regress.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 0,
        "width": 12
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "source": [
    "## Use case 2: hourly rates as a function of skills and other profile variables.  Regression on numeric predictors plus cluster categories provides a statistically significant but very weak model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 4,
        "width": 12
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "source": [
    "## Why?  Omitted variable bias: the most crucial variable missing for each freelancer is active time since joining Upwork.  Hourly rates should be correlated with experience on Upwork: more proven ability, better ability to negotiate with employers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import copy\n",
    "import glob\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import statsmodels.api as sm\n",
    "from sklearn.model_selection import StratifiedKFold, KFold\n",
    "from sklearn.preprocessing import StandardScaler, scale\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score\n",
    "from sklearn.neighbors import KNeighborsRegressor\n",
    "from sklearn.linear_model import LinearRegression, LogisticRegression, Ridge\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.naive_bayes import GaussianNB, MultinomialNB, BernoulliNB\n",
    "from sklearn.tree import DecisionTreeRegressor, DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestRegressor, RandomForestClassifier, AdaBoostRegressor, AdaBoostClassifier\n",
    "from sklearn.metrics import mean_squared_error, r2_score, accuracy_score, f1_score, roc_curve, auc, log_loss\n",
    "from keras.models import Sequential\n",
    "from keras import initializers\n",
    "from keras.layers import Dense\n",
    "from keras import optimizers\n",
    "from kmodes.kmodes import KModes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df = pd.concat([pd.read_csv(f) for f in glob.glob('./data/*')], ignore_index = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df = df.drop(['Unnamed: 0'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 8,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4950, 10)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 12,
        "hidden": false,
        "row": 8,
        "width": 8
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "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>bio</th>\n",
       "      <th>earned</th>\n",
       "      <th>edu</th>\n",
       "      <th>exp</th>\n",
       "      <th>headline</th>\n",
       "      <th>hours</th>\n",
       "      <th>jobs</th>\n",
       "      <th>rate</th>\n",
       "      <th>skills</th>\n",
       "      <th>tests</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Google Adwords &amp; Bing Ads Certified MBA market...</td>\n",
       "      <td>$200k+</td>\n",
       "      <td>3.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>Marketing, Strategic Mgmt/Writer/SEO, SEM/Mark...</td>\n",
       "      <td>10,654</td>\n",
       "      <td>40</td>\n",
       "      <td>$45.00</td>\n",
       "      <td>['Strategic Planning', 'Business Analysis', 'A...</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>I have been transcribing for more than 7 years...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Transcriptionist</td>\n",
       "      <td>1,043</td>\n",
       "      <td>60</td>\n",
       "      <td>$12.00</td>\n",
       "      <td>['Transcription', 'Medical Transcription']</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Throughout my undergraduate years, I was a res...</td>\n",
       "      <td>$60k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Transcription (98WPM), Virtual Assistant</td>\n",
       "      <td>2,840</td>\n",
       "      <td>254</td>\n",
       "      <td>$15.00</td>\n",
       "      <td>['Biology', 'English', 'Transcription', 'Virtu...</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>As a digital marketing professional with a foc...</td>\n",
       "      <td>$30k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Full-Stack Digital Marketer | Facebook Ads Spe...</td>\n",
       "      <td>673</td>\n",
       "      <td>21</td>\n",
       "      <td>$75.00</td>\n",
       "      <td>['Facebook Marketing', 'Inbound Marketing', 'M...</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>WordPress Development + Theme Customization + ...</td>\n",
       "      <td>$40k+</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Turn-Key Responsive Website Design &amp; Development</td>\n",
       "      <td>625</td>\n",
       "      <td>131</td>\n",
       "      <td>$45.00</td>\n",
       "      <td>['WordPress', 'Woocommerce', 'Website Developm...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Raising money as a startup is hard work! The a...</td>\n",
       "      <td>$100k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Startup Specialist, Pitch Decks, Business Plan...</td>\n",
       "      <td>655</td>\n",
       "      <td>32</td>\n",
       "      <td>$150.00</td>\n",
       "      <td>['Startup Consulting', 'Business Modeling', 'B...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Hire me to make sure that your business runs s...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Tech Savvy Assistant for Busy Entrepreneur</td>\n",
       "      <td>434</td>\n",
       "      <td>34</td>\n",
       "      <td>$27.00</td>\n",
       "      <td>['HTML', 'Customer Service', 'English', 'WordP...</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>I am a medical student at Northwestern Univers...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Medical Writer and Researcher</td>\n",
       "      <td>347</td>\n",
       "      <td>13</td>\n",
       "      <td>$50.00</td>\n",
       "      <td>['Customer Support', 'Medical Writing', 'Micro...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>My expertise in writing, editing, and designin...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Writer, Editor, and Instructional Designer</td>\n",
       "      <td>587</td>\n",
       "      <td>118</td>\n",
       "      <td>$37.50</td>\n",
       "      <td>['Academic Writing', 'Copyediting', 'Editing',...</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>👋 Hi, I'm Brandon! Top-Rated Wix Designer &amp; De...</td>\n",
       "      <td>$40k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WIX PROFESSIONAL: Expert Wix Website Designer ...</td>\n",
       "      <td>241</td>\n",
       "      <td>162</td>\n",
       "      <td>$50.00</td>\n",
       "      <td>['WiX', 'Web Design', 'Website Development', '...</td>\n",
       "      <td>26.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>I have over 20 years experience in data wareho...</td>\n",
       "      <td>$100k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Data Scientist and Technical Writer</td>\n",
       "      <td>1,790</td>\n",
       "      <td>99</td>\n",
       "      <td>$100.00</td>\n",
       "      <td>['Data Science', 'Data Modeling', 'SQL', 'Tech...</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>This is a bold statement, but I stand behind m...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Search Engine Optimization Specialist | Data V...</td>\n",
       "      <td>532</td>\n",
       "      <td>38</td>\n",
       "      <td>$55.00</td>\n",
       "      <td>['Internet Marketing', 'Google Website Optimiz...</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Let me be your creative chameleon. I can write...</td>\n",
       "      <td>$50k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Creative Writer and Editor</td>\n",
       "      <td>825</td>\n",
       "      <td>91</td>\n",
       "      <td>$85.00</td>\n",
       "      <td>['Academic Writing', 'Article Writing', 'Copye...</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>I am a widely skilled researcher and writer dr...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Creative Writer and Historical Researcher</td>\n",
       "      <td>969</td>\n",
       "      <td>12</td>\n",
       "      <td>$20.00</td>\n",
       "      <td>['History', 'Research Papers', 'Fiction Writin...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>I was a part time Virtual Assistant as well as...</td>\n",
       "      <td>$500k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Project &amp; Office Manager</td>\n",
       "      <td>59,063</td>\n",
       "      <td>105</td>\n",
       "      <td>$20.00</td>\n",
       "      <td>['Virtual Assistant', 'Customer Support', 'Pro...</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Like to test your app in iPhone S7 to SX Max, ...</td>\n",
       "      <td>$5k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>iPhone Developer &amp; Tester (Swift) | Google Ske...</td>\n",
       "      <td>655</td>\n",
       "      <td>2</td>\n",
       "      <td>$25.00</td>\n",
       "      <td>['Swift', 'iOS Development', '3D Modeling', 'M...</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>My previous work experience includes: Mortgage...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Compliance and Administrative Specialist</td>\n",
       "      <td>1,729</td>\n",
       "      <td>7</td>\n",
       "      <td>$15.00</td>\n",
       "      <td>['Accounting', 'Bookkeeping', 'Customer Servic...</td>\n",
       "      <td>14.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Over Twenty years of management experience inc...</td>\n",
       "      <td>$200k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>Experienced Operations Manager for Global Supp...</td>\n",
       "      <td>8,209</td>\n",
       "      <td>12</td>\n",
       "      <td>$24.44</td>\n",
       "      <td>['Customer Service', 'Data Analytics', 'Google...</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>\"Top level writing! She communicates well and ...</td>\n",
       "      <td>$30k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>B2B Copywriter: Website and Marketing Content ...</td>\n",
       "      <td>268</td>\n",
       "      <td>34</td>\n",
       "      <td>$95.00</td>\n",
       "      <td>['Content Strategy', 'Copywriting', 'Content W...</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Overall management and financial responsibilit...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Project Management</td>\n",
       "      <td>977</td>\n",
       "      <td>1</td>\n",
       "      <td>$49.00</td>\n",
       "      <td>['Project Management']</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>I am a professional programmer working on Powe...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>PowerBuilder 2017 / SQL Server T-SQL / Oracle ...</td>\n",
       "      <td>540</td>\n",
       "      <td>2</td>\n",
       "      <td>$65.00</td>\n",
       "      <td>['Microsoft SQL Server Programming', 'SQL Serv...</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>I am a responsible and dynamic person with a s...</td>\n",
       "      <td>$30k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>Customer Service, Data Entry, Nursing, Adminis...</td>\n",
       "      <td>2,560</td>\n",
       "      <td>68</td>\n",
       "      <td>$14.67</td>\n",
       "      <td>['Data Entry', 'Customer Service', 'Internet R...</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Audio Engineering, Sound Design, Original Musi...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Audio Post Engineer, Sound Designer, Video Edi...</td>\n",
       "      <td>1,420</td>\n",
       "      <td>92</td>\n",
       "      <td>$31.25</td>\n",
       "      <td>['Audio Editing', 'Audio Mixing', 'Audio Post ...</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>I have over 5 years of experience in Manual &amp; ...</td>\n",
       "      <td>$4k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ISTQB Certified Quality Engineer - Manual &amp; Au...</td>\n",
       "      <td>251</td>\n",
       "      <td>2</td>\n",
       "      <td>$20.00</td>\n",
       "      <td>['Selenium WebDriver', 'Black Box Testing', 'A...</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>I am a detail oriented, well organized adminis...</td>\n",
       "      <td>$5k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Detail Oriented Administrative Assistant</td>\n",
       "      <td>342</td>\n",
       "      <td>5</td>\n",
       "      <td>$20.00</td>\n",
       "      <td>['Proofreading', 'Editing', 'Social Media Mark...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>I have been in the Consulting and Strategy fie...</td>\n",
       "      <td>$7k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Financial Analyst</td>\n",
       "      <td>235</td>\n",
       "      <td>2</td>\n",
       "      <td>$45.00</td>\n",
       "      <td>['Statistics', 'Mathematics', 'Microsoft Excel...</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>My freelancing career started with Textbroker....</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Data entry/Transcription specialist</td>\n",
       "      <td>1,379</td>\n",
       "      <td>101</td>\n",
       "      <td>$20.00</td>\n",
       "      <td>['Data Entry', 'Article Writing', 'Receptionis...</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>Kevin brings more than 15 years of marketing a...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Marketing Fixer &amp; Award-Winning Graphic Designer</td>\n",
       "      <td>244</td>\n",
       "      <td>6</td>\n",
       "      <td>$50.00</td>\n",
       "      <td>['Marketing Strategy', 'Digital Marketing', 'S...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>I have extensive experience in many areas acro...</td>\n",
       "      <td>$100k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Project Engineer</td>\n",
       "      <td>5,903</td>\n",
       "      <td>15</td>\n",
       "      <td>$50.00</td>\n",
       "      <td>['Unix System Administration', 'Linux System A...</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>I am a web developer with over 15 years of exp...</td>\n",
       "      <td>$40k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Web Developer</td>\n",
       "      <td>391</td>\n",
       "      <td>52</td>\n",
       "      <td>$80.00</td>\n",
       "      <td>['PHP', 'WordPress', 'Drupal', 'Ruby', 'Django...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>I have extensive experience working as a paral...</td>\n",
       "      <td>$7k+</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Paralegal - English and U.S. law, HR generalist</td>\n",
       "      <td>251</td>\n",
       "      <td>2</td>\n",
       "      <td>$25.00</td>\n",
       "      <td>['Employment Law', 'Contract Law', 'HRM', 'Adm...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>I have a BFA in Animation, and have created An...</td>\n",
       "      <td>$8k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>Motion Graphics Designer and 3D Artist/Animator</td>\n",
       "      <td>252</td>\n",
       "      <td>8</td>\n",
       "      <td>$37.50</td>\n",
       "      <td>['3D Animation', '2D Animation', '3D Modeling'...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>I have a B.A in Marketing, I am a native Engli...</td>\n",
       "      <td>$7k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Content writer, Blogger, and SEO Expert</td>\n",
       "      <td>215</td>\n",
       "      <td>7</td>\n",
       "      <td>$25.00</td>\n",
       "      <td>['WordPress', 'Sports Writing', 'News Writing ...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>Hello, My name is Isa. I have 4 years of Tax P...</td>\n",
       "      <td>$4k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>Tax Professional/Bookkeeper/ Admin Manager Wit...</td>\n",
       "      <td>355</td>\n",
       "      <td>5</td>\n",
       "      <td>$15.00</td>\n",
       "      <td>['Data Entry', 'Bookkeeping', 'Customer Servic...</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>Intelligent, creative and detail oriented indi...</td>\n",
       "      <td>$6k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>Graphic Artist and Marketing Professional</td>\n",
       "      <td>239</td>\n",
       "      <td>49</td>\n",
       "      <td>$21.00</td>\n",
       "      <td>['Graphic Design', 'Print Design', 'Brochure D...</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>Hi! I'm Rammy! \\n\\nI'm a creative consultant, ...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>Naming Specialist, Expert Writer, Brand Consul...</td>\n",
       "      <td>285</td>\n",
       "      <td>25</td>\n",
       "      <td>$150.00</td>\n",
       "      <td>['Copywriting', 'Brand Consulting', 'Communica...</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>13+ year of front-end web developer experience...</td>\n",
       "      <td>$40k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>Senior Front End Web Developer</td>\n",
       "      <td>741</td>\n",
       "      <td>1</td>\n",
       "      <td>$85.00</td>\n",
       "      <td>['CSS3', 'AJAX', 'WordPress', 'PSD to Wordpres...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>I'm an experienced technology veteran with ove...</td>\n",
