rbc_notebook_sprint2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1 style='color:#00868b'>Read, balance and clean dataset<span class=\"tocSkip\"></span></h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook produces 6 different CSV files. \n",
    "\n",
    "**CSVs**:\n",
    "\n",
    "There are **3 types CSV files (cp, cps, cpsi)** that each have **2 subtypes**, indicating whether or not the dataset has been balanced or not. In other words, there are **6 CSVs in total** that serve as the starting point for every unsupervised algorithm.\n",
    "\n",
    "*Imbalanced*\n",
    "* <code>corpus_sprint2_imbalanced_cp</code>: unbalanced, cleaned dataset consisting of the *consumer complaint narrative* (**c**) and the *product* (**p**).\n",
    "* <code>corpus_sprint2_imbalanced_cps</code>: unbalanced, cleaned dataset consisting of the *consumer complaint narrative* (**c**), the *product* (**p**) and the *sub-product* (**s**).\n",
    "* <code>corpus_sprint2_imbalanced_cpsi</code>: unbalanced, cleaned dataset consisting of the *consumer complaint narrative* (**c**), the *product* (**p**), the *sub-product* (**s**) and the *Issue* and *Sub-issue*, combined into one (**i**).\n",
    "\n",
    "*Balanced*\n",
    "* <code>corpus_sprint2_balanced_cp</code>: balanced, cleaned dataset consisting of the *consumer complaint narrative* (**c**) and the *product* (**p**).\n",
    "* <code>corpus_sprint2_balanced_cps</code>: balanced, cleaned dataset consisting of the *consumer complaint narrative* (**c**), the *product* (**p**) and the *sub-product* (**s**).\n",
    "* <code>corpus_sprint2_balanced_cpsi</code>: balanced, cleaned dataset consisting of the *consumer complaint narrative* (**c**), the *product* (**p**), the *sub-product* (**s**) and the *Issue* and *Sub-issue*, combined into one (**i**).\n",
    "\n",
    "We have chosen this approach because we are also interested in viewing what the imbalanced dataset yields as a result. A new column is added over time to gauge whether or not it improves the clustering algorithms or not."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Start"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Read dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'pd' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-1-5306a7565ae5>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0m_df\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread_csv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"complaints-2020-01-22_08_24.csv\"\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mencoding\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m\"utf-8\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m: name 'pd' is not defined"
     ]
    }
   ],
   "source": [
    "_df = pd.read_csv(\"complaints-2020-01-22_08_24.csv\", encoding=\"utf-8\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Date received</th>\n",
       "      <th>Product</th>\n",
       "      <th>Sub-product</th>\n",
       "      <th>Issue</th>\n",
       "      <th>Sub-issue</th>\n",
       "      <th>Consumer complaint narrative</th>\n",
       "      <th>Company public response</th>\n",
       "      <th>Company</th>\n",
       "      <th>State</th>\n",
       "      <th>ZIP code</th>\n",
       "      <th>Tags</th>\n",
       "      <th>Consumer consent provided?</th>\n",
       "      <th>Submitted via</th>\n",
       "      <th>Date sent to company</th>\n",
       "      <th>Company response to consumer</th>\n",
       "      <th>Timely response?</th>\n",
       "      <th>Consumer disputed?</th>\n",
       "      <th>Complaint ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>07/23/19</td>\n",
       "      <td>Credit reporting, credit repair services, or o...</td>\n",
       "      <td>Credit reporting</td>\n",
       "      <td>Credit monitoring or identity theft protection...</td>\n",
       "      <td>Problem canceling credit monitoring or identif...</td>\n",
       "      <td>I have complained many times that the credit r...</td>\n",
       "      <td>Company has responded to the consumer and the ...</td>\n",
       "      <td>Experian Information Solutions Inc.</td>\n",
       "      <td>CA</td>\n",
       "      <td>926XX</td>\n",
       "      <td>None</td>\n",
       "      <td>Consent provided</td>\n",
       "      <td>Web</td>\n",
       "      <td>07/23/19</td>\n",
       "      <td>Closed with explanation</td>\n",
       "      <td>Yes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3315279</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>07/26/19</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>I do not know</td>\n",
       "      <td>False statements or representation</td>\n",
       "      <td>Attempted to collect wrong amount</td>\n",
       "      <td>please review the current fraud account and al...</td>\n",
       "      <td>Company believes it acted appropriately as aut...</td>\n",
       "      <td>Ideal Collection Services, Inc.</td>\n",
       "      <td>FL</td>\n",
