "      <td>42</td>\n",
       "      <td>19084.200</td>\n",
       "      <td>4732.520</td>\n",
       "      <td>-3463.320</td>\n",
       "      <td>26820.410</td>\n",
       "      <td>-7736.210</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0   race             gender  earnings_start  earnings_end  \\\n",
       "13134       13135  Other               Male       34179.000     42249.900   \n",
       "10111       10112  Black  Other/unspecified       23225.000     22868.400   \n",
       "15202       15203  White             Female       22376.000     46803.680   \n",
       "10827       10828  Black             Female       34726.000     37235.880   \n",
       "8320         8321  White  Other/unspecified       17815.000     19084.200   \n",
       "\n",
       "       age       program  program_start  program_end  education_level  ...  \\\n",
       "13134   33   Art History           2012         2015              124  ...   \n",
       "10111   61  Anthropology           2013         2016              111  ...   \n",
       "15202   54     Economics           2016         2018              125  ...   \n",
       "10827   40    Philosophy           2015         2021              111  ...   \n",
       "8320    18  Anthropology           2014         2018               92  ...   \n",
       "\n",
       "       end_month  earnings_change  year_month_entry  year_month_exit  \\\n",
       "13134          6         8070.900           2012-08          2015-06   \n",
       "10111          7         -356.600           2013-09          2016-07   \n",
       "15202          7        24427.680           2016-10          2018-07   \n",
       "10827          5         2509.880           2015-08          2021-05   \n",
       "8320           7         1269.200           2014-08          2018-07   \n",
       "\n",
       "      state_fips earnings_end_adjusted macro_wage_change  \\\n",
       "13134         26             44763.424          2792.920   \n",
       "10111        nan             23927.279               nan   \n",
       "15202         48             46803.680          2241.720   \n",
       "10827         06                   nan               nan   \n",
       "8320          42             19084.200          4732.520   \n",
       "\n",
       "      trend_adjusted_change  predicted_wage  earnings_premium  \n",
       "13134              5277.980       39463.487          2786.413  \n",
       "10111                   nan       19485.773          3382.627  \n",
       "15202             22185.960       20746.926         26056.754  \n",
       "10827                   nan       42246.742         -5010.862  \n",
       "8320              -3463.320       26820.410         -7736.210  \n",
       "\n",
       "[5 rows x 31 columns]"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fc333dd3610>"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzQAAAFtCAYAAADRbC9hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzde7xVVbnw8d8jN1Fk4x1MPVASUnm/Z3nNNF9NyzdNLYMwO5qpqcfyqIiX1yytvHdRkeqY5glvHRPN46UEQ0E33ggVQbBQvAGCAgLP+8ecmxbLvblsNnvvBb/v5zM/rDXGmGM8cy77tJ49xhwrMhNJkiRJqkVrtXUAkiRJktRcJjSSJEmSapYJjSRJkqSaZUIjSZIkqWaZ0EiSJEmqWSY0kiRJkmqWCY0kSZKkmmVCI0mSJKlmdWzrAKS2EBEBbAa829axSJIkqUnrAf/MzGyqgQmN1lSbAa+2dRCSJElaps2BfzRVaUKjNdW7AFOnTqV79+5tHYskSZKqzJo1iy222AKWsaLGhEZrtO7du5vQSJIk1TA3BZAkSZJUs5yh0Rptr3NvoUOXrm0dhqTV2NjLjmvrECRpteYMjSRJkqSa5QyNJEmSak5msmDBAhYuXNjWoaiZOnToQMeOHSl+TaP5TGgkSZJUU+bPn8+0adN477332joUraR11lmHXr160blz52b3YUIjSZKkmrFo0SImTZpEhw4d2GyzzejcufNK/4VfrS8zmT9/Pm+88QaTJk2ib9++rLVW856GMaGRJElSzZg/fz6LFi1iiy22YJ111mnrcLQSunbtSqdOnXjllVeYP38+a6+9drP6cVMASZIk1Zzm/jVf7UtLfI7+lyBJkiSpZpnQSJIkSapZJjSSJElSO9K7d2+uuOKKtg6jZpjQSJIkSapZJjSSJEmSapYJTY2LiIyIw9s6jqZExOSIOK2t45AkSWoNv/rVr9hss81YtGjREuWHHXYY3/zmN5k4cSKHHXYYm266Kd26dWOXXXbhgQceaLK/yZMnExHU19cvLpsxYwYRwcMPP7y47Nlnn+ULX/gC3bp1Y9NNN+XrX/86b775ZotfX3tkQtOKImKPiFgYEfc049whEVG/7JaSJElqK1/5yld46623eOihhxaXvf3224wYMYJjjz2W2bNnc/DBB/O///u/PPXUUxx00EEceuihTJkypdljzpgxg/32248ddtiBMWPGMGLECF5//XWOPPLIlrikds8f1mxdg4CrgUERsVlm/rO1A4jip3Q7ZOaC1h5bkiRpdbf++uvzhS98gd/97nfsv//+APzhD39go402Yt9992WttdZiu+22W9z+oosu4o477uDuu+/m5JNPbtaY11xzDTvssAOXXHLJ4rKhQ4eyxRZb8MILL/Dxj3985S6qnXOGppVERDfgKODnwD3AgIq6fcqlY/tHxJiIeC8iRkVEv7J+AHA+sF3ZLsuyBhtFxB3leS9GxBcb6fsLETEWmAd8JiK6RMRVETE9IuZGxKMRsUsj5/2fiHi6bPO3iPhU1XUdERHPRcS8cnnZGcu4D1tGxF0RMTsiZkXEbRGxaVWbc8u43o2IGyLi0obZqYjYKyI+iIieVedcERF/XdbnIEmStKode+yxDB8+nHnz5gFw880389WvfpW11lqL2bNnc+aZZ9K/f3969OhBt27dGD9+/ErN0IwbN46HHnqIbt26LT623nprACZOnNgi19SemdC0niOBv2fmBOC/gG+WsyWV/h9wBrAzsAAYWpb/HvgJ8BzQqzx+X3He+cBtwLbAn4CbI2KDqr4vBX4A9AeeBn4MHAF8A9gReAm4r5HzLitj2gV4A/hjRHQCiIidynFvBbYBhgAXVSVbi0XEWsBdwAbA3sABwEcrryUijgXOAb4P7ARMAU5sqM/MvwAvA1+vOKcTcGzF/Wps7C4R0b3hANZrqq0kSdLKOPTQQ8lM7rnnHqZOncpf//pXjj32WADOPPNM7rjjDi655BL++te/Ul9fzzbbbMP8+fMb7WuttYqv65m5uOyDDz5Yos3s2bM59NBDqa+vX+J48cUX2WuvvVbRVbYfLjlrPYMoEhmAEUAdxZf6hyvanJOZjwBExKXAPRGxdma+HxGzgQWZ+VojfQ/LzFvK8/4TOAXYtRynweDM/HPZZl2KJGFAZt5bln2LIsEYRJHENLig4rxvAK8CX6JIZE4H/jczLyrbvhARnwD+AxjWSJz7UyQ+fTJzatnnccBzEbFLZj4BfBe4MTNvKs+5MCI+D3Sr6OdGYGBFnIcCa5cxNeVsisRvCTd3u5L11u6wlNMkaeVMufCyZTdahbYc/Eybji+tidZee22+/OUvc/PNN/PSSy/Rr18/dtxxRwBGjhzJgAED+NKXvgQUycjkyZOb7GvjjTcGYNq0aeywww4AS2wQALDjjjsyfPhwevfuTceOa97Xe2doWkG5dGxX4BaA8vmV31MkD5Werng9rfx3k+UYYvF5mTkHmNXIeWMqXn8M6ASMrDjvA+BxihmcSo9VtHkbmFDRpn9lH6WRQN+IaCxL6A9MbUhmyj6fB2ZU9NmvjKNS9fthwFYRsXv5fgBwW3ntTfkhRRLZcGy+lLaSJEkr5dhjj+Wee+5h6NChi2dnAPr27cvtt99OfX0948aN45hjjvnQjmiVunbtyu67786ll17K+PHjeeSRRzj33HOXaPOd73yHt99+m6OPPponnniCiRMnct999zFw4EAWLly4yq6xvTChaR2DKGbD/hkRCyJiAcUMyRERUVfRrnL+sGFecXk+ow+q3mcj5y3ty35NyczpwB+BgeXzN19gKcvNynPmZeashgN4txVClSRJa6j99tuPDTbYgAkTJnDMMccsLv/pT3/K+uuvz6c//WkOPfRQDjzwwMWzN00ZOnQoCxYsYKedduK0007j4osvXqJ+s802Y+TIkSxcuJDPf/7zbLPNNpx22mn06NFj8ZK11dmaNyfVyiKiI3AcxXMo91dV3wkcDfx9ObqaD7TU2qiJZX97Aq+UcXaieE7miqq2u1M8x0JErA98HBhf1o0v+6i0J/BCZjb254DxwBYRsUXFkrNPAD2A58s2E8o4flNx3i582A0UM16vAhMzs3qmSJIkqc2stdZa/POfH97Qtnfv3jz44INLlH3nO99Z4n31ErT+/fszatSoJcoqn6mBf838rIlMaFa9Q4D1KZ4LmVlZERHDKWZv/mM5+pkM9ImI7Sm+xL+bmfOaE1BmzomInwOXRcTbFAnLWcA6FM+nVBocEW8Br1NsWvAmRSIGxUYFT0TEeRRL6PYATgZOamLoB4BnKDYtOI3iv7/rgEcys2FJ3NXA9RExBhhFsTPcthQbAVS6j2Jp3bnA4BW7A5IkSVpdrP5zUG1vEPBAdTJTGk6xo9m2y9HPcIqH/B+i2G3s6JWM6wdln78FngS2Ag7MzHcaaXclMBboCRyamfMBMvNJit3bvgo8C1xIsfnAsMYGzOJPCYcB7wB/oUhwXqZIWhra3EzxvMvlZVx9KJ6ZmVvV16KyvANLzuZIkiRpDeIMzSqWmYcupe5xoGHr5quq6uor6ihnY/5vI31Ub/1MZvaoeP1wZT8V5XMpdkM7ZRmX8GhmfqqpyswcTpEYNVXfu+r9FIqkpknlrmkNO6cREX+m2Fa62keAP2XmtEbqJEmStAYwoVG7EhHrAP9OsaRsIcVM1OcotpRuaFNHsf3zMcAXG+lGkiRJawgTGrU3CRxM8eOaa1NsEnBEZj5Q0eYuim2wf9HwGzmSJElaM5nQqFFNLVVrhXHfp5iRWVqbfVonGkmSJLV3bgogSZIkqWaZ0EiSJEmqWSY0kiRJ0hpswIABHH744W0dRrP5DI0kSZJq3k7/0bo/Szf2suNadbwhQ4Zw5513Ul9f36rj1gJnaCRJkiTVLBMaSZIkqRWMGDGCz3zmM/To0YMNN9yQQw45hIkTJy6uf/XVVzn66KPZYIMNWHfdddl5550ZPXo0w4YN44ILLmDcuHFEBBHBsGHDmDx5MhGxxKzNjBkziAgefvhhABYuXMigQYPo06cPXbt2pV+/flx55ZWtfemrlEvOJEmSpFYwZ84cTj/9dLbddltmz57N4MGD+dKXvkR9fT3vvfcee++9Nx/5yEe4++676dmzJ08++SSLFi3iqKOO4tlnn2XEiBE88EDx03x1dXW8/vrryxxz0aJFbL755vz3f/83G264IaNGjeKEE06gV69eHHnkkav6kluFCY0kSZLUCo444ogl3g8dOpSNN96Y559/nlGjRvHGG2/wxBNPsMEGGwCw1VZbLW7brVs3OnbsSM+ePVdozE6dOnHBBRcsft+nTx8ee+wxbrvtNhMaaXWwxQ/+Rvfu3ds6DEmStAZ48cUXGTx4MKNHj+bNN99k0aJFAEyZMoX6+np22GGHxclMS7r22msZOnQoU6ZM4f3332f+/Plsv/32LT5OW/EZGkmSJKkVHHroobz99ttcf/31jB49mtGjRwMwf/58unbtusL9rbVW8VU+MxeXffDBB0u0ufXWWznzzDMZNGgQ999/P/X19QwcOJD58+evxJW0LyY0kiRJ0ir21ltvMWHCBM4991z2339/+vfvzzvvvLO4ftttt6W+vp6333670fM7d+7MwoULlyjbeOONAZg2bdrisuptnUeOHMmnP/1pTjrpJHbYYQe22mqrJTYiWB2Y0EiSJEmr2Prrr8+GG27Ir371K1566SUefPBBTj/99MX1Rx99ND179uTwww9n5MiRvPzyywwfPpzHHnsMgN69ezNp0iTq6+t58803mTdvHl27dmX33Xfn0ksvZfz48TzyyCOce+65S4zbt29fxowZw3333ccLL7zAeeedxxNPPNGq176qmdBIkiRJq9haa63FrbfeytixY/nUpz7F9773PS677LLF9Z07d+b+++9nk0024eCDD2abbbbh0ksvpUOHDkCxocBBBx3Evvvuy8Ybb8wtt9wCFBsLLFiwgJ122onTTjuNiy++eIlxv/3tb/PlL3+Zo446it1224233nqLk046qfUuvBVE5Zo7aU0REd2BmTNnznRTAEmSasjcuXOZNGkSffr0Ye21127rcLSSlvZ5zpo1i7q6OoC6zJzVVB/O0EiSJEmqWSY0kiRJkmqWv0OjNdpe595Chy4rvk2iJKl2jL3suLYOQdIq5AyNJEmSpJplQiNJkiSpZpnQSJIkSapZJjSSJEmSapYJjSRJkqSaZUIjSZIkqWaZ0EiSJEmrmd69e3PFFVcsfh8R3Hnnna0ex5AhQ9h+++1X6Rj+Do0kSZJq3pQLt2nV8bYc/Eyrjreypk2bxvrrr79cbYcMGcKdd95JfX39Ko6qZZjQSJIkSe3Q/Pnz6dy5c4v01bNnzxbppz1yyZkkSZLUCvbZZx9OPvlkTj75ZOrq6thoo40477zzyEygWCZ20UUXcdxxx9G9e3dOOOEEAB599FE++9nP0rVrV7bYYgtOOeUU5syZs7jf6dOnc+ihh9K1a1f69OnDzTff/KGxq5ecvfrqqxx99NFssMEGrLvuuuy8886MHj2aYcOGccEFFzBu3Dgigohg2LBhAMyYMYPjjz+ejTfemO7du7Pffvsxbty4Jca59NJL2XTTTVlvvfUYNGgQc+fObenb+CEmNJIkSVIr+fWvf03Hjh15/PHHufLKK/npT3/KDTfcsLj+8ssvZ7vttuOpp57ivPPOY+LEiRx00EEcccQRPP300/z+97/n0Ucf5eSTT158zoABA5g6dSoPPfQQf/jDH7juuuuYPn16kzHMnj2bvffem3/84x/cfffdjBs3jrPOOotFixZx1FFHccYZZ/DJT36SadOmMW3aNI466igAvvKVrzB9+nTuvfdexo4dy4477sj+++/P22+/DcBtt93GkCFDuOSSSxgzZgy9evXiuuuuW0V38l9ccqZmi4h9gIeA9TNzRhuHI0mS1O5tscUW/OxnPyMi6NevH8888ww/+9nP+Na3vgXAfvvtxxlnnLG4/fHHH8+xxx7LaaedBkDfvn256qqr2Hvvvfn5z3/OlClTuPfee3n88cfZZZddALjxxhvp379/kzH87ne/44033uCJJ55ggw02AGCrrbZaXN+tWzc6duy4xDK1Rx99lMcff5zp06fTpUsXoEi+7rzzTv7whz9wwgkncMUVVzBo0CAGDRoEwMUXX8wDDzywymdpnKGpURGxR0QsjIh7VuCcIRGxzKe7mmoXEb0jIiOiYauKUUAvYOZy9LlPeW6P5Y1XkiRpdbP77rsTEYvf77HHHrz44ossXLgQgJ133nmJ9uPGjWPYsGF069Zt8XHggQeyaNEiJk2axPjx4+nYsSM77bTT4nO23nprevRo+itXfX09O+yww+JkZnmMGzeO2bNns+GGGy4Ry6RJk5g4cSIA48ePZ7fddlvivD322GO5x2guZ2hq1yDgamBQRGyWmf9sqmEU/6vp0NIBZOZ84LWW7ndpGq4lMxe05riSJEmtYd11113i/ezZs/n2t7/NKaec8qG2W265JS+88MIKj9G1a9cVPmf27Nn06tWLhx9++EN1S0ueWoMzNDUoIroBRwE/B+4BBlTVN8yGfCEixgLzgK8B5wPblXUZEQNYCdWzLhHxbxHxx4h4JyLmRMRzEXFwRPSmWJoG8E55zrDynC4RcVVETI+IuRHxaETssqxriYhFEbFzVTynRcQrEeF/15IkqV0aPXr0Eu//9re/0bdvXzp0aPxvzzvuuCPPP/88W2211YeOzp07s/XWW7NgwQLGjh27+JwJEyYwY0bTTwNsu+221NfXL372pVrnzp0XzxhVxvHaa6/RsWPHD8Wx0UYbAdC/f/9Gr29V84tfbToS+HtmTgD+C/hmVM5d/sulwA+A/sCfgZ8Az1EsE+sF/L6F47oW6ALsBWwDfB+YDUwFjijb9CvHPrV8/+Oy7hvAjsBLwH0RUT0HWnktdwMPAAOr2gwEhmXmourAysSpe8MBrLcS1ylJktQsU6ZM4fTTT2fChAnccsstXH311Zx66qlNtv/+97/PqFGjOPnkk6mvr+fFF1/krrvuWrwpQL9+/TjooIP49re/zejRoxk7dizHH3/8Umdhjj76aHr27Mnhhx/OyJEjefnllxk+fDiPPfYYUOy2NmnSJOrr63nzzTeZN28en/vc59hjjz04/PDDuf/++5k8eTKjRo3inHPOYcyYMQCceuqpDB06lJtuuokXXniB888/n+eee64F717jXHJWmwZRJDIAI4A6YG/g4ap2gzPzzw1vImI2sCAzl2eZ2DZl+0qNJU2VtgSGZ2bDL029XDF2w58ApjdsIBAR6wInAgMy896y7FvAARTXeNlSruUG4BcRcXpmzouIHSmSqMOaiO1sihmqJdzc7UrWW7vFV+NJUquqtR/4k9Zkxx13HO+//z677rorHTp04NRTT128PXNjtt12Wx555BHOOeccPvvZz5KZfOxjH1u88xjATTfdxPHHH8/ee+/NpptuysUXX8x5553XZJ+dO3fm/vvv54wzzuDggw9mwYIFfOITn+Daa68F4IgjjuD2229n3333ZcaMGdx0000MGDCAP/3pT5xzzjkMHDiQN954g549e7LXXnux6aabAnDUUUcxceJEzjrrLObOncsRRxzBiSeeyH333ddCd69x0bDvtWpDRPQDngU+kpnTy7JrgLrM/Hr5fh+KJV6bZ+Y/Ks4dAhyemdtX91s1xhCKWaAvVlV9hCJp2iEz66t3OYuI4ymWwT1OMYMyPDOfropp/YqEZltgHNA7M1+pGP8O4J3M/OZSrqUz8A/gu5l5a0RcBXwyM/dv4pq6UMweNVgPePXZs/ub0EiqeSY0WpPMnTuXSZMm0adPH9Zee+22DmeF7LPPPmy//fZcccUVbR1Ku7G0z3PWrFnU1dVB8T13VlN9uOSs9gyimFn7Z0QsiIgFFLMcR0REXVXbOR86e/nNz8yXKg/glaWdkJk3AB8FfksxWzImIr67EjFUWuJayg0JfgMMLJObY4ChS4ltXmbOajiAd1soLkmSJLUhE5oaEhEdgeOAM4DtK47tgH8CRy+ji/msgt3OKmXm1Mz8RWZ+meKZnW9VjE3V+BPL8j0bCiKiE7AL8PxyDHcD8DngJIok7/aVi16SJEm1xmdoasshwPrAjZm5xG+/RMRwitmbXyzl/MlAn/J3ZF4F3s3MeS0VXERcAdwLvFDGuS8wvqx+BUjgkIj4E/B+Zs6OiJ8Dl5XP2EwBzgLWAW5c1niZOT4i/gb8CBiame+31LVIkiS1tMa2PNbKc4amtgwCHqhOZkrDgZ3L51KaMpxiE4GHgDdY9ozOiupAsdPZ+HKcFyhmTyiffzmfYrey14FrynN+UMb1W+BJYCvgwMx8ZznHvBHozFKWm0mSJGn15aYAqmkRcR7wlcxcWiLX2HndgZluCiBpdeCmAFqT1PKmAPowNwXQGisiukXEp4CTgavbOh5JktS6/KP86qElPkcTGtWqa4CxFNtIu9xMkqQ1RKdOnQB477332jgStYSGz7Hhc20ONwVQTcrMAcCANg5DkiS1sg4dOtCjRw+mT58OwDrrrEPEsn77W+1NZvLee+8xffp0evToQYcOzX8EwIRGkiRJNaVnz54Ai5Ma1a4ePXos/jyby4RGkiRJNSUi6NWrF5tssgkffPBBW4ejZurUqdNKzcw0MKGRJElSTerQoUOLfCFWbXNTAEmSJEk1y4RGkiRJUs0yoZEkSZJUs0xoJEmSJNUsExpJkiRJNSsys61jkFpdRHQHZs6cOZPu3bu3dTiSJEmqMmvWLOrq6gDqMnNWU+2coZEkSZJUs0xoJEmSJNUsExpJkiRJNcuERpIkSVLNMqGRJEmSVLNMaCRJkiTVLBMaSZIkSTWrY1sHILWlvc69hQ5durZ1GJLUJsZedlxbhyBJK80ZGkmSJEk1y4RGkiRJUs0yoZEkSZJUs0xoJEmSJNUsExpJkiRJNcuERpIkSVLNMqGRJEmSVLNMaCRJkiTVLBMaSZIkSTXLhEaSJElSzTKhkSRJklSzTGi03CKid0RkRGy/Auc8HBFXrMq4JEmStOYyodFiETGsTFgajrciYkREbFs2mQr0Ap5twzAlSZKkxUxoVG0ERdLSC9gfWAD8D0BmLszM1zJzQRvGJ0mSJC1mQqNq88qk5bXMrAcuBbaIiI0bW3IWEXtHxOMRMS8ipkXEpRHRsanOI2L9iPhNRLwTEe9FxL0R0beqzbciYmpZf0dEnB4RM8q63hGxKCJ2rjrntIh4JSL8b1qSJGkN4pc/NSkiugFfA14C3mqk/iPAn4AngO2AE4FBwLlL6XYYsDPwRWAPIIA/RUSnss89gV8AVwLbA38Gzmk4OTMnAw8AA6v6HQgMy8xFK3aVkiRJqmVN/iVda6xDImJ2+XpdYBpwSGYuiojqtidRPFdzcmYm8PeI2Az4UURcWJ1clDMxXwT2zMxRZdmxZR+HA/8NfBe4NzMvL097ISI+DRxS0dUNwC8i4vTMnBcROwLbAIc1dVER0QXoUlG03vLcDEmSJLVvJjSq9hDFTAvA+hRJy70RsWsjbfsDj5XJTIORQDdgc2BKI+0XAKMbCjLzrYiYUNYB9APuqDrvcZZMaO4ErgW+BNwKDAAeKmdvmnI2cH514c3drmS9tTss5TRJWr1sOfiZtg5BklqUS85UbU5mvlQeTwDHU8zUfKuN41osM+cDvwEGRkRn4Bhg6DJO+yFQV3FsvkqDlCRJUqswodGyJLAI6NpI3Xhgj1hyLdqewLvAq0207wjs1lAQERtSzMo8XxZNAHapOq/6PRTLzj5HMYPUEbh9qReROS8zZzUcZYySJEmqcSY0qtYlInqWR3/gaoolZH9spO11wBbA1RGxdUQcBlwA/LSxh/Mz80XgLuD6iPhMRGwH/Bfwj7KccryDy53N+kbEt4EvUCRWlX2NB/4G/Ai4JTPfX/lLlyRJUq0xoVG1gyg2AphG8azLLsBXMvPh6oaZ+Q/gYGBXYBzF7mQ3Ahcvpf+BwFiK37Z5jGKXs4Mz84Oyz5HAvwOnl30eBPwMmNtIXzcCnVn2cjNJkiStptwUQItl5gCKB+ybqp9MkYBUlj1CkdA0dc4+Ve/fAY5bRhzXA9c3vI+I6ym2jq72EeCZ8lkfSZIkrYFMaNTuRMSZFL8/M4diudk3KJ6VaajvBvQGTmbpv3kjSZKk1ZxLztQe7UqR0DxDsfzslMy8oaL+Goplaw/jcjNJkqQ1mjM0ancy88hl1A9gKUvjJEmStOZwhkaSJElSzTKhkSRJklSzTGgkSZIk1SwTGkmSJEk1y4RGkiRJUs0yoZEkSZJUs0xoJEmSJNUsExpJkiRJNSsys3knRmwGfAbYhKrEKDOvWvnQpFUnIroDM2fOnEn37t3bOhxJkiRVmTVrFnV1dQB1mTmrqXYdm9N5RAwAfgnMB94CKrOiBExoJEmSJK1yzUpogIuAC4EfZuaiFoxHkiRJkpZbc5+hWQe41WRGkiRJUltqbkJzI/CVlgxEkiRJklZUc5ecnQ38T0QcBDwDfFBZmZmnr2xgkiRJkrQsK5PQHAhMKN9XbwogSZIkSatccxOaM4BvZuawFoxFkiRJklZIcxOaecDIlgxEagt7nXsLHbp0beswJElSE8Zedlxbh6B2rrmbAlwJfLclA5EkSZKkFdXcGZpdgf0i4hDgOT68KcCXVzYwSZIkSVqW5iY0M4DbWzIQSZIkSVpRzUpoMnNgSwciSZIkSSuquc/QSJIkSVKba+6SMyLi/wJHAlsCnSvrMnPHlYxLkiRJkpapWTM0EXEKcBPwOrAD8DjwFvBR4N4Wi06SJEmSlqK5S85OAk7IzO8C84EfZ+YBwFVAXUsFJ0mSJElL09yEZktgVPn6fWC98vVvgaNXNihJkiRJWh7NTWheAzYoX08Bdi9f9wFiZYOSJEmSpOXR3ITmQeCL5eubgJ9FxJ+B3wN3tERgkiRJkrQszd3l7ATKZCgzr42It4BPA3cDv2yh2CRJkiRpqVZ4hiYiOgLnAj0byjLz1sw8JTOvzsz5LRmg1gwR0TsiMiK2b+tYJEmSVDtWOKHJzAXAWazEb9io+cqjtqEAACAASURBVCJiWPnFv/oY0daxraSpQC/g2bYORJIkSbWjuUnJ/wJ7A5NbLhStgBHAwKqyeW0RSEvJzIUUm01IkiRJy625mwLcC1waEZdHxNER8cXKoyUDVKPmZeZrVcc7ABHRIyJ+GRGvR8TciHg2Ig5pODEijoiI5yJiXkRMjogzKjsuy/4zIoZGxLsRMSUiTqhqs01EPBgR70fEWxHxq4joVlE/LCLuLPt5PSJmRMTgiOgYEZdFxNsR8WpEDKw450NLziLikxHxPxExq4zlrxHxsbJun4h4PCLmlP2PjIh/a/lbLUmSpPasuTM015X/nt5IXQIdmtmvVkJErEWRbK4HfA2YCHwCWFjW7wTcBgyh2JHu08B1EfFWZg6r6OoM4DzgEuD/Aj+PiEcyc0JErAvcBzwG7AJsAtwAXAMMqOhjP+BVYC9gT+DGcry/ALsBRwG/jIg/Z+arjVzLR8q2D5d9zSr76Vg+x3UncD3F7x51Bnal+G+vqXvTBehSUbReU20lSZJUOyKzye+AaociYhhFsjK3quoSYAxFQtM/M19o5NybgY0z8/MVZT8G/k9mfrJ8Pxn4a2Z+vXwfFEvBzs/MX0TEt4AfAVtk5pyyzcHAH4HNMvP1MsZ9gI9m5qKyzd+B6Zm5V/m+AzATOD4zb42I3sAkYIfMrI+IS4CvAv0y84Oq69gAeAvYJzMfWc77NgQ4v7r82bP7s97a5t+S1mxbDn6mrUOQpA+ZNWsWdXV1AHWZOaupds1dcqa29RCwfdXxi/LfVxtLZkr9gZFVZSOBvmWC0eDphhdZZLyvUczENPQxriGZqehjLaBfRdlzDclM6XVg8f9jls/MvFXRb7XtKRKrD6orMvNtYBhwX0T8MSJOjYheTfTT4IdAXcWx+TLaS5IkqQY0a8lZRJzSRFVSzBy8BPyl/NKqljcnM1+qLoyI91uo/+okIlnx5LexPlak36VeS2YOjIirgIMolq9dHBEHZObfmmg/j4qNE4qJJ0mSJNW65j5D8z1gY2Ad4J2ybH3gPWA2xV/dX46IfTNz6kpHqeX1NLB5RHy8iVma8RTPoVTaE3hhBZLP8cCAiFi3YpZmT2ARMKE5QTfhaeAbEdGpsVkagMx8CngK+GFEPAYcAzSa0EiSJGn11NwlZ/8JPAH0zcwNM3ND4OPAaOBUYEuKZUo/a5EoVa1LRPSsOjYqnyf5CzA8Ig6IiD4R8YWIOKg87yfA/hFxXkR8PCK+AZwMXL4CY99MMQv364j4VETsC1wN/DYzX2/Ba7wG6A7cGhE7R0TfiPh6RPQrr+uHEbFHRPxbRHwe6EuRbEmSJGkN0tyE5mLge5k5saGgXAJ1JvDDcteqs/jwbIBaxkHAtKrj0bLuCIpk8xbgeeDHlLvOZeaTwJEUD9s/C1wIDK7a4WypMvM94EBgg3KcP1D8LtHJK3lN1eO8RbG7WTfgEWAs8C2KZWvvAVsDw4EXgF8B1wK/bMkYJEmS1P41a5eziHgP2Cszx1SV7wI8kpnrlLtWPZuZ3RrpQmpTEdEdmOkuZ5LkLmeS2qdVvcvZQxS/IbJDQ0H5+ufAg2XRNhTb8EqSJEnSKtHchGYQ8DYwtvzF+XkUv4HydlkHxeYAZzRxviRJkiSttGbtcpaZrwEHRMTWFJsBAEzIzAkVbR5qgfgkSZIkqUnN3bYZgMz8e0RMKF+v+MM4kiRJkrQSmrvkjIg4LiKeofgBxPcj4umI+HrLhSZJkiRJS9esGZqIOB24iOK3QkaWxZ8BflH+Hoq/PyNJkiRplWvukrPvAidm5m8qyu6OiOeAIfiDmpIkSZJaQXOXnPUCRjVSPqqskyRJkqRVrrkJzUsUvzhf7SjgxeaHI0mSJEnLr7lLzs4Hfh8Re/GvZ2j2BPan8URHkiRJklpcNHe35YjYETgd6F8WjQd+kplPtVBs0ioTEd2BmTNnzqR79+5tHY4kSZKqzJo1i7q6OoC6zJzVVLsVnqGJiE7AL4GLMvNrzQ9RkiRJklbOCj9Dk5kfAEesglgkSZIkaYU0d1OAO4HDWzIQSZIkSVpRzd0U4EVgcETsCYwF5lRWZuZVKxuYJEmSJC1LszYFiIhJS6nOzPxo80OSVj03BZAkSWrfVtmmAACZ2afhdUREWda87dIkSZIkqZma+wwNETEoIp4F5gJzI+LZiDi+5UKTJEmSpKVr1gxNRFxI8Rs0VwOPlcV7AD+LiC0zc3ALxSdJkiRJTWruMzRvAKdk5i1V5UcDV2fmRi0Un7RKNDxDs913f0GHLl3bOhxJkqR2bexlx7X6mMv7DE1zl5x1AsY0Uj6W5u+cJkmSJEkrpLkJzW+BExspPwG4ufnhSJIkSdLyW5nZlEER8Xngb+X73YAtgd9ExE8bGmXm6SsxhiRJkiQ1qbkJzaeAJ8vXHyv/fbM8PlXRzq2cJUmSJK0yzf0dmn1bOhBJkiRJWlHN/h0aSZIkSWprJjSSJEmSapYJjSRJkqSaZUIjSZIkqWaZ0EiSJEmqWSY0kiRJkmpWu09oIiIj4vA2GvvhiLiiBfsbFhF3tlR/K6Mt72u19nRfJEmSVFvaNKEpv8hmI8eIVTjminyR/zJw3qqKpVpE7NPE/ciI6NnCw/UC7m3hPpvrVGBAWwchSZKk2tOsH9ZsYSOAgVVl89oikAYR0Tkz52fm220UQj9gVlXZ9JYcIDNfa8n+miMiOhSh5My2jkWSJEm1qT0sOZuXma9VHe801TgitoiI2yJiRkS8HRF3RUTvqjbfjIjnImJeREyLiGvK8sllkzvKWY/JZfmQiKiPiOMjYhIwtyxfYslZRHSJiB9FxNSy75ciYlBZ1yEiboyISRHxfkRMiIhTm3lPpjdyTxaV4wyLiDsj4szy2t6KiGsjolNFnL0i4p4yjkkRcUxETI6I0yraLJ6pioje5fsvR8RDEfFeRIyLiD2q7utnIuKvZb9TI+KqiFi36v5cHhH/iIg5ETE6IvapqB9Qfm5fjIjnKRLXLauXnJX3/aqI+HH5Gb8WEUOqYtk6Ih6NiLkR8XxEfK49LaOTJElS62gPCc1yK7+03we8C3wW2BOYDYyIiM5lmxOBa4FfAdsAXwReKrvYpfx3IMWSq134l62AIyiWmW3fRAi/AY4GTgH6A98ux4fiXr4KfAX4BHAhcElEHNnsC27avsDHyn+/QbFca0BVnJsB+1Bc0wnAJsvR7/8DLqe4/heAWyKiI0BEfIxiNm04sC1wFPAZ4JqK868B9gC+Wrb5b4rPpm9Fm3WA7wPHA5+k6ZmnbwBzgN2As4DBEXFAGUsH4E7gvbL+hDL2JpXJVveGA1hvWTdDkiRJ7V97WHJ2SETMriq7JDMvaaTtURSJw/GZmQARMRCYQfHl/X7gXOAnmXllxXlPAGTmGxEBMKORJVedgeMy843GgoyIjwNHAgdk5gNl8csN9Zn5AXB+xSmTyhmOI4HbGutzKV4t42zwSmZ+suL9O8DJmbkQ+HtE3APsD1wfEVsDnwN2ycwxZezHAy8ux7iXZ+Y95TnnA89RJHp/B84Gbs7MhhmrFyPiFOCRMonchCJR3DIz/9nQX0QcVJb/Z1nWCTgpM8c1DFp1rQ2ezswLKsY6ubzGPwMHUCR0+zR8jhFxTlnXlLNZ8vMB4OZuV7Le2h2WcpokrRm2HPxMW4cgSc3SHhKah4ATq8qaenZlO4ov2O9WfQleG/hYRGxCMTPxv82I45WmkpnS9sBC4JGmGkTEd4BvAlsCXSmSpPpmxPJZilmoBh9U1T9XJjMNplHMRkHx/M0C4MmGysx8KSKaXMZX4emqPqFIVP5Oce+3jYhjK9oERYLZB/go0AF4oeqz6QK8VfF+ftU4yxNLQzwNs0z9gKlVSenjy+jvh8BPK96vRzGjJkmSpBrWHhKaOZn50rKbAdANGAsc20jdG8CilYljGfXvL60yIr5KsVzrDOAxioTkPyiWRK2oSZk5Yyn11QlO0jLLByv7zfLfhn67Ab8ErmrkvCkUS8wWAjuV/1aqnIF7v2F2bQViaYin2deYmfOo2GyiiVkhSZIk1Zj2kNCsiCcplp1Nz8zqXcCAxQ/+708x89OYDyhmElbUMxRfqPcGHmikfk9gVGZeVxHLx5oxzsqaQPG57kCR/BERWwHrr2S/TwKfaCr5jIinKO7rJpn515Uca1kmAFtExKaZ+XpZtsvSTpAkSdLqqT1sCtAlInpWHRs10fZm4E3groj4bET0ieK3W66KiM3LNkOAMyLilIjoGxE7RsR3K/qYDOxfjrPcX/IzczLwa2BoRBxeMXbDQ/8vAjtHxIER8fGIuIjmf8nepJF70mnZp0Fm/p0i4fpVROwaETtQbJDwPv+adWmOHwGfjohrImL78t4eFuUOcpn5AsXn85tyt7Q+5fhnR8T/WYlxG/NnYCLw64jYNiL2BC4u61bmGiVJklRj2kNCcxDF8xGVx6ONNczM94C9KJY43Q6MB26keIZmVtnm18BpwEkUD7X/D1C5y9YZFA+VTwWeWsFYTwT+AFxH8VzJ9UDDtsW/LGP6PTAa2LBs1xwT+PA92WkFzj8OeB34C3BHGee7lNtRN0dmPk0xO/Vx4K8U9+5C4J8VzQZS7LD2k/Ia7qRI6qY0d9wmYlkIHE6xDO4J4Ab+tctZs69RkiRJtSeW73EG1bJy9moq8LnMbM6GCe1eOUvzKLBVZk5cjvbdgZnPnt3fXc4kCXc5k9T+zJo1i7q6OoC6ph43gdp7hkbLISL2o5i9eIbi93Z+TLHU7i9tGFaLiogvUWw28CLFzndXAiOXJ5mRJEnS6sOEZvXUCbiEYivld4FRwLHlb+WsLtajeK5nS4rnqh6gWE4oSZKkNYgJzWooM+8D7mvrOFalzPwNxfM6kiRJWoO1h00BJEmSJKlZTGgkSZIk1SwTGkmSJEk1y4RGkiRJUs0yoZEkSZJUs0xoJEmSJNUsExpJkiRJNSsys61jkFpdRHQHZs6cOZPu3bu3dTiSJEmqMmvWLOrq6gDqMnNWU+2coZEkSZJUs0xoJEmSJNUsExpJkiRJNcuERpIkSVLNMqGRJEmSVLNMaCRJkiTVLBMaSZIkSTXLhEaSJElSzerY1gFIbWmvc2+hQ5eubR2GpFVk7GXHtXUIkqRVzBkaSZIkSTXLhEaSJElSzTKhkSRJklSzTGgkSZIk1SwTGkmSJEk1y4RGkiRJUs0yoZEkSZJUs0xoJEmSJNUsExpJkiRJNcuERpIkSVLNMqGRJEmSVLNMaNqZiBgQETNauM8hEVHfkn1KkiRJ7cEandBExLCIyEaOrdo6thUREb3LuLdvosnlwP4V7YdFxJ2tEx1ExD5lfD1aa0xJkiStGdbohKY0AuhVdUyqbhQRnVs5rhaTmbMz862W7rct7klEdGrtMSVJktR+mdDAvMx8repYGBEPR8Q1EXFFRLwJ3AcQEZ+KiHsjYnZEvB4Rv42IjRo6K8+7KiJ+HBFvR8RrETGkcsCI6BERvyzPnxsRz0bEIVVtDoyI8eU4IyKiV3MvsHLJWRnLN4DDKmak9inrtoiI2yJiRhn7XRHRu6KfYRFxZ0ScExH/BCaU5V+PiDER8W55vb+LiE3Kut7AQ2UX75TjDSvrJkfEaVWx1lfer7L9iRFxd0TMAc4pyw+LiCfL+/dyRJwfER2be48kSZJUm0xolu4bwHxgT+DfyyVTDwJPATsDBwGbArc1ct4cYDfgLGBwRBwAEBFrAfeWfX4N+ATwA2BhxfnrAGcCXwf2ArakWDbWEi4v462cmRpVznzcB7wLfLaMbzYwomomZn+gH3AA0JCEdQLOA7YDDgd6A8PKuqnAEeXrfuV4p65gzEOAO4BtgKER8VngN8CVFPfv28AAymSnMRHRJSK6NxzAeisYgyRJktoh/6INh0TE7Ir392bmV8rXL2bmWQ0VEXEu8FRm/mdF2TeBqRHx8cx8oSx+OjMvaOgjIk6mSAT+DHwO2BXoX9H+5aqYOgH/npkTyzGuAQav9JVSLD+LiPeBLpn5WsV1fI0iwT0+M7MsGwjMAPYB7i+bzinbzK/oc2jFEC9HxCnAExHRrRzv7bJuemY2Z8OD32XmTRWxDgUuzcxfV4x5HvBj4ILGOgDOBs6vLry525Wst3aHZoQkqRZMufCytg6h3dly8DNtHYIktSgTmmI51IkV7+dUvB5b1XY7YN+qBKjBx4DFCU1V3TRgk/L19sCrFclMY95rSGYaOX9V2Q7YCng