intuit.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Intuit QuickBook"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### INTRODUCTION"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "QuickBooks is an accounting software package developed and marketed by Intuit. The key event took place in 1995 with release of Version 3 of QuickBooks. QuickBooks products are geared mainly toward small and medium-sized businesses and offer on-premises accounting applications. Although, the products at that time could only be purchased through Intuit Direct (i.e., delivery by mail) or through “brick-and-mortar” retailers like Best Buy.\n",
    "Through this exercise, we do a detailed analysis of modeling the response to an upselling campaign by using testing and predictive analysis to target the \"right\" customers. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pyrsm as rsm\n",
    "import seaborn as sns\n",
    "import scipy.stats as stats\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.stats as sms\n",
    "#import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import statsmodels.formula.api as smf\n",
    "from sklearn import preprocessing\n",
    "from statsmodels.genmod.families import Binomial\n",
    "from statsmodels.genmod.families.links import logit\n",
    "from sklearn.inspection import permutation_importance\n",
    "from sklearn.neural_network import MLPClassifier\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>owntaxprod</th>\n",
       "      <th>upgraded</th>\n",
       "      <th>res1</th>\n",
       "      <th>training</th>\n",
       "      <th>res1_yes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>94553</td>\n",
       "      <td>18</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>109.5</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>53190</td>\n",
       "      <td>10</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>69.5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>93.0</td>\n",
       "      <td>14</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>02125</td>\n",
       "      <td>1</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>22.0</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id    zip  zip_bins      sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "0   1  94553        18     Male        0        2    109.5     5          12   \n",
       "1   2  53190        10  Unknown        0        1     69.5     4           3   \n",
       "2   3  37091         8     Male        0        4     93.0    14          29   \n",
       "3   4  02125         1     Male        0        1     22.0    17           1   \n",
       "4   5  60201        11     Male        0        1     24.5     2           3   \n",
       "\n",
       "   version1  owntaxprod  upgraded res1  training  res1_yes  \n",
       "0         0           0         0   No         1         0  \n",
       "1         0           0         0   No         0         0  \n",
       "2         0           0         1   No         0         0  \n",
       "3         0           0         0   No         1         0  \n",
       "4         0           0         0   No         0         0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## loading the data - this dataset must NOT be changed\n",
    "intuit75k = pd.read_pickle(\"data/intuit75k.pkl\")\n",
    "intuit75k[\"res1_yes\"] = rsm.ifelse(intuit75k.res1 == \"Yes\", 1, 0)\n",
    "intuit75k.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "## Intuit: Quickbooks upgrade\n",
       "\n",
       "The purpose of this exercise is to gain experience modeling the response to an upsell campaign. The `intuit75k.rds` file contains data on 75,000 (small) businesses that were selected randomly from the 801,821 that were sent the wave-1 mailing. The mailing contained an offer to upgrade to the latest version of the Quickbooks software. \n",
       "\n",
       "Variable `res1` denotes which of these businesses responded to the mailing by purchasing Quickbooks version 3.0 from Intuit Direct. Note that Intuit Direct sells products directly to its customers rather than through a retailer. Use the available data to predict which businesses that did not respond to the wave-1 mailing, are most likely to respond to the wave-2 mailing. Note that variables were added, deleted, and recoded so please ignore the variable descriptions in Exhibit 3 in the case in the course reader. Instead, use the variable descriptions below:\n",
       "\n",
       "## Variable description\n",
       "\n",
       "* id: Small business customer ID\n",
       "* zip: 5-Digit ZIP Code (00000=unknown, 99999=international ZIPs).\n",
       "* zip_bins: Zip-code bins (20 approx. equal sized bins from lowest to highest zip code number)\n",
       "* sex: \"Female\", \"Male\", or \"Unknown\"\n",
       "* bizflag: Business Flag.  Address contains a Business name (1 = yes, 0 = no or unknown).\n",
       "* numords: Number of orders from Intuit Direct in the previous 36 months\n",
       "* dollars: Total $ ordered from Intuit Direct in the previous 36 months\n",
       "* last: Time (in months) since last order from Intuit Direct in previous 36 months\n",
       "* sincepurch: Time (in months) since original (not upgrade) purchase of Quickbooks\n",
       "* version1: Is 1 if customer's current Quickbooks is version 1, 0 if version 2\n",
       "* owntaxprod: Is 1 if customer purchased tax software, 0 otherwise\n",
       "* upgraded: Is 1 if customer upgraded from Quickbooks vs. 1 to vs. 2\n",
       "* res1: Response to wave 1 mailing (\"Yes\" if responded else \"No\")\n",
       "* training: 70/30 split, 1 for training sample, 0 for validation sample\n"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# show dataset description\n",
    "rsm.describe(intuit75k)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>owntaxprod</th>\n",
       "      <th>upgraded</th>\n",
       "      <th>res1</th>\n",
       "      <th>training</th>\n",
       "      <th>res1_yes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>94553</td>\n",
       "      <td>18</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>109.5</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>02125</td>\n",
       "      <td>1</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>22.0</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>12309</td>\n",
       "      <td>3</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>20.0</td>\n",
       "      <td>17</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>55122</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>4</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>20036</td>\n",
       "      <td>5</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>73.5</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id    zip  zip_bins   sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "0   1  94553        18  Male        0        2    109.5     5          12   \n",
       "3   4  02125         1  Male        0        1     22.0    17           1   \n",
       "5   6  12309         3  Male        0        1     20.0    17          17   \n",
       "7   8  55122        11  Male        1        1     24.5     4          17   \n",
       "9  10  20036         5  Male        0        3     73.5    10           9   \n",
       "\n",
       "   version1  owntaxprod  upgraded res1  training  res1_yes  \n",
       "0         0           0         0   No         1         0  \n",
       "3         0           0         0   No         1         0  \n",
       "5         0           0         1   No         1         0  \n",
       "7         1           0         0   No         1         0  \n",
       "9         0           0         0   No         1         0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_training = intuit75k[intuit75k['training'] == 1]\n",
    "intuit75k_training.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    50002\n",
       "1     2498\n",
       "Name: res1_yes, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_training[\"res1_yes\"].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Percentage of customers responded (i.e., bought anything)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'4.76%'"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Percentage of people who bought the version 3\n",
    "perc_yes_train = str(round((intuit75k_training[\"res1_yes\"].value_counts()[1]/len(intuit75k_training))*100,2)) + '%'\n",
    "perc_yes_train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create independent quintile variables for recency, frequency and monetary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k = intuit75k.assign(rec_iq = rsm.xtile(intuit75k['last'], 5))\n",
    "intuit75k = intuit75k.assign(freq_iq = rsm.xtile(intuit75k['numords'], 5, rev = True))\n",
    "intuit75k = intuit75k.assign(mon_iq = rsm.xtile(intuit75k['dollars'], 5, rev = True))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "RFM is a widely used technique for doing analysis that uses data based on existing customer behaviour to predict how a new customer is likely to act in the future. This is built using three key factors - Recency - Frequency - Monetary (RFM), selecting significant customers and dividing them into different groups based on similarities found on this variables and past behavioral data. \n",
    "\n",
    "- Recency(R) is the number of days since the customer's last purchase. \n",
    "- Frequency(F) is the number of times the customer has made a purchase. \n",
    "- Monetary (M) is the total amount spent by the customer in the same period. \n",
    "\n",
    "RFM analysis enables marketers to increase revenue by targeting specific groups of existing customers (customer segmentation) by doing comparison between found segregation on the factors described above. It gives a sense of how much revenue comes from repeat customers (versus new customers). Each variable aims to segregate in 5 different bins\n",
    "\n",
    "- Recency: Bin 1 depicts the customers that were most recent and bin 5 depicts the least recent.\n",
    "- Frequency: Bin 1 depicts the customers that were more frequently made purchases and bin 5 depicts the least frequent.\n",
    "- Monetary: Bin 1 depicts the customers that have spent the most amount and bin 5 depicts the ones that spent the least."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create bar charts showing the response rate (i.e., the proportion of customers who bought something) for this deal per (independent) recency, frequency, and monetary quintile "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig1 = rsm.prop_plot(intuit75k[intuit75k['training']==1], \"rec_iq\", \"res1\", \"Yes\")\n",
    "fig1 = fig1.set_xlabel(\"Recency value quintiles (rec_iq)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Recency Chart : The above graph depicts the proportion of customers who responded \"yes\" to waive 1 mail that was send out, per independent recency. The customers in bin 1 had the highest response rate from the mail sent out as opposed to ones in bin 5. It shows that the response rate from the top quintile (20% group) was 9.00%, while the next quintile responded at a rate of ~5.4%. The last quintile had a response rate of ~2.3%. \n",
    "\n",
    "This graph shows that the recency factor has a constant downgrade at every level on the probability of purchasing the latest upgrade offer thru email. Although we can appreciate the biggest drop from bin1 to bin2, which would elevate the predictive probabilities to the group that pertains in bin1 the most."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig2 = rsm.prop_plot(intuit75k[intuit75k['training']==1], \"freq_iq\", \"res1\", \"Yes\")\n",
    "fig2 = fig2.set_xlabel(\"Frequency value quintiles (freq_iq)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Frequency Chart : The above graph depicts the proportion of customers who responded \"yes\" to waive 1 mail that was send out, per independent frequency. The customers in bin 1 had the highest frequency rate from the mail sent out as opposed to ones in bin 5. The customers in bin 1 had the highest response rate from the mail sent out as opposed to ones in bin 5. It shows that the response rate from the top quintile (20% group) was 8.00%, while the next quintile responded at a rate of 5.00%. The last quintile had a response rate of ~3.3%. \n",
    "\n",
    "This graph shows that the frequency factor has a drop of almost half from bin1 to bin2 on the probability of purchasing the latest upgrade offer thru email. Although it keeps downgrading from bin2 through bin5, not much drop is seen, meaning that in this case segmentation on bin1 will have a clearer advantage difference compare to the rest.\n",
    "\n",
    "Additionally we see that bin4 is skipped in this case, that is due to the fact that the amount of IDs found among bin3 and beyond were not enough to be scalable with the first two bins."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig3 = rsm.prop_plot(intuit75k[intuit75k['training']==1], \"mon_iq\", \"res1\", \"Yes\")\n",
    "fig3 = fig3.set_xlabel(\"Monetary value quintiles (mon_iq)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Monetory Chart : The above graph depicts the proportion of customers who responded \"yes\" to waive 1 mail that was send out, per independent monetory. The customers in bin 1 had the highest expenditure rate from the mail sent out as opposed to ones in bin 5. It shows that the response rate from the top quintile (20% group) was 7.00%, while the next quintile responded at a rate of 4.5%. The last quintile had a response rate of 3.2%. \n",
    "\n",
    "This graph shows that the monetary factor has again the most significant drop between bin1 and bin2. Then from bin2 through bin5 we appreciate a consant downgrade but with not much difference between them. Therefore segregation on bin1 will have the most probability of purchasing the latest upgrade thru email and with not much difference on the rest of the segmentations."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Generate Independent RFM index "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To perform RFM analysis, we bin the customers using Independent or Sequential Method. \n",
    "In independent method, the ranks of RFM are assigned independently. The interpretation of each rank becomes unambiguous. The monetory value of a bin 4 in 234 is interpreted the same way as that in 114. This is one advantage of interpretting the bin values in independent approach. One disadvantage of this approach is that incase of less samples, it can result in a less even distribution of the combined RFM scores i.e the quantities of customers in the RFM cells might be quite different."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k = intuit75k.assign(\n",
    "    rfm_iq=intuit75k.rec_iq.astype(str)\n",
    "    + intuit75k.freq_iq.astype(str)\n",
    "    + intuit75k.mon_iq.astype(str)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16, 7))\n",
    "# fig = rsm.prop_plot(bbb, \"rfm_iq\", \"buyer\", \"yes\", breakeven=breakeven)\n",
    "fig = rsm.prop_plot(intuit75k, \"rfm_iq\", \"res1\", \"Yes\")\n",
    "fig.set_xticklabels(fig.get_xticklabels(), rotation=90)\n",
    "fig = fig.set(xlabel=\"Independent RFM index (rfm_iq)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "RFM analysis depends on recency, frequency and monetary measures, but the real power of this technique comes from combining them into a three digit RFM \"cell code\". Hence, we now put all the customers in bins of 3 digits (555, 554, 553, 552, 551, 545,...down to 111). There are 125 different \"cells\" in all. \n",
    "\n",
    "We can observe that there are few bins with no customers in them like bin number 315 and 515. This is due to the nature of the customers in such bins. A customer in 515 is bound to be the least recent but more frequent and frugal in expenditure. Such a customer is generally very hard to find."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0235"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# state your assumptions relevant to calculate the breakeven response rate\n",
    "# breakeven should be of type `float`\n",
    "\n",
    "cost = 1.41 # each mail piece costs $1.41\n",
    "margin = 60  # the margin (or net revenue) from each responder, excluding the mailing cost, is $60\n",
    "breakeven =  cost/margin # float, length 1\n",
    "\n",
    "breakeven"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16, 7))\n",
    "fig_rfm_iq = rsm.prop_plot(intuit75k, \"rfm_iq\", \"res1\", \"Yes\", breakeven = breakeven)\n",
    "fig_rfm_iq.set_xticklabels(fig_rfm_iq.get_xticklabels(), rotation=90)\n",
    "fig_rfm_iq = fig_rfm_iq.set(xlabel=\"Independent RFM index (rsm_iq)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we calculate a breakeven point to further modify the customers we will be targetting. We will only send mails to those who's response rate lies above the breakeven point. Break even means that the net revenue (profit) from sales to members of the cell exactly paid for the cost of mailing to the cell members.\n",
    "\n",