       "      <td>333XX</td>\n",
       "      <td>None</td>\n",
       "      <td>Consent provided</td>\n",
       "      <td>Web</td>\n",
       "      <td>07/26/19</td>\n",
       "      <td>Closed with explanation</td>\n",
       "      <td>Yes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3319487</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>06/03/19</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>I do not know</td>\n",
       "      <td>Attempts to collect debt not owed</td>\n",
       "      <td>Debt was paid</td>\n",
       "      <td>Called multiple times over the years for a deb...</td>\n",
       "      <td>None</td>\n",
       "      <td>ONEMAIN FINANCIAL HOLDINGS, LLC.</td>\n",
       "      <td>FL</td>\n",
       "      <td>327XX</td>\n",
       "      <td>None</td>\n",
       "      <td>Consent provided</td>\n",
       "      <td>Web</td>\n",
       "      <td>06/07/19</td>\n",
       "      <td>Closed with explanation</td>\n",
       "      <td>Yes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3262794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>07/03/19</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>Other debt</td>\n",
       "      <td>Attempts to collect debt not owed</td>\n",
       "      <td>Debt was result of identity theft</td>\n",
       "      <td>I sent in a letter to the company to have them...</td>\n",
       "      <td>None</td>\n",
       "      <td>Diversified Consultants, Inc.</td>\n",
       "      <td>VA</td>\n",
       "      <td>232XX</td>\n",
       "      <td>None</td>\n",
       "      <td>Consent provided</td>\n",
       "      <td>Web</td>\n",
       "      <td>07/03/19</td>\n",
       "      <td>Closed with explanation</td>\n",
       "      <td>Yes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3295208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>07/14/19</td>\n",
       "      <td>Credit reporting, credit repair services, or o...</td>\n",
       "      <td>Credit reporting</td>\n",
       "      <td>Improper use of your report</td>\n",
       "      <td>Received unsolicited financial product or insu...</td>\n",
       "      <td>On XX/XX/19 I applied for a Debt Relief Produc...</td>\n",
       "      <td>None</td>\n",
       "      <td>ClearOne Advantage, LLC</td>\n",
       "      <td>PA</td>\n",
       "      <td>191XX</td>\n",
       "      <td>Older American, Servicemember</td>\n",
       "      <td>Consent provided</td>\n",
       "      <td>Web</td>\n",
       "      <td>07/18/19</td>\n",
       "      <td>Closed with explanation</td>\n",
       "      <td>Yes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3306130</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Date received                                            Product  \\\n",
       "0      07/23/19  Credit reporting, credit repair services, or o...   \n",
       "1      07/26/19                                    Debt collection   \n",
       "2      06/03/19                                    Debt collection   \n",
       "3      07/03/19                                    Debt collection   \n",
       "4      07/14/19  Credit reporting, credit repair services, or o...   \n",
       "\n",
       "        Sub-product                                              Issue  \\\n",
       "0  Credit reporting  Credit monitoring or identity theft protection...   \n",
       "1     I do not know                 False statements or representation   \n",
       "2     I do not know                  Attempts to collect debt not owed   \n",
       "3        Other debt                  Attempts to collect debt not owed   \n",
       "4  Credit reporting                        Improper use of your report   \n",
       "\n",
       "                                           Sub-issue  \\\n",
       "0  Problem canceling credit monitoring or identif...   \n",
       "1                  Attempted to collect wrong amount   \n",
       "2                                      Debt was paid   \n",
       "3                  Debt was result of identity theft   \n",
       "4  Received unsolicited financial product or insu...   \n",
       "\n",
       "                        Consumer complaint narrative  \\\n",
       "0  I have complained many times that the credit r...   \n",
       "1  please review the current fraud account and al...   \n",
       "2  Called multiple times over the years for a deb...   \n",
       "3  I sent in a letter to the company to have them...   \n",
       "4  On XX/XX/19 I applied for a Debt Relief Produc...   \n",
       "\n",
       "                             Company public response  \\\n",
       "0  Company has responded to the consumer and the ...   \n",
       "1  Company believes it acted appropriately as aut...   \n",
       "2                                               None   \n",
       "3                                               None   \n",
       "4                                               None   \n",
       "\n",