       "      <td>$3k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Experienced Technical Customer Support Veteran</td>\n",
       "      <td>271</td>\n",
       "      <td>4</td>\n",
       "      <td>$25.00</td>\n",
       "      <td>['Mac OS App Development', 'Customer Service',...</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>After 10 years of teaching grammar and academi...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Expert Writer + Editor</td>\n",
       "      <td>228</td>\n",
       "      <td>32</td>\n",
       "      <td>$30.00</td>\n",
       "      <td>['Editing', 'SEO Writing', 'Technical Writing'...</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>I develop competitive web and desktop applics ...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>SEM Manager, Developer</td>\n",
       "      <td>596</td>\n",
       "      <td>83</td>\n",
       "      <td>$20.00</td>\n",
       "      <td>['VBA', 'VB.NET', 'PHP', 'JavaScript', 'Adobe ...</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>Welcome to my profile! My name is Cassie. I wo...</td>\n",
       "      <td>$8k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Customer Service, Professional Data Entry and ...</td>\n",
       "      <td>891</td>\n",
       "      <td>10</td>\n",
       "      <td>$9.00</td>\n",
       "      <td>['English Grammar', 'English Spelling', 'Techn...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>Highly proficient project coordinator with exc...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Project Coordination and Recruiting Expert</td>\n",
       "      <td>1,088</td>\n",
       "      <td>3</td>\n",
       "      <td>$35.00</td>\n",
       "      <td>['Sales Management', 'Internet Marketing', 'Ma...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>I am an experienced virtual assistant, data en...</td>\n",
       "      <td>$100k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Customer Service Expert/Team Manager/Data Entr...</td>\n",
       "      <td>14,862</td>\n",
       "      <td>53</td>\n",
       "      <td>$13.75</td>\n",
       "      <td>['Customer Service', 'Live Chat Operator', 'Da...</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>My skills include designing , improving, or an...</td>\n",
       "      <td>$40k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Data Analyst and Spreadsheet Designer</td>\n",
       "      <td>2,373</td>\n",
       "      <td>234</td>\n",
       "      <td>$21.50</td>\n",
       "      <td>['Microsoft Excel', 'Microsoft PowerPoint']</td>\n",
       "      <td>26.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>I am an experienced and professional WordPress...</td>\n",
       "      <td>$8k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>WordPress &amp; Web Developer - Custom Websites, B...</td>\n",
       "      <td>247</td>\n",
       "      <td>29</td>\n",
       "      <td>$42.50</td>\n",
       "      <td>['WordPress', 'HTML5', 'CSS3', 'JavaScript', '...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>I am a Medical Doctor (MD) who graduated from ...</td>\n",
       "      <td>$100k+</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Physician Medical Writer and Consultant</td>\n",
       "      <td>222</td>\n",
       "      <td>76</td>\n",
       "      <td>$125.00</td>\n",
       "      <td>['Medical Writing', 'Content Writing', 'Scient...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>I have worked as a virtual assistant before. I...</td>\n",
       "      <td>$20k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Virtual Assistant</td>\n",
       "      <td>1,502</td>\n",
       "      <td>29</td>\n",
       "      <td>$18.00</td>\n",
       "      <td>['Sales', 'Microsoft Word', 'Microsoft Excel',...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>I have worked on presentations for Microsoft, ...</td>\n",
       "      <td>$30k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>PowerPoint &amp; Keynote Presentation Specialist</td>\n",
       "      <td>860</td>\n",
       "      <td>48</td>\n",
       "      <td>$58.00</td>\n",
       "      <td>['Adobe Creative Suite', 'Keynote', 'Microsoft...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>I am a writer and editor, although I would lov...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Writer and Editor</td>\n",
       "      <td>270</td>\n",
       "      <td>31</td>\n",
       "      <td>$28.00</td>\n",
       "      <td>['Editing', 'Copyediting', 'Copywriting', 'AP ...</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>Project Management &amp; Customer Support Speciali...</td>\n",
       "      <td>$100k+</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>Customer Service Retention Specialist and eCom...</td>\n",
       "      <td>5,285</td>\n",
       "      <td>10</td>\n",
       "      <td>$32.00</td>\n",
       "      <td>['Customer Service', 'Email Handling', 'Order ...</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  bio  earned  edu  exp  \\\n",
       "0   Google Adwords & Bing Ads Certified MBA market...  $200k+  3.0  8.0   \n",
       "1   I have been transcribing for more than 7 years...   $10k+  2.0  1.0   \n",
       "2   Throughout my undergraduate years, I was a res...   $60k+  1.0  3.0   \n",
       "3   As a digital marketing professional with a foc...   $30k+  2.0  2.0   \n",
       "4   WordPress Development + Theme Customization + ...   $40k+  NaN  1.0   \n",
       "5   Raising money as a startup is hard work! The a...  $100k+  1.0  5.0   \n",
       "6   Hire me to make sure that your business runs s...   $10k+  2.0  4.0   \n",
       "7   I am a medical student at Northwestern Univers...   $10k+  2.0  2.0   \n",
       "8   My expertise in writing, editing, and designin...   $20k+  3.0  3.0   \n",
       "9   👋 Hi, I'm Brandon! Top-Rated Wix Designer & De...   $40k+  1.0  1.0   \n",
       "10  I have over 20 years experience in data wareho...  $100k+  2.0  1.0   \n",
       "11  This is a bold statement, but I stand behind m...   $20k+  1.0  1.0   \n",
       "12  Let me be your creative chameleon. I can write...   $50k+  2.0  5.0   \n",
       "13  I am a widely skilled researcher and writer dr...   $10k+  2.0  4.0   \n",
       "14  I was a part time Virtual Assistant as well as...  $500k+  1.0  3.0   \n",
       "15  Like to test your app in iPhone S7 to SX Max, ...    $5k+  1.0  NaN   \n",
       "16  My previous work experience includes: Mortgage...   $20k+  1.0  4.0   \n",
       "17  Over Twenty years of management experience inc...  $200k+  1.0  8.0   \n",
       "18  \"Top level writing! She communicates well and ...   $30k+  2.0  2.0   \n",
       "19  Overall management and financial responsibilit...   $10k+  1.0  1.0   \n",
       "20  I am a professional programmer working on Powe...   $20k+  1.0  1.0   \n",
       "21  I am a responsible and dynamic person with a s...   $30k+  1.0  7.0   \n",
       "22  Audio Engineering, Sound Design, Original Musi...   $20k+  1.0  3.0   \n",
       "23  I have over 5 years of experience in Manual & ...    $4k+  1.0  NaN   \n",
       "24  I am a detail oriented, well organized adminis...    $5k+  1.0  2.0   \n",
       "25  I have been in the Consulting and Strategy fie...    $7k+  1.0  2.0   \n",
       "26  My freelancing career started with Textbroker....   $20k+  2.0  4.0   \n",
       "27  Kevin brings more than 15 years of marketing a...   $10k+  2.0  5.0   \n",
       "28  I have extensive experience in many areas acro...  $100k+  1.0  4.0   \n",
       "29  I am a web developer with over 15 years of exp...   $40k+  2.0  2.0   \n",
       "30  I have extensive experience working as a paral...    $7k+  4.0  4.0   \n",
       "31  I have a BFA in Animation, and have created An...    $8k+  1.0  6.0   \n",
       "32  I have a B.A in Marketing, I am a native Engli...    $7k+  1.0  4.0   \n",
       "33  Hello, My name is Isa. I have 4 years of Tax P...    $4k+  2.0  8.0   \n",
       "34  Intelligent, creative and detail oriented indi...    $6k+  2.0  7.0   \n",
       "35  Hi! I'm Rammy! \\n\\nI'm a creative consultant, ...   $20k+  2.0  8.0   \n",
       "36  13+ year of front-end web developer experience...   $40k+  1.0  7.0   \n",
       "37  I'm an experienced technology veteran with ove...    $3k+  1.0  3.0   \n",
       "38  After 10 years of teaching grammar and academi...   $10k+  2.0  3.0   \n",
       "39  I develop competitive web and desktop applics ...   $20k+  1.0  1.0   \n",
       "40  Welcome to my profile! My name is Cassie. I wo...    $8k+  1.0  4.0   \n",
       "41  Highly proficient project coordinator with exc...   $20k+  2.0  3.0   \n",
       "42  I am an experienced virtual assistant, data en...  $100k+  1.0  3.0   \n",
       "43  My skills include designing , improving, or an...   $40k+  1.0  1.0   \n",
       "44  I am an experienced and professional WordPress...    $8k+  2.0  2.0   \n",
       "45  I am a Medical Doctor (MD) who graduated from ...  $100k+  2.0  3.0   \n",
       "46  I have worked as a virtual assistant before. I...   $20k+  1.0  4.0   \n",
       "47  I have worked on presentations for Microsoft, ...   $30k+  1.0  3.0   \n",
       "48  I am a writer and editor, although I would lov...   $10k+  1.0  3.0   \n",
       "49  Project Management & Customer Support Speciali...  $100k+  3.0  6.0   \n",
       "\n",
       "                                             headline   hours jobs     rate  \\\n",
       "0   Marketing, Strategic Mgmt/Writer/SEO, SEM/Mark...  10,654   40   $45.00   \n",
       "1                                    Transcriptionist   1,043   60   $12.00   \n",
       "2            Transcription (98WPM), Virtual Assistant   2,840  254   $15.00   \n",
       "3   Full-Stack Digital Marketer | Facebook Ads Spe...     673   21   $75.00   \n",
       "4    Turn-Key Responsive Website Design & Development     625  131   $45.00   \n",
       "5   Startup Specialist, Pitch Decks, Business Plan...     655   32  $150.00   \n",
       "6          Tech Savvy Assistant for Busy Entrepreneur     434   34   $27.00   \n",
       "7                       Medical Writer and Researcher     347   13   $50.00   \n",
       "8          Writer, Editor, and Instructional Designer     587  118   $37.50   \n",
       "9   WIX PROFESSIONAL: Expert Wix Website Designer ...     241  162   $50.00   \n",
       "10                Data Scientist and Technical Writer   1,790   99  $100.00   \n",
       "11  Search Engine Optimization Specialist | Data V...     532   38   $55.00   \n",
       "12                         Creative Writer and Editor     825   91   $85.00   \n",
       "13          Creative Writer and Historical Researcher     969   12   $20.00   \n",
       "14                           Project & Office Manager  59,063  105   $20.00   \n",
       "15  iPhone Developer & Tester (Swift) | Google Ske...     655    2   $25.00   \n",
       "16           Compliance and Administrative Specialist   1,729    7   $15.00   \n",
       "17  Experienced Operations Manager for Global Supp...   8,209   12   $24.44   \n",
       "18  B2B Copywriter: Website and Marketing Content ...     268   34   $95.00   \n",
       "19                                 Project Management     977    1   $49.00   \n",
       "20  PowerBuilder 2017 / SQL Server T-SQL / Oracle ...     540    2   $65.00   \n",
       "21  Customer Service, Data Entry, Nursing, Adminis...   2,560   68   $14.67   \n",
       "22  Audio Post Engineer, Sound Designer, Video Edi...   1,420   92   $31.25   \n",
       "23  ISTQB Certified Quality Engineer - Manual & Au...     251    2   $20.00   \n",
       "24           Detail Oriented Administrative Assistant     342    5   $20.00   \n",
       "25                                  Financial Analyst     235    2   $45.00   \n",
       "26                Data entry/Transcription specialist   1,379  101   $20.00   \n",
       "27   Marketing Fixer & Award-Winning Graphic Designer     244    6   $50.00   \n",
       "28                                   Project Engineer   5,903   15   $50.00   \n",
       "29                                      Web Developer     391   52   $80.00   \n",
       "30    Paralegal - English and U.S. law, HR generalist     251    2   $25.00   \n",
       "31    Motion Graphics Designer and 3D Artist/Animator     252    8   $37.50   \n",
       "32            Content writer, Blogger, and SEO Expert     215    7   $25.00   \n",
       "33  Tax Professional/Bookkeeper/ Admin Manager Wit...     355    5   $15.00   \n",
       "34          Graphic Artist and Marketing Professional     239   49   $21.00   \n",
       "35  Naming Specialist, Expert Writer, Brand Consul...     285   25  $150.00   \n",
       "36                     Senior Front End Web Developer     741    1   $85.00   \n",
       "37     Experienced Technical Customer Support Veteran     271    4   $25.00   \n",
       "38                             Expert Writer + Editor     228   32   $30.00   \n",
       "39                             SEM Manager, Developer     596   83   $20.00   \n",
       "40  Customer Service, Professional Data Entry and ...     891   10    $9.00   \n",
       "41         Project Coordination and Recruiting Expert   1,088    3   $35.00   \n",
       "42  Customer Service Expert/Team Manager/Data Entr...  14,862   53   $13.75   \n",
       "43              Data Analyst and Spreadsheet Designer   2,373  234   $21.50   \n",
       "44  WordPress & Web Developer - Custom Websites, B...     247   29   $42.50   \n",
       "45            Physician Medical Writer and Consultant     222   76  $125.00   \n",
       "46                                  Virtual Assistant   1,502   29   $18.00   \n",
       "47       PowerPoint & Keynote Presentation Specialist     860   48   $58.00   \n",
       "48                                  Writer and Editor     270   31   $28.00   \n",
       "49  Customer Service Retention Specialist and eCom...   5,285   10   $32.00   \n",
       "\n",
       "                                               skills  tests  \n",
       "0   ['Strategic Planning', 'Business Analysis', 'A...    9.0  \n",
       "1          ['Transcription', 'Medical Transcription']    2.0  \n",
       "2   ['Biology', 'English', 'Transcription', 'Virtu...    4.0  \n",
       "3   ['Facebook Marketing', 'Inbound Marketing', 'M...    4.0  \n",
       "4   ['WordPress', 'Woocommerce', 'Website Developm...    1.0  \n",
       "5   ['Startup Consulting', 'Business Modeling', 'B...    1.0  \n",
       "6   ['HTML', 'Customer Service', 'English', 'WordP...    4.0  \n",
       "7   ['Customer Support', 'Medical Writing', 'Micro...    1.0  \n",
       "8   ['Academic Writing', 'Copyediting', 'Editing',...    6.0  \n",
       "9   ['WiX', 'Web Design', 'Website Development', '...   26.0  \n",
       "10  ['Data Science', 'Data Modeling', 'SQL', 'Tech...    4.0  \n",
       "11  ['Internet Marketing', 'Google Website Optimiz...    8.0  \n",
       "12  ['Academic Writing', 'Article Writing', 'Copye...    4.0  \n",
       "13  ['History', 'Research Papers', 'Fiction Writin...    NaN  \n",
       "14  ['Virtual Assistant', 'Customer Support', 'Pro...    6.0  \n",
       "15  ['Swift', 'iOS Development', '3D Modeling', 'M...    2.0  \n",
       "16  ['Accounting', 'Bookkeeping', 'Customer Servic...   14.0  \n",
       "17  ['Customer Service', 'Data Analytics', 'Google...    9.0  \n",
       "18  ['Content Strategy', 'Copywriting', 'Content W...    2.0  \n",
       "19                             ['Project Management']    NaN  \n",
       "20  ['Microsoft SQL Server Programming', 'SQL Serv...    3.0  \n",
       "21  ['Data Entry', 'Customer Service', 'Internet R...    8.0  \n",
       "22  ['Audio Editing', 'Audio Mixing', 'Audio Post ...    2.0  \n",
       "23  ['Selenium WebDriver', 'Black Box Testing', 'A...    2.0  \n",
       "24  ['Proofreading', 'Editing', 'Social Media Mark...    NaN  \n",
       "25  ['Statistics', 'Mathematics', 'Microsoft Excel...    3.0  \n",
       "26  ['Data Entry', 'Article Writing', 'Receptionis...    3.0  \n",
       "27  ['Marketing Strategy', 'Digital Marketing', 'S...    NaN  \n",
       "28  ['Unix System Administration', 'Linux System A...    5.0  \n",
       "29  ['PHP', 'WordPress', 'Drupal', 'Ruby', 'Django...    NaN  \n",
       "30  ['Employment Law', 'Contract Law', 'HRM', 'Adm...    NaN  \n",
       "31  ['3D Animation', '2D Animation', '3D Modeling'...    NaN  \n",
       "32  ['WordPress', 'Sports Writing', 'News Writing ...    1.0  \n",
       "33  ['Data Entry', 'Bookkeeping', 'Customer Servic...    2.0  \n",
       "34  ['Graphic Design', 'Print Design', 'Brochure D...    2.0  \n",
       "35  ['Copywriting', 'Brand Consulting', 'Communica...    4.0  \n",
       "36  ['CSS3', 'AJAX', 'WordPress', 'PSD to Wordpres...    1.0  \n",
       "37  ['Mac OS App Development', 'Customer Service',...    5.0  \n",
       "38  ['Editing', 'SEO Writing', 'Technical Writing'...    9.0  \n",
       "39  ['VBA', 'VB.NET', 'PHP', 'JavaScript', 'Adobe ...    6.0  \n",
       "40  ['English Grammar', 'English Spelling', 'Techn...    1.0  \n",
       "41  ['Sales Management', 'Internet Marketing', 'Ma...    NaN  \n",
       "42  ['Customer Service', 'Live Chat Operator', 'Da...    3.0  \n",
       "43        ['Microsoft Excel', 'Microsoft PowerPoint']   26.0  \n",
