diff --git a/lab-customer-analysis-round-1-checkpoint.ipynb b/lab-customer-analysis-round-1-checkpoint.ipynb
new file mode 100644
index 0000000..0fa7fff
--- /dev/null
+++ b/lab-customer-analysis-round-1-checkpoint.ipynb
@@ -0,0 +1,2543 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 101,
+   "id": "c43915d2",
+   "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>Customer</th>\n",
+       "      <th>ST</th>\n",
+       "      <th>GENDER</th>\n",
+       "      <th>Education</th>\n",
+       "      <th>Customer Lifetime Value</th>\n",
+       "      <th>Income</th>\n",
+       "      <th>Monthly Premium Auto</th>\n",
+       "      <th>Number of Open Complaints</th>\n",
+       "      <th>Policy Type</th>\n",
+       "      <th>Vehicle Class</th>\n",
+       "      <th>Total Claim Amount</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4003</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4004</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4005</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4006</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4007</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>4008 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     Customer          ST GENDER             Education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "4003      NaN         NaN    NaN                   NaN   \n",
+       "4004      NaN         NaN    NaN                   NaN   \n",
+       "4005      NaN         NaN    NaN                   NaN   \n",
+       "4006      NaN         NaN    NaN                   NaN   \n",
+       "4007      NaN         NaN    NaN                   NaN   \n",
+       "\n",
+       "     Customer Lifetime Value   Income  Monthly Premium Auto  \\\n",
+       "0                        NaN      0.0                1000.0   \n",
+       "1                 697953.59%      0.0                  94.0   \n",
+       "2                1288743.17%  48767.0                 108.0   \n",
+       "3                 764586.18%      0.0                 106.0   \n",
+       "4                 536307.65%  36357.0                  68.0   \n",
+       "...                      ...      ...                   ...   \n",
+       "4003                     NaN      NaN                   NaN   \n",
+       "4004                     NaN      NaN                   NaN   \n",
+       "4005                     NaN      NaN                   NaN   \n",
+       "4006                     NaN      NaN                   NaN   \n",
+       "4007                     NaN      NaN                   NaN   \n",
+       "\n",
+       "     Number of Open Complaints     Policy Type  Vehicle Class  \\\n",
+       "0                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "1                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "2                       1/0/00   Personal Auto   Two-Door Car   \n",
+       "3                       1/0/00  Corporate Auto            SUV   \n",
+       "4                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "...                        ...             ...            ...   \n",
+       "4003                       NaN             NaN            NaN   \n",
+       "4004                       NaN             NaN            NaN   \n",
+       "4005                       NaN             NaN            NaN   \n",
+       "4006                       NaN             NaN            NaN   \n",
+       "4007                       NaN             NaN            NaN   \n",
+       "\n",
+       "      Total Claim Amount  \n",
+       "0               2.704934  \n",
+       "1            1131.464935  \n",
+       "2             566.472247  \n",
+       "3             529.881344  \n",
+       "4              17.269323  \n",
+       "...                  ...  \n",
+       "4003                 NaN  \n",
+       "4004                 NaN  \n",
+       "4005                 NaN  \n",
+       "4006                 NaN  \n",
+       "4007                 NaN  \n",
+       "\n",
+       "[4008 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 101,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import pandas as pd\n",
+    "\n",
+    "tab1_data = pd.read_csv(r'C:\\Users\\david\\OneDrive\\Ambiente de Trabalho\\Iron Hack\\ficheiros para LAB PANDA 1\\file1.csv')\n",
+    "\n",
+    "tab1_data\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 102,
+   "id": "297404d5",
+   "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>Customer</th>\n",
+       "      <th>ST</th>\n",
+       "      <th>GENDER</th>\n",
+       "      <th>Education</th>\n",
+       "      <th>Customer Lifetime Value</th>\n",
+       "      <th>Income</th>\n",
+       "      <th>Monthly Premium Auto</th>\n",
+       "      <th>Number of Open Complaints</th>\n",
+       "      <th>Total Claim Amount</th>\n",
+       "      <th>Policy Type</th>\n",
+       "      <th>Vehicle Class</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>GS98873</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>323912.47%</td>\n",
+       "      <td>16061</td>\n",
+       "      <td>88</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>633.600000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>CW49887</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>462680.11%</td>\n",
+       "      <td>79487</td>\n",
+       "      <td>114</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>547.200000</td>\n",
+       "      <td>Special Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>MY31220</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>899704.02%</td>\n",
+       "      <td>54230</td>\n",
+       "      <td>112</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>537.600000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>UH35128</td>\n",
+       "      <td>Oregon</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2580706.30%</td>\n",
+       "      <td>71210</td>\n",
+       "      <td>214</td>\n",
+       "      <td>1/1/00</td>\n",
+       "      <td>1027.200000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Luxury Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>WH52799</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>380812.21%</td>\n",
+       "      <td>94903</td>\n",
+       "      <td>94</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>451.200000</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>991</th>\n",
+       "      <td>HV85198</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>847141.75%</td>\n",
+       "      <td>63513</td>\n",
+       "      <td>70</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>185.667213</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>992</th>\n",
+       "      <td>BS91566</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>543121.91%</td>\n",
+       "      <td>58161</td>\n",
+       "      <td>68</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>140.747286</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>993</th>\n",
+       "      <td>IL40123</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>568964.41%</td>\n",
+       "      <td>83640</td>\n",
+       "      <td>70</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>471.050488</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>994</th>\n",
+       "      <td>MY32149</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>368672.38%</td>\n",
+       "      <td>0</td>\n",
+       "      <td>96</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>28.460568</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>995</th>\n",
+       "      <td>SA91515</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>399258.39%</td>\n",
+       "      <td>0</td>\n",
+       "      <td>111</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>700.349052</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>996 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    Customer          ST GENDER Education Customer Lifetime Value  Income  \\\n",
+       "0    GS98873     Arizona      F  Bachelor              323912.47%   16061   \n",
+       "1    CW49887  California      F    Master              462680.11%   79487   \n",
+       "2    MY31220  California      F   College              899704.02%   54230   \n",
+       "3    UH35128      Oregon      F   College             2580706.30%   71210   \n",
+       "4    WH52799     Arizona      F   College              380812.21%   94903   \n",
+       "..       ...         ...    ...       ...                     ...     ...   \n",
+       "991  HV85198     Arizona      M    Master              847141.75%   63513   \n",
+       "992  BS91566     Arizona      F   College              543121.91%   58161   \n",
+       "993  IL40123      Nevada      F   College              568964.41%   83640   \n",
+       "994  MY32149  California      F    Master              368672.38%       0   \n",
+       "995  SA91515  California      M  Bachelor              399258.39%       0   \n",
+       "\n",
+       "     Monthly Premium Auto Number of Open Complaints  Total Claim Amount  \\\n",
+       "0                      88                    1/0/00          633.600000   \n",
+       "1                     114                    1/0/00          547.200000   \n",
+       "2                     112                    1/0/00          537.600000   \n",
+       "3                     214                    1/1/00         1027.200000   \n",
+       "4                      94                    1/0/00          451.200000   \n",
+       "..                    ...                       ...                 ...   \n",
+       "991                    70                    1/0/00          185.667213   \n",
+       "992                    68                    1/0/00          140.747286   \n",
+       "993                    70                    1/0/00          471.050488   \n",
+       "994                    96                    1/0/00           28.460568   \n",
+       "995                   111                    1/0/00          700.349052   \n",
+       "\n",
+       "        Policy Type  Vehicle Class  \n",
+       "0     Personal Auto  Four-Door Car  \n",
+       "1      Special Auto            SUV  \n",
+       "2     Personal Auto   Two-Door Car  \n",
+       "3     Personal Auto     Luxury Car  \n",
+       "4    Corporate Auto   Two-Door Car  \n",
+       "..              ...            ...  \n",
+       "991   Personal Auto  Four-Door Car  \n",
+       "992  Corporate Auto  Four-Door Car  \n",
+       "993  Corporate Auto   Two-Door Car  \n",
+       "994   Personal Auto   Two-Door Car  \n",
+       "995   Personal Auto            SUV  \n",
+       "\n",
+       "[996 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 102,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tab2_data = pd.read_csv(r'C:\\Users\\david\\OneDrive\\Ambiente de Trabalho\\Iron Hack\\ficheiros para LAB PANDA 1\\file2.csv')\n",
+    "\n",
+    "tab2_data\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 103,
+   "id": "5698462b",
+   "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>Customer</th>\n",
+       "      <th>State</th>\n",
+       "      <th>Customer Lifetime Value</th>\n",
+       "      <th>Education</th>\n",
+       "      <th>Gender</th>\n",
+       "      <th>Income</th>\n",
+       "      <th>Monthly Premium Auto</th>\n",
+       "      <th>Number of Open Complaints</th>\n",
+       "      <th>Policy Type</th>\n",
+       "      <th>Total Claim Amount</th>\n",
+       "      <th>Vehicle Class</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>SA25987</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>3479.137523</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>104</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>499.200000</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>TB86706</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>2502.637401</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>66</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>3.468912</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>ZL73902</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>3265.156348</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>F</td>\n",