       "                               Company State ZIP code  \\\n",
       "0  Experian Information Solutions Inc.    CA    926XX   \n",
       "1      Ideal Collection Services, Inc.    FL    333XX   \n",
       "2     ONEMAIN FINANCIAL HOLDINGS, LLC.    FL    327XX   \n",
       "3        Diversified Consultants, Inc.    VA    232XX   \n",
       "4              ClearOne Advantage, LLC    PA    191XX   \n",
       "\n",
       "                            Tags Consumer consent provided? Submitted via  \\\n",
       "0                           None           Consent provided           Web   \n",
       "1                           None           Consent provided           Web   \n",
       "2                           None           Consent provided           Web   \n",
       "3                           None           Consent provided           Web   \n",
       "4  Older American, Servicemember           Consent provided           Web   \n",
       "\n",
       "  Date sent to company Company response to consumer Timely response?  \\\n",
       "0             07/23/19      Closed with explanation              Yes   \n",
       "1             07/26/19      Closed with explanation              Yes   \n",
       "2             06/07/19      Closed with explanation              Yes   \n",
       "3             07/03/19      Closed with explanation              Yes   \n",
       "4             07/18/19      Closed with explanation              Yes   \n",
       "\n",
       "  Consumer disputed?  Complaint ID  \n",
       "0                NaN       3315279  \n",
       "1                NaN       3319487  \n",
       "2                NaN       3262794  \n",
       "3                NaN       3295208  \n",
       "4                NaN       3306130  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Select columns to keep"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Consumer complaint narrative + Product (cp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp = df.loc[:, ('Product', 'Consumer complaint narrative')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Product</th>\n",
       "      <th>Consumer complaint narrative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Credit reporting, credit repair services, or o...</td>\n",
       "      <td>I have complained many times that the credit r...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>please review the current fraud account and al...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>Called multiple times over the years for a deb...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>I sent in a letter to the company to have them...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Credit reporting, credit repair services, or o...</td>\n",
       "      <td>On XX/XX/19 I applied for a Debt Relief Produc...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             Product  \\\n",
       "0  Credit reporting, credit repair services, or o...   \n",
       "1                                    Debt collection   \n",
       "2                                    Debt collection   \n",
       "3                                    Debt collection   \n",
       "4  Credit reporting, credit repair services, or o...   \n",
       "\n",
       "                        Consumer complaint narrative  \n",
       "0  I have complained many times that the credit r...  \n",
       "1  please review the current fraud account and al...  \n",
       "2  Called multiple times over the years for a deb...  \n",
       "3  I sent in a letter to the company to have them...  \n",
       "4  On XX/XX/19 I applied for a Debt Relief Produc...  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Consumer complaint narrative + Product + Sub-product (cps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cps = df.loc[:, ('Product', 'Sub-product', 'Consumer complaint narrative')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Product</th>\n",
       "      <th>Sub-product</th>\n",
       "      <th>Consumer complaint narrative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Credit reporting, credit repair services, or o...</td>\n",
       "      <td>Credit reporting</td>\n",
       "      <td>I have complained many times that the credit r...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>I do not know</td>\n",
       "      <td>please review the current fraud account and al...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>I do not know</td>\n",
       "      <td>Called multiple times over the years for a deb...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>Other debt</td>\n",
       "      <td>I sent in a letter to the company to have them...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Credit reporting, credit repair services, or o...</td>\n",
       "      <td>Credit reporting</td>\n",
       "      <td>On XX/XX/19 I applied for a Debt Relief Produc...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             Product       Sub-product  \\\n",