    "Given, \n",
    "each mail piece costs 1.41 usd\n",
    "the margin (or net revenue) from each responder, excluding the mailing cost, is 60 usd\n",
    "So, breakeven =  cost/margin\n",
    "\n",
    "The graph above shows that we should be targetting only the customers in independent bins who lie above our breakeven point of 0.023 for the company to remain in profit."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Generate sequential quintiles"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To further modify the RFM analysis, we use sequential method to assign the ranks of RFM in order. In this method, first the Recency values are assigned. Then within each recency rank, customers are assigned a frequency rank, and within each frequency rank, customer are then assigned a monetary rank. This tends to take into account all the three factors in each bin value and hence provide a better even distribution of combined RFM scores. Hence the advantage of this method is that the numbe of customers in each RFM cell is approximately the same for all cells. The disadvantage is that the meaning of Frequecy and Monetory where the bins are same holds different meaning i.e. a customer with 234 and another with 534 will depict quite a different behaviour from eachother. Their frequency and monetory value might be different even though they have the same scores for them. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "intuit75k = intuit75k.assign(\n",
    "    freq_sq=intuit75k.groupby(\"rec_iq\").numords.transform(rsm.xtile, 5, rev=True)\n",
    ")\n",
    "fig_8 = rsm.prop_plot(intuit75k, \"freq_sq\", \"res1\", \"Yes\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "intuit75k = intuit75k.assign(\n",
    "    mon_sq=intuit75k.groupby([\"rec_iq\", \"freq_sq\"]).dollars.transform(rsm.xtile, 5, rev=True)\n",
    ")\n",
    "fig_9 = rsm.prop_plot(intuit75k, \"mon_sq\", \"res1\", \"Yes\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sequential model creates nested binning, a simple rank is assigned to recency values, which in this case t=is the same recency binning that we did in Independent RFM. However, in sequential, within each recency rank, customers are then assigned a frequency rank  and within each frequency rank, customers are assigned a monetary rank as depiceted above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>...</th>\n",
       "      <th>res1</th>\n",
       "      <th>training</th>\n",
       "      <th>res1_yes</th>\n",
       "      <th>rec_iq</th>\n",
       "      <th>freq_iq</th>\n",
       "      <th>mon_iq</th>\n",
       "      <th>rfm_iq</th>\n",
       "      <th>freq_sq</th>\n",
       "      <th>mon_sq</th>\n",
       "      <th>rfm_sq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>94553</td>\n",
       "      <td>18</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>109.5</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>132</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>53190</td>\n",
       "      <td>10</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>69.5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>93.0</td>\n",
       "      <td>14</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>312</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>02125</td>\n",
       "      <td>1</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>22.0</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>355</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>155</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id    zip  zip_bins      sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "0   1  94553        18     Male        0        2    109.5     5          12   \n",
       "1   2  53190        10  Unknown        0        1     69.5     4           3   \n",
       "2   3  37091         8     Male        0        4     93.0    14          29   \n",
       "3   4  02125         1     Male        0        1     22.0    17           1   \n",
       "4   5  60201        11     Male        0        1     24.5     2           3   \n",
       "\n",
       "   version1  ...  res1  training res1_yes  rec_iq  freq_iq  mon_iq  rfm_iq  \\\n",
       "0         0  ...    No         1        0       1        3       2     132   \n",
       "1         0  ...    No         0        0       1        5       3     153   \n",
       "2         0  ...    No         0        0       3        1       2     312   \n",
       "3         0  ...    No         1        0       3        5       5     355   \n",
       "4         0  ...    No         0        0       1        5       5     155   \n",
       "\n",
       "   freq_sq mon_sq  rfm_sq  \n",
       "0        4      2     142  \n",
       "1        5      2     152  \n",
       "2        1      5     315  \n",
       "3        5      4     354  \n",
       "4        5      4     154  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k = intuit75k.assign(\n",
    "    rfm_sq=intuit75k.rec_iq.astype(str)\n",
    "    + intuit75k.freq_sq.astype(str)\n",
    "    + intuit75k.mon_sq.astype(str)\n",
    ")\n",
    "\n",
    "intuit75k.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16, 7))\n",
    "# fig = rsm.prop_plot(bbb, \"rfm_iq\", \"buyer\", \"yes\", breakeven=breakeven)\n",
    "fig_rfm_sq = rsm.prop_plot(intuit75k, \"rfm_sq\", \"res1\", \"Yes\", breakeven = breakeven)\n",
    "fig_rfm_sq.set_xticklabels(fig_rfm_sq.get_xticklabels(), rotation=90)\n",
    "fig_rfm_sq = fig_rfm_sq.set(xlabel=\"Independent RFM index (rsm_sq)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The graph above shows that we should be targetting only the customers in sequencial bins who lie above our breakeven point of 0.023 for the company to remain in profit after sequential bining. We also observe that there are no empty bins in sequencial RFM indexing as opposed to independent one. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Calculation of performance without targetting // TEST DATA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_testdf = intuit75k[intuit75k['training'] == 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    22500.000000\n",
       "mean         0.049022\n",
       "Name: res1_yes, dtype: float64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_testdf[\"res1_yes\"].agg([\"count\", \"mean\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>...</th>\n",
       "      <th>res1</th>\n",
       "      <th>training</th>\n",
       "      <th>res1_yes</th>\n",
       "      <th>rec_iq</th>\n",
       "      <th>freq_iq</th>\n",
       "      <th>mon_iq</th>\n",
       "      <th>rfm_iq</th>\n",
       "      <th>freq_sq</th>\n",
       "      <th>mon_sq</th>\n",
       "      <th>rfm_sq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>53190</td>\n",
       "      <td>10</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>69.5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>93.0</td>\n",
       "      <td>14</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>312</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>155</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>22980</td>\n",
       "      <td>5</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>49.5</td>\n",
       "      <td>13</td>\n",
       "      <td>36</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>254</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>34950</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>44.5</td>\n",
       "      <td>15</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>354</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>353</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id    zip  zip_bins      sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "1   2  53190        10  Unknown        0        1     69.5     4           3   \n",
       "2   3  37091         8     Male        0        4     93.0    14          29   \n",
       "4   5  60201        11     Male        0        1     24.5     2           3   \n",
       "6   7  22980         5     Male        0        1     49.5    13          36   \n",
       "8   9  34950         8     Male        0        1     44.5    15           4   \n",
       "\n",
       "   version1  ...  res1  training res1_yes  rec_iq  freq_iq  mon_iq  rfm_iq  \\\n",
       "1         0  ...    No         0        0       1        5       3     153   \n",
       "2         0  ...    No         0        0       3        1       2     312   \n",
       "4         0  ...    No         0        0       1        5       5     155   \n",
       "6         1  ...    No         0        0       2        5       4     254   \n",
       "8         0  ...    No         0        0       3        5       4     354   \n",
       "\n",
       "   freq_sq mon_sq  rfm_sq  \n",
       "1        5      2     152  \n",
       "2        1      5     315  \n",
       "4        5      4     154  \n",
       "6        5      2     252  \n",
       "8        5      3     353  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_testdf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.049022222222222224"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_testdf['res1_yes'].value_counts(normalize = True)[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(34455.0, 1.0860520094562647)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "perc_emails = 1\n",
    "nr_emails_nt = len(intuit75k_testdf) # total number of emails that would be sent out\n",
    "email_cost_nt = cost*nr_emails_nt  # cost of sending an email (float)\n",
    "response_rate_nt = intuit75k_testdf['res1_yes'].value_counts(normalize = True)[1]\n",
    "nr_responses_nt = response_rate_nt * nr_emails_nt # total number of positive responses\n",
    "revenue_nt = margin*nr_responses_nt# total revenue \n",
    "profit_nt = margin * nr_responses_nt - email_cost_nt# total profit \n",
    "ROME_nt = profit_nt/email_cost_nt# Return on Marketing Expenditures expressed as a proportion (no rounding)\n",
    "\n",
    "profit_nt, ROME_nt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.049022222222222224"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "response_rate_nt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Quickbooks would sms all 22,500 (100%) customers. The response rate for the selected customers is predicted to be 4.9% or 1,103 buyers. The revenue is equal to $66,180. The expected profit is $34,455. The email cost is estimated to be $31,725 with a ROME of 108.61%\n"
     ]
    }
   ],
   "source": [
    "print(\n",
    "    f\"\"\"Quickbooks would sms all {int(nr_emails_nt):,} ({round((100 * perc_emails), 2)}%) customers. The response rate for the selected customers is predicted to be {round((100 * response_rate_nt), 2)}% or {int(nr_responses_nt):,} buyers. The revenue is equal to ${int(revenue_nt):,}. The expected profit is ${int(profit_nt):,}. The email cost is estimated to be ${int(email_cost_nt):,} with a ROME of {round((100 * ROME_nt), 2)}%\"\"\"\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Evaluate the performance implications of offering the deal to only those customers in RFM cells with a response rate greater than the breakeven response rate// TEST DATA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### IQ CELLS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>...</th>\n",
       "      <th>res1</th>\n",
       "      <th>training</th>\n",
       "      <th>res1_yes</th>\n",
       "      <th>rec_iq</th>\n",
       "      <th>freq_iq</th>\n",
       "      <th>mon_iq</th>\n",
       "      <th>rfm_iq</th>\n",
       "      <th>freq_sq</th>\n",
       "      <th>mon_sq</th>\n",
       "      <th>rfm_sq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>94553</td>\n",
       "      <td>18</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>109.5</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>132</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>53190</td>\n",
       "      <td>10</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>69.5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>93.0</td>\n",
       "      <td>14</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>312</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>02125</td>\n",
       "      <td>1</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>22.0</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>355</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>155</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id    zip  zip_bins      sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "0   1  94553        18     Male        0        2    109.5     5          12   \n",
       "1   2  53190        10  Unknown        0        1     69.5     4           3   \n",
       "2   3  37091         8     Male        0        4     93.0    14          29   \n",
       "3   4  02125         1     Male        0        1     22.0    17           1   \n",
       "4   5  60201        11     Male        0        1     24.5     2           3   \n",
       "\n",
       "   version1  ...  res1  training res1_yes  rec_iq  freq_iq  mon_iq  rfm_iq  \\\n",
       "0         0  ...    No         1        0       1        3       2     132   \n",
       "1         0  ...    No         0        0       1        5       3     153   \n",
       "2         0  ...    No         0        0       3        1       2     312   \n",
       "3         0  ...    No         1        0       3        5       5     355   \n",
       "4         0  ...    No         0        0       1        5       5     155   \n",
       "\n",
       "   freq_sq mon_sq  rfm_sq  \n",
       "0        4      2     142  \n",
       "1        5      2     152  \n",
       "2        1      5     315  \n",
       "3        5      4     354  \n",
       "4        5      4     154  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "def response_rate(x, lev=\"Yes\"):\n",
    "    return np.nanmean(x == lev)\n",
    "\n",
    "def emailto(x, lev=\"Yes\", breakeven=0):\n",
    "    return np.nanmean(x == \"Yes\") > breakeven"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_testdf_2 = intuit75k[intuit75k['training'] == 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_testdf_2 = intuit75k_testdf_2.assign(\n",
    "    rfm_iq_resp=(intuit75k_testdf_2.groupby(\"rfm_iq\").res1.transform(response_rate)),\n",
    "    emailto_iq=intuit75k_testdf_2.groupby(\"rfm_iq\").res1.transform(emailto, breakeven=breakeven*2)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>...</th>\n",
       "      <th>res1_yes</th>\n",
       "      <th>rec_iq</th>\n",
       "      <th>freq_iq</th>\n",
       "      <th>mon_iq</th>\n",
       "      <th>rfm_iq</th>\n",
       "      <th>freq_sq</th>\n",
       "      <th>mon_sq</th>\n",
       "      <th>rfm_sq</th>\n",
       "      <th>rfm_iq_resp</th>\n",
       "      <th>emailto_iq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>53190</td>\n",
       "      <td>10</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>69.5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>152</td>\n",
       "      <td>0.051370</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>93.0</td>\n",
       "      <td>14</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>312</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>315</td>\n",
       "      <td>0.067568</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>155</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "      <td>0.049226</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>22980</td>\n",
       "      <td>5</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>49.5</td>\n",
       "      <td>13</td>\n",
       "      <td>36</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>254</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>252</td>\n",
       "      <td>0.028623</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>34950</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>44.5</td>\n",
       "      <td>15</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>354</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>353</td>\n",
       "      <td>0.026408</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id    zip  zip_bins      sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "1   2  53190        10  Unknown        0        1     69.5     4           3   \n",
       "2   3  37091         8     Male        0        4     93.0    14          29   \n",
       "4   5  60201        11     Male        0        1     24.5     2           3   \n",
       "6   7  22980         5     Male        0        1     49.5    13          36   \n",
       "8   9  34950         8     Male        0        1     44.5    15           4   \n",
       "\n",
       "   version1  ...  res1_yes  rec_iq freq_iq  mon_iq  rfm_iq  freq_sq  mon_sq  \\\n",
       "1         0  ...         0       1       5       3     153        5       2   \n",
       "2         0  ...         0       3       1       2     312        1       5   \n",
       "4         0  ...         0       1       5       5     155        5       4   \n",
       "6         1  ...         0       2       5       4     254        5       2   \n",