+       "      <td>25820</td>\n",
+       "      <td>82</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>393.600000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>KX23516</td>\n",
+       "      <td>California</td>\n",
+       "      <td>4455.843406</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>F</td>\n",
+       "      <td>0</td>\n",
+       "      <td>121</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>699.615192</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>FN77294</td>\n",
+       "      <td>California</td>\n",
+       "      <td>7704.958480</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>M</td>\n",
+       "      <td>30366</td>\n",
+       "      <td>101</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>484.800000</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>California</td>\n",
+       "      <td>23405.987980</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>M</td>\n",
+       "      <td>71941</td>\n",
+       "      <td>73</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>California</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>College</td>\n",
+       "      <td>F</td>\n",
+       "      <td>21604</td>\n",
+       "      <td>79</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>California</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>85</td>\n",
+       "      <td>3</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>California</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>College</td>\n",
+       "      <td>M</td>\n",
+       "      <td>21941</td>\n",
+       "      <td>96</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>California</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>College</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>77</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>7070 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     Customer       State  Customer Lifetime Value             Education  \\\n",
+       "0     SA25987  Washington              3479.137523  High School or Below   \n",
+       "1     TB86706     Arizona              2502.637401                Master   \n",
+       "2     ZL73902      Nevada              3265.156348              Bachelor   \n",
+       "3     KX23516  California              4455.843406  High School or Below   \n",
+       "4     FN77294  California              7704.958480  High School or Below   \n",
+       "...       ...         ...                      ...                   ...   \n",
+       "7065  LA72316  California             23405.987980              Bachelor   \n",
+       "7066  PK87824  California              3096.511217               College   \n",
+       "7067  TD14365  California              8163.890428              Bachelor   \n",
+       "7068  UP19263  California              7524.442436               College   \n",
+       "7069  Y167826  California              2611.836866               College   \n",
+       "\n",
+       "     Gender  Income  Monthly Premium Auto  Number of Open Complaints  \\\n",
+       "0         M       0                   104                          0   \n",
+       "1         M       0                    66                          0   \n",
+       "2         F   25820                    82                          0   \n",
+       "3         F       0                   121                          0   \n",
+       "4         M   30366                   101                          2   \n",
+       "...     ...     ...                   ...                        ...   \n",
+       "7065      M   71941                    73                          0   \n",
+       "7066      F   21604                    79                          0   \n",
+       "7067      M       0                    85                          3   \n",
+       "7068      M   21941                    96                          0   \n",
+       "7069      M       0                    77                          0   \n",
+       "\n",
+       "         Policy Type  Total Claim Amount  Vehicle Class  \n",
+       "0      Personal Auto          499.200000   Two-Door Car  \n",
+       "1      Personal Auto            3.468912   Two-Door Car  \n",
+       "2      Personal Auto          393.600000  Four-Door Car  \n",
+       "3      Personal Auto          699.615192            SUV  \n",
+       "4      Personal Auto          484.800000            SUV  \n",
+       "...              ...                 ...            ...  \n",
+       "7065   Personal Auto          198.234764  Four-Door Car  \n",
+       "7066  Corporate Auto          379.200000  Four-Door Car  \n",
+       "7067  Corporate Auto          790.784983  Four-Door Car  \n",
+       "7068   Personal Auto          691.200000  Four-Door Car  \n",
+       "7069  Corporate Auto          369.600000   Two-Door Car  \n",
+       "\n",
+       "[7070 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 103,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tab3_data = pd.read_csv(r'C:\\Users\\david\\OneDrive\\Ambiente de Trabalho\\Iron Hack\\ficheiros para LAB PANDA 1\\file3.csv')\n",
+    "\n",
+    "tab3_data\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 104,
+   "id": "10a878b6",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(7070, 11)"
+      ]
+     },
+     "execution_count": 104,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    " # SHOW THE DATAFRAME'S SHAPE\n",
+    "\n",
+    "tab1_data.shape\n",
+    "tab2_data.shape\n",
+    "tab3_data.shape\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 105,
+   "id": "7eaccb15",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4003</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4004</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4005</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4006</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4007</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>4008 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer          st gender             education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "4003      NaN         NaN    NaN                   NaN   \n",
+       "4004      NaN         NaN    NaN                   NaN   \n",
+       "4005      NaN         NaN    NaN                   NaN   \n",
+       "4006      NaN         NaN    NaN                   NaN   \n",
+       "4007      NaN         NaN    NaN                   NaN   \n",
+       "\n",
+       "     customer_lifetime_value   income  monthly_premium_auto  \\\n",
+       "0                        NaN      0.0                1000.0   \n",
+       "1                 697953.59%      0.0                  94.0   \n",
+       "2                1288743.17%  48767.0                 108.0   \n",
+       "3                 764586.18%      0.0                 106.0   \n",
+       "4                 536307.65%  36357.0                  68.0   \n",
+       "...                      ...      ...                   ...   \n",
+       "4003                     NaN      NaN                   NaN   \n",
+       "4004                     NaN      NaN                   NaN   \n",
+       "4005                     NaN      NaN                   NaN   \n",
+       "4006                     NaN      NaN                   NaN   \n",
+       "4007                     NaN      NaN                   NaN   \n",
+       "\n",
+       "     number_of_open_complaints     policy_type  vehicle_class  \\\n",
+       "0                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "1                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "2                       1/0/00   Personal Auto   Two-Door Car   \n",
+       "3                       1/0/00  Corporate Auto            SUV   \n",
+       "4                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "...                        ...             ...            ...   \n",
+       "4003                       NaN             NaN            NaN   \n",
+       "4004                       NaN             NaN            NaN   \n",
+       "4005                       NaN             NaN            NaN   \n",
+       "4006                       NaN             NaN            NaN   \n",
+       "4007                       NaN             NaN            NaN   \n",
+       "\n",
+       "      total_claim_amount  \n",
+       "0               2.704934  \n",
+       "1            1131.464935  \n",
+       "2             566.472247  \n",
+       "3             529.881344  \n",
+       "4              17.269323  \n",
+       "...                  ...  \n",
+       "4003                 NaN  \n",
+       "4004                 NaN  \n",
+       "4005                 NaN  \n",
+       "4006                 NaN  \n",
+       "4007                 NaN  \n",
+       "\n",
+       "[4008 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 105,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    " # STANDARDIZE HEADER NAMES TABELA 1\n",
+    "    \n",
+    "cols = []\n",
+    "for i in range(len(tab1_data.columns)): \n",
+    "    cols.append(tab1_data.columns[i].lower().replace(' ', '_')) \n",
+    "tab1_data.columns = cols\n",
+    "tab1_data\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 106,
+   "id": "bb444eec",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>GS98873</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>323912.47%</td>\n",
+       "      <td>16061</td>\n",
+       "      <td>88</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>633.600000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>CW49887</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>462680.11%</td>\n",
+       "      <td>79487</td>\n",
+       "      <td>114</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>547.200000</td>\n",
+       "      <td>Special Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>MY31220</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>899704.02%</td>\n",
+       "      <td>54230</td>\n",
+       "      <td>112</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>537.600000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>UH35128</td>\n",
+       "      <td>Oregon</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2580706.30%</td>\n",
+       "      <td>71210</td>\n",
+       "      <td>214</td>\n",
+       "      <td>1/1/00</td>\n",
+       "      <td>1027.200000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Luxury Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>WH52799</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>380812.21%</td>\n",
+       "      <td>94903</td>\n",
+       "      <td>94</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>451.200000</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>991</th>\n",
+       "      <td>HV85198</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>847141.75%</td>\n",
+       "      <td>63513</td>\n",
+       "      <td>70</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>185.667213</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>992</th>\n",
+       "      <td>BS91566</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>543121.91%</td>\n",
+       "      <td>58161</td>\n",
+       "      <td>68</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>140.747286</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>993</th>\n",
+       "      <td>IL40123</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>568964.41%</td>\n",
+       "      <td>83640</td>\n",
+       "      <td>70</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>471.050488</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>994</th>\n",
+       "      <td>MY32149</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>368672.38%</td>\n",
+       "      <td>0</td>\n",
+       "      <td>96</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>28.460568</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>995</th>\n",
+       "      <td>SA91515</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>399258.39%</td>\n",
+       "      <td>0</td>\n",
+       "      <td>111</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>700.349052</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>996 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    customer          st gender education customer_lifetime_value  income  \\\n",