       "44  ['WordPress', 'HTML5', 'CSS3', 'JavaScript', '...    1.0  \n",
       "45  ['Medical Writing', 'Content Writing', 'Scient...    1.0  \n",
       "46  ['Sales', 'Microsoft Word', 'Microsoft Excel',...    NaN  \n",
       "47  ['Adobe Creative Suite', 'Keynote', 'Microsoft...    1.0  \n",
       "48  ['Editing', 'Copyediting', 'Copywriting', 'AP ...    5.0  \n",
       "49  ['Customer Service', 'Email Handling', 'Order ...    3.0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df = df.drop_duplicates()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 12,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4262, 10)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 7,
        "hidden": false,
        "row": 16,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bio          object\n",
       "earned       object\n",
       "edu         float64\n",
       "exp         float64\n",
       "headline     object\n",
       "hours        object\n",
       "jobs         object\n",
       "rate         object\n",
       "skills       object\n",
       "tests       float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 7,
        "hidden": false,
        "row": 20,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bio           17\n",
       "earned      1525\n",
       "edu          256\n",
       "exp          796\n",
       "headline      24\n",
       "hours       1523\n",
       "jobs        1494\n",
       "rate           2\n",
       "skills        70\n",
       "tests       2017\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df = df[df.rate.notna()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 20,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4260, 10)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['skills'] = df['skills'].fillna('[]')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['headline'] = df['headline'].fillna('')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['bio'] = df['bio'].fillna('')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "# sometimes hours and jobs will be zero but earned will be a positive number.\n",
    "# the below will allow one to check the earned column to see whether a freelancer has worked or not\n",
    "df = df.fillna('0')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['rate'] = df['rate'].map(lambda string: float(''.join([c for c in string if c.isdigit()]))/100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['hours'] = df['hours'].map(lambda h: int(''.join([c for c in str(h) if c.isdigit()])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['jobs'] = df['jobs'].map(lambda j: int(''.join([c for c in str(j) if c.isdigit()])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['exp'] = df['exp'].map(lambda e: int(e))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['edu'] = df['edu'].map(lambda e: int(e))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['tests'] = df['tests'].map(lambda t: int(t))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['skills'] = df['skills'].map(lambda sk_string: sk_string.replace('[', \"\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['skills'] = df['skills'].map(lambda sk_string: sk_string.replace(']', \"\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['skills'] = df['skills'].map(lambda sk_string: sk_string.replace(\"'\", \"\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['skills'] = df['skills'].map(lambda sk_string: sk_string.replace('\"', ''))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df['skills'] = df['skills'].map(lambda sk_string: sk_string.split(\", \"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "# ternary operator\n",
    "df['skills'] = df['skills'].map(lambda sk_list: (sk_list, [])[sk_list == ['']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 7,
        "hidden": false,
        "row": 23,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bio          object\n",
       "earned       object\n",
       "edu           int64\n",
       "exp           int64\n",
       "headline     object\n",
       "hours         int64\n",
       "jobs          int64\n",
       "rate        float64\n",
       "skills       object\n",
       "tests         int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 24,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4260"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 27,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4260"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nn = 0\n",
    "for i in range(len(df)):\n",
    "    nn += (type(df['skills'].iloc[55]) == list)\n",
    "nn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "# df = df[df.rate < 200]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df = df[df['earned'] != '0']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 28,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2735, 10)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "df = df.reset_index(drop = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "y = df.rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 30,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "38.606"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std = np.std(y)\n",
    "round(std, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 31,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "47.207"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "meany = np.mean(y)\n",
    "round(meany, 3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 32,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "source": [
    "### Hourly rate by skill:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "skills_df = df['skills'].str.join('|').str.get_dummies()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 34,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2735, 1888)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "skills_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 7,
        "hidden": false,
        "row": 35,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8\n",
      "2\n",
      "10\n",
      "5\n",
      "6\n",
      "3\n",
      "10\n",
      "7\n",
      "10\n",
      "10\n"
     ]
    }
   ],
   "source": [
    "for i in range(10):\n",
    "    print(sum(skills_df.iloc[i, :]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 7,
        "hidden": false,
        "row": 36,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8\n",
      "2\n",
      "10\n",
      "5\n",
      "6\n",
      "3\n",
      "10\n",
      "7\n",
      "10\n",
      "10\n"
     ]
    }
   ],
   "source": [
    "for i in range(10):\n",
    "    print(len(df.iloc[i, 8]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "rate = copy.deepcopy(df.rate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 5,
        "hidden": false,
        "row": 38,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    45.0\n",
       "1    12.0\n",
       "2    15.0\n",
       "3    75.0\n",
       "4    45.0\n",
       "Name: rate, dtype: float64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rate.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 15,
        "hidden": false,
        "row": 43,
        "width": 7
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>bio</th>\n",
       "      <th>earned</th>\n",
       "      <th>edu</th>\n",
       "      <th>exp</th>\n",
       "      <th>headline</th>\n",
       "      <th>hours</th>\n",
       "      <th>jobs</th>\n",
       "      <th>rate</th>\n",
       "      <th>skills</th>\n",
       "      <th>tests</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Google Adwords &amp; Bing Ads Certified MBA market...</td>\n",
       "      <td>$200k+</td>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>Marketing, Strategic Mgmt/Writer/SEO, SEM/Mark...</td>\n",
       "      <td>10654</td>\n",
       "      <td>40</td>\n",
       "      <td>45.0</td>\n",
       "      <td>[Strategic Planning, Business Analysis, Advert...</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>I have been transcribing for more than 7 years...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Transcriptionist</td>\n",
       "      <td>1043</td>\n",
       "      <td>60</td>\n",
       "      <td>12.0</td>\n",
       "      <td>[Transcription, Medical Transcription]</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Throughout my undergraduate years, I was a res...</td>\n",
       "      <td>$60k+</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Transcription (98WPM), Virtual Assistant</td>\n",
       "      <td>2840</td>\n",
       "      <td>254</td>\n",
       "      <td>15.0</td>\n",
       "      <td>[Biology, English, Transcription, Virtual Assi...</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>As a digital marketing professional with a foc...</td>\n",
       "      <td>$30k+</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>Full-Stack Digital Marketer | Facebook Ads Spe...</td>\n",
       "      <td>673</td>\n",
       "      <td>21</td>\n",
       "      <td>75.0</td>\n",
       "      <td>[Facebook Marketing, Inbound Marketing, Market...</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>WordPress Development + Theme Customization + ...</td>\n",
       "      <td>$40k+</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Turn-Key Responsive Website Design &amp; Development</td>\n",
       "      <td>625</td>\n",
       "      <td>131</td>\n",
       "      <td>45.0</td>\n",
       "      <td>[WordPress, Woocommerce, Website Development, ...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 bio  earned  edu  exp  \\\n",
       "0  Google Adwords & Bing Ads Certified MBA market...  $200k+    3    8   \n",
       "1  I have been transcribing for more than 7 years...   $10k+    2    1   \n",
       "2  Throughout my undergraduate years, I was a res...   $60k+    1    3   \n",
       "3  As a digital marketing professional with a foc...   $30k+    2    2   \n",
       "4  WordPress Development + Theme Customization + ...   $40k+    0    1   \n",
       "\n",
       "                                            headline  hours  jobs  rate  \\\n",
       "0  Marketing, Strategic Mgmt/Writer/SEO, SEM/Mark...  10654    40  45.0   \n",
       "1                                   Transcriptionist   1043    60  12.0   \n",
       "2           Transcription (98WPM), Virtual Assistant   2840   254  15.0   \n",
       "3  Full-Stack Digital Marketer | Facebook Ads Spe...    673    21  75.0   \n",
       "4   Turn-Key Responsive Website Design & Development    625   131  45.0   \n",
       "\n",
       "                                              skills  tests  \n",
       "0  [Strategic Planning, Business Analysis, Advert...      9  \n",
       "1             [Transcription, Medical Transcription]      2  \n",
       "2  [Biology, English, Transcription, Virtual Assi...      4  \n",
       "3  [Facebook Marketing, Inbound Marketing, Market...      4  \n",
       "4  [WordPress, Woocommerce, Website Development, ...      1  "
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "combined = pd.concat([skills_df, rate], axis = 1, sort = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 9,
        "hidden": false,
        "row": 58,
        "width": 12
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>.NET Core</th>\n",
       "      <th>.NET Framework</th>\n",
       "      <th>.NET Remoting</th>\n",
       "      <th>2D Animation</th>\n",
       "      <th>2D Design</th>\n",
       "      <th>3D Animation</th>\n",
       "      <th>3D Design</th>\n",
       "      <th>3D Modeling</th>\n",
       "      <th>3D Printing</th>\n",
       "      <th>3D Rendering</th>\n",
       "      <th>...</th>\n",
       "      <th>iPhone App Development</th>\n",
       "      <th>iPhone UI Design</th>\n",
       "      <th>insightly</th>\n",
       "      <th>jQuery</th>\n",
       "      <th>jQuery UI</th>\n",
       "      <th>libGDX</th>\n",
       "      <th>logstash</th>\n",
       "      <th>phpMyAdmin</th>\n",
       "      <th>soapUI</th>\n",
       "      <th>rate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>45.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>15.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>75.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>45.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 1889 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   .NET Core  .NET Framework  .NET Remoting  2D Animation  2D Design  \\\n",
       "0          0               0              0             0          0   \n",
       "1          0               0              0             0          0   \n",
       "2          0               0              0             0          0   \n",
       "3          0               0              0             0          0   \n",
       "4          0               0              0             0          0   \n",
       "\n",
       "   3D Animation  3D Design  3D Modeling  3D Printing  3D Rendering  ...   \\\n",
       "0             0          0            0            0             0  ...    \n",
       "1             0          0            0            0             0  ...    \n",
       "2             0          0            0            0             0  ...    \n",
       "3             0          0            0            0             0  ...    \n",
       "4             0          0            0            0             0  ...    \n",
       "\n",
       "   iPhone App Development  iPhone UI Design  insightly  jQuery  jQuery UI  \\\n",
       "0                       0                 0          0       0          0   \n",
       "1                       0                 0          0       0          0   \n",
       "2                       0                 0          0       0          0   \n",
       "3                       0                 0          0       0          0   \n",
       "4                       0                 0          0       0          0   \n",
       "\n",
       "   libGDX  logstash  phpMyAdmin  soapUI  rate  \n",
       "0       0         0           0       0  45.0  \n",
       "1       0         0           0       0  12.0  \n",
       "2       0         0           0       0  15.0  \n",
       "3       0         0           0       0  75.0  \n",
       "4       0         0           0       0  45.0  \n",
       "\n",
       "[5 rows x 1889 columns]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "combined.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 7,
        "height": 7,
        "hidden": false,
        "row": 43,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8.0\n",
      "2.0\n",
      "10.0\n",
      "5.0\n",
      "6.0\n",
      "3.0\n",
      "10.0\n",
      "7.0\n",
      "10.0\n",
      "10.0\n"
     ]
    }
   ],
   "source": [
    "for i in range(10):\n",
    "    print(sum(combined.iloc[i, :-1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 7,
        "height": 7,
        "hidden": false,
        "row": 50,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "45.0\n",
      "12.0\n",
      "15.0\n",
      "75.0\n",
      "45.0\n",
      "150.0\n",
      "27.0\n",
      "50.0\n",
      "37.5\n",
      "50.0\n"
     ]
    }
   ],
   "source": [
    "for i in range(10):\n",
    "    print(combined.iloc[i, -1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "sd_list = []\n",
    "for c in combined.columns[:-1]: \n",
    "    skill_dict = dict.fromkeys(['mean', 'std', 'skill'], [])\n",
    "    comb_g = combined.groupby(c)\n",
    "    group_rate = comb_g.get_group(1)['rate']\n",
    "    skill_dict['g_r'] = group_rate\n",
    "    skill_dict['mean'] = np.mean(group_rate)\n",
    "    skill_dict['std'] = np.std(group_rate)\n",
    "    skill_dict['skill'] = str(c)\n",
    "    skill_dict['count'] = len(group_rate)\n",
    "    sd_list.append(skill_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "skill_table = pd.DataFrame(sd_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 23,
        "hidden": false,
        "row": 67,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>g_r</th>\n",
       "      <th>mean</th>\n",
       "      <th>skill</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2711    30.0\n",
       "Name: rate, dtype: float64</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>.NET Core</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29</td>\n",
       "      <td>208     200.00\n",
       "264     150.00\n",
       "325      65.00\n",
       "3...</td>\n",
       "      <td>75.402069</td>\n",
       "      <td>.NET Framework</td>\n",
       "      <td>47.065003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>502    50.0\n",
       "Name: rate, dtype: float64</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>.NET Remoting</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>14</td>\n",
       "      <td>31      37.5\n",
       "63      27.0\n",
       "656     45.0\n",
       "751    ...</td>\n",
       "      <td>46.535714</td>\n",
       "      <td>2D Animation</td>\n",
       "      <td>11.760200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17</td>\n",
       "      <td>492     38.0\n",
       "568     55.0\n",
       "633     12.0\n",
       "650    ...</td>\n",
       "      <td>34.029412</td>\n",
       "      <td>2D Design</td>\n",
       "      <td>15.635998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>8</td>\n",