3IirL16a4tgbPVCYzABGxE8UsynbA+vxr5m9L4PkWiG1MI7HuGRGVMzIdgLUjYp3MfK+RPn4I/LTi/XrAqy0QmyRJktqQCQ3MycyXmqqret8N+CPw/UbaTqt4/UFVXfKvL/nvL0dMjZ0fjTVsQd0oErhjG6l7o+L1EvckItalWKp2X3nuGxSJzH3AsjYNWMSHr6uxh/4b+xzOB25vpO3cxgbKzHnAvIq4lxGaJEmSaoEJzYp5kuJ5kMmZuaCZfTwNbF61RK21zaeY0aj0JHAUxbKwWSvQ19bAhsAPMnMqQETs3Mh4NDLmGxTP1FCe1x3osxxjPgn0W0oiKkmSpDWEmwKsmGuBDYBbImKXiPhYuRvZTRGxXA9iZOYjwF+A4RFxQET0iYgvRMRBLRBfv4jYvupobMZjMrBtRPSLiI3KNjcDbwJ3RcRny7j2KXds23wpY06hSFi+GxEfjYgvUmwQUOkVilmmQyJi44joVpY/CHy9HG8b4NcsuTlCUy4Ejit3NvtkRPSPiK9GxMXLca4kSZJWIyY0KyAz/0mx+1cHiofknwGuoHhwftEKdHUE8ARwC8UzJj/mw7MXzXErxQ5slcemjbS7nmLL5TEUsyR7ls+d7EWRoNwOjAdupHiGpskZm8x8g2KHsa+U1/IDih3aKtv8g2KJ2KXA68A1ZdUPgUeA/wHuAe4EKp8damrM+yh2WPs8xX38G/A9isRJkiRJa5AoN7SS1ijl8raZz57d313OJK1R3OVMUq2YNWsWdXV1AHVLeyTCGRpJkiRJNcuERpIkSVLNMqGRJEmSVLNMaCRJkiTVLBMaSZIkSTXLhEaSJElSzTKhkSRJklSzTGgkSZIk1SwTGkmSJEk1y4RGkiRJUs2KzGzrGKRWFxHdgZkzZ86ke/fubR2OJEmSqsyaNYu6ujqAusyc1VQ7Z2gkSZIk1SwTGkmSJEk1y4RGkiRJUs0yoZEkSZJUs0xoJEmSJNUsExpJkiRJNcuERpIkSVLNMqGRJEmSVLM6tnUAUlva69xb6NCla1uHIUlaA4y97Li2DkFaLTlDI0mSJKlmmdBIkiRJqlkmNJIkSZJqlgmNJEmSpJplQiNJkiSpZpnQSJIkSapZJjSSJEmSapYJjSRJkqSaZUIjSZIkqWaZ0EiSJEmqWSY0kiRJkmqWCc1qJCIGRMSMZbQZFhF3Vrx/OCKuWPXRLb+ImBwRp7V1HJIkSWr/TGhqTJmQZHnMj4iXImJwRHRczi5OBQaswhAlSZKkVrO8X4LVvowABgJdgIOBa4EPgGnLOjEzZ67a0CRJkqTW4wxNbZqXma9l5iuZ+XPgAeCLDZURcWBEjI+I2RExIiJ6VdQtseSsWkSsHxG/iYh3IuK9iLg3IvpW1P9bRPyxrJ8TEc9FxMEV9XtHxOMRMS8ipkXEpZWzR+USt2vKY2ZEvBkRF0VEVIWyTkQMjYh3I2JKRJxQ0ceDEXFNVdwblzNW+6/gvZQkSVINM6FZPbwPdC5frwOcCXwd2AvYErh8BfoaBuxMkSDtAQTwp4joVNZfSzEztBewDfB9YDZARHwE+BPwBLAdcCIwCDi3aoxvAAuAXSmWwJ0OHF/V5gxgDLADcB3w84joV9bdABwTEV0q2n8N+AfwYGMXFRFdIqJ7wwGstxz3QpIkSe2cS85qWDmrsT9wIHB1WdwJ+PfMnFi2uQYYvJz99aVIZPbMzFFl2bHAVOBw4L8pEqThmflMedrLFV2cVLY9OTMT+HtEbAb8KCIuzMxFZbupwPfKNhMiYhvge8D1FX39KTOvK2P4UVm/LzABuB24BjgMuK1sPwAYVvbZmLOB86sLb+52Jeut3WGZ90aSatGWg59ZdiNJqnHO0NSmQyJiNjAXuBf4PTCkrHuvIZkpTQM2Wc5++1PMnIxuKMjMtyiSiP5l0VXAuRExMiIuiIhtq85/rCqpGAl0AzavKPtbVZvHgL4RUZlZPF0RQwKvNVxHZs4Ffgt8EyAidgQ+RTG71JQfAnUVx+ZLaStJkqQaYUJTmx4Ctgf6Al0z8xuZOaes+6CqbVIsG2sRmXkD8FGKhGIbYExEfLel+q/Q2HVU/vd6A3BARGxOsUHCg5n5SlOdZea8zJzVcADvtnjEkiRJanUmNLVpTma+lJlTMnNBC/Y7nmIZ4m4NBRGxIdAPeL6hLDOnZuYvMvPLwE+Ab1Wcv0fVA/57UiQPr1aU7caSdgdezMyFyxtoueRtTDn2McDQ5T1XkiRJqw8TGi2WmS8CdwHXR8RnImI74L8oHra/CyAirih3UetTLvXalyKRgeLh/f/f3t0H21WVdxz//kikUCAXEUmmxTQ0QkRedAyBUpRX0wBjfWm10nam4LRxtO+d1gL+hVarbS0NBQo6Uusrtk7VDlCkpm2mqWBaCC9JNUYkUCEQEUhuUgghZvWPte905/QmYHLuOXff+/3MrLl377XOPus852WfZ++113kZcHWSVyR5E/A+4MrW9TMAc5NcmWRBkl8Efgu4ah+6/HHgMuoZqC/tw+0lSZLUcSY06vUO4C7gZuq1LQEuLKWMDQGbQZ3p7JvU38NZT50MgFLKI9TfxTkVuBe4HrgB+EDPfXwKOBj4j2ZbVwEf24e+3ki95ufG5roaSZIkTTPZ86RQUv8lWQHcU0r53T5sax7wHWBRKWX1D3nbWcCWtZcf7yxnkqYsZzmT1GWjo6OMjIwAjDTXQI/LaZvVOc1v4ryEeubn6z9sMiNJkqSpwyFn6qIzqNNRLwLeNeS+SJIkaYg8Q6OBKqWc3YdtrKCPU1FLkiSpuzxDI0mSJKmzTGgkSZIkdZYJjSRJkqTOMqGRJEmS1FkmNJIkSZI6y4RGkiRJUmeZ0EiSJEnqrJRSht0HaeCSzAK2bNmyhVmzZg27O5IkSeoxOjrKyMgIwEgpZXRP7TxDI0mSJKmzTGgkSZIkdZYJjSRJkqTOmjnsDkjDNDq6x+GYkiRJGqIX+j3NSQE0LSWZB2wYcjckSZL0/I4upTyyp0rP0Gi6erL5ezSwdZgdmeIOAx7GOE804zwYxnniGePBMM6DYZz74zBg494amNBoutu6t2kAtX+SjP1rnCeQcR4M4zzxjPFgGOfBMM5987yxc1IASZIkSZ1lQiNJkiSps0xoNF09C7yv+auJY5wHwzgPhnGeeMZ4MIzzYBjnAXGWM0mSJEmd5RkaSZIkSZ1lQiNJkiSps0xoJEmSJHWWCY0kSZKkzjKh0bST5DeSPJhke5JVSU4ddp8miyRnJrkpycYkJcmbe+qT5P1JHk3yTJLlSY7taXNEks8mGU2yOckNSQ7taXNykpXNc/DdJH84Tl/elmRd02ZNkgsn5lEPXpLLk/xnkq1Jvpfky0kW9LQ5KMm1SZ5Isi3J3yeZ3dNmbpJbkjzdbOfPkszsaXN2ktVJnk1yf5JLxunPlHxPJHl3kvua1+JokjuSXNCqN8Z9luSy5rNjWWudcd5PSa5o4tou61r1xrhPkvx4ks80sXym2f+c0qpP3A9OPqUUi2XaFODt1OkT3wG8EvgY8BRw1LD7NhkKcAHwAeAtQAHe3FN/KbAZeBNwMvAPwAPAQa02twL3AKcBrwW+DXyuVT8LeAz4DHACcBHwNPDOVpufBnYC7wGOB/4I2AGcOOwY9SnOXwEuaR7/q4BbgIeAQ1ptrgP+GzgXWAjcAXytVT8DWAN8FXh189w9Dvxxq80xwP8Af97E8TebuC5ptZmy7wngZ4ELgWOB44APNq+jE4zxhMR7EbABuBdY5mu5r7G9AlgLzGmVI41x3+P8YuBB4BPAqU1MfgaY32rjfnASlqF3wGIZZAFWAde0lg8AHgEuG3bfJluhJ6EBAjwK/EFr3QiwHbioWT6+ud0prTbnA7uAH2uW3w08CRzYavNhYF1r+W+Bm3v683Xg+mHHZYJi/dImbme24roDeGurzSuaNj/VLF8A/ACY3WrzLmDLWGyBPwHW9tzX54GvtJan1Xuiee39qjHue1wPBdYDrwdW0CQ0xrlv8b0CuGcPdca4f3H+MLByL/XuBydpcciZpo0kB1KPXC0fW1dK2dUsnz6sfnXIMdSjgu34baHu4MbidzqwuZRyZ+t2y6kf5Ke12vxbKWVHq81twIIkL261Wc7ubmPqPk8jzd8nm78LgRexe6zXUY/AtmO9ppSyqbWd26hH/k5otdljHKfTeyLJjCQXAYdQj14b4/66FrillNIbC+PcP8emDgd+oBnONLdZb4z7543AnUm+0AzLuzvJ0la9+8FJyoRG08mR1NPum3rWb6J+QGnvxmK0t/jNAb7Xriyl7KR+UW+3GW8bvIA2U+55SnIAsIw6PGRts3oOsKOUsrmneW+s9zWOs5IczDR4TyQ5Kck26jCZ64G3lFK+gTHumyZRfA1w+TjVxrk/VlGHqZ5PPbp/DLAyyWEY4376SWp8vw0soQ7l+8skFzf17gcnqZnP30SSNIGuBU6kjrNW/32Les3ACPBW4JNJzhpul6aOJC8DrgIWl1K2D7s/U1Up5dbW4n1JVlGvu/sF4Jnh9GpKOgC4s5Ty3mb57iQnUofnfXJ43dLz8QyNppPv04wh7lk/m3pxnvZuLEZ7i99jwFHtymYWnSN62oy3DV5Amyn1PCW5BngDcE4p5eFW1WPAgUkO77lJb6z3NY6jpZRnmAbviVLKjlLK/aWUu0opl1MvWP8djHG/LKS+51cn2ZlkJ3AW8NvN/5swpgFzWAAABPlJREFUzn3XnI1ZD7wcX8v99CjwjZ513wTGhve5H5ykTGg0bTRjVe8Czhtb1wz3OY86pl57t4H6QdqO3yzqmOCx+N0BHJ5kYet251I/a1a12pyZ5EWtNouBb5VSnmq1OY/dLWaKPE/NtJ/XUGeTO7eUsqGnyV3Ac+we6wXUnWo71iclae84FwOj/N8Oea9xnKbviQOAH8EY98s/AydRz4KNlTuBz7b+N8591kwBPJ/6BdzXcv98DVjQs+446tkwcD84eQ17VgKLZZCFOuXkduBi6kwkH6VOOTl72H2bDIU6U9HYl5IC/F7z/9ym/tImXm+kfon5MuNPV7maOuXlGdSjiO3pKkeoO4RPUS9GfTt1qtDe6SqfA36fOlvPFUyh6SqBv6JO+3kWu0/DenCrzXXUneg51KPgtwO3t+rHpmG9jTr18xLquO3xpmH90yaOv87407BOyfcE8CHgTGBe83r9EPXC3MXGeELjvoL/P22zcd6/mH6k+byY13w+fpU67fJLjXFf47yIuu95L/Xs1y81MfnlVhv3g5OwDL0DFsugC3Vu/YeoFwmvAk4bdp8mSwHOpiYyveVvmvoA728+iLdTZ2A5rmcbRwCfA7ZSpwT9a+DQnjYnAyubbTwMXDpOX95Gvf7hWervL1w47Pj0Mc7jxbgAl7TaHES9vubJZkf3RWBOz3Z+AvhH6u8XPE790jNznOf07iaO32nfR6vNlHxPADdQf1PiWeqXt+U0yYwxntC4r2D3hMY4739MPw9sbB7Xw81y+7dRjHH/Yv0GavK3nTrcbGlPvfvBSVjSBEySJEmSOsdraCRJkiR1lgmNJEmSpM4yoZEkSZLUWSY0kiRJkjrLhEaSJElSZ5nQSJIkSeosExpJkiRJnWVCI0mSJKmzTGgkSZIkdZYJjSRJkqTOMqGRJGmIkqxIcnWSZUmeSrIpydIkhyT5RJKtSe5PckHrNicmuTXJtqb9p5Mc2ao/P8m/J9mc5IkkNyeZ36qfl6Qk+bkk/5rk6ST3Jjl90I9fkvaXCY0kScN3MfB94FTgauA64AvA7cBrgH8CPp3kR5McDvwLcDdwCnA+MBv4u9b2DgGubOrPA3YBX0rSu9//IPAR4NXAeuDGJDMn4gFK0kRJKWXYfZAkadpKsgKYUUp5XbM8A9gCfLGU8ivNujnAo8DpwOuB15VSlrS2cTTwXWBBKWX9OPdxJPA4cFIpZW2SecAG4NdKKTc0bV4J/BdwfCll3cQ8WknqP8/QSJI0fPeN/VNK+QHwBLCmVb+p+XsU8CrgnGa42bYk24CxBGQ+QJJjk9yY5IEko8CDTf3cPd0vNWEauw9J6gxPK0uSNHzP9SyX9rpSSkkC9UDkocBNwKXjbGcsKbkJeAhYCmxsbrcWOHAv9zs2ZMODnZI6xYRGkqRuWQ38PPBgKWVnb2WSlwALgKWllJXNutcOtouSNDgehZEkqVuuBY6gXsC/KMn8JEuaGdFmAE9Rh6y9M8nLk5xLnSBAkqYkExpJkjqklLIROAOYQZ39bA2wDNgM7Cql7AIuAhZSh5n9BfCe4fRWkiaes5xJkiRJ6izP0EiSJEnqLBMaSZIkSZ1lQiNJkiSps0xoJEmSJHWWCY0kSZKkzjKhkSRJktRZJjSSJEmSOsuERpIkSVJnmdBIkiRJ6iwTGkmSJEmdZUIjSZIkqbNMaCRJkiR11v8CP+GHLsf+yNIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Mean predicted vs actual wage by program\n",
    "original_summary = utilities.multiple_describe(data, 'program', 'earnings_end')\n",
    "adjusted_summary = utilities.multiple_describe(data, 'program', 'predicted_wage')\n",