       "0  Credit reporting, credit repair services, or o...  Credit reporting   \n",
       "1                                    Debt collection     I do not know   \n",
       "2                                    Debt collection     I do not know   \n",
       "3                                    Debt collection        Other debt   \n",
       "4  Credit reporting, credit repair services, or o...  Credit reporting   \n",
       "\n",
       "                        Consumer complaint narrative  \n",
       "0  I have complained many times that the credit r...  \n",
       "1  please review the current fraud account and al...  \n",
       "2  Called multiple times over the years for a deb...  \n",
       "3  I sent in a letter to the company to have them...  \n",
       "4  On XX/XX/19 I applied for a Debt Relief Produc...  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cps.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Consumer complaint narrative + Product + Issue + Sub-issue (cpsi)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cpsi = df.loc[:, ('Product', 'Sub-product', 'Issue', 'Sub-issue', 'Consumer complaint narrative')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": 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>Product</th>\n",
       "      <th>Sub-product</th>\n",
       "      <th>Issue</th>\n",
       "      <th>Sub-issue</th>\n",
       "      <th>Consumer complaint narrative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Credit reporting, credit repair services, or o...</td>\n",
       "      <td>Credit reporting</td>\n",
       "      <td>Credit monitoring or identity theft protection...</td>\n",
       "      <td>Problem canceling credit monitoring or identif...</td>\n",
       "      <td>I have complained many times that the credit r...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>I do not know</td>\n",
       "      <td>False statements or representation</td>\n",
       "      <td>Attempted to collect wrong amount</td>\n",
       "      <td>please review the current fraud account and al...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>I do not know</td>\n",
       "      <td>Attempts to collect debt not owed</td>\n",
       "      <td>Debt was paid</td>\n",
       "      <td>Called multiple times over the years for a deb...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Debt collection</td>\n",
       "      <td>Other debt</td>\n",
       "      <td>Attempts to collect debt not owed</td>\n",
       "      <td>Debt was result of identity theft</td>\n",
       "      <td>I sent in a letter to the company to have them...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Credit reporting, credit repair services, or o...</td>\n",
       "      <td>Credit reporting</td>\n",
       "      <td>Improper use of your report</td>\n",
       "      <td>Received unsolicited financial product or insu...</td>\n",
       "      <td>On XX/XX/19 I applied for a Debt Relief Produc...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             Product       Sub-product  \\\n",
       "0  Credit reporting, credit repair services, or o...  Credit reporting   \n",
       "1                                    Debt collection     I do not know   \n",
       "2                                    Debt collection     I do not know   \n",
       "3                                    Debt collection        Other debt   \n",
       "4  Credit reporting, credit repair services, or o...  Credit reporting   \n",
       "\n",
       "                                               Issue  \\\n",
       "0  Credit monitoring or identity theft protection...   \n",
       "1                 False statements or representation   \n",
       "2                  Attempts to collect debt not owed   \n",
       "3                  Attempts to collect debt not owed   \n",
       "4                        Improper use of your report   \n",
       "\n",
       "                                           Sub-issue  \\\n",
       "0  Problem canceling credit monitoring or identif...   \n",
       "1                  Attempted to collect wrong amount   \n",
       "2                                      Debt was paid   \n",
       "3                  Debt was result of identity theft   \n",
       "4  Received unsolicited financial product or insu...   \n",
       "\n",
       "                        Consumer complaint narrative  \n",
       "0  I have complained many times that the credit r...  \n",
       "1  please review the current fraud account and al...  \n",
       "2  Called multiple times over the years for a deb...  \n",
       "3  I sent in a letter to the company to have them...  \n",
       "4  On XX/XX/19 I applied for a Debt Relief Produc...  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cpsi.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "import string\n",
    "\n",
    "def clean_document(complaint):\n",
    "    # turn text to lowercase\n",
    "    complaint = complaint.lower()\n",