       "      <td>31      37.5\n",
       "173     32.0\n",
       "967     25.0\n",
       "1118   ...</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>3D Animation</td>\n",
       "      <td>10.773231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>17</td>\n",
       "      <td>427     30.00\n",
       "547     25.00\n",
       "568     55.00\n",
       "769 ...</td>\n",
       "      <td>38.514706</td>\n",
       "      <td>3D Design</td>\n",
       "      <td>18.010306</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>28</td>\n",
       "      <td>15      25.0\n",
       "31      37.5\n",
       "173     32.0\n",
       "249    ...</td>\n",
       "      <td>44.785714</td>\n",
       "      <td>3D Modeling</td>\n",
       "      <td>17.342366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>14</td>\n",
       "      <td>142     70.00\n",
       "215     13.00\n",
       "502     50.00\n",
       "644 ...</td>\n",
       "      <td>46.375000</td>\n",
       "      <td>3D Printing</td>\n",
       "      <td>21.383561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>20</td>\n",
       "      <td>97      100.0\n",
       "121      20.0\n",
       "173      32.0\n",
       "568 ...</td>\n",
       "      <td>48.125000</td>\n",
       "      <td>3D Rendering</td>\n",
       "      <td>23.279752</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   count                                                g_r       mean  \\\n",
       "0      1            2711    30.0\n",
       "Name: rate, dtype: float64  30.000000   \n",
       "1     29  208     200.00\n",
       "264     150.00\n",
       "325      65.00\n",
       "3...  75.402069   \n",
       "2      1             502    50.0\n",
       "Name: rate, dtype: float64  50.000000   \n",
       "3     14  31      37.5\n",
       "63      27.0\n",
       "656     45.0\n",
       "751    ...  46.535714   \n",
       "4     17  492     38.0\n",
       "568     55.0\n",
       "633     12.0\n",
       "650    ...  34.029412   \n",
       "5      8  31      37.5\n",
       "173     32.0\n",
       "967     25.0\n",
       "1118   ...  38.000000   \n",
       "6     17  427     30.00\n",
       "547     25.00\n",
       "568     55.00\n",
       "769 ...  38.514706   \n",
       "7     28  15      25.0\n",
       "31      37.5\n",
       "173     32.0\n",
       "249    ...  44.785714   \n",
       "8     14  142     70.00\n",
       "215     13.00\n",
       "502     50.00\n",
       "644 ...  46.375000   \n",
       "9     20  97      100.0\n",
       "121      20.0\n",
       "173      32.0\n",
       "568 ...  48.125000   \n",
       "\n",
       "            skill        std  \n",
       "0       .NET Core   0.000000  \n",
       "1  .NET Framework  47.065003  \n",
       "2   .NET Remoting   0.000000  \n",
       "3    2D Animation  11.760200  \n",
       "4       2D Design  15.635998  \n",
       "5    3D Animation  10.773231  \n",
       "6       3D Design  18.010306  \n",
       "7     3D Modeling  17.342366  \n",
       "8     3D Printing  21.383561  \n",
       "9    3D Rendering  23.279752  "
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "skill_table.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "skill_table = skill_table.sort_values(by = ['mean'], ascending = False)\n",
    "skill_table = skill_table.reset_index(drop = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "skill_tab = skill_table[skill_table['count'] > 10]\n",
    "skill_tab = skill_tab.reset_index(drop = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "ds = skill_table[skill_table['skill'] == 'Data Science']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 5,
        "hidden": false,
        "row": 67,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>g_r</th>\n",
       "      <th>mean</th>\n",
       "      <th>skill</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>332</th>\n",
       "      <td>21</td>\n",
       "      <td>10      100.00\n",
       "54      150.00\n",
       "197      75.00\n",
       "2...</td>\n",
       "      <td>74.428095</td>\n",
       "      <td>Data Science</td>\n",
       "      <td>35.115206</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     count                                                g_r       mean  \\\n",
       "332     21  10      100.00\n",
       "54      150.00\n",
       "197      75.00\n",
       "2...  74.428095   \n",
       "\n",
       "            skill        std  \n",
       "332  Data Science  35.115206  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 41,
        "hidden": false,
        "row": 67,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>skill</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>26</td>\n",
       "      <td>105.346154</td>\n",
       "      <td>Startup Consulting</td>\n",
       "      <td>55.338776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>19</td>\n",
       "      <td>90.877895</td>\n",
       "      <td>Ruby</td>\n",
       "      <td>59.125761</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21</td>\n",
       "      <td>90.500000</td>\n",
       "      <td>Management Consulting</td>\n",
       "      <td>53.494548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11</td>\n",
       "      <td>87.272727</td>\n",
       "      <td>Financial Writing</td>\n",
       "      <td>26.744892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>86.750000</td>\n",
       "      <td>Business Coaching</td>\n",
       "      <td>55.793985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>16</td>\n",
       "      <td>86.031250</td>\n",
       "      <td>DevOps</td>\n",
       "      <td>30.397995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>12</td>\n",
       "      <td>83.000000</td>\n",
       "      <td>Organizational Behavior</td>\n",
       "      <td>95.311769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>11</td>\n",
       "      <td>82.159091</td>\n",
       "      <td>Docker</td>\n",
       "      <td>28.900746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>20</td>\n",
       "      <td>81.950000</td>\n",
       "      <td>Travel Writing</td>\n",
       "      <td>210.969468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>11</td>\n",
       "      <td>81.409091</td>\n",
       "      <td>Rapid Prototyping</td>\n",
       "      <td>48.681160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>24</td>\n",
       "      <td>81.176667</td>\n",
       "      <td>Machine Learning</td>\n",
       "      <td>35.392992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>18</td>\n",
       "      <td>80.666667</td>\n",
       "      <td>R</td>\n",
       "      <td>62.420616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>29</td>\n",
       "      <td>80.070690</td>\n",
       "      <td>Business Planning</td>\n",
       "      <td>46.749927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>22</td>\n",
       "      <td>79.643636</td>\n",
       "      <td>Product Management</td>\n",
       "      <td>34.900764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>32</td>\n",
       "      <td>78.522500</td>\n",
       "      <td>Ruby on Rails</td>\n",
       "      <td>48.707970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>30</td>\n",
       "      <td>78.316333</td>\n",
       "      <td>Digital Marketing</td>\n",
       "      <td>53.705465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>14</td>\n",
       "      <td>78.102143</td>\n",
       "      <td>Conversion Rate Optimization</td>\n",
       "      <td>17.093104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>40</td>\n",
       "      <td>77.516750</td>\n",
       "      <td>Amazon Web Services</td>\n",
       "      <td>45.544594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>24</td>\n",
       "      <td>77.197917</td>\n",
       "      <td>Content Strategy</td>\n",
       "      <td>43.388925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>17</td>\n",
       "      <td>75.588235</td>\n",
       "      <td>VB.NET</td>\n",
       "      <td>60.021622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>29</td>\n",
       "      <td>75.402069</td>\n",
       "      <td>.NET Framework</td>\n",
       "      <td>47.065003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>21</td>\n",
       "      <td>74.428095</td>\n",
       "      <td>Data Science</td>\n",
       "      <td>35.115206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>11</td>\n",
       "      <td>74.272727</td>\n",
       "      <td>Program Management</td>\n",
       "      <td>60.320494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>11</td>\n",
       "      <td>73.909091</td>\n",
       "      <td>FileMaker</td>\n",
       "      <td>34.589685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>31</td>\n",
       "      <td>73.515806</td>\n",
       "      <td>Marketing Automation</td>\n",
       "      <td>40.198434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>23</td>\n",
       "      <td>73.500000</td>\n",
       "      <td>Brand Consulting</td>\n",
       "      <td>38.787885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>25</td>\n",
       "      <td>73.400000</td>\n",
       "      <td>Django</td>\n",
       "      <td>45.327254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>43</td>\n",
       "      <td>73.323023</td>\n",
       "      <td>Business Analysis</td>\n",
       "      <td>42.767763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>13</td>\n",
       "      <td>72.730769</td>\n",
       "      <td>PCB Design</td>\n",
       "      <td>29.522883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>20</td>\n",
       "      <td>72.650000</td>\n",
       "      <td>API Development</td>\n",
       "      <td>34.991820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>25</td>\n",
       "      <td>72.318800</td>\n",
       "      <td>Inbound Marketing</td>\n",
       "      <td>44.167893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>17</td>\n",
       "      <td>72.058824</td>\n",
       "      <td>Product Design</td>\n",
       "      <td>46.504548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>44</td>\n",
       "      <td>71.635000</td>\n",
       "      <td>Laravel Framework</td>\n",
       "      <td>34.267312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>12</td>\n",
       "      <td>71.416667</td>\n",
       "      <td>Cloud Computing</td>\n",
       "      <td>45.345816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>34</td>\n",
       "      <td>71.351765</td>\n",
       "      <td>Objective-C</td>\n",
       "      <td>28.273441</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>18</td>\n",
       "      <td>71.185556</td>\n",
       "      <td>Business Proposal Writing</td>\n",
       "      <td>27.271052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>37</td>\n",
       "      <td>70.972973</td>\n",
       "      <td>User Interface Design</td>\n",
       "      <td>52.110971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>36</td>\n",
       "      <td>70.611111</td>\n",
       "      <td>Mobile App Development</td>\n",
       "      <td>42.439681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>13</td>\n",
       "      <td>70.538462</td>\n",
       "      <td>Ionic Framework</td>\n",
       "      <td>36.659274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>26</td>\n",
       "      <td>70.480769</td>\n",
       "      <td>Strategic Planning</td>\n",
       "      <td>31.180764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>11</td>\n",
       "      <td>70.386364</td>\n",
       "      <td>Digital Electronics</td>\n",
       "      <td>33.705166</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    count        mean                         skill         std\n",
       "0      26  105.346154            Startup Consulting   55.338776\n",
       "1      19   90.877895                          Ruby   59.125761\n",
       "2      21   90.500000         Management Consulting   53.494548\n",
       "3      11   87.272727             Financial Writing   26.744892\n",
       "4      16   86.750000             Business Coaching   55.793985\n",
       "5      16   86.031250                        DevOps   30.397995\n",
       "6      12   83.000000       Organizational Behavior   95.311769\n",
       "7      11   82.159091                        Docker   28.900746\n",
       "8      20   81.950000                Travel Writing  210.969468\n",
       "9      11   81.409091             Rapid Prototyping   48.681160\n",
       "10     24   81.176667              Machine Learning   35.392992\n",
       "11     18   80.666667                             R   62.420616\n",
       "12     29   80.070690             Business Planning   46.749927\n",
       "13     22   79.643636            Product Management   34.900764\n",
       "14     32   78.522500                 Ruby on Rails   48.707970\n",
       "15     30   78.316333             Digital Marketing   53.705465\n",
       "16     14   78.102143  Conversion Rate Optimization   17.093104\n",
       "17     40   77.516750           Amazon Web Services   45.544594\n",
       "18     24   77.197917              Content Strategy   43.388925\n",
       "19     17   75.588235                        VB.NET   60.021622\n",
       "20     29   75.402069                .NET Framework   47.065003\n",
       "21     21   74.428095                  Data Science   35.115206\n",
       "22     11   74.272727            Program Management   60.320494\n",
       "23     11   73.909091                     FileMaker   34.589685\n",
       "24     31   73.515806          Marketing Automation   40.198434\n",
       "25     23   73.500000              Brand Consulting   38.787885\n",
       "26     25   73.400000                        Django   45.327254\n",
       "27     43   73.323023             Business Analysis   42.767763\n",
       "28     13   72.730769                    PCB Design   29.522883\n",
       "29     20   72.650000               API Development   34.991820\n",
       "30     25   72.318800             Inbound Marketing   44.167893\n",
       "31     17   72.058824                Product Design   46.504548\n",
       "32     44   71.635000             Laravel Framework   34.267312\n",
       "33     12   71.416667               Cloud Computing   45.345816\n",
       "34     34   71.351765                   Objective-C   28.273441\n",
       "35     18   71.185556     Business Proposal Writing   27.271052\n",
       "36     37   70.972973         User Interface Design   52.110971\n",
       "37     36   70.611111        Mobile App Development   42.439681\n",
       "38     13   70.538462               Ionic Framework   36.659274\n",
       "39     26   70.480769            Strategic Planning   31.180764\n",
       "40     11   70.386364           Digital Electronics   33.705166"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "skill_tab.loc[0:40, skill_tab.columns != 'g_r']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 72,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "source": [
    "### Clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 30,
        "hidden": false,
        "row": 76,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 1, iteration: 1/100, moves: 187, cost: 20372.0\n",
      "Run 1, iteration: 2/100, moves: 91, cost: 20327.0\n",
      "Run 1, iteration: 3/100, moves: 15, cost: 20327.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 2, iteration: 1/100, moves: 478, cost: 20038.0\n",
      "Run 2, iteration: 2/100, moves: 85, cost: 20024.0\n",
      "Run 2, iteration: 3/100, moves: 5, cost: 20024.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 3, iteration: 1/100, moves: 473, cost: 19771.0\n",
      "Run 3, iteration: 2/100, moves: 93, cost: 19771.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 4, iteration: 1/100, moves: 176, cost: 20137.0\n",
      "Run 4, iteration: 2/100, moves: 0, cost: 20137.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 5, iteration: 1/100, moves: 187, cost: 20117.0\n",
      "Run 5, iteration: 2/100, moves: 106, cost: 20042.0\n",
      "Run 5, iteration: 3/100, moves: 21, cost: 20042.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 6, iteration: 1/100, moves: 356, cost: 19690.0\n",
      "Run 6, iteration: 2/100, moves: 107, cost: 19690.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 7, iteration: 1/100, moves: 257, cost: 20100.0\n",
      "Run 7, iteration: 2/100, moves: 15, cost: 20100.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 8, iteration: 1/100, moves: 171, cost: 20532.0\n",
      "Run 8, iteration: 2/100, moves: 16, cost: 20532.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 9, iteration: 1/100, moves: 150, cost: 19875.0\n",
      "Run 9, iteration: 2/100, moves: 0, cost: 19875.0\n",
      "Init: initializing centroids\n",
      "Init: initializing clusters\n",
      "Starting iterations...\n",
      "Run 10, iteration: 1/100, moves: 277, cost: 19791.0\n",
      "Run 10, iteration: 2/100, moves: 61, cost: 19779.0\n",
      "Run 10, iteration: 3/100, moves: 0, cost: 19779.0\n",
      "Best run was number 6\n",
      "[[0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]\n",
      " ...\n",
      " [0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]]\n"
     ]
    }
   ],
   "source": [
    "km = KModes(n_clusters=15, init='Huang', n_init=10, verbose=1)\n",
    "clusters = km.fit_predict(skills_df)\n",
    "print(km.cluster_centroids_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "kmodes = km.cluster_centroids_\n",
    "shape = kmodes.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 33,
        "hidden": false,
        "row": 90,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "cluster 0: \n",
      "Microsoft PowerPoint\n",
      "\n",
      "cluster 1: \n",
      "WordPress\n",
      "\n",
      "cluster 2: \n",
      "Adobe Illustrator\n",
      "Adobe Photoshop\n",
      "\n",
      "cluster 3: \n",
      "no-skills cluster\n",
      "\n",
      "cluster 4: \n",
      "Editing\n",
      "Proofreading\n",
      "\n",
      "cluster 5: \n",
      "Intuit QuickBooks\n",
      "\n",
      "cluster 6: \n",
      "English\n",
      "\n",
      "cluster 7: \n",
      "Microsoft Excel\n",
      "Microsoft PowerPoint\n",
      "Microsoft Word\n",
      "\n",
      "cluster 8: \n",
      "CSS\n",
      "HTML\n",
      "JavaScript\n",
      "PHP\n",
      "WordPress\n",
      "\n",
      "cluster 9: \n",
      "Article Writing\n",
      "Blog Writing\n",
      "\n",
      "cluster 10: \n",
      "Microsoft Excel\n",
      "SQL\n",
      "\n",
      "cluster 11: \n",
      "Accounting\n",
      "Bookkeeping\n",
      "Intuit QuickBooks\n",
      "Microsoft Excel\n",
      "\n",
      "cluster 12: \n",
      "Windows Media Connect\n",
      "\n",
      "cluster 13: \n",
      "Copywriting\n",
      "\n",
      "cluster 14: \n",
      "Data Entry\n"
     ]
    }
   ],
   "source": [
    "for i in range(shape[0]):\n",
    "    if sum(kmodes[i,:]) == 0:\n",
    "        print(\"\\ncluster \" + str(i) + \": \")\n",
    "        print(\"no-skills cluster\")\n",
    "    else:\n",
    "        print(\"\\ncluster \" + str(i) + \": \")\n",
    "        cent = kmodes[i,:]\n",
    "        for j in skills_df.columns[np.nonzero(cent)]:\n",
    "            print(j)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "clust_assigned = km.predict(skills_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 10,
        "hidden": false,
        "row": 106,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0: 35,\n",
       " 1: 296,\n",
       " 2: 226,\n",
       " 3: 1281,\n",
       " 4: 117,\n",
       " 5: 32,\n",
       " 6: 63,\n",
       " 7: 119,\n",
       " 8: 104,\n",
       " 9: 101,\n",
       " 10: 16,\n",
       " 11: 55,\n",
       " 12: 7,\n",
       " 13: 138,\n",
       " 14: 145}"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unique, counts = np.unique(clust_assigned, return_counts=True)\n",
    "dict(zip(unique, counts))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "restrict = copy.deepcopy(skills_df)\n",
    "for name in restrict.columns:\n",
    "    if sum(restrict[name]) < 50:\n",
    "        restrict = restrict.drop([name], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "restrict.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "km2 = KModes(n_clusters=15, init='Huang', n_init=11, verbose=1)\n",
    "clusters = km2.fit_predict(restrict)\n",
    "print(km2.cluster_centroids_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "kmodes2 = km2.cluster_centroids_\n",
    "shape2 = kmodes2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    },
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "for i in range(shape2[0]):\n",
    "    if sum(kmodes2[i,:]) == 0:\n",
    "        print(\"\\ncluster \" + str(i) + \": \")\n",
    "        print(\"no-skills cluster\")\n",
    "    else:\n",
    "        print(\"\\ncluster \" + str(i) + \": \")\n",
    "        cent = kmodes2[i,:]\n",
    "        for j in restrict.columns[np.nonzero(cent)]:\n",
    "            print(j)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "clust_assigned2 = km2.predict(restrict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "unique2, counts2 = np.unique(clust_assigned2, return_counts=True)\n",
    "dict(zip(unique2, counts2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 108,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "source": [
    "### Cluster Boxplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 112,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2735, 10)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "for_box = np.stack((clust_assigned, y), axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "box_df = pd.DataFrame(for_box)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "box_df.columns = ['cluster', 'rate']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 11,
        "hidden": false,
        "row": 116,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0, 200)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dfg = box_df.groupby('cluster')\n",
    "counts = [len(v) for k, v in dfg]\n",
    "total = float(sum(counts))\n",
    "cases = len(counts)\n",
    "widths = [c/total for c in counts]  \n",
    "cax = box_df.boxplot(column='rate', by='cluster', figsize=(15,10))\n",
    "cax.set_xticklabels(['%s\\n$n$=%d'%(k, len(v)) for k, v in dfg])\n",
    "cax.set_ylim([0,200])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 116,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "source": [
    "### Predicting with cluster means"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "box_means = dfg.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 120,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2735, 10)"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "y_and_cluster_means = pd.merge(box_means, box_df, how='inner', on='cluster', left_on=None, right_on=None,\n",
    "         left_index=False, right_index=True, sort=False,\n",
    "         suffixes=('_x', '_y'), copy=True, indicator=False,\n",
    "         validate=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 8,
        "hidden": false,
        "row": 123,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>rate_x</th>\n",
       "      <th>rate_y</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cluster</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>42.986286</td>\n",
       "      <td>21.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>42.986286</td>\n",
       "      <td>58.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>42.986286</td>\n",
       "      <td>125.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>42.986286</td>\n",
       "      <td>50.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>42.986286</td>\n",
       "      <td>70.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            rate_x  rate_y\n",
       "cluster                   \n",
       "0.0      42.986286    21.5\n",
       "0.0      42.986286    58.0\n",
       "0.0      42.986286   125.0\n",
       "0.0      42.986286    50.0\n",
       "0.0      42.986286    70.0"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_and_cluster_means.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 124,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0677674609173502"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r2_score(y_and_cluster_means.rate_y, y_and_cluster_means.rate_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 127,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.3310563798059647"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(mean_squared_error(y_and_cluster_means.rate_y, y_and_cluster_means.rate_x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 128,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "source": [
    "### Making the predictor matrix (more feature engineering)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "X = df.drop(['rate'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "X.skills = X.skills.map(lambda li: len(li))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "fn = lambda row: len(row.headline) # define a function for the new column\n",
    "X = X.assign(head_len = X.apply(fn, axis=1).values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "fn = lambda row: len(row.headline) // len(row.headline.split(' '))\n",
    "X = X.assign(hd_wrd_len = X.apply(fn, axis=1).values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "fn = lambda row: len(row.bio)\n",
    "X = X.assign(bio_len = X.apply(fn, axis=1).values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "fn = lambda row: len(row.bio) // len(row.bio.split(' ')) \n",
    "X = X.assign(bio_wrd_len = X.apply(fn, axis=1).values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 15,
        "hidden": false,
        "row": 132,
        "width": 9
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>bio</th>\n",
       "      <th>earned</th>\n",
       "      <th>edu</th>\n",
       "      <th>exp</th>\n",
       "      <th>headline</th>\n",
       "      <th>hours</th>\n",
       "      <th>jobs</th>\n",
       "      <th>skills</th>\n",
       "      <th>tests</th>\n",
       "      <th>head_len</th>\n",
       "      <th>hd_wrd_len</th>\n",
       "      <th>bio_len</th>\n",
       "      <th>bio_wrd_len</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Google Adwords &amp; Bing Ads Certified MBA market...</td>\n",
       "      <td>$200k+</td>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>Marketing, Strategic Mgmt/Writer/SEO, SEM/Mark...</td>\n",
       "      <td>10654</td>\n",
       "      <td>40</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>66</td>\n",
       "      <td>11</td>\n",
       "      <td>2608</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>I have been transcribing for more than 7 years...</td>\n",
       "      <td>$10k+</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Transcriptionist</td>\n",
       "      <td>1043</td>\n",
       "      <td>60</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>16</td>\n",
       "      <td>305</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Throughout my undergraduate years, I was a res...</td>\n",
       "      <td>$60k+</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Transcription (98WPM), Virtual Assistant</td>\n",
       "      <td>2840</td>\n",
       "      <td>254</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>40</td>\n",
       "      <td>10</td>\n",
       "      <td>430</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>As a digital marketing professional with a foc...</td>\n",
       "      <td>$30k+</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>Full-Stack Digital Marketer | Facebook Ads Spe...</td>\n",
       "      <td>673</td>\n",
       "      <td>21</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>53</td>\n",
       "      <td>7</td>\n",
       "      <td>2050</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>WordPress Development + Theme Customization + ...</td>\n",
       "      <td>$40k+</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Turn-Key Responsive Website Design &amp; Development</td>\n",
       "      <td>625</td>\n",
       "      <td>131</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>8</td>\n",
       "      <td>664</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 bio  earned  edu  exp  \\\n",
       "0  Google Adwords & Bing Ads Certified MBA market...  $200k+    3    8   \n",
       "1  I have been transcribing for more than 7 years...   $10k+    2    1   \n",
       "2  Throughout my undergraduate years, I was a res...   $60k+    1    3   \n",
       "3  As a digital marketing professional with a foc...   $30k+    2    2   \n",
       "4  WordPress Development + Theme Customization + ...   $40k+    0    1   \n",
       "\n",
       "                                            headline  hours  jobs  skills  \\\n",
       "0  Marketing, Strategic Mgmt/Writer/SEO, SEM/Mark...  10654    40       8   \n",
       "1                                   Transcriptionist   1043    60       2   \n",
       "2           Transcription (98WPM), Virtual Assistant   2840   254      10   \n",
       "3  Full-Stack Digital Marketer | Facebook Ads Spe...    673    21       5   \n",
       "4   Turn-Key Responsive Website Design & Development    625   131       6   \n",
       "\n",
       "   tests  head_len  hd_wrd_len  bio_len  bio_wrd_len  \n",
       "0      9        66          11     2608            6  \n",
       "1      2        16          16      305            6  \n",
       "2      4        40          10      430            6  \n",
       "3      4        53           7     2050            6  \n",
       "4      1        48           8      664            7  "
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "X = X.drop(['bio', 'earned', 'headline',], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 7,
        "hidden": false,
        "row": 147,
        "width": 6
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>edu</th>\n",
       "      <th>exp</th>\n",
       "      <th>hours</th>\n",
       "      <th>jobs</th>\n",
       "      <th>skills</th>\n",
       "      <th>tests</th>\n",
       "      <th>head_len</th>\n",
       "      <th>hd_wrd_len</th>\n",
       "      <th>bio_len</th>\n",
       "      <th>bio_wrd_len</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>10654</td>\n",
       "      <td>40</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>66</td>\n",
       "      <td>11</td>\n",
       "      <td>2608</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1043</td>\n",
       "      <td>60</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>16</td>\n",
       "      <td>305</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2840</td>\n",
       "      <td>254</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>40</td>\n",
       "      <td>10</td>\n",
       "      <td>430</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>673</td>\n",
       "      <td>21</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>53</td>\n",
       "      <td>7</td>\n",
       "      <td>2050</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>625</td>\n",
       "      <td>131</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>8</td>\n",
       "      <td>664</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   edu  exp  hours  jobs  skills  tests  head_len  hd_wrd_len  bio_len  \\\n",
       "0    3    8  10654    40       8      9        66          11     2608   \n",
       "1    2    1   1043    60       2      2        16          16      305   \n",
       "2    1    3   2840   254      10      4        40          10      430   \n",
       "3    2    2    673    21       5      4        53           7     2050   \n",
       "4    0    1    625   131       6      1        48           8      664   \n",
       "\n",
       "   bio_wrd_len  \n",
       "0            6  \n",
       "1            6  \n",
       "2            6  \n",
       "3            6  \n",
       "4            7  "
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 6,
        "height": 7,
        "hidden": false,
        "row": 147,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "edu            int64\n",
       "exp            int64\n",
       "hours          int64\n",
       "jobs           int64\n",
       "skills         int64\n",
       "tests          int64\n",
       "head_len       int64\n",
       "hd_wrd_len     int64\n",
       "bio_len        int64\n",
       "bio_wrd_len    int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "clust = pd.Series(clust_assigned)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 154,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2735,)"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clust.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "# clust = clust.reset_index(drop = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "# X = X.reset_index(drop = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "result = pd.concat([X, clust], axis = 1, sort = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "result.rename(columns={0: 'cluster'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 7,
        "hidden": false,
        "row": 154,
        "width": 7
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>edu</th>\n",
       "      <th>exp</th>\n",
       "      <th>hours</th>\n",
       "      <th>jobs</th>\n",
       "      <th>skills</th>\n",
       "      <th>tests</th>\n",
       "      <th>head_len</th>\n",
       "      <th>hd_wrd_len</th>\n",
       "      <th>bio_len</th>\n",
       "      <th>bio_wrd_len</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>10654</td>\n",
       "      <td>40</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>66</td>\n",
       "      <td>11</td>\n",
       "      <td>2608</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1043</td>\n",
       "      <td>60</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>16</td>\n",
       "      <td>305</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2840</td>\n",
       "      <td>254</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>40</td>\n",
       "      <td>10</td>\n",
       "      <td>430</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>673</td>\n",
       "      <td>21</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>53</td>\n",
       "      <td>7</td>\n",
       "      <td>2050</td>\n",
       "      <td>6</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>625</td>\n",