    "original_summary['value'] = \"actual\"\n",
    "adjusted_summary['value'] = \"predicted\"\n",
    "\n",
    "sns.barplot(x=\"mean\", y=\"program\", hue=\"value\", data=pd.concat([original_summary, adjusted_summary]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fc333e87d50>"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=\"earnings_premium\", y=\"program\", data=data, estimator=np.mean)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 8: Calculate employment premium__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Students and decisionmakers may also ask the question: \"If I participate in a given program, how much likelier am I to get a job than if I hadn't participated?\"\n",
    "\n",
    "The `metrics.Employment_Likelihood()` method helps to answer that question in the same fashion as the trend-adjusted earnings premium calculated above in Section 7. In addition to YYYY-MM dates of entry and exit, this method also takes two binary variables-- one denoting whether or not an individual was employed before program entry, and one indicating whether they were employed after graduation.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This method uses the `state`, `year_month_entry` and `year_month_exit` arguments to correct for trend -- for each individual in the dataset, it uses BLS data to calculate the overall change in the employment rate in their state over the period of time they were in the program, and incorporates it into the change in employment rate for program completers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mattlerner/roi/sbir-roi/roi/utilities.py:135: UserWarning: State codes, though integers, should be passed as strings. Something else was passed. Attempting to coerce to string.\n",
      "  warnings.warn(\"State codes, though integers, should be passed as strings. Something else was passed. Attempting to coerce to string.\")\n"
     ]
    }
   ],
   "source": [
    "# library method\n",
    "employment = metrics.Employment_Likelihood(data, 'program', 'year_month_entry', 'year_month_exit', 'employed_at_end', 'employed_at_start', 'state')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# library method\n",
    "premium = employment.employment_premium"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Average employment premium')"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=\"mean_employment_premium\", y=\"program\", data=premium).set_title(\"Average employment premium\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Seciton 9: Calculate time to completion and completion rate statistics__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `metrics.Completion()` class is simple -- it's a wrapper for arithmetic calculations about completion rate and time-to-completion across programs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# library method\n",
    "completion = metrics.Completion(data, 'program', 'program_start', 'program_end', 'completer')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "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>program</th>\n",
       "      <th>n</th>\n",
       "      <th>mean</th>\n",
       "      <th>median</th>\n",
       "      <th>sd</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2288</td>\n",
       "      <td>3.101</td>\n",
       "      <td>3</td>\n",
       "      <td>1.764</td>\n",
       "      <td>-1</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Art History</td>\n",
       "      <td>2370</td>\n",
       "      <td>3.361</td>\n",
       "      <td>3</td>\n",
       "      <td>1.711</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Biology</td>\n",
       "      <td>2404</td>\n",
       "      <td>3.192</td>\n",
       "      <td>3</td>\n",
       "      <td>1.725</td>\n",
       "      <td>-1</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Economics</td>\n",
       "      <td>2327</td>\n",
       "      <td>3.713</td>\n",
       "      <td>3</td>\n",
       "      <td>1.792</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2386</td>\n",
       "      <td>4.009</td>\n",
       "      <td>4</td>\n",
       "      <td>1.732</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>French Literature</td>\n",
       "      <td>2339</td>\n",
       "      <td>3.041</td>\n",
       "      <td>3</td>\n",
       "      <td>1.709</td>\n",
       "      <td>-1</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Philosophy</td>\n",
       "      <td>2297</td>\n",
       "      <td>3.527</td>\n",
       "      <td>3</td>\n",
       "      <td>1.713</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  program     n  mean  median    sd  min  max\n",
       "0            Anthropology  2288 3.101       3 1.764   -1   10\n",
       "1             Art History  2370 3.361       3 1.711    0   10\n",
       "2                 Biology  2404 3.192       3 1.725   -1   11\n",
       "3               Economics  2327 3.713       3 1.792    0   12\n",
       "4  Electrical Engineering  2386 4.009       4 1.732    0   11\n",
       "5       French Literature  2339 3.041       3 1.709   -1   10\n",
       "6              Philosophy  2297 3.527       3 1.713    0   11"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "completion_time = completion.time_to_completion\n",
    "completion_time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>program</th>\n",
       "      <th>n</th>\n",
       "      <th>mean</th>\n",
       "      <th>median</th>\n",
       "      <th>sd</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2288</td>\n",
       "      <td>0.669</td>\n",
       "      <td>1</td>\n",
       "      <td>0.471</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Art History</td>\n",
       "      <td>2370</td>\n",
       "      <td>0.682</td>\n",
       "      <td>1</td>\n",
       "      <td>0.466</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Biology</td>\n",
       "      <td>2404</td>\n",
       "      <td>0.671</td>\n",
       "      <td>1</td>\n",
       "      <td>0.470</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Economics</td>\n",
       "      <td>2327</td>\n",
       "      <td>0.634</td>\n",
       "      <td>1</td>\n",
       "      <td>0.482</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2386</td>\n",
       "      <td>0.592</td>\n",
       "      <td>1</td>\n",
       "      <td>0.492</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>French Literature</td>\n",
       "      <td>2339</td>\n",
       "      <td>0.702</td>\n",
       "      <td>1</td>\n",
       "      <td>0.458</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Philosophy</td>\n",
       "      <td>2297</td>\n",