+       "0    GS98873     Arizona      F  Bachelor              323912.47%   16061   \n",
+       "1    CW49887  California      F    Master              462680.11%   79487   \n",
+       "2    MY31220  California      F   College              899704.02%   54230   \n",
+       "3    UH35128      Oregon      F   College             2580706.30%   71210   \n",
+       "4    WH52799     Arizona      F   College              380812.21%   94903   \n",
+       "..       ...         ...    ...       ...                     ...     ...   \n",
+       "991  HV85198     Arizona      M    Master              847141.75%   63513   \n",
+       "992  BS91566     Arizona      F   College              543121.91%   58161   \n",
+       "993  IL40123      Nevada      F   College              568964.41%   83640   \n",
+       "994  MY32149  California      F    Master              368672.38%       0   \n",
+       "995  SA91515  California      M  Bachelor              399258.39%       0   \n",
+       "\n",
+       "     monthly_premium_auto number_of_open_complaints  total_claim_amount  \\\n",
+       "0                      88                    1/0/00          633.600000   \n",
+       "1                     114                    1/0/00          547.200000   \n",
+       "2                     112                    1/0/00          537.600000   \n",
+       "3                     214                    1/1/00         1027.200000   \n",
+       "4                      94                    1/0/00          451.200000   \n",
+       "..                    ...                       ...                 ...   \n",
+       "991                    70                    1/0/00          185.667213   \n",
+       "992                    68                    1/0/00          140.747286   \n",
+       "993                    70                    1/0/00          471.050488   \n",
+       "994                    96                    1/0/00           28.460568   \n",
+       "995                   111                    1/0/00          700.349052   \n",
+       "\n",
+       "        policy_type  vehicle_class  \n",
+       "0     Personal Auto  Four-Door Car  \n",
+       "1      Special Auto            SUV  \n",
+       "2     Personal Auto   Two-Door Car  \n",
+       "3     Personal Auto     Luxury Car  \n",
+       "4    Corporate Auto   Two-Door Car  \n",
+       "..              ...            ...  \n",
+       "991   Personal Auto  Four-Door Car  \n",
+       "992  Corporate Auto  Four-Door Car  \n",
+       "993  Corporate Auto   Two-Door Car  \n",
+       "994   Personal Auto   Two-Door Car  \n",
+       "995   Personal Auto            SUV  \n",
+       "\n",
+       "[996 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 106,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# STANDARDIZE HEADER NAMES TABELA 2\n",
+    "    \n",
+    "cols = []\n",
+    "for i in range(len(tab2_data.columns)): \n",
+    "    cols.append(tab2_data.columns[i].lower().replace(' ', '_')) \n",
+    "tab2_data.columns = cols\n",
+    "tab2_data\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 107,
+   "id": "0fcf53af",
+   "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>customer</th>\n",
+       "      <th>state</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>education</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>SA25987</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>3479.137523</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>104</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>499.200000</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>TB86706</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>2502.637401</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>66</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>3.468912</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>ZL73902</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>3265.156348</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>F</td>\n",
+       "      <td>25820</td>\n",
+       "      <td>82</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>393.600000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>KX23516</td>\n",
+       "      <td>California</td>\n",
+       "      <td>4455.843406</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>F</td>\n",
+       "      <td>0</td>\n",
+       "      <td>121</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>699.615192</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>FN77294</td>\n",
+       "      <td>California</td>\n",
+       "      <td>7704.958480</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>M</td>\n",
+       "      <td>30366</td>\n",
+       "      <td>101</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>484.800000</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>California</td>\n",
+       "      <td>23405.987980</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>M</td>\n",
+       "      <td>71941</td>\n",
+       "      <td>73</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>California</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>College</td>\n",
+       "      <td>F</td>\n",
+       "      <td>21604</td>\n",
+       "      <td>79</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>California</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>85</td>\n",
+       "      <td>3</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>California</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>College</td>\n",
+       "      <td>M</td>\n",
+       "      <td>21941</td>\n",
+       "      <td>96</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>California</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>College</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>77</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>7070 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer       state  customer_lifetime_value             education  \\\n",
+       "0     SA25987  Washington              3479.137523  High School or Below   \n",
+       "1     TB86706     Arizona              2502.637401                Master   \n",
+       "2     ZL73902      Nevada              3265.156348              Bachelor   \n",
+       "3     KX23516  California              4455.843406  High School or Below   \n",
+       "4     FN77294  California              7704.958480  High School or Below   \n",
+       "...       ...         ...                      ...                   ...   \n",
+       "7065  LA72316  California             23405.987980              Bachelor   \n",
+       "7066  PK87824  California              3096.511217               College   \n",
+       "7067  TD14365  California              8163.890428              Bachelor   \n",
+       "7068  UP19263  California              7524.442436               College   \n",
+       "7069  Y167826  California              2611.836866               College   \n",
+       "\n",
+       "     gender  income  monthly_premium_auto  number_of_open_complaints  \\\n",
+       "0         M       0                   104                          0   \n",
+       "1         M       0                    66                          0   \n",
+       "2         F   25820                    82                          0   \n",
+       "3         F       0                   121                          0   \n",
+       "4         M   30366                   101                          2   \n",
+       "...     ...     ...                   ...                        ...   \n",
+       "7065      M   71941                    73                          0   \n",
+       "7066      F   21604                    79                          0   \n",
+       "7067      M       0                    85                          3   \n",
+       "7068      M   21941                    96                          0   \n",
+       "7069      M       0                    77                          0   \n",
+       "\n",
+       "         policy_type  total_claim_amount  vehicle_class  \n",
+       "0      Personal Auto          499.200000   Two-Door Car  \n",
+       "1      Personal Auto            3.468912   Two-Door Car  \n",
+       "2      Personal Auto          393.600000  Four-Door Car  \n",
+       "3      Personal Auto          699.615192            SUV  \n",
+       "4      Personal Auto          484.800000            SUV  \n",
+       "...              ...                 ...            ...  \n",
+       "7065   Personal Auto          198.234764  Four-Door Car  \n",
+       "7066  Corporate Auto          379.200000  Four-Door Car  \n",
+       "7067  Corporate Auto          790.784983  Four-Door Car  \n",
+       "7068   Personal Auto          691.200000  Four-Door Car  \n",
+       "7069  Corporate Auto          369.600000   Two-Door Car  \n",
+       "\n",
+       "[7070 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 107,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# STANDARDIZE HEADER NAMES TABELA 3\n",
+    "    \n",
+    "cols = []\n",
+    "for i in range(len(tab3_data.columns)): \n",
+    "    cols.append(tab3_data.columns[i].lower().replace(' ', '_')) \n",
+    "tab3_data.columns = cols\n",
+    "tab3_data\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 90,
+   "id": "552a0b18",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# REARRANGE THE COLUMNS IN THE DATAFRAME AS NEEDED\n",
+    "\n",
+    "\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 108,
+   "id": "a57f4cef",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>state</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>23405.98798</td>\n",
+       "      <td>71941.0</td>\n",
+       "      <td>73.0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>21604.0</td>\n",
+       "      <td>79.0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>85.0</td>\n",
+       "      <td>3</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>21941.0</td>\n",
+       "      <td>96.0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>77.0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>12074 rows × 12 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer          st gender             education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "7065  LA72316         NaN      M              Bachelor   \n",
+       "7066  PK87824         NaN      F               College   \n",
+       "7067  TD14365         NaN      M              Bachelor   \n",
+       "7068  UP19263         NaN      M               College   \n",
+       "7069  Y167826         NaN      M               College   \n",
+       "\n",
+       "     customer_lifetime_value   income  monthly_premium_auto  \\\n",
+       "0                        NaN      0.0                1000.0   \n",
+       "1                 697953.59%      0.0                  94.0   \n",
+       "2                1288743.17%  48767.0                 108.0   \n",
+       "3                 764586.18%      0.0                 106.0   \n",
+       "4                 536307.65%  36357.0                  68.0   \n",
+       "...                      ...      ...                   ...   \n",
+       "7065             23405.98798  71941.0                  73.0   \n",
+       "7066             3096.511217  21604.0                  79.0   \n",
+       "7067             8163.890428      0.0                  85.0   \n",
+       "7068             7524.442436  21941.0                  96.0   \n",
+       "7069             2611.836866      0.0                  77.0   \n",
+       "\n",
+       "     number_of_open_complaints     policy_type  vehicle_class  \\\n",
+       "0                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "1                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "2                       1/0/00   Personal Auto   Two-Door Car   \n",
+       "3                       1/0/00  Corporate Auto            SUV   \n",
+       "4                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "...                        ...             ...            ...   \n",
+       "7065                         0   Personal Auto  Four-Door Car   \n",