       "8         0  ...         0       3       5       4     354        5       3   \n",
       "\n",
       "   rfm_sq rfm_iq_resp  emailto_iq  \n",
       "1     152    0.051370        True  \n",
       "2     315    0.067568        True  \n",
       "4     154    0.049226        True  \n",
       "6     252    0.028623       False  \n",
       "8     353    0.026408       False  \n",
       "\n",
       "[5 rows x 24 columns]"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_testdf_2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    13323\n",
       "True      9177\n",
       "Name: emailto_iq, dtype: int64"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_testdf_2.emailto_iq.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>...</th>\n",
       "      <th>res1_yes</th>\n",
       "      <th>rec_iq</th>\n",
       "      <th>freq_iq</th>\n",
       "      <th>mon_iq</th>\n",
       "      <th>rfm_iq</th>\n",
       "      <th>freq_sq</th>\n",
       "      <th>mon_sq</th>\n",
       "      <th>rfm_sq</th>\n",
       "      <th>rfm_iq_resp</th>\n",
       "      <th>emailto_iq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>53190</td>\n",
       "      <td>10</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>69.5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>152</td>\n",
       "      <td>0.051370</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>93.0</td>\n",
       "      <td>14</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>312</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>315</td>\n",
       "      <td>0.067568</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>155</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "      <td>0.049226</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>91942</td>\n",
       "      <td>17</td>\n",
       "      <td>Male</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>79.0</td>\n",
       "      <td>5</td>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>113</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>115</td>\n",
       "      <td>0.180451</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>27262</td>\n",
       "      <td>6</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>5</td>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>0.057208</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    id    zip  zip_bins      sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "1    2  53190        10  Unknown        0        1     69.5     4           3   \n",
       "2    3  37091         8     Male        0        4     93.0    14          29   \n",
       "4    5  60201        11     Male        0        1     24.5     2           3   \n",
       "11  12  91942        17     Male        1        5     79.0     5          21   \n",
       "12  13  27262         6     Male        0        1     38.0     5          22   \n",
       "\n",
       "    version1  ...  res1_yes  rec_iq freq_iq  mon_iq  rfm_iq  freq_sq  mon_sq  \\\n",
       "1          0  ...         0       1       5       3     153        5       2   \n",
       "2          0  ...         0       3       1       2     312        1       5   \n",
       "4          0  ...         0       1       5       5     155        5       4   \n",
       "11         0  ...         1       1       1       3     113        1       5   \n",
       "12         1  ...         0       1       5       4     154        5       3   \n",
       "\n",
       "    rfm_sq rfm_iq_resp  emailto_iq  \n",
       "1      152    0.051370        True  \n",
       "2      315    0.067568        True  \n",
       "4      154    0.049226        True  \n",
       "11     115    0.180451        True  \n",
       "12     153    0.057208        True  \n",
       "\n",
       "[5 rows x 24 columns]"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_testdf_2a = intuit75k_testdf_2[intuit75k_testdf_2['emailto_iq'] == True]\n",
    "intuit75k_testdf_2a.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(30380.43, 2.3478701378793887)"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "perc_emails_iq = len(intuit75k_testdf_2a)/len(intuit75k_testdf_2)\n",
    "nr_emails_iq = len(intuit75k_testdf_2a) # total number of emails that would be sent out\n",
    "email_cost_iq = cost*nr_emails_iq # cost of sending an email (float)\n",
    "response_rate_iq = intuit75k_testdf_2a['res1_yes'].value_counts(normalize = True)[1]\n",
    "nr_responses_iq = response_rate_iq * nr_emails_iq # total number of positive responses\n",
    "revenue_iq = margin*nr_responses_iq# total revenue \n",
    "profit_iq = margin * nr_responses_iq - email_cost_iq# total profit \n",
    "ROME_iq = profit_iq/email_cost_iq# Return on Marketing Expenditures expressed as a proportion (no rounding)\n",
    "\n",
    "profit_iq, ROME_iq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Quickbooks would sms all 9,177 (40.79%) customers. The response rate for the selected customers is predicted to be 7.87% or 722 buyers. The revenue is equal to $43,320. The expected profit is $30,380. The emai; cost is estimated to be $12,939 with a ROME of 234.79%\n"
     ]
    }
   ],
   "source": [
    "print(\n",
    "    f\"\"\"Quickbooks would sms all {int(nr_emails_iq):,} ({round((100 * perc_emails_iq), 2)}%) customers. The response rate for the selected customers is predicted to be {round((100 * response_rate_iq), 2)}% or {int(nr_responses_iq):,} buyers. The revenue is equal to ${int(revenue_iq):,}. The expected profit is ${int(profit_iq):,}. The emai; cost is estimated to be ${int(email_cost_iq):,} with a ROME of {round((100 * ROME_iq), 2)}%\"\"\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "316543999.4574"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#wave 2- scaled\n",
    "scale1_iq = round((100 * perc_emails_iq), 2)\n",
    "scale2_iq = round((100 * response_rate_iq), 2) \n",
    "tot_emails = 801821 - 38487\n",
    "total_cost_iq = scale1_iq*tot_emails*cost\n",
    "total_revenue2 = tot_emails*scale2_iq*margin\n",
    "profit2 = total_revenue2 - total_cost_iq\n",
    "profit2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### SQ CELLS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_testdf_2 = intuit75k_testdf_2.assign(\n",
    "    rfm_sq_resp=(intuit75k_testdf_2.groupby(\"rfm_sq\").res1.transform(response_rate)),\n",
    "    emailto_sq=intuit75k_testdf_2.groupby(\"rfm_sq\").res1.transform(emailto, breakeven=breakeven * 2)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    12921\n",
       "True      9579\n",
       "Name: emailto_sq, dtype: int64"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_testdf_2.emailto_sq.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>...</th>\n",
       "      <th>freq_iq</th>\n",
       "      <th>mon_iq</th>\n",
       "      <th>rfm_iq</th>\n",
       "      <th>freq_sq</th>\n",
       "      <th>mon_sq</th>\n",
       "      <th>rfm_sq</th>\n",
       "      <th>rfm_iq_resp</th>\n",
       "      <th>emailto_iq</th>\n",
       "      <th>rfm_sq_resp</th>\n",
       "      <th>emailto_sq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>93.0</td>\n",
       "      <td>14</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>312</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>315</td>\n",
       "      <td>0.067568</td>\n",
       "      <td>True</td>\n",
       "      <td>0.069307</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>155</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "      <td>0.049226</td>\n",
       "      <td>True</td>\n",
       "      <td>0.050532</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>91942</td>\n",
       "      <td>17</td>\n",
       "      <td>Male</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>79.0</td>\n",
       "      <td>5</td>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>113</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>115</td>\n",
       "      <td>0.180451</td>\n",
       "      <td>True</td>\n",
       "      <td>0.144231</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>27262</td>\n",
       "      <td>6</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>5</td>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>0.057208</td>\n",
       "      <td>True</td>\n",
       "      <td>0.061644</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>93535</td>\n",
       "      <td>18</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>40.5</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>234</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>235</td>\n",
       "      <td>0.036765</td>\n",
       "      <td>False</td>\n",
       "      <td>0.054622</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    id    zip  zip_bins   sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "2    3  37091         8  Male        0        4     93.0    14          29   \n",
       "4    5  60201        11  Male        0        1     24.5     2           3   \n",
       "11  12  91942        17  Male        1        5     79.0     5          21   \n",
       "12  13  27262         6  Male        0        1     38.0     5          22   \n",
       "14  15  93535        18  Male        0        2     40.5    10           6   \n",
       "\n",
       "    version1  ...  freq_iq  mon_iq rfm_iq  freq_sq  mon_sq  rfm_sq  \\\n",
       "2          0  ...        1       2    312        1       5     315   \n",
       "4          0  ...        5       5    155        5       4     154   \n",
       "11         0  ...        1       3    113        1       5     115   \n",
       "12         1  ...        5       4    154        5       3     153   \n",
       "14         0  ...        3       4    234        3       5     235   \n",
       "\n",
       "    rfm_iq_resp  emailto_iq rfm_sq_resp  emailto_sq  \n",
       "2      0.067568        True    0.069307        True  \n",
       "4      0.049226        True    0.050532        True  \n",
       "11     0.180451        True    0.144231        True  \n",
       "12     0.057208        True    0.061644        True  \n",
       "14     0.036765       False    0.054622        True  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_testdf_2b = intuit75k_testdf_2[intuit75k_testdf_2['emailto_sq'] == True]\n",
    "intuit75k_testdf_2b.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(31373.61, 2.322871618545)"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "perc_emails_sq = len(intuit75k_testdf_2b)/len(intuit75k_testdf_2)\n",
    "nr_emails_sq = len(intuit75k_testdf_2b)# total number of emails that would be sent out\n",
    "email_cost_sq = cost*nr_emails_sq # cost of sending an email (float)\n",
    "response_rate_sq = intuit75k_testdf_2b['res1_yes'].value_counts(normalize = True)[1]\n",
    "nr_responses_sq = response_rate_sq * nr_emails_sq # total number of positive responses\n",
    "revenue_sq = margin*nr_responses_sq# total revenue \n",
    "profit_sq = margin * nr_responses_sq - email_cost_sq# total profit \n",
    "ROME_sq = profit_sq/email_cost_sq# Return on Marketing Expenditures expressed as a proportion (no rounding)\n",
    "\n",
    "profit_sq, ROME_sq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Quickbooks would sms all 9,579 (42.57%) customers. The response rate for the selected customers is predicted to be 7.81% or 748\n",
      "buyers. The revenue is equal to $44,880. The expected profit is $31,373. The emai; cost is estimated to be $13,506 with a ROME of 232.29%\n"
     ]
    }
   ],
   "source": [
    "print(\n",
    "    f\"\"\"Quickbooks would sms all {int(nr_emails_sq):,} ({round((100 * perc_emails_sq), 2)}%) customers. The response rate for the selected customers is predicted to be {round((100 * response_rate_sq), 2)}% or {int(nr_responses_sq):,}\n",
    "buyers. The revenue is equal to ${int(revenue_sq):,}. The expected profit is ${int(profit_sq):,}. The emai; cost is estimated to be ${int(email_cost_sq):,} with a ROME of {round((100 * ROME_sq), 2)}%\"\"\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "311880181.3842"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#wave2\n",
    "scale1_sq = round((100 * perc_emails_sq), 2)\n",
    "scale2_sq = round((100 * response_rate_sq), 2) \n",
    "tot_emails = 801821 - 38487\n",
    "total_cost_sq = scale1_sq*tot_emails*cost\n",
    "total_revenue2 = tot_emails*scale2_sq*margin\n",
    "profit2 = total_revenue2 - total_cost_sq\n",
    "profit2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " After running analysis on the data, we find the response rate, expected revenue and profit for the three scenarios: \n",
    "\n",
    "Without RFM analysis, Quickbooks would mail all 22,500 (100%) customers. The response rate for the selected customers is predicted to be 4.9% or 1,103 buyers. The revenue is equal to  64,624.The expected profit is 32,899. The email cost is estimated to be  31,725 with a ROME of 103.7%. \n",
    "\n",
    "With IQ targeting, Quickbooks would mail all 17,983 (79.92%) customers. The response rate for the selected customers is predicted to be 5.67% or 1,020\n",
    "buyers. The revenue is equal to 61,200. The expected profit is  35,843. The email cost is estimated to be 25,356 with a ROME of 141.36%\n",
    "\n",
    "With SQ targeting, Quickbooks would mail all 18,188 (80.84%) customers. The response rate for the selected customers is predicted to be 5.62% or 1,022\n",
    "buyers. The revenue is equal to 61,320. The expected profit is 35,674. The email cost is estimated to be 25,645 with a ROME of 139.11%\n",
    "\n",
    "We can observe that without RFM Quickbooks would mail all customer without RFM analysis whereas with IQ and SQ cells it is 80%. IQ and SQ analysis result in nearly the same target percentage rate, revenue and expected profit. \n",
    "Hence it is indeed beneficial to use RFM analysis for targetting. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Why try models apart from RFM?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The result of modeling using RFM is that Intuit will mail to their most responsive customers, and will neglect all of the others. That means that they will get absolutely no attention, and will eventually end up losing them. This is OK for few customers, who are worthless anyway. But Intuit would want to try to hang on to the rest of their customers who, with a little attention can be persuaded to move up to a more profitable RFM cell. Our marketing should be designed to encourage customers in some cells to do just that."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### LOGISTIC REGRESSION"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Logistic regression is an statistical analysis used for predictive analytics and modeling. This approach is known to be little more exact than the previous model we explored (RFM) given that it serves to understand the relationship between a dependent variable (response variable), which is 'res1_yes' in our case and one or more independent variables. \n",
    "\n",
    "This allow us to predict the likelihood of an event happening or choice being made, based on the the observed data. In this case will allow us to determine accordingly which variables in the explored data will provide weight (positively or negatively) toward making a decision on purchasing the latest version of the Quickbooks software."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Standardize values\n",
    "\n",
    "When dealing with multiple variables it is important to standardize the values in order to better scale the impact that the variables with greater scale will come to affect the response variable and therefore be able to compare the impact that each variable assigned in the logistic regression more precisely."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "# select variables to standardize\n",
    "to_std = intuit75k.loc[:, \"numords\":\"sincepurch\"].columns\n",
    "intuit75k_std = intuit75k.copy()\n",
    "intuit75k_std[to_std] = rsm.scale_df(\n",
    "    intuit75k[to_std], sf=1, train=intuit75k.training == 1\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>...</th>\n",
       "      <th>res1</th>\n",
       "      <th>training</th>\n",
       "      <th>res1_yes</th>\n",
       "      <th>rec_iq</th>\n",
       "      <th>freq_iq</th>\n",
       "      <th>mon_iq</th>\n",
       "      <th>rfm_iq</th>\n",
       "      <th>freq_sq</th>\n",
       "      <th>mon_sq</th>\n",
       "      <th>rfm_sq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>94553</td>\n",
       "      <td>18</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.057079</td>\n",
       "      <td>0.207902</td>\n",
       "      <td>-1.137183</td>\n",
       "      <td>-0.360809</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>132</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>53190</td>\n",
       "      <td>10</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.863720</td>\n",
       "      <td>-0.286561</td>\n",
       "      <td>-1.241930</td>\n",
       "      <td>-1.258669</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1.556202</td>\n",
       "      <td>0.003936</td>\n",