       "      <td>131</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>8</td>\n",
       "      <td>664</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   edu  exp  hours  jobs  skills  tests  head_len  hd_wrd_len  bio_len  \\\n",
       "0    3    8  10654    40       8      9        66          11     2608   \n",
       "1    2    1   1043    60       2      2        16          16      305   \n",
       "2    1    3   2840   254      10      4        40          10      430   \n",
       "3    2    2    673    21       5      4        53           7     2050   \n",
       "4    0    1    625   131       6      1        48           8      664   \n",
       "\n",
       "   bio_wrd_len  cluster  \n",
       "0            6        3  \n",
       "1            6        3  \n",
       "2            6        6  \n",
       "3            6       13  \n",
       "4            7        1  "
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "result_dum = pd.get_dummies(result, columns = ['cluster'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 8,
        "hidden": false,
        "row": 161,
        "width": 12
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>edu</th>\n",
       "      <th>exp</th>\n",
       "      <th>hours</th>\n",
       "      <th>jobs</th>\n",
       "      <th>skills</th>\n",
       "      <th>tests</th>\n",
       "      <th>head_len</th>\n",
       "      <th>hd_wrd_len</th>\n",
       "      <th>bio_len</th>\n",
       "      <th>bio_wrd_len</th>\n",
       "      <th>...</th>\n",
       "      <th>cluster_5</th>\n",
       "      <th>cluster_6</th>\n",
       "      <th>cluster_7</th>\n",
       "      <th>cluster_8</th>\n",
       "      <th>cluster_9</th>\n",
       "      <th>cluster_10</th>\n",
       "      <th>cluster_11</th>\n",
       "      <th>cluster_12</th>\n",
       "      <th>cluster_13</th>\n",
       "      <th>cluster_14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>10654</td>\n",
       "      <td>40</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>66</td>\n",
       "      <td>11</td>\n",
       "      <td>2608</td>\n",
       "      <td>6</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1043</td>\n",
       "      <td>60</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>16</td>\n",
       "      <td>305</td>\n",
       "      <td>6</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2840</td>\n",
       "      <td>254</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>40</td>\n",
       "      <td>10</td>\n",
       "      <td>430</td>\n",
       "      <td>6</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>673</td>\n",
       "      <td>21</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>53</td>\n",
       "      <td>7</td>\n",
       "      <td>2050</td>\n",
       "      <td>6</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>625</td>\n",
       "      <td>131</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>8</td>\n",
       "      <td>664</td>\n",
       "      <td>7</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   edu  exp  hours  jobs  skills  tests  head_len  hd_wrd_len  bio_len  \\\n",
       "0    3    8  10654    40       8      9        66          11     2608   \n",
       "1    2    1   1043    60       2      2        16          16      305   \n",
       "2    1    3   2840   254      10      4        40          10      430   \n",
       "3    2    2    673    21       5      4        53           7     2050   \n",
       "4    0    1    625   131       6      1        48           8      664   \n",
       "\n",
       "   bio_wrd_len     ...      cluster_5  cluster_6  cluster_7  cluster_8  \\\n",
       "0            6     ...              0          0          0          0   \n",
       "1            6     ...              0          0          0          0   \n",
       "2            6     ...              0          1          0          0   \n",
       "3            6     ...              0          0          0          0   \n",
       "4            7     ...              0          0          0          0   \n",
       "\n",
       "   cluster_9  cluster_10  cluster_11  cluster_12  cluster_13  cluster_14  \n",
       "0          0           0           0           0           0           0  \n",
       "1          0           0           0           0           0           0  \n",
       "2          0           0           0           0           0           0  \n",
       "3          0           0           0           0           1           0  \n",
       "4          0           0           0           0           0           0  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_dum.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 169,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2735, 25)"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_dum.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 16,
        "hidden": false,
        "row": 169,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "edu            0\n",
       "exp            0\n",
       "hours          0\n",
       "jobs           0\n",
       "skills         0\n",
       "tests          0\n",
       "head_len       0\n",
       "hd_wrd_len     0\n",
       "bio_len        0\n",
       "bio_wrd_len    0\n",
       "cluster_0      0\n",
       "cluster_1      0\n",
       "cluster_2      0\n",
       "cluster_3      0\n",
       "cluster_4      0\n",
       "cluster_5      0\n",
       "cluster_6      0\n",
       "cluster_7      0\n",
       "cluster_8      0\n",
       "cluster_9      0\n",
       "cluster_10     0\n",
       "cluster_11     0\n",
       "cluster_12     0\n",
       "cluster_13     0\n",
       "cluster_14     0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_dum.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 169,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "source": [
    "### Normalizing the predictor variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "norm_x = scale(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "pca = PCA()\n",
    "nx = pca.fit_transform(norm_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 173,
        "width": 4
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.81303002, 1.22582857, 1.10477938, 1.01006683, 0.98951616,\n",
       "       0.88813037, 0.79751108, 0.76819701, 0.73194161, 0.67465662])"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.explained_variance_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "nx = pd.DataFrame(nx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 7,
        "hidden": false,
        "row": 185,
        "width": 8
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.115624</td>\n",
       "      <td>2.632597</td>\n",
       "      <td>-0.101756</td>\n",
       "      <td>1.316947</td>\n",
       "      <td>-1.653452</td>\n",
       "      <td>0.231310</td>\n",
       "      <td>1.627862</td>\n",
       "      <td>-0.594969</td>\n",
       "      <td>0.095583</td>\n",
       "      <td>0.396956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.726502</td>\n",
       "      <td>1.271565</td>\n",
       "      <td>2.402553</td>\n",
       "      <td>1.452574</td>\n",
       "      <td>0.502881</td>\n",
       "      <td>-1.715443</td>\n",
       "      <td>0.858174</td>\n",
       "      <td>0.456867</td>\n",
       "      <td>0.444865</td>\n",
       "      <td>0.489602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.136515</td>\n",
       "      <td>0.797284</td>\n",
       "      <td>0.162744</td>\n",
       "      <td>0.474670</td>\n",
       "      <td>0.298927</td>\n",
       "      <td>0.099244</td>\n",
       "      <td>-1.225123</td>\n",
       "      <td>-0.269011</td>\n",
       "      <td>-0.047908</td>\n",
       "      <td>0.241261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.408673</td>\n",
       "      <td>-0.119873</td>\n",
       "      <td>-0.698414</td>\n",
       "      <td>0.013913</td>\n",
       "      <td>-0.101541</td>\n",
       "      <td>-0.898067</td>\n",
       "      <td>1.534765</td>\n",
       "      <td>0.814199</td>\n",
       "      <td>-0.054682</td>\n",
       "      <td>0.056092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.341179</td>\n",
       "      <td>-0.293220</td>\n",
       "      <td>-0.635489</td>\n",
       "      <td>0.966307</td>\n",
       "      <td>1.205188</td>\n",
       "      <td>0.312058</td>\n",
       "      <td>-0.050191</td>\n",
       "      <td>0.097328</td>\n",
       "      <td>-0.576039</td>\n",
       "      <td>-0.606702</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3         4         5         6  \\\n",
       "0  4.115624  2.632597 -0.101756  1.316947 -1.653452  0.231310  1.627862   \n",
       "1 -1.726502  1.271565  2.402553  1.452574  0.502881 -1.715443  0.858174   \n",
       "2  0.136515  0.797284  0.162744  0.474670  0.298927  0.099244 -1.225123   \n",
       "3  0.408673 -0.119873 -0.698414  0.013913 -0.101541 -0.898067  1.534765   \n",
       "4 -1.341179 -0.293220 -0.635489  0.966307  1.205188  0.312058 -0.050191   \n",
       "\n",
       "          7         8         9  \n",
       "0 -0.594969  0.095583  0.396956  \n",
       "1  0.456867  0.444865  0.489602  \n",
       "2 -0.269011 -0.047908  0.241261  \n",
       "3  0.814199 -0.054682  0.056092  \n",
       "4  0.097328 -0.576039 -0.606702  "
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nx.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "norm_result = pd.concat([nx, clust], axis = 1, sort = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "norm_result.columns = result.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 7,
        "hidden": false,
        "row": 192,
        "width": 9
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>edu</th>\n",
       "      <th>exp</th>\n",
       "      <th>hours</th>\n",
       "      <th>jobs</th>\n",
       "      <th>skills</th>\n",
       "      <th>tests</th>\n",
       "      <th>head_len</th>\n",
       "      <th>hd_wrd_len</th>\n",
       "      <th>bio_len</th>\n",
       "      <th>bio_wrd_len</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.115624</td>\n",
       "      <td>2.632597</td>\n",
       "      <td>-0.101756</td>\n",
       "      <td>1.316947</td>\n",
       "      <td>-1.653452</td>\n",
       "      <td>0.231310</td>\n",
       "      <td>1.627862</td>\n",
       "      <td>-0.594969</td>\n",
       "      <td>0.095583</td>\n",
       "      <td>0.396956</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.726502</td>\n",
       "      <td>1.271565</td>\n",
       "      <td>2.402553</td>\n",
       "      <td>1.452574</td>\n",
       "      <td>0.502881</td>\n",
       "      <td>-1.715443</td>\n",
       "      <td>0.858174</td>\n",
       "      <td>0.456867</td>\n",
       "      <td>0.444865</td>\n",
       "      <td>0.489602</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.136515</td>\n",
       "      <td>0.797284</td>\n",
       "      <td>0.162744</td>\n",
       "      <td>0.474670</td>\n",
       "      <td>0.298927</td>\n",
       "      <td>0.099244</td>\n",
       "      <td>-1.225123</td>\n",
       "      <td>-0.269011</td>\n",
       "      <td>-0.047908</td>\n",
       "      <td>0.241261</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.408673</td>\n",
       "      <td>-0.119873</td>\n",
       "      <td>-0.698414</td>\n",
       "      <td>0.013913</td>\n",
       "      <td>-0.101541</td>\n",
       "      <td>-0.898067</td>\n",
       "      <td>1.534765</td>\n",
       "      <td>0.814199</td>\n",
       "      <td>-0.054682</td>\n",
       "      <td>0.056092</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.341179</td>\n",
       "      <td>-0.293220</td>\n",
       "      <td>-0.635489</td>\n",
       "      <td>0.966307</td>\n",
       "      <td>1.205188</td>\n",
       "      <td>0.312058</td>\n",
       "      <td>-0.050191</td>\n",
       "      <td>0.097328</td>\n",
       "      <td>-0.576039</td>\n",
       "      <td>-0.606702</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        edu       exp     hours      jobs    skills     tests  head_len  \\\n",
       "0  4.115624  2.632597 -0.101756  1.316947 -1.653452  0.231310  1.627862   \n",
       "1 -1.726502  1.271565  2.402553  1.452574  0.502881 -1.715443  0.858174   \n",
       "2  0.136515  0.797284  0.162744  0.474670  0.298927  0.099244 -1.225123   \n",
       "3  0.408673 -0.119873 -0.698414  0.013913 -0.101541 -0.898067  1.534765   \n",
       "4 -1.341179 -0.293220 -0.635489  0.966307  1.205188  0.312058 -0.050191   \n",
       "\n",
       "   hd_wrd_len   bio_len  bio_wrd_len  cluster  \n",
       "0   -0.594969  0.095583     0.396956        3  \n",
       "1    0.456867  0.444865     0.489602        3  \n",
       "2   -0.269011 -0.047908     0.241261        6  \n",
       "3    0.814199 -0.054682     0.056092       13  \n",
       "4    0.097328 -0.576039    -0.606702        1  "
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "norm_result.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "norm_dum = pd.get_dummies(norm_result, columns = ['cluster'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 8,
        "hidden": false,
        "row": 199,
        "width": 12
       },
       "report_default": {
        "hidden": true
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>edu</th>\n",
       "      <th>exp</th>\n",
       "      <th>hours</th>\n",
       "      <th>jobs</th>\n",
       "      <th>skills</th>\n",
       "      <th>tests</th>\n",
       "      <th>head_len</th>\n",
       "      <th>hd_wrd_len</th>\n",
       "      <th>bio_len</th>\n",
       "      <th>bio_wrd_len</th>\n",
       "      <th>...</th>\n",
       "      <th>cluster_5</th>\n",
       "      <th>cluster_6</th>\n",
       "      <th>cluster_7</th>\n",
       "      <th>cluster_8</th>\n",
       "      <th>cluster_9</th>\n",
       "      <th>cluster_10</th>\n",
       "      <th>cluster_11</th>\n",
       "      <th>cluster_12</th>\n",
       "      <th>cluster_13</th>\n",
       "      <th>cluster_14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.115624</td>\n",
       "      <td>2.632597</td>\n",
       "      <td>-0.101756</td>\n",
       "      <td>1.316947</td>\n",
       "      <td>-1.653452</td>\n",
       "      <td>0.231310</td>\n",
       "      <td>1.627862</td>\n",
       "      <td>-0.594969</td>\n",
       "      <td>0.095583</td>\n",
       "      <td>0.396956</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.726502</td>\n",
       "      <td>1.271565</td>\n",
       "      <td>2.402553</td>\n",
       "      <td>1.452574</td>\n",
       "      <td>0.502881</td>\n",
       "      <td>-1.715443</td>\n",
       "      <td>0.858174</td>\n",
       "      <td>0.456867</td>\n",
       "      <td>0.444865</td>\n",
       "      <td>0.489602</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.136515</td>\n",
       "      <td>0.797284</td>\n",
       "      <td>0.162744</td>\n",
       "      <td>0.474670</td>\n",
       "      <td>0.298927</td>\n",
       "      <td>0.099244</td>\n",
       "      <td>-1.225123</td>\n",
       "      <td>-0.269011</td>\n",
       "      <td>-0.047908</td>\n",
       "      <td>0.241261</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.408673</td>\n",
       "      <td>-0.119873</td>\n",
       "      <td>-0.698414</td>\n",
       "      <td>0.013913</td>\n",
       "      <td>-0.101541</td>\n",
       "      <td>-0.898067</td>\n",
       "      <td>1.534765</td>\n",
       "      <td>0.814199</td>\n",
       "      <td>-0.054682</td>\n",
       "      <td>0.056092</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.341179</td>\n",
       "      <td>-0.293220</td>\n",
       "      <td>-0.635489</td>\n",
       "      <td>0.966307</td>\n",
       "      <td>1.205188</td>\n",
       "      <td>0.312058</td>\n",
       "      <td>-0.050191</td>\n",
       "      <td>0.097328</td>\n",
       "      <td>-0.576039</td>\n",
       "      <td>-0.606702</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        edu       exp     hours      jobs    skills     tests  head_len  \\\n",
       "0  4.115624  2.632597 -0.101756  1.316947 -1.653452  0.231310  1.627862   \n",
       "1 -1.726502  1.271565  2.402553  1.452574  0.502881 -1.715443  0.858174   \n",
       "2  0.136515  0.797284  0.162744  0.474670  0.298927  0.099244 -1.225123   \n",
       "3  0.408673 -0.119873 -0.698414  0.013913 -0.101541 -0.898067  1.534765   \n",
       "4 -1.341179 -0.293220 -0.635489  0.966307  1.205188  0.312058 -0.050191   \n",
       "\n",
       "   hd_wrd_len   bio_len  bio_wrd_len     ...      cluster_5  cluster_6  \\\n",
       "0   -0.594969  0.095583     0.396956     ...              0          0   \n",
       "1    0.456867  0.444865     0.489602     ...              0          0   \n",
       "2   -0.269011 -0.047908     0.241261     ...              0          1   \n",
       "3    0.814199 -0.054682     0.056092     ...              0          0   \n",
       "4    0.097328 -0.576039    -0.606702     ...              0          0   \n",
       "\n",
       "   cluster_7  cluster_8  cluster_9  cluster_10  cluster_11  cluster_12  \\\n",
       "0          0          0          0           0           0           0   \n",