+       "7066                         0  Corporate Auto  Four-Door Car   \n",
+       "7067                         3  Corporate Auto  Four-Door Car   \n",
+       "7068                         0   Personal Auto  Four-Door Car   \n",
+       "7069                         0  Corporate Auto   Two-Door Car   \n",
+       "\n",
+       "      total_claim_amount       state  \n",
+       "0               2.704934         NaN  \n",
+       "1            1131.464935         NaN  \n",
+       "2             566.472247         NaN  \n",
+       "3             529.881344         NaN  \n",
+       "4              17.269323         NaN  \n",
+       "...                  ...         ...  \n",
+       "7065          198.234764  California  \n",
+       "7066          379.200000  California  \n",
+       "7067          790.784983  California  \n",
+       "7068          691.200000  California  \n",
+       "7069          369.600000  California  \n",
+       "\n",
+       "[12074 rows x 12 columns]"
+      ]
+     },
+     "execution_count": 108,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# CONCATENATE THE THREE DATAFRAMES\n",
+    "\n",
+    "datatotal_data = pd.concat([tab1_data,tab2_data,tab3_data], axis = 0) \n",
+    "\n",
+    "datatotal_data\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 92,
+   "id": "a2367b94",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "customer                      object\n",
+       "st                            object\n",
+       "gender                        object\n",
+       "education                     object\n",
+       "customer_lifetime_value       object\n",
+       "income                       float64\n",
+       "monthly_premium_auto         float64\n",
+       "number_of_open_complaints     object\n",
+       "policy_type                   object\n",
+       "vehicle_class                 object\n",
+       "total_claim_amount           float64\n",
+       "state                         object\n",
+       "dtype: object"
+      ]
+     },
+     "execution_count": 92,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# QUESTION: WICH COLUMNS ARE NUMERICAL?\n",
+    "# ANSWER: Income, monthly_premium_auto, total_claim_amount\n",
+    "\n",
+    "# QUESTION: WICH COLUMNS ARE CATEGORICAL?\n",
+    "# ANSWER: customer, st, gender, education, customer_lifetime_value, number_of_open_complaints, policy_type, vehicle_class,state\n",
+    "\n",
+    "# UNDERSTAND THE MEANING OF ALL COLUMNS\n",
+    "# ANSWER:\n",
+    "\n",
+    "datatotal_data.dtypes\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 93,
+   "id": "32de16d5",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>state</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>23405.98798</td>\n",
+       "      <td>71941.0</td>\n",
+       "      <td>73.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>21604.0</td>\n",
+       "      <td>79.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>85.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>21941.0</td>\n",
+       "      <td>96.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>77.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>12074 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer          st gender             education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "7065  LA72316         NaN      M              Bachelor   \n",
+       "7066  PK87824         NaN      F               College   \n",
+       "7067  TD14365         NaN      M              Bachelor   \n",
+       "7068  UP19263         NaN      M               College   \n",
+       "7069  Y167826         NaN      M               College   \n",
+       "\n",
+       "     customer_lifetime_value   income  monthly_premium_auto     policy_type  \\\n",
+       "0                        NaN      0.0                1000.0   Personal Auto   \n",
+       "1                 697953.59%      0.0                  94.0   Personal Auto   \n",
+       "2                1288743.17%  48767.0                 108.0   Personal Auto   \n",
+       "3                 764586.18%      0.0                 106.0  Corporate Auto   \n",
+       "4                 536307.65%  36357.0                  68.0   Personal Auto   \n",
+       "...                      ...      ...                   ...             ...   \n",
+       "7065             23405.98798  71941.0                  73.0   Personal Auto   \n",
+       "7066             3096.511217  21604.0                  79.0  Corporate Auto   \n",
+       "7067             8163.890428      0.0                  85.0  Corporate Auto   \n",
+       "7068             7524.442436  21941.0                  96.0   Personal Auto   \n",
+       "7069             2611.836866      0.0                  77.0  Corporate Auto   \n",
+       "\n",
+       "      vehicle_class  total_claim_amount       state  \n",
+       "0     Four-Door Car            2.704934         NaN  \n",
+       "1     Four-Door Car         1131.464935         NaN  \n",
+       "2      Two-Door Car          566.472247         NaN  \n",
+       "3               SUV          529.881344         NaN  \n",
+       "4     Four-Door Car           17.269323         NaN  \n",
+       "...             ...                 ...         ...  \n",
+       "7065  Four-Door Car          198.234764  California  \n",
+       "7066  Four-Door Car          379.200000  California  \n",
+       "7067  Four-Door Car          790.784983  California  \n",
+       "7068  Four-Door Car          691.200000  California  \n",
+       "7069   Two-Door Car          369.600000  California  \n",
+       "\n",
+       "[12074 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 93,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    " # DELETE THE COLUMN EDUCATION AND THE NUMBER OF OPEN COMPLAINTS FROM THE DATAFRAME\n",
+    "    \n",
+    "totaldata_data = datatotal_data.drop(['education'], axis = 1)\n",
+    "totaldata_data\n",
+    "\n",
+    "totaldata_data = datatotal_data.drop(['number_of_open_complaints'], axis = 1)\n",
+    "totaldata_data\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 96,
+   "id": "854a8c46",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>state</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>23405.98798</td>\n",
+       "      <td>71941.0</td>\n",
+       "      <td>73.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>21604.0</td>\n",
+       "      <td>79.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>85.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>21941.0</td>\n",
+       "      <td>96.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>77.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>9135 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer          st gender             education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "7065  LA72316         NaN      M              Bachelor   \n",
+       "7066  PK87824         NaN      F               College   \n",
+       "7067  TD14365         NaN      M              Bachelor   \n",
+       "7068  UP19263         NaN      M               College   \n",
+       "7069  Y167826         NaN      M               College   \n",
+       "\n",
+       "     customer_lifetime_value   income  monthly_premium_auto     policy_type  \\\n",
+       "0                        NaN      0.0                1000.0   Personal Auto   \n",
+       "1                 697953.59%      0.0                  94.0   Personal Auto   \n",
+       "2                1288743.17%  48767.0                 108.0   Personal Auto   \n",
+       "3                 764586.18%      0.0                 106.0  Corporate Auto   \n",
+       "4                 536307.65%  36357.0                  68.0   Personal Auto   \n",
+       "...                      ...      ...                   ...             ...   \n",
+       "7065             23405.98798  71941.0                  73.0   Personal Auto   \n",
+       "7066             3096.511217  21604.0                  79.0  Corporate Auto   \n",
+       "7067             8163.890428      0.0                  85.0  Corporate Auto   \n",
+       "7068             7524.442436  21941.0                  96.0   Personal Auto   \n",
+       "7069             2611.836866      0.0                  77.0  Corporate Auto   \n",
+       "\n",
+       "      vehicle_class  total_claim_amount       state  \n",
+       "0     Four-Door Car            2.704934         NaN  \n",
+       "1     Four-Door Car         1131.464935         NaN  \n",
+       "2      Two-Door Car          566.472247         NaN  \n",
+       "3               SUV          529.881344         NaN  \n",
+       "4     Four-Door Car           17.269323         NaN  \n",
+       "...             ...                 ...         ...  \n",
+       "7065  Four-Door Car          198.234764  California  \n",
+       "7066  Four-Door Car          379.200000  California  \n",
+       "7067  Four-Door Car          790.784983  California  \n",
+       "7068  Four-Door Car          691.200000  California  \n",
+       "7069   Two-Door Car          369.600000  California  \n",
+       "\n",
+       "[9135 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 96,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# CHECK FOR DUPLICATE ROWS IN THE DATA AND REMOVE IF ANY\n",
+    "\n",
+    "totaldata_data.duplicated() \n",
+    "\n",
+    "totaldata_data[totaldata_data.duplicated()] \n",
+    "\n",
+    "totaldata_data = totaldata_data.drop_duplicates()\n",
+    "\n",
+    "totaldata_data\n",
+    "\n",
+    "\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 111,
+   "id": "90edf4ee",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>state</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "Empty DataFrame\n",
+       "Columns: [customer, st, gender, education, customer_lifetime_value, income, monthly_premium_auto, number_of_open_complaints, policy_type, vehicle_class, total_claim_amount, state]\n",
+       "Index: []"
+      ]
+     },
+     "execution_count": 111,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    " # Filter out the data for customers who have an income of 0 or less.\n",
+    "    \n",
+    "datatotal_data[datatotal_data['income'] < 0]\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.11.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
diff --git a/lab-customer-analysis-round-1.ipynb b/lab-customer-analysis-round-1.ipynb
new file mode 100644
index 0000000..0fa7fff
--- /dev/null
+++ b/lab-customer-analysis-round-1.ipynb
@@ -0,0 +1,2543 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 101,
+   "id": "c43915d2",
+   "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>Customer</th>\n",
+       "      <th>ST</th>\n",
+       "      <th>GENDER</th>\n",
+       "      <th>Education</th>\n",
+       "      <th>Customer Lifetime Value</th>\n",
+       "      <th>Income</th>\n",
+       "      <th>Monthly Premium Auto</th>\n",
+       "      <th>Number of Open Complaints</th>\n",
+       "      <th>Policy Type</th>\n",
+       "      <th>Vehicle Class</th>\n",
+       "      <th>Total Claim Amount</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4003</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4004</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4005</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4006</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4007</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>4008 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     Customer          ST GENDER             Education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "4003      NaN         NaN    NaN                   NaN   \n",
+       "4004      NaN         NaN    NaN                   NaN   \n",
+       "4005      NaN         NaN    NaN                   NaN   \n",
+       "4006      NaN         NaN    NaN                   NaN   \n",
+       "4007      NaN         NaN    NaN                   NaN   \n",
+       "\n",
+       "     Customer Lifetime Value   Income  Monthly Premium Auto  \\\n",
+       "0                        NaN      0.0                1000.0   \n",
+       "1                 697953.59%      0.0                  94.0   \n",