       "      <td>-0.194459</td>\n",
       "      <td>1.335150</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>312</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>02125</td>\n",
       "      <td>1</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.863720</td>\n",
       "      <td>-0.873735</td>\n",
       "      <td>0.119783</td>\n",
       "      <td>-1.458194</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>355</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.863720</td>\n",
       "      <td>-0.842831</td>\n",
       "      <td>-1.451425</td>\n",
       "      <td>-1.258669</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>155</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id    zip  zip_bins      sex  bizflag   numords   dollars      last  \\\n",
       "0   1  94553        18     Male        0 -0.057079  0.207902 -1.137183   \n",
       "1   2  53190        10  Unknown        0 -0.863720 -0.286561 -1.241930   \n",
       "2   3  37091         8     Male        0  1.556202  0.003936 -0.194459   \n",
       "3   4  02125         1     Male        0 -0.863720 -0.873735  0.119783   \n",
       "4   5  60201        11     Male        0 -0.863720 -0.842831 -1.451425   \n",
       "\n",
       "   sincepurch  version1  ...  res1  training res1_yes  rec_iq  freq_iq  \\\n",
       "0   -0.360809         0  ...    No         1        0       1        3   \n",
       "1   -1.258669         0  ...    No         0        0       1        5   \n",
       "2    1.335150         0  ...    No         0        0       3        1   \n",
       "3   -1.458194         0  ...    No         1        0       3        5   \n",
       "4   -1.258669         0  ...    No         0        0       1        5   \n",
       "\n",
       "   mon_iq  rfm_iq  freq_sq mon_sq  rfm_sq  \n",
       "0       2     132        4      2     142  \n",
       "1       3     153        5      2     152  \n",
       "2       2     312        1      5     315  \n",
       "3       5     355        5      4     354  \n",
       "4       5     155        5      4     154  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_std.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_std_df1 = intuit75k_std[intuit75k_std['training'] == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        index     OR     OR%   2.5%  97.5% p.values     \n",
      "1    zip_bins  0.946   -5.4%  0.939  0.953   < .001  ***\n",
      "2     numords  1.320   32.0%  1.262  1.382   < .001  ***\n",
      "3     dollars  1.082    8.2%  1.037  1.129   < .001  ***\n",
      "4        last  0.666  -33.4%  0.637  0.696   < .001  ***\n",
      "5    version1  2.156  115.6%  1.948  2.385   < .001  ***\n",
      "6  owntaxprod  1.361   36.1%  1.118  1.659    0.002   **\n",
      "7    upgraded  2.631  163.1%  2.387  2.900   < .001  ***\n"
     ]
    }
   ],
   "source": [
    "lr_std = smf.glm(\n",
    "    formula=\"res1_yes ~ zip_bins + numords + dollars + last + version1 + owntaxprod + upgraded\",\n",
    "    family=Binomial(link=logit()),\n",
    "    data=intuit75k_std_df1,\n",
    ").fit()\n",
    "print(rsm.or_ci(lr_std))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "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>variable</th>\n",
       "      <th>vif</th>\n",
       "      <th>Rsq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>numords</td>\n",
       "      <td>1.558</td>\n",
       "      <td>0.358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dollars</td>\n",
       "      <td>1.527</td>\n",
       "      <td>0.345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>upgraded</td>\n",
       "      <td>1.082</td>\n",
       "      <td>0.075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>version1</td>\n",
       "      <td>1.081</td>\n",
       "      <td>0.075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>owntaxprod</td>\n",
       "      <td>1.027</td>\n",
       "      <td>0.026</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>last</td>\n",
       "      <td>1.017</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>zip_bins</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     variable    vif    Rsq\n",
       "0     numords  1.558  0.358\n",
       "1     dollars  1.527  0.345\n",
       "2    upgraded  1.082  0.075\n",
       "3    version1  1.081  0.075\n",
       "4  owntaxprod  1.027  0.026\n",
       "5        last  1.017  0.017\n",
       "6    zip_bins  1.000  0.000"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rsm.vif(lr_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "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>index</th>\n",
       "      <th>OR</th>\n",
       "      <th>OR%</th>\n",
       "      <th>2.5%</th>\n",
       "      <th>97.5%</th>\n",
       "      <th>p.values</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>zip_bins</td>\n",
       "      <td>0.946</td>\n",
       "      <td>-5.4%</td>\n",
       "      <td>0.939</td>\n",
       "      <td>0.953</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>numords</td>\n",
       "      <td>1.320</td>\n",
       "      <td>32.0%</td>\n",
       "      <td>1.262</td>\n",
       "      <td>1.382</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>dollars</td>\n",
       "      <td>1.082</td>\n",
       "      <td>8.2%</td>\n",
       "      <td>1.037</td>\n",
       "      <td>1.129</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>last</td>\n",
       "      <td>0.666</td>\n",
       "      <td>-33.4%</td>\n",
       "      <td>0.637</td>\n",
       "      <td>0.696</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>version1</td>\n",
       "      <td>2.156</td>\n",
       "      <td>115.6%</td>\n",
       "      <td>1.948</td>\n",
       "      <td>2.385</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>owntaxprod</td>\n",
       "      <td>1.361</td>\n",
       "      <td>36.1%</td>\n",
       "      <td>1.118</td>\n",
       "      <td>1.659</td>\n",
       "      <td>0.002</td>\n",
       "      <td>**</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>upgraded</td>\n",
       "      <td>2.631</td>\n",
       "      <td>163.1%</td>\n",
       "      <td>2.387</td>\n",
       "      <td>2.900</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        index     OR     OR%   2.5%  97.5% p.values     \n",
       "1    zip_bins  0.946   -5.4%  0.939  0.953   < .001  ***\n",
       "2     numords  1.320   32.0%  1.262  1.382   < .001  ***\n",
       "3     dollars  1.082    8.2%  1.037  1.129   < .001  ***\n",
       "4        last  0.666  -33.4%  0.637  0.696   < .001  ***\n",
       "5    version1  2.156  115.6%  1.948  2.385   < .001  ***\n",
       "6  owntaxprod  1.361   36.1%  1.118  1.659    0.002   **\n",
       "7    upgraded  2.631  163.1%  2.387  2.900   < .001  ***"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "odd_ratio_std_df = rsm.or_ci(lr_std)\n",
    "odd_ratio_std_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Odds-ratio'>"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rsm.or_plot(lr_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Pseudo R-squared (McFadden): 0.071\n",
      "Pseudo R-squared (McFadden adjusted): 0.07\n",
      "Area under the RO Curve (AUC): 17.485\n",
      "Log-likelihood: -9332.69, AIC: 18681.379, BIC: 18752.328\n",
      "Chi-squared: 52777.974 df(7), p.value < 0.001 \n",
      "Nr obs: 52,500\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\pyrsm\\perf.py:678: RuntimeWarning: overflow encountered in long_scalars\n",
      "  U = np.sum(rankdata(pred)[rvar == False]) - n1 * (n1 + 1) / 2\n"
     ]
    }
   ],
   "source": [
    "rsm.model_fit(lr_std)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After running this standardized logistic regression we are able to appreciate that from all the variables selected they all seem to be statistically significant to describe the response variable given that their p-values are found to be really low, except for one exemption - 'owntaxprod' which even thought it has a p-value of 0.002 (which is still significant) does not perform as well as the rest of the variables."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Standardize values - Taking out owntaxprod, given the p-value found above"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      index     OR     OR%   2.5%  97.5% p.values     \n",
      "1  zip_bins  0.946   -5.4%  0.940  0.953   < .001  ***\n",
      "2   numords  1.327   32.7%  1.269  1.389   < .001  ***\n",
      "3   dollars  1.083    8.3%  1.038  1.129   < .001  ***\n",
      "4      last  0.666  -33.4%  0.637  0.696   < .001  ***\n",
      "5  version1  2.132  113.2%  1.927  2.358   < .001  ***\n",
      "6  upgraded  2.665  166.5%  2.419  2.937   < .001  ***\n"
     ]
    }
   ],
   "source": [
    "lr_std_df2 = smf.glm(\n",
    "    formula=\"res1_yes ~ zip_bins + numords + dollars + last + version1 + upgraded\",\n",
    "    family=Binomial(link=logit()),\n",
    "    data=intuit75k_std_df1,\n",
    ").fit()\n",
    "print(rsm.or_ci(lr_std_df2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After taking out 'owntaxprod' variable we now get a model with all values being strongly statistically significant to describe the response variable given that their p-values are found to be really low."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>variable</th>\n",
       "      <th>vif</th>\n",
       "      <th>Rsq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>numords</td>\n",
       "      <td>1.544</td>\n",
       "      <td>0.353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dollars</td>\n",
       "      <td>1.527</td>\n",
       "      <td>0.345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>version1</td>\n",
       "      <td>1.076</td>\n",
       "      <td>0.071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>upgraded</td>\n",
       "      <td>1.076</td>\n",
       "      <td>0.071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>last</td>\n",
       "      <td>1.017</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>zip_bins</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   variable    vif    Rsq\n",
       "0   numords  1.544  0.353\n",
       "1   dollars  1.527  0.345\n",
       "2  version1  1.076  0.071\n",
       "3  upgraded  1.076  0.071\n",
       "4      last  1.017  0.017\n",
       "5  zip_bins  1.000  0.000"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rsm.vif(lr_std_df2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Multicollinearity reduces the precision of the estimated coefficients, which weakens the statistical power of the regression model. This reduces trust on the p-values identified to be actually statistically significant. This being said, after running the variance influence factor, we can see that the variables selected in this logistic regression seem to be low, demonstrating that there is no concerns of multicollinearity found on the variables included in the model.\n",
    "\n",
    "Additionally we can explore that the variables with the highest R-squared are 'dollars' and 'numords' which indicates they have the highest influence and representation on the response variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>OR</th>\n",
       "      <th>OR%</th>\n",
       "      <th>2.5%</th>\n",
       "      <th>97.5%</th>\n",
       "      <th>p.values</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>zip_bins</td>\n",
       "      <td>0.946</td>\n",
       "      <td>-5.4%</td>\n",
       "      <td>0.940</td>\n",
       "      <td>0.953</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>numords</td>\n",
       "      <td>1.327</td>\n",
       "      <td>32.7%</td>\n",
       "      <td>1.269</td>\n",
       "      <td>1.389</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>dollars</td>\n",
       "      <td>1.083</td>\n",
       "      <td>8.3%</td>\n",
       "      <td>1.038</td>\n",
       "      <td>1.129</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>last</td>\n",
       "      <td>0.666</td>\n",
       "      <td>-33.4%</td>\n",
       "      <td>0.637</td>\n",
       "      <td>0.696</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>version1</td>\n",
       "      <td>2.132</td>\n",
       "      <td>113.2%</td>\n",
       "      <td>1.927</td>\n",
       "      <td>2.358</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>upgraded</td>\n",
       "      <td>2.665</td>\n",
       "      <td>166.5%</td>\n",
       "      <td>2.419</td>\n",
       "      <td>2.937</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      index     OR     OR%   2.5%  97.5% p.values     \n",
       "1  zip_bins  0.946   -5.4%  0.940  0.953   < .001  ***\n",
       "2   numords  1.327   32.7%  1.269  1.389   < .001  ***\n",
       "3   dollars  1.083    8.3%  1.038  1.129   < .001  ***\n",
       "4      last  0.666  -33.4%  0.637  0.696   < .001  ***\n",
       "5  version1  2.132  113.2%  1.927  2.358   < .001  ***\n",
       "6  upgraded  2.665  166.5%  2.419  2.937   < .001  ***"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "odd_ratio_std_df2 = rsm.or_ci(lr_std_df2)\n",
    "odd_ratio_std_df2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Odd-Ratio Interpretation\n",
    "\n",
    "##### The most important variable seem to be 'version1' (2.132), 'upgraded' (2.665) . They seem to be the ones causing the greatest reaction to the response variable.\n",
    "\n",
    "- 'version1' causing an increase of 113.2% on response variable 'res1-yes'. This means that if a customer has the version 1 is more likely to go ahead with the offer compared to the one's that have version 2.\n",
    "\n",
    "- 'upgraded' causing an increase of 166.5% on response variable 'res1-yes'. This means that a customer upgraded from version 1 to version 2 is more likely to go ahead with the offer compared to those who did not previously participated in the upgrade.\n",
    "\n",
    "\n",
    "##### While the least important variables seem to be 'zip_bins' (0.946) and 'dollars' (1.083)\n",
    "\n",
    "- 'zip_bins' causing an decrease of 5.4% on response variable 'res1-yes'. This means binned zip code variable will not be as relevant to dictate the likelihood of going ahead with the offer\n",
    "\n",
    "- 'dollars' causing an increase of 8.3% on response variable 'res1-yes'. This means that the total amount ordered from Intuit Direct n the previous months would not be as relevant for the customer to go ahead with the offer.\n",
    "\n",
    "##### The variables 'numords' (1.389), 'last' (0.637) will have a medium importance\n",
    "\n",
    "- 'numords' causing an increase of 38.9% on response variable 'res1-yes'. This means that customers with the higher number of orders from Intuit Direct in the previous 36 months are more likely to go ahead with the offer compared to the one's that have version 2.\n",
    "\n",
    "- 'last' causing an decrease of 33.4% on response variable 'res1-yes'. This means that the more time it has been since the customers last ordered from Intuit Direct the less likely to go ahead with the offer compared to the one's that have version 2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Odds-ratio'>"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rsm.or_plot(lr_std_df2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Pseudo R-squared (McFadden): 0.07\n",
      "Pseudo R-squared (McFadden adjusted): 0.07\n",
      "Area under the RO Curve (AUC): 17.485\n",
      "Log-likelihood: -9337.08, AIC: 18688.16, BIC: 18750.24\n",
      "Chi-squared: 52843.925 df(6), p.value < 0.001 \n",
      "Nr obs: 52,500\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\pyrsm\\perf.py:678: RuntimeWarning: overflow encountered in long_scalars\n",
      "  U = np.sum(rankdata(pred)[rvar == False]) - n1 * (n1 + 1) / 2\n"
     ]
    }
   ],
   "source": [
    "rsm.model_fit(lr_std_df2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The model interpretation\n",
    "\n",
    "The R-squared for the model was found to be 0.07 which is small in order to consider this model a good estimator for identifying patterns on the odds of responding yes to the email with the offer to upgrade to the latest version of thee QuickBooks software. This means the model only explains 7.0% of the variation of customer responding 'Yes' to the email."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Logistic Regression with Non-Standardize values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        index     OR     OR%   2.5%  97.5% p.values     \n",
      "1    zip_bins  0.946   -5.4%  0.939  0.953   < .001  ***\n",
      "2     numords  1.300   30.0%  1.262  1.340   < .001  ***\n",
      "3        last  0.958   -4.2%  0.954  0.963   < .001  ***\n",
      "4    version1  2.156  115.6%  1.949  2.386   < .001  ***\n",
      "5  owntaxprod  1.368   36.8%  1.123  1.666    0.002   **\n",
      "6    upgraded  2.628  162.8%  2.384  2.897   < .001  ***\n"
     ]
    }
   ],
   "source": [
    "lr = smf.glm(\n",
    "    formula =\"res1_yes ~ zip_bins + numords + last + version1 + owntaxprod + upgraded\",\n",
    "    family=Binomial(link=logit()),\n",