    "    # remove URLs\n",
    "    complaint = re.sub('https?://(?:[-\\w.]|(?:%[\\da-fA-F]{2}))+', '', complaint)\n",
    "    complaint = re.sub('https? ?: ?// ?(?:[-\\w.]|(?:%[\\da-fA-F]{2}))+', '', complaint)\n",
    "    # removed censored words\n",
    "    complaint = re.sub('[xxxx]{4}', '', complaint)\n",
    "    complaint = re.sub('[XXXX]{4}', '', complaint)\n",
    "    # remove  non-sensical characters\n",
    "    complaint = re.sub(\"[,#'-\\(\\):$;\\?%}{]\",' ', complaint)\n",
    "    complaint = re.sub('\\n', '', complaint)\n",
    "    complaint = re.sub('\\t', '', complaint)\n",
    "    # normalise spaces to just one space\n",
    "    complaint = re.sub(\" +\", \" \", complaint);\n",
    "    # remove normal dates and censored dates\n",
    "    complaint = re.sub('[\\dx]{1,2}/[\\dx]{1,2}/[\\dx]{2,4}', '', complaint)\n",
    "    #Get sentences from the tokenizer, remove the dot in each.\n",
    "    return complaint\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp[\"Consumer complaint narrative\"] = df_cp[\"Consumer complaint narrative\"].apply(clean_document)\n",
    "df_cps[\"Consumer complaint narrative\"] = df_cps[\"Consumer complaint narrative\"].apply(clean_document)\n",
    "df_cpsi[\"Consumer complaint narrative\"] = df_cpsi[\"Consumer complaint narrative\"].apply(clean_document)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Export to these 3 imbalanced datasets to csv for later use\n",
    "df_cp.to_csv(\"corpus_sprint3_imbalanced_cp.csv\", index=False)\n",
    "df_cps.to_csv(\"corpus_sprint3_imbalanced_cps.csv\", index=False)\n",
    "df_cpsi.to_csv(\"corpus_sprint3_imbalanced_cpsi.csv\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Balancing the dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now produce 3 more CSV files based on a balanced dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'categories')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "diff_products = df_cp.groupby(\"Product\")[\"Product\"].count().sort_values(ascending=False)\n",
    "fig = plt.figure()\n",
    "plt.rcParams['figure.figsize'] = [30,15]\n",
    "ax = fig.add_axes([0,0,0.4,0.5])\n",
    "\n",
    "ax.barh(diff_products.index,diff_products,color='teal')\n",
    "ax.invert_yaxis()\n",
    "plt.title('Complaints per Products')\n",
    "plt.xlabel('amount')\n",
    "plt.ylabel('categories')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Produce 3 dataframes that will contain the balanced dataset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp_balanced1 = df_cp.copy()\n",
    "df_cps_balanced1 = df_cps.copy()\n",
    "df_cpsi_balanced1 = df_cpsi.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Remove columns with the least complaints"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Balancing round 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp_balanced1 = df_cp_balanced1[df_cp_balanced1[\"Product\"] != \"Virtual currency\"]\n",
    "df_cp_balanced1 = df_cp_balanced1[df_cp_balanced1[\"Product\"] != \"Other financial service\"]\n",
    "df_cp_balanced1 = df_cp_balanced1[df_cp_balanced1[\"Product\"] != \"Prepaid card\"]\n",
    "df_cp_balanced1 = df_cp_balanced1[df_cp_balanced1[\"Product\"] != \"Money transfers\"]\n",
    "df_cp_balanced1 = df_cp_balanced1[df_cp_balanced1[\"Product\"] != \"Payday loan\"]\n",
    "\n",
    "df_cps_balanced1 = df_cps_balanced1[df_cps_balanced1[\"Product\"] != \"Virtual currency\"]\n",
    "df_cps_balanced1 = df_cps_balanced1[df_cps_balanced1[\"Product\"] != \"Other financial service\"]\n",
    "df_cps_balanced1 = df_cps_balanced1[df_cps_balanced1[\"Product\"] != \"Prepaid card\"]\n",
    "df_cps_balanced1 = df_cps_balanced1[df_cps_balanced1[\"Product\"] != \"Money transfers\"]\n",
    "df_cps_balanced1 = df_cps_balanced1[df_cps_balanced1[\"Product\"] != \"Payday loan\"]\n",
    "\n",
    "df_cpsi_balanced1 = df_cpsi_balanced1[df_cpsi_balanced1[\"Product\"] != \"Virtual currency\"]\n",
    "df_cpsi_balanced1 = df_cpsi_balanced1[df_cpsi_balanced1[\"Product\"] != \"Other financial service\"]\n",
    "df_cpsi_balanced1 = df_cpsi_balanced1[df_cpsi_balanced1[\"Product\"] != \"Prepaid card\"]\n",
    "df_cpsi_balanced1 = df_cpsi_balanced1[df_cpsi_balanced1[\"Product\"] != \"Money transfers\"]\n",
    "df_cpsi_balanced1 = df_cpsi_balanced1[df_cpsi_balanced1[\"Product\"] != \"Payday loan\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make a copy and iteratively drop the products with the most complaints, adding a random sample % of them back afterwards"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Balancing round 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp_balanced2 = df_cp_balanced1.copy()\n",
    "df_cps_balanced2 = df_cps_balanced1.copy()\n",
    "df_cpsi_balanced2 = df_cpsi_balanced1.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "indexNames = df_cp_balanced2[df_cp_balanced2[\"Product\"] == \"Mortgage\"].index\n",
    "df_cp_balanced2.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced2[df_cp_balanced2[\"Product\"] == \"Debt collection\"].index\n",