       "      <td>0.654</td>\n",
       "      <td>1</td>\n",
       "      <td>0.476</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  program     n  mean  median    sd  min  max\n",
       "0            Anthropology  2288 0.669       1 0.471    0    1\n",
       "1             Art History  2370 0.682       1 0.466    0    1\n",
       "2                 Biology  2404 0.671       1 0.470    0    1\n",
       "3               Economics  2327 0.634       1 0.482    0    1\n",
       "4  Electrical Engineering  2386 0.592       1 0.492    0    1\n",
       "5       French Literature  2339 0.702       1 0.458    0    1\n",
       "6              Philosophy  2297 0.654       1 0.476    0    1"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "completion_rate = completion.completion_rates\n",
    "completion_rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Average time to complete')"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=\"mean\", y=\"program\", data=completion_time).set_title(\"Average time to complete\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Average completion rate')"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=\"mean\", y=\"program\", data=completion_rate).set_title(\"Average completion rate\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 9: Try out loan amounts__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `cost` submodule currently contains just one class, `Compound_Interest_Loan`, which is a parent class for several separate methods related to calculating statistics about compound interest loans. These methods take either scalars or numpy arrays. Here, we merge programs data with student data and calculate the average period payment, including interest, on the assumption that each student takes out a loan with 5% interest and a term of 10 years. However, instead of the `0.05` and `10` that we give as arguments, we could also supply additional columns of the dataframe corresponding to the differing loan terms faced by each student."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_programs = data.merge(programs, left_on=\"program\", right_on=\"programs\", how=\"left\")\n",
    "data_programs['total_cost'] = data_programs['cost_per_semester'] * data_programs['length']\n",
    "\n",
    "# library method\n",
    "data_programs['period_payment'] = cost.Compound_Interest_Loan.calculate_period_payment(data_programs['total_cost'], 0.05, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0_x</th>\n",
       "      <th>race</th>\n",
       "      <th>gender</th>\n",
       "      <th>earnings_start</th>\n",
       "      <th>earnings_end</th>\n",
       "      <th>age</th>\n",
       "      <th>program</th>\n",
       "      <th>program_start</th>\n",
       "      <th>program_end</th>\n",
       "      <th>education_level</th>\n",
       "      <th>...</th>\n",
       "      <th>trend_adjusted_change</th>\n",
       "      <th>predicted_wage</th>\n",
       "      <th>earnings_premium</th>\n",
       "      <th>Unnamed: 0_y</th>\n",
       "      <th>programs</th>\n",
       "      <th>degree</th>\n",
       "      <th>length</th>\n",
       "      <th>cost_per_semester</th>\n",
       "      <th>total_cost</th>\n",
       "      <th>period_payment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12509</th>\n",
       "      <td>12510</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>20251.000</td>\n",
       "      <td>38710.800</td>\n",
       "      <td>18</td>\n",
       "      <td>Philosophy</td>\n",
       "      <td>2012</td>\n",
       "      <td>2016</td>\n",
       "      <td>123</td>\n",
       "      <td>...</td>\n",
       "      <td>14346.600</td>\n",
       "      <td>29610.231</td>\n",
       "      <td>9100.569</td>\n",
       "      <td>7</td>\n",
       "      <td>Philosophy</td>\n",
       "      <td>BA</td>\n",
       "      <td>8</td>\n",
       "      <td>8000</td>\n",
       "      <td>64000</td>\n",
       "      <td>8288.293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9379</th>\n",
       "      <td>9380</td>\n",
       "      <td>Other</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>19990.000</td>\n",
       "      <td>48732.625</td>\n",
       "      <td>38</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2015</td>\n",
       "      <td>2018</td>\n",
       "      <td>111</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>22274.278</td>\n",
       "      <td>26458.347</td>\n",
       "      <td>1</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>BA</td>\n",
       "      <td>8</td>\n",
       "      <td>5000</td>\n",
       "      <td>40000</td>\n",
       "      <td>5180.183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15682</th>\n",
       "      <td>15683</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Male</td>\n",
       "      <td>26940.000</td>\n",
       "      <td>35998.200</td>\n",
       "      <td>48</td>\n",
       "      <td>Biology</td>\n",
       "      <td>2015</td>\n",
       "      <td>2018</td>\n",
       "      <td>125</td>\n",
       "      <td>...</td>\n",
       "      <td>6594.440</td>\n",
       "      <td>26093.741</td>\n",
       "      <td>9904.459</td>\n",
       "      <td>5</td>\n",
       "      <td>Biology</td>\n",
       "      <td>AS</td>\n",
       "      <td>3</td>\n",
       "      <td>3000</td>\n",
       "      <td>9000</td>\n",
       "      <td>1165.541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6365</th>\n",
       "      <td>6366</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Male</td>\n",
       "      <td>33560.000</td>\n",
       "      <td>33022.400</td>\n",
       "      <td>57</td>\n",
       "      <td>Anthropology</td>\n",
       "      <td>2010</td>\n",
       "      <td>2012</td>\n",
       "      <td>81</td>\n",
       "      <td>...</td>\n",
       "      <td>-2503.720</td>\n",
       "      <td>31185.951</td>\n",
       "      <td>1836.449</td>\n",
       "      <td>4</td>\n",
       "      <td>Anthropology</td>\n",
       "      <td>BA</td>\n",
       "      <td>4</td>\n",
       "      <td>4500</td>\n",
       "      <td>18000</td>\n",
       "      <td>2331.082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7942</th>\n",
       "      <td>7943</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>33839.000</td>\n",
       "      <td>93123.800</td>\n",
       "      <td>43</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2016</td>\n",
       "      <td>2021</td>\n",
       "      <td>92</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>38683.523</td>\n",
       "      <td>54440.277</td>\n",
       "      <td>3</td>\n",
       "      <td>Economics</td>\n",