    "    data=intuit75k.query(\"training == 1\"),\n",
    ").fit()\n",
    "print(rsm.or_ci(lr))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Generalized Linear Model Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>       <td>res1_yes</td>     <th>  No. Observations:  </th>  <td> 52500</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                  <td>GLM</td>       <th>  Df Residuals:      </th>  <td> 52493</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model Family:</th>        <td>Binomial</td>     <th>  Df Model:          </th>  <td>     6</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Link Function:</th>         <td>logit</td>      <th>  Scale:             </th> <td>  1.0000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>                <td>IRLS</td>       <th>  Log-Likelihood:    </th> <td> -9339.1</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>            <td>Thu, 10 Feb 2022</td> <th>  Deviance:          </th> <td>  18678.</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                <td>10:53:29</td>     <th>  Pearson chi2:      </th> <td>5.28e+04</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>No. Iterations:</th>          <td>7</td>        <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>     <td>nonrobust</td>    <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td></td>         <th>coef</th>     <th>std err</th>      <th>z</th>      <th>P>|z|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>  <td>   -2.9271</td> <td>    0.069</td> <td>  -42.384</td> <td> 0.000</td> <td>   -3.062</td> <td>   -2.792</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>zip_bins</th>   <td>   -0.0552</td> <td>    0.004</td> <td>  -14.918</td> <td> 0.000</td> <td>   -0.062</td> <td>   -0.048</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>numords</th>    <td>    0.2624</td> <td>    0.015</td> <td>   17.159</td> <td> 0.000</td> <td>    0.232</td> <td>    0.292</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>last</th>       <td>   -0.0427</td> <td>    0.002</td> <td>  -17.997</td> <td> 0.000</td> <td>   -0.047</td> <td>   -0.038</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>version1</th>   <td>    0.7684</td> <td>    0.052</td> <td>   14.882</td> <td> 0.000</td> <td>    0.667</td> <td>    0.870</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>owntaxprod</th> <td>    0.3131</td> <td>    0.101</td> <td>    3.112</td> <td> 0.002</td> <td>    0.116</td> <td>    0.510</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>upgraded</th>   <td>    0.9662</td> <td>    0.050</td> <td>   19.464</td> <td> 0.000</td> <td>    0.869</td> <td>    1.064</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                 Generalized Linear Model Regression Results                  \n",
       "==============================================================================\n",
       "Dep. Variable:               res1_yes   No. Observations:                52500\n",
       "Model:                            GLM   Df Residuals:                    52493\n",
       "Model Family:                Binomial   Df Model:                            6\n",
       "Link Function:                  logit   Scale:                          1.0000\n",
       "Method:                          IRLS   Log-Likelihood:                -9339.1\n",
       "Date:                Thu, 10 Feb 2022   Deviance:                       18678.\n",
       "Time:                        10:53:29   Pearson chi2:                 5.28e+04\n",
       "No. Iterations:                     7                                         \n",
       "Covariance Type:            nonrobust                                         \n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "Intercept     -2.9271      0.069    -42.384      0.000      -3.062      -2.792\n",
       "zip_bins      -0.0552      0.004    -14.918      0.000      -0.062      -0.048\n",
       "numords        0.2624      0.015     17.159      0.000       0.232       0.292\n",
       "last          -0.0427      0.002    -17.997      0.000      -0.047      -0.038\n",
       "version1       0.7684      0.052     14.882      0.000       0.667       0.870\n",
       "owntaxprod     0.3131      0.101      3.112      0.002       0.116       0.510\n",
       "upgraded       0.9662      0.050     19.464      0.000       0.869       1.064\n",
       "==============================================================================\n",
       "\"\"\""
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "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>variable</th>\n",
       "      <th>vif</th>\n",
       "      <th>Rsq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>upgraded</td>\n",
       "      <td>1.082</td>\n",
       "      <td>0.075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>version1</td>\n",
       "      <td>1.081</td>\n",
       "      <td>0.075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>numords</td>\n",
       "      <td>1.032</td>\n",
       "      <td>0.031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>owntaxprod</td>\n",
       "      <td>1.027</td>\n",
       "      <td>0.026</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>last</td>\n",
       "      <td>1.017</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>zip_bins</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     variable    vif    Rsq\n",
       "0    upgraded  1.082  0.075\n",
       "1    version1  1.081  0.075\n",
       "2     numords  1.032  0.031\n",
       "3  owntaxprod  1.027  0.026\n",
       "4        last  1.017  0.017\n",
       "5    zip_bins  1.000  0.000"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rsm.vif(lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "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>index</th>\n",
       "      <th>OR</th>\n",
       "      <th>OR%</th>\n",
       "      <th>2.5%</th>\n",
       "      <th>97.5%</th>\n",
       "      <th>p.values</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>zip_bins</td>\n",
       "      <td>0.946</td>\n",
       "      <td>-5.4%</td>\n",
       "      <td>0.939</td>\n",
       "      <td>0.953</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>numords</td>\n",
       "      <td>1.300</td>\n",
       "      <td>30.0%</td>\n",
       "      <td>1.262</td>\n",
       "      <td>1.340</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>last</td>\n",
       "      <td>0.958</td>\n",
       "      <td>-4.2%</td>\n",
       "      <td>0.954</td>\n",
       "      <td>0.963</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>version1</td>\n",
       "      <td>2.156</td>\n",
       "      <td>115.6%</td>\n",
       "      <td>1.949</td>\n",
       "      <td>2.386</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>owntaxprod</td>\n",
       "      <td>1.368</td>\n",
       "      <td>36.8%</td>\n",
       "      <td>1.123</td>\n",
       "      <td>1.666</td>\n",
       "      <td>0.002</td>\n",
       "      <td>**</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>upgraded</td>\n",
       "      <td>2.628</td>\n",
       "      <td>162.8%</td>\n",
       "      <td>2.384</td>\n",
       "      <td>2.897</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        index     OR     OR%   2.5%  97.5% p.values     \n",
       "1    zip_bins  0.946   -5.4%  0.939  0.953   < .001  ***\n",
       "2     numords  1.300   30.0%  1.262  1.340   < .001  ***\n",
       "3        last  0.958   -4.2%  0.954  0.963   < .001  ***\n",
       "4    version1  2.156  115.6%  1.949  2.386   < .001  ***\n",
       "5  owntaxprod  1.368   36.8%  1.123  1.666    0.002   **\n",
       "6    upgraded  2.628  162.8%  2.384  2.897   < .001  ***"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "odd_ratio_df = rsm.or_ci(lr)\n",
    "odd_ratio_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Odds-ratio'>"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rsm.or_plot(lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Pseudo R-squared (McFadden): 0.07\n",
      "Pseudo R-squared (McFadden adjusted): 0.07\n",
      "Area under the RO Curve (AUC): 17.485\n",
      "Log-likelihood: -9339.093, AIC: 18692.186, BIC: 18754.266\n",
      "Chi-squared: 52849.948 df(6), p.value < 0.001 \n",
      "Nr obs: 52,500\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\pyrsm\\perf.py:678: RuntimeWarning: overflow encountered in long_scalars\n",
      "  U = np.sum(rankdata(pred)[rvar == False]) - n1 * (n1 + 1) / 2\n"
     ]
    }
   ],
   "source": [
    "rsm.model_fit(lr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After running this standardized logistic regression we are able to appreciate that from all the variables selected they all seem to be statistically significant to describe the response variable given that their p-values are found to be really low, except for one exemption - 'owntaxprod' which even thought it has a p-value of 0.002 (which is still significant) does not perform as well as the rest of the variables."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Non-Standardize logistic regression - Taking out owntaxprod, given the p-value found above"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      index     OR     OR%   2.5%  97.5% p.values     \n",
      "1  zip_bins  0.946   -5.4%  0.940  0.953   < .001  ***\n",
      "2   numords  1.257   25.7%  1.212  1.303   < .001  ***\n",
      "3   dollars  1.001    0.1%  1.000  1.002   < .001  ***\n",
      "4      last  0.958   -4.2%  0.954  0.963   < .001  ***\n",
      "5  version1  2.132  113.2%  1.927  2.358   < .001  ***\n",
      "6  upgraded  2.665  166.5%  2.419  2.937   < .001  ***\n"
     ]
    }
   ],
   "source": [
    "lr_df2 = smf.glm(\n",
    "    formula=\"res1_yes ~ zip_bins + numords + dollars + last + version1 + upgraded\",\n",
    "    family=Binomial(link=logit()),\n",
    "    data=intuit75k.query(\"training == 1\"),\n",
    ").fit()\n",
    "print(rsm.or_ci(lr_df2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After taking out 'owntaxprod' variable we now get a model with all values being strongly statistically significant to describe the response variable given that their p-values are found to be really low."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>variable</th>\n",
       "      <th>vif</th>\n",
       "      <th>Rsq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>numords</td>\n",
       "      <td>1.544</td>\n",
       "      <td>0.353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dollars</td>\n",
       "      <td>1.527</td>\n",
       "      <td>0.345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>version1</td>\n",
       "      <td>1.076</td>\n",
       "      <td>0.071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>upgraded</td>\n",
       "      <td>1.076</td>\n",
       "      <td>0.071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>last</td>\n",
       "      <td>1.017</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>zip_bins</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   variable    vif    Rsq\n",
       "0   numords  1.544  0.353\n",
       "1   dollars  1.527  0.345\n",
       "2  version1  1.076  0.071\n",
       "3  upgraded  1.076  0.071\n",
       "4      last  1.017  0.017\n",
       "5  zip_bins  1.000  0.000"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rsm.vif(lr_df2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Multicollinearity reduces the precision of the estimated coefficients, which weakens the statistical power of the regression model. This reduces trust on the p-values identified to be actually statistically significant. This being said, after running the variance influence factor, we can see that the variables selected in this logistic regression seem to be low, demonstrating that there is no concerns of multicollinearity found on the variables included in the model.\n",
    "\n",
    "Additionally we can explore that the variables with the highest R-squared are 'dollars' and 'numords' which indicates they have the highest influence and representation on the response variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>OR</th>\n",
       "      <th>OR%</th>\n",
       "      <th>2.5%</th>\n",
       "      <th>97.5%</th>\n",
       "      <th>p.values</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>zip_bins</td>\n",
       "      <td>0.946</td>\n",
       "      <td>-5.4%</td>\n",
       "      <td>0.940</td>\n",
       "      <td>0.953</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>numords</td>\n",
       "      <td>1.257</td>\n",
       "      <td>25.7%</td>\n",
       "      <td>1.212</td>\n",
       "      <td>1.303</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>dollars</td>\n",
       "      <td>1.001</td>\n",
       "      <td>0.1%</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1.002</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>last</td>\n",
       "      <td>0.958</td>\n",
       "      <td>-4.2%</td>\n",
       "      <td>0.954</td>\n",
       "      <td>0.963</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>version1</td>\n",
       "      <td>2.132</td>\n",
       "      <td>113.2%</td>\n",
       "      <td>1.927</td>\n",
       "      <td>2.358</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>upgraded</td>\n",
       "      <td>2.665</td>\n",
       "      <td>166.5%</td>\n",
       "      <td>2.419</td>\n",
       "      <td>2.937</td>\n",
       "      <td>&lt; .001</td>\n",
       "      <td>***</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      index     OR     OR%   2.5%  97.5% p.values     \n",
       "1  zip_bins  0.946   -5.4%  0.940  0.953   < .001  ***\n",
       "2   numords  1.257   25.7%  1.212  1.303   < .001  ***\n",
       "3   dollars  1.001    0.1%  1.000  1.002   < .001  ***\n",
       "4      last  0.958   -4.2%  0.954  0.963   < .001  ***\n",
       "5  version1  2.132  113.2%  1.927  2.358   < .001  ***\n",
       "6  upgraded  2.665  166.5%  2.419  2.937   < .001  ***"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "odd_ratio_df2 = rsm.or_ci(lr_df2)\n",
    "odd_ratio_df2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Odd-Ratio Interpretation\n",
    "\n",
    "##### The most important variable seem to be 'version1' (2.132), 'upgraded' (2.665) . They seem to be the ones causing the greatest reaction to the response variable.\n",
    "\n",
    "- 'version1' causing an increase of 113.2% on response variable 'res1-yes'. This means that if a customer has the version 1 is more likely to go ahead with the offer compared to the one's that have version 2.\n",
    "\n",
    "- 'upgraded' causing an increase of 166.5% on response variable 'res1-yes'. This means that a customer upgraded from version 1 to version 2 is more likely to go ahead with the offer compared to those who did not previously participated in the upgrade.\n",
    "\n",
    "\n",
    "##### While the least important variables seem to be 'zip_bins' (0.946), 'last' (0.954) and 'dollars' (1.001)\n",
    "\n",
    "- 'zip_bins' causing an decrease of 5.4% on response variable 'res1-yes'. This means binned zip code variable will not be as relevant to dictate the likelihood of going ahead with the offer\n",
    "\n",
    "- 'dollars' causing an increase of 8.3% on response variable 'res1-yes'. This means that the total amount ordered from Intuit Direct n the previous months would not be as relevant for the customer to go ahead with the offer.\n",
    "\n",
    "- 'last' causing an decrease of 4.2% on response variable 'res1-yes'. This means that the more time it has been since the customers last ordered from Intuit Direct the less likely to go ahead with the offer compared to the one's that have version 2.\n",
    "\n",
    "##### The variables 'numords' (1.212) will have a medium importance\n",
    "\n",
    "- 'numords' causing an increase of 25.7% on response variable 'res1-yes'. This means that customers with the higher number of orders from Intuit Direct in the previous 36 months are more likely to go ahead with the offer compared to the one's that have version 2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Odds-ratio'>"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rsm.or_plot(lr_df2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Pseudo R-squared (McFadden): 0.07\n",
      "Pseudo R-squared (McFadden adjusted): 0.07\n",
      "Area under the RO Curve (AUC): 17.485\n",
      "Log-likelihood: -9337.08, AIC: 18688.16, BIC: 18750.24\n",
      "Chi-squared: 52843.925 df(6), p.value < 0.001 \n",
      "Nr obs: 52,500\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\pyrsm\\perf.py:678: RuntimeWarning: overflow encountered in long_scalars\n",
      "  U = np.sum(rankdata(pred)[rvar == False]) - n1 * (n1 + 1) / 2\n"
     ]
    }
   ],
   "source": [
    "rsm.model_fit(lr_df2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The model interpretation\n",
    "\n",