    "df_cp_balanced2.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced2[df_cp_balanced2[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cp_balanced2.drop(indexNames, inplace=True)\n",
    "\n",
    "indexNames = df_cps_balanced2[df_cps_balanced2[\"Product\"] == \"Mortgage\"].index\n",
    "df_cps_balanced2.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced2[df_cps_balanced2[\"Product\"] == \"Debt collection\"].index\n",
    "df_cps_balanced2.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced2[df_cps_balanced2[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cps_balanced2.drop(indexNames, inplace=True)\n",
    "\n",
    "indexNames = df_cpsi_balanced2[df_cpsi_balanced2[\"Product\"] == \"Mortgage\"].index\n",
    "df_cpsi_balanced2.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced2[df_cpsi_balanced2[\"Product\"] == \"Debt collection\"].index\n",
    "df_cpsi_balanced2.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced2[df_cpsi_balanced2[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cpsi_balanced2.drop(indexNames, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp_balanced2 = df_cp_balanced2.append(df_cp_balanced1.loc[df_cp_balanced1[\"Product\"] == 'Mortgage'].sample(frac=0.5))\n",
    "df_cp_balanced2 = df_cp_balanced2.append(df_cp_balanced1.loc[df_cp_balanced1[\"Product\"] == 'Debt collection'].sample(frac=0.4))\n",
    "df_cp_balanced2 = df_cp_balanced2.append(df_cp_balanced1.loc[df_cp_balanced1[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.3))\n",
    "\n",
    "df_cps_balanced2 = df_cps_balanced2.append(df_cps_balanced1.loc[df_cps_balanced1[\"Product\"] == 'Mortgage'].sample(frac=0.5))\n",
    "df_cps_balanced2 = df_cps_balanced2.append(df_cps_balanced1.loc[df_cps_balanced1[\"Product\"] == 'Debt collection'].sample(frac=0.4))\n",
    "df_cps_balanced2 = df_cps_balanced2.append(df_cps_balanced1.loc[df_cps_balanced1[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.3))\n",
    "\n",
    "df_cpsi_balanced2 = df_cpsi_balanced2.append(df_cpsi_balanced1.loc[df_cpsi_balanced1[\"Product\"] == 'Mortgage'].sample(frac=0.5))\n",
    "df_cpsi_balanced2 = df_cpsi_balanced2.append(df_cpsi_balanced1.loc[df_cpsi_balanced1[\"Product\"] == 'Debt collection'].sample(frac=0.4))\n",
    "df_cpsi_balanced2 = df_cpsi_balanced2.append(df_cpsi_balanced1.loc[df_cpsi_balanced1[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'categories')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "diff_products = df_cp_balanced2.groupby(\"Product\")[\"Product\"].count().sort_values(ascending=False)\n",
    "fig = plt.figure()\n",
    "plt.rcParams['figure.figsize'] = [30,15]\n",
    "ax = fig.add_axes([0,0,0.4,0.5])\n",
    "\n",
    "ax.barh(diff_products.index,diff_products,color='teal')\n",
    "ax.invert_yaxis()\n",
    "plt.title('Complaints per Products')\n",
    "plt.xlabel('amount')\n",
    "plt.ylabel('categories')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Balancing round 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp_balanced3 = df_cp_balanced2.copy()\n",
    "df_cps_balanced3 = df_cps_balanced2.copy()\n",
    "df_cpsi_balanced3 = df_cpsi_balanced2.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "indexNames = df_cp_balanced3[df_cp_balanced3[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cp_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced3[df_cp_balanced3[\"Product\"] == \"Debt collection\"].index\n",
    "df_cp_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced3[df_cp_balanced3[\"Product\"] == \"Credit reporting\"].index\n",
    "df_cp_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced3[df_cp_balanced3[\"Product\"] == \"Credit card or prepaid card\"].index\n",
    "df_cp_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced3[df_cp_balanced3[\"Product\"] == \"Mortgage\"].index\n",
    "df_cp_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced3[df_cp_balanced3[\"Product\"] == \"Student loan\"].index\n",
    "df_cp_balanced3.drop(indexNames, inplace=True)\n",
    "\n",
    "indexNames = df_cps_balanced3[df_cps_balanced3[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cps_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced3[df_cps_balanced3[\"Product\"] == \"Debt collection\"].index\n",
    "df_cps_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced3[df_cps_balanced3[\"Product\"] == \"Credit reporting\"].index\n",
    "df_cps_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced3[df_cps_balanced3[\"Product\"] == \"Credit card or prepaid card\"].index\n",
    "df_cps_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced3[df_cps_balanced3[\"Product\"] == \"Mortgage\"].index\n",
    "df_cps_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced3[df_cps_balanced3[\"Product\"] == \"Student loan\"].index\n",
    "df_cps_balanced3.drop(indexNames, inplace=True)\n",
    "\n",