+       "2                1288743.17%  48767.0                 108.0   \n",
+       "3                 764586.18%      0.0                 106.0   \n",
+       "4                 536307.65%  36357.0                  68.0   \n",
+       "...                      ...      ...                   ...   \n",
+       "4003                     NaN      NaN                   NaN   \n",
+       "4004                     NaN      NaN                   NaN   \n",
+       "4005                     NaN      NaN                   NaN   \n",
+       "4006                     NaN      NaN                   NaN   \n",
+       "4007                     NaN      NaN                   NaN   \n",
+       "\n",
+       "     Number of Open Complaints     Policy Type  Vehicle Class  \\\n",
+       "0                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "1                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "2                       1/0/00   Personal Auto   Two-Door Car   \n",
+       "3                       1/0/00  Corporate Auto            SUV   \n",
+       "4                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "...                        ...             ...            ...   \n",
+       "4003                       NaN             NaN            NaN   \n",
+       "4004                       NaN             NaN            NaN   \n",
+       "4005                       NaN             NaN            NaN   \n",
+       "4006                       NaN             NaN            NaN   \n",
+       "4007                       NaN             NaN            NaN   \n",
+       "\n",
+       "      Total Claim Amount  \n",
+       "0               2.704934  \n",
+       "1            1131.464935  \n",
+       "2             566.472247  \n",
+       "3             529.881344  \n",
+       "4              17.269323  \n",
+       "...                  ...  \n",
+       "4003                 NaN  \n",
+       "4004                 NaN  \n",
+       "4005                 NaN  \n",
+       "4006                 NaN  \n",
+       "4007                 NaN  \n",
+       "\n",
+       "[4008 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 101,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import pandas as pd\n",
+    "\n",
+    "tab1_data = pd.read_csv(r'C:\\Users\\david\\OneDrive\\Ambiente de Trabalho\\Iron Hack\\ficheiros para LAB PANDA 1\\file1.csv')\n",
+    "\n",
+    "tab1_data\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 102,
+   "id": "297404d5",
+   "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>Customer</th>\n",
+       "      <th>ST</th>\n",
+       "      <th>GENDER</th>\n",
+       "      <th>Education</th>\n",
+       "      <th>Customer Lifetime Value</th>\n",
+       "      <th>Income</th>\n",
+       "      <th>Monthly Premium Auto</th>\n",
+       "      <th>Number of Open Complaints</th>\n",
+       "      <th>Total Claim Amount</th>\n",
+       "      <th>Policy Type</th>\n",
+       "      <th>Vehicle Class</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>GS98873</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>323912.47%</td>\n",
+       "      <td>16061</td>\n",
+       "      <td>88</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>633.600000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>CW49887</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>462680.11%</td>\n",
+       "      <td>79487</td>\n",
+       "      <td>114</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>547.200000</td>\n",
+       "      <td>Special Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>MY31220</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>899704.02%</td>\n",
+       "      <td>54230</td>\n",
+       "      <td>112</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>537.600000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>UH35128</td>\n",
+       "      <td>Oregon</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2580706.30%</td>\n",
+       "      <td>71210</td>\n",
+       "      <td>214</td>\n",
+       "      <td>1/1/00</td>\n",
+       "      <td>1027.200000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Luxury Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>WH52799</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>380812.21%</td>\n",
+       "      <td>94903</td>\n",
+       "      <td>94</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>451.200000</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>991</th>\n",
+       "      <td>HV85198</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>847141.75%</td>\n",
+       "      <td>63513</td>\n",
+       "      <td>70</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>185.667213</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>992</th>\n",
+       "      <td>BS91566</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>543121.91%</td>\n",
+       "      <td>58161</td>\n",
+       "      <td>68</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>140.747286</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>993</th>\n",
+       "      <td>IL40123</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>568964.41%</td>\n",
+       "      <td>83640</td>\n",
+       "      <td>70</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>471.050488</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>994</th>\n",
+       "      <td>MY32149</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>368672.38%</td>\n",
+       "      <td>0</td>\n",
+       "      <td>96</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>28.460568</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>995</th>\n",
+       "      <td>SA91515</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>399258.39%</td>\n",
+       "      <td>0</td>\n",
+       "      <td>111</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>700.349052</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>996 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    Customer          ST GENDER Education Customer Lifetime Value  Income  \\\n",
+       "0    GS98873     Arizona      F  Bachelor              323912.47%   16061   \n",
+       "1    CW49887  California      F    Master              462680.11%   79487   \n",
+       "2    MY31220  California      F   College              899704.02%   54230   \n",
+       "3    UH35128      Oregon      F   College             2580706.30%   71210   \n",
+       "4    WH52799     Arizona      F   College              380812.21%   94903   \n",
+       "..       ...         ...    ...       ...                     ...     ...   \n",
+       "991  HV85198     Arizona      M    Master              847141.75%   63513   \n",
+       "992  BS91566     Arizona      F   College              543121.91%   58161   \n",
+       "993  IL40123      Nevada      F   College              568964.41%   83640   \n",
+       "994  MY32149  California      F    Master              368672.38%       0   \n",
+       "995  SA91515  California      M  Bachelor              399258.39%       0   \n",
+       "\n",
+       "     Monthly Premium Auto Number of Open Complaints  Total Claim Amount  \\\n",
+       "0                      88                    1/0/00          633.600000   \n",
+       "1                     114                    1/0/00          547.200000   \n",
+       "2                     112                    1/0/00          537.600000   \n",
+       "3                     214                    1/1/00         1027.200000   \n",
+       "4                      94                    1/0/00          451.200000   \n",
+       "..                    ...                       ...                 ...   \n",
+       "991                    70                    1/0/00          185.667213   \n",
+       "992                    68                    1/0/00          140.747286   \n",
+       "993                    70                    1/0/00          471.050488   \n",
+       "994                    96                    1/0/00           28.460568   \n",
+       "995                   111                    1/0/00          700.349052   \n",
+       "\n",
+       "        Policy Type  Vehicle Class  \n",
+       "0     Personal Auto  Four-Door Car  \n",
+       "1      Special Auto            SUV  \n",
+       "2     Personal Auto   Two-Door Car  \n",
+       "3     Personal Auto     Luxury Car  \n",
+       "4    Corporate Auto   Two-Door Car  \n",
+       "..              ...            ...  \n",
+       "991   Personal Auto  Four-Door Car  \n",
+       "992  Corporate Auto  Four-Door Car  \n",
+       "993  Corporate Auto   Two-Door Car  \n",
+       "994   Personal Auto   Two-Door Car  \n",
+       "995   Personal Auto            SUV  \n",
+       "\n",
+       "[996 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 102,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tab2_data = pd.read_csv(r'C:\\Users\\david\\OneDrive\\Ambiente de Trabalho\\Iron Hack\\ficheiros para LAB PANDA 1\\file2.csv')\n",
+    "\n",
+    "tab2_data\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 103,
+   "id": "5698462b",
+   "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>Customer</th>\n",
+       "      <th>State</th>\n",
+       "      <th>Customer Lifetime Value</th>\n",
+       "      <th>Education</th>\n",
+       "      <th>Gender</th>\n",
+       "      <th>Income</th>\n",
+       "      <th>Monthly Premium Auto</th>\n",
+       "      <th>Number of Open Complaints</th>\n",
+       "      <th>Policy Type</th>\n",
+       "      <th>Total Claim Amount</th>\n",
+       "      <th>Vehicle Class</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>SA25987</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>3479.137523</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>104</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>499.200000</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>TB86706</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>2502.637401</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>66</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>3.468912</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>ZL73902</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>3265.156348</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>F</td>\n",
+       "      <td>25820</td>\n",
+       "      <td>82</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>393.600000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>KX23516</td>\n",
+       "      <td>California</td>\n",
+       "      <td>4455.843406</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>F</td>\n",
+       "      <td>0</td>\n",
+       "      <td>121</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>699.615192</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>FN77294</td>\n",
+       "      <td>California</td>\n",
+       "      <td>7704.958480</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>M</td>\n",
+       "      <td>30366</td>\n",
+       "      <td>101</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>484.800000</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>California</td>\n",
+       "      <td>23405.987980</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>M</td>\n",
+       "      <td>71941</td>\n",
+       "      <td>73</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>California</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>College</td>\n",
+       "      <td>F</td>\n",
+       "      <td>21604</td>\n",
+       "      <td>79</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>California</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>85</td>\n",
+       "      <td>3</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>California</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>College</td>\n",
+       "      <td>M</td>\n",
+       "      <td>21941</td>\n",
+       "      <td>96</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>California</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>College</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>77</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>7070 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     Customer       State  Customer Lifetime Value             Education  \\\n",