    "The R-squared for the model was found to be 0.07 which is small in order to consider this model a good estimator for identifying patterns on the odds of responding yes to the email with the offer to upgrade to the latest version of thee QuickBooks software. This means the model only explains 7.0% of the variation of customer responding 'Yes' to the email."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Model selected as the predictor in order to assign groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'rsm' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp/ipykernel_5380/1082953010.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mrsm\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpredict_ci\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlr_df2\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mintuit75k\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m: name 'rsm' is not defined"
     ]
    }
   ],
   "source": [
    "rsm.predict_ci(lr_df2, intuit75k)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Assignning a variable pred logit according to what is found on the logistic regression and the different variables that each row has"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k[\"pred_logit\"] = rsm.predict_ci(lr, intuit75k).prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Assignning a deciles according to pred logit in order to divide them into 10 groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_df2 = intuit75k#[intuit75k['training'] == 0]\n",
    "\n",
    "intuit75k_df2['deciles'] = lr.predict(intuit75k_df2.query(\"training == 0\"))\n",
    "intuit75k_df2['deciles']\n",
    "\n",
    "intuit75k_df2[\"dec_logit\"] = rsm.xtile(intuit75k_df2['deciles'], 10, rev = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "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>dec_logit</th>\n",
       "      <th>Number of Emails</th>\n",
       "      <th>Responded Yes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>2249</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>2252</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dec_logit  Number of Emails  Responded Yes\n",
       "0        1.0              2247            336\n",
       "1        2.0              2253            157\n",
       "2        3.0              2250            151\n",
       "3        4.0              2250            113\n",
       "4        5.0              2249            100\n",
       "5        6.0              2248             66\n",
       "6        7.0              2248             57\n",
       "7        8.0              2250             55\n",
       "8        9.0              2253             39\n",
       "9       10.0              2252             29"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_df2.groupby(\"dec_logit\")[\"res1_yes\"].agg([\"count\", \"sum\"]).reset_index().rename(columns = {\"count\" : \"Number of Emails\",\n",
    "                                                                                        \"sum\" : \"Responded Yes\"})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "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>dec_logit</th>\n",
       "      <th>Number of Emails</th>\n",
       "      <th>Responded Yes</th>\n",
       "      <th>Requested Upgrade Rate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "      <td>0.149533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>157</td>\n",
       "      <td>0.069685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>151</td>\n",
       "      <td>0.067111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>113</td>\n",
       "      <td>0.050222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>2249</td>\n",
       "      <td>100</td>\n",
       "      <td>0.044464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>66</td>\n",
       "      <td>0.029359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>57</td>\n",
       "      <td>0.025356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>55</td>\n",
       "      <td>0.024444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>39</td>\n",
       "      <td>0.017310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>2252</td>\n",
       "      <td>29</td>\n",
       "      <td>0.012877</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dec_logit  Number of Emails  Responded Yes  Requested Upgrade Rate\n",
       "0        1.0              2247            336                0.149533\n",
       "1        2.0              2253            157                0.069685\n",
       "2        3.0              2250            151                0.067111\n",
       "3        4.0              2250            113                0.050222\n",
       "4        5.0              2249            100                0.044464\n",
       "5        6.0              2248             66                0.029359\n",
       "6        7.0              2248             57                0.025356\n",
       "7        8.0              2250             55                0.024444\n",
       "8        9.0              2253             39                0.017310\n",
       "9       10.0              2252             29                0.012877"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dec_logit = intuit75k_df2.groupby(\"dec_logit\")[\"res1_yes\"].agg([\"count\", \"sum\"]).reset_index().rename(columns = {\"count\" : \"Number of Emails\",\n",
    "                                                                                        \"sum\" : \"Responded Yes\"})\n",
    "\n",
    "df_dec_logit[\"Requested Upgrade Rate\"] = df_dec_logit[\"Responded Yes\"]/df_dec_logit[\"Number of Emails\"]\n",
    "\n",
    "df_dec_logit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.barplot(data = df_dec_logit, x = \"dec_logit\", y = \"Requested Upgrade Rate\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Lift, Gains, and Profit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "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>dec_logit</th>\n",
       "      <th>Number of Emails</th>\n",
       "      <th>Responded Yes</th>\n",
       "      <th>Requested Upgrade Rate</th>\n",
       "      <th>Cum Number of Emails</th>\n",
       "      <th>Cum Responded Yes</th>\n",
       "      <th>Cum Prop of Number of Emails</th>\n",
       "      <th>Cum Requested Upgrade Rate</th>\n",
       "      <th>Lift</th>\n",
       "      <th>Cum Lift</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "      <td>0.149533</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "      <td>0.018160</td>\n",
       "      <td>0.149533</td>\n",
       "      <td>3.050305</td>\n",
       "      <td>3.050305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>157</td>\n",
       "      <td>0.069685</td>\n",
       "      <td>4500</td>\n",
       "      <td>493</td>\n",
       "      <td>0.054530</td>\n",
       "      <td>0.109556</td>\n",
       "      <td>1.421495</td>\n",
       "      <td>2.234814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>151</td>\n",
       "      <td>0.067111</td>\n",
       "      <td>6750</td>\n",
       "      <td>644</td>\n",
       "      <td>0.109083</td>\n",
       "      <td>0.095407</td>\n",
       "      <td>1.368994</td>\n",
       "      <td>1.946207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>113</td>\n",
       "      <td>0.050222</td>\n",
       "      <td>9000</td>\n",
       "      <td>757</td>\n",
       "      <td>0.181822</td>\n",
       "      <td>0.084111</td>\n",
       "      <td>1.024479</td>\n",
       "      <td>1.715775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>2249</td>\n",
       "      <td>100</td>\n",
       "      <td>0.044464</td>\n",
       "      <td>11249</td>\n",
       "      <td>857</td>\n",
       "      <td>0.272737</td>\n",
       "      <td>0.076185</td>\n",
       "      <td>0.907021</td>\n",
       "      <td>1.554082</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dec_logit  Number of Emails  Responded Yes  Requested Upgrade Rate  \\\n",
       "0        1.0              2247            336                0.149533   \n",
       "1        2.0              2253            157                0.069685   \n",
       "2        3.0              2250            151                0.067111   \n",
       "3        4.0              2250            113                0.050222   \n",
       "4        5.0              2249            100                0.044464   \n",
       "\n",
       "   Cum Number of Emails  Cum Responded Yes  Cum Prop of Number of Emails  \\\n",
       "0                  2247                336                      0.018160   \n",
       "1                  4500                493                      0.054530   \n",
       "2                  6750                644                      0.109083   \n",
       "3                  9000                757                      0.181822   \n",
       "4                 11249                857                      0.272737   \n",
       "\n",
       "   Cum Requested Upgrade Rate      Lift  Cum Lift  \n",
       "0                    0.149533  3.050305  3.050305  \n",
       "1                    0.109556  1.421495  2.234814  \n",
       "2                    0.095407  1.368994  1.946207  \n",
       "3                    0.084111  1.024479  1.715775  \n",
       "4                    0.076185  0.907021  1.554082  "
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dec_logit[\"Cum Number of Emails\"] = df_dec_logit[\"Number of Emails\"].cumsum()#/df_dec_logit[\"Number of Impressions\"].sum()\n",
    "df_dec_logit[\"Cum Responded Yes\"] = df_dec_logit[\"Responded Yes\"].cumsum()\n",
    "df_dec_logit[\"Cum Prop of Number of Emails\"] = (df_dec_logit[\"Cum Number of Emails\"]/df_dec_logit[\"Cum Number of Emails\"].sum()).cumsum()\n",
    "\n",
    "aaa = df_dec_logit[\"Responded Yes\"].sum()/df_dec_logit[\"Number of Emails\"].sum()\n",
    "aaa\n",
    "df_dec_logit[\"Cum Requested Upgrade Rate\"] = df_dec_logit[\"Cum Responded Yes\"]/df_dec_logit[\"Cum Number of Emails\"]\n",
    "\n",
    "\n",
    "df_dec_logit[\"Lift\"] = df_dec_logit[\"Requested Upgrade Rate\"]/aaa\n",
    "df_dec_logit[\"Cum Lift\"] = df_dec_logit[\"Cum Requested Upgrade Rate\"]/aaa\n",
    "\n",
    "\n",
    "df_dec_logit.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig2 = sns.lineplot(data = df_dec_logit, x = \"Cum Prop of Number of Emails\", y = \"Cum Lift\" , marker = 'o')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "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>dec_logit</th>\n",
       "      <th>Number of Emails</th>\n",
       "      <th>Responded Yes</th>\n",
       "      <th>Requested Upgrade Rate</th>\n",
       "      <th>Cum Number of Emails</th>\n",
       "      <th>Cum Responded Yes</th>\n",
       "      <th>Cum Prop of Number of Emails</th>\n",
       "      <th>Cum Requested Upgrade Rate</th>\n",
       "      <th>Lift</th>\n",
       "      <th>Cum Lift</th>\n",
       "      <th>Gains</th>\n",
       "      <th>Cum Gains</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "      <td>0.149533</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "      <td>0.018160</td>\n",
       "      <td>0.149533</td>\n",
       "      <td>3.050305</td>\n",
       "      <td>3.050305</td>\n",
       "      <td>0.304624</td>\n",
       "      <td>0.304624</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>157</td>\n",
       "      <td>0.069685</td>\n",
       "      <td>4500</td>\n",
       "      <td>493</td>\n",
       "      <td>0.054530</td>\n",
       "      <td>0.109556</td>\n",
       "      <td>1.421495</td>\n",
       "      <td>2.234814</td>\n",
       "      <td>0.142339</td>\n",
       "      <td>0.446963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>151</td>\n",
       "      <td>0.067111</td>\n",
       "      <td>6750</td>\n",
       "      <td>644</td>\n",
       "      <td>0.109083</td>\n",
       "      <td>0.095407</td>\n",
       "      <td>1.368994</td>\n",
       "      <td>1.946207</td>\n",
       "      <td>0.136899</td>\n",
       "      <td>0.583862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>113</td>\n",
       "      <td>0.050222</td>\n",
       "      <td>9000</td>\n",
       "      <td>757</td>\n",
       "      <td>0.181822</td>\n",
       "      <td>0.084111</td>\n",
       "      <td>1.024479</td>\n",
       "      <td>1.715775</td>\n",
       "      <td>0.102448</td>\n",
       "      <td>0.686310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>2249</td>\n",
       "      <td>100</td>\n",
       "      <td>0.044464</td>\n",
       "      <td>11249</td>\n",
       "      <td>857</td>\n",
       "      <td>0.272737</td>\n",
       "      <td>0.076185</td>\n",
       "      <td>0.907021</td>\n",
       "      <td>1.554082</td>\n",
       "      <td>0.090662</td>\n",
       "      <td>0.776972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>66</td>\n",
       "      <td>0.029359</td>\n",
       "      <td>13497</td>\n",
       "      <td>923</td>\n",
       "      <td>0.381820</td>\n",
       "      <td>0.068386</td>\n",
       "      <td>0.598900</td>\n",
       "      <td>1.394991</td>\n",
       "      <td>0.059837</td>\n",
       "      <td>0.836809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>57</td>\n",
       "      <td>0.025356</td>\n",
       "      <td>15745</td>\n",
       "      <td>980</td>\n",
       "      <td>0.509072</td>\n",
       "      <td>0.062242</td>\n",
       "      <td>0.517232</td>\n",
       "      <td>1.269669</td>\n",
       "      <td>0.051677</td>\n",
       "      <td>0.888486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>55</td>\n",
       "      <td>0.024444</td>\n",
       "      <td>17995</td>\n",
       "      <td>1035</td>\n",
       "      <td>0.654509</td>\n",
       "      <td>0.057516</td>\n",
       "      <td>0.498640</td>\n",
       "      <td>1.173263</td>\n",
       "      <td>0.049864</td>\n",
       "      <td>0.938350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>39</td>\n",
       "      <td>0.017310</td>\n",
       "      <td>20248</td>\n",
       "      <td>1074</td>\n",
       "      <td>0.818154</td>\n",
       "      <td>0.053042</td>\n",
       "      <td>0.353110</td>\n",
       "      <td>1.082005</td>\n",
       "      <td>0.035358</td>\n",
       "      <td>0.973708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>2252</td>\n",
       "      <td>29</td>\n",
       "      <td>0.012877</td>\n",
       "      <td>22500</td>\n",
       "      <td>1103</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.049022</td>\n",
       "      <td>0.262686</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.026292</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dec_logit  Number of Emails  Responded Yes  Requested Upgrade Rate  \\\n",
       "0        1.0              2247            336                0.149533   \n",
       "1        2.0              2253            157                0.069685   \n",
       "2        3.0              2250            151                0.067111   \n",
       "3        4.0              2250            113                0.050222   \n",
       "4        5.0              2249            100                0.044464   \n",
       "5        6.0              2248             66                0.029359   \n",
       "6        7.0              2248             57                0.025356   \n",
       "7        8.0              2250             55                0.024444   \n",
       "8        9.0              2253             39                0.017310   \n",
       "9       10.0              2252             29                0.012877   \n",
       "\n",
       "   Cum Number of Emails  Cum Responded Yes  Cum Prop of Number of Emails  \\\n",
       "0                  2247                336                      0.018160   \n",
       "1                  4500                493                      0.054530   \n",
       "2                  6750                644                      0.109083   \n",
       "3                  9000                757                      0.181822   \n",
       "4                 11249                857                      0.272737   \n",
       "5                 13497                923                      0.381820   \n",
       "6                 15745                980                      0.509072   \n",
       "7                 17995               1035                      0.654509   \n",
       "8                 20248               1074                      0.818154   \n",
       "9                 22500               1103                      1.000000   \n",
       "\n",
       "   Cum Requested Upgrade Rate      Lift  Cum Lift     Gains  Cum Gains  \n",
       "0                    0.149533  3.050305  3.050305  0.304624   0.304624  \n",
       "1                    0.109556  1.421495  2.234814  0.142339   0.446963  \n",
       "2                    0.095407  1.368994  1.946207  0.136899   0.583862  \n",
       "3                    0.084111  1.024479  1.715775  0.102448   0.686310  \n",
       "4                    0.076185  0.907021  1.554082  0.090662   0.776972  \n",
       "5                    0.068386  0.598900  1.394991  0.059837   0.836809  \n",
       "6                    0.062242  0.517232  1.269669  0.051677   0.888486  \n",