    "indexNames = df_cpsi_balanced3[df_cpsi_balanced3[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cpsi_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced3[df_cpsi_balanced3[\"Product\"] == \"Debt collection\"].index\n",
    "df_cpsi_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced3[df_cpsi_balanced3[\"Product\"] == \"Credit reporting\"].index\n",
    "df_cpsi_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced3[df_cpsi_balanced3[\"Product\"] == \"Credit card or prepaid card\"].index\n",
    "df_cpsi_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced3[df_cpsi_balanced3[\"Product\"] == \"Mortgage\"].index\n",
    "df_cpsi_balanced3.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced3[df_cpsi_balanced3[\"Product\"] == \"Student loan\"].index\n",
    "df_cpsi_balanced3.drop(indexNames, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp_balanced3 = df_cp_balanced3.append(df_cp_balanced2.loc[df_cp_balanced2[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.5))\n",
    "df_cp_balanced3 = df_cp_balanced3.append(df_cp_balanced2.loc[df_cp_balanced2[\"Product\"] == 'Debt collection'].sample(frac=0.5))\n",
    "df_cp_balanced3 = df_cp_balanced3.append(df_cp_balanced2.loc[df_cp_balanced2[\"Product\"] == 'Credit reporting'].sample(frac=0.6))\n",
    "df_cp_balanced3 = df_cp_balanced3.append(df_cp_balanced2.loc[df_cp_balanced2[\"Product\"] == 'Credit card or prepaid card'].sample(frac=0.6))\n",
    "df_cp_balanced3 = df_cp_balanced3.append(df_cp_balanced2.loc[df_cp_balanced2[\"Product\"] == 'Mortgage'].sample(frac=0.6))\n",
    "df_cp_balanced3 = df_cp_balanced3.append(df_cp_balanced2.loc[df_cp_balanced2[\"Product\"] == 'Student loan'].sample(frac=0.85))\n",
    "\n",
    "df_cps_balanced3 = df_cps_balanced3.append(df_cps_balanced2.loc[df_cps_balanced2[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.5))\n",
    "df_cps_balanced3 = df_cps_balanced3.append(df_cps_balanced2.loc[df_cps_balanced2[\"Product\"] == 'Debt collection'].sample(frac=0.5))\n",
    "df_cps_balanced3 = df_cps_balanced3.append(df_cps_balanced2.loc[df_cps_balanced2[\"Product\"] == 'Credit reporting'].sample(frac=0.6))\n",
    "df_cps_balanced3 = df_cps_balanced3.append(df_cps_balanced2.loc[df_cps_balanced2[\"Product\"] == 'Credit card or prepaid card'].sample(frac=0.6))\n",
    "df_cps_balanced3 = df_cps_balanced3.append(df_cps_balanced2.loc[df_cps_balanced2[\"Product\"] == 'Mortgage'].sample(frac=0.6))\n",
    "df_cps_balanced3 = df_cps_balanced3.append(df_cps_balanced2.loc[df_cps_balanced2[\"Product\"] == 'Student loan'].sample(frac=0.85))\n",
    "\n",
    "df_cpsi_balanced3 = df_cpsi_balanced3.append(df_cpsi_balanced2.loc[df_cpsi_balanced2[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.5))\n",
    "df_cpsi_balanced3 = df_cpsi_balanced3.append(df_cpsi_balanced2.loc[df_cpsi_balanced2[\"Product\"] == 'Debt collection'].sample(frac=0.5))\n",
    "df_cpsi_balanced3 = df_cpsi_balanced3.append(df_cpsi_balanced2.loc[df_cpsi_balanced2[\"Product\"] == 'Credit reporting'].sample(frac=0.6))\n",
    "df_cpsi_balanced3 = df_cpsi_balanced3.append(df_cpsi_balanced2.loc[df_cpsi_balanced2[\"Product\"] == 'Credit card or prepaid card'].sample(frac=0.6))\n",
    "df_cpsi_balanced3 = df_cpsi_balanced3.append(df_cpsi_balanced2.loc[df_cpsi_balanced2[\"Product\"] == 'Mortgage'].sample(frac=0.6))\n",
    "df_cpsi_balanced3 = df_cpsi_balanced3.append(df_cpsi_balanced2.loc[df_cpsi_balanced2[\"Product\"] == 'Student loan'].sample(frac=0.85))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'categories')"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "diff_products = df_cp_balanced3.groupby(\"Product\")[\"Product\"].count().sort_values(ascending=False)\n",
    "fig = plt.figure()\n",
    "plt.rcParams['figure.figsize'] = [30,15]\n",
    "ax = fig.add_axes([0,0,0.4,0.5])\n",
    "\n",
    "ax.barh(diff_products.index,diff_products,color='teal')\n",
    "ax.invert_yaxis()\n",
    "plt.title('Complaints per Products')\n",
    "plt.xlabel('amount')\n",
    "plt.ylabel('categories')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Balancing round 4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp_balanced4 = df_cp_balanced3.copy()\n",
    "df_cps_balanced4 = df_cps_balanced3.copy()\n",
    "df_cpsi_balanced4 = df_cpsi_balanced3.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Debt collection\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Credit reporting\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Credit card or prepaid card\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Mortgage\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Student loan\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Credit card\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Checking or savings account\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cp_balanced4[df_cp_balanced4[\"Product\"] == \"Bank account or service\"].index\n",
    "df_cp_balanced4.drop(indexNames, inplace=True)\n",