+       "0     SA25987  Washington              3479.137523  High School or Below   \n",
+       "1     TB86706     Arizona              2502.637401                Master   \n",
+       "2     ZL73902      Nevada              3265.156348              Bachelor   \n",
+       "3     KX23516  California              4455.843406  High School or Below   \n",
+       "4     FN77294  California              7704.958480  High School or Below   \n",
+       "...       ...         ...                      ...                   ...   \n",
+       "7065  LA72316  California             23405.987980              Bachelor   \n",
+       "7066  PK87824  California              3096.511217               College   \n",
+       "7067  TD14365  California              8163.890428              Bachelor   \n",
+       "7068  UP19263  California              7524.442436               College   \n",
+       "7069  Y167826  California              2611.836866               College   \n",
+       "\n",
+       "     Gender  Income  Monthly Premium Auto  Number of Open Complaints  \\\n",
+       "0         M       0                   104                          0   \n",
+       "1         M       0                    66                          0   \n",
+       "2         F   25820                    82                          0   \n",
+       "3         F       0                   121                          0   \n",
+       "4         M   30366                   101                          2   \n",
+       "...     ...     ...                   ...                        ...   \n",
+       "7065      M   71941                    73                          0   \n",
+       "7066      F   21604                    79                          0   \n",
+       "7067      M       0                    85                          3   \n",
+       "7068      M   21941                    96                          0   \n",
+       "7069      M       0                    77                          0   \n",
+       "\n",
+       "         Policy Type  Total Claim Amount  Vehicle Class  \n",
+       "0      Personal Auto          499.200000   Two-Door Car  \n",
+       "1      Personal Auto            3.468912   Two-Door Car  \n",
+       "2      Personal Auto          393.600000  Four-Door Car  \n",
+       "3      Personal Auto          699.615192            SUV  \n",
+       "4      Personal Auto          484.800000            SUV  \n",
+       "...              ...                 ...            ...  \n",
+       "7065   Personal Auto          198.234764  Four-Door Car  \n",
+       "7066  Corporate Auto          379.200000  Four-Door Car  \n",
+       "7067  Corporate Auto          790.784983  Four-Door Car  \n",
+       "7068   Personal Auto          691.200000  Four-Door Car  \n",
+       "7069  Corporate Auto          369.600000   Two-Door Car  \n",
+       "\n",
+       "[7070 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 103,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tab3_data = pd.read_csv(r'C:\\Users\\david\\OneDrive\\Ambiente de Trabalho\\Iron Hack\\ficheiros para LAB PANDA 1\\file3.csv')\n",
+    "\n",
+    "tab3_data\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 104,
+   "id": "10a878b6",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(7070, 11)"
+      ]
+     },
+     "execution_count": 104,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    " # SHOW THE DATAFRAME'S SHAPE\n",
+    "\n",
+    "tab1_data.shape\n",
+    "tab2_data.shape\n",
+    "tab3_data.shape\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 105,
+   "id": "7eaccb15",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4003</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4004</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4005</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4006</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4007</th>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>4008 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer          st gender             education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "4003      NaN         NaN    NaN                   NaN   \n",
+       "4004      NaN         NaN    NaN                   NaN   \n",
+       "4005      NaN         NaN    NaN                   NaN   \n",
+       "4006      NaN         NaN    NaN                   NaN   \n",
+       "4007      NaN         NaN    NaN                   NaN   \n",
+       "\n",
+       "     customer_lifetime_value   income  monthly_premium_auto  \\\n",
+       "0                        NaN      0.0                1000.0   \n",
+       "1                 697953.59%      0.0                  94.0   \n",
+       "2                1288743.17%  48767.0                 108.0   \n",
+       "3                 764586.18%      0.0                 106.0   \n",
+       "4                 536307.65%  36357.0                  68.0   \n",
+       "...                      ...      ...                   ...   \n",
+       "4003                     NaN      NaN                   NaN   \n",
+       "4004                     NaN      NaN                   NaN   \n",
+       "4005                     NaN      NaN                   NaN   \n",
+       "4006                     NaN      NaN                   NaN   \n",
+       "4007                     NaN      NaN                   NaN   \n",
+       "\n",
+       "     number_of_open_complaints     policy_type  vehicle_class  \\\n",
+       "0                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "1                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "2                       1/0/00   Personal Auto   Two-Door Car   \n",
+       "3                       1/0/00  Corporate Auto            SUV   \n",
+       "4                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "...                        ...             ...            ...   \n",
+       "4003                       NaN             NaN            NaN   \n",
+       "4004                       NaN             NaN            NaN   \n",
+       "4005                       NaN             NaN            NaN   \n",
+       "4006                       NaN             NaN            NaN   \n",
+       "4007                       NaN             NaN            NaN   \n",
+       "\n",
+       "      total_claim_amount  \n",
+       "0               2.704934  \n",
+       "1            1131.464935  \n",
+       "2             566.472247  \n",
+       "3             529.881344  \n",
+       "4              17.269323  \n",
+       "...                  ...  \n",
+       "4003                 NaN  \n",
+       "4004                 NaN  \n",
+       "4005                 NaN  \n",
+       "4006                 NaN  \n",
+       "4007                 NaN  \n",
+       "\n",
+       "[4008 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 105,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    " # STANDARDIZE HEADER NAMES TABELA 1\n",
+    "    \n",
+    "cols = []\n",
+    "for i in range(len(tab1_data.columns)): \n",
+    "    cols.append(tab1_data.columns[i].lower().replace(' ', '_')) \n",
+    "tab1_data.columns = cols\n",
+    "tab1_data\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 106,
+   "id": "bb444eec",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>GS98873</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>323912.47%</td>\n",
+       "      <td>16061</td>\n",
+       "      <td>88</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>633.600000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>CW49887</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>462680.11%</td>\n",
+       "      <td>79487</td>\n",
+       "      <td>114</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>547.200000</td>\n",
+       "      <td>Special Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>MY31220</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>899704.02%</td>\n",
+       "      <td>54230</td>\n",
+       "      <td>112</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>537.600000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>UH35128</td>\n",
+       "      <td>Oregon</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2580706.30%</td>\n",
+       "      <td>71210</td>\n",
+       "      <td>214</td>\n",
+       "      <td>1/1/00</td>\n",
+       "      <td>1027.200000</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Luxury Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>WH52799</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>380812.21%</td>\n",
+       "      <td>94903</td>\n",
+       "      <td>94</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>451.200000</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>991</th>\n",
+       "      <td>HV85198</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>847141.75%</td>\n",
+       "      <td>63513</td>\n",
+       "      <td>70</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>185.667213</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>992</th>\n",
+       "      <td>BS91566</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>543121.91%</td>\n",
+       "      <td>58161</td>\n",
+       "      <td>68</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>140.747286</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>993</th>\n",
+       "      <td>IL40123</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>568964.41%</td>\n",
+       "      <td>83640</td>\n",
+       "      <td>70</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>471.050488</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>994</th>\n",
+       "      <td>MY32149</td>\n",
+       "      <td>California</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>368672.38%</td>\n",
+       "      <td>0</td>\n",
+       "      <td>96</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>28.460568</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>995</th>\n",
+       "      <td>SA91515</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>399258.39%</td>\n",
+       "      <td>0</td>\n",
+       "      <td>111</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>700.349052</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>996 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    customer          st gender education customer_lifetime_value  income  \\\n",
+       "0    GS98873     Arizona      F  Bachelor              323912.47%   16061   \n",
+       "1    CW49887  California      F    Master              462680.11%   79487   \n",
+       "2    MY31220  California      F   College              899704.02%   54230   \n",
+       "3    UH35128      Oregon      F   College             2580706.30%   71210   \n",
+       "4    WH52799     Arizona      F   College              380812.21%   94903   \n",
+       "..       ...         ...    ...       ...                     ...     ...   \n",
+       "991  HV85198     Arizona      M    Master              847141.75%   63513   \n",
+       "992  BS91566     Arizona      F   College              543121.91%   58161   \n",
+       "993  IL40123      Nevada      F   College              568964.41%   83640   \n",
+       "994  MY32149  California      F    Master              368672.38%       0   \n",
+       "995  SA91515  California      M  Bachelor              399258.39%       0   \n",
+       "\n",
+       "     monthly_premium_auto number_of_open_complaints  total_claim_amount  \\\n",
+       "0                      88                    1/0/00          633.600000   \n",
+       "1                     114                    1/0/00          547.200000   \n",
+       "2                     112                    1/0/00          537.600000   \n",
+       "3                     214                    1/1/00         1027.200000   \n",
+       "4                      94                    1/0/00          451.200000   \n",
+       "..                    ...                       ...                 ...   \n",
+       "991                    70                    1/0/00          185.667213   \n",
+       "992                    68                    1/0/00          140.747286   \n",