       "7                    0.057516  0.498640  1.173263  0.049864   0.938350  \n",
       "8                    0.053042  0.353110  1.082005  0.035358   0.973708  \n",
       "9                    0.049022  0.262686  1.000000  0.026292   1.000000  "
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dec_logit[\"Gains\"] = df_dec_logit[\"Responded Yes\"]/df_dec_logit[\"Responded Yes\"].sum()\n",
    "df_dec_logit[\"Cum Gains\"] = df_dec_logit[\"Gains\"].cumsum()\n",
    "\n",
    "df_dec_logit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "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>dec_logit</th>\n",
       "      <th>Number of Emails</th>\n",
       "      <th>Responded Yes</th>\n",
       "      <th>Requested Upgrade Rate</th>\n",
       "      <th>Cum Number of Emails</th>\n",
       "      <th>Cum Responded Yes</th>\n",
       "      <th>Cum Prop of Number of Emails</th>\n",
       "      <th>Cum Requested Upgrade Rate</th>\n",
       "      <th>Lift</th>\n",
       "      <th>Cum Lift</th>\n",
       "      <th>Gains</th>\n",
       "      <th>Cum Gains</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "      <td>0.149533</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "      <td>0.018160</td>\n",
       "      <td>0.149533</td>\n",
       "      <td>3.050305</td>\n",
       "      <td>3.050305</td>\n",
       "      <td>0.304624</td>\n",
       "      <td>0.304624</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>157</td>\n",
       "      <td>0.069685</td>\n",
       "      <td>4500</td>\n",
       "      <td>493</td>\n",
       "      <td>0.054530</td>\n",
       "      <td>0.109556</td>\n",
       "      <td>1.421495</td>\n",
       "      <td>2.234814</td>\n",
       "      <td>0.142339</td>\n",
       "      <td>0.446963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>151</td>\n",
       "      <td>0.067111</td>\n",
       "      <td>6750</td>\n",
       "      <td>644</td>\n",
       "      <td>0.109083</td>\n",
       "      <td>0.095407</td>\n",
       "      <td>1.368994</td>\n",
       "      <td>1.946207</td>\n",
       "      <td>0.136899</td>\n",
       "      <td>0.583862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>113</td>\n",
       "      <td>0.050222</td>\n",
       "      <td>9000</td>\n",
       "      <td>757</td>\n",
       "      <td>0.181822</td>\n",
       "      <td>0.084111</td>\n",
       "      <td>1.024479</td>\n",
       "      <td>1.715775</td>\n",
       "      <td>0.102448</td>\n",
       "      <td>0.686310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5.0</td>\n",
       "      <td>2249</td>\n",
       "      <td>100</td>\n",
       "      <td>0.044464</td>\n",
       "      <td>11249</td>\n",
       "      <td>857</td>\n",
       "      <td>0.272737</td>\n",
       "      <td>0.076185</td>\n",
       "      <td>0.907021</td>\n",
       "      <td>1.554082</td>\n",
       "      <td>0.090662</td>\n",
       "      <td>0.776972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>66</td>\n",
       "      <td>0.029359</td>\n",
       "      <td>13497</td>\n",
       "      <td>923</td>\n",
       "      <td>0.381820</td>\n",
       "      <td>0.068386</td>\n",
       "      <td>0.598900</td>\n",
       "      <td>1.394991</td>\n",
       "      <td>0.059837</td>\n",
       "      <td>0.836809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>57</td>\n",
       "      <td>0.025356</td>\n",
       "      <td>15745</td>\n",
       "      <td>980</td>\n",
       "      <td>0.509072</td>\n",
       "      <td>0.062242</td>\n",
       "      <td>0.517232</td>\n",
       "      <td>1.269669</td>\n",
       "      <td>0.051677</td>\n",
       "      <td>0.888486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>55</td>\n",
       "      <td>0.024444</td>\n",
       "      <td>17995</td>\n",
       "      <td>1035</td>\n",
       "      <td>0.654509</td>\n",
       "      <td>0.057516</td>\n",
       "      <td>0.498640</td>\n",
       "      <td>1.173263</td>\n",
       "      <td>0.049864</td>\n",
       "      <td>0.938350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>39</td>\n",
       "      <td>0.017310</td>\n",
       "      <td>20248</td>\n",
       "      <td>1074</td>\n",
       "      <td>0.818154</td>\n",
       "      <td>0.053042</td>\n",
       "      <td>0.353110</td>\n",
       "      <td>1.082005</td>\n",
       "      <td>0.035358</td>\n",
       "      <td>0.973708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10.0</td>\n",
       "      <td>2252</td>\n",
       "      <td>29</td>\n",
       "      <td>0.012877</td>\n",
       "      <td>22500</td>\n",
       "      <td>1103</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.049022</td>\n",
       "      <td>0.262686</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.026292</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    dec_logit  Number of Emails  Responded Yes  Requested Upgrade Rate  \\\n",
       "0         0.0                 0              0                0.000000   \n",
       "1         1.0              2247            336                0.149533   \n",
       "2         2.0              2253            157                0.069685   \n",
       "3         3.0              2250            151                0.067111   \n",
       "4         4.0              2250            113                0.050222   \n",
       "5         5.0              2249            100                0.044464   \n",
       "6         6.0              2248             66                0.029359   \n",
       "7         7.0              2248             57                0.025356   \n",
       "8         8.0              2250             55                0.024444   \n",
       "9         9.0              2253             39                0.017310   \n",
       "10       10.0              2252             29                0.012877   \n",
       "\n",
       "    Cum Number of Emails  Cum Responded Yes  Cum Prop of Number of Emails  \\\n",
       "0                      0                  0                      0.000000   \n",
       "1                   2247                336                      0.018160   \n",
       "2                   4500                493                      0.054530   \n",
       "3                   6750                644                      0.109083   \n",
       "4                   9000                757                      0.181822   \n",
       "5                  11249                857                      0.272737   \n",
       "6                  13497                923                      0.381820   \n",
       "7                  15745                980                      0.509072   \n",
       "8                  17995               1035                      0.654509   \n",
       "9                  20248               1074                      0.818154   \n",
       "10                 22500               1103                      1.000000   \n",
       "\n",
       "    Cum Requested Upgrade Rate      Lift  Cum Lift     Gains  Cum Gains  \n",
       "0                     0.000000  0.000000  0.000000  0.000000   0.000000  \n",
       "1                     0.149533  3.050305  3.050305  0.304624   0.304624  \n",
       "2                     0.109556  1.421495  2.234814  0.142339   0.446963  \n",
       "3                     0.095407  1.368994  1.946207  0.136899   0.583862  \n",
       "4                     0.084111  1.024479  1.715775  0.102448   0.686310  \n",
       "5                     0.076185  0.907021  1.554082  0.090662   0.776972  \n",
       "6                     0.068386  0.598900  1.394991  0.059837   0.836809  \n",
       "7                     0.062242  0.517232  1.269669  0.051677   0.888486  \n",
       "8                     0.057516  0.498640  1.173263  0.049864   0.938350  \n",
       "9                     0.053042  0.353110  1.082005  0.035358   0.973708  \n",
       "10                    0.049022  0.262686  1.000000  0.026292   1.000000  "
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gains_tab = df_dec_logit\n",
    "gains_tab\n",
    "\n",
    "new_row = pd.DataFrame({'dec_logit':0, 'Number of Emails':0, 'Responded Yes':0,\n",
    "       'Requested Upgrade Rate':0, 'Cum Number of Emails':0, 'Cum Responded Yes':0,\n",
    "       'Cum Prop of Number of Emails':0, 'Cum Requested Upgrade Rate':0, 'Lift':0, 'Cum Lift':0,\n",
    "       'Gains':0, 'Cum Gains':0}, index=[0])\n",
    "\n",
    "gains_tab = pd.concat([new_row, gains_tab]).reset_index(drop = True)\n",
    "gains_tab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2994a5b6a90>]"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(gains_tab[\"Cum Prop of Number of Emails\"], gains_tab[\"Cum Gains\"], label = 'line 1', linestyle = 'solid')\n",
    "plt.plot([0,1], [0,1], label = 'line 2', linestyle = 'dashed')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "cost = 1.41 # each mail piece costs $1.41\n",
    "margin = 60 # the margin (or net revenue) from each responder, excluding the mailing cost, is $60\n",
    "breakeven =  cost/margin # float, length 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Rohan\\AppData\\Local\\Temp/ipykernel_105448/705169543.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  profit_tab['Cost'] = profit_tab['Number of Emails']*cost\n",
      "C:\\Users\\Rohan\\AppData\\Local\\Temp/ipykernel_105448/705169543.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  profit_tab['Revenue'] = profit_tab['Responded Yes']*margin\n",
      "C:\\Users\\Rohan\\AppData\\Local\\Temp/ipykernel_105448/705169543.py:4: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  profit_tab['Profit'] = profit_tab['Revenue'] - profit_tab['Cost']\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dec_logit</th>\n",
       "      <th>Number of Emails</th>\n",
       "      <th>Responded Yes</th>\n",
       "      <th>Cum Prop of Number of Emails</th>\n",
       "      <th>Cost</th>\n",
       "      <th>Revenue</th>\n",
       "      <th>Profit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2247</td>\n",
       "      <td>336</td>\n",
       "      <td>0.018160</td>\n",
       "      <td>3168.27</td>\n",
       "      <td>20160</td>\n",
       "      <td>16991.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>157</td>\n",
       "      <td>0.054530</td>\n",
       "      <td>3176.73</td>\n",
       "      <td>9420</td>\n",
       "      <td>6243.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>151</td>\n",
       "      <td>0.109083</td>\n",
       "      <td>3172.50</td>\n",
       "      <td>9060</td>\n",
       "      <td>5887.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>113</td>\n",
       "      <td>0.181822</td>\n",
       "      <td>3172.50</td>\n",
       "      <td>6780</td>\n",
       "      <td>3607.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>2249</td>\n",
       "      <td>100</td>\n",
       "      <td>0.272737</td>\n",
       "      <td>3171.09</td>\n",
       "      <td>6000</td>\n",
       "      <td>2828.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>66</td>\n",
       "      <td>0.381820</td>\n",
       "      <td>3169.68</td>\n",
       "      <td>3960</td>\n",
       "      <td>790.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7.0</td>\n",
       "      <td>2248</td>\n",
       "      <td>57</td>\n",
       "      <td>0.509072</td>\n",
       "      <td>3169.68</td>\n",
       "      <td>3420</td>\n",
       "      <td>250.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8.0</td>\n",
       "      <td>2250</td>\n",
       "      <td>55</td>\n",
       "      <td>0.654509</td>\n",
       "      <td>3172.50</td>\n",
       "      <td>3300</td>\n",
       "      <td>127.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2253</td>\n",
       "      <td>39</td>\n",
       "      <td>0.818154</td>\n",
       "      <td>3176.73</td>\n",
       "      <td>2340</td>\n",
       "      <td>-836.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>2252</td>\n",
       "      <td>29</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3175.32</td>\n",
       "      <td>1740</td>\n",
       "      <td>-1435.32</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dec_logit  Number of Emails  Responded Yes  Cum Prop of Number of Emails  \\\n",
       "0        1.0              2247            336                      0.018160   \n",
       "1        2.0              2253            157                      0.054530   \n",
       "2        3.0              2250            151                      0.109083   \n",
       "3        4.0              2250            113                      0.181822   \n",
       "4        5.0              2249            100                      0.272737   \n",
       "5        6.0              2248             66                      0.381820   \n",
       "6        7.0              2248             57                      0.509072   \n",
       "7        8.0              2250             55                      0.654509   \n",
       "8        9.0              2253             39                      0.818154   \n",
       "9       10.0              2252             29                      1.000000   \n",
       "\n",
       "      Cost  Revenue    Profit  \n",
       "0  3168.27    20160  16991.73  \n",
       "1  3176.73     9420   6243.27  \n",
       "2  3172.50     9060   5887.50  \n",
       "3  3172.50     6780   3607.50  \n",
       "4  3171.09     6000   2828.91  \n",
       "5  3169.68     3960    790.32  \n",
       "6  3169.68     3420    250.32  \n",
       "7  3172.50     3300    127.50  \n",
       "8  3176.73     2340   -836.73  \n",
       "9  3175.32     1740  -1435.32  "
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profit_tab = df_dec_logit[['dec_logit', 'Number of Emails', 'Responded Yes', 'Cum Prop of Number of Emails']]\n",
    "profit_tab['Cost'] = profit_tab['Number of Emails']*cost\n",
    "profit_tab['Revenue'] = profit_tab['Responded Yes']*margin\n",
    "profit_tab['Profit'] = profit_tab['Revenue'] - profit_tab['Cost']\n",
    "\n",
    "\n",
    "profit_tab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2994a627220>]"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig4 = sns.lineplot(data = profit_tab, x = \"Cum Prop of Number of Emails\", y = \"Profit\" , marker = 'o')\n",
    "\n",
    "plt.plot(profit_tab[\"Cum Prop of Number of Emails\"], profit_tab[\"Profit\"], label = 'line 1', linestyle = 'solid')\n",
    "plt.plot([0,1], [0,0], label = 'line 2', linestyle = 'dashed')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_test = intuit75k[intuit75k['training'] == 0]\n",
    "target_logit_df = intuit75k_df3[\"pred_logit\"] > breakeven"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pred_logit  res1_yes\n",
       "False       0            6406\n",
       "            1             116\n",
       "True        0           14991\n",
       "            1             987\n",
       "Name: res1_yes, dtype: int64"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k_df4 = intuit75k_df3.groupby(intuit75k[\"pred_logit\"] > breakeven)\n",
    "vvv2 = intuit75k_df4['res1_yes'].value_counts()\n",
    "vvv2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_df3\n",
    "\n",
    "\n",
    "tuango[\"smsto_sq\"] = tuango.rfm_sq_resp > breakeven"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "TP = vvv2[1][1]\n",
    "FP = vvv2[1][0]\n",
    "TN = vvv2[0][0]\n",
    "FN = vvv2[0][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "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>label</th>\n",
       "      <th>nr</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TP</td>\n",
       "      <td>987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>FP</td>\n",
       "      <td>14991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TN</td>\n",
       "      <td>6406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>FN</td>\n",
       "      <td>116</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  label     nr\n",
       "0    TP    987\n",
       "1    FP  14991\n",
       "2    TN   6406\n",
       "3    FN    116"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# confusion matrix 0.22% and 0.10% \n",
    "cm_logit = pd.DataFrame(\n",
    "    {\n",
    "        \"label\": [\"TP\", \"FP\", \"TN\", \"FN\"],\n",
    "        \"nr\": [TP, FP, TN, FN] # TP, FP, TN, and FN values in that order\n",
    "    }\n",
    ")\n",
    "cm_logit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3285777777777778"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_logit = (TP + TN)/cm_logit['nr'].sum()# float\n",
    "accuracy_logit"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion matrices\n",
    "\n",
    "After running the confusion matrix it is clear that the logit model has a high level of differentiation, knows how to target correctly and assign prediction based on this test. This model has a high prediction of True Positive (TP) values with 987, which means it detects quite sharp unfortunately it has a high production of False Positive (FP) values with 14991 which would end up rising the cost of resources (sending emails) of doing the marketting campaign."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "36691.020000000004"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "profit = (TP * margin) - (TP + FP) * 1.41\n",