       "      <td>MA</td>\n",
       "      <td>4</td>\n",
       "      <td>10000</td>\n",
       "      <td>40000</td>\n",
       "      <td>5180.183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13186</th>\n",
       "      <td>13187</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Female</td>\n",
       "      <td>20780.000</td>\n",
       "      <td>37186.930</td>\n",
       "      <td>37</td>\n",
       "      <td>Art History</td>\n",
       "      <td>2017</td>\n",
       "      <td>2021</td>\n",
       "      <td>124</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>23903.179</td>\n",
       "      <td>13283.751</td>\n",
       "      <td>2</td>\n",
       "      <td>Art History</td>\n",
       "      <td>BA</td>\n",
       "      <td>8</td>\n",
       "      <td>7000</td>\n",
       "      <td>56000</td>\n",
       "      <td>7252.256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13908</th>\n",
       "      <td>13909</td>\n",
       "      <td>Other</td>\n",
       "      <td>Male</td>\n",
       "      <td>19500.000</td>\n",
       "      <td>27138.600</td>\n",
       "      <td>57</td>\n",
       "      <td>Biology</td>\n",
       "      <td>2013</td>\n",
       "      <td>2014</td>\n",
       "      <td>124</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>18560.128</td>\n",
       "      <td>8578.472</td>\n",
       "      <td>5</td>\n",
       "      <td>Biology</td>\n",
       "      <td>AS</td>\n",
       "      <td>3</td>\n",
       "      <td>3000</td>\n",
       "      <td>9000</td>\n",
       "      <td>1165.541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11060</th>\n",
       "      <td>11061</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Male</td>\n",
       "      <td>31051.000</td>\n",
       "      <td>70767.600</td>\n",
       "      <td>64</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>2013</td>\n",
       "      <td>2018</td>\n",
       "      <td>123</td>\n",
       "      <td>...</td>\n",
       "      <td>34875.920</td>\n",
       "      <td>20807.428</td>\n",
       "      <td>49960.172</td>\n",
       "      <td>1</td>\n",
       "      <td>Electrical Engineering</td>\n",
       "      <td>BA</td>\n",
       "      <td>8</td>\n",
       "      <td>5000</td>\n",
       "      <td>40000</td>\n",
       "      <td>5180.183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8098</th>\n",
       "      <td>8099</td>\n",
       "      <td>Asian or Pacific Islander</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>22324.000</td>\n",
       "      <td>43047.060</td>\n",
       "      <td>22</td>\n",
       "      <td>Economics</td>\n",
       "      <td>2013</td>\n",
       "      <td>2016</td>\n",
       "      <td>92</td>\n",
       "      <td>...</td>\n",
       "      <td>17307.700</td>\n",
       "      <td>29585.206</td>\n",
       "      <td>13461.854</td>\n",
       "      <td>3</td>\n",
       "      <td>Economics</td>\n",
       "      <td>MA</td>\n",
       "      <td>4</td>\n",
       "      <td>10000</td>\n",
       "      <td>40000</td>\n",
       "      <td>5180.183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16100</th>\n",
       "      <td>16101</td>\n",
       "      <td>Black</td>\n",
       "      <td>Other/unspecified</td>\n",
       "      <td>34444.000</td>\n",
       "      <td>nan</td>\n",
       "      <td>25</td>\n",
       "      <td>French Literature</td>\n",
       "      <td>2017</td>\n",
       "      <td>2017</td>\n",
       "      <td>125</td>\n",
       "      <td>...</td>\n",
       "      <td>nan</td>\n",
       "      <td>34444.000</td>\n",
       "      <td>nan</td>\n",
       "      <td>6</td>\n",
       "      <td>French Literature</td>\n",
       "      <td>MA</td>\n",
       "      <td>5</td>\n",
       "      <td>6000</td>\n",
       "      <td>30000</td>\n",
       "      <td>3885.137</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 38 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0_x                       race             gender  \\\n",
       "12509         12510                      White               Male   \n",
       "9379           9380                      Other  Other/unspecified   \n",
       "15682         15683  Asian or Pacific Islander               Male   \n",
       "6365           6366  Asian or Pacific Islander               Male   \n",
       "7942           7943                      White             Female   \n",
       "13186         13187  Asian or Pacific Islander             Female   \n",
       "13908         13909                      Other               Male   \n",
       "11060         11061  Asian or Pacific Islander               Male   \n",
       "8098           8099  Asian or Pacific Islander  Other/unspecified   \n",
       "16100         16101                      Black  Other/unspecified   \n",
       "\n",
       "       earnings_start  earnings_end  age                 program  \\\n",
       "12509       20251.000     38710.800   18              Philosophy   \n",
       "9379        19990.000     48732.625   38  Electrical Engineering   \n",
       "15682       26940.000     35998.200   48                 Biology   \n",
       "6365        33560.000     33022.400   57            Anthropology   \n",
       "7942        33839.000     93123.800   43               Economics   \n",
       "13186       20780.000     37186.930   37             Art History   \n",
       "13908       19500.000     27138.600   57                 Biology   \n",
       "11060       31051.000     70767.600   64  Electrical Engineering   \n",
       "8098        22324.000     43047.060   22               Economics   \n",
       "16100       34444.000           nan   25       French Literature   \n",
       "\n",
       "       program_start  program_end  education_level  ...  \\\n",
       "12509           2012         2016              123  ...   \n",
       "9379            2015         2018              111  ...   \n",
       "15682           2015         2018              125  ...   \n",
       "6365            2010         2012               81  ...   \n",
       "7942            2016         2021               92  ...   \n",
       "13186           2017         2021              124  ...   \n",
       "13908           2013         2014              124  ...   \n",
       "11060           2013         2018              123  ...   \n",
       "8098            2013         2016               92  ...   \n",
       "16100           2017         2017              125  ...   \n",
       "\n",
       "       trend_adjusted_change  predicted_wage  earnings_premium  Unnamed: 0_y  \\\n",
       "12509              14346.600       29610.231          9100.569             7   \n",