+       "993                    70                    1/0/00          471.050488   \n",
+       "994                    96                    1/0/00           28.460568   \n",
+       "995                   111                    1/0/00          700.349052   \n",
+       "\n",
+       "        policy_type  vehicle_class  \n",
+       "0     Personal Auto  Four-Door Car  \n",
+       "1      Special Auto            SUV  \n",
+       "2     Personal Auto   Two-Door Car  \n",
+       "3     Personal Auto     Luxury Car  \n",
+       "4    Corporate Auto   Two-Door Car  \n",
+       "..              ...            ...  \n",
+       "991   Personal Auto  Four-Door Car  \n",
+       "992  Corporate Auto  Four-Door Car  \n",
+       "993  Corporate Auto   Two-Door Car  \n",
+       "994   Personal Auto   Two-Door Car  \n",
+       "995   Personal Auto            SUV  \n",
+       "\n",
+       "[996 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 106,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# STANDARDIZE HEADER NAMES TABELA 2\n",
+    "    \n",
+    "cols = []\n",
+    "for i in range(len(tab2_data.columns)): \n",
+    "    cols.append(tab2_data.columns[i].lower().replace(' ', '_')) \n",
+    "tab2_data.columns = cols\n",
+    "tab2_data\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 107,
+   "id": "0fcf53af",
+   "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>customer</th>\n",
+       "      <th>state</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>education</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>SA25987</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>3479.137523</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>104</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>499.200000</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>TB86706</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>2502.637401</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>66</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>3.468912</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>ZL73902</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>3265.156348</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>F</td>\n",
+       "      <td>25820</td>\n",
+       "      <td>82</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>393.600000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>KX23516</td>\n",
+       "      <td>California</td>\n",
+       "      <td>4455.843406</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>F</td>\n",
+       "      <td>0</td>\n",
+       "      <td>121</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>699.615192</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>FN77294</td>\n",
+       "      <td>California</td>\n",
+       "      <td>7704.958480</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>M</td>\n",
+       "      <td>30366</td>\n",
+       "      <td>101</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>484.800000</td>\n",
+       "      <td>SUV</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>California</td>\n",
+       "      <td>23405.987980</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>M</td>\n",
+       "      <td>71941</td>\n",
+       "      <td>73</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>California</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>College</td>\n",
+       "      <td>F</td>\n",
+       "      <td>21604</td>\n",
+       "      <td>79</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>California</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>85</td>\n",
+       "      <td>3</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>California</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>College</td>\n",
+       "      <td>M</td>\n",
+       "      <td>21941</td>\n",
+       "      <td>96</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>California</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>College</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>77</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>7070 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer       state  customer_lifetime_value             education  \\\n",
+       "0     SA25987  Washington              3479.137523  High School or Below   \n",
+       "1     TB86706     Arizona              2502.637401                Master   \n",
+       "2     ZL73902      Nevada              3265.156348              Bachelor   \n",
+       "3     KX23516  California              4455.843406  High School or Below   \n",
+       "4     FN77294  California              7704.958480  High School or Below   \n",
+       "...       ...         ...                      ...                   ...   \n",
+       "7065  LA72316  California             23405.987980              Bachelor   \n",
+       "7066  PK87824  California              3096.511217               College   \n",
+       "7067  TD14365  California              8163.890428              Bachelor   \n",
+       "7068  UP19263  California              7524.442436               College   \n",
+       "7069  Y167826  California              2611.836866               College   \n",
+       "\n",
+       "     gender  income  monthly_premium_auto  number_of_open_complaints  \\\n",
+       "0         M       0                   104                          0   \n",
+       "1         M       0                    66                          0   \n",
+       "2         F   25820                    82                          0   \n",
+       "3         F       0                   121                          0   \n",
+       "4         M   30366                   101                          2   \n",
+       "...     ...     ...                   ...                        ...   \n",
+       "7065      M   71941                    73                          0   \n",
+       "7066      F   21604                    79                          0   \n",
+       "7067      M       0                    85                          3   \n",
+       "7068      M   21941                    96                          0   \n",
+       "7069      M       0                    77                          0   \n",
+       "\n",
+       "         policy_type  total_claim_amount  vehicle_class  \n",
+       "0      Personal Auto          499.200000   Two-Door Car  \n",
+       "1      Personal Auto            3.468912   Two-Door Car  \n",
+       "2      Personal Auto          393.600000  Four-Door Car  \n",
+       "3      Personal Auto          699.615192            SUV  \n",
+       "4      Personal Auto          484.800000            SUV  \n",
+       "...              ...                 ...            ...  \n",
+       "7065   Personal Auto          198.234764  Four-Door Car  \n",
+       "7066  Corporate Auto          379.200000  Four-Door Car  \n",
+       "7067  Corporate Auto          790.784983  Four-Door Car  \n",
+       "7068   Personal Auto          691.200000  Four-Door Car  \n",
+       "7069  Corporate Auto          369.600000   Two-Door Car  \n",
+       "\n",
+       "[7070 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 107,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# STANDARDIZE HEADER NAMES TABELA 3\n",
+    "    \n",
+    "cols = []\n",
+    "for i in range(len(tab3_data.columns)): \n",
+    "    cols.append(tab3_data.columns[i].lower().replace(' ', '_')) \n",
+    "tab3_data.columns = cols\n",
+    "tab3_data\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 90,
+   "id": "552a0b18",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# REARRANGE THE COLUMNS IN THE DATAFRAME AS NEEDED\n",
+    "\n",
+    "\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 108,
+   "id": "a57f4cef",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>state</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>1/0/00</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>23405.98798</td>\n",
+       "      <td>71941.0</td>\n",
+       "      <td>73.0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>21604.0</td>\n",
+       "      <td>79.0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>85.0</td>\n",
+       "      <td>3</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>21941.0</td>\n",
+       "      <td>96.0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>77.0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>12074 rows × 12 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer          st gender             education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "7065  LA72316         NaN      M              Bachelor   \n",
+       "7066  PK87824         NaN      F               College   \n",
+       "7067  TD14365         NaN      M              Bachelor   \n",
+       "7068  UP19263         NaN      M               College   \n",
+       "7069  Y167826         NaN      M               College   \n",
+       "\n",
+       "     customer_lifetime_value   income  monthly_premium_auto  \\\n",
+       "0                        NaN      0.0                1000.0   \n",
+       "1                 697953.59%      0.0                  94.0   \n",
+       "2                1288743.17%  48767.0                 108.0   \n",
+       "3                 764586.18%      0.0                 106.0   \n",
+       "4                 536307.65%  36357.0                  68.0   \n",
+       "...                      ...      ...                   ...   \n",
+       "7065             23405.98798  71941.0                  73.0   \n",
+       "7066             3096.511217  21604.0                  79.0   \n",
+       "7067             8163.890428      0.0                  85.0   \n",
+       "7068             7524.442436  21941.0                  96.0   \n",
+       "7069             2611.836866      0.0                  77.0   \n",
+       "\n",
+       "     number_of_open_complaints     policy_type  vehicle_class  \\\n",
+       "0                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "1                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "2                       1/0/00   Personal Auto   Two-Door Car   \n",
+       "3                       1/0/00  Corporate Auto            SUV   \n",
+       "4                       1/0/00   Personal Auto  Four-Door Car   \n",
+       "...                        ...             ...            ...   \n",
+       "7065                         0   Personal Auto  Four-Door Car   \n",
+       "7066                         0  Corporate Auto  Four-Door Car   \n",
+       "7067                         3  Corporate Auto  Four-Door Car   \n",
+       "7068                         0   Personal Auto  Four-Door Car   \n",
+       "7069                         0  Corporate Auto   Two-Door Car   \n",
+       "\n",
+       "      total_claim_amount       state  \n",
+       "0               2.704934         NaN  \n",
+       "1            1131.464935         NaN  \n",
+       "2             566.472247         NaN  \n",
+       "3             529.881344         NaN  \n",
+       "4              17.269323         NaN  \n",
+       "...                  ...         ...  \n",
+       "7065          198.234764  California  \n",
+       "7066          379.200000  California  \n",
+       "7067          790.784983  California  \n",
+       "7068          691.200000  California  \n",
+       "7069          369.600000  California  \n",
+       "\n",
+       "[12074 rows x 12 columns]"
+      ]
+     },
+     "execution_count": 108,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# CONCATENATE THE THREE DATAFRAMES\n",
+    "\n",
+    "datatotal_data = pd.concat([tab1_data,tab2_data,tab3_data], axis = 0) \n",
+    "\n",
+    "datatotal_data\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 92,
+   "id": "a2367b94",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "customer                      object\n",
+       "st                            object\n",
+       "gender                        object\n",
+       "education                     object\n",
+       "customer_lifetime_value       object\n",
+       "income                       float64\n",
+       "monthly_premium_auto         float64\n",
+       "number_of_open_complaints     object\n",