    "profit"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Model Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    47421\n",
       "True     27579\n",
       "Name: target_logit, dtype: int64"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k[\"target_logit\"] = rsm.ifelse(intuit75k.pred_logit * 0.5 > breakeven, True, False)\n",
    "intuit75k[\"target_logit\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    25212\n",
       "1     2367\n",
       "Name: res1_yes, dtype: int64"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df1 = intuit75k[intuit75k[\"target_logit\"] == True]\n",
    "test_df1.res1_yes.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "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>id</th>\n",
       "      <th>zip</th>\n",
       "      <th>zip_bins</th>\n",
       "      <th>sex</th>\n",
       "      <th>bizflag</th>\n",
       "      <th>numords</th>\n",
       "      <th>dollars</th>\n",
       "      <th>last</th>\n",
       "      <th>sincepurch</th>\n",
       "      <th>version1</th>\n",
       "      <th>...</th>\n",
       "      <th>freq_iq</th>\n",
       "      <th>mon_iq</th>\n",
       "      <th>rfm_iq</th>\n",
       "      <th>freq_sq</th>\n",
       "      <th>mon_sq</th>\n",
       "      <th>rfm_sq</th>\n",
       "      <th>pred_logit</th>\n",
       "      <th>deciles</th>\n",
       "      <th>dec_logit</th>\n",
       "      <th>target_logit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>53190</td>\n",
       "      <td>10</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>69.5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>153</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>152</td>\n",
       "      <td>0.032690</td>\n",
       "      <td>0.032690</td>\n",
       "      <td>6.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37091</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>93.0</td>\n",
       "      <td>14</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>312</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>315</td>\n",
       "      <td>0.124466</td>\n",
       "      <td>0.124466</td>\n",
       "      <td>1.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60201</td>\n",
       "      <td>11</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24.5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>155</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>154</td>\n",
       "      <td>0.033659</td>\n",
       "      <td>0.033659</td>\n",
       "      <td>6.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>22980</td>\n",
       "      <td>5</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>49.5</td>\n",
       "      <td>13</td>\n",
       "      <td>36</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>254</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>252</td>\n",
       "      <td>0.061372</td>\n",
       "      <td>0.061372</td>\n",
       "      <td>3.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>34950</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>44.5</td>\n",
       "      <td>15</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>354</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>353</td>\n",
       "      <td>0.023049</td>\n",
       "      <td>0.023049</td>\n",
       "      <td>8.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id    zip  zip_bins      sex  bizflag  numords  dollars  last  sincepurch  \\\n",
       "1   2  53190        10  Unknown        0        1     69.5     4           3   \n",
       "2   3  37091         8     Male        0        4     93.0    14          29   \n",
       "4   5  60201        11     Male        0        1     24.5     2           3   \n",
       "6   7  22980         5     Male        0        1     49.5    13          36   \n",
       "8   9  34950         8     Male        0        1     44.5    15           4   \n",
       "\n",
       "   version1  ...  freq_iq  mon_iq rfm_iq  freq_sq  mon_sq  rfm_sq  pred_logit  \\\n",
       "1         0  ...        5       3    153        5       2     152    0.032690   \n",
       "2         0  ...        1       2    312        1       5     315    0.124466   \n",
       "4         0  ...        5       5    155        5       4     154    0.033659   \n",
       "6         1  ...        5       4    254        5       2     252    0.061372   \n",
       "8         0  ...        5       4    354        5       3     353    0.023049   \n",
       "\n",
       "    deciles dec_logit  target_logit  \n",
       "1  0.032690       6.0         False  \n",
       "2  0.124466       1.0          True  \n",
       "4  0.033659       6.0         False  \n",
       "6  0.061372       3.0          True  \n",
       "8  0.023049       8.0         False  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intuit75k[intuit75k['training']==0].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "35059.2"
      ]
     },
     "execution_count": 205,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nr_emails_logit = len(test_df1)\n",
    "perc_emails_logit = len(test_df1)/len(intuit75k_df3)\n",
    "\n",
    "response_rate_logit = test_df1.res1_yes.value_counts(normalize = True)[1]\n",
    "nr_responses_logit = response_rate_logit*nr_emails_logit\n",
    "\n",
    "cost_email = len(test_df1)*cost\n",
    "revenue_logit = test_df1.res1_yes.value_counts()[1]*margin\n",
    "profit_logit = revenue_logit - cost_email\n",
    "\n",
    "ROME_logit = profit_logit/cost_email\n",
    "\n",
    "profit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Targetting Quickbooks with logit would email all 27,579 (122.57%) customers. The response rate for the selected customers is predicted to be 8.58% or 2,367\n",
      "buyers. The revenue is equal to $142,020. The expected profit is $35,059. The email cost is estimated to be $38,886 with a ROME of 265.22%\n"
     ]
    }
   ],
   "source": [
    "print(\n",
    "    f\"\"\"Targetting Quickbooks with logit would email all {int(nr_emails_logit):,} ({round((100 * perc_emails_logit), 2)}%) customers. The response rate for the selected customers is predicted to be {round((100 * response_rate_logit), 2)}% or {int(nr_responses_logit):,}\n",
    "buyers. The revenue is equal to ${int(revenue_logit):,}. The expected profit is ${int(profit):,}. The email cost is estimated to be ${int(cost_email):,} with a ROME of {round((100 * ROME_logit), 2)}%\"\"\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "392964343.2"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scale1 = round((100 * perc_emails_logit), 2)\n",
    "scale2 = round((100 * response_rate_logit), 2)\n",
    "tot_emails = 801821 - 38487\n",
    "total_cost = scale1*tot_emails*cost\n",
    "total_revenue = tot_emails*round((100 * response_rate_logit), 2)*margin\n",
    "total_revenue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = rsm.gains_plot(intuit75k[intuit75k.training == 1], \"res1_yes\", 1, \"pred_logit\").legend(title=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "tpr = TP/(TP+FN)\n",
    "specificity = TN/(TN+FP)\n",
    "fpr = 1- specificity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = intuit75k.res1_yes\n",
    "y_prob = rsm.ifelse(intuit75k[\"pred_logit\"] * 0.5 > breakeven, 1 ,0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import metrics\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y, y_prob)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "auc = metrics.roc_auc_score(y, y_prob)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(fpr,tpr,label=\"AUC=\"+str(auc))\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.legend(loc=4)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "22500"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(intuit75k[intuit75k['training']==0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Rohan\\AppData\\Local\\Temp/ipykernel_105448/3167277150.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  intuit75k_csv[\"target\"] = rsm.ifelse(intuit75k_csv[\"pred_logit\"] * 0.5 > breakeven, True ,False)\n"
     ]
    }
   ],
   "source": [
    "intuit75k_csv = intuit75k[intuit75k['training']==0]\n",
    "\n",
    "intuit75k_csv[\"target\"] = rsm.ifelse(intuit75k_csv[\"pred_logit\"] * 0.5 > breakeven, True ,False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [],
   "source": [
    "intuit75k_csv[['id', 'target']].to_csv('Arnva_Akanksha_Renato_Rohan_Group24.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " After running analysis on the data, we find the response rate, expected revenue and profit for the three scenarios: \n",
    "\n",
    "- Targetting Quickbooks with logit would email all 15,978 (71.01%) customers. The response rate for the selected customers is predicted to be 6.18% or 987\n",
    "buyers. The revenue is equal to 59,220. The expected profit is 36,691. The email cost is estimated to be 22,528 with a ROME of 162.86%\n",
    "\n",
    "We can observe that with the Logistic Regression approach we are able to narrow down our targetted number of customers, reducing the amount of emails to be sent and therefore allowing us to increase the expected profit compared to the RFM model. Which when scalable to do a bigger marketing campaign, like the projected 763,334 customers to be targetted in the second-wave, we would get more profit."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Neural Networks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our RFM and Logistic models have given us some promosing results. In order to better evaluate our mailing prediction, we will use neural networks and build a model accordingly to come up with potential profit. \n",
    "\n",
    "Neural Networks usually consist of two major hyperparameters that controls its architecture : The number of layers and the numbers of nodes in each hidden layers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [],
   "source": [
    "def nn(X, y, hlsizes=(1,), max_iter=1000):\n",
    "    return MLPClassifier(\n",
    "        activation=\"tanh\",\n",
    "        solver=\"lbfgs\",\n",
    "        hidden_layer_sizes=hlsizes,\n",
    "        random_state=1234,\n",
    "        max_iter=max_iter,\n",
    "    ).fit(X, y)\n",
    "\n",
    "\n",
    "def importance(clf, X, y, cn):\n",
    "    imp = permutation_importance(\n",
    "        clf, X, y, scoring=\"roc_auc\", n_repeats=10, random_state=1234\n",
    "    )\n",
    "    data = pd.DataFrame(imp.importances.T)\n",
    "    data.columns = cn\n",
    "    order = data.agg(\"mean\").sort_values(ascending=False).index\n",
    "    fig = sns.barplot(\n",
    "        x=\"value\", y=\"variable\", color=\"slateblue\", data=pd.melt(data[order])\n",
    "    )\n",
    "    fig.set(title=\"Permutation Importances\", xlabel=None, ylabel=None)\n",
    "    return fig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = intuit75k.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to make our modelling feasible, we factorised two of our significant variables \"sex\"  and \"res1\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [],
   "source": [
    "X['sex'],_ = X['sex'].factorize()\n",
    "X['res1'],_ = X['res1'].factorize()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The 'X_NEW' dataframe contains all the information relevant to the significant variables mentioned below. These response variables were also used in making our logistic model and gave us the optimal result."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_NEW = X[['zip_bins','numords', 'last', 'version1', 'owntaxprod' , 'upgraded']].copy()\n",
    "y = intuit75k.res1_yes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After thorough analysis, we came up with the appropriate number of nodes and hidden layers. The right combination which we decided to work on was 7 nodes and 9 hidden layers. These hyperparameters gave us the appropriate amount of profit. Adding more modes and hidden layers adds felxibilty to our model. This helped us in fitting the training data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\neural_network\\_multilayer_perceptron.py:500: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = nn(X_NEW[intuit75k.training == 1], y[intuit75k.training == 1], (6,3))\n",
    "intuit75k[\"pred_nn1\"] = clf.predict_proba(X_NEW)[:, 1]  # predicting for the full dataset\n",
    "fig = importance(clf, X_NEW[intuit75k.training == 1], y[intuit75k.training == 1], X_NEW.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can observe in the above figure, the permutation importances for all the explanatory variables are given in the descending order. 'zip_bins' have the highest importance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pred_nn1  res1_yes\n",
       "False     0           15263\n",
       "          1             357\n",
       "True      0            6134\n",
       "          1             746\n",
       "Name: res1_yes, dtype: int64"
      ]
     },
     "execution_count": 190,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df2 = intuit75k[intuit75k['training'] == 0]\n",
    "test_df3 = test_df2.groupby(intuit75k[\"pred_nn1\"] * 0.5 > breakeven)\n",
    "vvv2 = test_df3['res1_yes'].value_counts()\n",
    "test_df3['res1_yes'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The above code was executed in order to find TP, FP, TN, FN. These variables will further be used to create a confusion matrix for the calculation of profit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [],
   "source": [
    "TP = vvv2[1][1]\n",
    "FP = vvv2[1][0]\n",
    "TN = vvv2[0][0]\n",
    "FN = vvv2[0][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "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>label</th>\n",
       "      <th>nr</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TP</td>\n",
       "      <td>746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>FP</td>\n",
       "      <td>6134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TN</td>\n",
       "      <td>15263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>FN</td>\n",
       "      <td>357</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  label     nr\n",
       "0    TP    746\n",
       "1    FP   6134\n",
       "2    TN  15263\n",
       "3    FN    357"
      ]
     },
     "execution_count": 192,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# confusion matrix\n",
    "cm_logit_recalc = pd.DataFrame(\n",
    "    {\n",
    "        \"label\": [\"TP\", \"FP\", \"TN\", \"FN\"],\n",
    "        \"nr\": [TP, FP, TN, FN]# TP, FP, TN, and FN values in that order\n",
    "    }\n",
    ")\n",
    "\n",
    "cm_logit_recalc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7115111111111111"
      ]
     },
     "execution_count": 193,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_logit_recalc = (TP + TN)/cm_logit_recalc['nr'].sum()# float\n",
    "accuracy_logit_recalc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The accuracy for our neural network model was calculated and it came out to be approximately 72.4%. Accuracy is truely not the right metric to measure the effectiveness of our model. This is because, in terms of business, we want to focus on metrics like profit which is usually computed using True Positives: Users whom the model predicted to send the mail and they were send. But at the same time, as we can observe, our False Positives are also very high which means that our model predicted to send mail to those users who were not supposed to be mails. This would likely result in incurring high amount of cost without any revenue from this cohort."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "35059.2"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profit = (TP * margin) - (TP + FP) * 1.41\n",
    "profit"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The profit for our neural netwrok model comes out to be 35059.2 USD."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [],
   "source": [
    "tpr = TP/(TP+FN)\n",
    "specificity = TN/(TN+FP)\n",
    "fpr = 1- specificity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.6763372620126926, 0.28667570220124317)"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tpr, fpr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_prob = rsm.ifelse(intuit75k[\"pred_nn1\"] * 0.5 > breakeven, 1 ,0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import metrics\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y, y_prob)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [],
   "source": [
    "auc = metrics.roc_auc_score(y, y_prob)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(fpr,tpr,label=\"AUC=\"+str(auc))\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.legend(loc=4)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = rsm.gains_plot(intuit75k[intuit75k.training == 1], \"res1_yes\", 1, \"pred_nn1\").legend(title=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#END "
   ]
  }
 ],
 "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}