    "\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Debt collection\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Credit reporting\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Credit card or prepaid card\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Mortgage\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Student loan\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Credit card\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Checking or savings account\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cps_balanced4[df_cps_balanced4[\"Product\"] == \"Bank account or service\"].index\n",
    "df_cps_balanced4.drop(indexNames, inplace=True)\n",
    "\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Credit reporting, credit repair services, or other personal consumer reports\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Debt collection\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Credit reporting\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Credit card or prepaid card\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Mortgage\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Student loan\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Credit card\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Checking or savings account\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)\n",
    "indexNames = df_cpsi_balanced4[df_cpsi_balanced4[\"Product\"] == \"Bank account or service\"].index\n",
    "df_cpsi_balanced4.drop(indexNames, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.44))\n",
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Debt collection'].sample(frac=0.44))\n",
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Credit reporting'].sample(frac=0.46))\n",
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Credit card or prepaid card'].sample(frac=0.44))\n",
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Mortgage'].sample(frac=0.45))\n",
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Student loan'].sample(frac=0.43))\n",
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Credit card'].sample(frac=0.5))\n",
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Checking or savings account'].sample(frac=0.46))\n",
    "df_cp_balanced4 = df_cp_balanced4.append(df_cp_balanced3.loc[df_cp_balanced3[\"Product\"] == 'Bank account or service'].sample(frac=0.6))\n",
    "\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.44))\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Debt collection'].sample(frac=0.44))\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Credit reporting'].sample(frac=0.46))\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Credit card or prepaid card'].sample(frac=0.44))\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Mortgage'].sample(frac=0.45))\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Student loan'].sample(frac=0.43))\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Credit card'].sample(frac=0.5))\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Checking or savings account'].sample(frac=0.46))\n",
    "df_cps_balanced4 = df_cps_balanced4.append(df_cps_balanced3.loc[df_cps_balanced3[\"Product\"] == 'Bank account or service'].sample(frac=0.6))\n",
    "\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Credit reporting, credit repair services, or other personal consumer reports'].sample(frac=0.44))\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Debt collection'].sample(frac=0.44))\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Credit reporting'].sample(frac=0.46))\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Credit card or prepaid card'].sample(frac=0.44))\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Mortgage'].sample(frac=0.45))\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Student loan'].sample(frac=0.43))\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Credit card'].sample(frac=0.5))\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Checking or savings account'].sample(frac=0.46))\n",
    "df_cpsi_balanced4 = df_cpsi_balanced4.append(df_cpsi_balanced3.loc[df_cpsi_balanced3[\"Product\"] == 'Bank account or service'].sample(frac=0.6))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'categories')"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "diff_products = df_cp_balanced4.groupby(\"Product\")[\"Product\"].count().sort_values(ascending=False)\n",
    "fig = plt.figure()\n",
    "plt.rcParams['figure.figsize'] = [30,15]\n",
    "ax = fig.add_axes([0,0,0.4,0.5])\n",
    "\n",
    "ax.barh(diff_products.index,diff_products,color='teal')\n",
    "ax.invert_yaxis()\n",
    "plt.title('Complaints per Products')\n",
    "plt.xlabel('amount')\n",
    "plt.ylabel('categories')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Export to these 3 more balanced datasets to csv for later use\n",
    "df_cp_balanced4.to_csv(\"corpus_sprint3_balanced_cp.csv\", index=False)\n",
    "df_cps_balanced4.to_csv(\"corpus_sprint3_balanced_cps.csv\", index=False)\n",
    "df_cpsi_balanced4.to_csv(\"corpus_sprint3_balanced_cpsi.csv\", index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}