       "1          0          0          0           0           0           0   \n",
       "2          0          0          0           0           0           0   \n",
       "3          0          0          0           0           0           0   \n",
       "4          0          0          0           0           0           0   \n",
       "\n",
       "   cluster_13  cluster_14  \n",
       "0           0           0  \n",
       "1           0           0  \n",
       "2           0           0  \n",
       "3           1           0  \n",
       "4           0           0  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "norm_dum.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 173,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "source": [
    "### Running the models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "cv = 4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "tr = DecisionTreeRegressor(max_depth = 3)\n",
    "trmodel = tr.fit(result_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 177,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.11288552800797969"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trmodel.score(result_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 177,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.05084381160230814"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(trmodel, result_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 181,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9988810252215856"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1*np.mean(cross_val_score(trmodel, result_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "ntr = DecisionTreeRegressor(max_depth = 3)\n",
    "ntrmodel = tr.fit(norm_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 181,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.1085991740546911"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ntrmodel.score(norm_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 185,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.01412761871132312"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(ntrmodel, norm_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 207,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.16075770656295418"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1*np.mean(cross_val_score(ntrmodel, norm_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "rf = RandomForestRegressor(n_estimators = 500, max_depth = 5)\n",
    "rfmodel = rf.fit(result_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 207,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.24354880219197772"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfmodel.score(result_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 207,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.10503422153554778"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(rfmodel, result_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 211,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.971421265251557"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1*np.mean(cross_val_score(rfmodel, result_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "nrf = RandomForestRegressor(n_estimators = 500, max_depth = 5)\n",
    "nrfmodel = rf.fit(norm_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 211,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.28980495376261994"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nrfmodel.score(norm_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 211,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.08247892261068604"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(nrfmodel, norm_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 215,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.5397921514130815"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1*np.mean(cross_val_score(nrfmodel, norm_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 14,
        "hidden": false,
        "row": 215,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 1  9  7  0  8 24  2  6  4 17 13  3  5 23 16 18 12 11 14 19 21 20 10 15\n",
      " 22]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_feature_importances(model):\n",
    "    importances = model.feature_importances_\n",
    "    std = np.std([tree.feature_importances_ for tree in model.estimators_], axis=0)\n",
    "    indices = np.argsort(importances)[::-1]\n",
    "    print(indices)\n",
    "    n_features = norm_dum.shape[1]\n",
    "    plt.figure(figsize=(8,8))\n",
    "    plt.barh(range(n_features), importances[indices], xerr=std[indices], align='center') \n",
    "    plt.yticks(np.arange(n_features), norm_dum.columns.values[indices]) \n",
    "    plt.xlabel(\"Feature importance\")\n",
    "    plt.ylabel(\"Feature\")\n",
    "\n",
    "plot_feature_importances(rfmodel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "ada = AdaBoostRegressor(n_estimators = 1000, learning_rate = 0.01)\n",
    "adamodel = ada.fit(result_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 215,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.25826506028577956"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "adamodel.score(result_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 219,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.007849528391269983"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(adamodel, result_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 219,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.02265542735789694"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1*np.mean(cross_val_score(adamodel, result_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "nada = AdaBoostRegressor(n_estimators = 1000, learning_rate = 0.01)\n",
    "nadamodel = nada.fit(norm_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 223,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3079031254362068"
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nadamodel.score(norm_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 223,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.009061864819616322"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(nadamodel, norm_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 227,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.22122705172412793"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1*np.mean(cross_val_score(nadamodel, norm_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "parameters = {'metric':('jaccard', 'matching', 'dice', 'kulsinski', 'rogerstanimoto', 'russellrao', \n",
    "                        'sokalmichener', 'sokalsneath'), 'n_neighbors':[2, 10, 25, 50, 100]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 11,
        "hidden": false,
        "row": 227,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=4, error_score='raise',\n",
       "       estimator=KNeighborsRegressor(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "          metric_params=None, n_jobs=1, n_neighbors=5, p=2,\n",
       "          weights='uniform'),\n",
       "       fit_params=None, iid=True, n_jobs=1,\n",
       "       param_grid={'metric': ('jaccard', 'matching', 'dice', 'kulsinski', 'rogerstanimoto', 'russellrao', 'sokalmichener', 'sokalsneath'), 'n_neighbors': [2, 10, 25, 50, 100]},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring='neg_mean_squared_error', verbose=0)"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neigh = KNeighborsRegressor()\n",
    "reg = GridSearchCV(neigh, parameters, cv = cv, scoring = 'neg_mean_squared_error')\n",
    "reg.fit(result_dum, y) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "bestknn = reg.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 229,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'metric': 'rogerstanimoto', 'n_neighbors': 100}"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bestknn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 231,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.1070244151798576"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knnbest = KNeighborsRegressor(n_neighbors = bestknn['n_neighbors'], metric = bestknn['metric'])\n",
    "std - np.sqrt(-1*np.mean(cross_val_score(knnbest, result_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 233,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.06314157945862667"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(knnbest, result_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 11,
        "hidden": false,
        "row": 235,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=4, error_score='raise',\n",
       "       estimator=KNeighborsRegressor(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "          metric_params=None, n_jobs=1, n_neighbors=5, p=2,\n",
       "          weights='uniform'),\n",
       "       fit_params=None, iid=True, n_jobs=1,\n",
       "       param_grid={'metric': ('jaccard', 'matching', 'dice', 'kulsinski', 'rogerstanimoto', 'russellrao', 'sokalmichener', 'sokalsneath'), 'n_neighbors': [2, 10, 25, 50, 100]},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring='neg_mean_squared_error', verbose=0)"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nneigh = KNeighborsRegressor()\n",
    "nreg = GridSearchCV(nneigh, parameters, cv = cv, scoring = 'neg_mean_squared_error')\n",
    "nreg.fit(norm_dum, y) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "nbestknn = nreg.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 237,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'metric': 'jaccard', 'n_neighbors': 100}"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nbestknn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 238,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8343229917787625"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nknnbest = KNeighborsRegressor(n_neighbors = nbestknn['n_neighbors'], metric = nbestknn['metric'])\n",
    "std - np.sqrt(-1*np.mean(cross_val_score(nknnbest, norm_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 241,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.04647942300074939"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(nknnbest, norm_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "ridgeReg = Ridge(alpha=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 242,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ridge(alpha=100, copy_X=True, fit_intercept=True, max_iter=None,\n",
       "   normalize=False, random_state=None, solver='auto', tol=0.001)"
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ridgeReg.fit(result_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 245,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "20.36711430741136"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(ridgeReg.coef_) - min(ridgeReg.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "coef = pd.Series(ridgeReg.coef_, result_dum.columns).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 12,
        "hidden": false,
        "row": 246,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1c33b169b0>"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "coef.plot(kind='bar', title='Model Coefficients')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 246,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.04490569474595607"
      ]
     },
     "execution_count": 224,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(Ridge(alpha=1), result_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 249,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.1382957487734942"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1 * np.mean(cross_val_score(Ridge(alpha=1), result_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "nridgeReg = Ridge(alpha=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 250,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ridge(alpha=0.01, copy_X=True, fit_intercept=True, max_iter=None,\n",
       "   normalize=False, random_state=None, solver='auto', tol=0.001)"
      ]
     },
     "execution_count": 211,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nridgeReg.fit(norm_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 253,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33.37522190150324"
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(nridgeReg.coef_) - min(nridgeReg.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "ncoef = pd.Series(nridgeReg.coef_, norm_dum.columns).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 12,
        "hidden": false,
        "row": 254,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1c33718c88>"
      ]
     },
     "execution_count": 214,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ncoef.plot(kind='bar', title='Model Coefficients')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 257,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.09225152966556757"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(Ridge(alpha=100), norm_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 258,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.8454828672642378"
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1 * np.mean(cross_val_score(Ridge(alpha=100), norm_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "lin = LinearRegression()\n",
    "linmodel = lin.fit(result_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 261,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.13340337760168108"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "linmodel.score(result_dum,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 262,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33.378094485284144"
      ]
     },
     "execution_count": 160,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(linmodel.coef_) - min(linmodel.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "coef2 = pd.Series(linmodel.coef_, result_dum.columns).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 12,
        "hidden": false,
        "row": 265,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1c3387cf60>"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "coef2.plot(kind='bar', title='Model Coefficients')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 266,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9.610973798263316"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "linmodel.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 266,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.04453410321086229"
      ]
     },
     "execution_count": 164,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(LinearRegression(), result_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 270,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.1313683473416205"
      ]
     },
     "execution_count": 165,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1 * np.mean(cross_val_score(LinearRegression(), result_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "nlin = LinearRegression()\n",
    "nlinmodel = nlin.fit(norm_dum, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 270,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.13340337760168097"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nlinmodel.score(norm_dum,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 274,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33.378094485284166"
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(nlinmodel.coef_) - min(nlinmodel.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "coef3 = pd.Series(nlinmodel.coef_, norm_dum.columns).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 12,
        "hidden": false,
        "row": 274,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x102968a20>"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "coef3.plot(kind='bar', title='Model Coefficients')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 277,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "41.477367676568655"
      ]
     },
     "execution_count": 171,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nlinmodel.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 278,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.04453410321085871"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(cross_val_score(LinearRegression(), norm_dum, y, scoring = 'r2', cv = cv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 281,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.1313683473415637"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - np.sqrt(-1 * np.mean(cross_val_score(LinearRegression(), norm_dum, y, scoring = 'neg_mean_squared_error', cv = cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "XX = np.array(copy.deepcopy(result_dum))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "XXT = XX.T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "prod = XXT @ XX"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 282,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25, 25)"
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prod.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "inv = np.linalg.inv(prod)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 285,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25, 25)"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inv.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "triple_X = inv @ XXT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 286,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(triple_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "z = np.array(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 286,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2735,)"