+       "policy_type                   object\n",
+       "vehicle_class                 object\n",
+       "total_claim_amount           float64\n",
+       "state                         object\n",
+       "dtype: object"
+      ]
+     },
+     "execution_count": 92,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# QUESTION: WICH COLUMNS ARE NUMERICAL?\n",
+    "# ANSWER: Income, monthly_premium_auto, total_claim_amount\n",
+    "\n",
+    "# QUESTION: WICH COLUMNS ARE CATEGORICAL?\n",
+    "# ANSWER: customer, st, gender, education, customer_lifetime_value, number_of_open_complaints, policy_type, vehicle_class,state\n",
+    "\n",
+    "# UNDERSTAND THE MEANING OF ALL COLUMNS\n",
+    "# ANSWER:\n",
+    "\n",
+    "datatotal_data.dtypes\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 93,
+   "id": "32de16d5",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>state</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>23405.98798</td>\n",
+       "      <td>71941.0</td>\n",
+       "      <td>73.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>21604.0</td>\n",
+       "      <td>79.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>85.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>21941.0</td>\n",
+       "      <td>96.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>77.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>12074 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer          st gender             education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "7065  LA72316         NaN      M              Bachelor   \n",
+       "7066  PK87824         NaN      F               College   \n",
+       "7067  TD14365         NaN      M              Bachelor   \n",
+       "7068  UP19263         NaN      M               College   \n",
+       "7069  Y167826         NaN      M               College   \n",
+       "\n",
+       "     customer_lifetime_value   income  monthly_premium_auto     policy_type  \\\n",
+       "0                        NaN      0.0                1000.0   Personal Auto   \n",
+       "1                 697953.59%      0.0                  94.0   Personal Auto   \n",
+       "2                1288743.17%  48767.0                 108.0   Personal Auto   \n",
+       "3                 764586.18%      0.0                 106.0  Corporate Auto   \n",
+       "4                 536307.65%  36357.0                  68.0   Personal Auto   \n",
+       "...                      ...      ...                   ...             ...   \n",
+       "7065             23405.98798  71941.0                  73.0   Personal Auto   \n",
+       "7066             3096.511217  21604.0                  79.0  Corporate Auto   \n",
+       "7067             8163.890428      0.0                  85.0  Corporate Auto   \n",
+       "7068             7524.442436  21941.0                  96.0   Personal Auto   \n",
+       "7069             2611.836866      0.0                  77.0  Corporate Auto   \n",
+       "\n",
+       "      vehicle_class  total_claim_amount       state  \n",
+       "0     Four-Door Car            2.704934         NaN  \n",
+       "1     Four-Door Car         1131.464935         NaN  \n",
+       "2      Two-Door Car          566.472247         NaN  \n",
+       "3               SUV          529.881344         NaN  \n",
+       "4     Four-Door Car           17.269323         NaN  \n",
+       "...             ...                 ...         ...  \n",
+       "7065  Four-Door Car          198.234764  California  \n",
+       "7066  Four-Door Car          379.200000  California  \n",
+       "7067  Four-Door Car          790.784983  California  \n",
+       "7068  Four-Door Car          691.200000  California  \n",
+       "7069   Two-Door Car          369.600000  California  \n",
+       "\n",
+       "[12074 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 93,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    " # DELETE THE COLUMN EDUCATION AND THE NUMBER OF OPEN COMPLAINTS FROM THE DATAFRAME\n",
+    "    \n",
+    "totaldata_data = datatotal_data.drop(['education'], axis = 1)\n",
+    "totaldata_data\n",
+    "\n",
+    "totaldata_data = datatotal_data.drop(['number_of_open_complaints'], axis = 1)\n",
+    "totaldata_data\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 96,
+   "id": "854a8c46",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>state</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>RB50392</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Master</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>2.704934</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>QZ44356</td>\n",
+       "      <td>Arizona</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>697953.59%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>94.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>1131.464935</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>AI49188</td>\n",
+       "      <td>Nevada</td>\n",
+       "      <td>F</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>1288743.17%</td>\n",
+       "      <td>48767.0</td>\n",
+       "      <td>108.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>566.472247</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>WW63253</td>\n",
+       "      <td>California</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>764586.18%</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>106.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>SUV</td>\n",
+       "      <td>529.881344</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GA49547</td>\n",
+       "      <td>Washington</td>\n",
+       "      <td>M</td>\n",
+       "      <td>High School or Below</td>\n",
+       "      <td>536307.65%</td>\n",
+       "      <td>36357.0</td>\n",
+       "      <td>68.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>17.269323</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7065</th>\n",
+       "      <td>LA72316</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>23405.98798</td>\n",
+       "      <td>71941.0</td>\n",
+       "      <td>73.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>198.234764</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7066</th>\n",
+       "      <td>PK87824</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>F</td>\n",
+       "      <td>College</td>\n",
+       "      <td>3096.511217</td>\n",
+       "      <td>21604.0</td>\n",
+       "      <td>79.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>379.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7067</th>\n",
+       "      <td>TD14365</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>Bachelor</td>\n",
+       "      <td>8163.890428</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>85.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>790.784983</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7068</th>\n",
+       "      <td>UP19263</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>7524.442436</td>\n",
+       "      <td>21941.0</td>\n",
+       "      <td>96.0</td>\n",
+       "      <td>Personal Auto</td>\n",
+       "      <td>Four-Door Car</td>\n",
+       "      <td>691.200000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7069</th>\n",
+       "      <td>Y167826</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>M</td>\n",
+       "      <td>College</td>\n",
+       "      <td>2611.836866</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>77.0</td>\n",
+       "      <td>Corporate Auto</td>\n",
+       "      <td>Two-Door Car</td>\n",
+       "      <td>369.600000</td>\n",
+       "      <td>California</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>9135 rows × 11 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     customer          st gender             education  \\\n",
+       "0     RB50392  Washington    NaN                Master   \n",
+       "1     QZ44356     Arizona      F              Bachelor   \n",
+       "2     AI49188      Nevada      F              Bachelor   \n",
+       "3     WW63253  California      M              Bachelor   \n",
+       "4     GA49547  Washington      M  High School or Below   \n",
+       "...       ...         ...    ...                   ...   \n",
+       "7065  LA72316         NaN      M              Bachelor   \n",
+       "7066  PK87824         NaN      F               College   \n",
+       "7067  TD14365         NaN      M              Bachelor   \n",
+       "7068  UP19263         NaN      M               College   \n",
+       "7069  Y167826         NaN      M               College   \n",
+       "\n",
+       "     customer_lifetime_value   income  monthly_premium_auto     policy_type  \\\n",
+       "0                        NaN      0.0                1000.0   Personal Auto   \n",
+       "1                 697953.59%      0.0                  94.0   Personal Auto   \n",
+       "2                1288743.17%  48767.0                 108.0   Personal Auto   \n",
+       "3                 764586.18%      0.0                 106.0  Corporate Auto   \n",
+       "4                 536307.65%  36357.0                  68.0   Personal Auto   \n",
+       "...                      ...      ...                   ...             ...   \n",
+       "7065             23405.98798  71941.0                  73.0   Personal Auto   \n",
+       "7066             3096.511217  21604.0                  79.0  Corporate Auto   \n",
+       "7067             8163.890428      0.0                  85.0  Corporate Auto   \n",
+       "7068             7524.442436  21941.0                  96.0   Personal Auto   \n",
+       "7069             2611.836866      0.0                  77.0  Corporate Auto   \n",
+       "\n",
+       "      vehicle_class  total_claim_amount       state  \n",
+       "0     Four-Door Car            2.704934         NaN  \n",
+       "1     Four-Door Car         1131.464935         NaN  \n",
+       "2      Two-Door Car          566.472247         NaN  \n",
+       "3               SUV          529.881344         NaN  \n",
+       "4     Four-Door Car           17.269323         NaN  \n",
+       "...             ...                 ...         ...  \n",
+       "7065  Four-Door Car          198.234764  California  \n",
+       "7066  Four-Door Car          379.200000  California  \n",
+       "7067  Four-Door Car          790.784983  California  \n",
+       "7068  Four-Door Car          691.200000  California  \n",
+       "7069   Two-Door Car          369.600000  California  \n",
+       "\n",
+       "[9135 rows x 11 columns]"
+      ]
+     },
+     "execution_count": 96,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# CHECK FOR DUPLICATE ROWS IN THE DATA AND REMOVE IF ANY\n",
+    "\n",
+    "totaldata_data.duplicated() \n",
+    "\n",
+    "totaldata_data[totaldata_data.duplicated()] \n",
+    "\n",
+    "totaldata_data = totaldata_data.drop_duplicates()\n",
+    "\n",
+    "totaldata_data\n",
+    "\n",
+    "\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 111,
+   "id": "90edf4ee",
+   "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>customer</th>\n",
+       "      <th>st</th>\n",
+       "      <th>gender</th>\n",
+       "      <th>education</th>\n",
+       "      <th>customer_lifetime_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>monthly_premium_auto</th>\n",
+       "      <th>number_of_open_complaints</th>\n",
+       "      <th>policy_type</th>\n",
+       "      <th>vehicle_class</th>\n",
+       "      <th>total_claim_amount</th>\n",
+       "      <th>state</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "Empty DataFrame\n",
+       "Columns: [customer, st, gender, education, customer_lifetime_value, income, monthly_premium_auto, number_of_open_complaints, policy_type, vehicle_class, total_claim_amount, state]\n",
+       "Index: []"
+      ]
+     },
+     "execution_count": 111,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    " # Filter out the data for customers who have an income of 0 or less.\n",
+    "    \n",
+    "datatotal_data[datatotal_data['income'] < 0]\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.11.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}