       "9379                     nan       22274.278         26458.347             1   \n",
       "15682               6594.440       26093.741          9904.459             5   \n",
       "6365               -2503.720       31185.951          1836.449             4   \n",
       "7942                     nan       38683.523         54440.277             3   \n",
       "13186                    nan       23903.179         13283.751             2   \n",
       "13908                    nan       18560.128          8578.472             5   \n",
       "11060              34875.920       20807.428         49960.172             1   \n",
       "8098               17307.700       29585.206         13461.854             3   \n",
       "16100                    nan       34444.000               nan             6   \n",
       "\n",
       "                     programs degree length cost_per_semester  total_cost  \\\n",
       "12509              Philosophy     BA      8              8000       64000   \n",
       "9379   Electrical Engineering     BA      8              5000       40000   \n",
       "15682                 Biology     AS      3              3000        9000   \n",
       "6365             Anthropology     BA      4              4500       18000   \n",
       "7942                Economics     MA      4             10000       40000   \n",
       "13186             Art History     BA      8              7000       56000   \n",
       "13908                 Biology     AS      3              3000        9000   \n",
       "11060  Electrical Engineering     BA      8              5000       40000   \n",
       "8098                Economics     MA      4             10000       40000   \n",
       "16100       French Literature     MA      5              6000       30000   \n",
       "\n",
       "       period_payment  \n",
       "12509        8288.293  \n",
       "9379         5180.183  \n",
       "15682        1165.541  \n",
       "6365         2331.082  \n",
       "7942         5180.183  \n",
       "13186        7252.256  \n",
       "13908        1165.541  \n",
       "11060        5180.183  \n",
       "8098         5180.183  \n",
       "16100        3885.137  \n",
       "\n",
       "[10 rows x 38 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_programs.sample(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_programs['premium_with_payments'] = data_programs['earnings_premium'] - data_programs['period_payment']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "summary_w_payments = utilities.multiple_describe(data_programs, 'program', 'premium_with_payments')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Return including loan payments')"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=\"mean\", y=\"program\", data=summary_w_payments).set_title(\"Return including loan payments\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 10: High school baselines__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `surveys.CPS_Ops()` class contains several methods. Here, we use the `.rudimentary_hs_baseline()` method, which takes a state and a list of numbers. It uses CPS data to provide a rudimentary forecast of expected wages for high school graduates in the state and over the time period provided. It does this simply by looking at historical wages for high school graduates over the past 10 years of CPS data. That's why __rudimentary__ is in the name -- this is a blunt instrument suitable only for providing a baseline forecast in a graph or other simple visualizaion. It should not be used for calculation or evaluation purposes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mattlerner/roi/sbir-roi/roi/utilities.py:135: UserWarning: State codes, though integers, should be passed as strings. Something else was passed. Attempting to coerce to string.\n",
      "  warnings.warn(\"State codes, though integers, should be passed as strings. Something else was passed. Attempting to coerce to string.\")\n"
     ]
    }
   ],
   "source": [
    "cps = surveys.CPS_Ops()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[15017.630256967966, 21688.246894625805, 34336.28062777556]"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# library method\n",
    "cps.rudimentary_hs_baseline('08', [1, 5, 10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "State_To_FIPS_series usually takes a pandas series. Something else was passed. Now trying under assumption that passed object is list or array\n"
     ]
    }
   ],
   "source": [
    "states = ['CA','TX','AK','MS','CO','NY','NJ','PA']\n",
    "fips = utilities.State_To_FIPS_series(states)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['06', '48', '02', '28', '08', '36', '34', '42']"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fips"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "baselines = pd.DataFrame([cps.rudimentary_hs_baseline(state, [1, 5, 10]) for state in fips])\n",
    "baselines['state'] = states\n",
    "baselines.columns = ['1','5','10','State']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "baselines_melted = pd.melt(baselines, id_vars=['State'], var_name=\"year\", value_name=\"earnings\")\n",
    "baselines_melted['year'] = baselines_melted['year'].astype(int)\n",
    "baselines_melted['earnings'] = baselines_melted['earnings'].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7ff183195a90>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lineplot(data=baselines_melted, x=\"year\", y=\"earnings\", hue=\"State\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Section 11: Completion rate vs earnings premium__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's an example of one of the ways this data can be presented: a scatterplot plotting completion rate against employment premium. No ROI or equity statistic is particularly useful in isolation: metrics can provide only clues or context for futher investigation, and need to be considered holistically."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "merged_rate_premium = completion_rate.merge(premium, on=\"program\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0, 0.5, 'Increased likelihood of employment'),\n",
       " Text(0.5, 0, 'Completion rate'),\n",
       " Text(0.5, 1.0, 'Completion rate vs employment premium')]"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "sns.scatterplot(data=merged_rate_premium, x=\"mean\", y=\"mean_employment_premium\", ax=ax, s=100)\n",
    "ax.set(xlabel='Completion rate', ylabel='Increased likelihood of employment', title='Completion rate vs employment premium')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}