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "theta = triple_X @ z"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 4,
        "hidden": false,
        "row": 289,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(XX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "preds = [theta @ XX[i,:] for i in range(len(z))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "err = [(y[i] - preds[i])for i in range(len(y))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "RMSE = (sum([e**2 for e in err])/len(y))**0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 290,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.6672119750817913"
      ]
     },
     "execution_count": 189,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - RMSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "results = sm.OLS(y, XX).fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 13,
        "hidden": false,
        "row": 290,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('edu', 0.2278061412293597), ('exp', 0.09927680162205871), ('hours', -0.0006173232632130546), ('jobs', 0.002127088785313749), ('skills', 1.0716686807190845), ('tests', -0.4095252509542218), ('head_len', 0.09691304230712537), ('hd_wrd_len', -0.804404966730424), ('bio_len', 0.00875239152388405), ('bio_wrd_len', 2.8970009969806743), ('cluster_0', 13.648727339038128), ('cluster_1', 13.431145279778724), ('cluster_2', 12.89633492308225), ('cluster_3', 21.809480776081784), ('cluster_4', 2.708339852305612), ('cluster_5', 4.503165977475874), ('cluster_6', 6.2957489140947125), ('cluster_7', -4.307110625938661), ('cluster_8', 26.44049263915564), ('cluster_9', 10.33995080731658), ('cluster_10', 18.89416567925813), ('cluster_11', 5.903768400432163), ('cluster_12', -2.5372280279314516), ('cluster_13', 21.075226885939674), ('cluster_14', -6.937601846128443)]\n"
     ]
    }
   ],
   "source": [
    "print(list(zip(list(result_dum.columns),list(theta))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 4,
        "height": 56,
        "hidden": false,
        "row": 293,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   rate   R-squared:                       0.133\n",
      "Model:                            OLS   Adj. R-squared:                  0.126\n",
      "Method:                 Least Squares   F-statistic:                     17.38\n",
      "Date:                Sun, 07 Oct 2018   Prob (F-statistic):           1.03e-67\n",
      "Time:                        16:32:13   Log-Likelihood:                -13677.\n",
      "No. Observations:                2735   AIC:                         2.740e+04\n",
      "Df Residuals:                    2710   BIC:                         2.755e+04\n",
      "Df Model:                          24                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "x1             0.2278      0.875      0.260      0.795      -1.489       1.944\n",
      "x2             0.0993      0.327      0.303      0.762      -0.543       0.741\n",
      "x3            -0.0006      0.000     -2.240      0.025      -0.001   -7.69e-05\n",
      "x4             0.0021      0.002      0.995      0.320      -0.002       0.006\n",
      "x5             1.0717      0.278      3.852      0.000       0.526       1.617\n",
      "x6            -0.4095      0.206     -1.984      0.047      -0.814      -0.005\n",
      "x7             0.0969      0.042      2.320      0.020       0.015       0.179\n",
      "x8            -0.8044      0.237     -3.392      0.001      -1.269      -0.339\n",
      "x9             0.0088      0.001      9.627      0.000       0.007       0.011\n",
      "x10            2.8970      0.669      4.329      0.000       1.585       4.209\n",
      "x11           13.6487      7.767      1.757      0.079      -1.581      28.878\n",
      "x12           13.4311      5.307      2.531      0.011       3.024      23.838\n",
      "x13           12.8963      5.274      2.445      0.015       2.555      23.238\n",
      "x14           21.8095      4.904      4.447      0.000      12.194      31.425\n",
      "x15            2.7083      5.924      0.457      0.648      -8.908      14.325\n",
      "x16            4.5032      7.905      0.570      0.569     -10.998      20.004\n",
      "x17            6.2957      6.691      0.941      0.347      -6.823      19.415\n",
      "x18           -4.3071      5.911     -0.729      0.466     -15.898       7.284\n",
      "x19           26.4405      5.961      4.436      0.000      14.752      38.129\n",
      "x20           10.3400      6.062      1.706      0.088      -1.547      22.227\n",
      "x21           18.8942     10.345      1.826      0.068      -1.390      39.178\n",
      "x22            5.9038      7.040      0.839      0.402      -7.901      19.708\n",
      "x23           -2.5372     14.343     -0.177      0.860     -30.661      25.586\n",
      "x24           21.0752      5.748      3.667      0.000       9.805      32.346\n",
      "x25           -6.9376      5.747     -1.207      0.227     -18.206       4.330\n",
      "==============================================================================\n",
      "Omnibus:                     4022.028   Durbin-Watson:                   1.996\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          4002790.709\n",
      "Skew:                           8.327   Prob(JB):                         0.00\n",
      "Kurtosis:                     189.675   Cond. No.                     8.57e+04\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.57e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    }
   ],
   "source": [
    "print(results.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 12,
        "hidden": false,
        "row": 294,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 2051 samples, validate on 684 samples\n",
      "Epoch 1/30\n",
      "2051/2051 [==============================] - 2s 1ms/step - loss: 3411.7151 - mean_squared_error: 3411.7151 - val_loss: 2324.5587 - val_mean_squared_error: 2324.5587\n",
      "Epoch 2/30\n",
      "2051/2051 [==============================] - 1s 321us/step - loss: 1894.7607 - mean_squared_error: 1894.7607 - val_loss: 1101.0096 - val_mean_squared_error: 1101.0096\n",
      "Epoch 3/30\n",
      "2051/2051 [==============================] - 1s 309us/step - loss: 1465.4591 - mean_squared_error: 1465.4591 - val_loss: 1059.2096 - val_mean_squared_error: 1059.2096\n",
      "Epoch 4/30\n",
      "2051/2051 [==============================] - 1s 315us/step - loss: 1439.3951 - mean_squared_error: 1439.3951 - val_loss: 1046.5157 - val_mean_squared_error: 1046.5157\n",
      "Epoch 5/30\n",
      "2051/2051 [==============================] - 1s 350us/step - loss: 1428.2610 - mean_squared_error: 1428.2610 - val_loss: 1037.6554 - val_mean_squared_error: 1037.6554\n",
      "Epoch 6/30\n",
      "2051/2051 [==============================] - 1s 335us/step - loss: 1419.5770 - mean_squared_error: 1419.5770 - val_loss: 1028.9508 - val_mean_squared_error: 1028.9508\n",
      "Epoch 7/30\n",
      "2051/2051 [==============================] - 1s 300us/step - loss: 1411.8987 - mean_squared_error: 1411.8987 - val_loss: 1024.0863 - val_mean_squared_error: 1024.0863\n",
      "Epoch 8/30\n",
      "2051/2051 [==============================] - 1s 344us/step - loss: 1408.4517 - mean_squared_error: 1408.4517 - val_loss: 1017.9654 - val_mean_squared_error: 1017.9654\n",
      "Epoch 9/30\n",
      "2051/2051 [==============================] - 1s 320us/step - loss: 1401.2229 - mean_squared_error: 1401.2229 - val_loss: 1015.3307 - val_mean_squared_error: 1015.3307\n",
      "Epoch 10/30\n",
      "2051/2051 [==============================] - 1s 303us/step - loss: 1399.9041 - mean_squared_error: 1399.9041 - val_loss: 1016.8087 - val_mean_squared_error: 1016.8087\n",
      "Epoch 11/30\n",
      "2051/2051 [==============================] - 1s 336us/step - loss: 1395.6145 - mean_squared_error: 1395.6145 - val_loss: 1007.6912 - val_mean_squared_error: 1007.6912\n",
      "Epoch 12/30\n",
      "2051/2051 [==============================] - 1s 313us/step - loss: 1393.5267 - mean_squared_error: 1393.5267 - val_loss: 1007.7543 - val_mean_squared_error: 1007.7543\n",
      "Epoch 13/30\n",
      "2051/2051 [==============================] - 1s 361us/step - loss: 1389.3108 - mean_squared_error: 1389.3108 - val_loss: 1005.6668 - val_mean_squared_error: 1005.6668\n",
      "Epoch 14/30\n",
      "2051/2051 [==============================] - 1s 314us/step - loss: 1386.6558 - mean_squared_error: 1386.6558 - val_loss: 1007.4148 - val_mean_squared_error: 1007.4148\n",
      "Epoch 15/30\n",
      "2051/2051 [==============================] - 1s 291us/step - loss: 1384.5403 - mean_squared_error: 1384.5403 - val_loss: 1007.6200 - val_mean_squared_error: 1007.6200\n",
      "Epoch 16/30\n",
      "2051/2051 [==============================] - 1s 308us/step - loss: 1381.7017 - mean_squared_error: 1381.7017 - val_loss: 1005.6663 - val_mean_squared_error: 1005.6663\n",
      "Epoch 17/30\n",
      "2051/2051 [==============================] - 1s 305us/step - loss: 1380.1686 - mean_squared_error: 1380.1686 - val_loss: 1004.1698 - val_mean_squared_error: 1004.1698\n",
      "Epoch 18/30\n",
      "2051/2051 [==============================] - 1s 332us/step - loss: 1380.1451 - mean_squared_error: 1380.1451 - val_loss: 1008.5462 - val_mean_squared_error: 1008.5462\n",
      "Epoch 19/30\n",
      "2051/2051 [==============================] - 1s 306us/step - loss: 1376.6911 - mean_squared_error: 1376.6911 - val_loss: 1001.4405 - val_mean_squared_error: 1001.4405\n",
      "Epoch 20/30\n",
      "2051/2051 [==============================] - 1s 316us/step - loss: 1373.7871 - mean_squared_error: 1373.7871 - val_loss: 1001.1340 - val_mean_squared_error: 1001.1340\n",
      "Epoch 21/30\n",
      "2051/2051 [==============================] - 1s 330us/step - loss: 1367.1886 - mean_squared_error: 1367.1886 - val_loss: 1011.1182 - val_mean_squared_error: 1011.1182\n",
      "Epoch 22/30\n",
      "2051/2051 [==============================] - 1s 314us/step - loss: 1372.8122 - mean_squared_error: 1372.8122 - val_loss: 1005.1369 - val_mean_squared_error: 1005.1369\n",
      "Epoch 23/30\n",
      "2051/2051 [==============================] - 1s 323us/step - loss: 1367.5431 - mean_squared_error: 1367.5431 - val_loss: 1005.6724 - val_mean_squared_error: 1005.6724\n",
      "Epoch 24/30\n",
      "2051/2051 [==============================] - 1s 316us/step - loss: 1364.4748 - mean_squared_error: 1364.4748 - val_loss: 1005.9284 - val_mean_squared_error: 1005.9284\n",
      "Epoch 25/30\n",
      "2051/2051 [==============================] - 1s 307us/step - loss: 1362.6013 - mean_squared_error: 1362.6013 - val_loss: 1006.7710 - val_mean_squared_error: 1006.7710\n",
      "Epoch 26/30\n",
      "2051/2051 [==============================] - 1s 310us/step - loss: 1361.3043 - mean_squared_error: 1361.3043 - val_loss: 1008.1988 - val_mean_squared_error: 1008.1988\n",
      "Epoch 27/30\n",
      "2051/2051 [==============================] - 1s 353us/step - loss: 1358.9207 - mean_squared_error: 1358.9207 - val_loss: 1005.8032 - val_mean_squared_error: 1005.8032\n",
      "Epoch 28/30\n",
      "2051/2051 [==============================] - 1s 311us/step - loss: 1355.8680 - mean_squared_error: 1355.8680 - val_loss: 1007.8423 - val_mean_squared_error: 1007.8423\n",
      "Epoch 29/30\n",
      "2051/2051 [==============================] - 1s 317us/step - loss: 1354.5584 - mean_squared_error: 1354.5584 - val_loss: 1011.2410 - val_mean_squared_error: 1011.2410\n",
      "Epoch 30/30\n",
      "2051/2051 [==============================] - 1s 330us/step - loss: 1353.7390 - mean_squared_error: 1353.7390 - val_loss: 1009.8339 - val_mean_squared_error: 1009.8339\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(20, input_dim=25, activation='relu'))\n",
    "model.add(Dense(10, input_dim=25, activation='relu'))\n",
    "model.add(Dense(1, activation = 'linear'))\n",
    "model.compile(optimizer= \"Adam\", loss='mse', metrics=['mse'])\n",
    "hist = model.fit(norm_dum, y, validation_split = 0.25, batch_size=10, \n",
    "                 epochs=30, verbose = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 303,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.956273762107024"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_nn = model.predict(norm_dum).reshape(-1)\n",
    "RMSE = np.sqrt(np.mean((pred_nn - y)**2))\n",
    "std - RMSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# summarize history for accuracy\n",
    "plt.plot(hist.history['loss'])\n",
    "plt.plot(hist.history['val_loss'])\n",
    "plt.title('model MSE')\n",
    "plt.ylabel('MSE')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'validation'], loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "kfold = KFold(n_splits=cv, shuffle=True, random_state=7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "norm_dum_copy = copy.deepcopy(norm_dum)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "y_copy = copy.deepcopy(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "cvscores = []\n",
    "cvscores_RMS = []\n",
    "nd_array = np.array(norm_dum)\n",
    "y_array = np.array(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "hidden": true
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [],
   "source": [
    "from keras import backend as K\n",
    "\n",
    "def coeff_determination(y_true, y_pred):\n",
    "    mean_y_true = K.mean(y_true)\n",
    "    SS_res = K.sum(K.square(y_true - y_pred)) \n",
    "    SS_tot = K.sum(K.square(y_true - mean_y_true))\n",
    "    return (1 - SS_res/SS_tot)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 5,
        "hidden": false,
        "row": 306,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "coeff_determination: -0.11730426201346325\n",
      "coeff_determination: 0.025980531297929106\n",
      "coeff_determination: -0.06799835541792083\n",
      "coeff_determination: -0.20565342920590982\n",
      "-0.11491858665167899 +/- 0.07971747621222382\n"
     ]
    }
   ],
   "source": [
    "for train, test in kfold.split(nd_array, y_array):\n",
    "    model = Sequential()\n",
    "    model.add(Dense(20, input_dim = 25, activation = 'relu'))\n",
    "    model.add(Dense(10, input_dim = 25, activation = 'relu'))\n",
    "    model.add(Dense(1, activation = 'linear'))\n",
    "    model.compile(optimizer = \"Adam\", loss = 'mse', metrics = [coeff_determination])\n",
    "    hist = model.fit(nd_array, y_array, batch_size = 10, \n",
    "                 epochs = 25, verbose = 0)\n",
    "    scores = model.evaluate(nd_array[test], y_array[test], verbose=0, batch_size = 10)\n",
    "    print('{}: {}'.format(model.metrics_names[1], scores[1]))\n",
    "    cvscores.append(scores[1])\n",
    "print('{} +/- {}'.format(np.mean(cvscores), np.std(cvscores)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 8,
        "height": 4,
        "hidden": false,
        "row": 307,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mean_squared_error: 29.665838537570735\n",
      "mean_squared_error: 30.6196659176269\n",
      "mean_squared_error: 30.334161569508204\n",
      "mean_squared_error: 48.641383244810065\n",
      "34.85530393523554 +/- 7.973885662588337\n"
     ]
    }
   ],
   "source": [
    "for train, test in kfold.split(nd_array, y_array):\n",
    "    model = Sequential()\n",
    "    model.add(Dense(8, input_dim=25, activation='relu'))\n",
    "    model.add(Dense(8, input_dim=25, activation='relu'))\n",
    "    model.add(Dense(1, activation = 'linear'))\n",
    "    model.compile(optimizer= \"Adam\", loss = 'mse', metrics=['mse'])\n",
    "    hist = model.fit(nd_array, y_array, batch_size=32, \n",
    "                 epochs=100, verbose = 0)\n",
    "    scores = model.evaluate(nd_array[test], y_array[test], verbose=0, batch_size = 10)\n",
    "    sc = np.sqrt(scores[1])\n",
    "    print('{}: {}'.format(model.metrics_names[1], sc))\n",
    "    cvscores_RMS.append(sc)\n",
    "RMS = np.mean(cvscores_RMS)\n",
    "print('{} +/- {}'.format(RMS, np.std(cvscores_RMS)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "extensions": {
     "jupyter_dashboards": {
      "version": 1,
      "views": {
       "grid_default": {
        "col": 0,
        "height": 4,
        "hidden": false,
        "row": 311,
        "width": 4
       },
       "report_default": {
        "hidden": false
       }
      }
     }
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.493632056493105"
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std - RMS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "extensions": {
   "jupyter_dashboards": {
    "activeView": "report_default",
    "version": 1,
    "views": {
     "grid_default": {
      "cellMargin": 10,
      "defaultCellHeight": 20,
      "maxColumns": 12,
      "name": "grid",
      "type": "grid"
     },
     "report_default": {
      "name": "report",
      "type": "report"
     }
    }
   }
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}